acta histochemica 113 (2011) 201–213

Contents lists available at ScienceDirect

acta histochemica

0065-12

doi:10.1

n Corrnn Cor

E-m

journal homepage: www.elsevier.de/acthis

Ultrastructural and immunohistochemical investigation on the gills of theteleost, Thalassoma pavo L., exposed to cadmium

Elvira Brunelli a, Angela Mauceri b, Maria Maisano b, Ilaria Bernab �o a, Alessia Giannetto b,Elena De Domenico b, Barbara Corapi a, Sandro Tripepi a,n, Salvatore Fasulo b,nn

a Abele Saita Electron Microscopy Laboratories, Department of Ecology, University of Calabria, Via P. Bucci, I-87036 Rende (Cosenza), Italyb Department of Animal Biology and Marine Ecology, University of Messina, Contrada Sperone 31 I-98166, S. Agata, Messina, Italy

a r t i c l e i n f o

Article history:

Received 9 August 2009

Received in revised form

30 September 2009

Accepted 1 October 2009

Keywords:

Gills

Ultrastructure

Electron microscopy

Cadmium

Immunohistochemistry

Teleost

Thalassoma pavo

81/$ - see front matter & 2009 Elsevier Gmb

016/j.acthis.2009.10.002

esponding author. Fax:+39 0984 492986.

responding author. Fax: +39 090 6765556.

ail addresses: s.tripepi@unical.it (S. Tripepi), s

a b s t r a c t

An investigation was conducted to determine the effects of the heavy metal, cadmium (Cd), on the gills

of the teleost fish, Thalassoma pavo Linnaeus, 1758. The fishes were exposed to several sublethal

concentrations of cadmium (10, 40, 60 and 120 mM (mg/L)) for a period of 48, 96 and 192 h. The value of

the LC50 after 96 h of cadmium exposure, determined using the System of Finney, was equal to

128.3 mM. The gills of the fishes were examined by light and electron microscopy. Toxic, apoptotic and

cadmium effects were analyzed using some neuropeptides, metallothioneins (MT), caspase 3, PCNA and

calmodulin, as bioindicators, respectively. The results showed that the alterations in the gills were

proportional to the exposure periods and concentrations of the metal, which were found to be both dose

and time dependent. The biological responses in the gills of the tested animals are discussed in relation

to results obtained by analysis of the biomarkers. These data may be used for the planning of a model to

determine biological risk in the marine environment and may be particularly useful to investigate

organisms exposed to cadmium.

& 2009 Elsevier GmbH. All rights reserved.

Introduction

Coastal seawater is easily contaminated by heavy metals due tohuman activities with heavy metal contamination reported inaquatic organisms (Olojo et al., 2005). Cadmium (Cd) is a fairlycommon pollutant of the smelting and metal-plating industriesand is also naturally found in trace amounts in aquatic systems.The effects of water-borne cadmium on fishes have been wellstudied, and include alterations of gill structures, the destructionof neural functions and ion disturbances (Wu et al., 2006). Thetoxicity of this metal on aquatic organisms is influenced bychemical features of water, such as pH and hardness (Mance,1987) and its bioaccumulation is directly related to its concentra-tion in seawater (Sadik, 1992). The problem has become moreserious for aquatic species that live close to the coastline whereheavy metals tend to accumulate (Migliarini et al., 2005). As oneof the key compartments of fishes, gills are physiologicallycomplex and are also the first and the most important target ofcadmium and water-borne toxicants (Tao et al., 2000) prior to thetransfer of the cadmium to internal organs (such as the liver,kidneys, intestine) via the circulatory system (De Smet and Blust,2001; Wu et al., 2006).

H. All rights reserved.

fasulo@unime.it (S. Fasulo).

In the present study, several specimens of Thalassoma pavo,also known as the Ornate Wrasse, a widespread fish species in theMediterranean and popular for culture in aquaria, were exposedto varied concentrations and times with cadmium to obtain datathat can be used as early markers of cadmium pollution (Langstonet al., 2002). The present study focuses on the ultrastructuralalterations of epithelial, mucus, pillar and chloride cells in thegills. The biological responses in the gills of the tested animals arediscussed in relation to results obtained by analysis of thebiomarkers.

Cadmium is biologically non-essential, but is an important metalfor industrial applications (Wu et al., 2009) and can damage gills.Chloride cells (CC) are the primary targets of water-borne cadmium,because this metal decreases the activity of gill Ca++-ATPase, whichleads to hypocalcemia in the fishes and causes problems in calciumhomeostasis (Wu et al., 2007). In addition cadmium has negativeeffects on respiratory functions and osmoregulation (Patrap andWendelaar Bonga, 1993; Castan~o et al., 1998).

Histopathological studies are commonly used to evaluate thetoxic effects of pollutants and provide a useful means to evaluatewater quality (FAO, 1981; EIFAC, 1983; Murty, 1986). In this paperthe metallothioneins (MTs) were detected immunohistochemi-cally. These metal-binding proteins play an important role inmetal metabolism or detoxification of heavy metals, includingcadmium (Wu et al., 2006; Fasulo et al., 2008). Induction of MTshas been proposed as useful biomarkers for eco-toxicologicalstudies on aquatic animals (Alvarado et al., 2007).

E. Brunelli et al. / acta histochemica 113 (2011) 201–213202

Immunohistochemical techniques were used to investigate theeffects of cadmium on the respiratory epithelia, with particularreference to the presence of neuronal nitric oxide synthase (nNOS)and vasoactive intestinal peptide (VIP). Nitric oxide is involved inregulation of a variety of processes including vascular tone,neurotransmission and ion balance in mammals and fishes(Mauceri et al., 1999; Zaccone et al., 2003; Hyndman et al.,2006). The expression of nNOS in the gills of the killifish (Fundulus

heteroclitus) suggests that the enzyme may regulate systemic iontransport as a paracrine-signaling molecule (Mauceri et al., 2005;Hyndman et al., 2006).

Cell proliferation was detected through immunolabeling forproliferating cell nuclear antigen (PCNA) and apoptosis byimmunolabeling for caspase-3. Cell death and proliferation havealso been proposed as biomarkers of exposure to heavy metals(Berntssen et al., 2001; Ferrando et al., 2005; Mauceri et al., 2005).PCNA is a highly conserved molecule, essential for the synthesis ofDNA in the S-phase of the cell cycle. It is known that serotonin(5-hydroxytryptamine, 5-HT) modulates cell renewal in differenttissues, and in particular in the proliferation and migration in avariety of cell types (Azmitia, 2001).

