Subfunctionalization of Expression and Peptide Domains Following the Ancient Duplication of the Proopiomelanocortin Gene in Teleost Fishes
http://www.100md.com
分子生物学进展 2005年第12期
* Instituto de Investigaciones en Ingeniería Genética y Biología Molecular, CONICET, FCEyN, Universidad de Buenos Aires, Argentina; Departamento de Fisiología, Biología Molecular y Celular, FCEyN, Universidad de Buenos Aires, Argentina; Vollum Institute, Oregon Health and Science University; Department of Behavioral Neuroscience, Oregon Health and Science University; || Center for the Study of Weight Regulation and Associated Disorders, Oregon Health and Science University; and ? Centro de Estudios Científicos, Valdivia, Chile
E-mail: mrubins@dna.uba.ar.
Abstract
The proopiomelanocortin gene (POMC) encodes several bioactive peptides, including adrenocorticotropin hormone, -, ?-, and -melanocyte-stimulating hormone, and the opioid peptide ?-endorphin, which play key roles in vertebrate physiology. In the human, mouse, and chicken genomes, there is only one POMC gene. By searching public genome projects, we have found that Tetraodon (Tetraodon nigroviridis), Fugu (Takifugu rubripes), and zebrafish (Danio rerio) possess two POMC genes, which we called POMC and POMC?, and we present phylogenetic and mapping evidence that these paralogue genes originated in the whole-genome duplication specific to the teleost lineage over 300 MYA. In addition, we present evidence for two types of subfunction partitioning between the paralogues. First, in situ hybridization experiments indicate that the expression domains of the ancestral POMC gene have been subfunctionalized in Tetraodon, with POMC expressed in the nucleus lateralis tuberis of the hypothalamus, as well as in the rostral pars distalis and pars intermedia (PI) of the pituitary, whereas POMC? is expressed in the preoptic area of the brain and weakly in the pituitary PI. Second, POMC? genes have a ?-endorphin segment that lacks the consensus opioid signal and seems to be under neutral evolution in tetraodontids, whereas POMC genes possess well-conserved peptide regions. Thus, POMC paralogues have experienced subfunctionalization of both expression and peptide domains during teleost evolution. The study of regulatory regions of fish POMC genes might shed light on the mechanisms of enhancer partitioning between duplicate genes, as well as the roles of POMC-derived peptides in fish physiology.
Key Words: POMC ? teleosts ? evolution ? Tetraodon ? ?-endorphin ? subfunctionalization
Introduction
Proopiomelanocortin (POMC) is a prohormone that gives rise to several different bioactive peptides through cell-type–specific posttranslational processing. POMC-derived peptides are produced mainly in the pituitary gland and the brain, being involved in processes as divergent as stress, pain, energy balance, and, in fishes and amphibians, adaptation to background color (reviewed in Hadley and Haskell-Luevano 1999; Volkoff et al. 2005). In the pituitary gland, the POMC gene is expressed in two types of cells: corticotrophs in the pars distalis, where POMC is processed into adrenocorticotropin hormone (ACTH) and ?-lipotropin and melanotrophs in the pars intermedia (PI), where it gives rise to -melanocyte-stimulating hormone (-MSH) and ?-endorphin. In the vertebrate brain, POMC is mainly expressed in the ventral hypothalamus, where POMC is processed into -MSH, ?-MSH, -MSH, and ?-endorphin.
Genes and cDNAs encoding POMC have been cloned and sequenced from representatives of all vertebrate classes, including several fishes (Takahashi et al. 2001). Birds and mammals have only one POMC gene, but some fishes and amphibians possess more than one. Gene duplicates may arise through a variety of mechanisms, ranging from single-gene to whole-genome duplication. The recent sequencing and assembly of the genome of the teleost Tetraodon nigroviridis (family Tetraodontidae), as well as studies on gene duplicates in the genome of another tetraodontid, Takifugu rubripes (Fugu), give strong support to the idea that an ancient whole-genome duplication occurred in the teleost lineage around 320 MYA (Taylor et al. 2001; Christoffels et al. 2004; Jaillon et al. 2004; Vandepoele et al. 2004). In addition, some fishes like carp, rainbow trout, and sturgeons have undergone recent polyploidization events (Volff 2005).
After a gene is duplicated, one of the copies usually accumulates nonsense mutations and becomes a pseudogene, although sometimes both paralogue copies are retained in the genome. In the case of tetraodontids, the percentage of duplicate genes retained after the proposed teleost-specific duplication has been calculated to be 3%–7% (Christoffels et al. 2004; Jaillon et al. 2004). The retention of the duplicated copies may be accomplished when one of the paralogues acquires a new function in relation to the original gene (neofunctionalization) or, alternatively, the functions of the original gene are partitioned between the new duplicates, as stated in the duplication-degeneration-complementation (DDC) model (Force et al. 1999; Postlethwait et al. 2004). The latter process, known as subfunctionalization, is most often evidenced by the partitioning of the original expression pattern. For instance, if the original gene was expressed in regions A and B, the differential loss of regulatory regions might cause one duplicate to be expressed in region A and the other in region B so that both paralogues are kept under selective pressure and become fixed in the genome. In teleosts, partitioning of expression domains has been observed in several paralogue genes, including the transcription factors Engrailed1 (Force et al. 1999), Hoxb1 (McClintock, Kheirbek, and Prince 2002), and Mitf (Altschmied et al. 2002), retinol-binding proteins (Liu et al. 2005), the neuropeptide somatostatin (Low 2003), among others. In addition, subfunctionalization may also involve the differential partitioning of peptide domains (Force et al. 1999).
Here, we present molecular and phylogenetic evidence that higher teleosts possess two POMC paralogue genes, which we named POMC and POMC?, that originated in the ancient teleost whole-genome duplication. In addition, we show that both Tetraodon POMC gene duplicates have undergone subfunctionalization, as evidenced by their complementary expression patterns in brain and pituitary. Moreover, the peptide repertoire of POMC paralogues has also been subfunctionalized, with the loss of ?-endorphin from many teleost POMC? genes.
Materials and Methods
Animals
Adult fish from the genus Tetraodon (T. nigroviridis and T. fluviatilis) and zebrafish (Danio rerio) were purchased from a local aquarium and processed immediately for nucleic acid extraction or tissue fixation. Experimental protocols were approved by the local animal care and use committees and were consistent with the Guide for the Care and Use of Laboratory Animals.
Reverse Transcriptase–Polymerase Chain Reactions and Sequencing
Total RNA from Tetraodon pituitaries and preoptic area was isolated using TRIzol (Invitrogen, Carlsbad, Calif.). For 5' and 3' rapid amplification of cDNA ends (RACE) POMC polymerase chain reactions (PCRs), cDNA was prepared using the MATCHMAKER Library Construction & Screening Kit (Clontech, Palo Alto, Calif.) incorporating a SMART oligo at the 5' end and using CDS III–modified oligo-dT, according to the instructions of the manufacturer. Nested PCRs were performed with the following primer pairs: for 5' RACE, 5'-ACGCAGAGTGGCCATTATGGCCGGG-3' (complementary to the SMART oligo)/5'-CATGCTGCTGGCCTCGGAGT-3' (complementary to exon 2); and for 3' RACE, 5'-GAGCGCCCACCTCCAACT-3' (complementary to exon 3)/5'-GTATCGATGCCCACCCTCTAGAGGCCGAGGCGGCCGACA-3' (complementary to CDS III oligo). To determine the exon-intron structure of POMC genes, PCRs were performed with the following primer pairs: for POMC, 5'-CCATCAGCAGCCAGCACGA-3' (complementary to exon 1)/5'-CCAACAACCGTCATGAGC-3' (complementary to exon 3); for POMC?, 5'-ACGGCGGGAGGAAAGCAA-3' (complementary to exon 1)/5'-CAGTATCCTCCTCTTGTC-3' (complementary to exon 2); and 5'-GCTGATGACGTACCTGTG-3' (complementary to exon 2)/5'-TGCCCTCGCTCGTCCCTGT-3' (complementary to exon 3). Gel-purified PCR bands were subjected to automated DNA sequencing.
Immunohistochemistry
Adult fishes were anaesthetized by immersion in a neutral buffered solution of 0.1% (w/v) tricaine methane sulfonate (Sigma, St. Louis, Mo.), and heads were fixed overnight in Bouin's fixative. Brains with attached pituitaries and part of the skull (cranium base) were dissected and washed with 70% (v/v) ethanol for 24 h, cryoprotected with 30% (w/v) sucrose, and frozen in embedding medium (Tissue Tek, Sakura Tek, Torrance, Calif.). Serial transversal and sagital sections (20 μm) were cut using a cryostat microtome (IEC Microtome, Walldorf, Germany) at –20°C. Sections were thaw mounted on Vectabond-coated (Vector Laboratories, Burlingame, Calif.) slides and stored at –70°C until use. Alternatively, some brains were embedded into paraffin and 4- or 10-μm slices were cut in a microtome (SM2000R, Leica, Nussloch, Germany). Immunohistochemistry proceeded as described previously (de Souza et al. 2005) with a polyclonal antihuman ACTH IC-1 antibody (1:300; National Hormone and Pituitary Program, National Institutes of Health).
