Effects of a sublethal concentration of sodium lauryl sulphate on the morphology and Na+/K+ ATPase activity in the gill of the ornate wrasse (Thalassoma pavo)

This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

Author's personal copy

Ecotoxicology and Environmental Safety 71 (2008) 436–445

Effects of a sublethal concentration of sodium lauryl sulphateon the morphology and Na+/K+ ATPase activity in the gill

of the ornate wrasse (Thalassoma pavo)

Elvira Brunelli�, Erminia Talarico, Barbara Corapi, Ida Perrotta, Sandro Tripepi

Dipartimento di Ecologia, Universita della Calabria Cubo 4B, Via P. Bucci 87036 Rende, CS, Italy

Received 16 June 2006; received in revised form 6 September 2007; accepted 21 September 2007

Available online 19 November 2007

Abstract

We analysed the morphology and ultrastructure of the gill apparatus of the ornate wrasse, Thalassoma pavo, under normal conditions

and after exposure to a sublethal concentration of sodium lauryl sulphate (3.5mg/l, which is one-third of the 96LC99 value). To identify

the biochemical mechanisms affected by this pollutant, we evaluated and compared the localisation of Na+/K+ ATPase in normal and

experimental conditions. Immunocytochemical analysis revealed that this enzyme was active in the chloride cells (CCs), which were

distributed in clusters in the interlamellar region of the filament. Ultrastructural analysis revealed conspicuous alterations on the

epithelium after 96 and 192 h of exposure to sodium lauryl sulphate: structural features of the surface cells were lost, the appearance of

intercellular lacunae changed, and cellular degeneration occurred. Statistical analysis comparing the number and dimensions of CCs in

normal conditions and after 96 h of exposure showed that the CC area decreased after exposure to the detergent.

r 2007 Elsevier Inc. All rights reserved.

Keywords: Gill; Ultrastructure; Electron microscopy; Sodium lauryl sulphate; Chloride cells; Na+/K+ ATPase; Confocal laser scanning microscopy;

Thalassoma pavo

1. Introduction

Some pollutants can induce alterations to the respiratoryapparatus in fish (Mallat, 1985; Van den Heuvel et al.,2000). The gill apparatus and the skin, which arecontinuously exposed to the aquatic medium, are theprincipal sites of activity of numerous toxic substances(Van den Heuvel et al., 2000). Anionic detergents representan important category of pollutants and are still widelyused today in the production of many consumer goodsbecause of their surface-active properties. In particular,sodium lauryl sulphate (SLS) is used in many personal careproducts and cosmetics. These products are disposed offthrough domestic and industrial waste discharges into thesea where they accumulate in seawater and sediments(Sigoillot and Nguyen, 1992). For example, �60,000 tonsof detergents are poured into the Mediterranean each year

(Della Croce et al., 2001); although biodegradation of SLSranged from 45% to 95% within 24 h, the continuousintroduction of SLS into the environment kept theconcentration of this pollutant high (Cserhati et al.,2002). A number of researchers have studied the noxiouseffects of anionic detergents on marine organisms (e.g.,Misra et al., 1985, 1991; Zaccone et al., 1985a, b, 1986;Rosas et al., 1988; Roy 1988a, b; Gupta et al., 1989;Ribelles et al., 1995). Such studies highlight the importanceof these substances in determining mortality or pathologyin marine populations. Anionic detergents have a strongtendency to bind to the lipid component of the membrane,and thus high concentrations can alter the structuralarrangement of the membrane and compromise itsfunctionality.In fish, gills play a fundamental role in gas exchange, and

they also represent an important site for osmotic andacid–base regulation (Laurent, 1984; Evans et al., 1999).The structural complexity of the gill apparatus reflectsthese multiple functions (Franchini et al., 1994; Ribelles

ARTICLE IN PRESS

www.elsevier.com/locate/ecoenv

0147-6513/$ - see front matter r 2007 Elsevier Inc. All rights reserved.

doi:10.1016/j.ecoenv.2007.09.010

�Corresponding author. Fax: +39 0984492986.

E-mail address: [email protected] (E. Brunelli).

Author's personal copy

et al., 1995). Pollutants can affect the gill apparatus and itsrole as a respiratory and osmoregulatory organ (Mallat,1985). Following chronic or acute exposure to differentpollutants, alterations in gill tissue can occur. Thesechanges represent a response to stressors that have oftenbeen interpreted as non-specific (Mallat, 1985; Evans,1987); in many cases, alterations at the cellular or sub-cellular level are not by themselves diagnostic of aparticular type of pollutant. Thus, a combination ofmorphological analysis, the evaluation of functionalactivity, and the application of statistical methods mightbe a better approach to determining a specific response to aparticular pollutant (Mallat, 1985).

In this paper, we analysed the morphology and ultra-structure of the gill epithelium of the ornate wrasse,Thalassoma pavo, under normal conditions and afterexposure to sublethal concentrations of SLS. Our goalwas to evaluate the histopathological and sub-cellularalterations induced by exposure to SLS. Furthermore, weused a confocal laser scanning microscope for Na+/K+

ATPase immunolocalisation (Witters et al., 1996; Dang etal., 2000) to determine where chloride cells (CCs) reside.We then compared the CC population in the control andtreated samples.

