Toxicology Letters 112–113 (2000) 289–301

Toxicology of environmental chemicals in the flounder(Platichthys flesus) with emphasis on the immune system:

field, semi-field (mesocosm) and laboratory studies

G.C.M. Grinwis a,*, A.D. Vethaak b, P.W. Wester c, J.G. Vos a,c

a Department of Pathology, Faculty of Veterinary Medicine, Utrecht Uni6ersity, P.O. Box 80158,NL-3508 TD Utrecht, The Netherlands

b National Institute for Coastal and Marine Management, Middelburg, The Netherlandsc National Institute of Public Health and the En6ironment, Biltho6en, The Netherlands

Abstract

European flounder (Platichthys flesus) has shown an increased prevalence of liver tumors and lymphocystis disease(a viral infection) that correlated with pollution in field research in Dutch coastal and estuarine waters. Semi-field ormesocosm experiments confirmed the supposed causality. Although these types of research are highly relevant for theferal population, laboratory experiments are necessary to establish causal relationships between specific chemicalpollutants and disease. Therefore, the effects on flounder of some of the potentially causative chemicals such asbenzo[a]pyrene (BaP), dimethyl-benz[a]anthracene (DMBA), 2,3,7,8 tetrachlorodibenzo-p-dioxin (TCDD), 3,3’,4,4’,5pentachlorobiphenyl (PCB-126), and bis(tri-n-butyltin)oxide (TBTO) were examined in several laboratory experi-ments. These effects were evaluated using general toxicological parameters and histopathology. For immune functionassessment, attempts to develop an infection model with the lymphocystis virus were made, but appeared unsuccessfuland immune function tests are not fully operational at the moment. Flounder has been successfully maintained andexposed to toxic substances in captivity in our laboratory. Short-term aqueous exposure to high levels of BaP orDMBA did not induce marked effects under our experimental conditions. Results of oral exposure of flounder to lowlevels of TCDD, PCB-126 or harbor sludge extract show significant induction of cytochrome P4501A (CYP1A) inhepatocytes. Oral exposure to high levels of TCDD or PCB-126 also significantly induced CYP1A immunoreactivityin epithelium in mesonephros and digestive tract and in endothelium in several organs. Remarkable was the inductionof CYP1A in a distinct population of mononuclear cells in the mesonephros. Moreover, oral exposure to TCDDresulted in an increased mitotic activity and an increase of the hepatosomatic index in the 20 and 500 mg TCDD/kggroup respectively. Therefore, exposure to TCDD and related substances may promote the development of livertumors in the field. Exposure to PCB-126 also significantly reduced the relative thymus volume, but other resultsindicate that flounder is relatively insensitive to this type of chemicals. Short-term aqueous exposure of flounder toTBTO, in concentrations that were in the same order of magnitude as upper TBT levels measured in the field, causedmortality after 7–12 days associated with gill lesions, and induced reduction of the non-specific resistance anddecrease of the relative thymus volume. From these results we therefore conclude that TBTO might play a causal role

www.elsevier.com/locate/toxlet

* Corresponding author. Tel.: +31-30-253-4296; fax: +31-30-251-6853.E-mail address: g.grinwis@pobox.uu.nl (G.C.M. Grinwis)

0378-4274/00/$ - see front matter © 2000 Elsevier Science Ireland Ltd. All rights reserved.

PII: S 0378 -4274 (99 )00239 -8

G.C.M. Grinwis et al. / Toxicology Letters 112–113 (2000) 289–301290

in, for instance, increased prevalence of lymphocystis virus infections in the field © 2000 Elsevier Science Ireland Ltd.All rights reserved.

Keywords: Flounder; Immunotoxicology; Immunohistochemistry; TBTO; TCDD; PCB; Cytochrome P450 1A

1. Introduction

Diseases and abnormalities of marine and estu-arine fish populations have received much atten-tion recently, partly because of concern over thepossible contribution of pollution-mediated dis-eases to fish mortality and observed populationdeclines, but also because fish diseases may besuitable indicators for monitoring anthropogenicenvironmental stress (Vethaak and ap Rheinallt,1992). Studies of fish pathology in relation topollution may attempt to answer the followingquestions:1. in general terms, are changes in disease pat-

terns associated with changes in environmentalquality? (field experiments),

2. can we identify cause-and-effect relationshipsbetween disease patterns and exposure to thecomplex mixture of chemical pollutantspresent in the environment? (semi-fieldexperiments),

3. can we identify cause-and-effect relationshipsfor specific chemical contaminants? (labora-tory experiments).

The answers to these questions are highly relevantto the potential use of disease as a biomarker ofcontaminant effects. The present review will focuson field, semi-field and laboratory studies in theeuryhaline fish species European flounder(Platichthys flesus) carried out in TheNetherlands.

2. Field experiments

The primary driving force for the studies wasprovided by reports of epidemics of grossly visibledisease signs in fish. This led to the initiation ofsystematic and routine epidemiological studies inDutch coastal fish populations in the early 1980s(Van Banning, 1987; Vethaak, 1987; Vethaak and

Jol, 1996; Vethaak and Wester, 1996). These fieldsurveys have shown relatively high prevalences ofthree grossly visible diseases in Europeanflounder: skin ulcers, lymphocystis disease andliver tumors, and in these studies attempts weremade to correlate the presence of fish disease withchemical pollution and other anthropogenic stres-sors (Vethaak and Jol, 1996). European flounder,a coastal and estuarine fish which is widespread inEurope, was chosen as indicator species primarilybecause of their high disease prevalence (Bucke etal., 1996). As a result of these studies, recently anumber of pollution-associated diseases/abnor-malities were incorporated within (inter)nationalfish disease monitoring programs. In 1991, thesestudies became part of the ongoing Joint Assess-ment and Monitoring Program (JAMP).

Various authors have attributed the develop-ment of skin ulceration in fish to a variety ofagents such as viruses, bacteria of a range ofgenera, and parasitic infections. In the case offlounder, bacteriological examination has resultedin the isolation of bacteria of the genera Vibrio,Pseudomonas, and Aeromonas from the lesionsand other tissues of affected fish (reviewed byWiklund, 1994). Vethaak (1993) carried out bacte-rial analyses at two sites adjacent to drainagesluices in the Dutch Wadden Sea, one with amuch higher prevalence of skin ulcers than theother. Results showed no clear association be-tween bacterial levels in the environment andbacterial infections of the fish, indicating thatother factors might be of influence. Nevertheless,experiments have shown that bacteria culturedfrom diseased fish can infect mechanicallyabraded skin and initiate the development of skinulcers (Wiklund, 1994).

