The Toxic Effects of Multiple Persistent Organic Pollutant Exposures on the Post-Hatch Immunity Maturation of Glaucous Gulls

PLEASE SCROLL DOWN FOR ARTICLE

This article was downloaded by: [University Library of Tromsoe]On: 30 July 2009Access details: Access Details: [subscription number 907435825]Publisher Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Journal of Toxicology and Environmental Health, Part APublication details, including instructions for authors and subscription information:http://www.informaworld.com/smpp/title~content=t713667303

The Toxic Effects of Multiple Persistent Organic Pollutant Exposures on thePost-Hatch Immunity Maturation of Glaucous GullsKjetil Sagerup a; Hans Jørgen S. Larsen b; Janneche Utne Skaare bc; Grethe M. Johansen b; Geir WingGabrielsen d

a Tromsø University Museum, Tromsø, Norway b Norwegian School of Veterinary Science, Oslo, Norway c

National Veterinary Institute, Oslo, Norway d The Polar Environmental Centre, Norwegian Polar Institute,Tromsø, Norway

Online Publication Date: 01 January 2009

To cite this Article Sagerup, Kjetil, Larsen, Hans Jørgen S., Skaare, Janneche Utne, Johansen, Grethe M. and Gabrielsen, GeirWing(2009)'The Toxic Effects of Multiple Persistent Organic Pollutant Exposures on the Post-Hatch Immunity Maturation of GlaucousGulls',Journal of Toxicology and Environmental Health, Part A,72:14,870 — 883

To link to this Article: DOI: 10.1080/15287390902959516

URL: http://dx.doi.org/10.1080/15287390902959516

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf

This article may be used for research, teaching and private study purposes. Any substantial orsystematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply ordistribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contentswill be complete or accurate or up to date. The accuracy of any instructions, formulae and drug dosesshould be independently verified with primary sources. The publisher shall not be liable for any loss,actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directlyor indirectly in connection with or arising out of the use of this material.

870

Journal of Toxicology and Environmental Health, Part A, 72: 870–883, 2009Copyright © Taylor & Francis Group, LLCISSN: 1528-7394 print / 1087-2620 online DOI: 10.1080/15287390902959516

UTEHThe Toxic Effects of Multiple Persistent Organic Pollutant Exposures on the Post-Hatch Immunity Maturation of Glaucous Gulls

Toxic Effect oF Pop on the Developing Immune SystemKjetil Sagerup1, Hans Jørgen S. Larsen2, Janneche Utne Skaare2,3, Grethe M. Johansen2, and Geir Wing Gabrielsen4

1Tromsø University Museum, Tromsø, Norway, 2Norwegian School of Veterinary Science, Oslo, Norway, 3National Veterinary Institute, Oslo, Norway, and 4The Polar Environmental Centre, Norwegian Polar Institute, Tromsø, Norway

This study tested whether the immune system of the glaucousgull (Larus hyperboreus) chicks became affected by existing envi-ronmental contaminants. An experimental group was given foodthat mimicked the natural contaminant mixture found in foodfrom the North Atlantic marine environment, while the controlgroup was given the equivalent of nearly clean food. All chickswere immunized with herpes virus (EHV), reovirus (REO), influ-enza virus (EIV), and tetanus toxoid (TET) in order to test theirability to respond to foreign specific antigens. At 8 wk, theexperimental group had 3- to 13-fold higher concentrations ofhexachlorobenzene (HCB), oxychlordane, p,p¢-DDE, and totalpolychlorinated biphenyls (SPCB) than did the control. Theexperimental group produced significantly lower antibody titeragainst EIV and had lower concentrations of immunoglobulin-G(IgG) and -M (IgM) in blood. Hematocrit percent and leukocytenumbers did not differ between the two groups. The ability oflymphocytes to proliferate in vitro was tested with three mitogens,phytohemagglutinin (PHA), concanavalin A (Con A), andpokeweed mitogen (PWM), and three antigens, keyhole limpethemocyanin (KLH), TET, and Mycobacterium avium subsp.paratuberculosis tuberculin purified protein derivative (PPD).The experimental group had a significantly higher peripheralblood lymphocyte response to PHA and to spleen lymphocytes invitro stimulated with Con A and PCB congeners 99 or 153, whilethe Con A, PWM, KLH, TET, PPD, and Con A plus PCB-156 or

-126 showed nonsignificant differences between groups. Dataindicate that the combined effect of multiple persistent organicpollution exposures occurring naturally in the Arctic negativelyaffect the immune system of the glaucous gull chick.

Persistent organic pollutants (POP) are well known to impairthe immune system, especially during the developmental phase(Hertz-Picciotto et al., 2008; Tryphonas, 1994). The impairmentcan be seen as reduced weight of immunological organs such asthe spleen and thymus, reduced numbers of white blood cells,reduced disease resistance, reduced antibody titers againstantigens, and changed response to stimulated lymphocytes(Tryphonas, 2005; Vos & Luster, 1989). These changes are dra-matic and affect disease resistance and survival. The contaminantcompositions and complexities in the wild are, however, differentfrom the individual congeners and technical mixtures used in mostexperimental studies. For example, POP in the environment aremodulated due to species-specific biomagnification patterns.Some are metabolized or excreted, while others are persistent andaccumulate (Walker, 1990). Congener-specific accumulation andbiomagnification lead to lower concentrations of less persistentsubstances and higher concentrations of persistent substancesupward through the food chain (Borgå et al., 2001). Animals in thewild therefore need to cope with mixtures of POP differently fromthose found in technical mixtures.

Studies of immune effects of POP mixtures in wild birdsand marine mammals are a relatively new branch of toxicology(Fairbrother et al., 2004). In a semi-field experimental study ofharbor seals (Phoca vitulina) where the two groups receivedherring (Clupea harengus) from a low-polluted versus pollutedarea, reduced natural killer (NK) cell activity and reducedmitogen-induced lymphocyte proliferate were associated withPOP exposure (deSwart et al., 1996, and references therein).In polar bears (Ursus maritimus) from Canada and Svalbardin the Norwegian Arctic, serum immunoglobulin G (IgG)

Received 22 October 2008; accepted 16 February 2009.We thank Grete Berntsen, Chris Bingham, Tine Borgen, Vegard

Bunes, Heather Coleman, Signe Haugen, Åse Krøkkje, AnuschkaPolder, Tor Gunnar Solvang, and Ole Gunnar Støen for technicalassistance during the experiment and with the analyses. We also thankRob Barrett and four anonymous reviewers for comments that greatlyimproved the article. The Transport and Effect Program and theNorwegian Pollution Control Authority were the main financialcontributors for the study. The Norwegian Polar Institute, theNorwegian School of Veterinary Science, the National VeterinaryInstitute, the Roald Amundsen Centre for Arctic Research, and theUniversity of Tromsø have also supported the study.

Address correspondence to Kjetil Sagerup, Tromsø UniversityMuseum, NO-9037 Tromsø, Norway. E-mail: [email protected]

Downloaded By: [University Library of Tromsoe] At: 23:24 30 July 2009

TOXIC EFFECT OF POP ON THE DEVELOPING IMMUNE SYSTEM 871

concentration was negatively correlated with polychlorinatedbiphenyls (PCB) levels (Bernhoft et al., 2000; Lie et al., 2004).Further, there were poorer antibody response to viruses and animpaired antigen- and mitogen-induced lymphocyte prolifera-tion found in the Svalbard population (Lie et al., 2004, 2005).Lie et al. (2004, 2005) attempted to determine whether higherlevels of PCB in the Svalbard population (Norstrom et al.,1998) were responsible for the observations seen, but due to(1) low sample size, (2) different gender distribution betweengroups, or (3) simply that there were no marked differences, nosignificant differences were noted in organochlorine levelsbetween Canada and Svalbard polar bears. In Caspian terns(Sterna caspia) and herring gulls (Larus argentatus), changesin antibody titers to sheep red blood cells (SRBC) and a strongnegative correlation for the phytohemagglutinin (PHA) skintest to organochlorine (OC) exposure were found when com-paring a polluted and a less polluted area (Grasman et al.,1996; Grasman & Fox, 2001). In glaucous gulls (Larushyperboreus) from Bjørnøya (Bear Island) in the Barents Sea,a positive correlation between nematode intensities and OCconcentrations (Sagerup et al., 2000) was interpreted asreduced ability of the immune system to fight parasites. A sig-nificantly reduced response to diphtheria toxoid vaccine withincreasing hexachlorobenzene (HCB) and oxychlordane levelsamong breeding female glaucous gulls (Bustnes et al., 2004)strongly supports this finding. In addition, the Bjørnøya popu-lation of glaucous gulls suffered from reduced reproductivesuccess, adult survival, and levels of thyroid hormones (T4 andT4:T3 ratio), fluctuating asymmetry in wing feathers, anddecreased feeding effectiveness as results of high POP levels(Bustnes et al., 2001, 2002, 2003, 2006; Verreault et al., 2004).

As top predators, the glaucous gull and the polar bear carrythe highest levels of POP in the marine ecosystem of the BarentsSea (Bernhoft et al., 1997; Bogan & Bourne, 1972; Mehlum &Daelemans, 1995; Norstrom & Muir, 1994). Glaucous gullsbreed throughout the Arctic and feed opportunistically in themarine food web (Barry & Barry, 1990; Erikstad, 1990). Theomnivorous feeding behavior and limited ability to metabolizepollutants, particularly the non-ortho and mono-ortho PCB con-geners (Henriksen et al., 2000; Walker, 1990), make the glau-cous gull at risk for elevated POP exposure (Bustnes, 2006).

This study investigated the influence of POP composition,as found in the wild, on the developing immune system ofglaucous gull chicks. Since the immune system is probably oneof the most sensitive target areas of the harmful effects of POPto the biota (Fairbrother et al., 2004; Tryphonas, 1994), thehumoral immune system was challenged with antigens to testits ability to produce specific antibodies. The total IgG andimmunoglobulin M (IgM) and the lymphocytes’ ability to pro-liferate in vitro were further measured. A reduced ability toproduce antibodies is a reliable indicator of an immunosup-pressive effect and a change in lymphocyte proliferation wouldfurther be a possible indicator of immune alterations by exist-ing POP of Norwegian Arctic.

