The toxic effect of phthalate esters on

PAE dissolved in corn oil was continuously fed to prawns for 8 days and five immune parameters total hemocyte count, THC; ratio of granulocytes to hyalinocytes, G/H; intrahemocytic total

The toxic effect of phthalate esters on immune responses of giant

freshwater prawn (Macrobrachium rosenbergii) via oral treatment

Wen-Liang Chen, Hung-Hung Sung∗

Department of Microbiology, Soochow University, Taipei 111, Taiwan, ROC

Received 23 November 2004; received in revised form 28 March 2005; accepted 2 May 2005

Abstract

A previous in vitro study has indicated that four phthalate esters (PAEs) could damage hemocytes and decreases the cellular immunity of prawns [Sung, H.H., Kao, W.Y., Su, Y.J., 2003 Effects and toxicity of phthalate esters to hemocytes of giant

freshwater prawn, Macrobranchium rosenbergii Aquat Toxicol 64, 25–37] The aim of this study was to investigate the in vivo

effect of four PAEs, diethyl phthalate (DEP), dihexyl phthalate (DHP), dipropyl phthalate (DPrP) and diphenyl phthalate (DPP)

on the defense system of the giant freshwater prawn, M rosenbergii PAE dissolved in corn oil was continuously fed to prawns

for 8 days and five immune parameters (total hemocyte count, THC; ratio of granulocytes to hyalinocytes, G/H; intrahemocytic total phenoloxidase activity, POT; intracellular superoxide anion (O2−) production; transglutaminase (TGase) activity) were separately detected on days 1, 4 and 8 In addition, mortality was determined on days 4 and 8 after challenging the prawns with

Lactobacillus garvieae In comparison with untreated prawns, the results showed that DHP demonstrated the lowest toxicity in that it only influenced the PO activity and O2−production before 4 days after treatment and caused 6.6% mortality on day 8 DEP decreased G/H, POTand TGase activity on day 1 and reduced THC, G/H and POTand caused 16.6% mortality on day 4; however, on day 8, it increased O2−production and caused no mortality In the DPrP-treated group, a reduction of all the immune reactions apart from TGase activity and 22.2% mortality were detected on day 4 As for the effect of DPP, results showed that

it decreased all the immune parameters apart from THC on days 1 and 4, but caused no mortality on day 4; but on day 8, an increase of O2−production and 17.7% mortality were detected These results indicated that the immune reactions of prawns were variable due to the different toxic effects of PAEs In addition, it was found that, on day 8 after treatment, the three PAEs, DHP, DPrP and DPP increased O2−production and did not influence the other four reactions, but mortality was detected in these groups These results suggest that other physiological responses may also be affected to increase the susceptibility of prawns to pathogens

© 2005 Elsevier B.V All rights reserved

Keywords: Phthalate esters; Endocrine disruptor (ED); Chemical pollutant; Prawn; Immunity; Toxicity

∗ Corresponding author Tel.: +886 2 28819471x6860;

fax: +886 2 28831193.

E-mail address:hhsung@scu.edu.tw (H.-H Sung).

1 Introduction

In many countries of the world (e.g Taiwan), prawn cultivation has an economic importance in the 0166-445X/$ – see front matter © 2005 Elsevier B.V All rights reserved.

doi:10.1016/j.aquatox.2005.05.008

aquaculture industry In Taiwan, production of the

giant freshwater prawn, Macrobrachium rosenbergii,

reached a peak (16,196 t) in 1991, but decreased in

1992 and 1993 and was only 7223 t in 1999 (Primavera,

1997; Rosenberry, 1998) The decline in production

of farmed prawns resulted from outbreaks caused by

yeast infection in the cool season and bacterial

infec-tion in the hot season (Hsu, 1993; Cheng and Chen,

1998) However, poor pond management or

deteriora-tion of water quality can also induce disease outbreaks

and thus, decrease production (Liao, 1989;

Chamber-lain, 1997) Previous studies have demonstrated that

the appearance of harmful factors, such as toxins and

changes in dissolved oxygen, salinity or temperature,

created an environment which increased the

suscep-tibility of prawns to disease (Lightner, 1988, 1996;

Lightner and Redman, 1998; Brock, 1992; Brock and

Lightner, 1990; Le Moullac et al., 1998; Cheng et al.,

2002) Other studies have reported that both the heavy

metals released from sediments and the contamination

of shrimp farms due to pesticides and pollutants from

agriculture or industrial activities may decrease the

resistance of the shrimp to disease (Primavera, 1993;

Flaherty and Karnjanakesorn, 1995)