The aim of this study was to assess the effects of exposure ofcadmium on the renewal of the gill epithelium of the fishes and torelate this to the presence and distribution of molecularbiomarkers that could be employed to monitor natural aquaticbioenvironments.

Materials and methods

Fish maintenance and holding conditions

The T. pavo specimens used in this study were collected from alocation on the Tyrrhenian coast (S. Lucido) using baited traps andthey were acclimatized to laboratory conditions for 5 days. Duringthis period, fishes were maintained in previously prepared aquariaof 150 l capacity (10 fishes per tank) with seawater taken from thecapture site and equipped with filtration and oxygenationsystems. During the acclimatization period, salinity (35%),density (1.027–1.028 g/cm3), temperature (18 –24 1C) and theconcentration of nitrites and nitrates were measured and keptconstant (dissolved oxygen 8–8.6 mg/L, hardness 100 mg CaCO3/Land absence of heavy metals). For the entire duration of theexperiment, the animals were maintained under a natural light/dark cycle and fed every 2 days with commercial fish food(Tetramin). Animal maintenance and experimental procedureswere in accordance with the ‘‘Guide for Care and Use ofLaboratory Animals’’ (European Communities Council Directive,1986).

Exposure to cadmium and sampling

A semi-static exposure system was used in accordance withstandard procedure guidelines (ASTM, 1997; EPA, 2002). In thismethod the experimental solution and the samples (i.e. fish) areput in a test chamber (i.e. aquarium) and the solutions wererenewed every 24 h. The contaminant was prepared by dissolvingCdCl2 �2 H2O (Sigma-Aldrich, Milan, Italy) in appropriate volumesof a vehicle that consisted simply of seawater. For the acutebioassay tests, 15 fishes were used for each concentration,including the control; a total of three replicates were carried outfor each dose. The number of dead fish were counted every 12 hand removed immediately from the aquaria. The lethal concen-tration for 50% of the animals (LC50) was determined by the use ofFinney’s Probit Analysis LC50 Determination Method (Finney,

1971). The computer analysis was conducted with LC50 1.00software developed by EPA (LC50 Software Program Version 1.00,1999; Centre for exposure assessment modeling – CEAM,Washington, DC, USA).

After identifying the lethal concentration, some fishes wereexposed to five non-lethal concentrations of Cd (10, 20, 40, 60and 120 mM) under the experimental conditions described above.The control group (n=8) was maintained in the same conditionswith the exception that the vehicle only was added to thetanks. Fishes of comparable body dimensions were randomlyassigned to the various exposure tanks. Two replicates, eachcontaining 5 fishes were used for the control and the exposuregroup.

The gills were removed after 48, 96 and 192 h and the controlsample gills were removed at the same time. The animals wereanesthetized with 2–4 g/l tricaine methane sulphonate (MS 222,Sandoz, Sigma-Aldrich, St. Louis, Mo) and killed by spinal cordtransection. All animal handling was performed according toEthical Committee recommendations and under supervision ofauthorized investigators.

Electron microscopy

The excised gills were fixed using Karnovsky’s fixative in acacodylate buffer (pH 7.4). After post-fixation with osmium tetroxide(1% in the same buffer) and dehydration in graded ethanols,specimens for light microscopy (LM) and transmission electronmicroscopy (TEM) were placed in propylene oxide and embeddedin Epon-Araldite. After heat polymerization, semi-thin sections(1–2 mm) were stained with Grimley’s dyes (toluidine bluemethod–malachite green–acid fuchsin) (Grimley, 1964) and ob-served using a Leitz Dialux 20 EB light microscope. Ultrathinsections (90 nm), cut on the ultramicrotome, were contrast stainedwith lead citrate and uranyl acetate and then examined in thetransmission electron microscope operating at 50 kV (ZeissEM 900).

The samples used for scanning electron microscope (SEM)observation, after dehydration, were subjected to the progressivesubstitution of ethanol with hexamethyldisilazane (HSDM),removed by complete evaporation (Nation, 1983), coated withgold in an Emitech K550 ion sputter unit (Quorum TechnologiesLtd., The Broyle Ringmer, East Sussex, UK) and then examinedunder a Zeiss DSM 940 scanning electron microscope.

Histological markers

Samples of gill tissues for histological assessment were fixedin 4% paraformaldehyde in 0.1 M phosphate buffered solution(pH 7.4) at 4 1C, dehydrated in ethanol and embedded in Paraplast(Bio-Optica, Milan, Italy). Histological sections (5 mm thick) werecut with a rotary automatic microtome (Leica Microsystems,Wetzlar, Germany). Sections were mounted on glass slides andstained with hematoxylin/eosin (Bio-Optica, Italy) to visualizetypical morphological features. The staining methods for neutralpolysaccharides (PAS, periodic acid-Schiff, staining) and acidmucopolysaccharides (Alcian blue, pH 2.5), respectively wereused to detect the presence of mucosubstances in the mucouscells in the gill epithelium. Quantification of acid and neutralmucous stained cells of 30 specimens was performed by countingthe positive cells with a Zeiss AxioVision image analysis system(release 4.5 software) and results were analyzed statistically withone-way analysis of variance (ANOVA) using GraphPad (InStat)software 3.0 (GraphPad Software, La Jolla, CA, USA).

E. Brunelli et al. / acta histochemica 113 (2011) 201–213 203

Immunohistochemistry

Some sections were processed by the immunoperoxidase andimmunofluorescence methods as previously reported (Mauceriet al., 1999; Zaccone et al., 2003) to reveal various antigens andpeptides (see Table 1). After inhibition of endogenous peroxidaseactivity using 1% H2O2 in PBS for 30 min, sections were incubated inthe presence of antisera overnight at 4 1C in a moist chamber.Binding sites of antibodies were IgG (Sigma-Aldrich) tetramethylrhodamine isothiocyanate (TRITC)-goat anti-rabbit (Sigma-Aldrich),and peroxidase, diluted 1:100 at room temperature. The peroxidasewas visualized using a fresh solution of 3,3-diaminobenzidine-4HCl(DAB, 0.3 mg ml�1) and H2O2 (0.01%) in PBS.

A Zeiss AxioImager Z1 microscope integrated with AxioVision4.5 software and an AxioCam digital camera (Zeiss) was used forimage acquisition. Sections were imaged using the appropriatefilter settings for the excitation of FITC (480–525 nm) and TRITC(515–590 nm).

