General and Comparative Endocrinology 210 (2015) 114–123

Contents lists available at ScienceDirect

General and Comparative Endocrinology

journal homepage: www.elsevier .com/locate /ygcen

Effects of nutritional status on plasma leptin levels and in vitroregulation of adipocyte leptin expression and secretion in rainbow trout

http://dx.doi.org/10.1016/j.ygcen.2014.10.0160016-6480/� 2014 Elsevier Inc. All rights reserved.

Abbreviations: AgRP, agouti-related protein; BL, body length; BSA, bovine serumalbumin; BW, body weight; CART, cocaine and amphetamine-regulated transcript;CF, condition factor; Ef1a, elongation factor-1a; EPA, eicosapentaenoic fatty acid;FFA, free fatty acids; GH, growth hormone; HSI, hepatosomatic index; LepA1, leptinA1 gene isoform; LepA2, leptin A2 gene isoform; MFI, mesenteric fat index; NPY,neuropeptide Y; PC, Pearson correlation; POMC, pro-opiomelanocortin; PPARc,peroxisome proliferator-activated receptor-c; qPCR, quantitative real-time PCR; RE,restricted fish; RIA, radioimmunoassay; SA, satiated fish; SGRL, standard growthrate for body length; SGRW, standard growth rate for body weight; TG, triglycerides;WGD, whole genome duplication.⇑ Corresponding author at: Department of Physiology and Immunology, Faculty

of Biology, University of Barcelona, Av. Diagonal 643, Barcelona 08028, Spain.Fax: +34 934110358.

E-mail address: ecapilla@ub.edu (E. Capilla).

Cristina Salmerón a, Marcus Johansson b, Anna R. Angotzi c, Ivar Rønnestad c, Elisabeth Jönsson b,Björn Thrandur Björnsson b, Joaquim Gutiérrez a, Isabel Navarro a, Encarnación Capilla a,⇑a Department of Physiology and Immunology, Faculty of Biology, University of Barcelona, Barcelona 08028, Spainb Fish Endocrinology Laboratory, Department of Biological and Environmental Sciences, University of Gothenburg, 40590 Gothenburg, Swedenc Department of Biology, University of Bergen, Bergen 5020, Norway

a r t i c l e i n f o

Article history:Received 2 July 2014Revised 1 October 2014Accepted 31 October 2014Available online 8 November 2014

Keywords:TeleostOncorhynchus mykissAdipose tissueGhrelinInsulinLeucine

a b s t r a c t

As leptin has a key role on appetite, knowledge about leptin regulation is important in order to under-stand the control of energy balance. We aimed to explore the modulatory effects of adiposity on plasmaleptin levels in vivo and the role of potential regulators on leptin expression and secretion in rainbowtrout adipocytes in vitro. Fish were fed a regular diet twice daily ad libitum or a high-energy diet oncedaily at two ration levels; satiation (SA group) or restricted (RE group) to 25% of satiation, for 8 weeks.RE fish had significantly reduced growth (p < 0.001) and adipose tissue weight (p < 0.001), and higherplasma leptin levels (p = 0.022) compared with SA fish. Moreover, plasma leptin levels negatively corre-lated with mesenteric fat index (p = 0.009). Adipocytes isolated from the different fish were treated withinsulin, ghrelin, leucine, eicosapentaenoic acid or left untreated (control). In adipocytes from fish fed reg-ular diet, insulin and ghrelin increased leptin secretion dose-dependently (p = 0.002; p = 0.033, respec-tively). Leptin secretion in control adipocytes was significantly higher in RE than in SA fish (p = 0.022)in agreement with the in vivo findings, indicating that adipose tissue may contribute to the circulatingleptin levels. No treatment effects were observed in adipocytes from the high-energy diet groups, neitherin leptin expression nor secretion, except that leptin secretion was significantly reduced by leucine in REfish adipocytes (p = 0.025). Overall, these data show that the regulation of leptin in rainbow trout adipo-cytes by hormones and nutrients seems to be on secretion, rather than at the transcriptional level.

� 2014 Elsevier Inc. All rights reserved.

1. Introduction

Leptin is a pleiotropic 16-kDa peptide hormone implicated inthe regulation of energy homeostasis, obesity, reproduction, boneformation, wound healing and immunity among other biological

functions (Peelman et al., 2006). Leptin in mammals is producedand secreted primarily by mature white adipocytes (Matsonet al., 1996; Zhang et al., 1994) and is considered to be a lipostaticand satiety signal, acting on hypothalamic orexigenic and anorex-igenic neurons via transmembrane receptors to regulate foodintake and energy balance (Coll et al., 2007; Harris, 2014). Whenfat mass decreases, circulating leptin is reduced, leading to stimu-lated appetite and suppressed energy expenditure (Ahima et al.,1996; Coppari and Bjørbæk, 2012).

In fish, the leptin gene was first discovered in pufferfish, Taki-fugu rubripes (Kurokawa et al., 2005), and since then, leptin Aand B isoforms (LepA and LepB) resulting from an ancient fishwhole genome duplication event (WGD; 3R) (Taylor et al., 2003;Volff, 2004) have been identified in zebrafish, Danio rerio(Gorissen et al., 2009), Japanese medaka, Oryzias latipes(Kurokawa and Murashita, 2009) and orange-spotted grouper, Epi-nephelus coioides (Zhang et al., 2012). Salmonids possess duplicatesfor both the LepA and LepB genes (Angotzi et al., 2013; Rønnestad

C. Salmerón et al. / General and Comparative Endocrinology 210 (2015) 114–123 115

et al., 2010), most probably generated by the lineage specific WGD(4R) resulting in its tetraploidization about 25–100 million yearsago (Allendorf and Thorgaard, 1984; Ohno, 1970). Nevertheless,some species such as the rainbow trout (Oncorhynchus mykiss)have later preserved only the LepB1 paralog (Angotzi et al.,2013). Leptin teleost primary structure is only 13–25% identicalto human leptin. However, predicted tertiary structures of fish lep-tin are similar to that of their mammalian counterparts (Gorissenet al., 2009; Huising et al., 2006; Kurokawa et al., 2005; Li et al.,2010; Murashita et al., 2008; Rønnestad et al., 2010), thus support-ing functional homology. In salmonids as in other teleost speciesstudied, the liver appears to be the main production site of leptin,as evidenced by high hepatic Lep gene expression (Gong et al.2013; Gorissen et al., 2009; Huising et al., 2006; Kling et al.,2012; Kurokawa et al., 2005; Kurokawa and Murashita, 2009;Murashita et al., 2008; Pfundt et al., 2009; Rønnestad et al.,2010). Although leptin is expressed at relatively low levels in adi-pose tissue of rainbow trout (Gong et al., 2013; Pfundt et al., 2009)and Atlantic salmon, Salmo salar (Rønnestad et al., 2010), it hasbeen immunohistochemically detected in primary cultured matureadipocytes of Atlantic salmon (Vegusdal et al., 2003) and in rain-bow trout adipose tissue (Pfundt et al., 2009) using mammalianantibodies.

The role of leptin as a satiety signal has been reported in mam-malian as well as non-mammalian species either administeringhomologous or heterologous leptin by different means (reviewedby Londraville et al. (2014)). In rainbow trout, short-term injec-tions with human or homologous leptin decreases food intake,and causes a reduction on the hypothalamic mRNA levels of theorexigenic neuropeptide Y (NPY), while it elevates the expressionof the anorexigenic neuropeptides pro-opiomelanocortin (POMC)A1 and A2 (Aguilar et al., 2010; Murashita et al., 2008). Moreover,in goldfish (Carassius auratus) an acute or chronic treatment withhuman leptin also decreases food intake modulating partly theorexigenic effects of NPY and orexin A (De Pedro et al., 2006;Vivas et al., 2011; Volkoff et al., 2003). Similar results with leptinreducing feeding have been reported in grass carp, Ctenopharyng-odon idellus (Li et al., 2010), Atlantic salmon (Murashita et al.,2011) and striped bass, Morone saxatilis (Won et al., 2012). Sup-porting the role of leptin suppressing appetite, increased foodintake and up-regulated mRNA levels of the orexigenic neuropep-tides NPY and agouti-related protein (AgRP), together with reducedexpression of POMC has been described in a leptin receptor-defi-cient Japanese medaka (Chisada et al., 2014). More recently also,in rainbow trout infected with a pathogenic hemoflagellate, Cryp-tobia salmositica, which causes anorexia, a reduction in food intakehas been observed associated with increased hepatic leptin expres-sion (LepA1) and plasma leptin levels, as well as decreased mRNAlevels of NPY and cocaine and amphetamine-regulated transcript(CART), and up-regulated AgRP and POMC-A2 in the hypothalamus(MacDonald et al., 2014). Furthermore, although certain contro-versy exists (reviewed by Londraville et al. (2014) and Won andBorski (2013)), a lipostatic model for food intake regulation hasbeen proposed also for fish, suggesting that the central nervoussystem senses the amount of lipid stores and modulates/adjustsfeeding behavior accordingly (Johansen et al., 2002). However, inthis regard, the regulation of leptin seems to be fundamentally dif-ferent between fish and mammals although being in both ananorexigenic peptide, as long-term fasting or restricted feedingincreases plasma leptin levels in many fish species, such as rain-bow trout, Atlantic salmon, Arctic charr (Salvelinus alpinus) and fineflounder (Paralichthys adspersus) (Frøiland et al., 2012; Fuenteset al., 2012; Johnsen et al., 2011; Kling et al., 2009, 2012;Rønnestad et al., 2010; Trombley et al., 2012).

