lable at ScienceDirect

Fish & Shellfish Immunology 28 (2010) 296e302

Contents lists avai

Fish & Shellfish Immunology

journal homepage: www.elsevier .com/locate/ fs i

Effects of stocking density and feed ration on growth and gene expression in theSenegalese sole (Solea senegalensis): Potential effects on the immune response

Emilio Salas-Leiton a, Victoria Anguis a, Beatriz Martín-Antonio a,1, Diego Crespo b,Josep V. Planas b, Carlos Infante a, Jose Pedro Cañavate a, Manuel Manchado a,*

a IFAPA Centro El Toruño, Consejería de Agricultura y Pesca, Junta de Andalucía. Apartado 16, 11500 El Puerto de Santa María, Cádiz, SpainbDepartament de Fisiologia, Facultat de Biologia, Universitat de Barcelona, 08028 Barcelona, Spain

a r t i c l e i n f o

Article history:Received 28 July 2009Received in revised form19 October 2009Accepted 3 November 2009Available online 10 November 2009

Keywords:Stocking densityRation sizeSolea senegalensisGrowthStressCortisolInnate immune system

* Corresponding author. Tel.: þ34 956 01 13 15; faxE-mail address: manuel.manchado@juntadeandalu

1 Present address: Hospital Universitario Virgen delEdif. Laboratorios 5a planta, 41013 Sevilla.

1050-4648/$ e see front matter � 2009 Elsevier Ltd.doi:10.1016/j.fsi.2009.11.006

a b s t r a c t

Stocking density and ration size are two major factors influencing aquaculture production. To evaluatetheir effects on growth and immune system in Senegalese sole (Solea senegalensis) juveniles, a 2 � 2experimental design using two rations (1.0% and 0.25% of the total fish biomass) and two different initialstocking densities (7 and 30 kg m�2) was performed throughout a 60 days culture period. Soles fed 1.0%showed a higher specific growth rate (SGR) than those fed 0.25% (3.3-fold). No differences in SGR at 60days were found between densities in spite of reduced values were detected at high density after 20 days(soles fed 0.25%) and 40 days (soles fed 1%) suggesting a compensatory growth. Physiologically, plasmacortisol levels were elevated in soles at high density (45-fold higher than at 7 kg m�2) whereas nodifferences associated to the feeding ration were observed. To assess the effects at a molecular level, themRNA levels of genes involved in cellular stress (heat shock proteins HSP70 and HSP90), growth (insulin-like growth factors IGF-I, the spliced variants IGF-Ia and IGFI-b, and IGF-II) and innate immune system(g-type lysozyme and hepcidin (HAMP1)) were quantified. No differences in HSP90 expression weredetected between densities or rations. In contrast, IGF-I, IGF-Ia and IGF-II showed reduced transcriptlevels in liver and HSP70 in liver and kidney at high density. Finally, g-type lysozyme and HAMP1expression was greatly affected by both factors exhibiting an important reduction in the transcript levelsat high density and low ration. Overall, our results show that S. senegalensis juveniles might exhibitsatisfactory SGR at high density although the high plasma cortisol levels indicate a crowding stress thatcould negatively affect the expression levels of some of the genes studied.

� 2009 Elsevier Ltd. All rights reserved.

1. Introduction

Ration sizes and stocking densities are two major factors in aqua-culture influencing growth, welfare, and health [1e4]. Restrictedrations are routinely associated with reduced growth rates [4e6].Similarly, high densities can also negatively affect growth dependingon the species and age. Adverse effects on growthperformance at highstocking densities were observed in California halibut (Paralichthyscalifornicus), Atlantic halibut (Hippoglossus hippoglossus), turbot (Psettamaxima) and Dover sole (Solea solea) [7e10]. In contrast, null effectsof high densities upon production were reported in Senegalese sole(Solea senegalensis Kaup), wedge sole (Dicologlossa cuneata), summerflounder (Paralichthys dentatus), winter flounder (Pseudopleuronectes

: þ34 956 01 13 24.cia.es (M. Manchado).Rocio, Dpto. de Hematología,

All rights reserved.

americanus) and turbot [11e15]. Whatever the effects on growth,changes inphysiological parameters have frequently been reported. InSenegalese sole, a stocking density of 9.1 kgm�2 for 63 days increasedcortisol levels and reduced free amino acids [16]. In the wedgesole, both cortisol and osmolality were also elevated at a densityof 9.0 kg m�2 for 22 days [13]. In Sparus aurata, plasma cortisol andglucose levels increased at high density (40.8 kg m�3 for 15 weeks)decreasing the alternative complement pathway [3]. Sea bass (Dicen-trarchus labrax, L.) stocked at high density (45 kg m�3 for 6 weeks)showedahigherconcentrationofnon-esterified fattyacids [17].Duetothe importance of stocking densities in commercial fish production, itis crucial to establish the physiological responses and the molecularmechanismsthatgovernadaptationtocrowding inorder towarrantanoptimal welfare and health status.