Phthalate esters (PAEs) are widely used industrial

chemicals which serve as important additives to impart

flexibility to polyvinyl chloride (PVC) resins and

become widely diffused in the environment (Jobling et

al., 1995) via the manufacturing process PAE

concen-trations have been reported in the range of 0.1–300 ␮g/l

for surface marine waters (Mayer et al., 1972; Giam et

al., 1978; Gledhill et al., 1980; Fatoki and Vermon,

1990) and freshwater sites (Gledhill et al., 1980) and

from 0.1 ng/g to 100 ␮g/g for river sediments (Thuren,

1986; Tan, 1995) In Taiwan, PAEs have been found to

be widely distributed in river water and sediment and

soil (Liu et al., 2000; Yuan et al., 2002) and four PAEs,

diethyl phthalate (DEP), dibutyl phthalate (DBP),

ben-zyl butyl phthalate (BBP) and di-(2-ethylhexyl)

phtha-late (DEHP), have also been found to accumuphtha-late in

fish (Chang et al., 2004)

The United States Environmental Protection

Agency (USEPA) and its counterparts in several other

countries have classified the most commonly

occur-ring PAEs as priority pollutants and endocrine

disrupt-ing compounds (ECPI, 1996) Numerous experiments

have shown that bioaccumulation of PAEs occurs in

the aquatic and terrestrial food chain (reviewed from

Staples et al., 1997a) DEHP has been shown to be absorbed, metabolized and largely accumulated in the tissue of a penaied shrimp via oral administration and this process had a linear relationship with the dose according to the dose range studied (Hobson et al.,

1984) Many studies have demonstrated the acute toxi-city and chronic toxitoxi-city of phthalate esters to microor-ganisms, algae, aquatic invertebrates and fish (reviewed

byStaples et al., 1997b)

In crustaceans, hemocytes play a crucial role in non-specific cellular immunity against pathogens and parasites, which function as the primary immune responses including phagocytosis, encapsulation, nod-ule formation and cytotoxic mediation (Anderson,

1992) The circulating phagocytic hemocytes have received considerable attention as the primary cell-mediated immunity mechanism.Sung and Song (1996) have indicated that phagocytosis is performed by hyalinocytes in prawns According to previous stud-ies (S¨oderh¨all, 1982; Ratcliffe et al., 1985; Smith and S¨oderh¨all, 1991), several proteins associated with the hemocyte prophenoloxidase-activating system (PAS), which is released from induced semigranular and gran-ular cells, play an important role in non-self recognition and host defense for elimination of foreign particles

in the body cavity of crayfish and other crustaceans (S¨oderh¨all et al., 1994) Furthermore, a coagulation system is essential in invertebrates to prevent excess blood loss from a wound and to obstruct microorgan-isms, which would otherwise invade the wound In the wound area, the clottable protein oligomerizes to pre-vent hemolymph loss through breaks in the exoskeleton and dissemination of bacteria throughout the body

In crayfish hemocytes, both semigranular and granu-lar cells, as well as the muscle tissues contain TGase (Hisanori et al., 1997) Therefore, the phagocytic activ-ity, the activation of PAS and TGase may be used as the defense indicators in crustaceans, including cultured prawns (Rodr´ıguez and Moullac, 2000)

Before the study ofSung et al (2003), which indi-cated that PAEs could damage hemocytes and reduce the cellular immunity of prawns by means of in vitro exposure experiments, no one knew the effects of PAEs and their derivates on the defense reactions

of crustaceans such as cultured prawns Therefore,

in this study, we further investigated the effects of four PAEs, diethyl phthalate (DEP), dihexyl phtha-late (DHP), dipropyl phthaphtha-late (DPrP) and diphenyl

phthalate (DPP), which are highly toxic to the

hemo-cytes of giant freshwater prawns (M rosenbergii)

(Sung et al., 2003) on the defense functions of prawns

given these PAEs orally for 8 days Five immune

parameters, comprising total hemocyte count (THC),

ratio of granulocytes to hyalinocytes (G/H),

intrahemo-cytic total phenoloxidase activity (POT), intracellular

superoxide anion (O2−) production and

transglutami-nase (TGase) activity, were used to evaluate the effect

of PAEs on prawns; this was further defined by

record-ing the resultant mortality when the prawns were

chal-lenged with a pathogen

2 Materials and methods

2.1 Chemicals and preparation

Four phthalate esters were used in this study as

shown inFig 1 These were diethyl phthalate (Chem

Service Co., O-525), dihexyl phthalate (Chem Service

Co., Pt-21), diphenyl phthalate (Aldrich Chem Co.,

10588-0) and dipropyl phthalate (Chem Service Co.,

F2158) The hemocyte-culture medium, M-199, was

purchased from Life Technologies Inc (GIBCO BRL 21200-076) The four PAEs were separately dissolved

in corn oil to a concentration of 10,000 ppm as a stock solution, which was stored at room temperature prior

to the experiments

2.2 Animals 2.2.1 Acclimation

pur-chased on separate days from local prawn farms, were acclimated in 360 l glass aquaria containing fresh pond water at 28◦C for at least 3 days prior to the exper-iments Prawns were fed with synthetic feed pellets twice a day, an amount equivalent to 5% of their body weight The stocking densities were maintained at 20 prawns per aquarium