In mammals, leptin production by adipocytes appears to be reg-ulated at the transcriptional and translational levels, but also

through storage, turnover and secretion (Lee and Fried, 2006). Sev-eral nutritional and hormonal signals have been found to modulateleptin release from isolated adipocytes, including food consump-tion (Lynch et al., 2006), leucine (Lynch et al., 2006; Roh et al.,2003), eicosapentaenoic fatty acid (EPA) (Pérez-Matute et al.,2005), insulin (Moreno-Aliaga et al., 2003; Ricci et al., 2005) andghrelin (Giovambattista et al., 2008), all of which stimulate leptinsecretion, whereas fasting (Szkudelski et al., 2004), fatty acids(e.g., oleic acid) (Margetic et al., 2002) and growth hormone (GH)(Margetic et al., 2002) appear to be leptin inhibitors. To date, noth-ing is known about the regulation of leptin expression and secre-tion by fish adipocytes.

The overall aim of this study was to further elucidate leptinendocrinology in teleost fish focusing on the role of adipose tissue,the contribution of which is uncertain due to its low expression ofleptin in comparison to the liver, by addressing two key questionsand using rainbow trout as an experimental model. Firstly, to eval-uate leptin as a potential endocrine signal of adiposity at theorganism level, two different states of adiposity were establishedin rainbow trout by means of feed restriction. Secondly, to shedlight on the hormonal and/or nutritional regulation of leptin tran-scription and secretion by the adipose tissue at the cellular level,isolated adipocytes from rainbow trout under different feedingregimes were used as an in vitro model (Albalat et al., 2005;Cruz-Garcia et al., 2009).

2. Materials and methods

2.1. Fish and in vivo experimental trial

For the in vivo studies two groups of rainbow trout were gener-ated with different levels of visceral adiposity by feeding them ahigh-energy diet once a day at two ration levels, satiation (SAgroup) or restricted (RE group) to 25% of satiation for 8 weeks.For the in vitro studies, adipocytes were isolated first from a groupof fish fed ad libitum twice daily a regular diet and secondly, fromthe two groups of fish fed the high-energy diet (SA and RE).

All animal handling procedures were approved by the Ethicsand Animal Care Committee of the University of Barcelona, follow-ing the European Union, Spanish and Catalan Governments estab-lished legislation (reference numbers CEEA 237/12 and DAAM6755).

2.1.1. Regular diet fishRainbow trout (O. mykiss, Walbaum 1792) used for the regular

diet experiment were obtained from Truites del Segre (Oliana, Lle-ida, Spain) and were held at the facilities of the University of Bar-celona (Barcelona, Spain) until sampling. A total of 48 adult fishwere equally distributed in 6 fiberglass tanks of 400 L each and,maintained in a recirculation system at 15 ± 1 �C with 12 h light:12 h dark photoperiod. The fish were acclimated during 2 weeksfed twice daily ad libitum with a commercial diet (38% protein,24% fat and 19.8 MJ/kg digestible energy; Trout evolution, DibaqDiproteg S.A., Segovia, Spain), after which the fish were anesthe-tized with MS-222 (0.1 g/L) and sacrificed by a blow to the headand medullar section. The adipose tissue was excised and the adi-pocytes isolated for the in vitro experiments, as described below inSection 2.2.

2.1.2. High-energy diet experiment: satiated and feed restricted fishRainbow trout from the fish farm Viveros de los Pirineos S.A. (El

Grado, Huesca, Spain) were held at the facilities of the University ofBarcelona (Barcelona, Spain) where the experimental trial was car-ried out. A total of 98 adult fish were distributed among 6 fiber-glass tanks of 400 L each at equal densities (16–17 trout/tank),

116 C. Salmerón et al. / General and Comparative Endocrinology 210 (2015) 114–123

maintained in a recirculation system kept at 15 ± 1 �C with 12 hlight: 12 h dark photoperiod. The fish were allowed to acclimatefor 2 weeks, during which they were once daily fed to satiationwith a commercial feed (Trout evolution, Dibaq Diproteg S.A.,Segovia, Spain). Then, fish were fasted 24 h whereupon 8 animalswere rapidly netted at time 0 (t = 0). Fish were anesthetized withMS-222 (0.1 g/L) and once opercular movements ceased and thefish began to float, they were sampled. Blood was obtained (1–2 mL per fish) by a heparinized syringe (sodium heparin 5000 IU/mL) drawn from the caudal vein. The blood obtained was quicklycentrifuged (5000 rpm, 10 min), and plasma stored at �80 �C untilanalysis. Then, the animals were quickly sacrificed by a blow to thehead and medullar section, and body weight (BW) and body length(BL) recorded. Fish were then externally sterilized with 70% etha-nol and tissue samples collected. Whole liver and visceral adiposetissue, except that surrounding the pyloric caeca, were collectedand weighed to calculate hepatosomatic index (HSI) and mesen-teric fat index (MFI), respectively. The remaining 90 fish were alsoanesthetized with MS-222 (0.1 g/L), weighed, measured, taggedintramuscularly below the dorsal fin with passive integrated tran-sponder tags (PIT-tags, ID100-B T.CRISTAL, Trovan, Madrid, Spain)after disinfecting the area with diluted povidone iodine (Betadine),and were let to recover.

Following recovery, the fish were divided into two equally-sizedgroups (three tanks each), which were fed two different rations ofcommercial feed with high energy content (40% protein, 30% fatand 22.1 MJ/kg digestible energy; Aquatex 30 HMD, Dibaq Dipro-teg S.A., Segovia, Spain) every morning. One group was fed untilvisual satiation (SA group) while the ration of the other groupwas restricted (RE group) to 25% of the feed given to the SA group.The rations were established during the first experimental week bydaily weighing of the feed eaten by the SA group, after which 25%of that amount was given to the RE group. The average daily rationestablished during week 1 was then used to feed the SA and RE fishduring weeks 2 and 3. On week 4 (t = 4), in order to monitor therationing effects on individual growth, all fish were fasted 24 h,then sedated with MS-222 (0.03–0.04 g/L), weighed and measuredand individually identified with a PIT-tag reader, after which thefish were returned to their tanks. During week 4, the procedureused during the first week was repeated to recalculate the rationto be given to the SA and RE groups during weeks 5 and 6. Duringthe last 2 weeks (weeks 7 and 8), the feed ration was recalculateddaily based on the number of animals remaining in each tank.

Fish were sampled on 7 different days over weeks 7 and 8 (t = 8)in order to minimize the stress associated with handling, and tocarry out the in vitro isolations of adipocytes. On each samplingday, 24 h-fasted fish from the SA and RE groups (n = 4–9) from asingle tank per treatment were sampled as described above fort = 0. The adipose tissue was taken to determine the MFI and usedfor the isolation of adipocytes.

2.2. Adipocyte isolation and in vitro studies

2.2.1. Regular diet experimentAdipocytes were isolated following the method described by

Albalat et al. (2005) with minor modifications. For each isolation,a pool of adipose tissue of �42 g was obtained from 7–10 fish(average weight 234.6 ± 5.5 g). The adipose tissue was kept inKrebs-Hepes buffer pre-gassed with a mixture of O2/CO2 (pH7.3–7.4), and then minced into small pieces and incubated for60 min in Krebs-Hepes buffer containing collagenase type II(130 U/mL) plus 1% bovine serum albumin (BSA) at 18 �C in a shak-ing water bath (Unitronic OR, JP Selecta, Abrera, Spain). The cellsuspension was then filtered through a 100 lm cell strainer andwashed by flotation 1–2 times with Krebs-Hepes buffer with 1%BSA and then 1–2 times with Krebs-Hepes buffer supplemented

with 2% BSA depending on the blood remaining. Finally, cells werecarefully resuspended at the desired concentration in Krebs-Hepesbuffer containing 2% BSA after counting using a Fuchs-Rosenthalchamber.

Approximately 2 million isolated adipocytes (�0.5 mL of cellcake and �0.5 mL of Krebs-Hepes buffer with 2% BSA) were incu-bated in triplicate during 3 h in polypropylene tubes in the shakingwater bath at 18 �C, in the absence (Control, C) or presence of dif-ferent concentrations of insulin at 10 nM (I10) or 100 nM (I100); orghrelin at 0.1 nM (G0.1), 1 nM (G1) or 10 nM (G10). Porcine insulin(Sigma–Aldrich, Tres Cantos, Spain) was used as piscine insulinwas not available. This source of porcine insulin has been used pre-viously on fish adipose cells and has been shown to be bioactive(Albalat et al., 2005; Bouraoui et al., 2010, 2012). Synthetic 20amino-acid octanoylated rainbow trout ghrelin (MW: 2206 Da)was obtained from the Peptide Institute Inc., Osaka. Japan. Theincubations were of 3 h as it has been demonstrated previouslythat for fish mature adipocytes this time is optimal to assess theregulatory action of hormones without compromising integrityand viability of the cells (Albalat et al., 2005). After incubation, cellsand media were quickly transferred to a sterile 1.5 mL tube, andfollowing a short centrifugation (2700g for 2 min at 4 �C) cell-freealiquots of each triplicate were immediately transferred to a newsterile 1.5 mL tube and kept at �80 �C until analysis of leptin secre-tion. After removing all the media, the remaining cells from eachtriplicate were recovered together with 1 mL of TRI Reagent solu-tion (Ambion, Alcobendas, Spain) and transferred to a singleRNase/DNase free 1.5 mL tube and kept at �80 �C until Lep geneexpression analysis. Results are the average of n = 6 independentadipocyte isolations.