2. Materials and methods

2.1. Fish maintenance and holding conditions

The T. pavo specimens used in this study were collected from a location

on the Tyrrhenian coast (S. Lucido) using baited traps. Adult specimens of

both sexes with a mean mass of 9.6970.48 g were transported to the

laboratory and kept in a 150 l aquaria (10 fishes per tank) filled with

seawater from the capture site and equipped with filter and oxygenation

systems. The fishes were kept in the aquaria for a 5-day acclimatisation

period, during which salinity (35%), density (1.027–1.028 g/cm3), tem-

perature (18–24 1C), and the nitrite and nitrate concentrations were

measured and kept constant. Other conditions included dissolved oxygen

at 8.0–8.6mg/l, hardness 100mg CO3Ca/l, and the absence of heavy

metals. Throughout the experiment, the animals were maintained under a

natural light/dark cycle and fed every 2 days with commercial fish food

(Tetramin).

2.2. Exposure to SLS and sampling

The experiments were carried out using a semi-static acute experimental

method, meaning that the experimental solution and the samples (i.e., fish)

were put in a test chamber (i.e., aquarium). To avoid variations in

detergent concentrations, solutions were renewed every 24 h (EPA, 2002).

From some preliminary evaluations (not shown) biodegradation levels

result to be less than 12% of the initial concentration which is very similar

to that obtained in other similar studies (Flores et al., 1980; Ribelles et al.,

1995).

The toxicant used in this study was SLS (CH3(CH2)11OSO3Na; Sigma,

Milano, Italy). The fish were randomly distributed in different concentra-

tions of SLS (2.5, 3, 3.5, 4, 4.5, 5, 5.5, 6, 6.5, and 7mg/l). For the acute

bioassay tests, 15 fishes were used per concentration per replicate. A total

of four replicates was used for each dose and for the control group. The

number of dead fishes was counted every 12 h, and they were removed

immediately from the aquaria.

In this study, the acute toxic effect of SLS on T. pavo was determined

using Finney’s Probit Analysis LC50 Determination Method (Finney,

1971). The computer analysis was performed using LC50 1.00 software

developed by the EPA (1999).

The data were also evaluated following the Behrens–Karber method

using the following formula (Klassen, 1991; Yilmaz et al., 2004):

LC50 ¼ LC100�abþ . . .þ ab

n,

where LC50 and LC100 indicate the lethal doses for 50% and 100% of the

samples, respectively; a is the difference between the two consecutive

doses; b is the arithmetic mean of the mortality caused by two consecutive

doses; and n is the number of samples in each group.

After identifying the lethal concentration, we identified the maximum

concentration in which all the subjects showed a normal swimming

position and normal feeding behaviour, and we carried out two sets of

experiments with a sublethal dose (SLS nominal concentration 3.5mg/l).

The gills were removed after 48, 96, and 192h and control sample gills

were removed at the same time. The animals were anaesthetised with

2–4 g/l tricaine methane sulphonate (MS 222, Sandoz, Sigma, St. Louis,

MO) and killed by spinal cord transection. The removed gills were

prepared for scanning electron microscopy (SEM) and transmission

electron microscopy (TEM) and for confocal microscopy. Animal

manipulation was performed according to Ethical Committee recommen-

dations and under the supervision of authorised investigators.

2.3. Electron microscopy

After removal, the gill samples were fixed using Karnovsky liquid in a

cacodylate buffer (pH 7.4). After post-fixation with osmium tetroxide and

dehydration in ethanol at increasing concentrations, the samples used for

TEM observation were enclosed in epon/araldite. Semi-thin sections

(1–2mm) were stained with Grimley’s dyes (toluidine blue method:

malachite green and acid fuchsine) and observed with a Leitz Dialux 20

EB light microscope. The ultrathin sections were observed under a Zeiss

EM 900 TEM. After dehydration, the samples used for SEM observation

were subjected to the progressive substitution of ethanol with hexam-

ethyldisilazane, removed by complete evaporation (Nation, 1983), coated

with gold in an Emitech K550 ion sputter unit, and then examined under a

Zeiss DSM 940 SEM.

2.4. Immunohistochemistry

The samples were immersed in Bouin liquid, in phosphate-buffered

saline (PBS, pH 7.1) for 24 h, and then in graded ethanol; after

clarification in xylene, they were embedded in paraffin wax with a mean

fusion point of 56 1C. The 10-mm sections were dewaxed and subjected to

the indirect immunofluorescence technique (Coons et al., 1955). To block

endogenous peroxidase, the sections were immersed in 2% H2O2 in 0.1

PBS for 20min and then incubated for 30min in a moist chamber with

20% normal goat serum to block non-specific sites. Unwashed sections

were then incubated overnight at 4 1C with a mouse monoclonal antibody

to Na+/K+ ATPase at working dilutions of 1:100 (IgGa5; Developmental

Studies Hybridoma Bank, Department of Biological Sciences, The

University of Iowa, Iowa City, IA, USA). After several washes in PBS,

fluorescein–isothiocyanate-conjugated g-globulin goat anti-mouse (Sigma)

was used as the second antiserum at a dilution of 1:200 for 1 h at room

temperature. To verify the specificity of the immunolabelling, the primary

antiserum was substituted with non-immune goat serum or with PBS. In

addition, propidium iodide was used for visualisation of the general tissue

structure (at working dilutions of 1:200); this phenanthrene colourant,

when inserted in the DNA, forms a sufficiently stable complex that emits a

red fluorescence when excited. The observations were carried out with a

Leica TCS SP2 Confocal Laser Scanning Microscope.