2.2.2 Oral treatment with phthalate esters

For each type of PAE, prawns were divided into three experimental groups and one control group The three experimental groups were continuously fed with the PAE for 1, 4 and 8 days, respectively Each prawn was fed with 100 ␮l of the PAE (1 ␮g/␮l) once a day

Fig 1 The chemical structures and properties of the four phthalate esters used in this experiment MW, molecular weight; (–) no detection Data

of MW and aqueous solubility are cited from the review of Staples et al (1997b) and the study of Cousins and Mackay (2000) , respectively.

using a syringe with a soft silicon tube The control

group was fed with an equal volume of corn oil

2.3 Preparation of hemolymph and hemocyte

samples

To evaluate the immune reactions of PAE-treated

prawns, both hemolymph and hemocyte samples were

prepared A hemolymph sample (0.5 ml) was drawn

from the first abdominal segment of each prawn with

a 25 G hypodermic needle containing 0.5 ml of

anti-coagulant (10 mM Tris–HCl, 100 mM trisodium

cit-rate, 10 mM EDTA, 82 mM glucose, 20 mM NaCl,

pH 7.56) with an osmolarity of 420 ± 20 mOsm/kg

Hemocyte suspensions were prepared according to

a procedure described by Song and Hsieh (1994)

Briefly, the hemolymph sample was centrifuged at

300 × g for 10 min at 4◦C and the resultant

hemo-cyte pellet was suspended in 1 ml of calcodylate (CAC)

buffer (pH 7.0) or M-199 medium with an osmolarity

of 420 ± 20 mOsm/kg Hemocyte concentrations were

adjusted according to different experiments Only, cell

suspensions with a viability of 85% or more, tested by

trypan blue exclusion (0.05% in 0.01 M PBS), were

used to determine the immune reactions in this study

The hemolymph sample was used to determine the

total hemocyte count, the differential hemocyte count

(DHC) and transglutaminase activity The hemocyte

sample in CAC buffer or M-199 medium was used to

examine the intrahemocytic total phenoloxidase

activ-ity and the production of superoxide (O2−),

respec-tively

2.4 Determination of defense responses

2.4.1 Total hemocyte count and differential

hemocyte count

The total number of hemocytes in a mixture of

10 ␮l of hemolymph and trypan blue (Sigma, T-6164;

0.05% trypan blue in 0.01 M PBS, pH 7.56,

osmolar-ity 420 ± 20 mOsm/kg) was counted with a

hemocy-tometer (Hausser Scientific, Bright-Line) under a light

microscope (NIKON, ECLIPSE, E800) at a

magnifica-tion of 100× As for determining the differential

hemo-cyte count, 10 ␮l of hemolymph was fixed with an equal

volume of 20% neutralized formaldehyde for 1 min at

room temperature After adding 20 ␮l of 0.5% Evans’

Blue (Sigma, E-2129), 20 ␮l of mixture was smeared

on a cover glass (18 mm × 18 mm) The numbers of both hyalinocytes (HC) and granulocytes (GC; com-posing semigranular and granular cells) were counted for a total of 50–100 cells with a light microscope at a magnification of 400× The ratio of GC to HC was cal-culated by the formula: G/H = number of GC/number

of HC The values given in this study were the means

of average relative THC or G/H ± the standard devia-tion (S.D.) of the mean of three replicates from more than 18 individuals All of the values of average rela-tive THC or G/H from both experimental (PAE-treated) group and control (corn oil-treated) group were com-pared to the untreated group The value of relative THC or G/H was calculated using the formula: THC or G/H of PAE-treated prawn/THC or G/H of untreated prawn

2.4.2 Intrahemocytic total phenoloxidase activity

Before assay of the phenoloxidase activity, hemo-cyte lysate supernatant (HLS) of prawn was prepared according to procedures described by Sung et al (1996) Briefly, the hemocyte suspension in 0.01 M CAC buffer was homogenized using a sonicator (Vibra cell, AC-600) equipped with a microtip and centrifuged

at 43,000 × g for 30 min at 4◦C and the HLS was then collected The resultant HLS was used as an enzyme source and its protein concentration was determined by the Protein assay Kit II (Bio-Rad, USA)

Intrahemocytic total phenoloxidase activity, which resulted from all the intrahemocytic prophenoloxidase (proPO) being completely catalyzed to form PO when HLS was treated with trypsin, was assayed as described

bySung et al (2004) After a mixture of 25 ␮l of HLS and an equal volume of trypsin solution (1 mg/ml of 0.01 M CAC buffer; Sigma, T-4665) was incubated at

