General and Comparative Endocrinology 175 (2012) 407–415

Contents lists available at SciVerse ScienceDirect

General and Comparative Endocrinology

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

Subfunctionalization of POMC paralogues in Senegalese sole (Solea senegalensis)

Yvette S. Wunderink a,b,⇑, Erik de Vrieze b, Juriaan R. Metz b, Silke Halm c, Gonzalo Martínez-Rodríguez c,Gert Flik b, Peter H.M. Klaren b, Juan M. Mancera a

a Department of Biology, Faculty of Marine and Environmental Sciences, University of Cádiz, 11510 Puerto Real, Cádiz, Spainb Department of Organismal Animal Physiology, Institute for Water and Wetland Research, Faculty of Science, Radboud University Nijmegen, Heyendaalseweg 135, 6525 AJNijmegen, The Netherlandsc Instituto de Ciencias Marinas de Andalucía (CSIC), Apartado Oficial, 11510 Puerto Real, Cádiz, Spain

a r t i c l e i n f o

Article history:Received 14 July 2011Revised 15 November 2011Accepted 16 November 2011Available online 28 November 2011

Keywords:EvolutionGene duplicationPOMCSolea senegalensisStressSubfunctionalization

0016-6480/$ - see front matter � 2011 Elsevier Inc. Adoi:10.1016/j.ygcen.2011.11.026

⇑ Corresponding author at: Department of BioloEnvironmental Sciences, University of Cádiz, 11510 P+34 956 016019.

E-mail address: yvette.wunderink@uca.es (Y.S. Wu

a b s t r a c t

The precursor protein proopiomelanocortin (POMC) gives rise to a variety of biologically active peptidesthrough cell-specific posttranslational processing. Two transcripts of pomc were found in the flatfish Soleasenegalensis (ssePOMC-A and ssePOMC-B), that most likely represent subfunctionalized paralogues: sse-POMC-A lacks the N-terminal cleavage site for b-MSH, whereas ssePOMC-B cannot yield ACTH and com-pletely lacks the opioid consensus sequence in the b-END region. An analysis of nucleotide substitutionrates shows that the POMC-derived peptides possess well-conserved regions under purifying selection,except the b-END derived from POMC-B, which has undergone positive selection. The calculated Ks valuesfor ssePOMC-A versus ssePOMC-B and zebrafish POMCa versus zebrafish POMCb are 0.40 and 0.72,respectively, indicating that the zebrafish POMC paralogues started to evolve almost twice as early inevolution, and that the Solea POMC paralogues arose independently from the whole genome duplicationevent that gave rise to the zebrafish paralogues. This makes ssePOMC-B the first identified POMCa ortho-logue that lacks the opioid consensus. Furthermore, pomc-a expression is down-regulated in chronicstressed S. senegalensis juveniles, whereas pomc-b expression levels remain unaffected, indicating differ-ent physiological roles for both POMC paralogues. The distribution of functional POMC-derived peptidehormones over two pomc genes in S. senegalensis suggests subfunctionalization of the paralogues, a rel-evant notion when studying POMC function in endocrine responses.

� 2011 Elsevier Inc. All rights reserved.

1. Introduction

Proopiomelanocortin (POMC) is a precursor protein that givesrise to a number of biologically active peptide hormones and iswidely studied in vertebrate lineages. Posttranslational processingof POMC is cell-type-specific and depends on proteolytic cleavageby prohormone convertases (PC1/3 and PC2). In the corticotrophsof the pituitary pars distalis, PC1/3 cleaves POMC to yieldadrenocorticotropic hormone (ACTH), b-lipotropin (b-LPH) andN-terminal peptide. In the melanotrophs located in the pituitarypars intermedia, PC2 is present as well, and processes ACTH furtherto a-melanocyte stimulating hormone (a-MSH) and corticotropin-like intermediate lobe peptide (CLIP), and cleaves b-LPH to b-MSHand b-endorphin (b-END). Furthermore, in some species, POMC isalso expressed in the ventral hypothalamus, where a-MSH andCLIP derive from ACTH, next to b-MSH and b-END from b-LPH[10,27,33].

ll rights reserved.

gy, Faculty of Marine anduerto Real, Cádiz, Spain. Fax:

nderink).

The pomc gene is well-conserved among vertebrates, and thepresence of two paralogues has been reported for carp, zebrafishand flounder [2,15,20,31]. Duplications are not uncommon in theevolution of the pomc gene family. In fact, it is thought that theplethora of biologically active POMC peptides has arisen from sev-eral duplication events. The pomc gene is considered to have orig-inated from an ancestral opioid-coding gene after the first round ofwhole genome duplication (1R). Following the second genomeduplication event (2R), the c-MSH sequence was gained, but later,among the jawed vertebrates, three trends in pomc evolution havebeen proposed of which the loss of the c-MSH sequence for theray-finned fish is one [12,13]. A third genome duplication event(3R) occurred only in the lineage of the ray-finned fish around350 MYA, which most likely gave rise to duplicate copies of thepomc gene [24]. Following a gene duplication, one paralogue canaccumulate nonsense mutations, lose its function and become apseudogene, or can be deleted completely from the genome [36].However, mutations can also lead to neofunctionalization whenone of the paralogues obtains a new function. In addition, subfunc-tionalization occurs when the function of the original gene is di-vided among the two copies [10,14,26]. At least in some teleostspecies, POMC sequences can be subdivided in POMCa and POMCb.