MATERIALS AND METHODS

Experimental DesignGlaucous gull eggs were collected along the southwest

coast of Spitsbergen in June 1999. The eggs were incubatedand hatched at the Norwegian Polar Institute’s Research Sta-tion in Ny-Ålesund, Svalbard. As they hatched, chicks werealternately placed in the experimental and the control group.This selection was used instead of a complete random alloca-tion, since hatching date or time spent in the incubator couldbias the sample. In addition, late-hatched chicks have often areduced chance of survival compared to early-hatched chicks(Catry et al., 1998; Perrins, 1970). Of 40 chicks, one died dur-ing the first day and three were removed from the experimentdue to injuries. The study therefore included 19 and 17 chicksin the experimental and control group, respectively.

The glaucous gull chicks were held in indoor cages equippedwith heat lamps for the first 2 d after hatching. They were thenplaced in outdoor pens with chicks of the same age. The penswere protected against wind and rain and a heat lamp replacedthe adult gull as a heat source. The pens were big enough forchicks to freely choose their surrounding temperature. As thechicks grew, the heat lamp was removed and all chicks wereplaced in one big cage to ensure the same environmental condi-tions for all chicks. The cage was cleaned twice each day.

Both groups of chicks were hand-fed homogenates of eggs3 times per day during the first 26 d. The experimental groupreceived naturally contaminated gull eggs collected onHornøya at the northeast coast of Norwegian mainland. Theeggs were collected from herring gulls (Larus argentatus) andgreat black-backed gulls (L. marius) nests. The eggs representa Barents Sea POP cocktail of organochlorinated pesticides(OCPs), PCBs, brominated flame retardants, and low levels ofmercury (Barrett et al., 1996; Helgason et al., 2008a, 2008b).The control group received hen (Gallus domesticus) eggs. Inaddition to egg homogenates, the chicks were offered ad libi-tum access to fish (polar cod Boreogadus saida and herringClupea harengus, Table 1) 3 times per day during the first 3wk and thereafter twice per day. They had free access to water.From 1 wk of age, the chicks received 1 vitamin tablet (Fish-eater tablets, Mazuri Zoo Foods, England) every fourth day.Each glaucous gull chick received about 3 L (mean = 2990 ml,SD = 149) of contaminated or “clean” egg homogenate.

All sampling and bird handling were done in accordance tocurrent regulations of the Norwegian Animal Welfare Act. Per-mission was approved by the Norwegian Animal ResearchAuthority and the Governor of Svalbard (ref. 99/00324-2).

Analyses of POP in Feeds and ChicksThe analyses of POP were carried out at the Laboratory of

Environmental Toxicology at the Norwegian School of VeterinaryScience in Oslo. The laboratory is accredited for determinationof POP in biological material according to NS-EN ISO/IEC

Downloaded By: [University Library of Tromsoe] At: 23:24 30 July 2009

872 K. SAGERUP ET AL.

17025 (TEST 137). POP were quantified in blood, egg, andfish samples using the methods described by Brevik (1978)with modification by Bernhoft et al. (1997) and Andersen et al.(2001). Briefly, the samples were weighed and internalstandards were added (PCB-29, -112, and -207). The lipidswere extracted twice with cyclohexane and acetone, and lipiddetermination was done gravimetrically using an aliquot fromthe fish and egg sample extracts and the whole extracts fromthe blood samples. The blood extracts were then re-dissolvedin cyclohexane. For cleanup (i.e., removal of lipids) thelipid extracts were treated with ultraclean (purity 98.8%)

concentrated H2SO4 (Scanpure, Chemscan AS, Elverum, Nor-way). The sample concentrates were transferred to gas chroma-tography (GC) vials before analysis on a high-resolution GC(Agilent 6890 series gas chromatography system; AgilentTechnologies, Pennsylvania). The GC was equipped with an autosampler (Agilent 7683 series; Agilent Technologies) and a dualcolumn system with specifications SPB-5 and SPB-1701, both 60m, 0.25 mm internal diameter, and 0.25 μm film thickness(Supelco, Bellefonte, Pennsylvania) coupled to 2 63Ni m-electroncapture detectors (Agilent 6890 μ-ECD). Five-point linear calibra-tion curves (r2 ≥ 0.99) were used for quantification.

TABLE 1 POP Concentrations (ng/g wet wt) in Hen Eggs, Gull Eggs, Polar Cod, and Herring Fed to Glaucous

Gull Chicks in the Experiment

Hen egg, n = 2 a

Gull egg, n = 4 a

Polar cod, n = 6 a Herring b

Compound (mean ± SD) (mean ± SD) () (mean)

Lipid (%) 10.9 ± 0.5 10.1 ± 0.5 4.4 ± 2.4 6.9HCB 0.4 ± 0.1 40.9 ± 1.5 2.8 ± 1 1.1b-HCH n.d. 2.7 ± 0.2 n.a. 1.0c

Oxychlordane n.d. 34.3 ± 2.1 0.5 ± 0.2 <1.3d

trans-Nonachlor n.d. 22.3 ± 1.7 n.a. <0.5d

p,p′-DDE 0.4 ± 0.02 279.6 ± 11.1 1.3 ± 0.5 n.g.p,p′-DDT n.d. 18.5 ± 1.2 n.a. n.g.ΣDDTs 0.4 ± 0.02 298.1 ± 10.1 1.3 ± 0.5 10Mirex n.d. 3.4 ± 0.2 n.a. n.g.PCB-28 n.d. 5.5 ± 0.1 0.2 ± 0.1 n.g.PCB-52 n.d. 2.7 ± 0.1 0.5 ± 0.2 n.g.PCB-101 n.d. 8 ± 0.3 0.6 ± 0.2 n.g.PCB-99 n.d. 61.6 ± 1.6 0.6 ± 0.3 n.g.PCB-118 n.d. 141.7 ± 4.1 0.5 ± 0.2 n.g.PCB-153 0.6 ± 0.1 309.1 ± 14.9 0.5 ± 0.2 n.g.PCB-138 0.5 ± 0.1 254.5 ± 11.3 0.4 ± 0.1 n.g.PCB-156 n.d. 17.9 ± 0.6 n.a. n.g.PCB-180 1.7 ± 0.4 73.2 ± 10.9 0.3 ± 0.1 n.g.PCB-170 n.d. 21 ± 2.8 0.1 ± 0.04 n.g.Σ7PCBs e — — — 9.1Σ9PCBs f 2.8 ± 0.5 877.3 ± 43 3.8 ± 1.3 —Σ33PCBs g — 1141 ± 56.3 — —Exposure h 0.8 ± 0.2 324.1 ± 99.2 n.d. ± n.d. n.d.

Note. n.a., Not analyzed; n.g., not given; n.d., not detected.aHomogenized batches of about 100 eggs. Homogenized whole fish of polar cod.bData from NIFES (2008). Mean of 25 samples from years 1995, 1999, 2001, 2002, 2003, and 2006.cData from 2002 only (NIFES, 2008).dData from 2006 only (NIFES, 2008).eΣ7PCBs is the sum of PCB-28, -52, -101, -118, -138, -153, and -180 (NIFES, 2008).fΣ9PCBs is the sum of PCB-28, -52, -101, -99, -118, -153, -138, -180, and -170.gΣ33PCBs is the sum of PCB-28, -31, -52, -47, -74, -66, -56, -101, -99, -87, -136, -110, -151, -149, -118, -114, -

153, -105, -141, -137, -138, -187, -183, -128, -156, -157, -180, -170, -199, -196, -189, -194, and -206.hIn ng/g of Σ33PCB/g/d for d 1–26.

Downloaded By: [University Library of Tromsoe] At: 23:24 30 July 2009

TOXIC EFFECT OF POP ON THE DEVELOPING IMMUNE SYSTEM 873

In fish and blood samples, concentrations of HCB, oxychlor-dane, 1,1-dichloroethylene bis[p-chlorophenyl] (p,p¢-DDE),dichlorodiphenyltrichloroethane (p,p¢-DDT), and nine poly-chlorinated biphenyl (PCB) congeners with IUPAC numbers28, 52 (only fish), 99, 101, 118, 138, 153, 156 (only blood),170, and 180 were quantified. In addition, the following OCwere quantified in the egg homogenates: HCB, 1,2,3,4,5,6-hexachlorocyclohexanes (a-HCH, b-HCH, g-HCH), oxychlor-dane, trans-chlordane, cis-chlordane, trans-nonachlor, p,p¢-DDE, o,p¢-DDD, p,p¢-DDD, p,p¢-DDT, mirex, and 33 PCB con-geners: 28, 31, 52, 47, 74, 66, 56, 101, 99, 87, 136, 110, 151,149, 118, 114, 153, 105, 141, 137, 138, 187, 183, 128, 156, 157,180, 170, 199, 196, 189, 194, and 206. The a-HCH, g-HCH,trans-chlordane, o,p¢-DDD, p,p¢-DDD, and PCB-31, -87, -136,-151, and -199 were not detected in any of the samples.

The laboratory accredited analytical quality was approvedin several international intercalibration tests. Standard proce-dures were used to ensure adequate quality assurance and con-trol, and the precision, linearity, and sensitivity of the analyseswere within the laboratory accredited requirements. The detec-tion limits for individual compounds were determined as 3times the noise level and ranged from 0.004 to 1.38 ng/g wetweight (wet wt). Blank samples were analyzed continuouslyduring the experiment (2 blank/10 samples). The recoveries ofthe POP analyzed in the present study ranged from 80 to 115%.A more detailed description of the methods for extraction andcleanup, detection, calculation of congener concentrations, andquality control is given by Bernhoft et al. (1997), Andersenet al. (2001), and Murvoll et al. (2005).

ImmunizationEach of the 36 chicks was immunized at 15 and 43 d of age.