30◦C for 15 min, 200 ␮l of freshly prepared substrate solution, 0.01 M of l-3,4-dihydroxyphenylalanine (l-DOPA, Sigma, D-9628) in CAC buffer, was added and reacted for 1 min The optical absorbance at 490 nm was measured One unit of enzyme activity was defined

as an increase in absorbance of 0.001/(min mg) of pro-tein (S¨oderh¨all and Unestam, 1979) The values given

in this study were the means of the average relative POT activity (RA) ± the standard deviation of the mean of three replicates from more than 18 individuals The value of RA was calculated using the formula: POT

of HLS from PAE-treated prawn/POT of HLS from untreated prawn

2.4.3 NBT assay

Since the production of superoxide anions (O2−)

and H2O2contribute to the initiation of a

proinflam-matory event, in this study, O2−production assayed by

NBT reduction was used as a defense parameter This

assay was conducted as described bySong and Hsieh

(1994) Reactions occurred in flat-bottomed 96-well

microtiter plates, with each well coated with 100 ␮l of

poly-l-lysine solution (0.2%, Sigma P-1274), at room

temperature for 30 min Hemocyte suspension (100 ␮l)

was added to each well (106cells/well) and

cytocen-trifuged (Kubota, KN-70) at 300 × g for 10 min at 4◦C

After removing the supernatant, 100 ␮l of zymosan A

(from Saccharomyces cerevisiae; Sigma Z-4250)

sus-pension (107particles/well) was added and the

mix-ture was incubated at 28◦C for 30 min After washing

with M-199, the hemocytes were stained with 200 ␮l

nitroblue tetrazolium solution (NBT, 0.15% in M-199)

for 30 min at 28◦C The staining reaction was

termi-nated by removing the NBT solution and then adding

200 ␮l of absolute methanol (Merck) After three

wash-ings with 70% methanol, the hemocytes were air-dried

and coated with a solution of 120 ␮l KOH (2 M) and

140 ␮l dimethyl sulfoxide (DMSO) to dissolved

cyto-plasmic formazan, and then, the mixture was measured

at 630 nm with a microplate reader (Molecular Device,

Emax) In order to determine the reproducibility of the

results, hemocytes collected from more than 18 prawns

were individually assayed The ratio of OD630 from

the treated hemocytes to the OD630 of the untreated

hemocytes was used as an index for comparing the

effects of different PAEs on both O2−generation and

reductase activity, since either O2−or cellular

reduc-tase can reduce NBT to the monoformazan (Tarpey and

Fridovich, 2001)

2.4.4 Transglutaminase activity

The biotin-labeled casein used as a substrate in the

transglutaminase activity assay was prepared

accord-ing to procedures described by Song et al (2003)

Briefly, a mixture of 200 ␮l of biotinamidocaproyl

hydrazide (Sigma, B-3770) solution (100 mg/ml of

DMSO) and 10 ml of casein (Sigma, C-5890)

solu-tion was stirred at room temperature for 12 h and

then dialyzed in 50 mM Tris–HCl buffer (pH 7.4) in

a 14 K dialysis tube at 4◦C overnight After dialysis,

the biotin-labeled casein solution was diluted 20× with

50 mM Tris–HCl buffer (pH 7.4) and stocked at 4◦C

The TGase activity assay was performed according

to procedures described bySeiving et al (1991) First,

200 ␮l of the casein solution (1 mg/ml of sodium car-bonate buffer at pH 9.8) was added to each well of a flat-bottomed 96-well microtiter plate Before washing with washing buffer (50 mM Tris–HCl, 0.15 M NaCl and 0.1% Tween-80) containing 1 mg/ml of dithiothre-itol (DTT; Amersham Pharmacia Biotech), they were incubated at room temperature overnight Thereafter, a hemolymph sample (100 ␮l) was added to one casein-coated well and then serially diluted two-fold with

50 mM Tris–HCl (osmolarity 420 ± 20 mOsm/kg) After supplementation with 100 ␮l of reagent Ca2+ (6 ml of 5.8 U/ml of thrombin, 2 ml of 50× diluted biolinated casein, 1 ml CaCl2), the mixture was incu-bated at 37◦C for 20 min Following washing twice with washing buffer, 100 ␮l of streptavidin-labeled alkaline phosphatase (Sigma, S2890) solution was added and then incubated at 28◦C for 45 min After

washing to remove the unbound biotin, 100 ␮l of

p-nitrophenylphosphate solution (1 mg/ml) was added and incubated at 37◦C for 30 min Color develop-ment was measured at a wavelength of 405 nm Instead

of the hemolymph sample, a guinea pig liver TGase (Sigma, T5398) solution was added as a standard enzyme to calculate the standard curve of OD405versus enzyme activity (unit/mg) The protein concentration

of hemolymph samples was determined by the Pro-tein Assay Kit II (Bio-Rad, USA) The values given in this study were the means of average relative TGase activity ± the standard deviation of the mean of three replicates from more than 18 individuals The value

of relative TGase activity was calculated using the formula: TGase activity of PAE-treated prawn/TGase activity of untreated prawn