Lymphocystis disease is known to be of viralorigin. Although its pathology (Wolf, 1962, 1988)and epidemiology are well documented, relatively

G.C.M. Grinwis et al. / Toxicology Letters 112–113 (2000) 289–301 291

little is known about the immunological aspects ofthe disease (Russel, 1974). Lymphocystis diseasewas found in an average prevalence of 14.3% inthe field (Vethaak and Jol, 1996). Some evidencefor a role of contaminants in disease causationwas found for this disease. However, the hypothe-sis that contaminants may contribute to the devel-opment of lymphocystis disease throughimmunosuppression needs to be addressed.

Liver tumors were reported in flounder in theearly 1980s (Vethaak, 1987). Liver tumors werepresent in 1.0% of the fish sampled during thestudy, prevalences rising steeply with age andlocally attaining values of up to 30% in the 6-yeargroup and older fish (Vethaak and Jol, 1996). Themajority of these tumors were diagnosed as hepa-tocellular adenomas after histopathological exam-ination, 13.1% were diagnosed as hepatocellularcarcinomas (Vethaak and Wester, 1996). Routinehistopathological examination of livers showingno gross pathology resulted in the identification ofa range of lesions including hepatocellular ade-noma, foci of cellular alteration, inflammatorylesions, focal necrosis, regenerative foci, hydropicvacuolated hepatocytes and bile duct epithelium,and fibrillar hepatocytes (Vethaak and Wester,1996).

From these retrospective field studies a possiblecontribution of pollutants to multifactorial dis-ease causation could not be adequately assessed

due to interfering factors such as fish migration,salinity, and fishing impact. Disease prevalencesshowed a strong correlation with fishing activity(possibly indicating an effect of damage by fishinggear) and appeared to be positively related tosalinity, but there appeared to be little correlationwith concentrations of contaminants in sedimentsor tissues. However, when only strictly marinesites were considered, a relation with pollutioncould not be ruled out (Vethaak and Jol, 1996).

In an attempt to overcome the shortcomings ofepidemiological field studies, and to investigatethe role of environmental contamination, a large-scale mesocosm study of disease in flounder wasinitiated in 1988 on the island of Texel in theNetherlands.

3. Semi-field (mesocosm) experiments

This three-year (1990–1993) prospective semi-field study involved the exposure of single cohortsof European flounder, obtained from a pristinesite, to contaminated dredge spoil (representativefor the general pollution load of the river Rhine)in three large-scale mesocosms (Vethaak et al.1996). The advantage of using mesocosms is thatthey act as an intermediate between controlledlaboratory studies and field studies (Fig. 1). Themesocosm approach allows us to study the long-

Fig. 1. Integrated approach to study the causes of wildlife diseases (According to Vethaak, 1993).

G.C.M. Grinwis et al. / Toxicology Letters 112–113 (2000) 289–301292

term development of diseases within a single co-hort of fish under ‘natural’ conditions with re-duced or eliminated interference from otherfactors including migratory behavior of the fish,fishing activity and salinity changes (Vethaak etal. 1996). Two of the mesocosms used in theexperiment contained clean sand whilst the third,sharing a common water circulation with one ofthe clean sand mesocosms, was stocked with con-taminated dredged spoil from Rotterdam Harbor.In this way, one of the clean sand mesocosms wasindirectly polluted via the water phase, and analy-sis of contaminant concentrations showed it tohave a status intermediate between the other two,with polychlorinated biphenyls (PCBs) and se-lected polycyclic aromatic hydrocarbons (PAHs)showing a clear concentration gradient across thethree mesocosms (Vethaak et al., 1996). Analysisof the PAH metabolite 1-hydroxy (1-OH) pyrenein the bile of the flounder, using synchronousfluorescence spectrometry, showed a similar gradi-ent (Ariese et al., 1993). Random samples offlounder from the indirectly polluted and refer-ence mesocosms were examined every 2 monthsfor epidermal diseases (lymphocystis, skin ulcersand fin rot) and then released back into themesocosms. In addition, some fish from all threemesocosms were sacrificed every 6 months forhistological and other investigations.

The prevalences of skin-related diseases(lymphocystis disease and skin ulcers) in the refer-ence and indirectly polluted mesocosms showed asimilar pattern of temporal variation, and wererather high compared with values observed in thefield. Little difference in the prevalence of skinulcers was found between the reference and theindirectly polluted mesocosm. In contrast preva-lences of lymphocystis disease were consistentlyhigher in the indirectly polluted mesocosm duringthe latter half of the 3-year period. In flounderfrom the directly polluted mesocosm, a significantreduction of vitamin A levels was found accompa-nied by high PCB tissue-levels. Strongest evidencefor induction by contaminants was found for livertumors and preneoplastic foci. Evidence for achemical carcinogenesis in the mesocosm experi-ment was further substantiated by a correlationbetween biliary hydroxy-metabolites of carcino-

genic PAHs, such as benzo[a]pyrene and the oc-currence of PAH-DNA adducts (Vethaak et al.,1996).

4. Laboratory experiments

From both field and mesocosm experiments thefollowing hypothesis was formulated: xenobioticspresent in polluted spoil can induce(pre)neoplastic liver lesions, and can cause im-munotoxicity making European flounder moresusceptible for infectious diseases. In order toprove causal links between specific pollutants andfish disease laboratory experiments were incorpo-rated in the integrated study. Therefore Europeanflounder had to be kept and raised in captivity inorder to perform experiments under laboratoryconditions.

4.1. Collection of fish

Several authors have obtained fertilized eggs offlounder by artificial fertilization of eggs strippedfrom mature specimen caught on spawninggrounds (Solemdal, 1967; Von Westernhagen,1970), but to our knowledge no attempts havebeen made to reproduce and rear flounder on alarge scale in captivity (Grinwis et al., 1995). Forlaboratory research we therefore relied completelyon collection of fingerlings from feral populations.

Animals used in several laboratory experimentswere caught with a dip-net in a relatively cleanestuary near Southampton (England) during ashort period in spring. The flounder obtained inEngland belonged to the 0+ group and measured:1 cm upon arrival at the laboratory.

4.2. Water quality and holding facilities

European flounder is a euryhaline fish specieswhich normally spends a large part of its life-cyclein fresh or brackish water. Flounder can thereforebe kept in fresh as well as sea water. A bottom-substrate is essential in a tank with flounders. Fishkept in tanks without a bottom-substrate showedcontinuous bacterial and parasitic problems thatcould interfere with the experiments and were

G.C.M. Grinwis et al. / Toxicology Letters 112–113 (2000) 289–301 293

difficult to treat. Moving fish to a sand-substrateremoved these disease problems almostimmediately.