Different biomarkers can be evaluated to determine the effectsof feed ration and stocking density on growth and health. Insulin-like growth factors (IGF-I and IGF-II) are single-chain polypeptideswith a structural homology similar to proinsulin. They have beenassociated with multiple functions in fish including somatic

E. Salas-Leiton et al. / Fish & Shellfish Immunology 28 (2010) 296e302 297

growth, metabolism, development, cell differentiation, reproduc-tion, osmoregulation and immune response (reviewed in [18]).Heat shock proteins (HSPs) are highly conserved molecular chap-erones that also participate in a wide spectrum of functions. Theirsynthesis is induced under stressful cellular conditions such as hightemperature, anoxia, tissue trauma, ischemia, heavy metals, radi-ation, calcium increase, glucose deprivation, cancer, and bacterialinfection (reviewed in [19e21]). HSPs protect the cells by pre-venting protein aggregation, facilitating folding of newly anddamaged proteins, and targeting misfolded proteins to specificdegradative pathways. According to their molecular weight, theycan be classified into several major categories, e.g., HSP100, HSP90,HSP70, HSP60, and low molecular weight. Recently, the inducibleform HSP90AA has been reported as a useful marker in response tothermal stress in Senegalese sole [22]. Finally, g-type lysozyme andhepcidin antimicrobial peptides (HAMPs) are molecules thatparticipate in the innate immune response by acting as a firstbarrier against a bacterial infection and their expression can bemodulated by different stressors including lipopolysaccharide (LPS)or bacteria [23e26].

Senegalese sole is a flatfish highly demanded in SouthernEuropean aquaculture markets. Important advances on its cultiva-tion cycle have been achieved, especially in reproduction [27] andlarval rearing and weaning [28e30]. Initial ongrowing resultsindicated that this species could be adequately cultivated at highstocking densities [15]. Nevertheless, little information about themolecular responses promoted by stocking density and ration size isavailable. This work aimed at studying the molecular responsesunleashed by both factors. For this purpose, two densities and tworations were assayed. Growth rates and cortisol levels were evalu-ated. Expression patterns of six genes involved in growth regula-tion, stress adaptation and innate immune system were quantifiedin juveniles. The importance of ration and density in Senegalese soleaquaculture is discussed.

2. Material and methods

2.1. Fish and rearing conditions

A total of 560 S. senegalensis juveniles averaging 148.32 � 50.3 gwere cultivated for 60 days. Soles were initially stocked in 0.5 m2

rectangular fibreglass tanks at two different densities: 7 and30 kgm�2. For the lowest stocking conditionwe employed six tanks(N ¼ 24 animals/tank) while for 30 kg m�2 only four tanks wereused (N¼ 104 animals/tank).Within each stocking, half of the tankswere fed a ration of 1.0 or 0.25% of total fish biomass. The rationof 1% was chosen because it was close to satiation, allowed anoptimum growth and no excess feed appeared on the bottom of thetanks. Animals were fed on dry commercial feed (LE-5 Elite fromSKRETTING Co.) six days per week. Therefore, daily siphoningout was not needed, resulting in a reduction of stress by handling.Pellets were supplied at night by automatic feeders. Mean watertemperature was 18.4 � 0.09 �C, salinity varied between 32&and 38& and pH between 8.0 and 8.2. A 12L:12D photoperiod wasfixed. During the light phase, the intensity on water surface was80 lx. A flow-through systemwas used, that ensured a water inflowsufficient to achieve 25 exchanges per day. This water exchangemaintained dissolved oxygen levels at 6.6 � 0.16 mg L�1

(86.2 � 2.02% saturation). Neither chemical nor antibiotic treat-ments were used at any time.

2.2. Sampling

Growth was monitored by measuring weight and total length at0, 20, 40 and 60 days after the commencement of the assay. Total

sample size represented 100 and 55% of the whole population (24and 60 individuals of each replicate) for densities of 7 and30 kg m�2, respectively.

To perform cortisol plasma determinations and tissue geneexpression analyses, 9 individuals of each experimental group(stocking density and ration size) were euthanized at day 60. Fishwere carefully taken from their respective tanks and transferred toa new tank containing 2-phenoxyethanol (0.5 ml L�1). Bloodsamples were taken by caudal vein puncture and immediatelycentrifuged 2 min at 1500 � g at room temperature to separate theplasma, which was kept at �80 �C until analysis. Livers and kidneyswere rapidly dissected, frozen in liquid nitrogen, and stored at�80 �C until use.

2.3. Studied growth parameters

Specific growth rates (1) were calculated from individualweights recorded on days 0, 20, 40 and 60 [31].

SGR�% day�1� ¼ ðeg � 1Þ*100

g ¼ ðln Wt � ln W0Þ=ðtt � t0Þ(1)

where, SGR¼ specific growth rate (% d�1);Wt ¼ averaged weight atsampling day (g); W0 ¼ averaged weight at day 0 (g); tt ¼ Numberof days between sampling day and day 0 (d); t0 ¼ day 0 (d).

The corresponding stocking densities (2) were also obtained.

SD ¼ ðN*WtÞ=S (2)

where, SD ¼ stocking density (Kg biomass/m2); N ¼ numberof individuals; Wt ¼ average weight at sampling day (g); S ¼surface (m2).

2.4. Cortisol determinations

The plasma levels of cortisol were measured using a commercialspecific radioimmunoassay (CORT-CT2, Iba Molecular Spain, S.A.)according to the manufacturer's instructions.

2.5. RNA isolation and gene expression analysis

Homogenization of liver and kidney (n ¼ 9 specimens/group)was carried out in the Fastprep FG120 instrument (Bio101) usingLysing Matrix D (Q-BioGene) for 40 s at speed setting 6. Total RNAwas isolated from 50 mg of S. senegalensis tissues using the RNeasyMini Kit (Qiagen). RNA isolation procedure was carried out inaccordance with the manufacturer's protocol. In all cases, total RNAwas treated twice with DNase I using the RNase-Free DNase kit(Qiagen) for 30min. RNA sample quality was checked on an agarosegel, and quantification was spectrophotometrically determined.Total RNA (1 mg) from each sample was reverse-transcribed usingthe iScript� cDNA Synthesis kit (Bio-Rad). Lack of genomic DNAcontaminationwas confirmed by PCR amplification of RNA samplesin the absence of cDNA synthesis.