408 Y.S. Wunderink et al. / General and Comparative Endocrinology 175 (2012) 407–415

In general, POMCb is characterized by a loss of a functional b-END,which remains conserved in the POMCa paralogues and it has beenproposed that they may have undergone subfunctionalization [10].

In fish, POMC-derived peptides exert a variety of functions. Theregulation of skin pigmentation is mostly mediated by a-MSH, b-END is involved in the brain’s reward system, and ACTH plays animportant role in the neuroendocrine stress response. In the latter,hypothalamic corticotropin-releasing hormone (CRH) stimulatesthe release of ACTH, secreted from the corticotroph cells in the parsdistalis, which activates the interrenal cells of the head kidney toproduce and release cortisol, which is considered to be the mainstress hormone [37]. In fish aquaculture, chronic exposure tostressors can lead to allostatic overload, which negatively affectsreproduction, growth and immune functions, leading to diseasesand reduced animal welfare [4]. Hence, the characterization ofPOMC as a key player in the neuroendocrine stress response, willcontribute to understand these processes [21,37].

The Senegalese sole (Solea senegalensis Kaup) is a flatfish withhigh economic value in North Africa and the Iberian Peninsulaand it is cultured at a commercial scale [11,19]. In the last fewyears, several studies have focussed on various physiological pro-cess of this species, including the stress response [3,9,25,29,38].Previous studies on this species have shown that high stockingdensity imposes stress as indicated by elevated plasma cortisol lev-els and an upregulation of crh gene expression [8,29,38].

In this study, two POMC paralogues were characterized in S.senegalensis, and their structural differences and role in the chronicstress response, viz. high stocking density, investigated. We discussfrom an evolutionary perspective the hypothesis that subfunction-alized POMC-derived peptide hormones are distributed over twoparalogue S. senegalensis pomc genes.

Table 1Primer oligonucleotide sequences used for RQ-PCR analysis.

0 0

2. Material and methods

2.1. Animals and experimental design

Animals and experimental design used in this study were iden-tical to those described by Wunderink et al. [38] and are as follow-ing: Senegalese sole juveniles (S. senegalensis Kaup) with anaverage body weight of 163 ± 30 g were provided by Planta de Cul-tivos Marinos (CASEM, University of Cádiz, Puerto Real (Cádiz),Spain), transferred to the facilities at the Faculty of Marine andEnvironmental Sciences (Puerto Real (Cádiz), Spain) and accli-mated in 200-L tanks with continuous seawater inflow at 18 �C un-der natural photoperiod conditions (October–November, 2007).Fish were fed a daily ration of 1% of their estimated body weightand fasted for 24 h before sampling. Fish were kept at two densi-ties: ‘low density’ (LD, 1.9 kg/m2) and ‘high density’ (HD, 9.8 kg/m2); each density was performed in two tanks to avoid tank ef-fects. Stocking densities were chosen according to previous studieson flatfish [8,17]. After 33 days, 12 individuals were sampled fromeach stocking density. Animals were anaesthetized in 0.1% (v/v)2-phenoxyethanol (Sigma–Aldrich, St. Louis, USA). After spinaltransection, pituitary tissues were sampled. All samples wereimmediately frozen in liquid nitrogen and stored at �80 �C untilfurther analysis. All experimental procedures complied with theGuidelines of the European Union (2010/63/EU) and theSpanish legislation (RD 1201/2005 and law 32/2007) for the useof laboratory animals.

Gene Primer Sequence (5 ? 3 )

pomc-a qssePA_Fw AAGGCAAAGAGGCGTTGTATqssePA_Rv TTCTTGAACAGCGTGAGCAG

pomc-b qssePB_Fw GTCGAGCAACACAAGTTCCAqssePB-Rv GTCAGCTCGTCGTAGCGTTT

b-actin qsseBact_Fw TCTTCCAGCCATCCTTCCTCGqsseBact_Rv TGTTGGCATACAGGTCCTTACGG

2.2. RNA extraction and cDNA synthesis

Total RNA was extracted using the commercial kit RNeasy�

(Qiagen) according to manufacturer’s instructions. Incubation withDNAse I (Qiagen) eliminated potential genomic DNA contamination.

RNA concentrations were measured by spectrophotometry and thequality of the extract was ensured using the Agilent RNA 6000 Picoor Nano Assay Kit on an Agilent 2100 Bioanalyzer (Agilent Technol-ogies). First strand cDNA synthesis was carried out using theqScript™ cDNA Synthesis kit (Quanta BioSciences).

2.3. Cloning and cDNA library screening

PCR was carried out on S. senegalensis brain cDNA with primersdesigned on conserved regions of common sole (DQ845108), floun-der-I (AF184066), flounder-II (AF191593) and European seabass(AY691808) pomc sequences. Two PCR products were generated,both partial sequences of ssepomc which were used as probes forscreening a kZAP cDNA library of S. senegalensis brain to obtainthe complete coding sequences of pomc-a and pomc-b. XL1-BlueMRF’ and SOLR bacteria were used for in vivo excision of positivek phages. Plasmid DNA was isolated by use of QIAprep miniprepkit (Qiagen) and outsourced for sequencing.