2.2.2. High-energy diet experimentAdipocytes were isolated on each sampling day as described

above in Section 2.2.1. For each experiment a pool of adipose tissueof 20–30 g from 4–9 fish from each group (SA and RE) wasobtained. Approximately 2 million isolated adipocytes were incu-bated in triplicate tubes during 3 h in absence (Control, Ctrl) orpresence of different treatments: insulin 100 nM (Ins), ghrelin10 nM (Ghrel), leucine 5 mM (Leu) and EPA 100 lM (EPA). Thedose used for insulin and ghrelin was the dose found to be mosteffective in the regular diet experiment, and for leucine and EPA,the concentrations used were based on previous studies(Cammisotto et al., 2003; Murata et al., 2000; Pérez-Matute et al.,2005; Roh et al., 2003). Leucine and EPA were supplied bySigma–Aldrich (Tres Cantos, Spain). After incubation, cells andmedia were collected separately for leptin secretion and Lep mRNAexpression analyses as explained above in Section 2.2.1. Results arethe average of n = 6–7 independent adipocyte isolations, with eachtreatment performed 2–4 times.

2.3. Plasma metabolites

Plasma glucose, triglycerides (TG) and free fatty acids (FFA)were analyzed with commercially available kits according to themanufacturers’ instructions. Plasma glucose concentration wasdetermined by a glucose oxidase colorimetric method (Spinreact,Sant Esteve d’en Bas, Spain). TG and FFA concentrations were mea-sured enzymatically with the Serum Triglyceride DeterminationKit (Sigma–Aldrich, Tres Cantos, Spain) and the Non-EsterifiedFatty Acid kit (NEFA-HR2, Wako Chemicals GmbH, Neuss,Germany), respectively.

2.4. Leptin radioimmunoassay (RIA)

Leptin was measured in plasma and media using the homolo-gous RIA protocol established to perform plasma analyses by

C. Salmerón et al. / General and Comparative Endocrinology 210 (2015) 114–123 117

Kling et al. (2009). The salmonid RIA was developed using asequence of 14 amino acids as antigen, and a polyclonal antibodyraised in rabbit against that same peptide. The peptide used basedon LepA1 is identical between rainbow trout (AB354909) andAtlantic salmon (BI468126) and, shares 71% and 21% sequenceidentity with rainbow trout LepA2 (JX123129) and LepB1(JX131306), respectively (Angotzi et al., 2013). Thus, the assay pri-marily measures the LepA1 isoform, although based on sequencesimilarities some cross-reactivity might exist with LepA2, but notLepB1. The RIA was validated for leptin detection in adipocytemedia samples determining parallelism between leptin standardcurves and media dilutions (data not shown). Plasma samples wereused undiluted and media samples from isolated adipocytes werediluted 5.7 times.

2.5. Quantitative real-time PCR (qPCR)

Total RNA from isolated adipocytes was extracted with TRIReagent solution (Ambion, Alcobendas, Spain) according to themanufacturer’s protocol, and cDNA reversely transcribed using2 lg of total RNA in conjunction with oligo(dT)12–18 primer andSuperscript III (Invitrogen, Carlsbad, CA, USA) following standardprocedures. Relative expression of leptin was quantified on theCFX-96 Real-Time PCR detection system platform (Bio-Rad, Hercu-les, CA, USA) using SYBR Green (QuantiTec SYBR Green PCR kit, Qia-gen, Hilden, Germany). Each qPCR reaction comprised 12.5 ll2xSYBR Green PCR Master Mix, 300 nM of forward and reverseprimers, 100 ng cDNA template and nuclease-free water up to afinal volume of 25 ll. To visualize potential genomic amplificationin qPCR samples, the primer assay was designed in regions spanningexon-exon boundaries (LepA1 F: TTGCTCAAACCATGGTGATTAGGA,LepA1 R: GTCCATGCCCTCGATCAGGTTA; LepA2 F: TGGAAACCAAAAAGCTCCCTTCCTCTT, LepA2 R GCCTTCTATAGGCTGGTCTCCTGCA; and elongation factor-1a (Ef1a) (AF498320) F:ATTAACATTGTGGTCATTGGCCATGTC, Ef1a R: ATCTCAGCTGCTTCCTTCTCGAACTTTT). The qPCR conditions were as follows: 3 minat 95 �C, amplification for 45 cycles at 95 �C for 15 s and 60 �C for1 min. A serial dilution in nuclease-free water of cDNA derivedfrom a RNA pool of experimental samples was amplified to con-struct standard curves (2-fold dilutions) for both target and refer-ence genes. Standard curves were included in each run todetermine amplification efficiency (E) calculated as the slope fromthe plot of log RNA concentration versus threshold cycle (Ct) valuesusing the following formula: E = 10(�1/slope) � 1, which was 98.4%,90.4% and 82.9% for Ef1a, LepA1 and LepA2, respectively. However,only LepA1, but not LepA2 was sufficiently expressed in the adipo-cytes to be analyzed; therefore, we only measured LepA1. Foldchange in target gene expression was determined using the ‘‘2-ddCT’’ method (Livak and Schmittgen, 2001), using Ef1a as endog-enous reference gene after confirming that its expression did notdiffer between treatments (p > 0.05). Melting curves were recorded

Table 1Biometric parameters of satiated (SA) and feed restricted (RE) rainbow trout. Body weight (Bt = 4 and t = 8), and specific growth rate for BW (SGRW) and BL (SGRL) during three differen(n = 42–44). Different letters indicate significant differences among sampling times or betwindicate significant differences between the SA and RE groups for comparisons within sam

Sampling time (week) t = 0 t = 4

Group SA RE SA

BW (g) 205.84 ± 5.42A 204.60 ± 6.06a 362.8BL (cm) 26.90 ± 0.21A 26.83 ± 0.27a 30.0CF 1.05 ± 0.02A 1.05 ± 0.02 1.3Sampling period (weeks) P 0–4 P 4–8

Group SA RE SA

SGRW 1.97 ± 0.06A 0.49 ± 0.05⁄ 1.1SGRL 0.38 ± 0.01 0.15 ± 0.01⁄ 0.3

to evaluate specificity of amplification and lack of primer–dimers.Product specificity was also confirmed by agarose gel analysis andthe amplicons were sequence verified.

2.6. Calculations and statistical analyses

Specific growth rate for BW (SGRW) and BL (SGRL) were deter-mined according to the formulas (ln(BWf-BWi)D�1)100 and (ln(BLf-BLi)D�1)100, respectively, where i and f denote initial and final size,and D is the number of days between measurements. Three differ-ent sampling periods (P) were studied for SGRW and SGRL: P 0–4, P4–8 and P 0–8, where the number indicates the sampling week.Condition factor (CF) was calculated as (BW � BL�3)100. HSI andMFI were calculated as (X � BW�1)100, where X represents liveror adipose weight, respectively.

Differences in BW, SGRW and SGRL were analyzed with a two-way ANOVA adjusting for repeated measures using group (SA andRE) and sampling time (t = 0, t = 4 and t = 8) or period (P 0–4 andP 4–8) respectively, as fixed factors. This was followed by a one-way ANOVA and Tukey’s post hoc test for each group to analyzedifferences through sampling times and a Student’s t-test or aMann–Whitney U test between periods or groups. BL and CFdid not adhere to the ANOVA assumption for normal distribution.Therefore, these variables were analyzed using non-parametrictests, using Friedman and Wilcoxon test within each groupthrough sampling time for repeated measures, and a Mann–Whitney U test between groups at each sampling time. Compar-isons of adipose tissue and liver weight, MFI and HSI, and plasmaparameters between time (t = 0 and t = 8) or group (SA and RE) att = 8 were performed for each parameter separately by Student’st-test. Linear correlation analyses were carried out at the individ-ual level for BW, HSI and MFI versus plasma leptin using a Pear-son Correlation (PC). Differences in leptin secretion into themedia and LepA1 relative expression by treatments from isolatedadipocytes within each group were analyzed by one-way ANOVAfollowed by a Tukey’s post hoc test, and a Student’s t-testbetween groups for each treatment. In all cases, normality andhomoscedasticity were confirmed by Shapiro–Wilk and Levene’stests, respectively. Numerical values are expressed asmean ± standard error of the mean (SEM), and differences wereaccepted as statistically different at the level of p < 0.05. Statisti-cal analyses were performed using the SPSS Statistics 20 software(IBM, Chicago, USA).