Real-time PCR analysis was carried out using an iCycler (Bio-Rad). Reactions were done in a total volumen of 25 mL containingcDNA generated from 10 ng of original RNA template, 300 nM eachof specific forward and reverse primers (Table 1), and 12.5 mL of iQ�SYBR Green Supermix (Bio-Rad). Primers for IGF-I, IGF-II, HSP90AAand ubiquitin (UB52) were reported previously [22,32,33]. Splicevariants of IGF-1 were predicted and named as previously reported[34]. Sequences for HAMP1 (Acc No AB455099), g-type lysozyme(Acc No AB428773) and a putative HSP70 (Acc No AB513855) wereretrieved from GenBank/EMBL/DDBJ. Matching oligonucleotideprimers (Table 1) were designed using the Oligo v6.89 software(Medprobe). The amplification protocol used was as follows: initial

Table 1List of primers used in this study. Reverse primer for IGF-Ia, IGF-Ib and IGF-Ic was thesame. Amplicon size for each gene or transcript variant is indicated.

Target Primers (50 / 30) Fragment size (bp)

HAMP1 F TGAATCATGTCAGACACAGACGCTGGAGT 93R TGCAGCAGTAAACCCAGGCGTCA

LYSG F ACTGCTCGCGGTGAATGGGACA 95R CCTGAAAATTTATTACGGATTCGGCCAATG

IGF-Ia F GCACAAGGCGGACAAGGGCACA 144IGF-Ib F CGAGGGCACCGAAGAGACCTTTACCTG 91IGF-Ic F GCCGAGGGCACCGAAGGATGT 57

R CCCCCCGTGTTGCCTCGACTHSP70 F GCTATACCAGGGAGGGATGGAAGGAGGG 119

R CGACCTCCTCAATATTTGGGCCAGCA

Fig. 1. Plasma cortisol levels in soles initially stocked at 7 and 30 kg m�2 under 1.0(grey) and 0.25% (black) rations for 60 days. Each bar represents the mean � SEM ofnine fish for each group, with each fish assayed in duplicate. Asterisks represent

E. Salas-Leiton et al. / Fish & Shellfish Immunology 28 (2010) 296e302298

denaturation and enzyme activation for 7 min at 95 �C, followed by40 cycles of 95 �C for 15 s, and 70 �C for 30 s. Each assay was done induplicate. For normalization of cDNA loading, all samples were runin parallel with the housekeeping gene ubiquitin. Relative mRNAexpression was determined using the 2(�DDCt) method [35].

statistically significant differences between stocking densities.

2.6. Statistical analysis

Differences in growth rate due to stocking density and feedration were determined by a one-way ANOVA, taking into accountthe normality of the data distribution and the homogeneity ofvariances. Significance was established at P < 0.05.

To determine significant differences in gene expression due tostocking densities and rations, a two-way ANOVA was performed.KolmogoroveSmirnov and ShapiroeWilk testing confirmeda normal distribution of the log-transformed data. Further, theinteraction between the two main effects (P < 0.05) was assessed.

3. Results

3.1. Growth parameters and plasma cortisol levels

Table 2 depicts specific growth rates (SGR) for each density andration size (0.25% or 1.0% of biomass). No differences in SGRbetween densities at day 60 were detected (Table 2). Only soles fed0.25% at day 20 and 1.0% at day 40 (P < 0.05) showed statisticallysignificant lower SGR values under high density. As expected,animals fed 1.0% exhibited a significant higher SGR (mean w3.5-fold) than those fed 0.25%, irrespective of stocking densities.

During the experiment, no disease outbreaks or other signs ofdisease were observed. Only 3 specimens from the 30 kg m�2 and0.25% ration group died asymptomatically (M ¼ 0.9%).

Cortisol plasma levels were estimated in sole exposed todifferent stocking densities and feed rations (Fig. 1). Animals grownat 30 kg m�2 showed elevated plasma cortisol levels (w45-foldhigher than at 7 kg m�2). No differences associated to the rationswere observed. Moreover, very high intra-population variabilitywas detected in juveniles under the same experimental conditions.

Table 2Progression of specific growth rates (SGR) under different stocking densities (SD) (kg mwithin each feed ration and sampling day.

1.0% Biomass

SD SGR SD SGR

Mean SEM Mean SEM

day 0 7.45 30.97day 20 8.69 0.68 0.01 36.20 0.68 0.0day 40 10.19 0.73* 0.02 38.65 0.56* 0.0day 60 10.99 0.65 0.01 43.32 0.59 0.0

3.2. Effect of stocking density and feed rationon IGFs gene expression

Total IGF-I mRNA levels decreased significantly at 30 kg m�2 inliver but not in kidney (Fig. 2). In contrast, no differences betweenrations were found at any of the two stocking densities and tissues.

Similarly to IGF-I, IGF-II mRNA levels only decreased signifi-cantly in the liver at the highest stocking density. No significantdifferences were detected between stocking densities in kidney orbetween feeding regimes in any tissue.

To determine if the splice variants IGF-Ia, IGF-Ib and IGF-Ic couldcontribute differentially to total IGF-I mRNA levels, we quantifiedtheir transcript levels in liver (Fig. 3A). IGF-Ia was the mostexpressed splice variant with levels similar to those of total IGF-I.IGFI-b was expressed at very low levels (wthree orders of magni-tude lower than total IGF-I) and IGF-Ic expressionwas not detected.IGF-II transcript levels were approximately two orders of magni-tude lower than total IGF-I. Only IGFIa showed significant lowertranscript levels at high density (Fig. 3B).