Table 3Estimated LC values and confidence limits.

Point Conc.

Lower Upper

Controls

Negative controls for immunohistochemical labelling wereperformed by incubating without the primary and secondaryantibody. Incubating of some peptide labelling was verified byprocessing sections with antiserum preabsorbed with respectiveantigen (10–100 mg/ml) overnight at 4 1C. VIP was provided byBachem, nNOS enzyme, and 5-HT were purchased from Biomol(Enzo Life Sciences Int., Plymouth Meeting, USA) and Sigma-Aldrich, respectively.

LC 1.00 57.389 37.704 71.245

LC 5.00 72.655 53.500 85.544

LC 10.00 82.391 64.312 94.542

LC 15.00 89.690 72.684 101.333

LC 50.00 128.394 116.054 142.720LC 85.00 183.802 161.641 230.437

LC 90.00 200.084 173.118 260.633

LC 95.00 226.896 191.195 313.525

LC 99.00 287.252 229.427 45.150

Exposure 95% confidence limits.

Statistical analyses

Histochemical and immunohistochemical results were per-formed using AxioVision Release 4.5 software. This programallowed to count the positive cells and the obtained data werestatistically processed with ANOVA system by Graphpad Instatsoftware (GraphPad Software, La Jolla, CA, USA).

Table 1Primary antibodies used.

Antigen Animal source

Neuronal nitric oxide synthase (nNOS) Rabbit

5-Hydroxytryptamine (5-HT) Mouse

VIP Rabbit

MT Rabbit

PCNA Rabbit

Caspase 3 Rabbit

Calmodulin Rabbit

Table 2Relationship between cadmium concentration and mortality rate of Thalassoma pavo.

Concentration

(mM)

Number of exposed

fish

Number of dead

fish

Observed proportion

responding

20.0000 15 0 0.0000

40.0000 15 0 0.0000

60.0000 15 0 0.0000

80.0000 15 2 0.1333

110.0000 15 3 0.2000

120.0000 15 7 0.4667

130.0000 15 9 0.6000

150.0000 15 10 0.6667

200.0000 15 13 0.8667

300.0000 15 15 1.0000

Results

Acute toxicity

Table 2 shows the relationship between the cadmium concen-tration and the mortality rate of T. pavo according to Finney’s ProbitAnalysis using EPA Computer Program (LC50 Software ProgramVersion 1.00, 1999; Centre for exposure assessment modeling,CEAM – Washington, DC, USA). The nominal LC50-96 h value forcadmium was 128.3 mM (28.16 mg/l). No mortality occurred in thecontrol group. Estimated LC50 values and confidence limits for 96 hcadmium exposure are presented in Table 3.

Histopathology

The histopathological effects of cadmium on the gills of T. pavo

were analyzed by light microscopy. The general gill morphology istypical of other teleost fishes (Evans et al., 2005). The gill archescarry typical filaments and lamellae (Fig. 1A). The gill epitheliumis composed of several cell types: pavement cells (PVCs),mitochondrion-rich cells (MRCs), neuroepithelial cells (NECs)and mucous cells (Wilson and Laurent, 2002). The mucous

Dilution Distributor

1:100 Biomol, Milan, Italy

1:50 Dako Cytomation, Milan, Italy

1:50 Sigma-Aldrich, St. Louis, MO, USA

1:300 Peninsula Labs, San Carlos, CA, USA

1:100 Sigma-Aldrich, St. Louis, MO, USA

1:100 Sigma-Aldrich, St. Louis, MO, USA

1:100 Sigma-Aldrich, St. Louis, MO, USA

Proportion responding adjusted for

controls

Predicted proportion

responding

0.0000 0.0000

0.0000 0.0004

0.0000 0.0140

0.1333 0.0859

0.2000 0.3275

0.4667 0.4226

0.6000 0.5143

0.6667 0.6734

0.8667 0.8998

1.0000 0.9929

Fig. 1. Hematoxylin and Eosin (H&E) and alcian blue – periodic acid Schiff (AB-PAS) staining in the gill of Thalassoma pavo exposed to cadmium compared with control

specimens characterized by damaged structures (B) and numerous mucous cells (D). Mean and standard deviation (SD) calculated from the number of cells positive to the

AB-PAS reaction in 48 h (E), 96 h (F) and 192 h (G) of treatment. The values are statistically different (Po0.001) for all times and concentrations considered. Scale bar,

20 mm.

E. Brunelli et al. / acta histochemica 113 (2011) 201–213204

or goblet cells present in the epithelium of gill filamentswere revealed by the Alcian blue-PAS reaction (AB/PAS)(Fig. 1C).

Morphological analyses of gills, after 48 h of exposure tocadmium, showed histopathological alterations at all testedconcentrations with evident lifting of the lamellar epithelium.

Filament and lamellar telangiectasis was found with increasedtime of exposure and concentration (Fig. 1B). It was followed by aconcomitant thinning of the gill epithelium.

The AB/PAS reaction detected a reduced number of mucousacidophilic cells in relation to exposure times and pollutantconcentration (Fig. 1D–G).

Fig. 2. . Immunohistochemical labelling for 5-HT (A–C), VIP (D–F) and nNOS (G–I) in the gills of Thalassoma pavo treated with cadmium. Scale bar, 20 mm.

E. Brunelli et al. / acta histochemica 113 (2011) 201–213 205

Immunohistochemistry

5-HT

In the gills of control fishes, numerous serotonin immunor-eactive cells were detected along the branchial filament (Fig. 2A).At 48 h, after an initial inhibition at lower concentrations 10 mM,at higher levels numerous serotonin immunoreactive fibers andsome NECs were observed in the epithelium of the filament. After96 h serotonin immunoreactive NECs were present at 10 and20 mM concentrations, while few serotonin immunoreactivevaricose nerve fibers were observed at 120 mM (Fig. 2B). After192 h at 40–60 mM concentrations numerous NECs and fiberswere revealed and at 120 mM these appeared distributed along thefilament and the lamellae (Fig. 2C).

VIP and nNOS

In the controls, numerous VIP (Fig. 2D) and nNOS-immunos-tained NECs were detected in the interlamellar region, near theCCs (Fig. 2G). At every concentration and at each exposure timethe VIP did not show significant variations (Fig. 2E and F).

nNOS-immunostained NECs were present along the gillfilament in the specimens exposed to several concentrations

during the 48 and 96 h (Fig. 2H). At 192 h there was evidence of agradual inhibition of immunopositivity with a total lack at120 mM concentration (Fig. 2I).