3. Results

3.1. In vivo effects of nutritional status on biometric parameters

BW and BL significantly increased during the time of the exper-iment in both groups of fish, as well as the CF in SA fish (p < 0.001,Table 1). Moreover at all times, all the biometric parameters

W), body length (BL) and condition factor (CF) at different sampling times (week t = 0,t sampling periods (P 0–4, P 4–8 and P 0–8 in weeks). Data are shown as means ± SEM

een periods within the SA (upper-case) or RE (lower-case) groups; whereas asteriskspling times or periods (p < 0.05).

t = 8

RE SA RE

3 ± 7.45B 235.13 ± 5.81b⁄ 480.39 ± 10.95C 256.10 ± 6.79c⁄

0 ± 0.25B 28.00 ± 0.27b⁄ 32.49 ± 0.27C 28.68 ± 0.25c⁄

4 ± 0.02B 1.07 ± 0.02⁄ 1.40 ± 0.02B 1.08 ± 0.02⁄

P 0–8

RE SA RE

9 ± 0.13B 0.42 ± 0.18⁄ 1.61 ± 0.06 0.45 ± 0.08⁄

4 ± 0.05 0.12 ± 0.05⁄ 0.36 ± 0.02 0.13 ± 0.02⁄

A

C

A

B

*

02468

1012

t=0 t=8

Adip

ose

tissu

e (g

)

SA RE

Aa

B

b*

01234567

t=0 t=8

Live

r (g)

SA REB

Da

b*

0.0

0.5

1.0

1.5

2.0

2.5

t=0 t=8

MFI

Sampling time (week)

Aa

B

b*

0.00.20.40.60.81.01.21.41.6

t=0 t=8H

SISampling time (week)

Fig. 1. Biometric parameters of satiated (SA) and feed restricted (RE) rainbow trout. (A) Adipose tissue weight, (B) liver weight, (C) mesenteric fat index (MFI) and (D)hepatosomatic index (HSI) at two different sampling times (week t = 0 and t = 8). Data are shown as means ± SEM (n = 7–8 for t = 0 and n = 41–44 for t = 8). Different lettersindicate significant differences between sampling times within the SA (upper-case) or RE (lower-case) groups; whereas asterisks indicate significant differences between theSA and RE groups at t = 8 (p < 0.05).

Table 2Plasma parameters of satiated (SA) and feed restricted (RE) rainbow trout. Glucose,triglycerides (TG) and free fatty acids (FFA) at two different sampling times (weekt = 0 and t = 8). Data are shown as means ± SEM (n = 7–8 for t = 0 and n = 41–44 fort = 8). Different letters indicate significant differences between sampling times withinthe SA (upper-case) or RE (lower-case) groups; whereas asterisks indicate significantdifferences between the SA and RE groups at t = 8 (p < 0.05).

Sampling time (week) t = 0 t = 8

Group SA RE

Glucose (mM) 5.44 ± 0.94 4.74 ± 0.11 4.98 ± 0.15TG (mM) 1.66 ± 0.07Aa 4.15 ± 0.49B 2.55 ± 0.14b

118 C. Salmerón et al. / General and Comparative Endocrinology 210 (2015) 114–123

studied including SGRW and SGRL, were always significantly higherin SA than RE fish (p < 0.001, Table 1). SGRW data also showed thatSA fish grew significantly faster during the first experimental per-iod (P 0–4) than during the second (P 4–8) (Table 1). After 8 weeksof feeding the high-energy diet, SA fish had significantly higheradipose tissue and liver weights, MFI and HSI than the RE groupat t = 8, as well as in comparison to t = 0 (p < 0.001, Fig. 1A, B andD), except for the MFI (Fig. 1C). Furthermore, RE fish had a signifi-cantly larger liver, and higher HSI at t = 8 than at t = 0 (p < 0.001,Fig. 1B and D), although their adipose tissue weight did not changewith time (p = 0.200, Fig. 1A) and the MFI was significantly reduced(p < 0.001, Fig. 1C).

FFA (mg/dL) 9.83 ± 0.48A 16.05 ± 0.92B 11.88 ± 0.48⁄

3.2. In vivo effects of nutritional status on plasma leptin andmetabolites

Plasma leptin levels did not change with time (at t = 8 com-pared to t = 0), neither in SA nor in RE fish. However, significant

*

0

200

400

600

800

1000

1200

t=0 t=8

Plas

ma

lept

in (p

M)

Sampling time (week)

SA RE

Fig. 2. Plasma leptin levels in satiated (SA) and feed restricted (RE) rainbow trout.Data at different sampling times (week t = 0 and t = 8) are shown as means ± SEM(n = 7–8 for t = 0 and n = 42–43 for t = 8). No significant differences were observedbetween sampling times within the SA or RE groups; whereas the asterisk indicatessignificant differences between the SA and RE groups at t = 8 (p < 0.05).

differences were observed between the two groups (SA and RE)at t = 8 (p = 0.022, Fig. 2).

After 8 weeks of feeding the high-energy diet, plasma glucoselevels remained unaffected by time and nutritional treatment(Table 2). Plasma TG levels significantly increased with the high-energy diet for both groups at t = 8 in comparison to t = 0. How-ever, plasma TG levels did not differ between the SA and RE groups(Table 2). Plasma FFA levels significantly increased during theexperimental period in SA fish, and were significantly higher att = 8 when compared with RE fish (Table 2).

3.3. Plasma leptin correlations

To further evaluate the possible role of leptin as an adipositysignal, correlation analyses were carried out between plasma lep-tin levels and MFI and HSI, as well as BW. There was a significantnegative correlation between plasma leptin levels and MFI(PC = �0.282, p = 0.009, Fig. 3A). On the other hand, plasma leptinlevels correlated negatively, but not significantly, with HSI(PC = �0.153, p = 0.165, Fig. 3B) and BW (PC = �0.197, p = 0.070).

C. Salmerón et al. / General and Comparative Endocrinology 210 (2015) 114–123 119

3.4. Leptin expression and secretion in isolated adipocytes

3.4.1. In vitro regulation of leptin production in adipocytes from fish onregular diet

Leptin secretion of adipocytes isolated from fish fed regular dietincreased in a dose-dependent manner after incubation with insu-lin or ghrelin (Fig. 4A and B). The effects were significantly differ-ent in comparison to untreated control adipocytes (incubated inthe absence of peptides) at 100 nM insulin concentration(p = 0.001, Fig. 4A) and ghrelin concentration at 10 nM (p = 0.033,Fig. 4B). LepA1 gene expression remained unaffected in all condi-tions tested either with insulin or ghrelin (Fig. 4C and D), althougha trend towards a reduced expression was observed in the pres-ence of insulin (p = 0.062) (Fig. 4C).

R² = 0,080PC = -0,282

p = 0,009

0

500

1000

1500

2000

2500

0 1 2 3 4

Plas

ma

lept

in (p

M)

MFI

A B

Fig. 3. Correlations between plasma leptin levels and biometric parameters of satiated an(B) hepatosomatic index (HSI) at week t = 8. R2 of the linear regression and correlation

0

1000

2000

3000

4000

5000

6000

C I 10 I 100

Lept

inco

ncen

trat

ion

(pM

) baa

0.00.20.40.60.81.01.21.41.6

C I 10 I 100

LepA

1 re

lativ

e ex

pres

sion

Treatment

A

C

B

D

Fig. 4. Leptin secretion and expression in isolated adipocytes from regular fed rainbow trexpression of LepA1 normalized to Ef1a. Adipocytes were isolated as explained in Section(G) at 0.1, 1 or 10 nM for 3 h. Data are shown as means ± SEM (n = 5–6). Different letter

3.4.2. In vitro regulation of leptin production in adipocytes from fish onhigh-energy diet

Leptin secretion of adipocytes isolated from fish given differentrations of high-energy diet was significantly affected both, by thein vivo condition (p = 0.002) as well as the in vitro treatment(p = 0.001). Leptin secretion in vitro was significantly higher fromuntreated cells and from insulin-treated adipocytes of RE than SAfish (p = 0.022 and p = 0.005, respectively, Fig. 5A). In adipocytesfrom SA fish, leptin secretion remained unaffected by the differenttreatments (p = 0.556, Fig. 5A). In adipocytes isolated from RE fish,leptin secretion was significantly reduced by leucine in comparisonto the control treatment (p = 0.025, Fig. 5A). Nevertheless, no sig-nificant effects were observed after EPA or insulin treatments(p = 0.690 and p = 0.226, respectively, Fig. 5A). Regarding LepA1

R² = 0,023PC = -0,153

p = 0,165

0

500

1000

1500

2000

2500

0.0 0.5 1.0 1.5 2.0

Plas

ma

lept

in (p

M)

HSI

d feed restricted rainbow trout. Plasma leptin and (A) mesenteric fat index (MFI) andcoefficient (PC) and p value (p) from the Pearson correlation are indicated.

0.00.20.40.60.81.01.21.41.6

C G 0.1 G 1 G 10

LepA

1 re

lativ

eex

pres

sion

Treatment

0

1000

2000

3000

4000

5000

6000

C G 0.1 G 1 G 10

Lept

inco

ncen

trat

ion

(pM

) bababa

out. (A and B) Leptin concentration in the media and (C and D) quantitative relative2 and left untreated (C) or treated either with insulin (I) at 10 or 100 nM or ghrelin

s indicate significant differences among treatments (p < 0.05).