At the Dutch National Institute of PublicHealth and the Environment (RIVM), where mostlaboratory experiments were performed, the ani-mals were kept in glass aquariums filled with freshwater (Dutch Standard Water (DSW) at 1992°Cunder a 16 h light and 8 h dark regimen. Thewater was circulated and aerated continuously,and passed over a biological filter. Temperaturewas checked daily, pH, oxygen and NO2

− contentsof the water were checked weekly. The populationdensity depended on the size of the animals, andranged from maximum of 30 animals per 25 l foranimals between 2 and 5 cm, to 10 animals per100 l for the larger category. The bottoms of theaquaria were covered with a 2 cm layer of silversand in order to decrease the above mentionedproblems. During a quarantine period of severalweeks the animals were monitored in particularfor signs of illness. Clinically sick animals weretreated once in a water basin containing 324 mlformalin (35% formaldehyde solution) and 0.5 mgmalachite-green per liter DSW for 30 min, againstbacterial-, mycotic- and parasitic infections. Theanimals were checked daily for gross abnormali-ties, condition and behavior. Following this pro-cedure flounder was kept and raised in thelaboratory without major problems.

4.3. Feeding and growth

Flounder can be fed with a variety of naturalfood items. Artificial compound feed is usuallynot attractive for flounder although flounder hassuccessfully been adapted to commercial sheath-fish feed. During the laboratory experiments theanimals were fed Artemia salina (Rasbora,Veenendaal, The Netherlands).

5. Immune system

It is known from mammalian studies that theimmune system is sensitive for the effects of pollu-tion. Immune parameters could therefore be valu-able to evaluate the health condition of fish and

to monitor the possible effects of pollutants onthe organism. To address the putative im-munotoxic effects of xenobiotics, attempts weremade to adapt ex vivo/in vitro immune functiontest for the use in flounder, and experiments wereconducted to develop a viral infection model.

The fish immune system shows, to a certainextent, a resemblance with the defense system ofmammals. Nonspecific immune reactions are usu-ally the first defense mechanism and can be pro-voked by invasion by foreign organisms and/ortrauma. Phagocytic cells (granulocytes andmononuclear phagocytes) and several mediatorsplay an important role (Zelikoff, 1994). In fish,non-specific cytotoxic cells (NCCs), the counter-part of mammalian natural killer cells, have beenidentified. These cells belong to a leukocyte sub-population of large granular lymphocytes andplay an important role in immunosurveillance andare able to lyse tumor cells and cells infected byviruses. The NCCs are found mainly in the headkidney, spleen and peripheral blood (Evans et al.,1989; Faisal et al., 1989; Greenly et al., 1991).

The specific immune defense mechanisms of fishinclude cellular- and humoral responses and arecomparable to the mammalian mechanisms (VanMuiswinkel et al., 1991; Zelikoff, 1994). In fishesantibodies are produced by B-lymphocytes andcan be demonstrated in the serum, bile, and mu-cus of the skin. The antibodies are of a singleclass, comparable with IgM of mammals, but arestructurally tetrameric instead of pentameric as inmammals (Rombout et al., 1990). At present,there are no antibodies available to distinguishlymphocyte subpopulations in European flounder.Thymus, spleen and kidney (the latter containingthe bone marrow homologue in fish) are the ma-jor organs in the fish immune system, while bonemarrow and lymph nodes are absent. Like inmammals the bursa of Fabricius, the centrallymphoid organ for B-cells in birds, is lacking.

5.1. Thymus

The thymus can be found in all vertebrates,except agnatha, and shows no essential histologicvariations between vertebrates. In Europeanflounder, the thymus is a paired structure local-

G.C.M. Grinwis et al. / Toxicology Letters 112–113 (2000) 289–301294

ized in close association with the gill cavity (Grin-wis et al., 1995). Due to its irregular shape andsmall size, weighing the thymus, a procedure oftenused in mammals to detect thymotoxicity is prac-tically not feasible. Also, no clear distinction be-tween cortex and medulla, as seen in mammals, ispresent in the thymus of European flounder.Therefore, we used morphometric analysis of se-rial sections of the thymus in our experimentsfrom which the thymus volume was calculated(Grinwis et al., 1998).

5.2. Spleen

The spleen is a rounded organ with a dark redto black color. It is situated in the mesentery nextto the stomach, guts, liver, and pancreas (Grinwiset al., 1995). The histology of the spleen is com-parable with that of mammals: red pulp, whitepulp, vessels and ellipsoids can be distinguished(Ellis et al., 1989; Zapata and Cooper, 1990). Thered pulp constitutes the major part of the spleenwith a cellular reticulum, hemopoietic tissue(which usually shows little activity) and sinusoids.The white pulp is typically not prominent in ourexperimental animals. Ellipsoids are the terminalcapillaries surrounded by a cuff of fibrous tissuewith mainly macrophages. In the red pulp largergroups of pigment containing macrophages, themelano-macrophage centers (MMCs) can befound.

5.3. Kidney

The kidney of teleosts is a multifunctional or-gan in which four functional aspects can be distin-guished. The kidney has an excretory- and ahematopoietic function, plays a role in the im-mune system and produces hormones (Ellis et al.,1989; Zapata and Cooper, 1990). The kidney canbe divided into two parts, the head kidney and thetrunk kidney, and is situated retroperitoneallyventral to the vertebral column. The main portionof the head kidney consists of lymphoid-, hemo-poietic- and adrenal (suprarenal) tissue,lymphoid- and hemopoietic tissue is also presentin the trunk kidney. MMCs can be encountered inboth parts of the kidney.

5.4. Blood

Teleosts have a small blood volume in relationto the body weight. Usually May–GrunwaldGiemsa stained blood smears are used for cyto-logic evaluation of the blood. The main cellularcomponent of the blood is the erythrocyte. Like inamphibians, reptiles and birds the red blood cellsof fish contain nuclei. The cells and nuclei aremore or less oval shaped. Several leucocytes canbe found in the blood of fish: lymphocytes, granu-locytes monocytes, and thrombocytes. An impor-tant problem in using blood smears forlymphocyte quantification in fish is the difficultyof distinguishing lymphocytes and thrombocytesas they can have a similar morphology. Thismight be the main reason for the widely differingdata on numbers of circulating lymphocyteswithin the same species reported in the literature(Ellis, 1977; Rowley, 1990). Data on lymphocytecounts from May–Grunwald Giemsa stainedblood smears are therefore rather unreliable. Inorder to use blood (smears) for the quantificationof lymphocytes, specific immunological reagentsare required, but those are not yet available forthe flounder.