2.4. Real-time quantitative PCR

Primers for use in real-time quantitative PCR (RQ-PCR) were de-signed by use of Primer3 software (v. 0.4.0.) available at http://fok-ker.wi.mit.edu/primer3/input.htm, accessed in July 2011 [28].Primer oligonucleotide sequences are shown in Table 1. To performRQ-PCR reactions, 8 ll cDNA, forward and reverse primer and 10 llPerfeCta™ SYBR� Green Fastmix™ (Quanta BioSciences) was used.RQ-PCR (10 min at 95 �C, 40 cycles of 30 s 95 �C and 45 s 60 �C) wasperformed on a Mastercycler� EPgradient S RealPlex2 (Eppendorf).S. senegalensis b-actin (accession number: DQ485686) remainedstable during the experiment and was used as an internal controlamplicon. A calibrator sample was measured on every RQ-PCRplate to correct for inter-assay differences. PCR amplificationefficiencies were similar for all genes, all between 1.9 and 2.1.Expression levels R were calculated with the DDCt method [23].

2.5. Sequences

Sequences were retrieved from the NCBI protein database(www.ncbi.nlm.nih.gov/protein, accessed in July 2011) and theensemble genome browser (release 62). Accession numbers forthe amino acid sequences used in this study: S. senegalensis (Sene-galese sole) POMC-A (FR851915) and POMC-B (FR851916), Homosapiens (human) POMC (NP_000930), Gallus gallus (chicken) POMC(NP_001026269), Danio rerio (zebrafish) POMC-A (AAI33875) andPOMC-B (CAQ15072), Paralichthys olivaceus (Japanese flounder)POMC-I (AAG16978) and POMC-II (AAG28378), Solea solea (com-mon sole) POMC (ABI26718), Verasper moseri (barfin flounder)POMC-A (BAB18467), POMC-B (BAB18468) and POMC-C (BAB18469), Salmo salar (Atlantic salmon) POMC-A1 (NP_001185504)and POMC-A2 (ABF60231), Cyprinus carpio (common carp)POMC-I (CAA74968) and POMC-II (CAA74967), Xenopus tropicalis(western clawed frog) POMC (AAH88054), Canis familiaris (dog)POMC (XP_549460), Rattus novergicus (rat) POMC (NP_647542),Mus musculus (mouse) POMC (NP_032921), Tetraodon nigroviridis

Y.S. Wunderink et al. / General and Comparative Endocrinology 175 (2012) 407–415 409

POMC-1 (ENSTNIP00000003749) and POMC-2 (ENSTNIP00000004097), Takifugu rubripes (Fugu) POMC-1 (ENSTRUP00000024485)and POMC-2 (ENSTRUP00000036308) and gasterosteus aculeatus(stickleback) POMC-1 (ENSGACP00000012552) and POMC-2(ENSGACP00000013340). Sequences of closely related proorphanin(PNOC) were used as outgroup: X. tropicalis PNOC (NP-001106534),M. musculus PNOC (AAI44866) and H. sapiens PNOC (CAG46763).

2.6. Phylogenetic and evolutionary analyses

Multiple sequence alignments were made using the ClustalWsoftware [22] and POMC proteins of key species. Phylogenetic anal-ysis of POMC amino acid sequences was done using MEGA4 soft-ware [32] with neighbour-joining algorithm, based on amino aciddifferences (p-distance) and complete deletion. Reliability of thetree was assessed by bootstrapping (10,000 replications). To char-acterize the nature of selective pressure on POMC-derived peptides,the ratio of the number of non-synonymous substitutions per non-synonymous site (Ka) to the number of synonymous substitutionsper synonymous site (Ks) was calculated using the MEGA4 software[32]. A nucleotide substitution is synonymous when it does not re-sult in a codon coding for a different amino acid. That is to say, ami-no acid sequences are well-conserved when most substitutions aresynonymous, and the Ka/Ks ratio <1, designated purifying selection.On the other hand, non-synonymous substitutions change the ami-no acid encoded by the original codon. Selective pressure acts onamino acid sequences, and thus are only affected by non-synony-mous substitutions. Ka/Ks � 1 is considered as neutral selection[18]. Synonymous substitutions in a population occur and are fixedin a population at constant rate, since no selective pressure is actingon these nucleotides [16]. Therefore, Ks values are also calculated toestimate the age of pomc duplication events. Selective pressure wascalculated based on nucleotide alignments of POMC derivates,corrected manually to overt mismatches, guided by the correspond-

Fig. 1. Nucleotide and deduced amino acid sequences of S. senegalensis POMC-A cDNA. Tdisplayed above the nucleotide sequence. The predicted start of the N-terminal peptideThe K to G165 substitution is boxed. Accession number: FR851915.

ing amino acid alignments. Comparisons were made with the clo-sely related Japanese flounder, zebrafish (resultant of 3R) andhuman (resultant of 2R) POMC sequences.

2.7. Statistical analyses

Differences between two treatments at same sample point weretested using a Student’s t-test for unpaired data. Statistical signifi-cance was accepted at P < 0.05. Ka/Ks ratios were statistically ana-lyzed by a Z-test, testing the null hypothesis H0: Ka/Ks = 1, i.e.whether the level of purifying or positive selection is different fromneutral selection. Statistical significance was accepted at P < 0.05.