2.5 Susceptibility of prawns to pathogen 2.5.1 Preparation of bacterial suspension

The pathogen Lactococcus garvieae used in the study was isolated from diseased M rosenbergii with

whitish musculature syndrome (Cheng and Chen, 1998; Chen et al., 2001) and was a kind gift from Dr Winton Cheng (Department of Aquaculture, National Pingtung University of Science and Technology) A stock kept in 50% glycerol at −20◦C was thawed at

37◦C for 5 min; 1 ml of the stock was inoculated to a

250 ml flask containing 50 ml of brain heart infusion

broth (BHIB, Difco) and incubated overnight at 28◦C,

150 rpm Subsequently, 1 ml of bacterial solution was

subcultured into 50 ml of BHIB and the mixture was

incubated at 28◦C until bacterial growth reached the

late-log phase Following centrifugation at 3000 × g

and 4◦C for 15 min, the pellet was washed once and

suspended in sterile 0.01 M phosphate-buffered

solu-tion (0.01 M PBS, osmolarity 420 ± 20 mOsm/kg, pH

7.56) The concentration of bacterial cells was adjusted

to 1 × 108cells/ml via a cell counting method (Petroff

Hausser Counting Chamber, Hausser Scientific Co.,

USA) using a light microscope at a magnification of

1000×

2.5.2 Challenge experiment

In the challenge experiment for each PAE, each

batch of prawns was divided into four groups with 12

or 15 prawns in each group; two groups were

con-tinuously fed with PAE for 4 days and the other two

groups were fed for 8 days On day 4 after treatment,

the two 4-day-treated groups were divided into one

experimental group, which was injected with 100 ␮l

of the bacterial suspension at the dose of 105cells/g

of prawn (1/10 LD50) and one corresponding control

group, which was injected with an equal volume of

ster-ile PBS The injection dose was able to cause infection

of healthy untreated prawns, but not induce death (Sung

and Sun, 2002) The same challenge experiment was

performed on day 8 after treatment After injection,

prawns were held in aquaria at 28◦C with aeration

The number of dead prawns was recorded twice daily,

until no prawns died for 2 days Mortality percentages

were calculated using the formula: (total number of

dead prawns number of non-specific death)/(total

num-ber of prawns numnum-ber of non-specific death) × 100%,

where the non-specific death represents the

num-ber of prawns that died within the first 12 h after

injection

2.6 Detection of anorexia

To evaluate whether the effect of PAEs on prawn

immunity is caused indirectly by an influence on prawn

appetite, in both PAE-treated and untreated groups, the

amount of synthetic feed pellets taken by prawns was

recorded After feeding, the quantity of remaining

syn-thetic feed pellets was calculated at 1 h intervals for

12 h

2.7 Statistics

Outbreed prawns were used as samples in this study and the physiological status of each sample used was significantly different All results from the experi-ments, including the five immune parameters, total hemocyte count, ratio of granulocytes to hyalinocytes, intrahemocytic total phenoloxidase activity, intracellu-lar superoxide (O2−) production and transglutaminase activity, as well as the mortality of prawns challenged with the pathogen, also showed great variations among individuals Therefore, the data from the means of the average relative value of three replicates from at least

15 individuals were statistically analyzed with regard

to the effects of PAEs on immune reactions and the sus-ceptibility of prawns by using ANOVA and Duncan’s multiple range tests with a specified significance level

of p < 0.05.

3 Results

To determine whether prawn appetite is affected

by corn oil or PAE treatment, the amount of syn-thetic feed pellets taken by prawns was recorded during

12 h In this observation period, corn oil- and PAE-treated prawns took fewer pellets in the first 6 h and recovery of normal appetite was observed during the second 6 h Furthermore, we still observed but did not record the data during the whole experimental period Appetite did not differ between the treated and untreated groups Therefore, in this study, the effects of PAEs on immune responses are apparently independent

of prawn appetite

Following the feeding of prawns with PAEs, five immune parameters were detected on days 1, 4 and

8 The total hemocyte counts were not different from corresponding control groups on days 1 and 8 after treatment with the four PAEs; however, a significant reduction in the percentage of THC was detected in the

DEP- and DPrP-treated groups (p < 0.05) on day 4, the

figures being 77.7 and 72.7%, respectively (Table 1)