3.3. Effect of stocking density and feed rationon HSP gene expression

HSP70 transcript levels decreased significantly at the highestdensity both in liver and kidney (Fig. 4). In contrast, no differencesassociated to rations were observed (P > 0.05). However, a signifi-cant interaction (Pinteraction¼ 0.0042) between stocking density andfeed ration was found after two-way ANOVA analysis.

HSP90 transcripts were not detected in liver at any densityor feed ration. Moreover, no differences in HSP90 mRNA levelsbetween stocking densities or rations were found in kidney(P > 0.05) (Fig. 4).

�2) and feeding rations. Asterisks denote significant differences between densities

0.25% Biomass

SD SGR SD SGR

Mean SEM Mean SEM

7.17 31.811 7.47 0.18* 0.03 31.58 �0.04* 0.052 7.72 0.17 0.02 33.76 0.14 0.041 8.13 0.21 0.03 34.99 0.17 0.03

Fig. 2. Relative gene expression of IGF-I and IGF-II in liver and kidney of soles initiallystocked at 7 and 30 kg m�2 under 1.0 (grey) and 0.25% (black) rations. Data areexpressed as the mean fold change (mean � SEM, n ¼ 9) from the calibrator group(30 kg m�2; 0.25%). Asterisks represent statistically significant differences betweenstocking densities.

Fig. 4. Relative gene expression of HSP70 and HSP90 in liver and kidney of solesinitially stocked at 7 and 30 kg m�2 under 1.0 (grey) and 0.25% (black) rations. Data areexpressed as the mean fold change (mean � SEM, n ¼ 9) from the calibrator group(30 kg m�2; 0.25%). HSP90 transcripts in liver were not amplified (N/A). Asterisksrepresent statistically significant differences between stocking densities.

E. Salas-Leiton et al. / Fish & Shellfish Immunology 28 (2010) 296e302 299

3.4. Effect of stocking density and feed ration on g-typelysozyme and HAMP gene expression

g-type lysozyme transcripts were affected by stocking densityand feed ration. In liver and kidney, g-type lysozyme mRNA levelsdecreased at the highest density (mean 1.7 and 1.6-fold in liver andkidney, respectively) and at the lowest ration (mean 2.4- and 1.5-fold in liver and kidney, respectively) (Fig. 5).

In a similar way, HAMP1 expressionwas also influenced by bothfactors. HAMP1 mRNA levels decreased significantly at the higheststocking density in liver (w5-fold) and at the lowest ration inliver and kidney. A clear interaction (Pinteraction ¼ 0.0001) between

Fig. 3. A. Relative abundance of IGF-I, IGF-Ia, IGF-Ib and IGF-II in liver at 7 kg m�2 and 1.0%initially stocked at 7 and 30 kg m�2 under 1.0 (grey) and 0.25% (black) rations. Data are(30 kg m�2; 0.25%). Asterisks represent statistically significant differences between stockin

stocking density and ration was observed for HAMP1 in liver(Fig. 5).

4. Discussion

Stocking densities and ration size are two important factorsaffecting growth in fish. Determination of optimal values for thesetwo factors is a prerequisite to warrant the economic viability ofsole production. However, commercial production must also takeinto account other factors such as animal welfare, stress and health.Hence, the effects of density and ration on the immune system andphysiology need to be elucidated. In this study, we demonstratethat feed ration is a major factor influencing growth as reported inother flatfish species [4e6]. Sole fed the 0.25% ration showeda reduced SGR (3.3-fold lower than 1% at 60 days) without affectingsurvival. Nevertheless, the effects of ration size on growth can bemodulated by factors such as temperature or feeding time [6,36]. Inthe present study, we show that stocking density can also interactwith ration to modulate growth performance. Sole grown at highdensity and fed the 0.25% ration reduced SGR (after 20 days) earlierthan those fed the 1% ration (after 40 days). In both cases, theywereable to develop a compensatory growth with no significant differ-ences in SGR between densities at 60 days. We hypothesize that thehighest stocking density could provoke a crowding stress with thesubsequent enhancement of energy requirements andmobilizationof energy stores. The limited energy provided by the low rationsize would have affected SGR earlier than the higher ration.

ration. B. Relative gene expression of spliced variants IGF-Ia and IGF-Ib in liver of solesexpressed as the mean fold change (mean � SEM, n ¼ 9) from the calibrator groupg densities.

Fig. 5. Relative gene expression of g-type lysozyme and HAMP1 in liver and kidney ofsoles initially stocked at 7 and 30 kg m�2 under 1.0 (grey) and 0.25% (black) ration.Data are expressed as the mean fold change (mean � SEM, n ¼ 9) from the calibratorgroup (30 kg m�2; 0.25%). Asterisks represent statistically significant differencesbetween stocking densities. Letters indicate statistically significant differences causedby ration.

E. Salas-Leiton et al. / Fish & Shellfish Immunology 28 (2010) 296e302300

The optimization of the metabolic rate, e.g. reducing locomotoryactivity and promoting feed ingestion, would have contributed toincrease the proportion of the energy available for growth [15,37].Our growth data and plasma cortisol levels at 60 days were similarto previous data reported for this species [15,16], suggesting thatS. senegalensis may be cultivated, in principle, at densities between7 and 30 kg m�2 being the ration size a critical factor for main-taining a satisfactory SGR. Finally, we would like to point out thehigh variation in cortisol levels observed among fish exposed toa high stocking density, as reported by Costas et al. [16], revealingimportant inter-individual responses to high density, an issue thatdeserves to be investigated.