3. Results

3.1. Cloning and characterization of S. senegalensis pomc-a and pomc-b

Fig. 1 shows the full-length nucleotide and deduced amino acidsequence of the ssepomc-a precursor peptide. The pomc-a transcriptcontains an open reading frame encoding a 216 amino acid protein,including sequences for a signal peptide (M1-S21), N-terminal pep-tide (Q22-T74), ACTH (S76-M116) which contains the sequences ofa-MSH (S76-G90) and CLIP (P95-M116), b-LPH (E120-K216) and b-END(Y185-K216). The N-terminal cleavage site of b-MSH contains aLys165-to-Gly165 substitution compared to general POMC cleavageconsensus. The complete coding sequence of ssepomc-b is presentedin Fig. 2. A precursor protein of 186 amino acids is derived from thenucleotide sequence and includes a signal peptide (M1-N22), N-terminal peptide (Q23-A83), a-MSH (A86-G99), b-MSH (D143-S159)and a degenerated b-END (Y162-R186). In the predicted cleavage sitebetween ACTH and b-LPH/b-MSH, a substitution in the cleavageconsensus of Arg123 to Gln123 is observed.

he start and stop codon are presented in bold. The deduced amino acid sequence is(Q22-W24) is presented in bold capitals. The predicted cleavage sites are underlined.

410 Y.S. Wunderink et al. / General and Comparative Endocrinology 175 (2012) 407–415

Conserved regions of the ssePOMC peptides were identifiedbased on a multiple sequence alignment comparing deduced ami-no acid sequences of several vertebrate POMC peptides. These re-gions showed that POMC-derived peptides are well-conservedamong vertebrates (Fig. 3, see Section 2.5. for protein accessioncodes).

3.2. Phylogenetic and evolutionary analyses of POMC peptides

Phylogenetic analysis of the POMC sequences of key speciesshow two main branches of teleost POMCs: POMCa and POMCb[10] (Fig. 4). POMCb is characterized by a degenerated b-END se-quence in which the opioid signal (YGGF) is absent. Changes incleavage sites are also observed in zebrafish, Tetraodon, and fuguPOMCbs. Both sequences of ssePOMC cluster closely together inthe POMCa branch. ssePOMC-B amino acid sequence shows a97% identity to S. solea POMC, both containing a degenerated b-END sequence that lacks the opioid signal. However, phylogeneticanalysis does not classify ssePOMC-B and S. solea POMC as POMCb;both peptides are clustered as a separate branch in the POMCagroup.

To characterize the selective force that acts on the genomic se-quence of the various POMC-derived peptides, the Ka/Ks ratios werecalculated compared to species from different branches in the phy-logenetic analysis (Tables 2 and 3). For ssePOMC-A, all Ka/Ks valuescalculated for POMC-derived peptides are <1 (Table 2). Most of thePOMC-derived peptides are under strong purifying selection,especially ACTH and a-MSH. Only the ssePOMC-A signal peptidesequence is under neutral selection, except when compared tozebrafish POMCa. Derivates of ssePOMC-B are much less conservedbased on their Ka/Ks ratios (Table 3). Ka/Ks ratios of b-END of

Fig. 2. Nucleotide and deduced amino acid sequences of S. senegalensis POMC-B cDNA. Tdisplayed above the nucleotide sequence. The predicted start of the N-terminal peptideThe R to Q123 substitution is boxed. Accession number: FR851916.

ssePOMC-B are even >1 compared to orthologues of Japaneseflounder-II and human, indicating a positive selective force.

The calculated Ks values for ssePOMC-A versus ssePOMC-B andzebrafish POMCa versus zebrafish POMCb are 0.40 and 0.72 respec-tively, indicating that zebrafish POMC paralogues contain almostdouble the number of synonymous substitutions per synonymoussite as ssePOMC paralogues do.

3.3. Effects of chronic stress on pituitary pomc-a and b expression

When S. senegalensis juveniles are kept at high stocking density(HD) for 33 days, a chronic stressor known to significantly increaseplasma cortisol and crh expression levels [38], pomc-a expressionwas significantly decreased by 45% compared to control animalskept at low density (LD). Expression levels of pomc-b were not af-fected by stocking density (Fig. 5A and B).

4. Discussion

In this study, we demonstrate the existence of two paraloguesof the precursor peptide POMC in S. senegalensis: pomc-a andpomc-b. Sequence analysis of the two genes points towards a sub-functionalization comparable to POMCa and b. The genome dupli-cations result, following neighbour-joining analysis of POMCamino acid sequences, in three distinct clades of POMCs. Twoclades belong to the teleostean fishes, the third to tetrapods.Although the differentiation between POMCa and b resulted froma whole genome duplication event, clustering of the ssePOMC par-alogues suggests that this specific duplication occurred more re-cently. Interestingly, one of these paralogues also displays a lossof functional b-END (characteristic for POMCb), and likely reflects

he start and stop codon are presented in bold. The deduced amino acid sequence is(Q23-W25) is presented in bold capitals. The predicted cleavage sites are underlined.

H

Fig. 3. Multiple alignment of deduced amino acid sequences of POMC. The melanocortin (HFRW) and opioid (YGGF) signals are underlined. The predicted cleavage sites arepresented in bold. Deviating amino acid substitutions are boxed. See Section 2.5 for protein accession codes.