2.5. Quantification and statistical analysis

The CCs in the filament and lamellae were identified by Na+/K+

ATPase immunohistochemistry (according to Dang et al., 2000) and then

ARTICLE IN PRESSE. Brunelli et al. / Ecotoxicology and Environmental Safety 71 (2008) 436–445 437

Author's personal copy

were quantified using an image analysis program (NIH, developed at the

National Institutes of Health, a part of the US Department of Health and

Human Services). For each animal, 10mm thick parasagittal sections were

made and placed on glass slides. For each section, the cells belonging to

625� 625 pixel fields (500� 500mm) were counted and measured. Two to

three fields were analysed for each slide. The analysed sections came from

20 animals (10 control and 10 treated, randomly selected). The results are

expressed as the number of cells per mm2 and as an area in mm2. The data

are presented as mean7S.E.M. The differences between the two groups

were tested using the InStat (GraphPad Software, Inc., San Diego, CA,

USA) statistical program; we applied the Student’s two-tailed t-test for

unpaired data, where appropriate, with Po0.05 as the fiducial limit.

3. Results

3.1. Acute toxicity

Table 1 shows the relationship between the SLSconcentration and the mortality rate of T. pavo accordingto Finney’s Probit Analysis using the EPA computerprogram. Estimated LC50 values and confidence limitsdetermined in semi-static tests for 96 h are presented inTable 2.

We also evaluated the data using the Behrens–Karbermethod (Klassen, 1991; Yilmaz et al., 2004), and the LC50value was estimated to be 5.32mg/l. The two methods arein good agreement.

3.2. Morphological and ultrastructural observations

3.2.1. Controls (group A)

The T. pavo gill apparatus macroscopic organisationshowed structural characteristics typical for Teleostei. Eachgill was supported by four gill arches, creating an insertionpoint for a double series of filaments (Fig. 1a); from thefilaments, the lamellae rose perpendicularly from both sides(Fig. 1a and b).

The filaments were covered with a multilayered epithe-lium (primary epithelium) and the lamellae were covered bya secondary epithelium. In the primary epithelium (Fig.1c), four cell types could be identified: basal cells, mucouscells, CCs along with accessory cells, and pavement cells(PVCs). Under SEM (Fig. 1d), we recognised these

polygon-shaped PVCs, which are characterised by acomplicated but regularly distributed system of superficialconcentric microridges; the margins where these cells joinwere well defined, and we could see the opening of theunderlying cells on the surface. Mucous cells were situatedbeneath the PVCs, in particular along the filament margin(Fig. 1e), and generally were covered by the adjacent PVCedge, which reduced the portion that was directly incontact with the external medium. These round-shapedcells were characterised by the presence of large electron-clear granules that occupied the entire cytoplasm. The CCs(Fig. 1f) (also called mitochondria-rich cells) were oftendistributed in clusters in the filament interlamellar regionand were characterised by the presence of accessory cellswith which they interdigitated, particularly in the apicalportion. CCs showed a typical apical crypt and a cytoplasmfilled by numerous mitochondria associated with a com-plex, ramified tubular system. The innermost layer wasmade up of non-differentiated basal cells that were incontact with the underlying connective tissue (Fig. 1e).The organisation of the lamellae was very simple; an

external layer of PVCs lay on the inner layer ofundifferentiated basal cells (Fig. 1(g)). In section, the pillarcells controlling the blood flow were visible.

ARTICLE IN PRESS

Table 1

The relationship between the SLS concentration and the mortality rate of Thalassoma pavo

Concentration (mg/l) No. of exposed fish No. of dead fish Observed proportion

responding

Proportion responding

adjusted for controls

Predicted proportion

responding

2.5 15 0 0.0000 0.0000 0.0002

3.0 15 0 0.0000 0.0000 0.0037

3.5 15 1 0.0667 0.0667 0.0266

4 15 1 0.0667 0.0667 0.0983

4.5 15 3 0.2000 0.2000 0.2344

5 15 8 0.5333 0.5333 0.4138

5.5 15 8 0.5333 0.5333 0.5951

6 15 9 0.6000 0.6000 0.7451

6.5 15 13 0.8667 0.8667 0.8518

7 15 15 1.0000 1.0000 0.9193

Table 2

Estimated LC values and confidence limits determined with Finney’s

Probit Analysis LC50 (EPA, 2002)

Point Concentration (mg/l) 95% confidence limits

Lower Upper

LC 1.00 3.225 2.613 3.637

LC 5.00 3.716 3.179 4.075

LC 10.00 4.008 3.525 4.335

LC 15.00 4.217 3.775 4.525

LC 50.00 5.232 4.934 5.544

LC 85.00 6.490 6.053 7.239

LC 90.00 6.829 6.318 7.753

LC 95.00 7.365 6.722 8.595

LC 99.00 8.486 7.531 10.455

E. Brunelli et al. / Ecotoxicology and Environmental Safety 71 (2008) 436–445438

Author's personal copy

3.2.2. SLS group: 3.5 mg/l, 96 h (group B)

Exposure to SLS induced changes in the gill apparatus.Under SEM, structural modifications were evident. Atseveral points (Fig. 2a), lamellar fusion occurred; this wasparticularly evident in the distal portion and along themargins, and it led to occlusion of the interlamellar space.