Radioactive In Situ Hybridization
Cryosections obtained as described above were treated with 10 μg/ml proteinase K in 10 x TE buffer (100 mM Tris, 10 mM ethylenediaminetetraacetic acid [EDTA]) at room temperature (RT) for 15 min and then washed twice with the same buffer. Slides were then treated with 0.1 M triethanolamine (TEA) pH 8.0 for 3 min, incubated in 0.0025% (v/v) acetic anhydride in 0.1 M TEA for 10 min, washed twice in 2 x standard saline citrate (SSC) (NaCl, sodium citrate), and finally dehydrated quickly in ascending ethanol concentrations. For prehybridization, slides were incubated under parafilm coverslips with the hybridization buffer (66% [v/v] formamide, 260 mM NaCl, 1.3 x Denhardt's, 1.3 mM EDTA, 13 mM Tris-HCl pH 8.0, 13% [w/v] dextran sulfate) for 1 h/57°C. [35S]-labeled riboprobes were heated to 65°C/5 min (with 0.5 mg/ml tRNA, 10 mM dithiothreitol [DTT] in water) and added to the hybridization buffer in a concentration of 5 x 106–107 cpm/ml. Hybridization followed overnight at 57°C. Next day, slices were washed four times in 4 x SSC and subjected to ribonuclease (RNase) A digestion (20 μg/ml RNase A, 0.5 M NaCl, 10 mM Tris-HCl pH 8.0, 1 mM EDTA pH 8.0) at 37°C/30 min, desalted in 2 x, 1 x, 0.5 x SSC/1 mM DTT, and subjected to a final wash in 0.1 x SSC/1 mM DTT at 30 min/65°C. After a quick rinse in 0.1 x SSC/1 mM DTT RT, slides were dehydrated through 5-min steps in ethanol at 70%, 95%, and 100% (v/v), dried at room temperature, dipped in Kodak N-TB2 emulsion (diluted 1:1 in distilled water), and exposed for 3–6 days. Slides were developed, counterstained with cresyl violet (Sigma), dehydrated with ethanol, and coverslipped under Permount (Fisher Scientific, Pittsburgh, Pa.). To prepare riboprobes, PCR fragments of exon 2 of Tetraodon POMC (123-bp product; TE2SE: 5'-CCAGAATGAGTCCTGCGTGGCT-3' and TE2AN: 5'-CATGCTGCTGGCCTCGGAGT-3') and exon 3 of POMC? (302-bp product; 2TPE3SE4: 5' CGGAGCAGCAACAGGGAC 3' and 2TPE3AN2: 5' AACGCCTGCGGGCACGGA-3') were cloned into pGEMT-Easy vector (Promega, Madison, Wisc.). Antisense RNA probes were synthesized by incubating 1 μg of linearized plasmid with transcription buffer (40 mM Tris-HCl, pH 8.0, 15 mM MgCl2, 5 mM DTT, 0.05 mg/ml bovine serum albumin), 10 mM DTT, 60 U RNase out (Invitrogen), 0.5 mM of adenosine triphosphate, cytidine triphosphate, and guanosine triphosphate, 1,000 μCi [35S]UTP (1,250 Ci/mmol; PerkinElmer, Wellesley, Mass.), and 16 U of T7 polymerase (Amersham Biosciences, Piscataway, N.J.). After 70 min of incubation at 37°C, plasmid was digested with DNAse I (Invitrogen). Riboprobes were precipitated with 10 μg tRNA, 0.1 vol 5 M NH4 acetate, and 2.5 vol 100% (v/v) ethanol for 1 h at –70°C, washed with 70% (v/v) ethanol, and resuspended in RNAse-free water.
Sequences and Databases
The genomic regions around POMC loci from Fugu and Tetraodon were located via Blast searches (Altschul et al. 1990) in the Ensembl database Web site (http://www.ensembl.org/index.html) using POMC sequences from other species. POMC loci are present in the scaffolds 970 in Fugu and 10884 in Tetraodon, while POMC? loci are located in scaffold 130 in Fugu and chromosome 14 in Tetraodon. A complete cDNA sequence for Fugu POMC was assembled from expressed sequence tag (EST) clones EFRz001apsG3, EFRa090apsE11, and EFRz002apsD6 retrieved from the Fugu Home Page of the Rosalind Franklin Centre for Genomic Research (http://www.hgmp.mrc.ac.uk). A cDNA sequence for Tetraodon POMC? (clone name FD0ADA50CB12AAP1) was retrieved from the Genoscope Home Page (http://www.genoscope.cns.fr/externe/tetranew). Data on the POMC neighborhood in chicken (Gallus gallus) and humans were obtained from the Ensembl and Mapviewer Web sites (http://www.ncbi.nih.gov/mapview/), respectively. Zebrafish POMC and POMC? were located, respectively, in sequences AL590149 (from linkage group [LG] 17) and AL845420 (from LG 20) deposited in GenBank (http://www.ncbi.nih.gov). We also retrieved from GenBank the POMC peptide and cDNA sequences from the following species and accession numbers: Verasper moseri (barfin flounder) POMC-A (AB051424); POMC-B (AB051425) and POMC-C (AB051426); Oncorhynchus mykiss (rainbow trout) POMC-A (X69808) and POMC-B (X69809); Oncorhynchus keta (chum salmon) (X01122); Polypterus senegalus (bichir) (AF465781); Acipenser trasmontanus (sturgeon) POMC-A (AF092937) and POMC-B (AF092936); Polyodon spathula (paddlefish) POMC-A (AF117302) and POMC-B (AF117303); Cyprinus carpio (common carp) POMC-I (Y14618) and POMC-II (Y14617); D. rerio (zebrafish) (NM_181438); Homo sapiens (man) (NM_000939); Carassius auratus (goldfish) (AJ431209); Ictalurus puntactus (channel catfish) (AY174050); Paralichthys olivaceus (bastard halibut) POMC-I (AF184066) and POMC-II (AF191593); Anguilla japonica (Japanese eel) (AY158010); and Anguilla rostrata (American eel) (AF194969).
Evolutionary Analyses
POMC peptide sequences were aligned with the ClustalW program (http://www.ebi.ac.uk/clustalw) (Thompson, Higgins, and Gibson 1994). The alignment used for phylogenetic inference using the PHYLIP 3.62 Program Package (Felsenstein 2004) is shown in Supplementary Figure 1 (Supplementary Material online). For maximum-likelihood (ML) phylogenetic inference, the PROTML program was used with the amino acid substitution model of Jones, Taylor, and Thornton (1992) using 100 replicates generated with the SEQBOOT program. A consensus tree was created with CONSENSE. Ka/Ks ratios were calculated using the DnaSP 4.0 program (J. Rozas and R. Rozas 1999) downloaded from http://www.ub.es/dnasp.
Results
Identification of Duplicate POMC Genes in Teleost Fishes
We searched the public genome projects of Tetraodon (Jaillon et al. 2004) and Fugu (Aparicio et al. 2002) using the Ensembl database Web site. In each tetraodontid genome, we identified two POMC genes which we named POMC and POMC?, according to their identity to mammalian POMC at the peptide level (45% and 35%, respectively). In the Tetraodon genome project, POMC? has been assigned to chromosome 14, while POMC is found in scaffold 10884 of unknown location. In the Fugu genomic database, where scaffolds have not yet been assigned to chromosomes, POMC and POMC? are located in scaffolds 970 and 130, respectively.
The complete cDNA from Fugu POMC was assembled from ESTs retrieved from the Fugu database of the Rosalind Franklin Centre for Genome Research, and this sequence was used to deduce the exon-intron structure of both Fugu and Tetraodon POMC genes. Reverse transcriptase (RT)–PCR and 5'/3'-RACE reactions using Tetraodon pituitary cDNA were done to confirm exon-intron boundaries and the transcriptional start site. The structure of tetraodontid POMC? genes was defined using a Tetraodon POMC? cDNA clone from the Genoscope Web site, and exon-intron boundaries were confirmed by sequencing RT-PCR fragments. Both tetraodontid POMC genes have three exons: exon 1 is noncoding, exon 2 contains the starting ATG, and most of the coding region lies in exon 3 (fig. 1A). This organization is remarkably similar to that of POMC genes of all tetrapods studied to date.
FIG. 1.— (A) Organization of Tetraodon POMC paralogues compared to the human POMC gene. Note the shorter size of introns in Tetraodon. (B) Alignment of Tetraodon POMC amino acid sequences with those of different vertebrates. Peptides derived from POMC are shown on top of the sequences. Putative dibasic cleavage sites are indicated by arrowheads. Asterisks and dots below the sequences indicate identical and conservative residues, respectively. Melanocortin and opioid core sequences are underlined. Note that both teleost POMC and POMC? lack -MSH and that Tetraodon and Fugu POMC? lack the opioid signal.
A gene and a cDNA encoding zebrafish POMC have been described previously (Hansen et al. 2003; Liu et al. 2003). In addition, González-Nu?ez, González-Sarmiento, and Rodríguez (2003) have reported a second, incomplete cDNA corresponding to a more divergent POMC sequence, isolated from a brain cDNA library. Based on the phylogeny of the genes (see below), we propose to call these zebrafish genes POMC and POMC?, respectively. Searching the GenBank database, we identified zebrafish genomic scaffolds of LGs 17 and 20 that contain POMC and POMC?, respectively. Comparing the reported partial POMC? cDNA (González-Nu?ez, González-Sarmiento, and Rodríguez 2003) with genomic sequence, we identified two exons which, as judged from sequence alignments, contain the whole coding region of zebrafish POMC?. These exons correspond probably to exons 2 and 3 because all POMC genes studied to date contain a noncoding exon 1. A putative exon 1 could not be identified from sequence comparisons.
Characterization of Duplicate POMC Peptides
Figure 1B shows an alignment of deduced Tetraodon, Fugu, and zebrafish POMC and POMC? amino acid sequences and those of other representative vertebrates. We have included in the alignment the sequences of the reported POMC-B from rainbow trout (O. mykiss—Salbert et al. 1992) and the POMC-C from barfin flounder (Verasper moseri—Takahashi et al. 2005), which we have assigned to the POMC? group. In general, conservation of typical POMC features is high between teleost POMC sequences but lower in POMC? sequences. The amino acid sequence of -MSH is highly conserved (93%–100%) in POMC and POMC? in all species (fig. 1B). ?-MSH is also conserved, but zebrafish POMC? lacks a dibasic cleavage site at the prospective amino-terminal of the peptide (fig. 1B). The hormone ACTH, which encompasses -MSH and CLIP, is potentially derived from POMC and POMC? in all species except zebrafish POMC?, which lacks a dibasic cleavage site at the C-terminal of the prospective peptide (fig. 1B; González-Nu?ez, González-Sarmiento, and Rodríguez 2003). Amino acid identity between human and teleost ACTH is 74%–77% for POMC and 59% for POMC? sequences. Both teleost POMC and POMC? sequences lack the region corresponding to -MSH (fig. 1B), a third melanocortin which is present in other vertebrate POMC sequences (Dores and Lecaude 2005).