120 C. Salmerón et al. / General and Comparative Endocrinology 210 (2015) 114–123

mRNA expression, there was a close to significant main effect oftreatment (p = 0.051), specifically in adipocytes isolated from SAfish (Fig. 5B).

4. Discussion

In the present study, the ad libitum and restricted access tohigh-energy diet over 8 weeks created two very different trajecto-ries in terms of growth and energy balance.

The SA fish gained weight rapidly (1.6% day�1), becoming 130%heavier than at the onset of the study. The weight gain was greaterthan the length gain, resulting in the fish becoming bulkier(increased CF). This indicates that the SA fish were in a highly ana-bolic state, rapidly converting excess energy intake into soft tissuegrowth. Some of the ingested energy was being stored in the liver,with an 85% increase in relative liver size (HSI). What is notable,however, is that there was no increase in the relative mesentericfat reserve (MFI). Thus, the soft tissue growth was mostly that ofmuscle, probably by both protein and fat deposition.

In sharp contrast, the RE fish, with highly restricted access tofood, grew slower in both weight and length over the 8-week per-iod, but managed to maintain the same size proportions (CF). Someof this growth, including some energy deposition in the liver(increased HSI) was clearly at the expense of mesenteric fat. In rel-ative terms (MFI), this reserve decreased almost by half during thestudy, and was reduced even in absolute terms of mesenteric fat

0.0

0.5

1.0

1.5

2.0

2.5

Ctrl Ins Ghrel Leu EPA

LepA

1 re

lativ

e ex

pres

sion

Treatment

ab*a*

abc

cbc

0100020003000400050006000700080009000

Ctrl Ins Ghrel Leu EPA

Lept

inco

ncen

trat

ion

(pM

)

SA REA

B

Fig. 5. Leptin secretion and expression in satiated (SA) and feed restricted (RE)rainbow trout isolated adipocytes. (A) Leptin concentration in the media and (B)quantitative relative expression of LepA1 normalized to Ef1a of isolated adipocytesfrom SA and RE fed rainbow trout at week t = 8. Adipocytes were isolated asexplained in Section 2 and left untreated (Ctrl) or treated either with insulin 100 nM(Ins), ghrelin 10 nM (Ghrel), leucine 5 mM (Leu) or EPA 100 lM (EPA) for 3 h. Dataare shown as means ± SEM (n = 6–7 with 2–4 treatment replicates per group).Different letters indicate significant differences among treatments within the SA(upper-case) or RE (lower-case) groups; whereas asterisks indicate significantdifferences between the SA and RE groups within each treatment (p < 0.05).

mass. This indicates that the visceral adipose tissue was beingcatabolized in the RE fish, while in the SA fish, this tissue was ina normal, anabolic growth phase.

4.1. Nutritional modulatory effects on plasma leptin

In the current study, plasma leptin levels in the RE fish were sig-nificantly higher than in the SA fish, which is in agreement withprevious studies in rainbow trout and other teleost species, whereplasma leptin levels increased with fasting or feed restriction(Frøiland et al., 2012; Fuentes et al., 2012; Johnsen et al., 2011;Kullgren et al., 2013; Trombley et al., 2012); and supports thehypothesis of Fuentes and colleagues (2012), who suggested thatthe increased leptin levels observed in fish during periods of foodshortage may be linked to a survival behavior, lowering appetiteand limiting energy-wasting foraging activity. Similarly, plasmaleptin levels in the burbot (Lota lota) increase in the wild afterspawning, when the energy stores of the fish are small(Mustonen et al., 2002). An exception to these observations is theabsence of effect of a reduced ration (60% of fully fed controls) inplasma leptin levels reported in one study performed in Atlanticsalmon, despite differences in growth rate (Rønnestad et al.,2010), which could be due to differences in animal size or durationof the feeding trial. The increase in the levels of plasma leptin infish in response to situations where food is limited or absent isopposite to that observed in many mammalian species, includingmice, rats and humans, where leptin levels decrease during fasting,often interpreted as a consequence of decreased adipose tissuemass, as this is the principal leptin-secreting tissue (Ahima et al.,1996; Coppari and Bjørbæk, 2012). In the subantarctic fur seal (Arc-tocephalus tropicalis) leptin is also reduced in response to pro-longed fasting (Verrier et al., 2012), whereas in the northernelephant seal (Mirounga angustirostris) no effect of fasting on circu-lating leptin levels is found (Ortiz et al., 2001), in both cases how-ever, without correlation observed with regards to the adipositylevel; therefore, reflecting differences in the role of leptin in fatreserves between terrestrial and marine mammals. Furthermore,in hibernating mammalian species such as the mink (Mustelavison), the raccoon dog (Nyctereutes procyonoides), the woodchuck(Marmota monax) or the common shrew (Sorex araneus), the high-est plasma leptin levels are found in animals with the lowest bodyadiposity (Nieminen et al., 2000, 2002; Nieminen and Hyvärinen,2000; Concannon et al., 2001). Thus, it can be hypothesized thatsome fundamental differences may exist in the regulation ofplasma leptin levels between ‘‘continuously’’ feeding animals,and species that naturally may experience seasonal periods oflow food availability, such as salmonids, marine mammals or ter-restrial hibernating mammals.

In line with these previous observations, plasma leptin levelscorrelate inversely with the relative amount of mesenteric fat(MFI). Although this correlation only explains a small part(R2 = 0.08) of the variation in leptin levels, it supports the generalconclusion that plasma leptin levels are higher in salmonids undercatabolic conditions than in those with increased adiposity; as pre-viously seen in Atlantic salmon where plasma leptin levels nega-tively correlated with body lipid content (Trombley et al., 2012).Contrary to these observations, a positive correlation has beendescribed between circulating leptin levels and visceral adiposetissue and belly flap lipid content in juvenile rainbow trout treatedwith GH, although the GH-treated fish had lower CF than thesham-treated controls (Kling et al., 2012). Therefore, it may bethe nutritional status and/or the interplay of metabolic hormoneswhich determines these relationships, rather than adiposity perse. Moreover, it should be noted that in mammals, plasma leptinlevels are not only regulated as a function of long-term adiposity,but rapid regulatory mechanisms also exist, seen as fast (hours)

C. Salmerón et al. / General and Comparative Endocrinology 210 (2015) 114–123 121

changes in plasma leptin levels in response to short-term changesin nutritional status (e.g., fasting and re-feeding) (Chelikani et al.,2004; Ishioka et al., 2005). Recent studies have shown short-termleptin responses also in fish, such us an increase in the hepaticexpression of LepA genes in common carp (Cyprinus carpio) andzebrafish after exposure to hypoxic conditions (Bernier et al.,2012; Chu et al., 2010), and in the Mozambique tilapia (Oreochr-omis mossambicus) after an hyperosmotic challenge (Baltzegaret al., 2014).

4.2. Regulation of leptin expression and secretion in isolatedadipocytes

In teleosts, leptin mRNA expression is low in adipose tissue, rel-ative to the liver expression (Douros et al., 2014; Gorissen et al.,2009; Huising et al., 2006; Kurokawa and Murashita, 2009;Kurokawa et al., 2005; Murashita et al., 2008). On the other hand,the liver is only about 1% of salmonid body mass and, data on thespecific contribution of the different tissues to the circulating lev-els of plasma leptin in fish are still missing. Notwithstanding, thepresent study demonstrates for the first time that isolated rainbowtrout adipocytes can express and secrete leptin as well as respondto nutritional and hormonal stimuli.

Adipocytes isolated from rainbow trout fed regular dietresponded to insulin and ghrelin incubations by increasing leptinsecretion. A similar stimulatory effect has been found in acutelytreated mammalian isolated adipocytes, although the extent ofthe response was lower when compared with cultured adipocytesincubated for a period of 24–48 h (Giovambattista et al., 2008;Moreno-Aliaga et al., 2003; Ricci et al., 2005). It has to be consid-ered that, compared to mammals, the incubation temperatures offish cells is normally lower in order to mimic in vivo temperatures,which translates into reduced metabolic rates (e.g., lipolysis)(Albalat et al., 2005). Changes in plasma insulin levels affect leptinmRNA expression in adipose tissue and plasma leptin levels, e.g., instreptozotocin-treated rats, where reduced leptin mRNA expres-sion was quickly reversed by insulin administration or, during fast-ing and re-feeding (MacDougald et al., 1995; Patel et al., 1998).Hepatic LepA1 and LepA2 expression increases transiently in com-mon carp following the postprandial plasma glucose peak (Huisinget al., 2006), suggesting that leptin expression can be regulatedindirectly by hormones such as insulin and cholecystokinin. Inthe present study, although insulin stimulated leptin secretionafter the 3 h of incubation time, LepA1 mRNA expression was unaf-fected. In agreement with our results, in 3T3-L1 adipocytes, insulinhas been shown to promote leptin secretion by enhancing leptinsynthesis and/or vesicle trafficking without affecting gene tran-scription in the first 2 h of incubation (Wang et al., 2014). There-fore, the current findings suggest that in rainbow trout, insulincould act as in mammals, regulating leptin post-transcriptionallyeither by inducing stabilization of the mRNA (Moreno-Aliagaet al., 2003) or by increasing adipocyte leptin secretion from theendoplasmic reticulum (Barr et al., 1997).