6. Immune function tests

Experiments for the development of immunefunction test focused on the adaptation of existingprotocols for mammalian species for the assess-ment of the mitogen-induced proliferation oflymphocytes and for the assessment of the non-specific cytotoxic cell (NCC, the fish-homologueof natural killer cells) (Boonstra et al., 1996).

Like in mammals, in fish specific immunologicalreactions are mediated by B and T lymphocytes.Flounder lymphocytes were tested for their abilityto proliferate following stimulation with the B cellmitogen LPS, or the T cell mitogen PHA. Theleukocytes were cultured in the presence of LPSor PHA for 4 days; on the third day 3H-thymidinewas added. Leukocytes from the spleen,mesonephros and peripheral blood were stimu-lated with 0.03, 0.08, 0.17 mg/ml LPS (final con-centration). As a reference, cells were incubated

G.C.M. Grinwis et al. / Toxicology Letters 112–113 (2000) 289–301 295

with medium alone. The highest degree of prolif-eration was observed in peripheral bloodlymphocytes (PBL) compared to the response ofspleen and mesonephros leukocytes, with an opti-mum of 0.08 mg/ml LPS. The mitogen responsesof flounder leukocytes to the T cell mitogen PHAshowed a completely different pattern than LPS.The spleen showed the highest degree of prolifera-tion following incubation for 4 days with PHA.Its optimum concentration was 0.12 mg/ml PHA.PHA also stimulated the proliferation of themesonephros leukocytes, but only at the highestconcentration (0.30 mg/ml PHA). PBL did notshow any response when stimulated with this Tcell mitogen.

The activity of nonspecific cytotoxic (NCC)cells was measured in vitro by determining theirability to lyse a radiolabeled target cell. Theamount of radioactivity released is a measure forthe number of cells killed by the NCC cells. TheNCC activity was determined in flounder at theage of 2.5 years (weighing :50 g) at differentincubation temperatures (20 vs. 26°C). There ap-peared to be a trend that the lower incubationtemperature augmented the NCC activity as com-pared to the higher temperature. The effector cellswere leukocytes isolated from the spleen,mesonephros and pronephros of flounder. Afterco-incubation of effector and target cells for 4 h,the mesonephros leukocytes were clearly capableof lysing YAC-1. At an effector to target (E/T)ratio of 25, 19% of the YAC-1 cells were lysed bythe mesonephros leukocytes. A ratio of 100 re-sulted in killing of 37% of the target cells duringthe 4 h incubation period. The spleen andpronephros leukocytes on the other hand, showeda low NCC activity (B4%) even at an E/T ratioof 100. Similar experiments performed using P815or K562 as target cells in stead of YAC-1 showedthat these cells were less sensitive targets for deter-mining the cytotoxic activity of NCC cells ascompared to YAC-1. To ascertain whether thespecific release measured could be attributed tothe activity of NCC cells various inhibition andstimulation experiments were performed. Espe-cially the results obtained from the inhibitionstudies strongly suggest that the observed specificrelease following incubation of flounder

mesonephros leukocytes with the target cellsmight indeed be contributed to the activity ofNCC cells. However, in all immune function teststhe standard deviations were very high, whichmakes incorporation of these tests in experimentsdifficult.

7. Lymphocystis infection model

Experiments were performed to develop an in-fection model with the lymphocystis virus. Thisvirus belongs to the group of iridoviridae and cancause a chronic, self-limiting infection in fish.Several different virus strains have been recog-nized, and different strains are infectious to differ-ent fish species. The virus induces remarkablehypertrophy of infected fibroblast. Infectedfibroblasts can become as large as 1 mm in diame-ter. The typical lymphocystis lesions are oftenfound in the subcutaneous connective tissue in thefins, but lesions have also been described in inter-nal organs (Nigrelli and Ruggieri, 1965). Al-though several papers have been published onmorphological and virological aspects of the dis-ease, little is known about the involvement of theimmune system in the protection against infectionwith the virus (Russel 1974).

In our experiments, histologically verifiedlymphocystis virus-induced lesions were harvestedfrom freshly caught European flounder and pro-cessed into a cell-free suspension. Ten fish wereinjected in the subcutaneous tissue with 0.1 ml ofthe virus suspension at a standardized site justdorsal of the vertebrae on a line connecting thedorsal and ventral fin, ten control animals wereinjected with 0.1 suspension medium only at thesame site. Since no grossly visible lesions devel-oped during a 3 month period, a second experi-ment was initiated. The animals were stressed byalternating the water salinity every 3 days, andanimals were fed only once weekly instead ofdaily. Also, the virus suspension was applied by amulti-needle tattooing device in the tail and bothdorsal and ventral fin. Again, no grossly visiblelesions were noted during 3 months, rendering theused protocol not suitable as a host-resistancemodel for our purposes. Histological examination

G.C.M. Grinwis et al. / Toxicology Letters 112–113 (2000) 289–301296

of the injection-site, however, did show typicallymphocystis cells indicating that the virus sus-pension used did contain infectious viral particles.

8. Chemical exposure experiments

From analyses of sediments and fish combinedwith data on the effects in other (mammalian andnon-mammalian) species three (groups) of xenobi-otics were identified as being of specific interest:1. polycyclic aromatic hydrocarbons (PAHs; ben-

zo[a]pyrene (BaP) and dimethylbenz[a]anthracene (DMBA),

2. tetrachlorodibenzo-p-dioxin and 3,3’,4,4’,5pentachlorobiphenyl (PCB-12),

3. bis(tri-n-butyltin)oxide (TBTO).

8.1. PAHs

BaP and DMBA are formed by combustion offossil fuels for instance. They occur widely andare therefore important environmental contami-nants. They have both carcinogenic and im-munotoxic properties (Zedeck, 1980; Ward et al.,1985; White et al., 1994).

Five short term pilot exposure studies wereperformed in which European flounders were ex-posed to benzo[a]pyrene (BaP) and 7,12-dimethyl-benz[a]anthracene (DMBA). The fish wereexposed through the water-phase with 28 mg BaP/lor 190 mg DMBA/l. These levels were muchhigher than the solubility in water, but the realbioavailability of the PAHs was not determined.Also, fish were exposed by intraperitoneal injec-tion (25 mg BaP/kg body weight (bw) or 12 mgDMBA/kg bw). Effects on behavior and appear-ance were monitored, and the effects on gills,thymus, hepatopancreas, digestive tract, pro- andmesonephros, spleen, and ovary or testicles wereexamined histopathologically. Induction of theenzyme cytochrome P450 1A1 (CYP1A) as a mea-sure for the biotransformation of PAHs and theformation of biologically active metabolites wasdetermined immunohistochemically.