The region corresponding to the ?-endorphin peptide is more conserved in teleost POMC than POMC? sequences (fig. 1B). Amino acid identity between human and teleost ?-endorphins ranges from 59% to 67% for POMC and only 20%–44% for POMC?. The N-terminal met-enkephalin sequence of ?-endorphin, which contains the opioid motif YGGFM, presents several degrees of degeneration in teleost POMC? sequences. The motif is changed to YGGLM in zebrafish POMC? (González-Nu?ez, González-Sarmiento, and Rodríguez 2003) and to SGRFM in barfin flounder POMC-C (Takahashi et al. 2005; fig. 1B). Strikingly, the opioid motif is completely absent from POMC? ?-endorphin sequences in Tetraodon and Fugu (fig. 1B). Because the opioid signal is necessary for stimulation of the opioid receptors (Reinscheid et al. 1995; Waldhoer, Bartlett, and Whistler 2004), it is likely that POMC? from tetraodontids and, possibly, barfin flounder and zebrafish are incapable of generating a ?-endorphin with opioid activity. Nevertheless, ?-endorphin peptides containing the canonical opioid signal are present in the POMC-B sequences of trout (fig. 1B) and salmon, indicating that the peptide repertoire of POMC? sequences varies among teleosts.
The absence of opioid activity in ?-endorphin of Tetraodon and Fugu POMC? indicates that the C-terminal of POMC? might be nonfunctional and, consequently, might be accumulating a high number of nonsynonymous mutations. To check this possibility, we calculated the ratio of the number of nonsynonymous substitutions per nonsynonymous sites (Ka) to the number of synonymous substitutions per synonymous sites (Ks) (Hurst 2002) of the ?-endorphin regions of tetraodontid POMC and POMC? genes. ?-endorphins from POMC showed a Ka/Ks much lower than 1 (Ka/Ks = 0.12), indicating that this region is under purifying selection in POMC genes. In contrast, ?-endorphins of POMC? have a Ka/Ks slightly higher than 1 (Ka/Ks = 1.26), suggesting that the ?-endorphin region of POMC? came under relaxed selection in the tetraodontid lineage. ACTH regions, in contrast, seem to be under purifying selection in both tetraodontid POMC (Ka/Ks = 0.36) and POMC? (Ka/Ks = 0.17) genes. The unique POMC gene of distantly related sarcopterygians like humans and Australian lungfish (Neoceratodus forsteri, Dores et al. 1999) is under strong purifying selection in both the ?-endorphin (Ka/Ks = 0.15) and ACTH (Ka/Ks = 0.07) regions. We found no strong evidence for relaxed or positive selection in ACTH or ?-endorphin regions in other fish lineages.
Phylogenetic Analysis of Teleost POMC Genes
POMC cDNAs have been sequenced from several phylogenetically diverse species of fishes, and POMC duplicates have been identified in some of them. To understand the origin of POMC paralogues in teleost genomes, we constructed phylogenetic trees of POMC amino acid sequences using ML analyses. Figure 2B shows an ML phylogenetic tree of POMC peptides of ray-finned fishes (Actinopterygii). Supplementary Figure 1 (Supplementary Material online) shows the alignment used to build this tree. To serve as reference, figure 2A shows the phylogenetic relationships between selected fishes adapted from Kikugawa et al. (2004) and Inoue et al. (2004). Recent molecular phylogenetic studies agree on the relationships of most groups shown, but there is controversy over the relationship between Semionotiformes (gar) and Acipenseriformes (sturgeons and paddlefishes). For a similar but slightly different molecular phylogeny, see Inoue et al. (2003). The POMC phylogenetic tree shown in figure 2B corresponds roughly to the phylogenetic relationships between the fish species. POMC and POMC? sequences from euteleosts (higher teleosts), as well as POMC sequences from two eel species (which are basal teleosts), are grouped in a branch with 100% bootstrap support (fig. 2B). Importantly, the analysis supports (99% bootstrap) a branch with the POMC? sequences of zebrafish, Tetraodon, and Fugu, together with rainbow trout POMC-B, barfin flounder POMC-C, and a POMC sequence from chum salmon. POMC sequences from euteleosts are grouped in another branch, being monophyletic in the tree, although with relatively low statistical support (74%). The available eel POMC sequences cluster with the POMC? branch with moderate statistical support (84% bootstrap).
FIG. 2.— Phylogenetic relationships between POMC sequences. (A) Putative phylogenetic relationships between relevant vertebrate groups adapted from Kikugawa et al. (2004) and Inoue et al. (2004). Putative timing of teleost whole-genome duplication (WGD) (Hoegg et al. 2004) is indicated by an arrow. (B) ML phylogenetic tree of POMC amino acid sequences of ray-finned fishes. Percentage bootstrap values (100 replicates) are indicated. The node grouping teleost POMC sequences is highlighted with a black circle, and POMC and POMC? clades are indicated. See Materials and Methods for details on the species.
Our phylogenetic results indicate that the duplication event that originated the POMC and POMC? paralogues happened in the teleost lineage, after the branching of the longnose gar lineage. Within euteleosts, the duplication seems to have happened before the separation of the superorders Ostariophysii (zebrafish, goldfish, and carp), Protacanthopterygii (salmon and trout), and Acanthopterygii (Tetraodon, Fugu, barfin flounder, and halibut). POMC sequences from eels, which are not considered euteleleosts, are grouped in the same branch as euteleost POMC and POMC?, suggesting that the POMC gene duplication happened before the branching of elopomorphs (eels and relatives) from the lineage that led to euteleosts (see Discussion).
Apart from the duplication that originated POMC and POMC?, there have been other duplications involving the POMC gene in the history of teleosts. As evidenced in fig. 2B, there are duplicates of POMC (POMC-I and -II) in the halibut and carp. In the case of barfin flounder, there are three POMC sequences reported: two belonging to the POMC group (POMC-A and -B) and another to the POMC? group (POMC-C). In addition, there are two similar POMC duplicates in the primitive chondrostean fishes sturgeons and paddlefish (POMC-A and -B). These very similar pairs of POMC genes in teleosts and chondrosteans may have arisen in recent, independent genome duplications in these lineages (see Discussion).
Neighborhood of POMC Genes
To determine whether POMC and POMC? originated in a single-gene duplication or in a higher order event (segmental, chromosomal, or whole-genome duplication), we compared the gene content around POMC loci from several species. POMC and POMC? genes from fish, as well as the POMC genes from mammals, chicken, and the frog Xenopus tropicalis are all flanked at the 3' end by a 21-exon gene of unknown function, dubbed KIAA0953 (fig. 3). KIAA0953 encodes a protein similar to a Drosophila transmembrane lipase (Huang et al. 2004). Phylogenetic analyses show that the KIAA0953 genes downstream of all POMC genes are homologues (not shown). At the 5' side, both zebrafish POMC genes and the unique chicken POMC gene in chromosome 3 are flanked by a gene encoding an angiopoietin-like factor (ANGPL, fig. 3). Tetraodon and Fugu POMC? genes also have an ANGPL gene located at the 5' side, although separated from POMC? by a few intervening genes. Phylogenetic analyses show that the angiopoietin-like factors near the POMC genes from zebrafish, tetraodontids, and chicken are true homologues (data not shown). In addition, the gene encoding sortin nexin 9 (SNX9) is conserved at the 5' side of teleost POMC genes (fig. 3).
FIG. 3.— Analysis of POMC neighborhood in different vertebrate genomes. Genomic scaffolds and chromosomes where the loci are located in each species are indicated. Gene neighborhoods around Fugu POMC loci (not shown) are identical to that of Tetraodon. Genes were named according to the human orthologs. Only genes showing homology to others are shown (black bars). White square denotes omitted intervening genes. The dotted line denotes change in gene order. Arrowheads show the direction of transcription of POMC genes. Hypothetical reconstructions of POMC gene neighborhood in the ancestor of Tetraodon and zebrafish (teleost ancestor) and teleosts and tetrapods (bony fish ancestor) are shown.
We conclude that the gene order ANGPL-POMC-KIAA0953 was likely found in the last common ancestor of ray-finned and lobe-finned fishes (fig. 3), which lived 400–450 MYA (Kumar and Hedges 1998). In addition, the gene order SNX9-ANGPL-POMC-KIAA0953 was probably present in the ancestor of higher teleosts. The conservation of gene content around duplicated POMC loci in teleosts indicates that these paralogues are not the result of a single-gene duplication, but rather of a higher order duplication event. The possibility of a chromosomal or genome duplication is further supported by the fact that the zebrafish POMC loci are located in different chromosomes.
Partitioning of Expression Domains of POMC in Tetraodon
The persistence of two ancient POMC duplicates in teleosts suggests that each of them performs a unique function. Because it has been observed that gene duplicates have frequently partitioned their expression domains, we decided to check this possibility for POMC paralogues. At first, to verify which regions express the POMC protein, we performed immunohistochemistry on slices of brain and pituitary of adult Tetraodon using an anti-ACTH antibody that also recognizes the whole POMC prohormone. Because the ACTH region is very similar in both Tetraodon POMC paralogues (fig. 1B), it is expected that the antibody will recognize the products of both genes.
As shown in figure 4, ACTH-positive perykaria were found in three regions: (1) the pituitary gland, (2) the nucleus lateralis tuberis (NLT) in the ventral hypothalamus, and (3) the preoptic nucleus area (POA). In the pituitary, POMC is found in the rostral pars distalis (RPD), presumably in corticotrophs, as well as in melanotrophs of the PI. No staining was observed in the pars nervosa (PN) or in the posterior pars distalis (PPD). In the ventral hypothalamus, we detected ACTH-like immuoreactivity in a few cells of the NLT. This pattern of expression in the pituitary and NLT of Tetraodon is similar to that of other fish like the barfin flounder (Amano et al. 2005), zebrafish (Herzog et al. 2003; Liu et al. 2003), and the cichlid Cichlastoma dimerus (Pandolfi et al. 2003). In addition to expression in these two areas, we observed ACTH immunoreactivity in the preoptic area, both in the anterior periventricular area as well as in the more dorsoposterior magnocellular region (POm). ACTH immunoreactivity in the preoptic area has already been observed in the carp (Metz et al. 2004) and other nonteleost vertebrates (see Discussion).
FIG. 4.— ACTH immunoreactivity in Tetraodon brain and pituitary. The ACTH antibody recognizes the whole POMC prohormone. Top drawing illustrates the location of POMC-expressing cell groups. Anterior is to the left in the drawing and panels (C) and (D). Arrowheads indicate the planes of the transversal cuts in panels (A) and (B). (A) Transversal cut at the level of the anterior preoptic area (POa) showing ACTH-immunoreactivity cell groups (arrows) around the diencephalic ventricle (V). (B) Transversal cut showing ACTH-positive cells in the magnocellular part of the preoptic area (POm, arrows). (C) Sagital cut showing ACTH immunoreactivity in the rostral pars distalis (RPD) and pars intermedia (PI) of the pituitary. (D) ACTH-positive cells (arrows) in a sagital cut of the nucleus lateralis tuberis (NLT). Other abbreviations: Cb, cerebellum; Ep, epiphysis; OfB, olfactory bulb; OpN, optic nerve; OpTec, optic tectum; MO, medulla oblongata; PPD, posterior pars distalis; PN, pars nervosa; Tel, telencephalon; and Thal, thalamus.