Y.S. Wunderink et al. / General and Comparative Endocrinology 175 (2012) 407–415 411

subfunctionalization in the Solea POMC system. This raises thequestion whether or not ssePOMC-B should be classified as aPOMCb. Lack of the opioid consensus and thus a non-functionalb-END favours classification as a POMCb. Nevertheless, it has notbeen reported thus far that duplicated paralogues in the POMCa

clade present a b-END region lacking the opioid signal [6]. Thenucleotide substitution rate of ssePOMC genes is almost twice aslow compared to zebrafish POMC genes. Assuming that the nucle-otide substitution rate is constant over time, and equal in both lin-eages, we conclude that the differentiation of sole paralogues arose

JapaneseflounderIBrednuolf

nifraB

SenegalesesoleA

Tigerpufferfish 1

Green sp

ottedpuffe

rfish1

Senegalese

soleB

Common sole

ArednuolfnifraB

Japanese flounder II

Stickleback 1

Atlantic salmon A1Atlantic

salmonA2

ZebrafishA

CommoncarpI

CommoncarpII

Tropical clawedfrog

Chicken

Human

Dog

Rat

Mouse

Zebrafish B

Tiger pufferfish 2

Green spotted pufferfish 2

Stickleback 2

Cred

nuolf

nifra

BTropical claw

edfrog

PNOC

HumanPNOC

Mouse

PNOC

0.1

Placental mammals

Birds

Amphibians

Teleost fishes

>95

70-95

POMCα

POMCβ

Fig. 4. Phylogenetic tree of vertebrate POMC sequences using neighbour-joining analysis based on amino acid difference (p-distance). Reliability of the tree was assessed bybootstrapping (10,000 replications). See Section 2.5 for protein accession codes.

412 Y.S. Wunderink et al. / General and Comparative Endocrinology 175 (2012) 407–415

independently from the whole genome duplication that gave riseto the zebrafish paralogues. Roughly estimated, the zebrafishduplication occurred almost twice as early in evolution, makingthe ssePOMC-B the first POMCa orthologue characterized with anon-functional b-END.

Sequence differences between the two forms of ssePOMC arealso seen in conspicuous endoproteolytic cleavage sites. Indeed,the C-terminal cleavage site of ssePOMC-B ACTH is substituted.POMC’s cleavage sites, consisting of the amino acids lysine andarginine, are highly conserved. It has been shown that both aminoacids are required for PC1/3 and 2 to act on POMC [7]. Hence, it is

very likely that any substitution of lysine or arginine in the cleav-age sites directly inhibit the capacity of PC1/3 and 2 to cleave sse-POMC-B into ACTH and CLIP. Thus, ssePOMC-B probably not onlyhas a non-functional b-END, but also a non-functional ACTHsequence. Altered cleavage sites are also seen in ssePOMC-A, aslysine of the N-terminal cleavage site of b-MSH has been substi-tuted for glycine. This indicates that b-MSH cannot derive fromssePOMC-A. Although ssePOMC-A cannot produce b-MSH, and sse-POMC-B does not produce a functional b-END and cannot yieldACTH and CLIP, both ssePOMCs together can potentially yield allPOMC-derived peptides, indicating a distribution of functional

Table 2Ka/Ks ratios for Senegalese sole POMC-A versus zebrafish, Japanese flounder and human POMCs for POMC-derived peptides. Ka and Ks values were calculated with MEGA4 software[32]. Z-tests were conducted to determine purifying or positive selection.

Senegalese sole POMC-A Zebrafish POMCa Zebrafish POMCb Japanese flounder POMC-I (a) Japanese flounder POMC-II (a) Human POMC

Signal peptide 0.63⁄ 0.71 0.28 0.62 0.83N-terminal peptide 0.43⁄ 0.64⁄ 0.36⁄ 0.33⁄ 0.60⁄

ACTH 0.11⁄ 0.46⁄ 0.13⁄ 0.13⁄ 0.30⁄

a-MSH 0.00⁄ 0.11⁄ 0.00⁄ 0.00⁄ 0.11⁄

CLIP 0.14⁄ 0.68 0.23⁄ 0.20⁄ 0.44⁄

b-LPH 0.31⁄ 0.71⁄ 0.31⁄ 0.44⁄ 0.66⁄

b-MSH n/c 0.56⁄ 0.25⁄ 0.40⁄ 0.50⁄

b-END 0.33⁄ 0.72 0.10⁄ 0.25⁄ 0.77

Asterisks indicate statistical significance (P < 0.05). Protein accession numbers are shown in Section 2.5.

Table 3Ka/Ks ratios for Senegalese sole POMC-B versus zebrafish, Japanese flounder and human POMCs for POMC-derived peptides. Further information as described in Table 2. Asterisksindicate statistical significance (P < 0.05).

Senegalese sole POMC-B Zebrafish POMCa Zebrafish POMCb Japanese flounder POMC-I (a) Japanese flounder POMC-II (a) Human POMC

Signal peptide 0.65⁄ 1.17 0.51 0.97 1.06N-terminal peptide 0.65⁄ 1.17 0.51 0.97 1.06ACTH 0.21⁄ 0.39⁄ 0.32⁄ 0.22⁄ 0.43⁄

a-MSH 0.16⁄ 0.18⁄ 0.22⁄ 0.16⁄ 0.05⁄

CLIP 0.25⁄ 0.58⁄ 0.41 0.25⁄ 0.73⁄

b-LPH 0.59⁄ 0.69⁄ 0.72 0.40⁄ 0.97⁄

b-MSH 0.24⁄ 0.42⁄ 0.15⁄ 0.17⁄ 1.22b-END 0.86 0.86 1.39⁄ 0.68 1.18

B

A

Fig. 5. Expression levels of pomc-a (A) and pomc-b (B) of S. senegalensis juvenileskept at a low density (LD) or high density (HD) for 33 days. Asterisks indicatestatistically significant differences between groups (P < 0.05). Data are expressed asmean ± SE, n = 5–6.