Calmodulin

Calmodulin (Cam) immunoreactivity was located along the gillfilament cells in the control samples (Fig. 3A). After 48 h of ex-posure to cadmium a reduction of immunopositive NECs wasobserved in the epithelium of specimens exposed to a concen-tration of 120 mM. After 96 h few immunopositive NECs were seenat all concentrations (Fig. 3B). At 192 h at a concentration of60 mM, numerous immunopositive NECs were seen distributed inthe very thin epithelium (Fig. 3C), while at higher concentrationsfew such cells were detected.

PCNA

In the control, PCNA was present in the basal portion of the gillfilament (Fig. 3D). During 48 h at all concentrations, a largenumber of positive elements were observed in comparison withthe control. At 96 h a slight inhibition of PCNA was noted in thecell nuclei distributed both along the filaments and the lamellae

Fig. 3. . Immunohistochemical labelling for CAM (A–C), PCNA (D–F) Caspase 3 (G–I) and MT (K–M) in the gills of Thalassoma pavo treated with cadmium. Scale bar, 20 mm.

E. Brunelli et al. / acta histochemica 113 (2011) 201–213206

(Fig. 3E) and this reduction increased after 192 h already at lowerconcentrations (Fig. 3F).

Caspase-3

In the control samples, caspase-3 was immunolocalized in thebasal portion of the gill filament (Fig. 3G). At 48 h an inhibition ofcaspase-3 at low concentrations was observed, while it increasedat 60 mM. After 96 h, numerous immunoreactive cells wererevealed at all concentrations (Fig. 3H). At 192 h there wasevidence of a gradual inhibition of immunoreactivity and a lack ofimmunoresponse at a concentration of 120 mM (Fig. 3I).

MT

MT immunoreactive cells were not detected in the gillepithelium of the control specimens (Fig. 3K). At 48 h of cadmiumexposure already at 40 mM concentration CC showing MTimmunoreactivity were observed along the filament. This im-munopositivity increased in the gill epithelium after 96 h (Fig. 3L).At 192 h of cadmium exposure at all concentrations numerousimmunopositive cells were present (Fig. 3M).

Statistical analyses

The values of mean and standard deviation of all histo-chemical and immunohistochemical results are shown in graphsin Table 4.

Morphological and ultrastructural observations

After 48 h of treatment with cadmium histopathologicalalterations were seen at all tested concentrations. At lowestconcentrations (10 and 20 mM) the gills showed a typicalepithelial organization; the first structural modification could beobserved with the SEM: PVCs showed irregularly formed micro-ridges (Fig. 4A). At intermediate concentrations (40–60 mM) theabsence of microridges was noted particularly in the proximalportion and along the margins of lamellae (Fig. 4B). At the highestcadmium concentration (120 mM) the gills showed an irregularappearance and the surface epithelium was infolded at severalpoints; a contraction of the PVCs provided an outlet to the surfaceof underlying cells (Fig. 4C). Secondary lamellae became thin andinfolded in the distal portion (Fig. 4D).

Table 4Mean and standard deviation (sd) of immunopositive cells.

E. Brunelli et al. / acta histochemica 113 (2011) 201–213 207

Goblet cells and chloride cells showed normal ultrastructuralcharacteristics, but they were not covered by the adjacent PVC andshowed an enlarged portion that was directly in contact with theexternal medium (Fig. 4E and F). The secondary lamellae appearedundamaged, while starting from 40 mM cadmium concentration,hypertrophic pavement cells with irregular shape, long cytoplasmic

processes and without microridges were observed (Fig. 4G). Ultra-structural analysis after 96 h of exposure to cadmium revealedconspicuous alterations, which were found on secondary lamellaewith profuse amounts of mucus (Fig. 5A and B); it was also possible tosee a thinning of the lamellae (Fig. 5C). At the highest concentrations,frequent swelling was seen in the distal portion of secondary lamellae

Fig. 4. . Thalassoma pavo gills after 48 h Cd exposure. (a and b) Scanning electron micrographs of gills exposed to 10 mM Cd. (F, gill filament; L, secondary lamellae). Note the

irregularly formed microridges of pavement cells (PVC) (a) and the loss of the microridges along the margins of lamellae (arrowhead) (b). (c and d) Scanning electron

micrograph of gills exposed to 120 mM Cd. Note the irregular arrangement of the epithelial surface, the outlet of underlying cells (arrowhead) (c) and the thinning of

secondary lamellae (d). (e and f) Transmission electron micrographs of primary epithelium exposed to 40 mM Cd. Goblet cells (GC) (e) and chloride cells (CC) (f) are not

covered by the adjacent PVC. (f) Transmission electron micrographs of secondary lamellae with hypertrophic pavement cells (P, pillar cell).

E. Brunelli et al. / acta histochemica 113 (2011) 201–213208

(Fig. 5D). PVCs lost their structural arrangement at all testedconcentrations giving the surface a wrinkled and non-homogenousappearance (Fig. 5E).

TEM observations also showed ultrastructural alterations in thesecondary lamellae with an alteration of cells at all concentrations(Fig. 5F). In the secondary lamellae degenerating or apoptotic cellswere present starting from 40 mM (Fig. 5G). After 192 h of treatment,alterations were already conspicuous at low concentrations. At aconcentration of 20 mM an accentuated surface epithelial separationfrom the lower layer was observed (Fig. 6A). At 40 mM mucoussecretion was abundant (Fig. 6B). At high concentrations the gillstructure appeared markedly modified with irregular formed lamellae

(Fig. 6C) and with hypertrophy of the surface epithelium, both infilaments and in lamellae (Fig. 6D).

Ultrastructural analysis at 20 mM confirmed the presence ofepithelial separation that, in the main filament, gave the appear-ance of deep invaginations bounding GCs and CCs, which werenumerous, but undamaged (Fig. 6E and F). The surface of secondarylamellae had an irregular appearance with lifting of the superficiallayer, with degenerative processes common in the pillar cells(Fig. 7A). At a concentration of 60 mM, disorganization of theepithelial layer in the main filament was observed, while insecondary lamellae a detachment of epithelium from the basallayer was detected accompanied by degeneration or disappearance

Fig. 5. . Thalassoma pavo gills after 96 h Cd exposure. (a–d) Scanning electron micrographs of secondary lamellae. (F, gill filament; L, secondary lamellae). A thick blanket of

mucus (n) is noticeable from 20 mM (a) to 60 mM (b) exposed groups. (c) Starting from 60 mM a thinning of the lamellae and swelling in distal portion were also observed

(d). (e) Scanning electron micrographs of gills exposed to 120 mM Cd; note the alterations of epithelium surface, which consist of the wrinkling of the pavement cells, loss of

the microridges, and thinning of the lamellae. (f and g) Transmission electron micrograph showing ultrastructural alterations in both primary filament (f) and secondary

lamellae (g); several degenerating cells (arrows) are evident on the lamellae surface. (CC, chloride cells).