To check whether the expression patterns of POMC might have been partitioned between the paralogues, we perfomed radioactive in situ hybridization with specific POMC and POMC? antisense riboprobes on slices of adult Tetraodon brains and pituitaries. As shown in figure 5, POMC is strongly expressed in pituitary RPD and PI, as well as in a few cells of the NLT. No expression was observed in preoptic nuclei or in other brain areas. POMC?, on the other hand, is expressed in the POA and in the pituitary PI, while no expression was detected in the pituitary RPD, NLT, or other brain area. In the pituitary PI, expression of POMC is much higher than that of POMC? (compare fig. 5A and B). The combined expression patterns of POMC and POMC? correspond to the distribution of ACTH-positive perikarya in Tetraodon pituitary and brain. Because there is strong evidence that POMC was expressed in the POA, NLT, and pituitary RPD and PI of ancestral ray-finned fishes (see Discussion), these results indicate that the expression domains of the POMC gene were partitioned between POMC and POMC? during the evolution of Tetraodon.
FIG. 5.— Radioactive in situ hybridization of POMC paralogues in the brain and pituitary of Tetraodon. Riboprobes are complementary to the exon 2 of POMC (left panels) and the exon 3 of POMC? (right panels). All panels show sagital cuts, anterior to the left, dorsal is up. (A, B) Pituitary cuts. RPD, rostral pars distalis; PI, pars intermedia. (C, D) Nucleus lateralis tuberis (NLT) cuts. (E, F) Preoptic area (POA) cuts; OpN, optic nerve. Note that POMC is expressed in the RPD, PI, and NLT, while POMC? is expressed in the POA and weakly in the PI.
Discussion
In this work, we show that: (1) Tetraodon and Fugu have two POMC genes, POMC and POMC?; (2) these two genes are the product of an ancient duplication and are present in all higher teleosts; (3) the expression domains of the primitive POMC gene have been partitioned, with POMC being expressed in the NLT and pituitary RPD and PI, and POMC? in the POA and pituitary PI; and (4) the peptide repertoire of POMC has also been subfunctionalised because a bona fide ?-endorphin is lacking in POMC? in tetraodontids and other lineages.
Phylogeny of POMC Duplicates in Teleosts
Molecular phylogenetic studies indicate that, among higher teleosts, one of the first groups to diverge from the lineage leading to the Acanthopterygii (which includes Tetraodon, Fugu, halibut, and barfin flounder) was the Ostariophysii (zebrafish, carps, catfishes, and their relatives), followed by the Protacanthopterygii (salmons and trouts) (fig. 2A). In our phylogenetic analyses, POMC? sequences from fishes as divergent as zebrafish, salmon, and Tetraodon are grouped together in a branch clearly separated from their paralogues, which also tend to group together. Thus, the duplication that originated the two POMC genes likely occurred in a common ancestor to all higher teleosts and, therefore, POMC and POMC? genes will probably be present in the genomes of all 22,000 higher teleost species. A recent inventory of the melanocortin system in Fugu identified only one POMC gene (Klovins et al. 2004), which corresponds to POMC. The low level of nucleotide identity of Fugu POMC? to other POMCs has probably prevented the identification of this gene in that study.
Recently, the hypothesis of an ancient, teleost-specific tetraploidization followed by rediploidization has gained much support (reviewed in Postlethwait et al. 2004; Volff 2005). In Fugu and zebrafish, phylogenetic analyses of duplicate paralogues suggest that an extensive gene duplication event, most likely a tetraploidization, happened around 320 MYA (Taylor et al. 2001; Christoffels et al. 2004; Vandepoele et al. 2004). In addition, analyses of Tetraodon and zebrafish genomes show that many syntenic segments (paralogons) occurring once in mammalian genomes are found duplicated in teleost genomes (Woods et al. 2000; Jaillon et al. 2004). Because POMC and POMC? seem to be common to all higher teleosts, it is likely that these genes are derived from the ancient teleost tetraploidization. This possibility is supported by the conservation of gene content around POMC paralogues in tetraodontids and zebrafish, which rules out the possibility of a single-gene or retrovirus-mediated duplication of POMC. In addition, the fact that the zebrafish POMC loci are found in different LGs, i.e., in chromosomes 17 and 20, is also compatible with a whole-genome duplication. Importantly, zebrafish LGs 17 and 20 have other genetic markers in common (snap25, bmp2, sox11), consistent with their being, to a great extent, sister chromosomes originated in the teleost genome duplication (fig. 4 in Woods et al. 2000; Taylor et al. 2003).
The phylogenetic timing of the genome duplication in the evolution of teleosts is still not known. Recently, Hoegg et al. (2004) analyzed duplicated genes from basal ray-finned fishes and basal teleosts (bony tongues and eels) and concluded that the genome duplication happened after the separation of the bichir, sturgeon, and gar, but before the separation of bony tongues (Osteoglossomorpha) and eels (Elopomorpha) from the higher teleost lineage (see fig. 2A), even though this hypothesis and the timing of 320 MYA for the genome duplication do not agree with the available actinopterygian fossil record (Benton 2000; see Discussion in Christoffels et al. 2004 and Hoegg et al. 2004). Our results are compatible with the phylogenetic dating proposed by Hoegg et al. (2004) because the POMC of eels (Anguilla) are grouped tightly within the branch of POMC and POMC? from higher teleosts (fig. 2B). In addition, there were duplications of the POMC gene which occurred independently from the duplication that originated POMC and POMC?. It is likely that many of these extra POMC genes are the result of more recent whole-genome duplications, which are known to have happened in sturgeons (Ludwig et al. 2001), salmonids (Palti et al. 2004), and some cyprinids like carp and goldfish (David et al. 2003).
Subfunctionalization of POMC Duplicates
The POMC gene plays several roles in vertebrate physiology. Among its functions are the regulation of serum glucocorticoid levels (mediated by pituitary ACTH), pain sensitivity (mediated by brain ?-endorphin), food intake (mediated by brain melanocortin peptides -MSH, ?-MSH, and -MSH), and the control of skin pigmentation (mediated by pituitary -MSH). This diversity of functions is, to a great extent, a consequence of the modular structure of the POMC prohormone, which can give rise to several bioactive peptides in different cell types. The POMC peptide repertoire has undergone different specializations in some vertebrate groups. For example, cartilaginous fishes have an extra melanocortin, -MSH, not present in other lineages (Dores and Lecaude 2005). In another case, the segment encoding -MSH, which is found in all vertebrate groups, has been lost from teleost POMC before the divergence of eels (Takahashi et al. 2001; Dores and Lecaude 2005). Both POMC and POMC? lack -MSH, suggesting that the degeneration of this melanocortin took place before the whole-genome duplication in the teleost lineage. Interestingly, mammalian -MSH binds preferentially to the MC3R, a receptor that is absent from the genome of tetraodontids (Klovins et al. 2004).
More importantly, we observed a high degree of degeneration of the ?-endorphin–coding sequence in teleost POMC? prohormones (fig 1B). ?-endorphin is characterized by the N-terminal sequence YGGFM (or close variants), which is necessary for opioid receptor binding (Reinscheid et al. 1995; Waldhoer, Bartlett, and Whistler 2004). The absence of this canonical motif suggests that tetraodontid POMC? peptides cannot stimulate opioid receptors. The missing opioid signal and the fact that these regions seem to be evolving neutrally (Ka/Ks around 1) suggest that the ?-endorphin regions of tetraodontid POMC? sequences are presently nonfunctional, in contrast to the -MSH/ACTH region, which has strong sequence conservation and low evolving rate (Ka/Ks << 1). In other teleosts, ?-endorphin peptides from POMC? show varying degrees of degeneracy: opioid signatures of POMC? are mutated in zebrafish and barfin flounder but are conserved in salmonids. POMC, on the other hand, has retained a conserved ?-endorphin peptide in all teleosts. Given that salmonids occupy an intermediate phylogenetic position among euteleosts (fig. 2A), it seems that the ?-endorphin region of POMC? started degenerating independently in several lineages, probably as a consequence of compensation provided by POMC and/or the partitioning of POMC expression domains between the two paralogues. The partitioning of the peptide repertoire between ancient POMC paralogues is one of the few examples of subfunctionalization of peptide functional domains, as suggested by Force et al. (1999). The unusual modular structure of the POMC prohormone, with its distinct functional peptides, makes it an ideal target for this sort of subfunctionalization. The evolution of POMC? ?-endorphin is reminiscent to that of -MSH, which also shows distinct levels of degeneration in different lineages of ray-finned fishes: sturgeons and gar have a recognizable but degenerating -MSH sequence, while teleosts lack this melanocortin completely (Dores and Lecaude 2005).
In addition to the subfunctionalization of peptide domains, the expression patterns of POMC and POMC? in the brain and pituitary of Tetraodon also indicate subfunctionalization. POMC mRNA is detected in the PI and rostral pars distalis in the pituitary and in the NLT in the hypothalamus. POMC? mRNA, on the other hand, is detected in the preoptic area (POA) in the brain and, weakly, in the PI of the pituitary. What about other teleosts? In the case of zebrafish, in situ hybridization using a POMC riboprobe detected expression in the pituitary and ventral hypothalamus but not in the POA (Herzog et al. 2003; Liu et al. 2003). However, ACTH- and ?-endorphin–like immunoreactivity is detected in the POA of a close species, the common carp (Metz et al. 2004). Thus, the partitioning of POMC expression domains might be a feature shared by other teleosts.