Cortisol can interact with the growth hormone (GH)/IGF-I axisby regulating IGF-I and IGF-I mRNA expression both in vitro and invivo. Cortisol decreases IGF-I gene expression in a dose-dependentmanner in sea bream hepatocytes [38]. Moreover, intraperitonealinjection of cortisol in tilapia (Oreochromis mossambicus) reducesplasma IGF-I levels and IGF-I mRNA levels in liver [39]. In channelcatfish (Ictalurus punctatus), an oral administration of cortisol alsodecreases circulating IGF-I but without affecting IGF-I transcripts inliver [40]. Suppression of IGF-I synthesis and gene expression byglucocorticoids have also been reported in mammals in vivo and invitro [41e44]. In this study, the mRNA levels of total IGF-I, the splicevariant IGF-Ia and IGF-II decreased significantly in liver at highstocking densities, probably associated to high plasma cortisollevels. However, the reduced IGF-I transcripts did not associatewith lower SGRs. In other teleost species, a compensatory growthhas been associated to a reduction of hepatic IGF-I mRNA levels andan increase of plasma IGF-I levels [45]. Unfortunately, due to thelack of specific assays, we were not able to measure the levels ofcirculating IGF-I in sole under the present experimental conditions

to investigate if the compensatory growth observed at day 60 mayhave been due to increased plasma IGF-I levels. However, wecannot rule out the involvement of other metabolic or hormonalfactors that could compensate for maintaining similar SGR underdifferent stocking densities.

Fish growth is highly influenced by environmental conditionsand nutrition through modulation of the GH-IGF axis (includingGH, GH receptor (GHR), IGF-I and IGF-II, IGF receptors and IGFbinding proteins) [46]. IGF-I levels correlate positively with dietaryration, dietary protein content and body growth rate (reviewed in[18,46]). Moreover, starvation increases plasma GH concentrationsand reduces hepatic GH-binding sites and circulating IGF-I levels.At the genomic level, hepatic IGF-I and GHR mRNA levels bothdecrease during fasting and refeeding restores the IGF-I steady-state expression levels and even results in an increase in circulatingIGF-I levels, promoting a compensatory growth (reviewed in[46e49]). In this study, we observed clear differences in the SGRsbetween feeding regimes. However, SGR was not correlated withhepatic IGF-I or IGF-II transcript levels at any stocking density. Asdescribed above for compensatory growth, circulating IGF-I levelscan be inversely correlated with mRNA levels [45]. Although IGF-Iplasma levels were not determined in this study, we cannot rule outthe possibility that differences in SGR between feeding regimeswere paralleled by changes in IGF-I plasma levels. Furthermore,some functions of IGF-I can be affected by the expression andproduction of IGF-I receptors and IGF binding proteins [18].Moreover, IGFs regulate not only somatic growth but also they canact as mediators of other numerous physiological processes such asmetabolism, reproduction, osmoregulation and immune system inan endocrine, autocrine and/or paracrine way [50e53]. Thecomplex regulation of IGF-I synthesis, secretion and function mayhelp explain the observed lack of correlation between IGF-I tran-script levels and SGR suggesting regulation at levels than other thetranscriptional.

Physiological and cellular stress responses represent twodefensive mechanisms to maintain homeostasis. The formercomprises the hypothalamic-pituitary-interrenal (HPI) axis, mainlyinvolving the production of cortisol. The latter depends on changesin proteins and genes playing an important role in the stressprocess, such as the molecular chaperones HSP70 and HSP90[54,55]. HSPs interact with the glucocorticoid receptor regulatingits correct folding, activation, intracellular transport, transcriptionalregulation and decay [56]. Stocking density can activate bothsystems. High stocking densities increase plasma cortisol levels inmost fish [3,16,57,58]. The mRNA levels of HSP70, but not of HSP90,increase after a crowding stress [57,59]. However, expression ofHSPs can be transient and modulated by certain hormonal factors.GH and prolactin negatively regulate hepatic HSP70 mRNA levels[60] whereas cortisol can suppress the heat stress-related induc-tion of HSP70 expression in vivo [61] and of HSP90 in vitro [62].Moreover, HSP90 responds rapidly to a thermal stress, recoveringto steady-state levels just after 24 h [22]. Our results failed to showany induction of HSP gene expression. In contrast, HSP70 expres-sion in liver and kidney was significantly lower at high stockingdensity whereas no differences in HSP90 mRNAs were detected. Itis possible that the high levels of cortisol at high density measuredin this study may have modulated HSP70 expression, supportingthe notion that cortisol may inhibit HSP70 expression.

g-type lysozyme and HAMP1 are genes known to participate,among other processes, in the response against bacterial patho-gens. Previous studies from our group in Scopththamus rhombusand Pagrus auriga showed that these genes activated their tran-scription intensely and rapidly in response to a bacterial challengeor LPS injection [25,63]. In this study, we demonstrate that mRNAlevels of both genes were highly reduced by a high stocking density

E. Salas-Leiton et al. / Fish & Shellfish Immunology 28 (2010) 296e302 301

and a low ration. A long-term crowding stress has been previouslyshown to decrease some parameters of the immune systemaffecting in some cases disease resistance [3,64,65]. This effect hasbeen routinely associated to elevated levels of cortisol due to theability of corticosteroids to modify negatively some immunesystem parameters after in vivo treatments (reviewed in [66]).Moreover, this study shows that an appropriate ration is essentialfor maintaining higher expression levels of these innate immunegenes. Although high densities (higher than 20 kg m�2) arerequired to warrant the economic viability of ongrowing sole [67],the optimization of ration sizes and stocking densities to maintainadequate expression levels of genes involved in the immune systemshould be a priority to reduce the impact of bacterial outbreaks atthe final of the production cycle [68,69].