E. Brunelli et al. / acta histochemica 113 (2011) 201–213 209

of pillar cells (Fig. 7B). This phenomenon was accentuated at120 mM, when degenerating pavement cells with electron-lucentcytoplasm and without organelles were observed (Fig. 7C). We alsofound cytoplasmic degeneration in chloride cells, seen as manylacunae that interrupted the tubular vesicular system (Fig. 7D).

Discussion

On the basis of the LC50 value and according to guidelines of OECD(1992), cadmium is classified as toxic for aquatic life. The resultsobtained in our study are in the range of values reported in the

literature (Table 5) and in particular they are similar to the resultsobtained by Robohm (1986) in the labridae, Tautogolabrus adspersus.According to the literature, it appears that freshwater fishes are moresensitive to heavy metals than seawater fishes.

The different LC50 values also depend on the different methodsused to determine it. We used Finney’s method as recommended byEPA, but in many studies other methods were used.

Gills of fishes are physiologically complex and are also animportant target of toxic metals in surrounding water (Evans,1987; Olsson et al., 1998; Tao et al., 2000). Cadmium mainlyaccumulates in three target organs: kidney, liver and gills (Noreyet al., 1990) and, among these, the gills are least capable of

Fig. 6. . Thalassoma pavo gills after 192 h Cd exposure. (a–d) Scanning electron micrographs of gill filaments. (F, gill filament; L, secondary lamellae). (a) The epithelial lifting

is evident starting from 20 mM Cd. (b) Mucous secretion cover epithelial surface at 40 mM Cd. (c and d) High Cd concentrations: note the irregular arrangement of lamellae,

and the alteration of gills surface; the hypertrophy of the surface epithelium is more evident at 120 mM-Cd. (e and f) Transmission electron micrographs of filament showing

observe the appearance of deep invaginations that bounds GCs and CCs at 20 mM-Cd; note that the ultrastrucural features are preserved.

E. Brunelli et al. / acta histochemica 113 (2011) 201–213210

sequestering or detoxifying cadmium (Hans et al., 2001; Moiseenkoand Kudryavtseva, 2001). The results of our study demonstrated thatexposure to sublethal concentrations of cadmium induces modifica-tions in gills even following short-term exposure. The degree ofhistopathological alterations of gills is closely linked to pollutantconcentrations and duration of exposure.

Gill damage and structural changes caused by water-bornecadmium have been reported for several fish species. Severalauthors reported histopathological lesions similar to that ob-served in T. pavo. In Macropsobrycon uruguayanae exposed tocadmium the reported effects included hyperplasia of bothprimary and secondary lamellar epithelium, fusion of adjacentsecondary lamellae, necrosis, telangiectasis, hyperplasia of chlor-ide cells and metaplasia of goblet cells (Randi et al., 1996).

Hypertrophy and hyperplasia of primary and secondary gilllamellae were also reported by Wangsongsak et al. (2007) in gillsof Common silver barb exposed to 0.06 mg/L cadmium for 60 d. InTilapia (Oreochromis mossambicus), Patrap et al. (1993) reportedan increase of CC as an initial response to cadmium exposure.However, it is difficult to determine a specific morphologicalresponse of gill epithelium to cadmium exposure and in fact mostof these morphological alterations (epithelial lifting, degenerationof pavement cells, microridges) were also found after exposure toother toxicants (Martinez et al., 2004). This could be related to thegeneral increase of epithelial layers aimed to make cadmiuminflux more difficult (Wendelaar Bonga et al., 1990).

The first defense mechanism in gills against exposure toheavy metals is the secretion of acidophilic mucus (McDonald

Fig. 7. . Thalassoma pavo gills after 192 h Cd exposure. (a and b) Transmission electron micrographs of secondary lamellae at 40–60 mM-Cd. The lifting of superficial layer is

evident (arrowhead); note regular (PC) and degenerating pillar cells. (c and d) Degenerating cells with electron-lucent cytoplasm and without organelles were observed at

120 mM-Cd. We also recognized the cytoplasm degeneration of a chloride cell.

Table 5The range of LC50 values reported in the literature.

Species LC50-96h Authors FW/SW

Oncorhyncus mykiss 0.022 mg/l Hollis et al. (1999) FW

Oncorhyncus mykiss 0.45 mg/l Oryan et al. (1997) FW

Oncorhyncus mykiss 0.018 mg/l Szebedinsky et al. (2001) FW

Salvelinus fontinalis 3.362 mg/l Holcombe et al. (1976) FW

Puntius gonionodus 2.30 mg/l Mungkung et al. (2001) FW

Poecilia reticulata 30.4 mg/l Yilmaz et al. (2004) FW

Cyprinus carpio 121.8 mg/l Muley et al. (2000) FW

Morone saxatilis 20 mg/l Robohm (1986) FW

Jordanella floridae 2500 mg/l Spehar (1976) estuarine

Labeo rohita 89.5 mg/l Dutta et al. (2000) estuarine

Rivulus marmoratus 24.48 mg/l Lin et al. (1993) estuarine

Dicentrarchus labrax 6.17 mg/l Gelli et al. (1999) SW

Tautogolabrus adspersus 26 mg/l Robohm (1986) SW

FW, freshwater; SW, seawater.

E. Brunelli et al. / acta histochemica 113 (2011) 201–213 211

and Wood, 1993; Wu et al., 2007), however, this phenomenonis reduced at higher exposures. This observation agrees withSanchez et al. (1997), who showed that the acidificationof the mucus layer appears to increase the protective function ofmucus.