In mammals, POMC mRNA is mainly detected in the pituitary and in the arcuate nucleus of the hypothalamus. In teleosts, the NLT is considered to be the topological and functional equivalent to the arcuate nucleus. In contrast, the teleost POA is considered to be the functional and topological equivalent to the paraventricular and supraoptic nuclei of the hypothalamus, which do not express POMC in mammals. Is POMC expression in the POA a teleost novelty? It seems not. In the lungfish Protopterus annectens, the pituitary, NLT, and POA show -MSH immunoreactivity (Vallarino, Tranchand Bunel, and Vaudry 1992). Similarly, studies on the frog Xenopus laevis show POMC mRNA expression in the same regions (Tuinhof et al. 1998). These data suggest that POMC was expressed in the pituitary, NLT, and POA of the last common ancestor of lobe- and ray-finned fishes. After the genome duplication in teleosts, the expression domains of the original POMC gene were partitioned between the two newly arisen paralogues so that each of them performs an essential subfunction and, therefore, is kept under selective pressure. In addition, perhaps after POMC? expression was restricted to the POA, its ?-endorphin module started degenerating in different lineages. Thus, it is likely that ?-endorphin does not play an essential role in POA neurons of tetraodontids and, possibly, other teleosts as well.
The subfunctionalization of POMC paralogues in teleosts has some similarities to what happened in a lower vertebrate, the marine lamprey Petromyzon marinus (a jawless fish). This lamprey has two POMC genes with an unique organization regarding their peptide content: proopiomelanotropin (POM) gives rise to two different MSHs and ?-endorphin but no ACTH, whereas proopiocortin (POC) produces ACTH, MSH, and a different ?-endorphin (Takahashi et al. 2001; Dores and Lecaude 2005). POM is expressed in the pituitary PI, while POC is expressed in the RPD and weakly in the PPD, with none of the paralogues being expressed in the lamprey brain (Ficele et al. 1998). Thus, there has also been expression and peptide subfunctionalization between the lamprey paralogues, underlining the tendency of the modular POMC prohormone for subfunction partitioning. The duplication that originated these paralogues is very ancient, being probably specific to the lamprey lineage (Dores and Lecaude 2005). In any event, the POMC duplication in lamprey occurred independently of the duplication we observed in teleosts.
What could be the mechanism for the partitioning of POMC expression domains? As stated in the DDC model (Force et al. 1999), differential degenerative mutations in the regulatory regions of each duplicate gene might have led to the loss of expression of the POMC paralogue in the POA and loss of POMC? expression from the pituitary RPD and NLT, provided that POMC regulatory regions are modular in nature. Figure 6 summarizes the most likely scenario for POMC evolution in the Tetraodon lineage, starting with the loss of -MSH and the origin of two POMC paralogues in the whole-genome duplication until the differential loss of enhancers and ?-endorphin. POMC enhancers are shown as modular blocks that work independently of each other. We have previously shown that the expression of the mammalian POMC gene is controlled by two independent regulatory regions: one is located near the promoter and directs expression to the pituitary (Rubinstein et al. 1993) and the other, composed of two enhancers located around 10 kb upstream of the gene, directs expression to the arcuate nucleus (de Souza et al. 2005). Further studies on the regulatory regions of tetraodontid POMC paralogues, taking advantage of the reduced amount of intergenic DNA in this species, might shed light on the mechanisms of subfunction partitioning of duplicate genes, a crucial process in the evolution of vertebrates.
FIG. 6.— Hypothetical scheme showing the evolution of the POMC gene in teleosts. The genes are shown with the peptide-coding domains and putative enhancers for expression in the nucleus lateralis tuberis (NLT), preoptic area (PO), and pituitary rostral pars distalis (RPD) and pars intermedia (PI). After -MSH is lost from the POMC gene, a whole-genome duplication gives rise to two identical POMC copies, which subsequently undergo enhancer partitioning and, in some teleost lineages including Tetraodon, peptide partitioning. The PI enhancer maintains residual activity in Tetraodon POMC? (denoted by white triangle). Enhancer and peptide partitioning did not occur necessarily in the order presented in the scheme.
Supplementary Material
Supplementary Figure 1 is available at Molecular Biology and Evolution online (http://www.mbe.oxfordjournals.org/).
Acknowledgements
We thank L. F. Franchini for thoughtful comments and suggestions, M. Pandolfi for help with fish anatomy, and A. E. Lima for help with histology. This work was supported in part by a Fogarty International Research Collaborative Award TW01233 (M.J.L., M.R.), NIH grant DK68400 (M.J.L., M.R.), International Research Scholar Grant of the Howard Hughes Medical Institute (M.R.), Agencia Nacional de Promoción Científica y Tecnológica (M.R.), and Universidad de Buenos Aires (M.R.). F.S.J.S. and V.F.B. are recipients of fellowships from the Consejo Nacional de Investigaciones Cientificas y Tecnicas (Argentina).
References
Altschmied, J., J. Delfgaauw, B. Wilde, J. Duschl, L. Bouneau, J. N. Volff, and M. Schartl. 2002. Subfunctionalization of duplicate mitf genes associated with differential degeneration of alternative exons in fish. Genetics 161:259–267.
Altschul, S. F., W. Gish, W. Miller, E. W. Myers, and D. J. Lipman. 1990. Basic local alignment search tool. J. Mol. Biol. 215:403–410.
Amano, M., A. Takahashi, T. Yamanome, Y. Oka, N. Amyia, H. Kawauchi, and K. Yamamori. 2005. Immunocytochemical localization and ontogenic development of -melanocyte stimulating hormone (-MSH) in the brain of a pleuronectiform fish, barfin flounder. Cell Tissue Res. 320:127–134.
Aparicio, S., J. Chapman, E. Stupka et al. (38 co-authors). 2002. Whole-genome shotgun assembly and analysis of the genome of Fugu rubripes. Science 297:1301–1310.
Benton, M. J. 2000. Vertebrate palaeontogy. 2nd edition. Blackwell Science, Oxford.
Christoffels, A., E. G. Koh, J. M. Chia, S. Brenner, S. Aparicio, and B. Venkatesh. 2004. Fugu genome analysis provides evidence for a whole-genome duplication early during the evolution of ray-finned fishes. Mol. Biol. Evol. 21:1146–1151.
David, L., S. Blum, M. W. Feldman, U. Lavi, and J. Hillel. 2003. Recent duplication of the common carp (Cyprinus carpio L.) genome as revealed by analyses of microsatellite loci. Mol. Biol. Evol. 20:1425–1434.
de Souza, F. S., A. M. Santangelo, V. Bumaschny, M. E. Avale, J. L. Smart, M. J. Low, and M. Rubinstein. 2005. Identification of neuronal enhancers of the proopiomelanocortin gene by transgenic mouse analysis and phylogenetic footprinting. Mol. Cell. Biol. 25:3076–3086.
Dores, R. M., and S. Lecaude. 2005. Trends in the evolution of the proopiomelanocortin gene. Gen. Comp. Endocrinol. 142:81–93.
Dores, R. M., C. Sollars, P. Danielson, J. Lee, J. Alrubaian, and J. M. Joss. 1999. Cloning of a proopiomelanocortin cDNA from the pituitary of the Australian lungfish, Neoceratodus forsteri: analyzing trends in the organization of this prohormone precursor. Gen. Comp. Endocrinol. 116:433–444.
Felsenstein, J. 2004. PHYLIP (phylogeny inference package). Version 3.6. Distributed by the author, Department of Genome Sciences, University of Washington, Seattle.
Ficele, G., J. A. Heinig, H. Kawauchi, J. H. Youson, F. W. Keeley, and G. M. Wright. 1998. Spatial and temporal distribution of proopiomelanotropin and proopiocortin mRNA during the life cycle of the sea lamprey: a qualitative and quantitative in situ hybridization study. Gen. Comp. Endocrinol. 110:212–225.
Force, A., M. Lynch, F. B. Pickett, A. Amores, Y. L. Yan, and J. H. Postlethwait. 1999. Preservation of duplicate genes by complementary, degenerative mutations. Genetics 151:1531–1545.
González-Nu?ez, V., R. González-Sarmiento, and R. E. Rodríguez. 2003. Identification of two proopiomelanocortin genes in zebrafish (Danio rerio). Brain Res. Mol. Brain Res. 120:1–8.
Hadley, M. E., and C. Haskell-Luevano. 1999. The proopiomelanocortin system. Ann. N. Y. Acad. Sci. 885:1–21.
Hansen, I. A., T. T. To, S. Wortmann, T. Burmester, C. Winkler, S. R. Meyer, C. Neuner, M. Fassnacht, and B. Allolio. 2003. The pro-opiomelanocortin gene of the zebrafish (Danio rerio). Biochem. Biophys. Res. Commun. 303:1121–1128.
Herzog, W., X. Zeng, Z. Lele, C. Sonntag, J. W. Ting, C. Y. Chang, and M. Hammerschmidt. 2003. Adenohypophysis formation in the zebrafish and its dependence on sonic hedgehog. Dev. Biol. 254:36–49.
Hoegg, S., H. Brinkmann, J. S. Taylor, and A. Meyer. 2004. Phylogenetic timing of the fish-specific genome duplication correlates with the diversification of teleost fish. J. Mol. Evol. 59:190–203.
Huang, F. D., H. J. Matthies, S. D. Speese, M. A. Smith, and K. Broadie. 2004. Rolling blackout, a newly identified PIP2-DAG pathway lipase required for Drosophila phototransduction. Nat. Neurosci. 7:1070–1078.
Hurst, L. D. 2002. The Ka/Ks ratio: diagnosing the form of sequence evolution. Trends Genet. 18:486.
Inoue, J. G., M. Miya, K. Tsukamoto, and M. Nishida. 2003. Basal actinopterygian relationships: a mitogenomic perspective on the phylogeny of the "ancient fish". Mol. Phylogenet. Evol. 26:110–120.
———. 2004. Mitogenomic evidence for the monophyly of elopomorph fishes (Teleostei) and the evolutionary origin of the leptocephalus larva. Mol. Phylogenet. Evol. 32:274–286.
Jaillon, O., J. M. Aury, F. Brunet et al. (47 co-authors). 2004. Genome duplication in the teleost fish Tetraodon nigroviridis reveals the early vertebrate proto-karyotype. Nature 431:946–957.
Jones, D. T., W. R. Taylor, and J. M. Thornton. 1992. The rapid generation of mutation data matrices from protein sequences. CABIOS 8:275–282.
Kikugawa, K., K. Katoh, S. Kuraku, H. Sakurai, O. Ishida, N. Iwabe, and T. Miyata. 2004. Basal jawed vertebrate phylogeny inferred from multiple nuclear DNA-coded genes. BMC Biol. 2:3.