In conclusion, our results confirm that S. senegalensis is a speciesthat tolerates, in productive terms, cultivation under crowdingconditions (range 30e45 kg m2). Nevertheless, responses associ-ated to a crowding stress and rations can modify expression levelsof IGF-I, HSP70, g-type lysozyme and HAMP1. Thus, finding theequilibrium between economically feasible, animal welfare andassumable risk factors seem to be the most reasonable option tocultivate this species nowadays. Further research on feeds toimprove the immunological status of S. senegalensis is required ifcrowding cultivation conditions are applied.

Acknowledgments

This work was supported by Spain's national developmentprogram for sole farming, “Advancing in the main processes toachieve the full rearing of the Senegalese sole (Solea senegalensis)”,financed by JACUMAR (the national aquaculture advisorycommittee). Authors are especially grateful to technicians AnielaCrespo and Eugenia Zuasti for their help in the laboratory.

References

[1] Ellis T, North B, Scott AP, Bromage NR, Porter M, Gadd D. The relationshipsbetween stocking density and welfare in farmed rainbow trout. J Fish Biol2005;61:493e531.

[2] Alcorn SW, Pascho RJ, Murray AL, Shearer KD. Effects of ration level onimmune functions in chinook salmon (Oncorhynchus tshawytscha). Aquacul-ture 2003;217:529e45.

[3] Montero D, Izquierdo MS, Tort L, Robaina L, Vergara JM. High stocking densityproduces crowding stress altering some physiological and biochemicalparameters in gilthead seabream, Sparus aurata, juveniles. Fish Physiol Bio-chem 1999;20:53e60.

[4] Sæther B-S, Jobling M. The effects of ration level on feed intake and growth,and compensatory growth after restricted feeding in turbot Scophthalmusmaximus L. Aquacult Res 1999;30:647e53.

[5] Puvanendram V, Boyce DL, Brown JA. Food ration requirements of0þ yellowtail flounder Limanda ferruginea (Storer) juveniles. Aquaculture2003;220:459e75.

[6] Verbeeten BE, Carter CG, Purser GJ. The combined effect of feeding time andration on growth performance and nitrogen metabolism of greenbackflounder. J Fish Biol 1999;55:6.

[7] Irwin S, O'Halloran J, Fitzgerald RD. Stocking density, growth and growthvariation in juvenile turbot, Scophthalmus maximus (Rafinesque). Aquaculture1999;178:77e88.

[8] Kristiansen TS, Fernö A, Holm JC, Privitera L, Bakke S, Fosseidengen JE.Swimming behaviour as an indicator of low growth rate and impaired welfarein Atlantic halibut (Hippoglossus hippoglossus L.) reared at three stockingdensities. Aquaculture 2004;230:137e51.

[9] Merino GE, Piedrahita RH, Conklin DE. The effect of fish stocking density onthe growth of California halibut (Paralichthys californicus) juveniles. Aquacul-ture 2007;265:176e86.

[10] Schram E, Van der Heul JW, Kamstra A, Verdegen MCJ. Stocking density-dependent growth of Dover sole (Solea solea. Aquaculture 2006;252:339e47.

[11] Fairchild EA, Howell WH. Optimal stocking density for juvenile winterflounder Pseudopleuronectes americanus. J World Aquac Soc 2001;32:300e8.

[12] King N, Howell WH, Fairchild E. The effect of stocking density on the growth ofjuvenile summer flounder Paralichthys dentatus. In: Howell WH, Keller BJ,Park PK, Mc Vey JP, Takayanagi K, Uekita Y, editors. Proceedings of the 26thUSeJapan aquaculture symp. Durham, NH, USA: University of New HampshireSea Grant Program; 1998. p. 173e80.

[13] Herrera M, Vargas-Chacoff L, Hachero I, Ruíz-Jarabo I, Rodiles A, Navas JI, et al.Physiological responses of juvenile wedge sole Dicologoglossa cuneata (Mor-eau) to high stocking density. Aquacult Res 2009;40:790e7.

[14] Martínez-Tapia C, Fernández-Pato CA. Influence of stock density on turbot(Scophthalmus maximus L.) growth. ICES CM 1991. 1991/F:20.

[15] Salas-Leiton E, Anguís V, Manchado M, Cañavate JP. Growth, feeding andoxygen consumption of Senegalese sole (Solea senegalensis) juveniles stockedat different densities. Aquaculture 2008;285:84e9.

[16] Costas B, Aragao C, Mancera JM, Dinis MT, Conceiçao LEC. High stockingdensity induces crowding stress and affects amino acid metabolism in Sene-galese sole Solea senegalensis (Kaup 1858) juveniles. Aquacult Res 2008;39:1e9.

[17] Di Marco P, Priori A, Finoia MG, Massari A, Mandich A, Marino G. Physiologicalresponses of European sea bass Dicentrarchus labrax to different stockingdensities and acute stress challenge. Aquaculture 2008;275:319e28.

[18] Reinecke M, Bjornsson BT, Dickhoff WW, McCormick SD, Navarro I, Power DM,et al. Growth hormone and insulin-like growth factors in fish: where we areand where to go. Gen Comp Endocrinol 2005;142:20e4.

[19] Feder ME, Hofmann GE. Heat-shock proteins, molecular chaperones, and thestress response: evolutionary and ecological physiology. Annu Rev Physiol1999;61:243e82.

[20] Iwama GK, Thomas PT, Forsyth RB, Vijayan MM. Heat shock protein expressionin fish. Rev Fish Biol Fish 1998;8:35e56.

[21] Robert J. Evolution of heat shock protein and immunity. Dev Comp Immunol2003;27:449e64.