As for the ratio of granulocytes to hyalinocytes, it was found that the percentage of G/H was decreased by 20.7 and 25.3% on days 1 and 4, respectively, after DEP treatment, by 23.5% on day 1 in the DPP-treated group

and by 34.5% on day 4 in the DPrP group (p < 0.05)

(Table 2)

Table 1

Changes in total hemocyte count (THC) of prawns given phthalate esters (PAEs) orally

Days of treatment Relative THC (×100%)

1 (n = 18) 100 ± 22.5 N 109 ± 22.6 N 113 ± 24.7 N 92.8 ± 19.1 N 121.7 ± 24.2 N

4 (n = 26) 100 ± 10.5 N 77.7 ± 18.5 b 98.6 ± 18.0 N 72.7 ± 19.3 a 104.0 ± 35.7 N

8 (n = 18) 100 ± 28.3 N 100.3 ± 27.0 N 102.9 ± 18.9 N 91.2 ± 19.4 N 92.1 ± 30.6 N

DEP, diethyl phthalate; DHP, dihexyl phthalate; DPrP, dipropyl phthalate; DPP, diphenyl phthalate; n, the number of prawns used in this

experiment for each treated group; N, no difference from the corresponding control.

a Prawns in control group were fed corn oil without PAE.

b The data from the means of relative value of three replicates from more than 18 prawns decrease were statistically analyzed using ANOVA

and Duncan’s multiple range tests with a specified significance level of p < 0.05; the error bars represent standard deviation (S.D.).

As shown inFig 2, the total phenoloxidase

activ-ity was reduced by 16.2 and 29.7% on days 1 and 4,

respectively, after DEP treatment (p < 0.05), but was

no different from the corresponding control group on

day 8 In the other three PAE-treated groups, a

signifi-cant reduction of POTwas detected on day 4 (p < 0.05),

but neither on days 1 nor 8 As for the expression

of superoxide (O2−) production and reductase

activ-ity (superoxide/reductase) analyzed by NBT assay, the

results showed that it was not affected on days 1 and

4 after DEP treatment, but was significantly enhanced

on day 8 (p < 0.05) (Fig 3) In the other three PAE-treated groups, the expression of superoxide/reductase was decreased on both days 1 and 4, but increased

on day 8 after PAE treatment (Fig 3) Transglutam-inase activity was significantly reduced on day 1 after DEP and DPP treatment, by 33.5 and 42%, respec-tively; however, there were no differences from the

corresponding control groups on days 4 and 8 (p < 0.05)

(Fig 4) No change in TGase activity was found in Table 2

Changes in ratio of granulocytes to hyalinocytes (G/H) of prawns given phthalate esters (PAEs) orally

Days of treatment Relative G/H (×100%)

1 (n = 18) 100 ± 26.6N 79.3 ± 16.3b 86.0 ± 18.3N 90.5 ± 25.0N 77.5 ± 19.1b

4 (n = 26) 100 ± 25.9N 74.7 ± 21.2b 86.2 ± 18.7N 65.5 ± 25.2b 87.4 ± 34.0N

8 (n = 18) 100 ± 21.7N 154.4 ± 108.8N 144 ± 124.2N 134.3 ± 76.3N 141.5 ± 121.5N

DEP, diethyl phthalate; DHP, dihexyl phthalate; DPrP, dipropyl phthalate; DPP, diphenyl phthalate; n, the number of prawns used in this

experiment for each treated group; N, no difference from the corresponding control.

a Prawns in control group were fed corn oil without PAE.

b The data from the means of relative value of three replicates from more than 18 prawns decrease were statistically analyzed using ANOVA

and Duncan’s multiple range tests with a specified significance level of p < 0.05; the error bars represent standard deviation (S.D.).

Table 3

Susceptibility of PAE-treated Macrobrachium rosenbergii to Lactococcus garvieae

Days of treatment Mortality a (%)

DEP, diethyl phthalate; DHP, dihexyl phthalate; DPrP, dipropyl phthalate; DPP, diphenyl phthalate; n, the total number of prawns used in the

challenge experiment for each PAE-treated group.

a Percentage of mortality was calculated by the following formula: death rate (%) = (total number of deaths − number of non-specific deaths)/(total number of prawns − number of non-specific deaths) The data were the means of mortality ± the standard deviation (S.D.) of three replicates from more than 12 prawns.

b PAE-treated prawns were injected with sterile PBS without bacterial cells.