The filament has a fairly irregular appearance and excessmucus was present.When we observed the filament surface at greater

magnification after removing the mucus (Fig. 2b), wenoted that at several points the regular microridgearrangement characteristic of PVCs was missing. Light

ARTICLE IN PRESS

Fig. 1. Thalassoma pavo CTRL. (a) Scanning electron micrograph of T. pavo gill arch (ga) under control conditions. Observe some gill filaments (pf) from

which arise the lamellae (sl). (b) Magnification of a filament (pf). Observe a series of lamellae (sl) interspaced in a very regular manner. (c) A low

magnification light micrograph of a section through a gill filament. The filament is covered with a multilayered epithelium (primary epithelium), which

extends in the space between the lamellae. Note the chloride cell (cc) and the mucous cell (gc). (d) Scanning electron micrograph of the primary epithelium

surface. Note the complicated system of concentric microridges on the pavement cell surface (pvc). (e) Electron micrographs of primary epithelium

showing pavement cells (pvc) with numerous microridges and a round-shaped mucous cell (gc); note that the entire cytoplasm is filled by large electron-

clear granules. (f) Detail of two chloride cells (cc) accompanied by an accessory cell (ac). The CCs show a typical apical crypt and a cytoplasm filled by

numerous mitochondria associated with a complex and ramified tubular system (pvc pavement cell). (g) Transmission electron micrograph showing a

transverse section of a gill lamella under control conditions; the bilayered epithelium is formed by pavement cells and basal cells. Note the pillar cell

interposed among blood vessels.

E. Brunelli et al. / Ecotoxicology and Environmental Safety 71 (2008) 436–445 439

Author's personal copy

microscope analysis (Fig. 2c) revealed the presence of someaneurysms, which were completely filled with erythrocytes;in section, it showed numerous fusion zones betweenadjacent lamellae and the appearance of wide spaces andintercellular lacunae.

High-magnification light microscope micrographs (Fig. 2d)showed that the fusion zone between adjacent lamellae

house many mainly CCs which were also present on thesurface of some lamellae. In some cases, these cells werelengthened and flattened, whereas in others they main-tained typical CC conformation. In these CCs, ultrastruc-tural organisation was maintained and it was possibleto recognise numerous mitochondria and sections ofthe complex tubular system. In no cases were the CCs

ARTICLE IN PRESS

Fig. 2. Thalassoma pavo after 96 h SLS exposure. (a) Scanning electron micrograph of T. pavo lamellae after 96 h SLS exposure. Note the irregular

arrangement and fusion of lamellae, which is particularly evident in the distal portion. (b) Detail of pavement cells at the epithelial surface. In some cases

the microridges disappear (*). (c, d) Light microscope micrographs of a cross-section through a filament. Note the fusion at the edge of lamellae

(arrowhead), the appearance of chloride cells (arrows), and the presence of aneurysms (*). (e) Transmission electron micrograph of fusion between two

lamellae after 96 h SLS exposure (cc—chloride cell). (f) Transmission electron micrograph showing the thickening of epithelium and the presence of a

mucous cell in the lamella (pvc—pavement cell).

E. Brunelli et al. / Ecotoxicology and Environmental Safety 71 (2008) 436–445440

Author's personal copy

accompanied by accessory cells. The CCs on the filament(Fig. 2e) observed through the TEM often appearedto have lost the apical crypt, and cytoplasmic lacunaeappeared that altered the typical tubular vesicular systemdistribution. Mucous cells, although fewer in number, wereidentified on the surface of the lamellae (Fig. 2f). Thesecells had a normal morphology and maintained theirtypical round shape; inside, typical electron-clear granuleswere distinguishable. The mucous cells on the primaryepithelium did not show signs of alteration. The PVCskept their ultrastructural characteristics in many cases,but in others we recognised degenerating or apoptoticcells.

3.2.3. SLS group: 3.5 mg/l, 192 h (group C)

The SEM observations of the gill epithelium after 192 hof exposure to SLS showed an accentuated disruptivephenomenon: the epithelium lost its structural arrangementat several points and the epithelium was lifted, giving thesurface a wrinkled, non-homogenous appearance that wasparticularly evident in the lamellae (Fig. 3a). In other cases,PVC groups with degenerated surface microridges werevisible (Fig. 3b). This phenomenon was particularlycommon in the interlamellar area, which, after 8 days ofexposure, appeared fused in different zones (Fig. 3c). Themost conspicuous cell alterations occurred in the primaryepithelium CCs, which, in some cases, underwent a

ARTICLE IN PRESS

Fig. 3. Thalassoma pavo after 192 h of SLS exposure. (a–c) Scanning electron micrographs of gill epithelia after 192 h of SLS exposure. Note the

alterations of epithelium surface, which consist of the wrinkling of the pavement cells (a), loss of the microridges (b), and fusion of the lamellae (c). (d)

Light microscope micrograph of a cross-section through the primary epithelium. Note regular (cc) and degenerating chloride cells (*). (e, f) Transmission

electron micrographs of primary epithelium. After 192 h of SLS exposure, several degenerating cells were observed. In the mucous cells (gc), mucous

granules became irregular in shape and varied in electron density (e), and in the chloride cells, cytoplasm and mitochondria lost their typical arrangement

(f) (pvc pavement cell).