Table 4The capacity of Senegalese sole POMC-A and B to yield POMC-derived peptides.

Peptide ssePOMC-A ssePOMC-B

ACTH + �a-MSH + +CLIP + �b-LPH + �b-MSH � +b-END + �

Y.S. Wunderink et al. / General and Comparative Endocrinology 175 (2012) 407–415 413

POMC-derived peptides over the two pomc paralogues (summa-rized in Table 4.).

If the precursor POMC cannot be cleaved as discussed above, thenon-functional domains are virtually useless. It is assumed thatthere is no longer purifying selection driving the evolution ofsequences that lost their function [18]. When this happens for thedifferent regions of the ssePOMC genes, differences in evolutionary

pressure will be observed. To investigate if sequences appear inac-tive or useless, evolutionary pressure acting on these domains wascalculated from the ratio of synonymous and non-synonymousnucleotide substitutions [18]. Comparing the individual peptide re-gions of ssePOMC-A and B, a strong sequence conservation and lowevolution rate for almost every POMC-derived peptide was present,especially for ACTH and a-MSH. However, when comparing thessePOMC-B b-END region with its orthologues, Ka/Ks ratios aroundor even >1 are seen, indicating a neutral selection directing topositive selection. It was already shown for b-END regions of tetra-odontid POMCb sequences that they evolve neutrally and appearnon-functional. However, thus far no strong evidence for relaxedor positive selection for ACTH/a-MSH or b-END regions was foundin other fish lineages [10]. Here, the Ka/Ks ratios in ssPOMC-Bindicate poor conservation and demonstrate the non-functionalityof its b-END in S. senegalensis. It has been proposed that the non-functional b-END started degenerating independently in severallineages, probably as a consequence of compensation provided bythe corresponding paralogue and/or the partitioning of pomc tis-sue distribution between the two paralogues [10]. Interesting isthe tendency of positive selection for the b-END region of sse-POMC-B. This suggests that the sequences are changing relativelyquickly and that there is a benefit for the loss of classical b-ENDfunction.

The sequence analyses point towards subfunctionalization ofpomc genes in S. senegalensis. We hypothesize that subfunctional-ization fine-tunes the production of POMC-derived peptides during

414 Y.S. Wunderink et al. / General and Comparative Endocrinology 175 (2012) 407–415

a stress response. We predict that ssePOMC-A (giving rise to ACTHbut not to b-MSH) is mainly involved in the endocrine stress re-sponse, whereas ssePOMC-B (giving rise to a/b-MSH but no ACTHand non-functional b-END) is primarily involved in the regulationof pigmentation (Table 4). Indeed, pomc-a expression is down-reg-ulated in S. senegalensis which experienced chronic stress, whereaspomc-b expression levels remain unaffected by high stocking den-sity, suggesting that ssePOMC-B is not involved in regulating thechronic stress response in S. senegalensis and indicates differentphysiological roles for both POMC paralogues.

During evolution, the POMC prohormone acquired an increasingnumber of derivate peptides [12]. To produce the required peptidehormone at the right time and place, extensive posttranslationalprocessing has evolved. For example, through the presence ofPC2 in the melanotropic cells, ACTH is further processed to a-MSH and CLIP, but also, b-LPH is cleaved into b-MSH and b-END.This means that when the production of MSH is needed, b-ENDis co-produced. The opioid b-END mediates a variety of functionsof which the involvement in both pain perception and rewardpathways is most important [30]. In physiological terms, theeffects of b-END may not always be required or can even beunwanted and thus suppressed. Indeed, it is already known thatb-END is post translationally processed by N-acetylation and C-ter-minal cleavage, which has significant effects on the physiologicalactions of b-END [1,5,34]. Moreover, the melanotroph cells containa mechanism to preferentially release MSH or b-END in commoncarp (Cyprinus carpio) [35]. The 3R whole genome duplicationgave the teleosts a second redundant copy of the POMC gene.The benefit of a simpler regulation, with e.g. a non-functional b-END, eliminates b-END’s side effects when only theproduction of MSH is needed and may have driven the subfunc-tionalization between the two copies. Specializations of pomc arealso reported for other vertebrates, suggesting an evolutionarytrend towards a different form of POMC peptide regulation. Beforethe divergence of eels, the region coding for a functional c-MSHhas been lost, whereas in cartilaginous fish an extra melanocyte-stimulating hormone, d-MSH, is found [10,13]. Like b-END,c-MSH shows distinct levels of degeneration in different lineagesof fish [10,13].

In conclusion, functional POMC-derived peptide hormones aredistributed over two S. senegalensis pomc genes which arose mostlikely independently of the 3R whole genome duplication event.Additionally, differential transcript expression following a chronicstressor indicate different physiological roles for both POMC para-logues. The subfunctionalization of POMC paralogues is a relevantnotion when studying POMC function in endocrine responses, notonly in S. senegalensis, but in all teleosts.