[22] Manchado M, Salas-Leiton E, Infante C, Ponce M, Asensio E, Crespo A, et al.Molecular characterization, gene expression and transcriptional regulation ofcytosolic HSP90 genes in the flatfish Senegalese sole (Solea senegalensis Kaup.Gene 2008;416:77e84.

[23] Douglas SE, Gallant JW, Liebscher RS, Dacanay A, Tsoi SC. Identification andexpression analysis of hepcidin-like antimicrobial peptides in bony fish. DevComp Immunol 2003;27:589e601.

[24] Hikima J, Minagawa S, Hirono I, Aoki T. Molecular cloning, expression andevolution of the Japanese flounder goose-type lysozyme gene, and the lyticactivity of its recombinant protein. Biochim Biophys Acta 2001;1520:35e44.

[25] Jiménez-Cantizano RM, Infante C, Martin-Antonio B, Ponce M, Hachero I,Navas JI, et al. Molecular characterization, phylogeny, and expression of c-typeand g-type lysozymes in brill (Scophthalmus rhombus). Fish Shellfish Immunol2008;25:57e65.

[26] Rodrigues PN, Vazquez-Dorado S, Neves JV, Wilson JM. Dual function of fishhepcidin: response to experimental iron overload and bacterial infection insea bass (Dicentrarchus labrax). Dev Comp Immunol 2006;30:1156e67.

[27] Anguís V, Cañavate JP. Spawning of captive Senegal sole (Solea senegalensis)under a naturally fluctuating temperature regime. Aquaculture 2005;243:133e45.

[28] Cañavate JP, Fernández-Díaz C. Influence of co-feeding larvae with live andinert diets on weaning the sole Solea senegalensis onto commercial dry feeds.Aquaculture 1999;174:255e63.

[29] Férnandez-Díaz C, Kopecka J, Cañavate JP, Sarasquete C, Solé M. Variations ondevelopment and stress defences in Solea senegalensis larvae fed on live andmicroencapsulated diets. Aquaculture 2006;251:573e84.

[30] Fernández-Díaz C, Yúfera M, Cañavate JP, Moyano FJ, Alarcón FJ, Díaz M.Growth and physiological changes during metamorphosis of Senegal solereared in the laboratory. J Fish Biol 2001;58:1e13.

[31] HoudeED, SchekterRC.Growth rates, rationsand cohort consumptionofmarinefish larvae in relation to prey concentrations. Rapports Procès-verbaux desRéunions Conseil international pour l'Exploration de la Mer 1981;178:441e53.

[32] Funes V, Asensio E, Ponce M, Infante C, Cañavate JP, Manchado M. Insulin-likegrowth factors I and II in the sole Solea senegalensis: cDNA cloning andquantitation of gene expression in tissues and during larval development. GenComp Endocrinol 2006;149:166e72.

[33] Infante C, Matsuoka MP, Asensio E, Canavate JP, Reith M, Manchado M.Selection of housekeeping genes for gene expression studies in larvae fromflatfish using real-time PCR. BMC Mol Biol 2008;9:28.

[34] Tiago DM, Laize V, Cancela ML. Alternatively spliced transcripts of Sparusaurata insulin-like growth factor 1 are differentially expressed in adult tissuesand during early development. Gen Comp Endocrinol 2008;157:107e15.

[35] Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2 (�DDCT) method. Methods 2001;25:402e8.

[36] Desai AS, Singh RK. The effects of water temperature and ration size ongrowth and body composition of fry of common carp, Cyprinus carpio. J ThermBiol 2009;34:276e80.

[37] Ali M, Nicieza A, Wootton RJ. Compensatory growth in fishes: a response togrowth depression. Fish Fish 2003;4:147e90.

[38] Leung LY, Kwong AK, Man AK, Woo NY. Direct actions of cortisol, thyroxineand growth hormone on IGF-I mRNA expression in sea bream hepatocytes.Comp Biochem Physiol A Mol Integr Physiol 2008;151:705e10.

[39] Kajimura S, Hirano T, Visitacion N, Moriyama S, Aida K, Grau EG. Dual mode ofcortisol action on GH/IGF-I/IGF binding proteins in the tilapia, Oreochromismossambicus. J Endocrinol 2003;178:91e9.

[40] Peterson BC, Small BC. Effects of exogenous cortisol on the GH/IGF-I/IGFBPnetwork in channel catfish. Domest Anim Endocrinol 2005;28:391e404.

[41] Unterman TG, Jentel JJ, Oehler DT, Lacson RG, Hofert JF. Effects of glucocorti-coids on circulating levels and hepatic expression of insulin-like growth factor

E. Salas-Leiton et al. / Fish & Shellfish Immunology 28 (2010) 296e302302

(IGF)-binding proteins and IGF-I in the adrenalectomized streptozotocin-dia-betic rat. Endocrinology 1993;133:2531e9.

[42] Delany AM, Canalis E. Transcriptional repression of insulin-like growth factor Iby glucocorticoids in rat bone cells. Endocrinology 1995;136:4776e81.

[43] Delany AM, Durant D, Canalis E. Glucocorticoid suppression of IGF I tran-scription in osteoblasts. Mol Endocrinol 2001;15:1781e9.

[44] Hammon HM, Zbinden Y, Sauerwein H, Breier BH, Blum JW, Donkin SS. Theresponse of the hepatic insulin-like growth factor system to growth hormoneand dexamethasone in calves. J Endocrinol 2003;179:427e35.

[45] Picha ME, Silverstein JT, Borski RJ. Discordant regulation of hepatic IGF-ImRNA and circulating IGF-I during compensatory growth in a teleost, thehybrid striped bass (Morone chrysopsxMorone saxatilis). Gen Comp Endocrinol2006;147:196e205.