Regarding ghrelin, nothing is known about its role controllingleptin in fish. Previous studies in rainbow trout in vivo have dem-onstrated that ghrelin appears to act through the anorexigenic cor-ticotropin-releasing hormone (CRH) neurons to suppress foodintake (Jönsson et al., 2010). Nonetheless contrary to mammals,fasting or feed restriction decrease plasma ghrelin levels (Jönssonet al., 2007), but increase circulating leptin (Kling et al., 2009). Inmammals, the incubation of cultured rat adipocytes for 24–48 hwith ghrelin causes an increase in leptin gene expression andsecretion (Giovambattista et al., 2008); and on the other hand lep-tin inhibits in vivo both the secretion of gastric ghrelin and thestimulation of feeding by ghrelin (Kalra et al., 2005). Althoughwe could not see significant differences in terms of LepA1 mRNA

expression due to the limitations of our in vitro system (i.e., freshlyisolated mature adipocytes), altogether, the results suggest thatinsulin, ghrelin and leptin may interact also in rainbow trout tomaintain energy homeostasis, and this interrelationship seems tobe reflected also in vitro.

Adipocytes isolated from fish fed high-energy diet appear toretain a ‘‘metabolic memory’’ of the energy balance of the fish.Thus, adipocytes isolated from the catabolic RE fish secrete moreleptin than adipocytes from the anabolic SA fish, both under con-trol conditions and after insulin treatments, echoing the differentplasma leptin levels of these fish groups. This observation supportsthe idea that the in vitro results reflect what occurs in vivo, and val-idates the use of the present experimental protocol to study theregulation of leptin mRNA expression and secretion in fish. More-over, it indicates that the adipose tissue contributes to the circulat-ing levels of leptin, although it is not known to what extent.Irrespective of whether the plasma leptin originates from adiposetissue as suggested by the current data, or from the liver as sug-gested by several other studies on fish (Douros et al., 2014), or fromboth tissues, it is clear that in the present study, plasma levelsincrease during feed restriction while both hepatic and adipose tis-sues are being reduced in size. Therefore, leptin secretion from anyof these tissues would be regulated rather than constitutive, andmechanisms such as post-transcriptional stabilization of the mRNAor peptide release from storage sites within the cells in response todifferent factors (e.g., catabolic signals) should be considered andexplored.

In the present study, it also appears that adipocytes isolatedfrom SA fish show a certain degree of resistance to the hormonalor nutritional stimuli studied, as similar LepA1 mRNA and medialeptin levels were detected in all treatments. This agrees with thehypothesis that animals fed with a high-energy diet ad libitummay develop metabolic disorders, which include the loss of abilityto respond to hormonal or nutritional external stimuli to regulateleptin, as observed in rodent models (Ceddia, 2005). On the otherhand, adipocytes isolated from RE fish secreted leptin in a treat-ment-specific manner. Intriguingly, when the adipocytes wereincubated with either insulin or ghrelin, the stimulatory effectsobserved in adipocytes from the regular diet fish were lost. Theseresults suggest that these hormones may partly lose their secretoryeffects when the animal is in a period of feed restriction, whichmay reflect what occurs in vivo, as both, insulin and ghrelindecrease in plasma in response to fasting in rainbow trout(Jönsson et al., 2007; Navarro and Gutiérrez, 1995). Moreover,when adipocytes from RE fish were incubated with nutrients suchas leucine, a decrease in leptin secretion occurred, contrary to whatis known in mammals (Lynch et al., 2006; Murata et al., 2000;Pérez-Matute et al., 2005; Roh et al., 2003). It can be speculatedthat increases in leucine or EPA levels could be interpreted at thecellular level as increased food availability, causing a decrease inleptin secretion. In turn, this would increase appetite and therebystimulate the feeding/foraging behavior of the fish.

At a transcriptional level, none of the treatments modifiedLepA1 mRNA in either group, indicating that they have less modu-latory effects on leptin transcription than secretion in fish adipo-cytes. In primary cultures of rat adipocytes and 3T3-L1 cells, EPAstimulates leptin gene expression and secretion (Pérez-Matuteet al., 2005), although other studies have reported that dietary n-3 fatty acids decrease leptin mRNA expression, maybe actingthrough the nuclear receptor peroxisome proliferator-activatedreceptor-c (PPARc) (Reseland et al., 2001). Therefore, the regula-tory role of fatty acids such as EPA on leptin production in fishdeserves to be further explored.

Altogether, the current findings demonstrate that the adiposetissue of rainbow trout produces leptin, contributing to someextent to the plasma levels of this hormone, which therefore may

122 C. Salmerón et al. / General and Comparative Endocrinology 210 (2015) 114–123

be in turn linked to the degree of adiposity. Furthermore, this lep-tin production from adipocytes seems to be under the influence ofthe nutritional history of the fish and is modulated at a post-tran-scriptional level by nutrients (e.g., leucine) and other importantmetabolic/appetite regulating hormones such as insulin andghrelin.

Authors’ contributions

I.R., E.J., B.T.B., J.G., I.N. and E.C. conceived the study. C.S., M.J.,A.R.A. and E.C. performed the in vivo and in vitro experimentsand the laboratory analyses. All authors analyzed the data, wrote,read and approved the final manuscript.

Acknowledgments

The authors thank Maryam Asaad and Parnaz Nik for their assis-tance during sampling. We thank Viveros de los Pirineos S.A. (ElGrado, Huesca, Spain) and Truites del Segre (Oliana, Lleida, Spain)for the rainbow trout used in these studies and the personnel fromthe animal facilities in the Faculty of Biology for its maintenance.EC is a Ramón y Cajal researcher fellow from the Spanish ‘‘Ministe-rio de Ciencia e Innovación’’ (MICINN). This work was supported byfunds from the MICINN (AGL2010-17324 and AGL2011-24961),the Catalonian Government (2009SGR-00402) and through the‘‘Xarxa de Referència d’R+D+I en Aqüicultura’’, the SwedishResearch Council for Environment, Agricultural Sciences and Spa-tial Planning (Formas) grant 223-2011-1356, and the 7th Frame-work Program of the European Union (project LIFECYCLE FP7-222719).

References

Aguilar, A.J., Conde-Sieira, M., Polakof, S., Míguez, J.M., Soengas, J.L., 2010. Centralleptin treatment modulates brain glucosensing function and peripheral energymetabolism of rainbow trout. Peptides 31, 1044–1054.

Ahima, R.S., Prabakaran, D., Mantzoros, C., Qu, D., Lowell, B., Maratos-Flier, E., Flier,J.S., 1996. Role of leptin in the neuroendocrine response to fasting. Nature 382,250–252.

Albalat, A., Gutiérrez, J., Navarro, I., 2005. Regulation of lipolysis in isolatedadipocytes of rainbow trout (Oncorhynchus mykiss): the role of insulin andglucagon. Comp. Biochem. Physiol. A Mol. Integr. Physiol. 142, 347–354.

Allendorf, F.W., Thorgaard, G.H., 1984. Tetraploidy and the evolution of salmonidfishes. In: Turner, B.J. (Ed.), Evolutionary Genetics of Fishes. Springer, US, pp. 1–53.

Angotzi, A.R., Stefansson, S., Nilsen, T.O., Rathore, R.M., Rønnestad, I., 2013.Molecular cloning and genomic characterization of novel leptin-like genes insalmonids provide new insight into the evolution of the leptin gene family. Gen.Comp. Endocrinol. 187, 48–59.

Baltzegar, D.A., Reading, B.J., Douros, J.D., Borski, R.J., 2014. Role for leptin inpromoting glucose mobilization during acute hyperosmotic stress in teleostfishes. J. Endocrinol. 220, 61–72.

Barr, V.A., Malide, D., Zarnowski, M.J., Taylor, S.I., Cushman, S.W., 1997. Insulinstimulates both leptin secretion and production by rat white adipose tissue.Endocrinology 138, 4463–4472.

Bernier, N.J., Gorissen, M., Flik, G., 2012. Differential effects of chronic hypoxia andfeed restriction on the expression of leptin and its receptor, food intakeregulation and the endocrine stress response in common carp. J. Exp. Biol. 215,2273–2282.

Bouraoui, L., Capilla, E., Gutiérrez, J., Navarro, I., 2010. Insulin and insulin-likegrowth factor I signaling pathways in rainbow trout (Oncorhynchus mykiss)during adipogenesis and their implication in glucose uptake. Am. J. Physiol.Regul. Integr. Comp. Physiol. 299, R33–R41.

Bouraoui, L., Cruz-Garcia, L., Gutiérrez, J., Capilla, E., Navarro, I., 2012. Regulation oflipoprotein lipase gene expression by insulin and troglitazone in rainbow trout(Oncorhynchus mykiss) adipocyte cells in culture. Comp. Biochem. Physiol. AMol. Integr. Physiol. 161, 83–88.

Cammisotto, P.G., Gélinas, Y., Deshaies, Y., Bukowiecki, L.J., 2003. Regulation ofleptin secretion from white adipocytes by free fatty acids. Am. J. Physiol.Endocrinol. Metab. 285, E521–E526.