Acknowledgments

This work was partially funded by grants AGL2007-61211/ACU(Ministerio Educación y Ciencia of Spain) and Proyecto de Excel-encia PO7-RNM-02843 (Consejería de Innovación, Ciencia yEmpresa, Junta de Andalucía) to J.M.M. The authors are gratefulto ‘‘Planta de Cultivos Marinos’’ (CASEM, Puerto Real, Spain) forthe donation of S. senegalensis. Y.W. is funded by Ministerio de Edu-cación y Ciencia (Spain) through the program ‘‘Formación de Pro-fesorado Universitario’’ (Ref: AP2006-03932). E.V. is funded bythe SmartMix Program of the Netherlands Ministry of EconomicAffairs and the Ministry of Education, Culture and Science.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.ygcen.2011.11.026.

References

[1] H. Akil, E. Young, S.J. Watson, D.H. Coy, Opiate binding properties of naturallyoccurring N- and C-terminus modified beta-endorphins, Peptides 2 (1981)289–292.

[2] R.J. Arends, H. Vermeer, G.J.M. Martens, J.A.M. Leunissen, S.E. WendelaarBonga, G. Flik, Cloning and expression of two proopiomelanocortin mRNAs inthe common carp (Cyprinus carpio L.), Mol. Cell. Endocrinol. 143 (1998) 23–31.

[3] F.J. Arjona, L. Vargas-Chacoff, M.P. Martín del Río, G. Flik, J.M. Mancera, P.H.M.Klaren, The involvement of thyroid hormones and cortisol in the osmoticacclimation of Solea senegalensis, Gen. Comp. Endocrinol. 155 (2008) 796–803.

[4] B.A. Barton, Stress in fishes: A diversity of responses with particular referenceto changes in circulating corticosteroids, Integr. Comp. Biol. 42 (2002) 517–525.

[5] N.J. Bernier, G. Flik, P.H.M. Klaren, Regulation and contribution of thecorticotropic, melanotropic and thyrotropic axes to the stress response infishes, in: N.J. Bernier, G. van der Kraak, A.P. Farrell, C.J. Brauner (Eds.), FishPhysiology, Academic Press, New York, 2009, pp. 235–311.

[6] J.C.R. Cardoso, R. Laíz-Carríon, B. Louro, N. Silva, A.V.M. Canario, J.M. Mancera,D.M. Power, Divergence of duplicate POMC genes in gilthead sea bream Sparusauratus, Gen. Comp. Endocrinol. doi: 10.1016/j.ygcen.2010.12.001.

[7] Q. Chen, K. Layton, K. Veo, J.K. Angleson, R.M. Dores, a-TC1.9 cells – A modelsystem for analyzing the endoproteolytic processing of POMC, Gen. Comp.Endocrinol. 172 (2011) 96–106.

[8] B. Costas, C. Aragão, J.M. Mancera, M.T. Dinis, L.E.C. Conceiçao, B.C. Refojos,High stocking density induces crowding stress and affects amino acidmetabolism in Senegalese sole Solea senegalensis (Kaup 1858) juveniles,Aquac. Res. 39 (2008) 1–9.

[9] B. Costas, L.E.C. Conceição, C. Aragão, J.A. Martos, I. Ruiz-Jarabo, J.M. Mancera, A.Afonso, Physiological responses of Senegalese sole (Solea senegalensis Kaup,1858) after stress challenge: Effects on non-specific immune parameters, Plasmafree amino acids and energy metabolism, Aquaculture 316 (2011) 68–76.

[10] F.S.J. de Souza, V.F. Bumaschny, M.J. Low, M. Rubinstein, Subfunctionalizationof expression and peptide domains following the ancient duplication of theproopiomelanocortin gene in teleost fishes, Mol. Biol. Evol. 22 (2005) 2417–2427.

[11] M.T. Dinis, L. Ribeiro, F. Soares, C. Sarasquete, A review on the cultivationpotential of Solea senegalensis in Spain and in Portugal, Aquaculture 176 (1999)27–38.

[12] R.M. Dores, A.J. Baron, Evolution of POMC: origin, phylogeny, posttranslationalprocessing, and the melacortins, Ann. N.Y. Acad. Sci. 1220 (2011) 34–48.

[13] R.M. Dores, S. Lecaude, Trends in the evolution of the proopiomelanocortingene, Gen. Comp. Endocrinol. 142 (2005) 81–93.

[14] A. Force, M. Lynch, F.B. Pickett, A. Amores, Y.L. Yan, J. Postlethwait,Preservation of duplicate genes by complementary, Degenerat. Mutat. Genet.151 (1999) 1531–1545.

[15] V. Gonzalez-Nunez, R. Gonzalez-Sarmiento, R.E. Rodríguez, Identification oftwo proopiomelanocortin genes in zebrafish (Danio rerio), Mol. Brain Res. 120(2003) 1–8.

[16] M. Gorissen, E. de Vrieze, G. Flik, M.O. Huising, STAT genes display differentialevolutionary rates that correlate with their roles in the endocrine and immunesystem, J. Endocrinol. 209 (2011) 175–184.

[17] M. Herrera, L. Vargas-Chacoff, I. Hachero, I. Ruíz-Jarabo, A. Rodiles, J.I. Navas,J.M. Mancera, Physiological responses of juvenile wedge sole Dicologoglossacuneata (Moreau) to high stocking density, Aquac. Res. 40 (2009) 790–797.

[18] L.D. Hurst, The Ka/Ks ratio: Diagnosing the form of sequence evolution, TrendsGenet. 18 (2002) 486–487.