Klovins, J., T. Haitina, D. Fridmanis, Z. Kilianova, I. Kapa, R. Fredriksson, N. Gallo-Payet, and H. B. Schi?th. 2004. The melanocortin system in Fugu: determination of POMC/AGRP/MCR gene repertoire and synteny, as well as pharmacology and anatomical distribution of the MCRs. Mol. Biol. Evol. 21:563–579.
Kumar, S., and S. B. Hedges. 1998. A molecular timescale for vertebrate evolution. Nature. 392:917–920.
Liu, N. A., H. Huang, Z. Yang, W. Herzog, M. Hammerschmidt, S. Lin, and S. Melmed. 2003. Pituitary corticotroph ontogeny and regulation in transgenic zebrafish. Mol. Endocrinol. 17:959–966.
Liu, R. Z., Q. Sun, C. Thisse, B. Thisse, J. M. Wright, and E. M. Denovan-Wright. 2005. The cellular retinol-binding protein genes are duplicated and differentially transcribed in the developing and adult zebrafish (Danio rerio). Mol. Biol. Evol. 22:469–477.
Low, M. J. 2003. Somatostatin. Pp. 379–388 in H. L. Henry and A. W. Norman, eds. Encyclopedia of hormones. Elsevier Science, San Diego, Calif.
Ludwig, A., N. M. Belfiore, C. Pitra, V. Svirsky, and I. Jenneckens. 2001. Genome duplication events and functional reduction of ploidy levels in sturgeon (Acipenser, Huso and Scaphirhynchus). Genetics 158:1203–1215.
McClintock, J. M., M. A. Kheirbek, and V. E. Prince. 2002. Knockdown of duplicated zebrafish hoxb1 genes reveals distinct roles in hindbrain patterning and a novel mechanism of duplicate gene retention. Development 129:2339–2354.
Metz, J. R., M. O. Huising, J. Meek, A. J. Taverne-Thiele, S. E. Wendelaar-Bonga, and G. Flik. 2004. Localization, expression and control of adrenocorticotropic hormone in the nucleus preopticus and pituitary gland of the common carp (Cyprinus carpio L.). J. Endocrinol. 182:23–31.
Palti, Y., S. A. Gahr, J. D. Hansen, and C. E. Rexroad III. 2004. Characterization of a new BAC library for rainbow trout: evidence for multi-locus duplication. Anim. Genet. 35:130–133.
Pandolfi, M., M. M. Canepa, M. A. Ravaglia, M. C. Maggese, D. A. Paz, and P. G. Vissio. 2003. Melanin-concentrating hormone system in the brain and skin of the cichlid fish Cichlasoma dimerus: anatomical localization, ontogeny and distribution in comparison to -melanocyte-stimulating hormone-expressing cells. Cell Tissue Res. 311:61–69.
Postlethwait, J., A. Amores, W. Cresko, A. Singer, and Y. L. Yan. 2004. Subfunction partitioning, the teleost radiation and the annotation of the human genome. Trends Genet. 20:481–490.
Reinscheid, R. K., H. P. Nothacker, A. Bourson, A. Ardati, R. A. Henningsen, J. R. Bunzow, D. K. Grandy, H. Langen, F. J. Monsma Jr., and O. Civelli. 1995. Orphanin FQ: a neuropeptide that activates an opioidlike G protein-coupled receptor. Science 270:792–794.
Rozas, J., and R. Rozas. 1999. DnaSP version 3: an integrated program for molecular population genetics and molecular evolution analysis. Bioinformatics 15:174–175.
Rubinstein, M., M. Mortrud, B. Liu, and M. J. Low. 1993. Rat and mouse proopiomelanocortin gene sequences target tissue-specific expression to the pituitary gland but not to the hypothalamus of transgenic mice. Neuroendocrinology 58:373–380.
Salbert, G., I. Chauveau, G. Bonnec, Y. Valotaire, and P. Jego. 1992. One of the two trout proopiomelanocortin messenger RNAs potentially encodes new peptides. Mol. Endocrinol. 6:1605–1613.
Takahashi, A., M. Amano, T. Itoh, A. Yasuda, T. Yamanome, Y. Amemyia, T. Sasaki, M. Sakai, K. Yamamori, and H. Kawauchi. 2005. Nucleotide sequence and expression of three subtypes of proopiomelanocortion RNA in barfin flounder. Gen. Comp. Endocrinol. 141:291–303.
Takahashi, A., Y. Amemiya, M. Nozaki, S. A. Sower, and H. Kawauchi. 2001. Evolutionary significance of proopiomelanocortin in agnatha and chondrichthyes. Comp. Biochem. Physiol. B Biochem. Mol. Biol. 129:283–289.
Taylor, J. S., I. Braasch, T. Frickey, A. Meyer, and Y. Van de Peer. 2003. Genome duplication, a trait shared by 22,000 species of ray-finned fish. Genome Res. 13:382–390.
Taylor, J. S., Y. Van de Peer, I. Braasch, and A. Meyer. 2001. Comparative genomics provides evidence for an ancient genome duplication event in fish. Philos. Trans. R. Soc. Lond. B Biol. Sci. 356:1661–1679.
Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22:4673–4680.
Tuinhof, R., R. Ubink, S. Tanaka, C. Atzori, F. J. van Strien, and E. W. Roubos. 1998. Distribution of pro-opiomelanocortin and its peptide end products in the brain and hypophysis of the aquatic toad, Xenopus laevis. Cell Tissue Res. 292:251–265.
Vallarino, M., D. Tranchand Bunel, and H. Vaudry. 1992. -melanocyte-stimulating hormone (-MSH) in the brain of the African lungfish, Protopterus annectens: immunohistochemical localization and biochemical characterization. J. Comp. Neurol. 322:266–274.
Vandepoele, K., W. De Vos, J. S. Taylor, A. Meyer, and Y. Van de Peer. 2004. Major events in the genome evolution of vertebrates: paranome age and size differ considerably between ray-finned fishes and land vertebrates. Proc. Natl. Acad. Sci. USA 101:1638–1643.
Volff, J. N. 2005. Genome evolution and biodiversity in teleost fish. Heredity 94:280–294.
Volkoff, H., L. F. Canosa, S. Unniappan, J. M. Cerda-Reverter, N. J. Bernier, S. P. Kelly, and R. E. Peter. 2005. Neuropeptides and the control of food intake in fish. Gen. Comp. Endocrinol. 142:3–19.
Waldhoer, M., S. E. Bartlett, and J. L. Whistler. 2004. Opioid receptors. Annu. Rev. Biochem. 73:953–990.
Woods, I. G., P. D. Kelly, F. Chu, P. Ngo-Hazelett, Y. L. Yan, H. Huang, J. H. Postlethwait, and W. S. Talbot. 2000. A comparative map of the zebrafish genome. Genome Res. 10:1903–1910.(Flávio S. J. de Souza*,1,)
E-mail: mrubins@dna.uba.ar.
Abstract
The proopiomelanocortin gene (POMC) encodes several bioactive peptides, including adrenocorticotropin hormone, -, ?-, and -melanocyte-stimulating hormone, and the opioid peptide ?-endorphin, which play key roles in vertebrate physiology. In the human, mouse, and chicken genomes, there is only one POMC gene. By searching public genome projects, we have found that Tetraodon (Tetraodon nigroviridis), Fugu (Takifugu rubripes), and zebrafish (Danio rerio) possess two POMC genes, which we called POMC and POMC?, and we present phylogenetic and mapping evidence that these paralogue genes originated in the whole-genome duplication specific to the teleost lineage over 300 MYA. In addition, we present evidence for two types of subfunction partitioning between the paralogues. First, in situ hybridization experiments indicate that the expression domains of the ancestral POMC gene have been subfunctionalized in Tetraodon, with POMC expressed in the nucleus lateralis tuberis of the hypothalamus, as well as in the rostral pars distalis and pars intermedia (PI) of the pituitary, whereas POMC? is expressed in the preoptic area of the brain and weakly in the pituitary PI. Second, POMC? genes have a ?-endorphin segment that lacks the consensus opioid signal and seems to be under neutral evolution in tetraodontids, whereas POMC genes possess well-conserved peptide regions. Thus, POMC paralogues have experienced subfunctionalization of both expression and peptide domains during teleost evolution. The study of regulatory regions of fish POMC genes might shed light on the mechanisms of enhancer partitioning between duplicate genes, as well as the roles of POMC-derived peptides in fish physiology.
Key Words: POMC ? teleosts ? evolution ? Tetraodon ? ?-endorphin ? subfunctionalization
Introduction
Proopiomelanocortin (POMC) is a prohormone that gives rise to several different bioactive peptides through cell-type–specific posttranslational processing. POMC-derived peptides are produced mainly in the pituitary gland and the brain, being involved in processes as divergent as stress, pain, energy balance, and, in fishes and amphibians, adaptation to background color (reviewed in Hadley and Haskell-Luevano 1999; Volkoff et al. 2005). In the pituitary gland, the POMC gene is expressed in two types of cells: corticotrophs in the pars distalis, where POMC is processed into adrenocorticotropin hormone (ACTH) and ?-lipotropin and melanotrophs in the pars intermedia (PI), where it gives rise to -melanocyte-stimulating hormone (-MSH) and ?-endorphin. In the vertebrate brain, POMC is mainly expressed in the ventral hypothalamus, where POMC is processed into -MSH, ?-MSH, -MSH, and ?-endorphin.
Genes and cDNAs encoding POMC have been cloned and sequenced from representatives of all vertebrate classes, including several fishes (Takahashi et al. 2001). Birds and mammals have only one POMC gene, but some fishes and amphibians possess more than one. Gene duplicates may arise through a variety of mechanisms, ranging from single-gene to whole-genome duplication. The recent sequencing and assembly of the genome of the teleost Tetraodon nigroviridis (family Tetraodontidae), as well as studies on gene duplicates in the genome of another tetraodontid, Takifugu rubripes (Fugu), give strong support to the idea that an ancient whole-genome duplication occurred in the teleost lineage around 320 MYA (Taylor et al. 2001; Christoffels et al. 2004; Jaillon et al. 2004; Vandepoele et al. 2004). In addition, some fishes like carp, rainbow trout, and sturgeons have undergone recent polyploidization events (Volff 2005).