Blood samples were obtained from the brachial vein beforeeach immunization and at the end of the study, d 56 post hatch.A subcutaneous injection of 0.25 ml Resequin Plus (Ch-Bnumber 0200211, Hoechst Roussel Vet GmbH, Wisbaden,Germany) containing inactivated equine herpes virus 1 and 4(EHV), inactivated equine reovirus, serotype 1 and 3 (REO),and inactivated equine influenza virus A/equi 1/Prag/1/56, A/equi 2/Miami/63 and A/equi 2/Fontainebleau/1/79 (EIV) weregiven on both days. The glaucous gull chicks also received asecond subcutaneous injection of 0.25 ml Ovivac P (lot num-ber N232019/1, Hoechst Roussel Vet GmbH, Wisbaden,Germany) containing toxoid of Clostridium tetani (TET), Cl.perfringens, Cl. septicum, and inactivated cells from Cl. Chau-voei, Mannheimia haemolytica, and Pasteurella trehalosi and athird subcutaneous injection of 1 ml water-in-oil emulsion con-taining 0.5 ml suspension of hemocyanin from keyhole limpets(KLH) (100 μg/ml in phosphate-buffered saline) (numberH7017) (Sigma-Aldrich, Inc., St. Louis, MO) and 400 μgMycobacterium avium subsp. paratuberculosis/ml (NationalVeterinary Institute, Oslo, Norway) mixed with 0.5 mlFreund’s incomplete adjuvant (Gibco Bio-cult, Paisly, UK).

Immunological AnalysesSpecific Antibody Response

The assessment of specific antibodies following immuniza-tion was carried out using a virus neutralization test for anti-bodies against EHV, virus hemaglutination inhibition test forantibodies against REO and EIV, and passive hemagglutina-tion test for antibodies to TET (Lie et al., 2004).

Immunoglobulin GTo measure the amount of IgG in blood, a polyclonal antise-

rum to glaucous gull IgG was produced. In a preliminary study,several commercial anti-chicken IgG sera were tested, but theydid not cross-react well with glaucous gull IgG. Further, thebinding of IgG serum protein to staphylococcal protein A andstreptococcal protein G affinity columns was not successful.Therefore, gel filtration and ion-exchange chromatography(Catty, 1998) were combined to produce protein solutions forimmunization from glaucous gull blood. By using a proteinsolution with partly purified glaucous gull serum immunoglob-ulins for immunization of rabbits it was not possible to obtainrabbit antibodies that one could use in gel precipitation.Further immunizations of rabbits based on isolated immuno-precipitates would induce specific rabbit antibodies to glau-cous gull immunoglobulins recognized as foreign. Rabbitswere therefore subcutaneously immunized with 1 mg of thisprotein solution in Freund’s complete adjuvant containing0.5 mg/ml Mycobacterium smegmatis (Gibco). Subsequently,4 inoculations with protein solution in Freund’s incompleteadjuvant were administered at 3-wk intervals. The purity of theimmunoglobulin preparations and the specificity of each anti-serum were tested by immune electrophoresis in agar. Toobtain glaucous gull IgG in a pure and immunogenic form,immune electrophoresis and crossed electrophoresis were usedto give precipitates that were cut out of wet gels. The gels wereextensively washed to remove nonprecipitated, contaminatingantigens from the gels, and used for the second immunization.After repeated immunization with precipitates, the rabbits pro-duced antibodies that gave only one precipitation line in thetests (data not shown). Serum from different blood samples ofeach rabbit was pooled and stored in aliquots at –20°C. This“polyclonal rabbit anti glaucous gull IgG” serum was used tomeasure serum IgG concentration in a single radial immunediffusion assay, as described by Mancini et al. (1964). Thediameters of the precipitation zones were measured by a “mea-suring viewer” (Behring Institut, Germany), and IgG concen-trations (mg/ml) were calculated from a standard curve derivedfrom 5 dilutions on each plate of pooled glaucous gull serum,defined to contain 5 mg IgG/ml. The use of pooled glaucousgull serum as standard IgG in the radial immune diffusionassay in the absence of highly purified glaucous gull IgG wasnecessary in order to measure the difference in the serum IgGcontent of the groups using the immune precipitation in gelmethod. Each sample was set up in duplicate on separate plates

Downloaded By: [University Library of Tromsoe] At: 23:24 30 July 2009

874 K. SAGERUP ET AL.

and the mean values were used. The mean variation of the twomeasurements of the samples was 7.5%.

Immunoglobulin MA goat anti chicken IgM (Fc) (batch 5222, Nordic immuno-

logical laboratories) was used for measurement of glaucousgulls serum IgM concentration using the single radial immunediffusion assay as described earlier. The IgM concentrations(mg/ml) were calculated from a standard curve derived from 5dilutions on each plate of pooled glaucous gull serum, definedto contain 2 mg IgM/ml. Each sample was set up in duplicateson separate plates and the mean values were used. The meanvariation of the two measurements of the samples was 7.7%.

Lymphocyte ProliferationIn a previous study of wild glaucous gulls, the preparation

of lymphocytes for in vitro testing of mitogen-induced prolifer-ation was studied and several parameters were measured(H. J. S. Larsen, personal communication). Unlike mammals,the glaucous gull red blood cells and thrombocytes are nucle-ated. Therefore, conventional methods of separation that areused in commercial livestock animals need to be modified tobe applicable in glaucous gulls. Different density gradient pro-cedures and slow-speed centrifugation were tested. Themethod developed is based on carefully purified mononucle-ated cells from the blood and spleen of the glaucous gulls. Thelymphocyte response was measured after in vitro stimulationwith three mitogens: PHA, concanavalin A (Con A), andpokeweed mitogen (PWM). The optimal mitogen concentra-tion for in vitro stimulation was tested for each mitogen. Themethod was based on stimulation of lymphocytes to undergoproliferation and measured by incorporation of [3H]thymidine.Viability staining was about 90–95% and the recovery variedbetween 30 and 60% on approximately 100% purified mono-nucleated cells (H. J. S. Larsen, personal communication). Thelymphocyte responses to the mitogens in the present studywere similar to the corresponding responses in the previousstudy on glaucous gulls (H. J. S. Larsen, personal communica-tion). The optimal concentration and the same batch of mito-gens found in the previous study were used in the presentstudy.

Briefly, heparinized (15 IU/ml) blood was diluted 1:2 inHanks balanced salt solution (BSS) (Gibco Bio-cult, Paisley,UK) in siliconized vials and 9 ml diluted blood was placed ontop of 3 ml Lymfoprep (Axis-Shield PoC AS, Oslo, Norway)(1.077 g/ml) in siliconized vials. Each gradient was centrifugedat 600 × g for 25 min and the lymphocyte layer was transferredto a siliconized vial with heparin (150 IU) and washed twice(430 × g in 10 min) with 9.8 ml Hanks balanced salt solution(HBSS). The cells were resuspended each time carefully toavoid aggregation.

Cells were counted and diluted in the cell culture mediumRPMI 1640 to 1 × 106 mononucleated cells/ml as described by

Lie et al. (2005). Two hundred microliters of the cell solutionwas added to each well of 96-well flat-bottomed microtitrationplates (Nunc, Roskilde, Denmark). To each well, one mitogen(20 μl/well)—PHA (20 μg/ml, Murex Biotech Limited, Dartford,UK), Con A (20 μg/ml, Pharmacia, Sweden), PWM (20 μg/ml,Gibco, Renfrewshire, UK)—or one antigen (20 μl/well)—KLH (135 μg/ml, Sigma, St. Louis, MO), TET (45 μg/ml,Centre for International Health, Copenhagen, Denmark), andMycobacterium avium subsp. paratuberculosis tuberculin puri-fied protein derivative (PPD) (10 μg/ml, National VeterinaryInstitute, Oslo, Norway)—were added. The RPMI 1640medium was used as diluent and was also added (20 μl/well) tounstimulated control cultures. Plates were incubated at 37ºC ina humidified atmosphere of 5% CO2 and 95% air for 3 d forcultures stimulated with mitogens and 5 d for cultures stimu-lated with antigens. Each well was treated with 1 μCi methyl[3H]thymidine (specific activity 5 Ci/mmol, Amersham Bio-sciences UK Limited, Buckinghamshire, UK) in 20 μl culturemedium, further incubated for 1 d, and stored frozen (Lie et al.,2005). The cells were harvested with a cell culture Filtermate96 harvester (Packard BioScience BV, Groningen, the Nether-lands) onto 96-point glass microfiber filter plates (UnifilterGF/C), dried, and counted on a Top Count microplate scintilla-tion counter (Packard BioScience BV, Groningen, the Nether-lands) after addition of scintillation liquid to the sealed filters.All assays were done in triplicate cultures. Lymphocyteresponse (Δ counts per minute, cpm) is therefore the differencebetween the mean stimulated cultures and mean unstimulatedcontrol cultures. The mean values of the unstimulated controlcultures for assaying mitogen and antigen response were 3405(SD = 3453) and 3687 (SD = 4708) cpm, respectively.

The lymphocyte response to mitogens and antigens was alsocarried out using splenocytes. The spleens were removed asep-tically into HBSS, washed twice with HBSS, and cut open ateach end. Spleen cells were suspended into the culturemedium. Single-cell suspensions of splenocytes were made bywashing the cells through a sterile tissue sieve equipped withan 80-mesh stainless-steel screen. The splenocyte suspensionsin 25 ml HBSS were placed in 50-ml tubes and clumps wereallowed to settle out for 5 min. The upper suspension was over-layered onto Lymfoprep for density gradient centrifugation topurify lymphocyte cultures for the lymphocyte proliferationassay as described earlier. The mean values of the unstimulatedcontrol cultures of splenocytes were 7748 (SD = 3104) and5300 (SD = 3113) cpm for mitogen and antigen response,respectively.

PCB, PCB-99 (200 ng/ml), PCB-126 (0.3 ng/ml), PCB-153(1500 ng/ml), and PCB-156 (50 ng/ml) resuspended in dime-thyl sulfoxide (DMSO; Sigma) were added (40 μl/well) to lym-phocyte cultures stimulated with Con A and incubated asdescribed earlier. This investigation was carried out to seewhether in vitro exposure with different PCB would influencethe Con A-induced proliferation of lymphocytes from glaucousgull chicks that were exposed to POP in vivo. For all

Downloaded By: [University Library of Tromsoe] At: 23:24 30 July 2009

TOXIC EFFECT OF POP ON THE DEVELOPING IMMUNE SYSTEM 875

compounds dissolved in DMSO the final DMSO concentrationwas prepared to be less than 0.5% in the medium. The 0.5%solution of the DMSO solvent was added to the control. Thelymphocyte response (Δ Con A, cpm) of unexposed culturewas expressed as the difference of the mean counts per minute(cpm) of Con A-stimulated lymphocyte and splenocyte cul-tures and unstimulated lymphocyte and splenocyte control cul-tures, respectively. The RPMI 1640 medium was added (20 μl/well) to unstimulated control cultures. The response of Con A-stimulated cultures exposed with PCB was expressed as ΔPCB(congeners 99, 156, 153, 126) (mean cpm) for Con A- andPCB-stimulated lymphocytes minus the mean cpm of controlcultures (containing DMSO).