Fig 2 Changes in intrahemocytic total phenoloxidase activity (PO T )

of prawns on days 1, 4 and 8 after PAE treatment DEP, diethyl

phtha-late; DHP, dihexyl phthaphtha-late; DPrP, dipropyl phthaphtha-late; DPP, diphenyl

phthalate The data from the means of average relative PO T ± the

standard deviation (S.D.) of three replicates from more than 18

prawns were statistically analyzed using ANOVA and Duncan’s

mul-tiple range tests with a specified significance level of p < 0.05; the

error bars represent standard deviation D, significant decrease

com-pared to activity of the corresponding control; N, no difference from

activity of corresponding control.

either the DHP- or the DPrP-treated group With

fur-ther statistical analysis of Pearson correlation, the POT,

superoxide/reductase and TGase activity in this study

were shown to have a significant positive

correla-tion with THC, percentage of granulocyte and G/H

(p < 0.01).

Finally, to evaluate the effects of PAEs on the

sus-ceptibility of prawns to pathogens, prawns were

chal-lenged with L garvieae (×105cells/g of prawn) via

injection on days 4 and 8 after PAE treatment and

the subsequent mortality was determined As shown in

Table 3, on day 4 after treatment, the mortality of

DEP-and DPrP-treated prawns was 16.6 DEP-and 22.2%,

respec-tively, and all of the DHP- and DPP-treated prawns

sur-vived However, on day 8, the DHP- and DPP-treated

groups had mortalities of 6.6 and 17.7%, respectively

In addition, the DPrP-treated group continued to have

a mortality of 20% In this experiment, the mortality of

all control groups was 0% (Table 3)

Fig 3 Changes in intrahemocytic superoxide production in prawns

on days 1, 4 and 8 after PAE treatment assessed by nitroblue tetra-zolium (NBT) assay DEP, diethyl phthalate; DHP, dihexyl phthalate; DPrP, dipropyl phthalate; DPP, diphenyl phthalate The data from the means of average relative OD 630 of three replicates from more than 18 individuals were statistically analyzed using ANOVA and Duncan’s multiple range tests with a specified significance level

of p < 0.05; the error bars represent standard deviation (S.D.) D,

significant decrease; E, enhancement, compared to activity of the corresponding control; N, no difference from activity of the corre-sponding control.

4 Discussion

Increasing evidence has indicated that many sub-stances, which are degraded from chemical pollutants but not biologically decomposed in sewage treatment works, are often not acutely toxic to exposed aquatic animals when they are emitted into water, but lead to

a chronic intoxication resulting in tissue alterations, including the formation of neoplasias There is a devel-oping awareness that in both fish and mollusks, diseases

in populations are linked to environmental changes or coastal marine pollution There is considerable evi-dence to support links between environmental changes (including contaminants), non-infectious diseases and

a depression of the immune system (Durnier and Siwicki, 1993; Pipe and Coles, 1995)

Fig 4 Changes in hemolymph transglutaminase (TGase) activity of

prawns on days 1, 4 and 8 after PAE treatment DEP, diethyl

phtha-late; DHP, dihexyl phthaphtha-late; DPrP, dipropyl phthaphtha-late; DPP, diphenyl

phthalate The data from the means of average relative TGase

activ-ity ± the standard deviation (S.D.) of three replicates from more than

18 prawns were statistically analyzed using ANOVA and Duncan’s

multiple range tests with a specified significance level of p < 0.05;

the error bars represent standard deviation D, significant decrease

compared to activity of the corresponding control; N, no difference

from activity of the corresponding control.

In crustaceans, environmental stress from

pollu-tants seems to be an important factor in determining

the reduction of immunocompetence and is signalled

by the appearance or increased prevalence of

dis-ease (Victor et al., 1990; Smith and Johnston, 1992)

Effects include infection pressure from facultative

microbial pathogens and reduced resistance to

infec-tion (Sindermann, 1979) Exposure of the common

shrimp, Crangon crangon, to contaminated sediments,

which contained numerous compounds including

poly-chlorinated biphenyls (PCBs), polynuclear aromatic

hydrocarbons (PAHs) or heavy metals, has shown that

the exposed shrimp displayed an elevation in

recov-erable hemolymph volume and a reduction in total

hemocyte count; biochemical assays also indicated

reduced hemocyte phenoloxidase (PO) activity (Smith

et al., 1995) In addition, the freshwater prawn,

Mac-robrachium idea,exposed to 1 ␮g/l of mercuric

chlo-ride for 30 days, exhibited hyperplastic gill lamellae

engorged with hemocytes; the hemocytes were released

into the interlamellar spaces through necrotic regions

and then covered the entire gill lamellae (Victor et al.,

1990) These findings indicate that chronic exposure

to contaminated sediment has a marked effect on host defense in marine crustaceans