Exposure of flounder to the PAHs did notresult in striking effects in spite of the high con-centrations. Only a slight increase in immunoreac-

tivity for CYP1A in hepatocytes of the injectedanimals was noted. Therefore we concluded thatflounder appears to be not very sensitive to PAHs,and that PAHs may play only a minor role in theinduction of some diseases in flounder as seen inthe field, taking into account the limitations of theexposure methods used in the present experi-ments. However, possible effects of PAHs ontumor induction could not be adequately assessedin our experiments.

8.2. TCDD and PCB-126

Polychlorinated dibenzo-p-dioxins (PCDDs)dibenzofurans (PCDFs), and polychlorinatedbiphenyls (PCBs) are important environmentalpollutants. They are highly persistent, highlylipophilic, and have shown to induce several toxiceffects in mammalian and non-mammalian spe-cies, including tumor-promotion andimmunotoxicity.

European flounder was orally exposed toTCDD, PCB-126 or to harbor sludge extract un-der controlled laboratory conditions. The effectson several organs (liver, gills, gastro-intestinaltract, thyroid gland, gonads, spleen andmesonephros) were examined histopathologically.Induction and localisation of CYP1A immunore-activity, and effects on hepatocyte-proliferationwere visualized using immunohistochemistry. Ef-fects on thymus size were examined by morpho-metric analysis.

Oral exposure of flounder weekly for 8 weeks to0.0125 or 0.3125 mg/kg bw TCDD, and to 0.3125mg TEQ of a harbor sludge extract weekly for 8weeks, induced increased immunoreactivityagainst CYP1A in hepatocytes. Single administra-tion of higher doses (20, 100 and 500 mg/kg bw) ofTCDD or 0, 0.5, 5 or 50 mg/kg bw PCB-126 alsoinduced significant CYP1A immunoreactivity inthe endothelium in all organs examined (exceptfor the sinusoidal endothelium in the liver), ep-ithelium of the digestive tract, liver, andmesonephros (Table 1). Remarkably, strong im-munoreactivity was noted in a distinct cell popu-lation of the hematopoietic tissue in themesonephros, a finding not previously describedin fish. In animals exposed to 0.3125 (total dose

G.C.M. Grinwis et al. / Toxicology Letters 112–113 (2000) 289–301 297

Table 1Immunoreactivity against CYP1A in endothelial cells, digestive tract, mesonephros epithelium, and mononuclear cells in thehematopoietic tissue of the mesonephros of floundera

PCB-126 dose (mg/kg body weight)Organ and TCDD dose (mg/kg body weight)

+/− + ++ − +/− + ++−

Endotheliumb

0 0 0 0 5 0 0 00 50 0 5 0.50 0100 1 1 30 0 5 5 0 0 1 4500 0

50 0 0 0 5

Digesti6e tract epitheliumb

0 0 0 05 50 0 0 00 1 3 0.5100 11 0 4 01 2 1 51 1500 1 1 2

50 0 0 0 5

Mesonephros epitheliumb

0 0 00 05 5 0 0 00 2 2 0.51 2100 1 2 0

0500 1 2 2 5 1 0 2 250 0 0 2 3

Mononuclear cells mesonephrosc

0 0 0 0 5 0 0 00 50 0 1 0.54 3100 0 1 10 3 0500 52 2 0 0 3

50 0 0 1 4

a Five animals per group, exposed to TCDD for 4 weeks or to PCB-126 for 16 days.b −, negative, +/−, some immunoreactivity, +, marked immunoreactivity, ++, strong immuno-reactivity.c Number of cells showing immunoreactivity; −, no immunoreactive cells, +/−, some immuno-reactive cells, +, 1–10

immunoreactive cells per field (400× magnification), ++, \10 immuno-reactive cells per field (400× magnification).

over 8 weeks: 2.5 mg) and 20 mg TCDD/kg bw, asignificant increase in mitotic activity of hepato-cytes, as was illustrated by an increased immunore-activity of hepatocyte nuclei against PCNA, wasnoted. Therefore, we presume that TCDD andrelated chemicals may play a role in tumor promo-tion in the field. However, after exposure to higherdoses this effect on mitotic activity was not seen.Single exposure to 500 mg TCDD/kg bw resulted ina significant 60% increase of the hepatosomaticindex after 4 weeks. A significant reduction inrelative thymus volume of almost 50% was notedin animals exposed to 50 mg PCB-126/kg bw, anda significant increase in hepatosomatic index wasnoted in animals exposed to 500 mg TCDD /kg. Therelative thymus volume of animals exposed to thishighest TCDD dose showed only a trend in thymussize reduction.

The present experiments show that Europeanflounder, under the present experimental condi-tions, is relatively insensitive to the toxic effects ofTCDD and PCB-126. Although the functionalimplications for the immune system of the signifi-cant reduction of the relative thymus size aftersingle exposure to a high dose of PCB-126 areunclear, a significant impact on the specific resis-tance against relevant infectious diseases, e.g. theviral lymphocystis disease, under field conditions isnot likely.

8.3. TBTO

One of the chemicals of interest found in pollutedwaters and sediments is the organotin compoundtributyltin (TBT), originating mainly from anti-fouling paints used on the hulls of ships.

G.C.M. Grinwis et al. / Toxicology Letters 112–113 (2000) 289–301298

European flounder was exposed to bis(tri-n-butyltin)oxide (TBTO) in the water under con-trolled laboratory conditions. The effects onseveral organs (gills, skin, eye, liver, mesonephros,ovary/testis, spleen, and gastrointestinal tract)were examined using histopathology, and mor-phometric analysis of the thymus was performedto assess the target organ(s) for TBTO in this fishspecies. Also the function of the non specific andspecific resistance was studied using ex vivo/ invitro immune function tests.

Exposure of flounder to TBTO, in concentra-tions which were in the same order of magnitudeas maximum TBT levels measured in the field(experiment: 17.3 mg TBT/ l; field: 7.2 mg TBT/l),caused mortality after 7–12 days, resulted in gilllesions. The total lymphocyte number in thespleen was decreased, and strong and significantdecrease of the non-specific cytotoxic cell activitywas observed (Table 2). Also, a significant de-crease of the relative thymus volume, but nomarked effects on the specific immune system,was noted after exposure to TBTO (Grinwis et al.,1998).