The relative effect of in vitro exposure on responding cellscompared with unexposed responding cells was expressed aspercent delta (%Δ) of cells stimulated with Con A and PCB (99,56, 153, 126) minus the control (containing DMSO), divided bythe response difference between Con A-stimulated and unstim-ulated cell cultures (containing only RPMI 1640 medium).

HematologyThe total erythrocyte counts (1012 × RBC/L) were assayed

using Mini-Pak ERY (Ames) and a mini-photometer (Compur);hemoglobin (HGB g/L) content was measured using Mini-Pak

Hb (Ames); and the hematocrit (HCT%) was estimated by con-ventional capillary tube centrifugation. The relative numbers(%) of lymphocytes, monocytes, neutrophils (heterophils),basophils, and eosinophils were determined in stained (Diff-Quick, Dade Behring, Inc.) blood smears.

Statistical AnalysisThe POP levels in the food items (Table 1) and in blood

drawn at d 15 and d 56 (Table 2) are presented without valuesfor levels below detection limits. The statistical analyses werecarried out using the free statistical software R (R DevelopmentCore Team, 2008). The significance level was set at p = 0.05. Foreach of the two groups, the residuals from the immunologicalresponse variables were visually investigated for constant vari-ance and normal distribution through quantile–quantile (q-q)plots. If the linearity in the plot was skewed, the variable wasnaturally log-transformed prior to the analyses. Statistical ana-lyzing of responses was performed with analysis of variance(ANOVA) controlled with the body condition index (BCI). TheBCI was calculated using principal component analysis toobtain a single measure of size (Jolicoeur & Mosimann, 1960).The first principal component was calculated from the totalhead length and wing length as a size index. The size index wascalculated separately for each sex because glaucous gulls are

TABLE 2 POP Concentrations (ng/g wet wt) in Blood of Glaucous Gull Chicks Fed “Clean” (contr) and Naturally

Contaminated (exp) Food at d 15 and 56

Day 15 Day 56

Contr n = 6 Exp n = 9 Contr n = 18 Exp n = 19Exp / contr

Component Mean ± SD Rnge Mean ± SD RangeMean ±

SD RangeMean ±

SD Range d 15 d 56

Lipid (%) 0.8 ± 0.2 0.5–1 0.8 ± 0.2 0.6–1.1 0.6 ± 0.1 0.4–0.7 0.6 ± 0.1 0.3–0.8 1 1HCB 1.4 ± 0.6 0.6–2.4 10.5 ± 3 6.8–15.6 0.4 ± 0.1 0.2–0.5 1.2 ± .3 0.8–2 8 3Oxychlordane 1.6 ± 1.3 0.5–3.6 9.6 ± 3.6 5–15.9 0.2 ± 0.1 n.d.–0.4 1.1 ± 0.4 0.6–1.9 6 7p,p′-DDE 11.3 ± 9.5 3.4–26 72.4 ± 26 34.9–120.2 0.7 ± 0.5 0.4–2.2 6.1 ± 2.2 3.5–11.2 6 8p,p′-DDT n.d. — 4.3 ± 1.8 1.8–7.4 0.3 ± 0.04 n.d.–0.3 0.4 ± 0.1 n.d.–0.8 — 2PCB-28 0.2 ± 0.03 n.d.–0.2 1.1 ± 0.4 0.6–1.8 0.03 ± 0.0 n.d.–0.03 0.1 ± 0.03 0.05–0.1 7 3PCB-101 0.3 ± 0.1 0.1–0.4 1.5 ± 0.5 0.8–2.3 0.1 ± .03 n.d.–0.2 0.2 ± 0.1 n.d.–0.4 6 1PCB-99 2.1 ± 1.3 0.9–4.3 17.6 ± 5.7 8.9–27.1 0.2 ± 0.1 n.d.–0.4 1.7 ± 0.6 1–2 8 9PCB-118 5 ± 3.7 1.7–11.3 43.8 ± 15.6 20.9–71 0.4 ± 0.2 0.2–0.9 4 ± 1.4 2.3–7.2 9 11PCB-153 12.9 ± 11.9 3.5–33.7 97.3 ± 32.2 46.7–157.4 0.8 ± 0.5 0.4–2.1 8.7 ± 3 4.8–16.1 8 11PCB-138 7.6 ± 6.2 2.5–18.7 78.5 ± 25.8 37.4–122.8 0.5 ± 0.3 0.3–1.4 7.2 ± 2.5 4–13.5 10 13PCB-156 1.7 ± 0.6 n.d.–2.1 5.8 ± 2 2.6–9.4 n.d. — 0.5 ± 0.2 n.d.–1 3 —PCB-180 4.1 ± 4 0.9–10.6 21.2 ± 7.7 9.5–37.4 0.3 ± 0.2 n.d.–0.6 1.7 ± 0.6 0.9–3.1 5 7PCB-170 1.5 ± 1.1 n.d.–2.8 6.5 ± 2.2 3.1–11.1 0.1 ± 0.1 n.d.–0.2 0.6 ± 0.2 0.3–1.1 4 7Σ9PCBa 33.5 ± 29.3 9.7–84.2 273.2 ± 91.3 130.3–440.2 2.2 ± 1.3 1.1–5.4 24.5 ± 8.6 13.7–45.5 8 11

Note. n.d., Not detected. Exp to contr ratios = concentration ratios (experiment/control) for d 15 and 56, respectively.aΣ9PCB is the sum of PCB-28, -101, -99, -118, -153, -138, -156, -180, and -170.

Downloaded By: [University Library of Tromsoe] At: 23:24 30 July 2009

876 K. SAGERUP ET AL.

sexually dimorphic. The size index and the body mass variableswere standardized within each sex (mean = 0, SD = 1) to createa common BCI. Data for each sex were pooled, and the residu-als from the linear regression of standardized body mass andstandardized size index were used as the BCI (Jakob et al.,1996). The BCI was included in the final models due to expec-tations that sexual hormones and nutritional state might influ-ence the development of the immune system. No adjustmentswere made for multiple comparison, since such comparisonsdecrease the chances of rejecting the null hypothesis that are notnull (type II error) (Rothman, 1990).

RESULTS

GrowthThe growth curve of the glaucous gull chicks was sigmoid

in shape (Figure 1). There were significant differences betweensexes in body mass, beak length, and total head length after 1wk of age. In addition to the hand-fed egg homogenate, thechicks received fish ad libitum. The body mass gain in the lin-ear area of the growth (37.9 g/d between d 14 and 39) was 3.5g/d more in the present study than in wild glaucous gull chicksfrom Bjørnøya (34.4 g/d from Henriksen, personal communi-cation). This corroborates our postmortem observation that thechicks had deposited large stores of body fat. The growth curvealso showed that the two immunizations, which entailed extrahandling at d 15 and 43, only had a small impact on the growthpatterns. The weight drop from d 43 to 44 was only about 3%and the same patterns were seen for both groups and sexes(Figure 1).

OC in the FeedOC levels in the polar cod, herring (NIFES, 2008), and hen

eggs were much lower than in gulls eggs (Table 1). The meanHCB concentrations were 2.8, 1.1, 0.4, and 40.9 ng/g wet wt inthe polar cod, herring (NIFES, 2008), hen eggs, and gulls eggs,respectively. The corresponding concentrations of Σ9PCBs(sum of 9 PCB) or Σ7PCB for herring (NIFES, 2008), were3.8, 9.1, 2.8, and 877.3 ng/g wet wt, respectively. Only 5 of 45analyzed POP were detectable in hen eggs (Table 1).

OC in BloodOC concentrations in blood from glaucous gull chicks were

significantly higher in the experimental than in the controlgroup at both d 15 and 56 for each of the OC compounds andPCB congeners (Table 2). The concentration ratios (meanexperimental/mean control) were calculated for each OCPcompound and PCB congener. The ratio of HCB, PCB-28, andPCB-101 decreased from d 15 to 56 (Table 2). A decreasedratio indicates that more of the compound is eliminated fromthe polluted experimental group than from the cleaner controlgroup. The concentration ratios for the oxychlordane, p,p′-DDE, and PCB-99, -118, -153, -138, -156, -180, and -170increased from d 15 to 56 (Table 2). In contrast to HCB, PCB-28, and PCB-101, which were decreasing, these compoundswere more persistent.

The measured compounds were significantly correlated in theexperimental group, except for p,p′-DDT. In the control group,where the levels were much lower (Table 2), the intercorrela-tions were not clear. The oxychlordane, p,p′-DDE, and PCB-99,

FIG. 1. Mean daily growth increment of female and male glaucous gull chicks raised in captivity in Ny-Ålesund, Svalbard. Arrows indicate the times ofimmunization.

0

200

400

600

800

1000

1200

1400

1600

1800

0 5 10 15 20 25 30 35 40 45 50 55 60Days post-hatch

Gra

ms

M exp. n = 5

M contr. n = 10

F exp. n = 14

F contr. n = 7

Downloaded By: [University Library of Tromsoe] At: 23:24 30 July 2009

TOXIC EFFECT OF POP ON THE DEVELOPING IMMUNE SYSTEM 877

-118, -153, -138, and -180 were significantly correlated,whereas there was no association with HCB, p,p¢-DDT, andPCB-101. In many of the control group samples, concentrationsof PCB-28, -156, and -170 were below the detection limit. Theintercorrelation could therefore not be calculated for these PCB.

Specific Antibody ResponseThe experimental group produced a significantly lower anti-

body titer against the EIV antigens than the control group(Table 3). The antibody response to EHV, REO, and TET didnot differ between the groups (Table 3). Since the intercorrela-tions between OC were high, only individual OCP and thesummarized Σ9PCB concentrations were used to test the anti-body responses within the control and experimental groups.Linear models controlled for BCI revealed that EIV significantpositively correlate with p,p′-DDE and Σ9PCB in the experi-mental group.