TEM observations show that the CCs preserve their structuralintegrity throughout almost all the experimental period ofcadmium exposure. At 192 h and at highest concentrations theCCs appear modified. Hyperplasia of filament epithelium, withpartial or complete fusion of lamellae, is very frequent in T. pavo,mostly after 96 h. The general thickness of the filament epithe-

lium has been included in a group of later compensating changes,which appears to be associated with the repair of gill damage.Acute exposure to toxicants could first induce degeneration oflamellae (Heath, 1987). Lamellar fusion could be a defensemechanism since it reduces vulnerable gill surface (Heath, 1987;Randi et al., 1996; Au, 2004; Martinez et al., 2004; Garcia-Santoset al., 2006).

The results for Cam positive cells showed inhibition at higherconcentration (treatment 120 mM of Cd) after 96 and 192 h withconsequent inhibition of the activity Ca++-ATPase, as seen in otheraquatic organisms (Wong and Wong, 2000; Niyogi and Wood,2006). This disturbs the calcium balance and induces negativeeffects on respiratory functions and osmoregulation (Wu et al.,2007), whereas at the low concentrations (treatment 20 mM of Cd)and prolonged exposure (192 h) there was no alteration in thedensity of the Cam immunopositive cells.

The alterations of the gill epithelium and the neuroepithelialcells, considered to be oxygen receptors (Jonz et al., 2004; Evanset al., 2005), are caused by the acute toxicity of the cadmium andthey may provoke an oxygen deficit and hypoxia. The localreaction to hypoxia involves different neuromodulators andneurotransmitters such as NO and VIP, as observed in otherspecies subjected to environmental stress (Mauceri et al., 2005).The presence of intense nNOS immunoreactivity was located inthe interlamellar portion of the gills in the specimens of thecontrol as in other species of studied teleosts (Mauceri et al., 1999;Zaccone et al., 2003; Ebbeson et al., 2005; Hyndman et al., 2006).The expression of nNOS in the gill suggests that the enzyme mayregulate transport (Hyndman et al., 2006), the vasodilator tone

E. Brunelli et al. / acta histochemica 113 (2011) 201–213212

and the development and properties of neural function. Inaddition our observations concerning nNOS immunolocalizationin adjacent cells to CCs suggests that NO may act as a paracrinesignaling molecule to regulate systemic NaCl secretion (Tipsmarkand Madsen, 2003). The samples exposed to increasing concen-trations of Cd show a gradual inhibition of nNOS immunopositivecells and fibers, leading ultimately to a total absence ofimmunopositivity.

In contrast, a large number of VIP-immunopositive cells in theinterfilament and lamellar epithelium are present in fishesexposed to the highest concentration and duration. This may belinked to a possible role that the VIP develops in the control of thesmooth muscle cells of the blood vessel walls, including a possiblemodulation of ion transport across the gills (Ferrito et al., 2007).VIP caused relaxation in the muscle of Gadus morhua and inducedvasodilation of the swimbladder blood vessels (Schwerte et al.,1999). Also the distribution pattern of VIP in the gill plays arole in the regulation and/or modulation of secretory activities(De Girolamo et al., 1998).

Serotonin (5-HT) is an evolutionarily conserved neurotrans-mitter found in both invertebrates and vertebrates and is involvedin different physiological and behavioral roles. In the gills it has arole in regulating blood flow and in the turnover of the cell.Serotonin is located as in T. pavo control specimens in NECs. Thefrequency and the distribution pattern of these serotonin-immunopositive elements varied according to the concentrationsand duration of exposure to cadmium. As seen in differentorganisms exposed to the effects of heavy metals, the distributionof 5-HT may be involved in adaptive processes of gill epitheliumof fishes, leading to increased proliferation and apoptosis withmorphological alteration of gills (Ferrando et al., 2005; Mauceriet al., 2005).

The intracellular defense mechanisms against metals may alsodiffer in different teleost species. The MT immunolocalization wassignificantly induced and it showed time dependence aftercadmium exposure in T. pavo. It appears that cadmium exposureplays a role in regulation of MTs in fishes, and probably incadmium detoxification and protection of cells against metal iontoxicity (Linde et al., 1999; Urena et al., 2007; Wu et al., 2009). TheMTs were located in the chloride cells, which are important forionic transport through the gill epithelium (Dang et al., 1999).

However, these results indicate that cadmium is notcompletely chelated in the gills inducing apoptotic signals, suchas caspase-3, which is commonly defined as an effector caspase(Migliarini et al., 2005). PCNA-positive cells showed a significantincrease in the gills of exposed specimens to various times and inparticular to low concentrations; at the same time an increaseof caspase-3 reactivity was observed. This indicates an inductionof cell repair and apoptosis, respectively. The greatest damage wasobserved following the most prolonged exposure, 192 h andhighest concentrations, where there are indications of significantdamage to respiratory epithelium. In fact, the PCNA andcaspase-3-immunopositive cells are strongly inhibited.

Our results indicate that the exposure of T. pavo to sublethalcadmium concentrations affects gill morphology and ultrastruc-ture. The intensity of the histopathological phenomena appearsclosely related to the pollutant concentration and exposure time.The presence of acidic mucus may represent a defense mechanismand adaptation to exposure to non-lethal concentrations ofcadmium.

Through our experimental study it was possible to show thatfollowing a 192 h of exposure and concentrations of 60 and120 mM to cadmium, there are numerous histological changes inthe respiratory epithelium and an inhibition of adaptive processes,while the production of chelating molecules such as metallothio-neins is increased.

The results showed that the degree of gill distortion is propor-tional to the length of exposure and concentration of thecadmium. These data may be used for planning a modelto determine biological risk in the marine environment and maybe particularly useful to investigate organisms exposed tocadmium.

References

Alvarado NE, Cancio I, Hilland K, Marig �omez I, Soto M. Immunolocalization ofmetallothioneins in different tissues of turbot (Scophthalmus maximus)exposed to Cd. Histol Histopathol 2007;22:719–28.

American Society for Testing Material (ASTM). Standard practice for conductingacute toxicity tests with fishes, macroinvertebrates, and amphibians. E 729-790. Philadelphia: American Society for Testing and Materials Standards, 1997.

Au DWT. The application of histo-cytopathological biomarkers in marine pollutionmonitoring: a review. Mar Poll Bull 2004;42:817–34.

Azmitia EC. Modern views on an ancient chemical: serotonin effects on cellproliferation, maturation and apoptosis. Brain Res Bull 2001;56:413–24.

Berntssen MHG, Aspholm OO, Hylland K, Wendelaar Bonga SE, Lundebye AK.Tissue metallothionein, apoptosis and cell proliferation responses in Atlanticsalmon (Salmo salar L.) parr fed elevated dietary cadmium. Comp BiochemPhysiol 2001;128:299–310.