Ceddia, R.B., 2005. Direct metabolic regulation in skeletal muscle and fat tissue byleptin: implications for glucose and fatty acids homeostasis. Int. J. Obes. 29,1175–1183.

Chelikani, P., Ambrose, J., Keisler, D., Kennelly, J., 2004. Effect of short-term fastingon plasma concentrations of leptin and other hormones and metabolites indairy cattle. Domest. Anim. Endocrinol. 26, 33–48.

Chisada, S.-i., Kurokawa, T., Murashita, K., Rønnestad, I., Taniguchi, Y., Toyoda, A.,Sakaki, Y., Takeda, S., Yoshiura, Y., 2014. Leptin receptor-deficient (knockout)medaka, Oryzias latipes, show chronical up-regulated levels of orexigenicneuropeptides, elevated food intake and stage specific effects on growth andfat allocation. Gen. Comp. Endocrinol. 195, 9–20.

Chu, D.L., Li, V.W., Yu, R.M., 2010. Leptin: clue to poor appetite in oxygen-starvedfish. Mol. Cell. Endocrinol. 319, 143–146.

Coll, A.P., Farooqi, I.S., O’Rahilly, S., 2007. The hormonal control of food intake. Cell129, 251–262.

Concannon, P., Levac, K., Rawson, R., Tennant, B., Bensadoun, A., 2001. Seasonalchanges in serum leptin, food intake, and body weight in photoentrainedwoodchucks. Am. J. Physiol. Regul. Integr. Comp. Physiol. 281, R951–R959.

Coppari, R., Bjørbæk, C., 2012. Leptin revisited: its mechanism of action andpotential for treating diabetes. Nat. Rev. Drug Dis. 11, 692–708.

Cruz-Garcia, L., Saera-Vila, A., Navarro, I., Calduch-Giner, J., Pérez-Sánchez, J., 2009.Targets for TNFa-induced lipolysis in gilthead sea bream (Sparus aurata L.)adipocytes isolated from lean and fat juvenile fish. J. Exp. Biol. 212, 2254–2260.

De Pedro, N., Martínez-Álvarez, R., Delgado, M.J., 2006. Acute and chronic leptinreduces food intake and body weight in goldfish (Carassius auratus). J.Endocrinol. 188, 513–520.

Douros, J.D., Baltzegar, D.A., Breves, J.P., Lerner, D.T., Seale, A.P., Gordon Grau, E.,Borski, R.J., 2014. Prolactin is a major inhibitor of hepatic leptin A synthesis andsecretion: studies utilizing a homologous leptin A ELISA in the tilapia. Gen.Comp. Endocrinol. 207, 86–93.

Frøiland, E., Jobling, M., Björnsson, B.T., Kling, P., Ravuri, C.S., Jørgensen, E.H., 2012.Seasonal appetite regulation in the anadromous Arctic charr: evidence for a roleof adiposity in the regulation of appetite but not for leptin in signallingadiposity. Gen. Comp. Endocrinol. 178, 330–337.

Fuentes, E.N., Kling, P., Einarsdottir, I.E., Alvarez, M., Valdés, J.A., Molina, A.,Björnsson, B.T., 2012. Plasma leptin and growth hormone levels in the fineflounder (Paralichthys adspersus) increase gradually during fasting and declinerapidly after refeeding. Gen. Comp. Endocrinol. 177, 120–127.

Giovambattista, A., Gaillard, R.C., Spinedi, E., 2008. Ghrelin gene-related peptidesmodulate rat white adiposity. Vitam. Horm. 77, 171–205.

Gong, N., Einarsdottir, I.E., Johansson, M., Björnsson, B.T., 2013. Alternative splicevariants of the rainbow trout leptin receptor encode multiple circulating leptinbinding proteins. Endocrinology 154, 2331–2340.

Gorissen, M., Bernier, N.J., Nabuurs, S.B., Flik, G., Huising, M.O., 2009. Two divergentleptin paralogues in zebrafish (Danio rerio) that originate early in teleosteanevolution. J. Endocrinol. 201, 329–339.

Harris, R., 2014. Direct and indirect effects of leptin on adipocyte metabolism.Biochim. Biophys. Acta 1842, 414–423.

Huising, M.O., Geven, E.J.W., Kruiswijk, C., Nabuurs, S.B., Stolte, E.H., Spanings, F.A.T.,Verburg-van Kemenade, B.M.L., Flik, G., 2006. Increased leptin expression incommon carp (Cyprinus carpio) after food intake but not after fasting or feedingto satiation. Endocrinology 147, 5786–5797.

Ishioka, K., Hatai, H., Komabayashi, K., Soliman, M., Shibata, H., Honjoh, T., Kimura,K., Saito, M., 2005. Diurnal variations of serum leptin in dogs: effects of fastingand re-feeding. Vet. J. 169, 85–90.

Johansen, S., Ekli, M., Jobling, M., 2002. Is there lipostatic regulation of feed intake inAtlantic salmon Salmo salar L.? Aquacult. Res. 33, 515–524.

Johnsen, C.A., Hagen, Ø., Adler, M., Jönsson, E., Kling, P., Bickerdike, R., Solberg, C.,Björnsson, B.T., Bendiksen, E.Å., 2011. Effects of feed, feeding regime and growthrate on flesh quality, connective tissue and plasma hormones in farmed Atlanticsalmon (Salmo salar L.). Aquaculture 318, 343–354.

Jönsson, E., Forsman, A., Einarsdottir, I.E., Kaiya, H., Ruohonen, K., Björnsson, B.T.,2007. Plasma ghrelin levels in rainbow trout in response to fasting, feeding andfood composition, and effects of ghrelin on voluntary food intake. Comp.Biochem. Physiol. A Mol. Integr. Physiol. 147, 1116–1124.

Jönsson, E., Kaiya, H., Björnsson, B.T., 2010. Ghrelin decreases food intake in juvenilerainbow trout (Oncorhynchus mykiss) through the central anorexigeniccorticotropin-releasing factor system. Gen. Comp. Endocrinol. 166, 39–46.

Kalra, S.P., Ueno, N., Kalra, P.S., 2005. Stimulation of appetite by ghrelin is regulatedby leptin restraint: peripheral and central sites of action. J. Nutr. 135, 1331–1335.

Kling, P., Rønnestad, I., Stefansson, S.O., Murashita, K., Kurokawa, T., Björnsson, B.T.,2009. A homologous salmonid leptin radioimmunoassay indicates elevatedplasma leptin levels during fasting of rainbow trout. Gen. Comp. Endocrinol.162, 307–312.

Kling, P., Jönsson, E., Nilsen, T.O., Einarsdottir, I.E., Rønnestad, I., Stefansson, S.O.,Björnsson, B.T., 2012. The role of growth hormone in growth, lipid homeostasis,energy utilization and partitioning in rainbow trout: interactions with leptin,ghrelin and insulin-like growth factor I. Gen. Comp. Endocrinol. 175, 153–162.

Kullgren, A., Jutfelt, F., Fontanillas, R., Sundell, K., Samuelsson, L., Wiklander, K.,Kling, P., Koppe, W., Larsson, D.G., Björnsson, B.T., 2013. The impact oftemperature on the metabolome and endocrine metabolic signals in Atlanticsalmon (Salmo salar). Comp. Biochem. Physiol. A Mol. Integr. Physiol. 164, 44–53.

Kurokawa, T., Murashita, K., 2009. Genomic characterization of multiple leptingenes and a leptin receptor gene in the Japanese medaka, Oryzias latipes. Gen.Comp. Endocrinol. 161, 229–237.

Kurokawa, T., Uji, S., Suzuki, T., 2005. Identification of cDNA coding for a homologueto mammalian leptin from pufferfish, Takifugu rubripes. Peptides 26, 745–750.

Lee, M.-J., Fried, S.K., 2006. Multilevel regulation of leptin storage, turnover, andsecretion by feeding and insulin in rat adipose tissue. J. Lipid Res. 47, 1984–1993.

C. Salmerón et al. / General and Comparative Endocrinology 210 (2015) 114–123 123

Li, G.-G., Liang, X.-F., Xie, Q., Li, G., Yu, Y., Lai, K., 2010. Gene structure, recombinantexpression and functional characterization of grass carp leptin. Gen. Comp.Endocrinol. 166, 117–127.

Livak, K.J., Schmittgen, T.D., 2001. Analysis of relative gene expression data usingreal-time quantitative PCR and the 2�DDCT method. Methods 25, 402–408.

Londraville, R.L., Macotela, Y., Duff, R.J., Easterling, M.R., Liu, Q., Crespi, E.J., 2014.Comparative endocrinology of leptin: assessing function in a phylogeneticcontext. Gen. Comp. Endocrinol. 203, 146–157.

Lynch, C.J., Gern, B., Lloyd, C., Hutson, S.M., Eicher, R., Vary, T.C., 2006. Leucine infood mediates some of the postprandial rise in plasma leptin concentrations.Am. J. Physiol. Endocrinol. Metab. 291, E621–E630.

MacDonald, L.E., Alderman, S.L., Kramer, S., Woo, P.T.K., Bernier, N.J., 2014.Hypoxemia-induced leptin secretion: a mechanism for the control of foodintake in diseased fish. J. Endocrinol. 221, 441–455.