ImmunoglobulinThe levels of IgM decreased from d 15 to d 56, while the

levels of IgG increased during the same period. There were nodifferences between groups at d 15 for either IgM or IgG.

However, at the end of the study, when maternal immunity islow and self-produced antibody production is established, thecontrol group had higher levels of circulating IgG and IgMthan the experimental group (Table 4). The specific antibodytiters to EHV, REO, EIV, or TET did not correlate with IgM orIgG, neither all together nor separately within the control andexperimental group.

The immunoglobulin intercorrelation and correlation toOCP and Σ9PCB were tested separately within the control andexperimental group. This was done since (1) the group immu-noglobulin responses were significantly different and (2) OClevels were distinctly different. For the control group, whichhad the highest concentration of immunoglobulins, the chicksthat produced the highest level of IgG and IgM at d 15 were thesame that produced high IgG and IgM levels at d 56. There wasalso a positive, nonsignificant trend between IgG and IgM at d56. In the experimental group, IgM levels were positively cor-related between d 15 and 56, and IgG and IgM levels were pos-itively correlated at d 56. The immunoglobulin levels were notsignificantly correlated to individual OCP or Σ9PCB concen-trations, neither in the experimental nor in the control group.The PCB levels could therefore not explain the within-groupimmunoglobulin variance.

TABLE 3 The Secondary Antibody Response Titer of Glaucous Gull Chicks Fed Naturally Contaminated (Experimental Group) and “Clean” (Control Group) Food to Immunizations of Herpes Virus (EHV), Reovirus (REO), Influenza Virus (EIV),

and Tetanus Toxoid (TET) Measured 56 d Post-Hatch

Experimental group Control group

n Mean ± SD Range n Mean ± SD Range F p

EHV 16 3.3 ± 0.8 2–5.5 15 3.2 ± 0.6 2–4 0.2 0.7REO 17 3.0 ± 0.0 3–3 15 3.1 ± 0.5 3–5 1.2 0.3EIV 17 4.4 ± 0.8 2.3–5.3 15 5.5 ± 1.3 3.3–7.3 10.7 0.003a

TET 17 5.7 ± 2.2 2–9 15 5.5 ± 2.3 2–9 0.05 0.8

aStatistically significant.

TABLE 4 Levels of IgG and IgM in Blood (mg/ml) (SD) of Glaucous Gull Chicks Fed naturally Contaminated

(Experimental Group) and “Clean” (Control Group) Food at d 15 or 56 Post-Hatch

Experimental group Control group

Age Ig n Mean ± SD Range n Mean ± SD Range F p

d 15 IgG 17 0.4 ± 0.3 0.1–1.3 17 0.4 ± 0.2 0.1–0.8 0.3 0.6IgM 19 0.9 ± 0.4 0.2–1.8 17 0.9 ± 0.5 0.0–1.8 0.2 0.7

d 56 IgG 19 4.6 ± 1.1 2.8–7.2 16 5.9 ± 2.1 2.4–10.6 5.1 0.03a

IgM 19 0.5 ± 0.3 0.0–1 16 0.7 ± 0.3 0.2–1.2 4.3 0.05a

aStatistically significant.

Downloaded By: [University Library of Tromsoe] At: 23:24 30 July 2009

878 K. SAGERUP ET AL.

Lymphocyte ProliferationMitogen-Induced Lymphocyte Proliferation

All glaucous gulls showed moderate to high nonspecific pro-liferation response of peripheral blood T-lymphocytes to PHA at56 d post-hatch, with ΔPHA ranging from 9800 to 120,500 cpm.The mean ΔPHA response was significantly higher in the experi-mental group (Table 5), indicating that a larger part of peripheralblood lymphocytes respond to PHA in vitro in the experimentalgroup compared to the controls. Almost all glaucous gull chicks(97%) showed a lymphocyte response to Con A (ΔCon A, notdetected [ND]–103,600 cpm) and about 67% responded to PWM(ΔPWM, ND–11,700 cpm). The mean response of peripheral

blood lymphocytes to both Con A and PWM tended to be higherin the experimental than in the control group, but due to highvariance, the difference was not significant (Table 5).

In all cultures from spleen cells, PHA induced a similarstrong response as in blood, with ΔPHA ranging from 14,800to 101,800 cpm. The response to Con A was weaker (ΔCon A,ND–31,400 cpm) and about 75% of the cultures responded toCon A. In the spleen lymphocyte cultures stimulated withPWM, only 59% of the cell cultures responded (ΔPWM, ND -8,300 cpm). There was no difference between the groupsregarding the mean delta response induced by mitogen onspleen lymphocyte cultures (Table 5).

TABLE 5 The Mean and Standard Deviation (SD) for In Vitro Lymphocyte Proliferation (Δcpm = cpm after – cmp before stimulation)

in Glaucous Gull Chicks Fed Naturally Contaminated (Experimental Group) and “Clean” (Control Group) Food Measured at 56 d Post-Hatch

Experimental group Control group

Parameter n Mean ± SD n Mean ± SD F p

Lymphocyte proliferation ΔPHA 19 63,127 ± 30,814 17 44,051 ± 18,306 4.8 0.04c

Blood cells ΔCon A 19 30,633 ± 31,335 17 19,161 ± 11,984 2.2 0.2ΔPWM 19 2188 ± 3081 17 1354 ± 2078 0.8 0.4ΔKLH 19 4810 ± 4867 17 4333 ± 4891 0.3 0.6ΔTET 19 3735 ± 3347 17 2425 ± 2766 1.6 0.2ΔPPD 19 2406 ± 2635 17 1730 ± 1497 0.4 0.5Δ99a 19 26,106 ± 24,115 17 21,025 ± 13,363 1.4 0.2Δ156a 19 29,701 ± 28,975 17 22,863 ± 13,348 0.6 0.5Δ153a 19 29,182 ± 28,817 17 22,088 ± 13,789 0.2 0.6Δ126a 19 28,828 ± 29,022 17 20,854 ± 13,678 0.4 0.5%Δ99Con Ab 18 130 ± 110 17 123 ± 72 0.3 0.6%Δ156Con Ab 18 129 ± 96 17 139 ± 93 0.4 0.5%Δ153Con Ab 18 131 ± 109 17 118 ± 52 0.0 1.0%Δ126Con Ab 18 136 ± 116 17 113 ± 50 1.3 0.3

Lymphocyte proliferation ΔPHA 17 53,399 ± 23,237 15 45,864 ± 21,540 0.9 0.4Spleen cells ΔCon A 17 4813 ± 7431 15 2883 ± 6303 1.0 0.3

ΔPWM 17 1238 ± 2675 15 999 ± 3305 0.0 0.8ΔKLH 17 195 ± 379 15 899 ± 2644 1.2 0.3ΔTET 17 1120 ± 1386 15 1369 ± 1730 0.1 0.8ΔPPD 17 2063 ± 2739 15 2645 ± 4267 0.0 0.9Δ99a 19 4998 ± 7994 17 2602 ± 5860 5.3 0.03c

Δ156a 17 5342 ± 8063 15 2140 ± 5873 3.8 0.01Δ153a 17 4797 ± 6939 15 2890 ± 6727 4.4 0.04c

Δ126a 17 5935 ± 7902 15 2378 ± 5518 3.5 0.1%Δ99Con Ab 15 123 ± 115 9 37 ± 42 3.5 0.1%Δ156Con Ab 15 130 ± 107 9 29 ± 38 3.7 0.1%Δ153Con Ab 15 137 ± 151 9 42 ± 54 4.3 0.05c

%Δ126Con Ab 15 126 ± 96 9 42 ± 45 2.6 0.1

aΔPCB (99, 156, 153 or 126) = (Con A + PCB – control).b% delta (%Δ) = 100 × (Con A + PCB (99, 156, 153 or 126) – control)/(Con A – control).cStatistically significant.

Downloaded By: [University Library of Tromsoe] At: 23:24 30 July 2009

TOXIC EFFECT OF POP ON THE DEVELOPING IMMUNE SYSTEM 879

Antigen-Induced Lymphocyte ProliferationMost of the gulls showed a specific response of peripheral

blood T lymphocytes to KLH (83%), TET (92%), and PPD(72%) 2 wk after the secondary immunization, with a ΔKLHresponse ranging from ND to 16,800 cpm, ΔTET from ND to10,300 cpm, and ΔPPD from ND to 8000 cpm. There were nodifferences between the groups regarding the mean response ofthe blood lymphocytes (Table 5). In spleen cell cultures from thegulls, the response to KLH (31%), TET (72%), and PPD (66%),ΔKLH, ranged from ND to 10,300 cpm, ΔTET from ND to 6000cpm and ΔPPD from ND to 13,900 cpm. The spleen lympho-cytes did not respond differently in the two groups (Table 5).

Correlation Patterns in Lymphocyte Response ParametersIntercorrelations for the mitogen- and antigen-induced lym-

phocyte proliferation of peripheral blood lymphocytes for eachof the experimental and control groups were analyzed. ThePHA response was significantly positively correlated toresponses from Con A in both groups and to PWM and TETresponses in the experimental group. The TET response wasalso positively correlated to Con A and PWM responses in theexperimental group, and KLH response to PPD response inboth groups. The KLH-stimulated spleen lymphocyte responsewas positively correlated with TET and PPD responses, andTET with PPD in both groups.

As with antibody responses, linear models of the within-group lymphocyte proliferation responses from peripheralblood and spleen lymphocytes to Σ9PCBs were calculated. Theperipheral blood lymphocytes stimulated with PWM correlatedpositively with Σ9PCB in the experimental group. There wasalso a negative correlation between blood lymphocyte PHAresponse and Σ9PCB in the control group.

Effect of In Vitro Exposure on Lymphocyte ProliferationTo test the effect of in vitro exposure to different PCB, the

congeners (PCB-99, -156, -153, and -126) were added to lym-phocyte cultures from peripheral blood and spleen stimulatedwith Con A. In all tests, the experimental group had a highermean response than control. However, only the experimentalspleen lymphocyte responses to PCB-99 (Δ99) and PCB-153(Δ153 and %Δ153Con A) were significantly different from thecontrol (Table 5).