After a gene is duplicated, one of the copies usually accumulates nonsense mutations and becomes a pseudogene, although sometimes both paralogue copies are retained in the genome. In the case of tetraodontids, the percentage of duplicate genes retained after the proposed teleost-specific duplication has been calculated to be 3%–7% (Christoffels et al. 2004; Jaillon et al. 2004). The retention of the duplicated copies may be accomplished when one of the paralogues acquires a new function in relation to the original gene (neofunctionalization) or, alternatively, the functions of the original gene are partitioned between the new duplicates, as stated in the duplication-degeneration-complementation (DDC) model (Force et al. 1999; Postlethwait et al. 2004). The latter process, known as subfunctionalization, is most often evidenced by the partitioning of the original expression pattern. For instance, if the original gene was expressed in regions A and B, the differential loss of regulatory regions might cause one duplicate to be expressed in region A and the other in region B so that both paralogues are kept under selective pressure and become fixed in the genome. In teleosts, partitioning of expression domains has been observed in several paralogue genes, including the transcription factors Engrailed1 (Force et al. 1999), Hoxb1 (McClintock, Kheirbek, and Prince 2002), and Mitf (Altschmied et al. 2002), retinol-binding proteins (Liu et al. 2005), the neuropeptide somatostatin (Low 2003), among others. In addition, subfunctionalization may also involve the differential partitioning of peptide domains (Force et al. 1999).
Here, we present molecular and phylogenetic evidence that higher teleosts possess two POMC paralogue genes, which we named POMC and POMC?, that originated in the ancient teleost whole-genome duplication. In addition, we show that both Tetraodon POMC gene duplicates have undergone subfunctionalization, as evidenced by their complementary expression patterns in brain and pituitary. Moreover, the peptide repertoire of POMC paralogues has also been subfunctionalized, with the loss of ?-endorphin from many teleost POMC? genes.
Materials and Methods
Animals
Adult fish from the genus Tetraodon (T. nigroviridis and T. fluviatilis) and zebrafish (Danio rerio) were purchased from a local aquarium and processed immediately for nucleic acid extraction or tissue fixation. Experimental protocols were approved by the local animal care and use committees and were consistent with the Guide for the Care and Use of Laboratory Animals.
Reverse Transcriptase–Polymerase Chain Reactions and Sequencing
Total RNA from Tetraodon pituitaries and preoptic area was isolated using TRIzol (Invitrogen, Carlsbad, Calif.). For 5' and 3' rapid amplification of cDNA ends (RACE) POMC polymerase chain reactions (PCRs), cDNA was prepared using the MATCHMAKER Library Construction & Screening Kit (Clontech, Palo Alto, Calif.) incorporating a SMART oligo at the 5' end and using CDS III–modified oligo-dT, according to the instructions of the manufacturer. Nested PCRs were performed with the following primer pairs: for 5' RACE, 5'-ACGCAGAGTGGCCATTATGGCCGGG-3' (complementary to the SMART oligo)/5'-CATGCTGCTGGCCTCGGAGT-3' (complementary to exon 2); and for 3' RACE, 5'-GAGCGCCCACCTCCAACT-3' (complementary to exon 3)/5'-GTATCGATGCCCACCCTCTAGAGGCCGAGGCGGCCGACA-3' (complementary to CDS III oligo). To determine the exon-intron structure of POMC genes, PCRs were performed with the following primer pairs: for POMC, 5'-CCATCAGCAGCCAGCACGA-3' (complementary to exon 1)/5'-CCAACAACCGTCATGAGC-3' (complementary to exon 3); for POMC?, 5'-ACGGCGGGAGGAAAGCAA-3' (complementary to exon 1)/5'-CAGTATCCTCCTCTTGTC-3' (complementary to exon 2); and 5'-GCTGATGACGTACCTGTG-3' (complementary to exon 2)/5'-TGCCCTCGCTCGTCCCTGT-3' (complementary to exon 3). Gel-purified PCR bands were subjected to automated DNA sequencing.
Immunohistochemistry
Adult fishes were anaesthetized by immersion in a neutral buffered solution of 0.1% (w/v) tricaine methane sulfonate (Sigma, St. Louis, Mo.), and heads were fixed overnight in Bouin's fixative. Brains with attached pituitaries and part of the skull (cranium base) were dissected and washed with 70% (v/v) ethanol for 24 h, cryoprotected with 30% (w/v) sucrose, and frozen in embedding medium (Tissue Tek, Sakura Tek, Torrance, Calif.). Serial transversal and sagital sections (20 μm) were cut using a cryostat microtome (IEC Microtome, Walldorf, Germany) at –20°C. Sections were thaw mounted on Vectabond-coated (Vector Laboratories, Burlingame, Calif.) slides and stored at –70°C until use. Alternatively, some brains were embedded into paraffin and 4- or 10-μm slices were cut in a microtome (SM2000R, Leica, Nussloch, Germany). Immunohistochemistry proceeded as described previously (de Souza et al. 2005) with a polyclonal antihuman ACTH IC-1 antibody (1:300; National Hormone and Pituitary Program, National Institutes of Health).
Radioactive In Situ Hybridization
Cryosections obtained as described above were treated with 10 μg/ml proteinase K in 10 x TE buffer (100 mM Tris, 10 mM ethylenediaminetetraacetic acid [EDTA]) at room temperature (RT) for 15 min and then washed twice with the same buffer. Slides were then treated with 0.1 M triethanolamine (TEA) pH 8.0 for 3 min, incubated in 0.0025% (v/v) acetic anhydride in 0.1 M TEA for 10 min, washed twice in 2 x standard saline citrate (SSC) (NaCl, sodium citrate), and finally dehydrated quickly in ascending ethanol concentrations. For prehybridization, slides were incubated under parafilm coverslips with the hybridization buffer (66% [v/v] formamide, 260 mM NaCl, 1.3 x Denhardt's, 1.3 mM EDTA, 13 mM Tris-HCl pH 8.0, 13% [w/v] dextran sulfate) for 1 h/57°C. [35S]-labeled riboprobes were heated to 65°C/5 min (with 0.5 mg/ml tRNA, 10 mM dithiothreitol [DTT] in water) and added to the hybridization buffer in a concentration of 5 x 106–107 cpm/ml. Hybridization followed overnight at 57°C. Next day, slices were washed four times in 4 x SSC and subjected to ribonuclease (RNase) A digestion (20 μg/ml RNase A, 0.5 M NaCl, 10 mM Tris-HCl pH 8.0, 1 mM EDTA pH 8.0) at 37°C/30 min, desalted in 2 x, 1 x, 0.5 x SSC/1 mM DTT, and subjected to a final wash in 0.1 x SSC/1 mM DTT at 30 min/65°C. After a quick rinse in 0.1 x SSC/1 mM DTT RT, slides were dehydrated through 5-min steps in ethanol at 70%, 95%, and 100% (v/v), dried at room temperature, dipped in Kodak N-TB2 emulsion (diluted 1:1 in distilled water), and exposed for 3–6 days. Slides were developed, counterstained with cresyl violet (Sigma), dehydrated with ethanol, and coverslipped under Permount (Fisher Scientific, Pittsburgh, Pa.). To prepare riboprobes, PCR fragments of exon 2 of Tetraodon POMC (123-bp product; TE2SE: 5'-CCAGAATGAGTCCTGCGTGGCT-3' and TE2AN: 5'-CATGCTGCTGGCCTCGGAGT-3') and exon 3 of POMC? (302-bp product; 2TPE3SE4: 5' CGGAGCAGCAACAGGGAC 3' and 2TPE3AN2: 5' AACGCCTGCGGGCACGGA-3') were cloned into pGEMT-Easy vector (Promega, Madison, Wisc.). Antisense RNA probes were synthesized by incubating 1 μg of linearized plasmid with transcription buffer (40 mM Tris-HCl, pH 8.0, 15 mM MgCl2, 5 mM DTT, 0.05 mg/ml bovine serum albumin), 10 mM DTT, 60 U RNase out (Invitrogen), 0.5 mM of adenosine triphosphate, cytidine triphosphate, and guanosine triphosphate, 1,000 μCi [35S]UTP (1,250 Ci/mmol; PerkinElmer, Wellesley, Mass.), and 16 U of T7 polymerase (Amersham Biosciences, Piscataway, N.J.). After 70 min of incubation at 37°C, plasmid was digested with DNAse I (Invitrogen). Riboprobes were precipitated with 10 μg tRNA, 0.1 vol 5 M NH4 acetate, and 2.5 vol 100% (v/v) ethanol for 1 h at –70°C, washed with 70% (v/v) ethanol, and resuspended in RNAse-free water.
Sequences and Databases
The genomic regions around POMC loci from Fugu and Tetraodon were located via Blast searches (Altschul et al. 1990) in the Ensembl database Web site (http://www.ensembl.org/index.html) using POMC sequences from other species. POMC loci are present in the scaffolds 970 in Fugu and 10884 in Tetraodon, while POMC? loci are located in scaffold 130 in Fugu and chromosome 14 in Tetraodon. A complete cDNA sequence for Fugu POMC was assembled from expressed sequence tag (EST) clones EFRz001apsG3, EFRa090apsE11, and EFRz002apsD6 retrieved from the Fugu Home Page of the Rosalind Franklin Centre for Genomic Research (http://www.hgmp.mrc.ac.uk). A cDNA sequence for Tetraodon POMC? (clone name FD0ADA50CB12AAP1) was retrieved from the Genoscope Home Page (http://www.genoscope.cns.fr/externe/tetranew). Data on the POMC neighborhood in chicken (Gallus gallus) and humans were obtained from the Ensembl and Mapviewer Web sites (http://www.ncbi.nih.gov/mapview/), respectively. Zebrafish POMC and POMC? were located, respectively, in sequences AL590149 (from linkage group [LG] 17) and AL845420 (from LG 20) deposited in GenBank (http://www.ncbi.nih.gov). We also retrieved from GenBank the POMC peptide and cDNA sequences from the following species and accession numbers: Verasper moseri (barfin flounder) POMC-A (AB051424); POMC-B (AB051425) and POMC-C (AB051426); Oncorhynchus mykiss (rainbow trout) POMC-A (X69808) and POMC-B (X69809); Oncorhynchus keta (chum salmon) (X01122); Polypterus senegalus (bichir) (AF465781); Acipenser trasmontanus (sturgeon) POMC-A (AF092937) and POMC-B (AF092936); Polyodon spathula (paddlefish) POMC-A (AF117302) and POMC-B (AF117303); Cyprinus carpio (common carp) POMC-I (Y14618) and POMC-II (Y14617); D. rerio (zebrafish) (NM_181438); Homo sapiens (man) (NM_000939); Carassius auratus (goldfish) (AJ431209); Ictalurus puntactus (channel catfish) (AY174050); Paralichthys olivaceus (bastard halibut) POMC-I (AF184066) and POMC-II (AF191593); Anguilla japonica (Japanese eel) (AY158010); and Anguilla rostrata (American eel) (AF194969).