Phthalate esters, considered as endocrine disrupting chemicals (EDCs), are found in various environmental and biological samples (Mayer et al., 1972; Giam et al., 1978; Gledhill et al., 1980; Thuren, 1986; Fatoki and Vermon, 1990; Tan, 1995; Yin and Su, 1996; Sta-ples et al., 1997a) Previous studies have demonstrated that acute toxicity and chronic toxicity of phthalate esters were limited in lower PAEs with alkyl chain lengths <C6, a feature related to their water solubil-ity (reviewed by Staples et al., 1997b) An in vitro study ofSung et al (2003)indicated that eight PAEs, four lower PAEs and four higher PAEs (≥C6), influ-enced the immune responses of prawns Three of these eight PAEs, DEP (C2), DPrP (C3) and DHP (C6), not only damaged hemocytes but also reduced hemocyte immunity, including hemocytic adhesion, pseudopo-dia formation, PO activity and superoxide anion (O2−) production In this study, the in vivo effect of PAEs

on the immune responses of prawns was examined

in hemolymph and hemocytes from prawns after oral treatment with four PAEs, comprising two lower PAEs (DEP and DPrP) and two higher PAEs (DHP and DPP) The results from the detection of five immune parame-ters and an assessment of susceptibility to infection are summarized inTable 4

Several studies have shown that the THC can vary greatly in response to infection, environmental stress and endocrine activity during the molting cycle (Smith and Ratcliffe, 1980; Persson et al., 1987; Smith and Johnston, 1992).Le Moullac et al (1998)found that,

in the case of hypoxia, the decrease in hemocyte number was associated with a significant decrease of hyalinocytes and semigranular cells, while the number

of large granular cells changed only a little In mercury-exposed prawns, the number of circulating hemocytes decreased and this result could be a consequence of hemocyte immobilization in the gills (Victor et al.,

1990) In addition, our previous study has demonstrated the PAE could damage prawn hemocytes (Sung et al.,

2003) In this study, a decrease in THC was found in prawns treated with the two lower PAEs, DEP and DPrP (Table 1), but not in those treated with the two higher PAEs; the decrease in THC was significantly associated with a decrease of granulocytes (Table 2) These results

Table 4

Summary of the effect of PAEs on the immune responses of prawns via oral treatment

DEP, diethyl phthalate; DHP, dihexyl phthalate; DPrP, dipropyl phthalate; DPP, diphenyl phthalate; D, decrease; E, enhancement; N, no difference from corresponding control; (–) no detection.

may explain why the total intrahemocytic proPO (POT)

was also found to be decreased in both the DEP- and

DPrP-treated groups (Fig 2) However, in the

DPrP-treated group, superoxide (O2−) production, which is

largely produced via phagocytosis of hyalinocytes in

prawns (Bach`ere et al., 1995; Sung and Song, 1996),

was also decreased on day 4 after treatment (Table 4)

Our results suggest that, in prawns, DPrP can decrease

both subpopulations of hemocytes, thereby reducing

the two subpopulation-related immune responses In

addition, combining the results of THC and G/H from

days 1 and 4 in prawns treated with DEP and DPP,

it appears that a decrease in THC may be due to a

reduction of granulocytes, hyalinocytes or both

sub-populations caused by the damage effect of PAE (Sung

et al., 2003)

Previous studies have demonstrated that

intrahemo-cytic O2− production in prawns was reduced under

hypoxic stress (Cheng et al., 2002; Le Moullac et al.,

1998) and after hemocytes were treated in vitro with

PAEs (Sung et al., 2003); but an increase in production

can be influenced in prawns either by an

immunos-timulant (Sung et al., 1998) or by infection with a virus

(Song et al., 2003) In this study, we found that, in three

PAE-treated groups, a reduction in O2−production was

detected on days 1 and 4; however, in all PAE-treated

groups, the production was enhanced on day 8 (Fig 3)

Since the other results from this study cannot clarify

the enhancement effect of PAEs on O2− production, further study is necessary

A coagulation system is essential in invertebrates to prevent excess blood loss from a wound and to obstruct microorganisms, which would otherwise invade the wound.Song et al (2003)have shown that in the Pacific

white shrimp, Litopenaeus vannamei, TGase activity

was decreased in virus-infected shrimp In this study, the decrease of TGase activity was detected on day 1 after treatment with either DEP or DPP, and this was consistent with a decrease of granulocytes; as well as, the reduction the activity seemed to be a short-term effect

In this study, to further clarify whether the change in expression of various immune parameters is related to infection and disease outbreak in prawns, the

suscepti-bility of PAE-treated prawns to L garvieae was

deter-mined The results indicated that, in both lower PAE-treated groups (DEP and DPrP), the mortality was sig-nificantly higher than that of the control groups on day

4 (Table 3); furthermore, the expression of at least three immune parameters was found to be reduced (Table 4) However, the results from the two higher PAE-treated groups (DHP and DPP), and the DPrP-treated group detected on day 8 showed that prawns still died after the challenge, although immune reactions did not change

or had even recovered (Table 4) These results suggest that, in addition to immune responses, other