E. Brunelli et al. / Ecotoxicology and Environmental Safety 71 (2008) 436–445 441

Author's personal copy

progressive degeneration (Fig. 3d and f). Some mucouscells showed normal ultrastructural characteristics,whereas others showed signs of alteration of the secretiongranules. In fact, we observed wide spaces and lacunaebetween the electron-clear granules that characterise thecytoplasm of this cell type; under normal conditions, thesegranules always appear strongly compacted (Fig. 3e). Inmany cases, a contraction of the surface cell edge, whichgenerally covers the mucous cells, separated them from theexternal medium.

3.3. Na+/K+ ATPase immunohistochemistry/confocal laser

scanning microscopy

In the control samples, the localisation of Na+/K+

ATPase (Fig. 4a) revealed immunopositive cells only in theprimary epithelium; that is, along the filament and in theinterlamellar region. These CCs had a very regulardistribution and formation.

The distribution of the Na+/K+ ATPase positive cells inthe 96 and 192 h groups (Fig. 4b and c) showed aproliferation of CCs and their appearance on the secondaryepithelium, particularly in the fusion regions between theneighbouring lamellae. The shape of the lamellae cellsranged from round to oval, and sometimes they wereparticularly long and slender. Statistical analysis of CCdensity on the filament (expressed as number of cells persquare mm) was limited to the 96 h group, and nosignificant differences were found between this group andthe control group. In contrast, the comparison between theCC area on the filament of the control samples and thefilament of the samples after 96 h exposure to SLS showedsignificant modifications (Fig. 5; Po0.0001). In fact,the CC area in the samples decreased after exposure tothe detergent. In the 96 h samples, we then compared theCC area on the filament to that on the lamellae and foundno significant variations. The comparison between the CCarea in control samples and that of the CCs appearing onthe lamellae in the 96 h samples was significant, and adecrease was once again noted in the area of the newlyformed cells (Po0.0001; Fig. 5).

4. Discussion

4.1. Acute toxicity

According to its LC50 values, SLS should be classified astoxic to fish (OECD, 1992). The results obtained in ourstudy are in the range of values reported in literature,although the acute toxicity of detergents varies, dependingon fish species and water quality (Abel, 1976; Roy, 1988a;Ribelles et al., 1995; Rosety-Rodrıguez et al., 2002).

4.2. Subacute toxicity

The results of our investigations demonstrated that theeffects of short-term exposure to a 96LC1concentration of

SLS on the gill apparatus of T. pavo are clearly visible;exposure resulted in the modification of morphology andcell composition (Table 3). The concept that fish areparticularly sensitive to the action of pollutants (Swedmarket al., 1971), probably due to their physiological andanatomical characteristics, is widely accepted and sup-ported by this study.Due to their chemical properties, anionic detergents tend

to influence gaseous exchange. Their action is particularly

ARTICLE IN PRESS

Fig. 4. Confocal micrographs of gill sections labelled with the mouse

monoclonal antibody IgGa5 raised against the a subunit of the Na+/K+

ATPase. Cell nuclei were labelled with propidium iodide. (a) Control

sample: immunolocalisations revealed positive cells (FITC conjugated)

distributed along the filament and in the interlamellar region. (b, c) Cross-

sections of a filament showing the presence of Na+/K+ ATPase positive

cells also on the lamellae after 96 h (b) and 192 h (c) of SLS exposure.

E. Brunelli et al. / Ecotoxicology and Environmental Safety 71 (2008) 436–445442

Author's personal copy

noxious for the gill apparatus of fish, which is coveredonly by a thin mucous layer and not by a cuticle, suchas that which protects crustacean gills. The mechanism oftoxic action of anionic detergents has been studiedextensively. According to some authors, mortality in fishfollowing a prolonged exposure to anionic detergents is dueto hypoxia or to the loss of osmotic or ionic stability (Lockand Van Overbeeke, 1981; Roy, 1988a; Wendelaar Bongaand Van der Meij, 1989; Franchini et al., 1994; Zacconeet al., 1985a). The primary target of SLS toxicity onmembrane structures has been reviewed (Singer andTjeerdema, 1993); membrane alterations seems to be aconsequence of interaction between the detergent and theopposite-charged membrane ions It has also been sug-gested that SDS causes lipid peroxidation, increasedglutathione production (Bindu and Babu, 2001), andchanges in carbon metabolism (Nickerson and Aspedon,1992).

This would cause cytolysis/autolysis or deactivation ofthe metabolic systems; it is therefore supposed that asimilar mechanism could act on fish gills.

The morpho-functional alterations that we observed inT. pavo after 96 and 192 h of exposure to SLS are slightlydifferent from those reported by Abel (1976) for thefreshwater teleost Salmo trutta exposed to the samedetergent. The different experimental conditions justifythe different results. Abel used a concentration of 18mg/las a minimum level, which is considerably different fromthe 3.5mg/l used in our study. At the concentration we

used, all animals fed normally and survived for the entiretreatment period. At 18mg/l, Abel reported an epitheliallifting phenomenon and interlamellar space occlusion.Survival time under these conditions was 45 h and, as hehimself stressed, this time-lapse was not long enough toallow epithelium reorganisation. Our observations showedthe capacity of the gill epithelium to respond to a stressorfrom the environment by modifying its cell composition(i.e., a compensatory mechanism).The occlusion of the interlamellar space through which

water flows and the appearance of wide intercellularlacunae reported by Abel for S. trutta, were also observedin T. pavo, but the extent of the phenomenon differed. Inour case, it was limited to the distal portion of some gilllamellae. Pillar cell alterations were also present in bothspecies.Van den Heuvel et al. (2000) reported similar effects in

Perca flavescens after exposure to oil sand mining-associated waters, and they suggested that pillar celldegeneration caused the appearance of aneurysms. Theyobserved that when gill lamellae structural integrity waslost, blood cells could fill the tissue lacunae. Thedegeneration observed with regard to some cell types suchas CCs and mucous cells after 192 h of exposure may beconsidered a localised phenomenon; degeneration will leadto a compensatory response resulting in stimulateddifferentiation of new cells. Both cell types (CC and MC)are present on the lamellae after 96 and 192 h of exposureto SLS.