The intercorrelations between the different measurements ofthe PCB-stimulated cells were high. All measurements in eachof the cell types (peripheral blood/spleen lymphocytes) and thecalculation (Δ/%Δ) were significantly and positively corre-lated. This indicates that the blood and spleen lymphocytesreacted similarly to the PCB stimuli and that different PCBcongeners (99, 156, 153, and 126) produced similar responses.

HematologyThe RBC counts among individual glaucous gulls at 56 d

post-hatch varied between 5 and 7 × 1012 cells/L, but there was

no difference between the experimental (mean 6 × 1012 cells/L)and control group (mean 5.8 × 1012 cells/L). The hemoglobincontent varied between 10.4 and 15.7 g/L, but there was againno difference between experimental (mean 12.8 g/L) and con-trol group (mean 12.7 g/L). Furthermore, the mean hematocritvalue of the control group (39.1%) was similar to that of theexperimental group (38.2%), and the relative numbers (%) oflymphocytes, monocytes, neutrophils (heterophils), basophils,and eosinophils did not differ between groups.

DISCUSSIONThe feeding experiment was designed to study the com-

bined toxic effects of naturally occurring POP exposure on theimmune system of growing glaucous gull chicks during thefirst 8 wk after hatching. The model used focused on theeffects of pollutants on the immune system during the matura-tion period and the results demonstrated that there was indeed anegative response. The ability to produce antibodies againstinfluenza virus following immunization was impaired and low-ered IgG and IgM serum levels were found. Furthermore, bothgroups respond to in vitro stimulation of peripheral blood andspleen lymphocytes. However, both in vitro PHA-induced pro-liferation of peripheral blood lymphocytes and in vitro Con Aplus PCB exposure of spleen lymphocytes were influenced byPOP contamination.

The small body mass changes following immunization indi-cated that the glaucous gull chicks did not react negatively toextra handling (Figure 1). The small change in body mass afterthe second immunization was about half the variation in bodymass. The change may have been produced by a short timereduction in appetite as a response to the extra handling or theimmunizations. It was therefore concluded that the chicks werenot exposed to handling stress or captivity stress that mightaffect the disease resistance or the immune responses. Treat-ments and response to immunization were the same in the twogroups (Figure 1) and did not confound the results.

At the end of the experiment, the chicks were at an agewhere they could have been flying (Løvenskiold, 1964). Com-pared to wild glaucous gull chicks, our chicks were probablyless active during the last part of the study. If less fat had beendeposited through increased activity, the blood concentrationsof OC would probably have been higher. Since biochemicaleffects of pollution are dependent on a certain level beingreached before effects are initiated (Walker et al., 2006),impaired immune responses may thus have been masked untilstarvation, at which point high bioavailable blood concentra-tions would occur.

The composition of the food given to the control and experi-mental group was chosen to reflect low- and high-contamina-tion exposure. The main reason to use herring gull and greatblack-backed gull eggs, instead of a commercial PCB product,was that the congener composition changes through the naturalfood chain (Borgå et al., 2001). The eggs from these two

Downloaded By: [University Library of Tromsoe] At: 23:24 30 July 2009

880 K. SAGERUP ET AL.

species were naturally contaminated and contained a SouthernBarents Sea POP mixture (Barrett et al., 1996). The glaucousgull only occasionally preys on eggs from other gull species(Barry & Barry, 1990). However, the POP composition in gulleggs is similar to that in auk eggs (Borgå et al., 2001), which arepart of the natural food of glaucous gulls on Bjørnøya (Bustneset al., 2000). By choosing gull eggs as feed, it was expected thatthe experimental group would receive a pollutant compositioncorresponding to that of wild glaucous gull chicks.

In order to compare blood POP levels in our laboratory-raised chicks to liver POP levels in wild glaucous gull chicks(Henriksen, 1999), the present blood POP levels were lipidnormalized and multiplied by a factor of 1.5 for ΣDDT andΣ9PCB and by 1 for HCB and oxychlordane (Henriksen et al.,1998). These factors compensate for the liver-to-blood POPconcentrations differences found in glaucous gull (Henriksenet al., 1998). The hepatic concentrations of OC from 37-d-oldglaucous gull chicks from Bjørnøya were within the adjustedlevels of this study’s experimental group (Table 6). The con-centrations of POPs in the experimental group were thereforecomparable with those in wild glaucous gull chicks fromBjørnøya.

Unfortunately, heavy metals, essential elements, vitamins,and fatty acids were not measured in the experimental- or con-trol food. While the lipid quantity of hen and gull eggs was thesame (Table 1), the compositions of the polyunsaturated fattyacids (n-3) probably differed, since the gull eggs came from amarine environment. The concentrations of heavy metals andessential metals in arctic seabirds, including the glaucous gull,do not, however, exceed the threshold levels for biologicaleffects (Borgå et al., 2006). The glaucous gull chicks were fed“Fish eater tablets” that contained protein, fat, vitamins, andessential elements. This might have leveled out some of themarine to terrestrial differences. However, food differences inthe two groups cannot be excluded as a bias.

At the end of the experiment, the OC concentrations in theexperimental group were about 10% of the OC concentrations

at d 15 (Table 2). The mean body mass almost doubled after d26 (Figure 1) when the egg feeding stopped. The decrease inOC concentrations was therefore mainly an effect of growthdilution. The reduction in the experimental to control grouplevel ratio (Table 2) indicates that the gulls metabolize and pre-sumably excrete these pollutants. This is congruent with foodchain studies that showed that these compounds were notenriched in seabirds and marine mammals (Borgå, 1997; Borgået al., 2005; Muir et al., 1988). However, the ratio increases(Table 2), indicating that these compounds may be more per-sistent (Henriksen et al., 2000).

Specific Antibody ProductionThe production of specific antibodies to influenza virus was

significantly lower following immunization in glaucous gullchicks exposed to POP compared with controls at d 56 post-hatch. This effect on antibody formation after the secondimmunization may reflect an impairment of the antibodyresponse to influenza virus in that the response is eitherdepressed or delayed. A rapid production of high titer of pro-tective antibodies against viruses, such as influenza and parain-fluenza virus, is crucial for survival during epidemics. It istherefore reasonable to assume that elevated POP levels maybe associated with a decreased resistance to some infections inglaucous gull chicks. In free-ranging polar bears feeding oncontaminated seal blubber, it was suggested that PCB and OCPmay exert a suppressive role in antibody response againstinfluenza virus and reovirus (Lie et al., 2004). In the polar bearstudy and in the present investigation the virus hemagglutina-tion inhibition test was used to assay the antibody titer ofblocking antibodies to both viruses. In the present study, onlylow levels of neutralizing antibodies against reovirus and her-pes virus were found, indicating that these antigens were lessimmunogenic in glaucous gull chicks. In humans, perinatalexposure to PCB through traditional diets including contami-nated whale blubber was shown to adversely impair the

TABLE 6 OC Levels (ng/g lipid Wt.) of Glaucous Gull Chicks Given Naturally

Contaminated (Experimental Group) Food at 56 d Post-Hatch From Present Study and Chicks Collected on Bjørnøya in 1996

OrganochlorineExperimental group, n = 19, mean (range)

Bjørnøya chick, n = 21, mean (range)c

HCB 229 (151–404)a 449 (162–1193)Oxychlordane 206 (106–367)a 395 (89–2498)ΣDDT 1796 (854–3176)b 1540 (393–8701)Σ9PCB 6804 (3359–12,018)b 4790 (1028–23,815)

aAdjustment factor 1.0 (Henriksen et al., 1998).bAdjustment factor 1.5 (Henriksen et al., 1998).cData from (Henriksen, 1999).

Downloaded By: [University Library of Tromsoe] At: 23:24 30 July 2009

TOXIC EFFECT OF POP ON THE DEVELOPING IMMUNE SYSTEM 881

immune response to childhood vaccinations (Heilmann et al.,2006). Heilmann et al. (2006) found a negative correlationbetween PCB exposure and antibody titer to diphtheria toxoidat 18 mo of age. The antibody response decreased by 24% foreach doubling of PCB exposure. However, the response todiphtheria toxoid was lower at an age of 7 yr and not associatedwith exposure to PCB. Heilmann et al. (2006) further foundthat the antibody response to another vaccine, the tetanus tox-oid, mainly was effective at 7 yr of age and not at 18 mo of age.The difference in association between PCB exposure and theantibody response to the two toxoid vaccines was explained bya relatively higher antigenicity of the tetanus toxoid (Heilmannet al., 2006). In the present study, no effect of POP exposure onthe antibody titer to tetanus toxoid at 56 d of age was noted.This is in accordance with a study on adult free-ranging glau-cous gulls (Bustnes et al., 2004), although female glaucousgulls with high HCB or oxychlordane levels showed adecreased antibody response to diphtheria toxoid 2 wk afterimmunization.

Contrary to the significantly higher antibody titer to EIV inthe control group, which received low levels of POP, a positiverelationship between p,p′-DDE and Σ9PCB levels and EIVantibody titer existed within the experimental group. Thesecontradictory results are difficult to explain, and data suggestthat they arose as a coincidence due to low sample size (n = 17)and that the two chicks with highest OC levels also have thisgroup’s maximum level of EIV (Table 3). Since the antibodyEIV titer was impaired in the experimental group, it is reason-able to assume that the elevated POP levels were associatedwith a decreased resistance to virus infection in glaucous gullchicks. This is supported by the numerous adverse effects,including immunological effects, related to OC levels in glau-cous gulls from Bjørnøya (Bustnes, 2006).

ImmunoglobulinDue to transfer of maternal IgG before hatching, the chicks

have high levels of IgG in their serum during the first week oflife. After 1 wk, the amount of maternal IgG falls, but the totalIgG increases as the chick produces intrinsic IgGs. The level ofserum IgG builds up over time after exposure to microorgan-isms in the environment. Since young individuals are repeat-edly exposed to new microbes, their IgG levels successivelyincrease until a steady state is reached. The IgG levels thereaf-ter fluctuate with infection status. Serum IgG concentration inthe experimental glaucous gull chicks, which were exposed toPOP, would therefore reflect the degree of influence POP exerton the immunoglobulin production when compared to the con-trol. In the present study, levels of IgG were low in all chicksand no difference was found between the 2 groups at d 15. Thismay be due to a combined effect of a low transfer and/or highmetabolism of maternal IgG. As expected, at 56 d, the levels ofIgG were higher in all chicks. The lower level of IgG in theexperimental group may be related to a POP-induced lower

ability to react against microbes in the environment and togiven immunizations. This observation corroborates the nega-tive relationship between IgG levels and PCB concentrationsfound in polar bears in two different experiments (Bernhoftet al., 2000; Lie et al., 2004).