9. Discussion and conclusions

The relatively high prevalence of the virallymphocytsis disease, skin ulcers and liver tumorsin European flounder caught in Dutch coastal andestuarine waters initiated a large-scale study inThe Netherlands. An integrated experimental ap-proach was chosen to combine experiments withhigh relevance for feral populations (field studies)with experiments using specific chemicals that can

provide causal evidence (laboratory studies). Inthe semi-field experiments, confounding factorslike salinity changes, migratory behavior andfishing activity were eliminated. The study focusedon the role of environmental pollution in diseaseinduction. Since field experiments could not provea causal relation between environmental pollu-tion, a multi-factorial etiology in the field wasanticipated and a semi-field experiment wasstarted to further investigate the role of pollution.Indirect evidence for a pollution-related diseaseinduction can also be drawn from recent datashowing a reduction of diseases in flounder com-bined with reduced pollution levels (Vethaak, per-sonal communication).

In the semi-field experiments, prevalences of(pre)neoplastic liver lesions were much higher inthe directly and indirectly (through the waterphase) polluted mesocosms compared to the refer-ence basin, demonstrating that chemical pollutionof dredged spoil was capable of inducing livertumors and associated lesions in Europeanflounder. PAHs have been proposed to be in-volved in the development of liver tumors inflatfish (Stein et al., 1987; Schiewe et al., 1991;Myers et al., 1994), in addition PCDDs, PCDFsand PCBs may also play a role in the promotionof preneoplastic liver lesions and liver tumors inEuropean flounder. Also, the semi-field experi-ments showed that pollution might be involved asa risk factor in the etiology of lymphocystis dis-ease (Vethaak et al., 1996) which may be sup-ported by the significant reduction in the numberof antibody-positive fish in the mesocosm contain-ing polluted harbor sludge, compared to fish inthe indirectly polluted mesocosm or referencemesocosm (Dixon et al., 1995).

Table 2Non-specific cytotoxic cell activity of mesonephros leukocytes at different effector to target (E/T) ratios of flounders exposed toTBTOa

E/T ratio

2010TBTO concentration 5

8.7910.5 (1–31)0 mg/l 19.6915.7 (5–50) 28.7918.3 (11–56)8.797.0 (1–18)*4.493.4 (0–8)*2.291.8 (0–4)*32 mg/l

a According to Grinwis et al., 1998. Mean values are expressed as % specific release9SD of 7 animals per group, the range isgiven in brackets.

* PB0.02 (Student’s t-test, 1-tailed) compared to controls.

G.C.M. Grinwis et al. / Toxicology Letters 112–113 (2000) 289–301 299

Several laboratory techniques have been de-scribed to assess the function of the immunesystem in fish (Stolen et al., 1990). Some of thesefunction tests have been adapted for Europeanflounder (Boonstra et al., 1996), but due to highinter-animal variation in test results, and prob-lems with adequate cell sorting, the adaptation ofthese tests did not result in fully operationaltests. Also, differentiation of lymphocyte subpop-ulations is hampered by the absence of specificantibodies in flounder, and infection with thelymphocystis virus, to be used as a model forhost-resistance, appeared impractical since nogrossly visible lesions were seen even after 3months. This latter finding is in contrast with thesuccessful propagation of an infection with thisvirus in the bluegill (Lepomis macrochirus ; Wolf,1962) and the plaice (Pleuronectes platessa ;Roberts, 1975). Laboratory exposure experimentswith several chemical pollutants indicated thatEuropean flounder appeared, in general, rela-tively insensitive to the individual pollutantsused. The organotin compound TBTO being anexception since aqueous exposure to this chemi-cal in concentrations in the same order of magni-tude as upper levels in the field (experiment 17.3mg TBT/l; field: 7.2 mg TBT/l) caused mortalityafter 7–12 days, resulted in gill lesions, and in-duced significant reduction of the non-specific re-sistance. Also, a significant decrease of therelative thymus volume was noted after exposureto TBTO (Grinwis et al., 1998). Lower exposurelevels of TBTO did not induce histopathologicallesions, which is in contrast with results fromexperiments with guppy (Poecilia reticulata) andmedaka (Oryzias latipes) (Wester and Canton,1987; Wester et al., 1990). From our results wetherefore conclude that TBTO might play acausal role in the increased prevalence of diseasesin the field, e.g. lymphocystis disease. Intraperi-toneal injection of five fish with 25 mg BaP/kg or15 mg DMBA/kg twice did not result in markedeffects in the treated animals. Also, no effectswere noted after aqueous exposure to PAHs inour laboratory but the real bioavailability of BaPand DMBA in these experiments was not ver-ified. Therefore we concluded that flounder ap-peared to be not very sensitive to PAHs, and that

PAHs may play only a minor role in the induc-tion of some diseases in flounder as seen in thefield, taking into account the limitations of theexposure methods used in the present experi-ments. However, possible effects of PAHs ontumor induction could not be adequately assessedin our experiments. Exposure to up to 500 mgTCDD/kg body weight or to up to 50 mg/kgbody weight of the structurally related PCB-126did not induce the marked pathology as de-scribed in several mammalian species (Kociba etal., 1978; McConnell et al., 1978; Turner andCollins, 1983). Marked pathology was also de-scribed in juvenile rainbow trout (Onchrorynchusmykiss) that showed 20% mortality, single cellnecrosis and inflammation in the liver after in-traperitoneal (ip) injection of 3.06 mg TCDD/kgbw (Van der Weiden et al., 1992). The LD50s ofTCDD in rainbow trout eggs were 0.331–0.489mg/kg egg following injection and 0.346–0.519mg/kg after waterborne exposure (Walker et al.,1992). In juvenile mirror carp (Cyprinus carpio) adecreased food intake, severe subcutaneous hem-orrhages and apathy were reported after ip injec-tion of 0.27 mg TCDD/kg bw, and 60% mortalitywas found after an intraperitoneal dose of 2.93mg TCDD/kg bw (Van der Weiden et al., 1994).However, in European flounder exposed to PCB-126 and higher doses of TCDD, a significantincreased induction of CYP1A was noted im-munohistochemically in several organs. Even ad-ministration of a low oral dose of TCDD orharbor sludge extract significantly increasedCYP1A induction in flounder hepatocytes. Re-markable in this context is the induction ofCYP1A in a distinct population of mononuclearcells in the hematopoietic tissue after exposure toTCDD or PCB-126. This induction in hemato-poietic cells has not been reported in fish before,and might be used as a ‘non-invasive’ biomarkerfor exposure to Ah-receptor binding chemicalssince these cells also appear in circulating blood.The increased mitotic activity and induction ofCYP1A in hepatocytes found in the present stud-ies might indicate that TCDD and related chemi-cals could play a role in tumor promotion in thefield.