Castan ~o A, Carbonell G, Carballo M, Fernandez C, Boleas S, Tarazona JV. Sublethaleffects of repeated intraperitoneal cadmium injections on rainbow trout(Oncorhynchus mykiss). Ecotoxicol Environ Saf 1998;41:29–35.

De Smet H, Blust R. Stress responses and changes in protein metabolism in carpCyprinus carpio during cadmium exposure. Ecotoxicol Environ Saf 2001;48:255–62.

Dang Z, Lock RAC, Flik G, Wendelaar Bonga SE. Metallothionein response in gills ofOreochromis mossambicus exposed to copper in fresh water. Am J Physiol1999;277:320–31.

De Girolamo P, Arcamone N, Rosica A, Gargiulo G. PACAP (pituitary adenylatecyclase-activating peptide )-like immunoreativity in the gill arch of thegoldfish, Carassius auratus: distribution and comparison with VIP. Cell TissueRes 1998;293:567–71.

Ebbesson LO, Tipsmark CK, Holmqvist B, Nilsen T, Anderson E, Stefansson SO, et al.Nitric oxide synthase in the gill of Atlantic salmon: colocalization with andinhibition of Na+, K+-ATPase. J Exp Biol 2005;208:1011–7.

EIFAC. Working party on toxic testing procedures, revised report on fish toxicitytesting procedures. EIFAC Technical Paper, vol. 24 1983. p. 1–37.

European Communities Council. NIH Guide for the Care and Use of LaboratoryAnimals Directive of November 24 (86/609/EEC). 1986 (NIH publication no. 80-22, Revised 1987).

EPA. LC50 Software Program, Version 1.00. Center for Exposure AssessmentModeling (CEAM) Distribution Center, 1999.

EPA. Methods for measuring the acute toxicity of effluents and receiving waters tofreshwater and marine organisms, fifth ed. Office of Water, Washington, DC,2002; EPA-821-R-02-012.

Evans DH. The fish gill: site of action and model for toxic effects of environmentalpollutants. Environ Health Perspect 1987;71:47–58.

Evans DH, Piermarini PM, Choe KP. The multifunctional fish gill: dominant site ofgas exchange, osmoregulation, acid–base regulation, and excretion of nitro-genous waste. Physiol Rev 2005;85:97–177.

FAO. Manual de metodos de investigation delmedio ambiente acuatico. Parte 4:bases para la eleccion de ensayos biologicos para evaular la contaminacionmarina. FAO, Documentos Tecnicos de Pesca 1981;164:1–34.

Fasulo S, Mauceri A, Giannetto A, Maisano M, Bianchi N, Parrino V. Expression ofmetallothionein mRNAs by in situ hybridization in the gills of Mytilusgalloprovincialis, from natural polluted environments. Aquat Toxicol 2008;88:62–8.

Ferrando S, Ferrando T, Girosi L, Mauceri A, Fasulo S, Tagliafierro G. Apoptosis, cellproliferation and serotonin immunoreactivity in gut of Liza aurata from naturalheavy metal polluted environments: preliminary observations. Eur J Histo-chem 2005;49:331–40.

Ferrito V, Mauceri A, Minniti F, Isaja M, Maisano M, Tigano C. Comparativemorphological studies of the neurocranium and the gills of two species ofblennies living in different habitats. Acta Histochem 2007;109:428–36.

Finney DJ. In: Probit analysis. New York: Cambridge University Press; 1971 p. 337.Garcia-Santos S, Fontaınhas-Fernandes A, Wilson JM. Cadmium tolerance in the

Nile Tilapia (Oreochromis niloticus) following acute exposure: assessment ofsome ionoregulatory parameters. Environ Toxicol 2006;21:33–46.

Grimley PM. A tribasic stain for thin sections of plastic-embedded OsO4-fixedtissues. Stain Techn 1964;39:229–33.

Hans DS, Bart DW, Richard L. Dynamics of (Cd, Zn)-metallothioneins in gills, liverand kidney of common carp Cyprinus carpio during cadmium exposure. AquatToxicol 2001;52:269–81.

Heath AG. In: Water pollution and fish physiology. Boca Raton, FL: CRC Press; 1987.Hyndman KA, Choe KP, Havird JC, Rose RE, Piermarini PM, Evans DH. Neuronal

nitric oxide synthase in the gill of the killifish, Fundulus heteroclitus. CompBiochem Physiol 2006;144:510–9.

Jonz M, Fearon IM, Nurse CA. Neuroepithelial oxygen chemoreceptors of theZebrafish gill. J Physiol 2004;560:737–52.

E. Brunelli et al. / acta histochemica 113 (2011) 201–213 213

Langston WJ, Chesman BS, Burt GR, Pope ND, McEvoy J. Metallothionein in liver ofeels Anguilla anguilla from the Thames Estuary: an indicator of environmentalquality?. Mar Environ Res 2002;53:263–93.

Linde AR, Sanchez-Galan S, Klein D, Garcia-Vazquez E, Summer KH, et al.Metallothionein and heavy metals in brown trout (Salmo trutta) and Europeaneel (Anguilla anguilla): a comparative study. Ecotoxicol Environ Saf 1999;44:168–73.

Mance G. Pollution threat of heavy metals in aquatic environments. Pollutionmonitoring series. Amsterdam/New York: Elsevier Applied Science; 1987.

Martinez CBR, Nagae MY, Zaia CTB, Zaia DAM. Acute morphological andphysiological effects of lead in the neotropical fish Prochilodus lineatus. Braz JBiol 2004;64:797–807.

McDonald DG, Wood CM. Branchial mechanisms of acclimation to metals offreshwater fish. In: Rankin JC, Jensen FB, editors. Fish ecophysiology. London:Chapman &Hall; 1993. p. 297–321.

Mauceri A, Fasulo S, Ainis L, Licata A, Lauriano ER, Martinez A, et al. Neuronal nitricoxide synthase (nNOS) expression in the epithelial neuroendocrine cell systemand nerve fibers in the gill of the catfish, Heteropneustes fossilis. ActaHistochem 1999;101:437–48.

Mauceri A, Fossi MC, Leonzio C, Ancora S, Minniti F, Maisano M, et al. Stress factorsin the gills of Liza aurata (Perciformes, Mugilidae) living in pollutedenvironments. Ital J Zool 2005;72:285–92.