[46] Duan C. Nutritional and developmental regulation of insulin-like growthfactors in fish. J Nutr 1998;128:306Se14S.

[47] Ayson FG, de Jesus-Ayson EG, Takemura A. mRNA expression patterns for GH,PRL, SL, IGF-I and IGF-II during altered feeding status in rabbitfish, Siganusguttatus. Gen Comp Endocrinol 2007;150:196e204.

[48] Imsland AK, Foss A, Roth B, Stefansson SO, Vikingstad E, Pedersen S, et al.Plasma insulin-like growth factor-I concentrations and growth in juvenilehalibut (Hippoglossus hippoglossus): effects of photoperiods and feedingregimes. Comp Biochem Physiol A Mol Integr Physiol 2008;151:66e70.

[49] Fukada H, Ozaki Y, Pierce AL, Adachi S, Yamauchi K, Hara A, et al. Salmongrowth hormone receptor: molecular cloning, ligand specificity, and responseto fasting. Gen Comp Endocrinol 2004;139:61e71.

[50] Maestro MA, Mendez E, Parrizas M, Gutierrez J. Characterization of insulin andinsulin-like growth factor-I ovarian receptors during the reproductive cycle ofcarp (Cyprinus carpio). Biol Reprod 1997;56:1126e32.

[51] McCormick SD. Effects of growth hormone and insulin-like growth factor I onsalinity tolerance and gill Naþ, Kþ-ATPase in Atlantic salmon (Salmo salar):interaction with cortisol. Gen Comp Endocrinol 1996;101:3e11.

[52] Norbeck LA, Kittilson JD, Sheridan MA. Resolving the growth-promoting andmetabolic effects of growth hormone: differential regulation of GH-IGF-Isystem components. Gen Comp Endocrinol 2007;151:332e41.

[53] Yada T. Growth hormone and fish immune system. Gen Comp Endocrinol2007;152:353e8.

[54] Basu N, Todgham AE, Ackerman PA, Bibeau MR, Nakano K, Schulte PM, et al.Heat shock protein genes and their functional significance in fish. Gene2002;295:173e83.

[55] Palmisano AN, Winton JR, Dickhoff WW. Tissue-specific induction of Hsp90mRNA and plasma cortisol response in chinook salmon following heat shock,seawater challenge, and handling challenge. Mar Biotechnol 2000;2:329e38.

[56] Grad I, Picard D. The glucocorticoid responses are shaped by molecularchaperones. Mol Cell Endocrinol 2007;275:2e12.

[57] Gornati R, Papis E, Rimoldi S, Terova G, Saroglia M, Bernardini G. Rearingdensity influences the expression of stress-related genes in sea bass (Dicen-trarchus labrax, L.). Gene 2004;341:111e8.

[58] Rotllant J, Tort L. Cortisol and glucose responses after acute stress by nethandling in the sparid red porgy previously subjected to crowding stress. JFish Biol 1997;51:21e8.

[59] Caipang CMA, Brinchmann MF, Berg I, Iversen M, Eliassen R, Kiron V. Changesin selected stress and immune-related genes in Atlantic cod, Gadus morhua,following overcrowding. Aquacult Res 2008;39:1533e40.

[60] Deane EE, Kelly SP, Lo CK, Woo NY. Effects of GH, prolactin and cortisol onhepatic heat shock protein 70 expression in a marine teleost Sparus sarba. JEndocrinol 1999;161:413e21.

[61] Basu N, Nakano T, Grau EG, Iwama GK. The effects of cortisol on heat shockprotein 70 levels in two fish species. Gen Comp Endocrinol 2001;124:97e105.

[62] Sathiyaa R, Campbell T, Vijayan MM. Cortisol modulates HSP90 mRNAexpression in primary cultures of trout hepatocytes. Comp Biochem Physiol BBiochem Mol Biol 2001;129:679e85.

[63] Martin-Antonio B, Jimenez-Cantizano RM, Salas-Leiton E, Infante C,Manchado M. Genomic characterization and gene expression analysis of fourhepcidin genes in the redbanded seabream (Pagrus auriga). Fish ShellfishImmunol 2009;26:483e91.

[64] Tort L, Sunyer JO, Gomez E, Molinero A. Crowding stress induces changes inserum haemolytic and agglutinating activity in the gilthead sea bream Sparusaurata. Vet Immunol Immunopathol 1996;51:179e88.

[65] Yin Z, Lam TJ, Sin YM. The effects of crowding stress on the non-specificimmuneresponse in fancy carp (Cyprinus carpio L.). Fish Shellfish Immunol1995;5:519e29.

[66] Weyts FAA, Cohen N, Flik G, Verburg-van Kemenade BML. Interactionsbetween the immune system and the hypothalamo-pituitary-interrenal axisin fish. Fish Shellfish Immunol 1999;9:1e20.

[67] García J, García B. An econometric viability model for ongrowing sole (Soleasenegalensis) in tanks using pumped well sea water. Span J Agric Res 2006;4:304e15.

[68] Zorrilla I, Balebona MC, Moriñigo MA, Sarasquete C, J.J.B. Isolation and char-acterization of the causative agent of pasteurellosis, Photobacterium damselaessp. piscicida, from sole, Solea senegalensis (Kaup). J Fish Dis 1999;22:167e72.

[69] Zorrilla I, Arijo S, Chabrillon M, Martínez-Manzanares E, Balebona MC,Moriñigo MA. Vibrio species isolated from diseased farmed sole, Solea sene-galensis (Kaup), and evaluation of the potential virulence role of their extra-cellular products. J Fish Dis 2003;26:103e8.