The production of IgM class antibodies starts soon afterimmunization. These antibodies are important in the earlystage of an infection and are replaced by IgG antibodies. TheIgM molecules are not as specific as IgG antibodies, butrespond to new microbes faster than IgG. Glaucous gull chicksthus have a high production of IgM class antibodies during thefirst weeks of life. The fall in IgM levels between d 15 and 56in both groups of chicks in this study is believed to be pro-duced by replacement of IgM by IgG. As with IgG, the lowerlevels of IgM in the experimental group at d 56 may be pro-duced by a POP-induced lower ability to produce IgM.

The total amount of immunoglobulin is produced by recentexposure to antigens from immunization and natural micro-pathogen exposure. None of the antibody titers to immunizedantigens were correlated with the total amount of IgG or IgM.This shows that the IgG and IgM levels were not influenced byone of the immunizations alone. The total IgG and IgM maytherefore act as markers for total response of natural andimmunized antigen exposures.

Lymphocyte ProliferationThe responses of both blood and spleen lymphocytes to

PHA were strong, whereas the response to PWM was poor.The response to Con A was weaker than to PHA, especially forthe spleen lymphocytes (Table 5). The significantly strongerblood lymphocyte response to PHA in the experimental groupindicates a stimulation of the immune system following ashort-term exposure to POP in glaucous gull chicks. In earlylife the immune system matures due to exposure to environ-mental microorganisms. The findings that a larger populationof cells in the cultures reacted to PHA could indicate that thelymphocytes in this early life exposure to POP were stimulatedmore strongly by microbes due to a POP impairment of theinnate immune system that comprises the first line of defense.However, the exposure of growing glaucous gull chicks toPOP did not impair the ability of lymphocytes to respond tomitogens, as was also demonstrated in chickens and free-rang-ing adult polar bears (Lavoie & Grasman, 2007; Lie et al.,2005). The lymphocyte proliferation is an important processfor clonal expansion in response to antigens (Janeway, 2005),and suggests that the experimental group in the present studymay have suffered from increased exposure to environmentalmicrobes. Although the unclear results from the lymphocytesproliferation tests are in agreement with Lavoie and Grasman(2007), who found that peripheral blood lymphocytes fromchickens exposed to PCB-126 produced a nonsignificantincreased response to PHA in one experiment, a significantdecrease was found in another experiment (Lavoie & Grasman,

Downloaded By: [University Library of Tromsoe] At: 23:24 30 July 2009

882 K. SAGERUP ET AL.

2007). In other studies in vitro T-dependent mitogen responseswere both enhanced (Segre et al., 2002) and suppressed (Rosset al., 1997) when mice or rats, respectively, were exposed toPOP. This suggests that the mitogen-induced lymphocyte pro-liferation may also be altered by severe POP exposure.

To test the effect of in vitro exposure to PCB on the generalresponse level of lymphocytes, different PCB congeners wereadded to Con A-stimulated lymphocyte cultures. All culturesgave a higher response in the experimental group, but becauseof great variation, these differences were only statistically sig-nificant for PCB-99 and PCB-153 in the spleen lymphocytecultures. The high intercorrelation between the responses fromCon A- and PCB-stimulated lymphocytes shows that the cellcultures reacted in the same way to different PCB. This meansthat the dioxin-like non-ortho Cl-substituted PCB-126 andmono-ortho Cl-substituted PCB-156 (Van den Berg et al.,2006) reacted similarly in the lymphocyte cultures, as did thenon-dioxin-like di-ortho Cl-substituted PCB-99 and PCB-153.This may indicate that the PCB influence on lymphocytes actsnot only through the aryl hydrocarbon receptor (AhR). It istherefore reasonable to suggest that immune-toxicity evalua-tion also needs to include non-dioxin-like PCB and that thetoxic equivalent factor (TEF) approach may underestimate theimmune toxicity from PCB exposure in wildlife (Lyche et al.,2004, 2006).

The demonstration that in vitro exposure to PCB did notsuppress the lymphocyte response to Con A in culture and thefact that chicks in the experimental group actually hadincreased lymphocyte proliferation following the addition ofPCB are difficult to explain. One should be aware, however,that only a small population of lymphocytes responded to ConA. Retrospectively, knowing the weak stimulation induced byCon A compared with PHA in glaucous gull chicks, PHA stim-ulation should have been used.

In conclusion, the present laboratory experiment showedthat the immune system of free-ranging glaucous gull chicksmight be altered and that the infection resistance may bereduced through POP exposures. Which part of the immunesystem is most affected by the POP exposure needs to be fur-ther investigated. This study demonstrated that the use of gullchicks in captivity in combination with different immunizationmodels might provide information concerning the combinedadverse effects of POP on specific immunity and on diseaseresistance.

REFERENCESAndersen, M., Lie, E., Derocher, A. E., Belikov, S. E., Bernhoft, A., Boltunov,

A. N., Garner, G. W., Skaare, J. U., and Wiig, Ø. 2001. Geographic varia-tion of PCB congeners in polar bears (Ursus maritimus) from Svalbard eastto the Chukchi Sea. Polar Biol. 24:231–238.

Barrett, R. T., Skaare, J. U., and Gabrielsen, G. W. 1996. Recent changes inlevels of persistent organochlorines and mercury in eggs of seabirds fromthe Barents Sea. Environ. Pollut. 92:13–18.

Barry, S. J., and Barry, T. W. 1990. Food habits of glaucous gulls in theBeaufort Sea. Arctic 43:43–49.

Bernhoft, A., Wiig, Ø., and Skaare, J. U. 1997. Organochlorines in polar bears(Ursus maritimus) at Svalbard. Environ. Pollut. 95:159–175.

Bernhoft, A., Skaare, J. U., Wiig, Ø., Derocher, A. E., and Larsen, H. J. S.2000. Possible immunotoxic effects of organochlorines in polar bear atSvalbard. J. Toxicol. Environ. Health A 59:561–574.

Bogan, J. A., and Bourne, W. R. P. 1972. Organochlorine levels in Atlanticseabirds. Nature 240:358.

Borgå, K. 1997. Bioaccumulation of organochlorines (OCs) in an Arcticmarine food chain. MSc thesis, University of Tromsø, Tromsø.

Borgå, K., Gabrielsen, G. W., and Skaare, J. U. 2001. Biomagnification oforganochlorines along a Barents Sea food chain. Environ. Pollut.113:187–198.

Borgå, K., Wolkers, H., Skaare, J. U., Hop, H., Muir, D. C. G., and Gabrielsen,G. W. 2005. Bioaccumulation of PCBs in Arctic seabirds: Influence ofdietary exposure and congener biotransformation. Environ. Pollut.134:397–409.

Borgå, K., Campbell, L., Gabrielsen, G. W., Norstrom, R. J., Muir, D. C. G.,and Fisk, A. T. 2006. Regional and species specific bioaccumulation ofmajor and trace elements in Arctic seabirds. Environ. Toxicol. Chem.25:2927–2936.

Brevik, E. M. 1978. Gas chromatographic method for the determination oforganochlorine pesticides in human milk. Bull. Environ. Contam. Toxicol.19:281–286.

Bustnes, J. O. 2006. Pinpointing potential causative agents in mixtures of per-sistent organic pollutants in observational field studies: A review of glau-cous gull studies. J. Toxicol. Environ. Health A 69:97–108.

Bustnes, J. O., Erikstad, K. E., Bakken, V., Mehlum, F., and Skaare, J. U.2000. Feeding ecology and the concentration of organochlorines in glau-cous gulls. Ecotoxicology 9:175–182.

Bustnes, J. O., Bakken, V., Erikstad, K. E., Mehlum, F., and Skaare, J. U.2001. Patterns of incubation and nest-site attentiveness in relation toorganochlorine (PCB) contamination in glaucous gulls. J. Appl. Ecol.38:791–801.

Bustnes, J. O., Folstad, I., Erikstad, K. E., Fjeld, M., Miland, Ø. O., andSkaare, J. U. 2002. Blood concentration of organochlorine pollutants andwing feather asymmetry in glaucous gulls. Funct. Ecol. 16:617–622.

Bustnes, J. O., Erikstad, K. E., Skaare, J. U., Bakken, V., and Mehlum, F.2003. Ecological effects of organochlorine pollutants in the Arctic: A studyof the glaucous gull. Ecol. Appl. 13:504–515.

Bustnes, J. O., Hanssen, S. A., Folstad, I., Erikstad, K. E., Hasselquist, D., andSkaare, J. U. 2004. Immune function and organochlorine pollutants in arc-tic breeding glaucous gulls. Arch. Environ. Contam. Toxicol. 47:530–541.

Bustnes, J. O., Erikstad, K. E., Hanssen, S. A., Tveraa, T., Folstad, I., andSkaare, J. U. 2006. Anti-parasite treatment removes negative effects ofenvironmental pollutants on reproduction in an Arctic seabird. Proc. R.Soc. Lond. B 273:3117–3122.

Catry, P., Ratcliffe, N., and Furness, R. W. 1998. The influence of hatchingdate on different life-history stages of Great Skuas Catharacta skua. J.Avian Biol. 29:299–304.

Catty, D. 1998. Antibodies: A practical approach. Oxford: IRL Press.deSwart, R. L., Ross, P. S., Vos, J. G., and Osterhaus, A. D. M. E. 1996.

Impaired immunity in harbour seals (Phoca vitulina) exposed to bioaccu-mulated environmental contaminants: Review of a long-term feedingstudy. Environ. Health Perspect. 104:823–828.

Erikstad, K. E. 1990. Winter diets of four seabird species in the Barents Seaafter a crash in the capelin stock. Polar Biol. 10:619–627.