ARTICLE IN PRESS

Fig. 5. Comparison between the chloride cell area present on the filament of the control and the chloride cells present after 96 h exposure to SLS. (chloride

cell area of control samples versus area of chloride cells of filament in treated samples, Po0.0001; chloride cell area on the filaments of controls and on the

lamellae of SLS-exposed fish, Po0.0001).

Table 3

Cellular composition of Thalassoma pavo gill epithelium under control conditions and after 96–192 h exposure to SLS

Cell types Primary epithelium control

group

Secondary epithelium control

group

Primary epithelium SLS group

96–192 h

Secondary epithelium SLS group

96–192 h

Pavement

cell

p p p p

Basal cell p p p pGoblet cell p p pChloride cell p p pAccessory

cell

p p

E. Brunelli et al. / Ecotoxicology and Environmental Safety 71 (2008) 436–445 443

Author's personal copy

4.3. Chloride cells

In the T. pavo gill epithelium, we identified only one typeof CC, which was recognisable as the aCC typical ofsaltwater teleosts. This cell always has an apical crypt andis accompanied by accessory cells. According to severalauthors (Pisam and Rambourg, 1991; Pisam et al., 1995;Evans et al., 1999), accessory cells are the precursors ofmature CCs and could therefore be responsible for thereported increase in CCs that occurred in response to anincrease in salinity, temperature variations, and variationsin food availability (Boyd et al., 1980; Karnaky, 1986). Inour experiments, all the animals were captured duringspring and the two groups (control and exposed to SLS)came from the same capture site. This excludes theinfluence of other parameters on the presence or distribu-tion of this cell type. The CCs we found on the T. pavo

lamellae after SLS exposure were always without accessorycells (Table 3) and without the typical apical crypt;therefore, we believe that these cells derived fromdifferentiating and migrating accessory cells.

Na+/K+ ATPase enzyme activity was present in all theCCs, including those situated on the lamellae, whichconfirmed the functional significance of the appearance ofthis cell type. Studies carried out on different teleost specieshave shown that water pollutants cause hypertrophy(Richman and Zaugg, 1987; Perry and Laurent, 1989;Goss et al., 1994; Dang et al., 2000). Dang hypothesisedthat the lamellar CCs originated from the basal cell layer orfrom the central epithelial region of the filament. Thefilament accessory cells could therefore have created theCCs that we found on the lamellar epithelium as a responseto the pollutants. No phenomena of increased size, such asthose observed by Van der Heijden et al. (1997) andcolleagues in Oreochromis mossambicus after acclimatisa-tion to saltwater, were found in our study.

4.4. Mucous cells

In Chionodraco hamatus and Trematomus bernacchii, twoAntarctic fish (Masini et al., 2000), the mucous cell isdistributed both on the filament and on the lamellae; in thelatter, however, they are distributed only on one side of thelamellae. The function of the mucus is probably to preventwater loss or ionic influx across the epithelium, but clearlythe presence of a continuous layer makes gaseousexchanges more difficult (Franklin, 1990; Shepherd,1994). Therefore, under normal conditions, the distributionof mucous cells, and, consequently, of the layer of mucussecreted by it, allows the water loss or the entrance of ionsto be controlled without limiting gaseous exchanges. If theappearance of intercellular spaces and the loss of structuralarrangement of the junctions cause an increased influx ofions along these paracellular paths, the presence of mucuscould be advantageous, despite the consequent reductionof the surface available for gaseous exchange. We suggestthat, in some cases, mucous cells appear only where

ultrastructural alterations are greater and where control ofpermeability is more necessary.

Acknowledgments

This work was supported by the MemoBioamarMURST Project. We thank Enrico Perrotta for technicalassistance.The authors declare that this study was conducted

according to national and institutional guidelines for theprotection of human subjects and animal welfare.

References

Abel, P.D., 1976. Toxic action of several lethal concentrations of an

anionic detergent on the gills of the brown trout (Salmo trutta L.).

J. Fish Biol. 9, 441–446.

Bindu, P.C., Babu, P., 2001. Surfactant-induced lipid peroxidation in a

tropical euryhaline teleost Oreochromis mossambicus (Tilapia) adapted

to fresh water. Indian J. Exp. Biol. 39 (11), 1118–1122.

Boyd, B., De Vries, A.L., Eastman, J.T., Pietra, G.G., 1980. The

secondary lamellae of the gills of cold water (high latitude) teleosts.

Cell Tissue Res. 213, 361–367.

Coons, A.H., Leduc, E.H., Connolly, J.M., 1955. Studies on antibody. I.