G.C.M. Grinwis et al. / Toxicology Letters 112–113 (2000) 289–301300

References

Ariese, F., Kok, S.J., Verkaik, M., Gooijer, C., Veldhorst,N.H., Hofstraat, J.W., 1993. Synchronous fluorescencespectrometry of fish bile: a rapid screening method for thebiomonitoring of PAH exposure. Aquat. Toxicol. 26, 271–286.

Boonstra, A., Grinwis, G.C.M., Wester, P.W., Van Loveren,H., Vaal, M.A., Vethaak, A.D., Vos, J.G., 1996. Develop-ment of ex vivo/in vitro immune function tests in flounder,Platichthys flesus. Report number 732501001. NationalInstitute of Public Health and the Environment, Bilthoven.

Bucke, D., Vethaak, A.D., Lang, T., Mellergaard, S., 1996.Training guide for marine fish diseases, Coop Rep. 155ICES, pp. 76.

Dixon, P., Vethaak, A.D., Bucke, D., Nicholson, M., 1995.Preliminary study of the detection of antibodies tolymphocystis disease virus in flounder, Platichthys flesus L.,exposed to contaminated harbor sludge. Fish ShellfishImmunol. 6, 123–133.

Ellis, A.E., 1977. The leucocytes of fish: a review. J. Fish Biol.11, 453–491.

Ellis, A.E., Roberts, R.J., Tytler, P., 1989. The anatomy andphysiology of teleosts. In: Roberts, R.J. (Ed.), Fish Pathol-ogy, second edition. Balliere Tindall, London, pp. 13–55.

Evans, D.L., Carlson, R.L., Graves, S.S., Hogan, K.T., 1989.Non specific cytotoxic cells in fish (Ictalurus punctatus). II.Parameters of target cell lysis and specificity. Dev. Comp.Immunol. 8, 303–312.

Faisal, M., Ahmed, I.I, Peters, G., Cooper, E.L., 1989. Natu-ral cytotoxicity of tilapia leucocytes. Dis. Aquatic Org. 7,17–22.

Greenly, A.R., Brown, R.A., Ristow, S.S., 1991. Nonspecificcytotoxic cell of rainbow trout (Onchorhynchus mykiss) killYAC-1 targets by both necrotic and apoptic mechanisms.Dev. Comp. Immunol. 29, 1181–1187.

Grinwis, G.C.M., P.W. Wester, A. Kamstra, E.J. Van denBrandhof, J.E. Van Dijk, R.J. Leewis, M.A. Vaal, A.D.Vethaak, Vos, J.G., 1995. Laboratory maintenance, topo-graphical anatomy and histology of flounder, Platichthysflesus. Report number 732402001. National Institute ofPublic Health and the Environment, Bilthoven.

Grinwis, G.C.M., Boonstra, A., van den Brandhof, E.J., Dor-mans, J.A.M.A., Engelsma, M., Kuiper, R.V., vanLoveren, H., Wester, P.W., Vaal, M.A., Vethaak, A.D.,Vos, J.G., 1998. Short-term toxicity of bis(tri-n-butyltin)oxide in flounder (Platichthys flesus): pathologyand immune function. Aquat. Toxicol. 42, 15–36.

Kociba, R.J., Keyes, D.G., Beyer, J.E., Carreon, R.M., Wade,C.E., Dittenber, D.A., Kalnins, R.P., Frauson, L.E.,Parks, C.N., Barnard, S.D., Hummel, R.A., Humiston,C.G., 1978. Results of a two-year chronic toxicity andoncogenicity study of 2,3,7,8-tetrachlorodibenzo-p-dioxinin rats. Toxicol. Appl. Pharmacol. 46, 279–303.

McConnell, E.E., Moore, J.A., Haseman, J.K., Harris, M.W.,1978. The comparative toxicity of chlorinated dibenzo-p-dioxins in mice and guinea pigs. Toxicol. Appl. Pharmacol.44, 335–356.

Myers, M.S., Stehr, C.M., Olson, O.P., Johnson, L.L., Mc-Cain, B.B., Chan, S.-L., Varanasi, U., 1994. Relationshipsbetween toxicopathic hepatic lesions and exposure tochemical contaminants in English sole (Pleuronectes 6e-tulus), starry flounder (Platichthys stellatus), and whitecroaker (Genyonemus lineatus) from selected marine siteson the Pacific coast, USA. Environ. Health Perspect. 102,200–215.

Nigrelli, R.F., Ruggieri, G.D., 1965. Spontaneous and experi-mentally induced lymphocystis disease in fishes of the NewYork Aquarium, with a report of new cases and an anno-tated bibliography. Zoologica 50, 83–95.

Roberts, R.J., 1975. Experimental pathogenesis of lymphocys-tis in the plaice (Pleuronectes platessa). In: Page, L.A.(Ed.), Wildlife Diseases. Plenum Press, New York, pp.431–441.

Rombout, J.H.W.M., Egberts, E., Taverne, N., van Diepen,J.C.E., 1990. Analysis of the immune system of fish withmonoclonal antibodies. In: Schots, A. (Ed.), MonoclonalAntibodies in Agriculture. Pudoc, Wageningen, pp. 81–86.

Rowley, A.F., 1990. Collection, separation and identificationof fish leucocytes. In: Stolen, J.S., Fletcher, T.C., Ander-son, D.P., Robertson, B.S., van Muiswinkel, W.B. (Eds.),Techniques in Fish Immunology. SOS, Fair Haven, pp.113–136.

Russel, P.H., 1974. Lymphocystis in wild plaice Pleuronectesplatessa (L), and flounder, Platichthys flesus (L.), in Britishcoastal waters: a histological and serological study. J. FishBiol. 6, 771–778.

Schiewe, M.H., Weber, D.D., Myers, M.S., Jacques, F.J.,Reichert, W.L., Krone, C.A., Malins, D.C., McCain, B.B.,Chan, S.-L., Varanasi, U., 1991. Induction of foci ofcellular alteration and other hepatic lesions in English sole(Parophrys 6etulis) exposed to an extract of urban marinesediment. Can. J. Fish. Aquat. Sci. 48, 1750–1760.

Solemdal, P., 1967. The effect of salinity on buoyancy, sizeand development of flounder eggs. Sarsia 29, 431–442.

Stein, J.E., Hom, T., Casillas, E., Friedman, A., Varanasi, U.,1987. Simultaneous exposure of English sole (Parophrys6etulis) to sediment-associated xenobiotics: part 2 —chronic exposure to an urban estuarine sediment withadded 3H-benzo[a]pyrene and 14C-polychlorinatedbiphenyls. Mar. Environ. Res. 22, 123–149.