Fairbrother, A., Smits, J., and Grasman, K. A. 2004. Avian immunotoxicology.J. Toxicol. Environ. Health B 7:105–137.

Grasman, K. A., Fox, G. A., Scanlon, P. F., and Ludwig, J. P. 1996. Orga-nochlorine-associated immunosuppression in prefledgling Caspian ternsand herring gulls from the Great Lakes: An ecoepidemiological study.Environ. Health Perspect. 104:829–842.

Grasman, K. A., and Fox, G. A. 2001. Associations between altered immunefunction and organochlorine contamination in young Caspian terns (Sternacaspia) from Lake Huron, 1997–1999. Ecotoxicology 10:101–114.

Heilmann, C., Grandjean, P., Weihe, P., Nielsen, F., and Budtz-Jorgensen, E.2006. Reduced antibody responses to vaccinations in children exposed topolychlorinated biphenyls. PLoS Med. 3:1352–1359.

Downloaded By: [University Library of Tromsoe] At: 23:24 30 July 2009

TOXIC EFFECT OF POP ON THE DEVELOPING IMMUNE SYSTEM 883

Helgason, L. B., Barrett, R., Lie, E., Polder, A., Skaare, J. U., and Gabrielsen,G. W. 2008a. Levels and temporal trends (1983–2003) of persistentorganic pollutants (POPs) and mercury (Hg) in seabird eggs from NorthernNorway. Environ. Pollut. 155:190–198.

Helgason, L. B., Polder, A., Føreid, S., Bæk, K., Lie, E., Gabrielsen, G. W., Bar-rett, R. T., and Skaare, J. U. 2008b. Levels and temporal trends (1983–2003)of polybrominated diphenyl ethers (PBDEs) and hexabromocyclododecane(HBCD) in seabird eggs from Northern Norway. Environ. Toxicol. Chem.2008 Dec 2022:2001. [Epub ahead of print]

Henriksen, E. O. 1999. Persistent organochlorine pollutants in two Arctic gullspecies: Levels and possible biochemical effects. PhD thesis, University ofTromsø, Tromsø.

Henriksen, E. O., Gabrielsen, G. W., and Skaare, J. U. 1998. Validation of theuse of blood samples to assess tissue concentrations of organochlorines inglaucous gulls, Larus hyperboreus. Chemosphere 37:2627–2643.

Henriksen, E. O., Gabrielsen, G. W., Trudeau, S., Wolkers, H., Sagerup, K.,and Skaare, J. U. 2000. Organochlorines and possible biochemical effectsin glaucous gulls (Larus hyperboreus) from Bjørnøya, the Barents Sea.Arch. Environ. Contam. Toxicol. 38:234–243.

Hertz-Picciotto, I., Park, H. Y., Dostal, M., Kocan, A., Trnovec, T., and Sram,R. 2008. Prenatal exposures to persistent and non-persistent organic com-pounds and effects on immune system development. Basic Clin. Pharma-col. Toxicol. 102:146–154.

Jakob, E. M., Marshall, S. D., and Uetz, G. W. 1996. Estimating fitness: Acomparison of body condition indices. Oikos 77:61–67.

Janeway, C. A. 2005. Immunobiology: the immune system in health and dis-ease. New York: Garland Science.

Jolicoeur, P., and Mosimann, J. E. 1960. Size and shape variation in thepainted turtle. A principal component analysis. Growth 24:339–354.

Lavoie, E. T., and Grasman, K. A. 2007. Effects of in ovo exposure to PCBs126 and 77 on mortality, deformities and post-hatch immune function inchickens. J. Toxicol. Environ. Health A 70:547–558.

Lie, E., Larsen, H. J. S., Larsen, S., Johnsen, G. M., Derocher, A. E., Lunn, N.J., Norstrom, R. J., Wiig, Ø., and Skaare, J. U. 2004. Does high orga-nochlorine (OC) exposure impair the resistance to infection in polar bears(Ursus maritimus)? Part 1: Effect of OCs on the humoral immunity. J. Tox-icol. Environ. Health A 67:555–582.

Lie, E., Larsen, H. J. S., Larsen, S., Johansen, G. M., Derocher, A. E., Lunn, N.J., Norstrom, R. J., Wiig, Ø., and Skaare, J. U. 2005. Does high orga-nochlorine (OC) exposure impair the resistance to infection in polar bears(Ursus maritimus)? Part II: Possible effect of OCs on mitogen- and anti-gen-induced lymphocyte proliferation. J. Toxicol. Environ. Health A68:457–484.

Lyche, J. L., Larsen, H. J. S., Skaare, J. U., Tverdal, A., Dahl, E., Johansen, G.,and Ropstad, E. 2004. Effects of perinatal exposure to low doses of PCB153 and PCB 126 on lymphocyte proliferation and hematology in goatkids. J. Toxicol. Environ. Health A 67:889–905.

Lyche, J. L., Larsen, H. J. S., Skaare, J. U., Tverdal, A., Johansen, G. M., andRopstad, E. 2006. Perinatal exposure to low doses of PCB 153 and PCB126 affects maternal and neonatal immunity in goat kids. J. Toxicol. Envi-ron. Health A 69:139–158.

Løvenskiold, H. L. 1964. Avifauna Svalbardensis: With a discussion on thegeographical distribution of the birds in Spitsbergen and adjacent islands.Oslo: Norwegian Polar Institute.

Mancini, G., Vaerman, J. P., Garbonara, A., and Heremans, J. F. 1964. A sin-gle-radial difusion method for the immunological quantification of pro-teins. In Protides of the biological fluids, Proceedings of the 11thColloquium, Bruges, 1963, ed. H. Peeters, pp. 370–373. Amsterdam:Elsevier.

Mehlum, F., and Daelemans, F. F. 1995. PCBs in Arctic seabirds from theSvalbard region. Sci. Total Environ. 160/161:441–446.

Muir, D. C. G., Norstrom, R. J., and Simon, M. 1988. Organochlorine contam-inants in Arctic marine food-chains—Accumulation of specific polychlori-nated-biphenyls and chlordane-related compounds. Environ. Sci. Technol.22:1071–1079.

Murvoll, K. M., Jenssen, B. M., and Skaare, J. U. 2005. Effects of pentabromi-nated diphenyl ether (PBDE-99) on vitamin status in domestic duck (Anasplatyrhynchos) hatchlings. J. Toxicol. Environ. Health A 68:515–533.

NIFES. 2008. Seafood data. National Institute of Nutrition and SeafoodResearch (NIFES). http://www.nifes.no [accessed 2008].

Norstrom, R. J., and Muir, D. C. G. 1994. Chlorinated hydrocarbon contami-nants in arctic marine mammals. Sci. Total Environ. 154:107–128.

Norstrom, R. J., Belikov, S. E., Born, E. W., Garner, G. W., Malone, B., Olpin-ski, S., Ramsay, M. A., Schliebe, S., Stirling, I., Stishov, M. S., Taylor, M.K., and Wiig, O. 1998. Chlorinated hydrocarbon contaminants in polarbears from eastern Russia, North America, Greenland, and Svalbard:Biomonitoring of Arctic pollution. Arch. Environ. Contam. Toxicol.35:354–367.

Perrins, C. M. 1970. Timing of birds breeding seasons. Ibis 112:242–255.R Development Core Team. 2008. R: A language and environment for statisti-

cal computing. Vienna: R Foundation for Statistical Computing.Ross, P. S., de Swart, R. L., van der Vliet, H., Willemsen, L., de Klerk, A., van

Amerongen, G., Groen, J., Brouwer, A., Schipholt, I., Morse, D. C., vanLovere, H., Osterhaus, A. D. M. E., and Vos, J. G. 1997. Impaired cellularimmune response in rats exposed perinatally to Baltic Sea herring oil or2,3,7,8-TCDD. Arch. Toxicol. 71:563–574.

Rothman, K. J. 1990. No adjustments are needed for multiple comparisons.Epidemiology 1:43–46.

Sagerup, K., Henriksen, E. O., Skorping, A., Skaare, J. U., and Gabrielsen, G.W. 2000. Intensity of parasitic nematodes is associated with organochlo-rine levels in the glaucous gull (Larus hyperboreus). J. Appl. Ecol. 37:1–9.

Segre, M., Arena, S. M., Greeley, E. H., Melancon, M. J., Graham, D. A.,and French, J. B. 2002. Immunological and physiological effects ofchronic exposure of Peromyscus leucopus to Aroclor 1254 at a concen-tration similar to that found at contaminated sites. Toxicology174:163–172.

Tryphonas, H. 1994. Immunotoxicity of polychlorinated biphenyls: Presentstatus and future considerations. Exp. Clin. Immunogenet. 11:149–162.

Tryphonas, H. 2005. Investigative immunotoxicology. Boca Raton, FL: Taylor& Francis.

Van den Berg,M., Birnbaum, L. S., Denison, M., De Vito, M., Farland, W.,Feeley, M., Fiedler, H., Hakansson, H., Hanberg, A., Haws, L., Rose,M., Safe, S., Schrenk, D., Tohyama, C., Tritscher, A., Tuomisto, J.,Tysklind, M., Walker, N., and Peterson, R. E. 2006. The 2005 WorldHealth Organization reevaluation of human and mammalian toxicequivalency factors for dioxins and dioxin-like compounds. Toxicol.Sci. 93:223–241.

Verreault, J., Skaare, J. U., Jenssen, B. M., and Gabrielsen, G. W. 2004.Effects of organochlorine contaminants on thyroid hormone levels in arcticbreeding glaucous gulls, Larus hyperboreus. Environ. Health Perspect.112:532–537.

Vos, J. G., and Luster, M. 1989. Immune alteration. In Halogenated biphenyls,terphenyls, naphthalenes, dibenzodioxins, and related products, eds. R.Kimbrough and A. Jensen, pp. 295–322. Amsterdam: Elsevier.

Walker, C. 1990. Persistent pollutants in fish-eating sea birds—Bioaccumula-tion, metabolism and effects. Aquat. Toxicol. 17:293–324.

Walker, C. H., Hopkin, S. P., Sibly, R. M., and Peakall, D. B. 2006. Principlesof ecotoxicology, 3rd ed. Boca Raton, FL: Taylor & Francis.

Downloaded By: [University Library of Tromsoe] At: 23:24 30 July 2009