[19] A.K. Imsland, A. Foss, L.E.C. Conceiçao, M.T. Dinis, D. Delbare, E. Schram, A.Kamstra, P. Rema, P. White, A review of the culture potential of Solea solea andS. senegalensis, Rev. Fish Biol. Fisher. 13 (2003) 379–407.

[20] K.S. Kim, H.W. Kim, T.T. Chen, Y.T. Kim, Molecular cloning, tissue distributionand quantitative analysis of two proopiomelanocortin mRNAs in Japaneseflounder (Paralichthys olivaceus), BMB Rep. 42 (2009) 206–211.

[21] S.M. Korte, J.M. Koolhaas, J.C. Wingfield, B.S. McEwen, The Darwinian conceptof stress: Benefits of allostasis and costs of allostatic load and the trade-offs inhealth and disease, Neurosci. Biobehav. R. 29 (2005) 3–38.

[22] M.A. Larkin, G. Blackshields, N.P. Brown, R. Chenna, P.A. McGettigan, H.McWilliam, F. Valentin, I.M. Wallace, A. Wilm, R. Lopez, J.D. Thompson, T.J.Gibson, D.G. Higgins, Clustal W and Clustal X version 2.0, Bioinformatics 23(2007) 2947–2948.

[23] K.J. Livak, T.D. Schmittgen, Analysis of relative gene expression data using Real-Time Quantitative PCR and the 2-DDCt method, Methods 25 (2001) 402–408.

[24] A. Meyer, Y. van de Peer, From 2R to 3R: Evidence for a fish-specific genomeduplication (FSGD), BioEssays 27 (2005) 937–945.

[25] M. Ponce, C. Infante, M. Manchado, Molecular characterization and geneexpression of thyrotropin receptor (TSHR) and a truncated TSHR-like inSenegalese sole, Gen. Comp. Endocrinol. 168 (2010) 431–439.

[26] J. Postlethwait, A. Amores, W. Cresko, A. Singer, Y.L. Yan, Subfunctionpartitioning, The teleost radiation and the annotation of the human genome,Trends Genet. 20 (2004) 481–490.

[27] L. Pritchard, A. Turnbull, A. White, Pro-opiomelanocortin processing in thehypothalamus: impact on melanocortin signalling and obesity, J. Endocrinol.172 (2002) 411–421.

[28] S. Rozen, H. Skaletsky, Primer3 on the WWW for general users and for biologistprogrammers, Method. Mol. Biol. (Clifton, N.J.) 132 (2000) 365-386.

Y.S. Wunderink et al. / General and Comparative Endocrinology 175 (2012) 407–415 415

[29] E. Salas-Leiton, V. Anguis, B. Martín-Antonio, D. Crespo, J.V. Planas, C. Infante,J.P. Cañavate, M. Manchado, Effects of stocking density and feed ration ongrowth and gene expression in the Senegalese sole (Solea senegalensis):Potential effects on the immune response, Fish Shellfish Immun. 28 (2010)296–302.

[30] G. Sundström, S. Dreborg, D. Larhammar, Concomitant duplications of opioidpeptide and receptor genes before the origin of jawed vertebrates, PLoS ONE 5(2010).

[31] A. Takahashi, M. Amano, T. Itoh, A. Yasuda, T. Yamanome, Y. Amemiya, K.Sasaki, M. Sakai, K. Yamamori, H. Kawauchi, Nucleotide sequence andexpression of three subtypes of proopiomelanocortin mRNA in barfinflounder, Gen. Comp. Endocrinol. 141 (2005) 291–303.

[32] K. Tamura, J. Dudley, M. Nei, S. Kumar, MEGA4: Molecular EvolutionaryGenetics Analysis (MEGA) software version 4.0, Mol. Biol. Evol. 24 (2007)1596–1599.

[33] S. Tanaka, Comparative aspects of intracellular proteolytic processing ofpeptide hormone precursors: Studies of proopiomelanocortin processing, Zool.Sci. 20 (2003) 1183–1198.

[34] E.H. van den Burg, J.R. Metz, R.J. Arends, B. Devreese, I. Vandenberghe, J. vanBeeumen, S.E. Wendelaar Bonga, G. Flik, Identification of b-endorphins in thepituitary gland and blood plasma of the common carp (Cyprinus carpio), J.Endocrinol. 169 (2001) 271–280.

[35] E.H. van den Burg, J.R. Metz, F.A.T. Spanings, S.E. Wendelaar Bonga, G. Flik,Plasma a-MSH, Plasma a-440 MSH and acetylated b-endorphin levelsfollowing stress vary according to CRH441 sensitivity of the pituitarymelanotropes in common carp, Cyprinus carpio, Gen. Comp. Endocrinol. 140(2005) 210–221.

[36] J.B. Walsh, How often do duplicated genes evolve new functions?, Genetics139 (1995) 421–428

[37] S.E. Wendelaar Bonga, The stress response in fish, Physiol. Rev. 77 (1997) 591–625.

[38] Y.S. Wunderink, S. Engels, S. Halm, M. Yúfera, G. Martínez-Rodríguez, G. Flik,P.H.M. Klaren, J.M. Mancera, Chronic and acute stress responses in Senegalesesole (Solea senegalensis): The involvement of cortisol, CRH and CRH-BP, Gen.Comp. Endocrinol. 171 (2011) 203–210.