A method for the histochemical demonstration of specific antibody

and its application to a study of the hyperimmune rabbit. J. Exp. Med.

102, 49–59.

Cserhati, T., Forgacs, E., Oros, G., 2002. Biological activity and

environmental impact of anionic surfactants. Environ. Int. 28,

337–348.

Dang, Z., Lock, R.A.C., Flik, G., Wendelaar Bonga, S.E., 2000. Na+/

K+-ATPase immunoreactivity in branchial chloride cells of Oreochro-

mis mossambicus exposed to copper. J. Exp. Biol. 203, 379–387.

Della Croce, N., Cattaneo Vietti, R., Danovaro, R., 2001. Ecologia e

Protezione Dell’ambiente Marino Costiero. UTET, Torino.

EPA, 1999. LC50 Software Program, Version 1.00. Center for Exposure

Assessment Modeling (CEAM) Distribution Center.

EPA, 2002. Methods for Measuring the Acute Toxicity of Effluents and

Receiving Waters to Freshwater and Marine Organisms, fifth ed.

Office of Water, Washington, DC (EPA-821-R-02-012).

Evans, D.H., 1987. The fish gill: site of action and model for toxic effects

of environmental pollutants. Environ. Health Perspect. 71, 47–58.

Evans, D.H., Piermarini, P.M., Potts, W.T.W., 1999. Ionic transport in

the fish gill epithelium. J. Exp. Zool. 283, 641–652.

Finney, D.J., 1971. Probit Analysis. Cambridge University Press, New

York, p. 337.

Flores, V., Sales, D., Establier, R., 1980. Contaminacion de las aguas de la

Bahıa de Cadiz (IV). Ensayos de biodegradabilidad con dodecil-sulfato

sodico. Ing. Quim. 131, 81–90.

Franchini, A., Alessandrini, F., Bolognani Fantin, A.M., 1994. Gill

morphology and ATPase activity in the goldfish Carassius carassius

var auratus exposed to experimental lead intoxication. Ital. J. Zool. 61,

29–37.

Franklin, G.E., 1990. Surface ultrastructure changes in the gills of sockeye

salmon (Teleostei: Oncorhynchus nerka) during seawater transfer:

comparison of successful and unsuccessful seawater adaptation.

J. Morphol. 206, 13–23.

Goss, G.G., Laurent, P., Perry, S.F., 1994. Gill morphology during

hypercapnia in brown bullhead (Ictalurus nebulosus): role of chloride

cells and pavement cells in acid–base regulation. J. Fish Biol. 45,

705–718.

Gupta, B.N., Mathur, A.K., Agarwal, C., Singh, A., 1989. In vitro effect

of linear alkylbenzene sulphonate (LAS) on some enzymes in liver and

gills of the teleost Channa punctatus. Bull. Environ. Contam. Toxicol.

42 (3), 375–381.

ARTICLE IN PRESSE. Brunelli et al. / Ecotoxicology and Environmental Safety 71 (2008) 436–445444

Author's personal copy

Karnaky, K.J., 1986. Structure and function of the chloride cell of

Fundulus heteroclitus and other teleosts. Am. Zool. 26, 209–224.

Klassen, C.D., 1991. Principles of toxicology. In: Gilman, A.G., Tall,

T.W., Nies, A.S., Taylor, P. (Eds.), Pharmacological Basis of

Therapeutics, eighth ed. McGraw-Hill, pp. 49–61.

Laurent, P., 1984. Gill internal morphology. In: Hoar, W.S., Randall, D.J.

(Eds.), Fish Physiology, vol. 10a. Academic Press, New York, pp.

73–183.

Lock, R.A.C., Van Overbeeke, P., 1981. Effects of mercuryc chloride and

methylmercuryc chloride on mucous secretion in rainbow trout, Salmo

gairdneri Richardson. Comp. Biochem. Physiol. 69C, 67–73.

Mallat, J., 1985. Fish gill structural changes induced by toxicants and

other irritants: a statistical review. Can. J. Fish. Aquat. Sci. 42,

630–648.

Masini, M.A., Sturla, M., Prato, P., Uva, B., 2000. Mitochondria-rich

cells in Antarctic fish gills. Polar Biol. 23, 250–256.

Misra, V., Lal, H., Chawla, G., Viswanathan, P.N., 1985. Pathomorpho-

logical changes in gills of fish fingerlings (Cirrhina mrigala) by linear

alkyl benzene sulfonate. Ecotoxicol. Environ. Saf. 10, 302–308.

Misra, V., Kumar, V., Pandey, S.D., Viswanathan, P.N., 1991.

Biochemical alterations in fish fingerlings (Cyprinus carpio) exposed

to sublethal concentration of linear alkyl benzene sulphonate. Arch.

Environ. Contam. Toxicol. 21 (4), 514–517.

Nation, J.L., 1983. A new method using hexamethil-disilazane for

preparation of soft insect tissues for scanning electron microscopy.

Stain Technol. 58, 347–351.

Nickerson, K.W., Aspedon, A., 1992. Detergent-shock response in enteric

bacteria. Mol. Microbiol. 6 (8), 957–961.

OECD, 1992. Harmonization of classification systems for chemicals. In:

Proceedings of the 18th joint meeting Chemical Group and Manage-

ment Committee, 1992. ENV/MC/CHEM (92) 6.