Stolen, J.S., Fletcher, T.C., Anderson, D.P., Robertson, B.S.,van Muiswinkel, W.B., 1990. Techniques in Fish Immunol-ogy 1. SOS, Fair Haven.

Turner, J.N., Collins, D.N., 1983. Liver morphology in guineapigs administered either pyrolysis products of a polychlori-nated biphenyl transformer fluid or 2,3,7,8-tetra-chlorodibenzo-p-dioxin. Toxicol. Appl. Pharmacol. 67,417–429.

Van Banning, P., 1987. Long-term recording of some fishdiseases using general fishery research surveys in the south-east part of the south-east North Sea. Dis. Aquat. Org. 3,1–11.

Van der Weiden, M.E.J., Kolk, J., Belumink, R., Seinen, W.,van den Berg, M., 1992. Concurrence of P450 1A1 induc-

G.C.M. Grinwis et al. / Toxicology Letters 112–113 (2000) 289–301 301

tion and toxic effects in the rainbow trout (Oncorhynchusmykiss) after administration of a low dose of 2,3,7,8-tetra-chlorodibenzo-p-dioxin. Aquat. Toxicol. 24, 123–142.

Van der Weiden, M.E.J., Bleumink, R., Seinen, W., van denBerg, M., 1994. Concurrence of P450 1A induction andtoxic effects in the mirror carp (Cyprinus carpio) afteradministration of a low dose of 2,3,7,8-tetrachlorodibenzo-p-dioxin. Aquat. Toxicol. 29 (3–4), 147–162.

Van Muiswinkel, W.B., Lamers, C.H.J., Rombout, J.H.W.M.,1991. Structural and functional aspects of the spleen inbony fish. Res. Immun. 142 (4), 362–366.

Vethaak, A.D., 1987. Fish diseases, signals for a diseasedenvironment? Second International North Sea Seminar,Rotterdam, October 1–3. Peet, G. (Ed.). Werkgroep No-ordzee Reasons for Concern, vol. 2, pp. 41–61.

Vethaak, A.D., 1993. Fish disease and marine pollution: a casestudy of the flounder (Platichthys flesus) in Dutch coastaland estuarine waters. PhD thesis, Amsterdam.

Vethaak, A.D., ap Rheinallt, T., 1992. Fish disease as amonitor for marine pollution: the case of the North Sea.Rev. Fish Biol. Fish. 2, 1–32.

Vethaak, A.D., Jol, J., 1996. Diseases of flounder (Platichthysflesus) in Dutch coastal and estuarine waters, with particu-lar reference to environmental stress factors, Part 1. Epi-zootiology of gross lesions. Dis. Aquat. Org. 26, 81–97.

Vethaak, A.D., Jol, J.G., Meijboom, A., Eggens, M.L., apRheinallt, T., Wester, P.W., van de Zande, T., Bergman,A., Dankers, N., Ariese, F., Baan, R.A., Everts, J.M.,Opperhuizen, A., Marquenie, J.M., 1996. Skin and liverdiseases induced in flounder (Platichthys flesus) after long-term exposure to contaminated sediments in large-scalemesocosms. Environ. Health Perspect. 104 (11), 1218–1229.

Vethaak, A.D., Wester, P.W., 1996. Diseases of flounderPlatichthys flesus in Dutch coastal and estuarine waters,with particular reference to environmental stress factors.II. Liver histopathology. Dis. Aquat. Org. 26, 99–116.

Von Westernhagen, H., 1970. Erbrutung der eier von dorsch(Gadus morhua), flunder (Platichthys flesus) und scholle(Pleuronectes platessa) unter combinierten temperatur- undsalzgehaltbedingungen. Helgol. Wiss. Meeresunters. 21,21–102.

Walker, M.K., Hufnagle, L.C. Jr, Clayton, M.K., Peterson,R.E., 1992. An egg injection method for assessing early lifestage mortality of polychlorinated dibenzo-p-dioxins,dibenzofurans, and biphenyls in rainbow trout,(Onchorhynchus mykiss). Aquat. Toxicol. 22, 15–38.

Ward, E.C., Murray, M.J., Dean, J.H., 1985. Immunotoxicityof nonhalogenated polycyclic aromatic hydrocarbons. In:Dean, J.H., Luster, M.I., Munson, A.E., Amos, H. (Eds.),Immunotoxicology and Immunopharmacology. RavenPress, New York, pp. 291–303.

Wester, P.W., Canton, J.H, 1987. Histopathological study ofPoecilia reticulata (guppy) after long-term exposure tobis(tri-n-butyltin)oxide (TBTO) and di-n-butyltinchloride(DBTC). Aquat. Toxicol. 10, 143–165.

Wester, P.W., Canton, J.H., Van Iersel, A.A.J., Krajnc, E.I.,Vaessen, H.A.M.G., 1990. The toxicity of bis(tri-n-butyltin)oxide (TBTO) in small fish species Oryzias latipes(medaka) and Poecilia reticulata (guppy). Aquat. Toxicol.16, 53–72.

White, K.L., Kawabata, T.T., Ladics, G.S., 1994. Mechanismsof polycyclic aromatic hydrocarbon immunotoxicity. In:Dean, J.H., Luster, M.I., Munson, A.E, Kimber, I. (Eds.),Immunotoxicology and Immunopharmacology, secondedition. Raven Press, New York, pp. 123–142.

Wiklund, T., 1994. Skin ulcer disease of flounder (Platichthysflesus) — a review. In: Bylung, G., Lonnstrom L.G.(Eds.), Diseases and Parasites of Flounder (Platichthysflesus) in the Baltic Sea. Baltic Marine Biologists Publica-tion 15, pp. 17–26

Wolf, K., 1962. Experimental propagation of lymphocystisdisease of fishes. Virology 18, 249–256.

Wolf, K., 1988. Lymphocystis diseases. In: Wolf, K. (Ed.),Fish Viruses and Fish Viral Diseases. Cornell UniversityPress, Ithaca, NY, pp. 268–291.

Zapata, A.G., Cooper, E.L., 1990. The Immune system: Com-parative Histophysiology. Wiley, Chichester.

Zedeck, M.S., 1980. Polycyclic aromatic hydrocarbons: a re-view. J. Environ. Pathol. Toxicol. 3, 537–567.

Zelikoff, J.T., 1994. Fish immunotoxicology. In: Dean, J.H.,Luster, M.I., Munson, A.E., Kimber, I. (Eds.), Im-munotoxicology and Immunopathology, second edition.Raven Press, New York, pp. 71–95.

.