MacDougald, O.A., Hwang, C.-S., Fan, H., Lane, M.D., 1995. Regulated expression ofthe obese gene product (leptin) in white adipose tissue and 3T3-L1 adipocytes.Proc. Natl. Acad. Sci. U.S.A. 92, 9034–9037.

Margetic, S., Gazzola, C., Pegg, G.G., Hill, R.A., 2002. Leptin: a review of its peripheralactions and interactions. Int. J. Obes. 26, 1407–1433.

Matson, C.A., Wiater, M.F., Weigle, D.S., 1996. Leptin and the regulation of bodyadiposity: a critical review. Diab. Rev. 4, 488–508.

Moreno-Aliaga, M.J., Stanhope, K.L., Gregoire, F.M., Warden, C.H., Havel, P.J., 2003.Effects of inhibiting transcription and protein synthesis on basal and insulin-stimulated leptin gene expression and leptin secretion in cultured ratadipocytes. Biochem. Biophys. Res. Commun. 307, 907–914.

Murashita, K., Uji, S., Yamamoto, T., Rønnestad, I., Kurokawa, T., 2008. Production ofrecombinant leptin and its effects on food intake in rainbow trout(Oncorhynchus mykiss). Comp. Biochem. Physiol. B Biochem. Mol. Biol. 150,377–384.

Murashita, K., Jordal, A.-E.O., Nilsen, T.O., Stefansson, S.O., Kurokawa, T., Björnsson,B.T., Moen, A.-G.G., Rønnestad, I., 2011. Leptin reduces Atlantic salmon growththrough the central pro-opiomelanocortin pathway. Comp. Biochem. Physiol. AMol. Integr. Physiol. 158, 79–86.

Murata, M., Kaji, H., Takahashi, Y., Iida, K., Mizuno, I., Okimura, Y., Abe, H., Chihara,K., 2000. Stimulation by eicosapentaenoic acids of leptin mRNA expression andits secretion in mouse 3T3-L1 adipocytes in vitro. Biochem. Biophys. Res.Commun. 270, 343–348.

Mustonen, A.-M., Nieminen, P., Hyvärinen, H., 2002. Leptin, ghrelin, and energymetabolism of the spawning burbot (Lota lota, L.). J. Exp. Zool. 293, 119–126.

Navarro, I., Gutiérrez, J., 1995. Fasting and starvation. In: Hochachka, P.W.,Mommsen, T.P. (Eds.), Biochemistry and Molecular Biology of Fishes, vol. 4.Elsevier Science, pp. 393–434.

Nieminen, P., Hyvärinen, H., 2000. Seasonality of leptin levels in the BAT of thecommon shrew (Sorex araneus). Z. Naturforsch. C 55, 455–460.

Nieminen, P., Hyvärinen, H., Käkelä, R., Asikainen, J., 2000. Plasma leptin andthyroxine of mink (Mustela vison) vary with gender, diet and subchronicexposure to PCBs. Comp. Biochem. Physiol. A Mol. Integr. Physiol. 127, 515–522.

Nieminen, P., Mustonen, A.-M., Asikainen, J., Hyvärinen, H., 2002. Seasonal weightregulation of the raccoon dog (Nyctereutes procyonoides): interactions betweenmelatonin, leptin, ghrelin, and growth hormone. J. Biol. Rhythms 17, 155–163.

Ohno, S., 1970. Evolution by Gene Duplication. George Alien & Unwin Ltd., London,Springer-Verlag, Berlin, Heidelberg, New York.

Ortiz, R.M., Noren, D.P., Litz, B., Ortiz, C.L., 2001. A new perspective on adiposity in anaturally obese mammal. Am. J. Physiol. Endocrinol. Metab. 281, E1347–E1351.

Patel, B.K., Koenig, J.I., Kaplan, L.M., Hooi, S.C., 1998. Increase in plasma leptin andLep mRNA concentrations by food intake is dependent on insulin. Metabolism47, 603–607.

Peelman, F., Couturier, C., Dam, J., Zabeau, L., Tavernier, J., Jockers, R., 2006.Techniques: new pharmacological perspectives for the leptin receptor. TrendsPharmacol. Sci. 27, 218–225.

Pérez-Matute, P., Marti, A., Martínez, J.A., Fernandez-Otero, M.P., Stanhope, K.L.,Havel, P.J., Moreno-Aliaga, M.J., 2005. Eicosapentaenoic fatty acid increases

leptin secretion from primary cultured rat adipocytes: role of glucosemetabolism. Am. J. Physiol. Regul. Integr. Comp. Physiol. 288, R1682–R1688.

Pfundt, B., Sauerwein, H., Mielenz, M., 2009. Leptin mRNA and proteinimmunoreactivity in adipose tissue and liver of rainbow trout (Oncorhynchusmykiss) and immunohistochemical localization in liver. Anat. Histol. Embryol.38, 406–410.

Reseland, J.E., Haugen, F., Hollung, K., Solvoll, K., Halvorsen, B., Brude, I.R., Nenseter,M.S., Christiansen, E.N., Drevon, C.A., 2001. Reduction of leptin gene expressionby dietary polyunsaturated fatty acids. J. Lipid Res. 42, 743–750.

Ricci, M.R., Lee, M.-J., Russell, C.D., Wang, Y., Sullivan, S., Schneider, S.H., Brolin, R.E.,Fried, S.K., 2005. Isoproterenol decreases leptin release from rat and humanadipose tissue through posttranscriptional mechanisms. Am. J. Physiol.Endocrinol. Metab. 288, E798–E804.

Roh, C., Han, J., Tzatsos, A., Kandror, K.V., 2003. Nutrient-sensing mTOR-mediatedpathway regulates leptin production in isolated rat adipocytes. Am. J. Physiol.Endocrinol. Metab. 284, E322–E330.

Rønnestad, I., Nilsen, T.O., Murashita, K., Angotzi, A.R., Gamst Moen, A.-G.,Stefansson, S.O., Kling, P., Björnsson, B.T., Kurokawa, T., 2010. Leptin andleptin receptor genes in Atlantic salmon: cloning, phylogeny, tissue distributionand expression correlated to long-term feeding status. Gen. Comp. Endocrinol.168, 55–70.

Szkudelski, T., Lisiecka, M., Nowicka, E., Kowalewska, A., Nogowski, L., Szkudelska,K., 2004. Short-term fasting and lipolytic activity in rat adipocytes. Horm.Metab. Res. 36, 667–673.

Taylor, J.S., Braasch, I., Frickey, T., Meyer, A., Van de Peer, Y., 2003. Genomeduplication, a trait shared by 22000 species of ray-finned fish. Genome Res. 13,382–390.

Trombley, S., Maugars, G., Kling, P., Björnsson, B.T., Schmitz, M., 2012. Effects oflong-term restricted feeding on plasma leptin, hepatic leptin expression andleptin receptor expression in juvenile Atlantic salmon (Salmo salar L.). Gen.Comp. Endocrinol. 175, 92–99.

Vegusdal, A., Sundvold, H., Gjøen, T., Ruyter, B., 2003. An in vitro method forstudying the proliferation and differentiation of Atlantic salmon preadipocytes.Lipids 38, 289–296.

Verrier, D., Atkinson, S., Guinet, C., Groscolas, R., Arnould, J.P., 2012. Hormonalresponses to extreme fasting in subantarctic fur seal (Arctocephalus tropicalis)pups. Am. J. Physiol. Regul. Integr. Comp. Physiol. 302, R929–R940.

Vivas, Y., Azpeleta, C., Feliciano, A., Velarde, E., Isorna, E., Delgado, M.J., De Pedro, N.,2011. Time dependent effects of leptin on food intake and locomotor activity ingoldfish. Peptides 32, 989–995.

Volff, J.N., 2004. Genome evolution and biodiversity in teleost fish. Heredity 94,280–294.

Volkoff, H., Eykelbosh, A.J., Peter, R.E., 2003. Role of leptin in the control of feeding ofgoldfish Carassius auratus: interactions with cholecystokinin, neuropeptide Yand orexin A, and modulation by fasting. Brain Res. 972, 90–109.

Wang, Y., Ali, Y., Lim, C.-Y., Hong, W., Pang, Z.P., Han, W., 2014. Insulin-stimulatedleptin secretion requires calcium and PI3K/Akt activation. Biochem. J. 458, 491–498.

Won, E.T., Borski, R.J., 2013. Endocrine regulation of compensatory growth in fish.Front. Endocrinol. 4, 74.

Won, E.T., Baltzegar, D.A., Picha, M.E., Borski, R.J., 2012. Cloning and characterizationof leptin in a Perciform fish, the striped bass (Morone saxatilis): control offeeding and regulation by nutritional state. Gen. Comp. Endocrinol. 178, 98–107.

Zhang, H., Chen, H., Zhang, Y., Li, S., Lu, D., Zhang, H., Meng, Z., Liu, X., Lin, H., 2012.Molecular cloning, characterization and expression profiles of multiple leptingenes and a leptin receptor gene in orange-spotted grouper (Epinepheluscoioides). Gen. Comp. Endocrinol. 181, 295–305.

Zhang, Y., Proenca, R., Maffei, M., Barone, M., Leopold, L., Friedman, J.M., 1994.Positional cloning of the mouse obese gene and its human homologue. Nature372, 425–432.