Migliarini B, Campisi AM, Maradona F, Truzzi C, Annibaldi A, Scarponi G, et al.Effects of cadmium exposure on testis apoptosis in the marine teleost Gobiusniger. Gen Comp Endocrinol 2005;142:241–7.

Moiseenko TI, Kudryavtseva LP. Trace metal accumulation and fish pathologies inareas affected by mining and metallurgical enterprises in the Koal Region.Russia Environ Pollut 2001;114:285–97.

Murty AS. Toxicity of pesticides to fish, vol. II. Boca Raton, FL: CRC Press; 1986.Nation JL. A new method using hexamethyl-disilazane for preparation of soft

insect tissues for scanning electron microscopy. Stain Technol 1983;58:347–51.Niyogi S, Wood CM. Interaction between dietary calcium supplementation and

chronic waterborne zinc exposure in juvenile rainbow trout, Oncorhynchusmykiss. Comp Biochem Physiol 2006;143:94–102.

Norey CG, Cryer A, Kay J. A comparison of cadmium induced metallothioneins geneexpression and Me2 + distribution in the tissues of cadmium-sensitive (rain-bow trout; Salmo gairdneri) and tolerant (stone loach; Noemacheilus barbatulus)species of freshwater fish. Comp Biochem Physiol 1990;97C:221–5.

OECD. Chemical Group and Management Committee: Harmonization of Classifica-tion Systems for Chemicals. In: 18th joint meeting; ENV/MC/CHEM, 1992, vol.92, p. 6.

Olojo EAA, Olurin KB, Mbaka G, Oluwemimo AD. Histopathology of the gill andliver tissues of the African catfish Clarias gariepinus exposed to lead. Afr JBiotechnol 2005;4:117–22.

Olsson P, Kling P, Kogstrand C. Mechanism of heavy metal accumulation andtoxicity in fish. In: Langston WJ, Bebianno MJ, editors. Metal metabolism inaquatic environments. London, UK: Chapman & Hall; 1998. p. 321–50.

Patrap HB, Wendelaar Bonga SE. Effect of ambient dietary cadmium on pavementcells, chloride cells, and Na+/K+-ATPase activity in the gills of the freshwaterteleost Oreochromis mossambicus at normal and high calcium levels in theambient water. Aquat Toxicol 1993;26:133–50.

Randi AS, Monserrat JM, Rodriguetz EM, Romano LA. Histological effects ofcadmium on the gills of the freshwater fish, Macropsobrycon uruguayanaeEigenmann (Pisces, Atherinidae). J Fish Dis 1996;19:311–22.

Robohm RA. Paradoxical effects of cadmium on antibacterial antibody responses intwo fish species: inhibition in cunners (Tautogolabrus adspersus) andenhancement in striped bass (Morone saxatilis). Veterinary Immunology andImmunopathology 1986;12:251–62.

Sadik M. Cadmium in toxic metal chemistry in marine environments. In: Toxicmetal chemistry in marine environments, 1st ed. M. Dekker Inc., King FahdUniversity of Petroleum and Minerals Dharan, Saudi Arabia Press, New York;1992. p. 130–2.

Sanchez JG, Speare DJ, Johnson GJ. Morphometric and histochemical assessment ofthe branchial tissue response of rainbow trout, Oncorhynchus mykiss(Walbbaum), associated with chloramines-T treatment. J Fish Dis 1997;20:375–81.

Schwerte T, Holmgren S, Pelster B. Vasodilatation of swimbladder vessels in theEuropean eel (Anguilla anguilla) induced by vasoactive intestinal polypeptide,nitric oxide, adenosine and protons. J Exp Biol 1999;202:1005–13.

Tao S, Liu C, Dawson R, Aiming L, Xu F. Uptake of cadmium adsorbed onparticulates by gills of goldfish (Carassius auratus). Ecotoxicol Environ Saf2000;47:306–13.

Tipsmark CK, Madsen SS. Regulation of Na+/K+-ATPase activity by nitric oxide inthe kidney and gill of the brown trout (Salmo trutta). J Exp Biol2003;206:1503–10.

Urena R, Peri S, del Ramo J, Torreblanca A. Metal and metallothionein content intissue from wild and farmed Anguilla anguilla at commercial size. Environ Int2007;33:532–9.

Wangsongsak A, Utarnpongsa S, Kruatrachue M, Ponglikitmongko M, Pokethi-tiyook P, Sumranwanich T. Alterations of organ histopathology and metal-lothionein mRNA expression in silver barb, Puntius gonionotus duringsubchronic cadmium exposure. J Environ Sci 2007;19:1341–8.

Wendelaar Bonga SE, Flik G, Balm PHM, van der Meij JCA. The ultrastructure ofchloride cells in the gills of the teleost Oreochromis mossambicus duringexposure to fresh water. Cell Tissue Res 1990;259:575–85.

Wilson JM, Laurent P. Fish gill morphology: inside out. J Exp Zool 2002;293:192–213.

Wong CKC, Wong MH. Morphological and biochemical changes in the gills ofTilapia (Oreochromis mossambicus) to ambient cadmium exposure. AquatToxicol 2000;48:517–27.

Wu JP, Chen HC, Huang DJ. Histopathological alterations in gills of white shrimpLitopenaeus vannamei (Boone) after exposure to cadmium and zinc. BullEnviron Contam Toxicol 2009;82:90–5.

Wu SM, Jong KJ, Lee YJ. Relationships among metallothionein, cadmiumaccumulation, and cadmium tolerance in three species of fish. Bull EnvironContam Toxicol 2006;76:595–600.

Wu SM, Shih MJ, Ho YC. Toxicological stress response and cadmium distribution inhybrid tilapia (Oreochromis sp.) upon cadmium exposure. Comp BiochemPhysiol C Toxicol Pharmacol 2007;145:218–26.

Zaccone G, Ainis L, Mauceri A, Lo Cascio P, Lo Giudice F, Fasulo S. NANC nerves inthe respiratory air sac and branchial vasculature of the Indian catfishHeteropneustes fossilis. Acta Histochem 2003;105:151–63.

Further reading

Bro-Rasmussen F, Calow P, Canton JH, Chambers PL, Silva Fernandes A, Hoffmann L,et al. water quality objectives for chemicals dangerous to aquatic environment(List 1). Rev Environ Contam Toxicol 1994;137:83–110.

Page AL, El-Anamy MM, Cheng AC. Cadmium in the environment and its entry intoterrestrial food chain crops. In: Foulkes EC, editor. Cadmium. Berlin: Springer-Verlag; 1986.