Perry, S.F., Laurent, P., 1989. Adaptational responses of rainbow trout to

lowered external NaCl: contribution of the branchial chloride cell.

J. Exp. Biol. 147, 147–168.

Pisam, M., Rambourg, A., 1991. Mitochondria-rich cells in the gill

epithelium of teleost fishes: an ultrastructural approach. Int. Rev.

Cytol. 130, 191–232.

Pisam, M., Le Moal, C., Auperin, B., Prunet, P., Rambourg, A., 1995.

Apical structures of ‘‘mitochondria-rich’’ a and b cells in euryhaline

fish gill: their behaviour in various living conditions. Anat. Rec. 241,

13–24.

Ribelles, A., Carrasco, C., Rosety, M., 1995. Morphological and

histochemical changes caused by sodium dodecyl sulphate in the gills

of giltheads (Sparus aurata, L.). Eur. J. Histochem. 39, 141–148.

Richman, N.H., Zaugg, W., 1987. Effects of cortisol and growth hormone

on osmoregulation in pre- and desmoltified coho salmon (Oncor-

hynchus kisutch). Gen. Comp. Endocrinol. 65, 189–198.

Rosas, C., Espina, S., Diaz., F., Curts., J., Rosas, I., 1988. Effect of

sublethal detergent concentration upon gill permeability of Ctenophar-

yngodon idella (pisces; cyprinidae). Water, Air, Soil Pollut. 24 (3/4),

253–258.

Rosety-Rodrıguez, M., Ordonez, F.J., Rosety, M., Rosety, J.M., Rosety,

I., Ribelles, A., Carrasco, C., 2002. Morpho-histochemical changes in

the gills of Turbot, Scophthalmus maximus L., induced by sodium

dodecyl sulfate. Ecotoxicol. Environ. Saf. 51, 223–228.

Roy, D., 1988a. Impact of detergents on the protein histochemistry of

various cell types of the gill epithelium of Rita rita. Ecotoxicol.

Environ. Saf. 15, 206–211.

Roy, D., 1988b. Statistical analysis of anionic detergent-induced changes

in the goblet mucous cells of opercular epidermis and gill epithelium of

Rita rita (Ham.) (Bagridae: Pisces). Ecotoxicol. Environ. Saf. 15,

260–271.

Shepherd, K.L., 1994. Functions for fish mucus. Rev. Fish Biol. Fisheries

4, 401–429.

Sigoillot, J.C., Nguyen, M.H., 1992. Complete oxidation of linear

alkylbenzene sulfonate by bacterial communities selected from coastal

seawater. Appl. Environ. Microbiol. 58 (4), 1308–1312.

Singer, M.M., Tjeerdema, R.S., 1993. Fate and effects of the surfactant

sodium dodecylsul fate. Rev. Environ. Contam. Toxicol. 133, 95–149.

Swedmark, M., Braaten, B., Emanuelsson, E., Granmo, A., 1971.

Biological effects of surface active agents on marine animals. Mar.

Biol. 9, 183–201.

Van der Heijden, A.J.H., Verbost, P.M., Eygensteyn, J., Li, J., Wendelaar

Bonga, S.E., Flik, G., 1997. Mitochondria-rich cells in gills of tilapia

(Oreochromis mossambicus) adapted to fresh water or sea water:

quantification by confocal laser scanning microscopy. J. Exp. Biol.

200, 55–64.

Van den Heuvel, M.R., Power, M., Richards, J., MacKinnon, M., Dixon,

D.G., 2000. Disease and gill lesion in yellow perch (Perca flavescens)

exposed to oil sands mining-associated waters. Ecotoxicol. Environ.

Saf. 46, 334–341.

Wendelaar Bonga, S.E., Van der Meij, C.J.M., 1989. Degeneration and death,

by apoptosis and necrosis, of the pavement and chloride cells in the gills of

the teleost Oreochromis mossambicus. Cell Tissue Res. 255, 235–243.

Witters, H., Berckmans, P., Vangenechten, C., 1996. Immunolocalization

of Na+/K+ ATPase in the gill epithelium of rainbow trout,

Oncorhynchus mikiss. Cell Tissue Res. 283, 461–468.

Yilmaz, M., Gul, A., Karacose, E., 2004. Investigation of acute toxicity

and the effect of cadmium chloride (CaCl2*2H2O) metal salt on

behavior of guppy (Poecilia reticulata). Chemosphere 56, 375–380.

Zaccone, G., Fasulo, S., Lo Cascio, P., Licata, A., 1985a. Patterns of

enzyme activities in the gills of the catfish Heteropneustes fossilis

(Bloch) exposed to the anionactive detergent Na-alkyl-benzenesulpho-

nate (LAS). Histochem. J. 82, 341–343.

Zaccone, G., Lo Cascio, P., Fasulo, S., Licata, A., 1985b. The effect of an

anionic detergent on complex carbohydrates and enzyme activities in

the epidermis of the catfish Heteropneustes fossilis (Bloch). Histochem.

J. 17, 453–466.

Zaccone, G., Fasulo, S., Lo Cascio, P., Licata, A., 1986. Effect of

sublethal concentrations of a zwitterionic detergent on oxidative

enzymes in the gills of the fresh catfish Heteropneustes fossilis (Bloch).

Arch. Biol. 97, 93–105.

ARTICLE IN PRESSE. Brunelli et al. / Ecotoxicology and Environmental Safety 71 (2008) 436–445 445