Evolutionary Analyses
POMC peptide sequences were aligned with the ClustalW program (http://www.ebi.ac.uk/clustalw) (Thompson, Higgins, and Gibson 1994). The alignment used for phylogenetic inference using the PHYLIP 3.62 Program Package (Felsenstein 2004) is shown in Supplementary Figure 1 (Supplementary Material online). For maximum-likelihood (ML) phylogenetic inference, the PROTML program was used with the amino acid substitution model of Jones, Taylor, and Thornton (1992) using 100 replicates generated with the SEQBOOT program. A consensus tree was created with CONSENSE. Ka/Ks ratios were calculated using the DnaSP 4.0 program (J. Rozas and R. Rozas 1999) downloaded from http://www.ub.es/dnasp.
Results
Identification of Duplicate POMC Genes in Teleost Fishes
We searched the public genome projects of Tetraodon (Jaillon et al. 2004) and Fugu (Aparicio et al. 2002) using the Ensembl database Web site. In each tetraodontid genome, we identified two POMC genes which we named POMC and POMC?, according to their identity to mammalian POMC at the peptide level (45% and 35%, respectively). In the Tetraodon genome project, POMC? has been assigned to chromosome 14, while POMC is found in scaffold 10884 of unknown location. In the Fugu genomic database, where scaffolds have not yet been assigned to chromosomes, POMC and POMC? are located in scaffolds 970 and 130, respectively.
The complete cDNA from Fugu POMC was assembled from ESTs retrieved from the Fugu database of the Rosalind Franklin Centre for Genome Research, and this sequence was used to deduce the exon-intron structure of both Fugu and Tetraodon POMC genes. Reverse transcriptase (RT)–PCR and 5'/3'-RACE reactions using Tetraodon pituitary cDNA were done to confirm exon-intron boundaries and the transcriptional start site. The structure of tetraodontid POMC? genes was defined using a Tetraodon POMC? cDNA clone from the Genoscope Web site, and exon-intron boundaries were confirmed by sequencing RT-PCR fragments. Both tetraodontid POMC genes have three exons: exon 1 is noncoding, exon 2 contains the starting ATG, and most of the coding region lies in exon 3 (fig. 1A). This organization is remarkably similar to that of POMC genes of all tetrapods studied to date.
FIG. 1.— (A) Organization of Tetraodon POMC paralogues compared to the human POMC gene. Note the shorter size of introns in Tetraodon. (B) Alignment of Tetraodon POMC amino acid sequences with those of different vertebrates. Peptides derived from POMC are shown on top of the sequences. Putative dibasic cleavage sites are indicated by arrowheads. Asterisks and dots below the sequences indicate identical and conservative residues, respectively. Melanocortin and opioid core sequences are underlined. Note that both teleost POMC and POMC? lack -MSH and that Tetraodon and Fugu POMC? lack the opioid signal.
A gene and a cDNA encoding zebrafish POMC have been described previously (Hansen et al. 2003; Liu et al. 2003). In addition, González-Nu?ez, González-Sarmiento, and Rodríguez (2003) have reported a second, incomplete cDNA corresponding to a more divergent POMC sequence, isolated from a brain cDNA library. Based on the phylogeny of the genes (see below), we propose to call these zebrafish genes POMC and POMC?, respectively. Searching the GenBank database, we identified zebrafish genomic scaffolds of LGs 17 and 20 that contain POMC and POMC?, respectively. Comparing the reported partial POMC? cDNA (González-Nu?ez, González-Sarmiento, and Rodríguez 2003) with genomic sequence, we identified two exons which, as judged from sequence alignments, contain the whole coding region of zebrafish POMC?. These exons correspond probably to exons 2 and 3 because all POMC genes studied to date contain a noncoding exon 1. A putative exon 1 could not be identified from sequence comparisons.
Characterization of Duplicate POMC Peptides
Figure 1B shows an alignment of deduced Tetraodon, Fugu, and zebrafish POMC and POMC? amino acid sequences and those of other representative vertebrates. We have included in the alignment the sequences of the reported POMC-B from rainbow trout (O. mykiss—Salbert et al. 1992) and the POMC-C from barfin flounder (Verasper moseri—Takahashi et al. 2005), which we have assigned to the POMC? group. In general, conservation of typical POMC features is high between teleost POMC sequences but lower in POMC? sequences. The amino acid sequence of -MSH is highly conserved (93%–100%) in POMC and POMC? in all species (fig. 1B). ?-MSH is also conserved, but zebrafish POMC? lacks a dibasic cleavage site at the prospective amino-terminal of the peptide (fig. 1B). The hormone ACTH, which encompasses -MSH and CLIP, is potentially derived from POMC and POMC? in all species except zebrafish POMC?, which lacks a dibasic cleavage site at the C-terminal of the prospective peptide (fig. 1B; González-Nu?ez, González-Sarmiento, and Rodríguez 2003). Amino acid identity between human and teleost ACTH is 74%–77% for POMC and 59% for POMC? sequences. Both teleost POMC and POMC? sequences lack the region corresponding to -MSH (fig. 1B), a third melanocortin which is present in other vertebrate POMC sequences (Dores and Lecaude 2005).
The region corresponding to the ?-endorphin peptide is more conserved in teleost POMC than POMC? sequences (fig. 1B). Amino acid identity between human and teleost ?-endorphins ranges from 59% to 67% for POMC and only 20%–44% for POMC?. The N-terminal met-enkephalin sequence of ?-endorphin, which contains the opioid motif YGGFM, presents several degrees of degeneration in teleost POMC? sequences. The motif is changed to YGGLM in zebrafish POMC? (González-Nu?ez, González-Sarmiento, and Rodríguez 2003) and to SGRFM in barfin flounder POMC-C (Takahashi et al. 2005; fig. 1B). Strikingly, the opioid motif is completely absent from POMC? ?-endorphin sequences in Tetraodon and Fugu (fig. 1B). Because the opioid signal is necessary for stimulation of the opioid receptors (Reinscheid et al. 1995; Waldhoer, Bartlett, and Whistler 2004), it is likely that POMC? from tetraodontids and, possibly, barfin flounder and zebrafish are incapable of generating a ?-endorphin with opioid activity. Nevertheless, ?-endorphin peptides containing the canonical opioid signal are present in the POMC-B sequences of trout (fig. 1B) and salmon, indicating that the peptide repertoire of POMC? sequences varies among teleosts.
The absence of opioid activity in ?-endorphin of Tetraodon and Fugu POMC? indicates that the C-terminal of POMC? might be nonfunctional and, consequently, might be accumulating a high number of nonsynonymous mutations. To check this possibility, we calculated the ratio of the number of nonsynonymous substitutions per nonsynonymous sites (Ka) to the number of synonymous substitutions per synonymous sites (Ks) (Hurst 2002) of the ?-endorphin regions of tetraodontid POMC and POMC? genes. ?-endorphins from POMC showed a Ka/Ks much lower than 1 (Ka/Ks = 0.12), indicating that this region is under purifying selection in POMC genes. In contrast, ?-endorphins of POMC? have a Ka/Ks slightly higher than 1 (Ka/Ks = 1.26), suggesting that the ?-endorphin region of POMC? came under relaxed selection in the tetraodontid lineage. ACTH regions, in contrast, seem to be under purifying selection in both tetraodontid POMC (Ka/Ks = 0.36) and POMC? (Ka/Ks = 0.17) genes. The unique POMC gene of distantly related sarcopterygians like humans and Australian lungfish (Neoceratodus forsteri, Dores et al. 1999) is under strong purifying selection in both the ?-endorphin (Ka/Ks = 0.15) and ACTH (Ka/Ks = 0.07) regions. We found no strong evidence for relaxed or positive selection in ACTH or ?-endorphin regions in other fish lineages.
Phylogenetic Analysis of Teleost POMC Genes
POMC cDNAs have been sequenced from several phylogenetically diverse species of fishes, and POMC duplicates have been identified in some of them. To understand the origin of POMC paralogues in teleost genomes, we constructed phylogenetic trees of POMC amino acid sequences using ML analyses. Figure 2B shows an ML phylogenetic tree of POMC peptides of ray-finned fishes (Actinopterygii). Supplementary Figure 1 (Supplementary Material online) shows the alignment used to build this tree. To serve as reference, figure 2A shows the phylogenetic relationships between selected fishes adapted from Kikugawa et al. (2004) and Inoue et al. (2004). Recent molecular phylogenetic studies agree on the relationships of most groups shown, but there is controversy over the relationship between Semionotiformes (gar) and Acipenseriformes (sturgeons and paddlefishes). For a similar but slightly different molecular phylogeny, see Inoue et al. (2003). The POMC phylogenetic tree shown in figure 2B corresponds roughly to the phylogenetic relationships between the fish species. POMC and POMC? sequences from euteleosts (higher teleosts), as well as POMC sequences from two eel species (which are basal teleosts), are grouped in a branch with 100% bootstrap support (fig. 2B). Importantly, the analysis supports (99% bootstrap) a branch with the POMC? sequences of zebrafish, Tetraodon, and Fugu, together with rainbow trout POMC-B, barfin flounder POMC-C, and a POMC sequence from chum salmon. POMC sequences from euteleosts are grouped in another branch, being monophyletic in the tree, although with relatively low statistical support (74%). The available eel POMC sequences cluster with the POMC? branch with moderate statistical support (84% bootstrap).
FIG. 2.— Phylogenetic relationships between POMC sequences. (A) Putative phylogenetic relationships between relevant vertebrate groups adapted from Kikugawa et al. (2004) and Inoue et al. (2004). Putative timing of teleost whole-genome duplication (WGD) (Hoegg et al. 2004) is indicated by an arrow. (B) ML phylogenetic tree of POMC amino acid sequences of ray-finned fishes. Percentage bootstrap values (100 replicates) are indicated. The node grouping teleost POMC sequences is highlighted with a black circle, and POMC and POMC? clades are indicated. See Materials and Methods for details on the species.
Our phylogenetic results indicate that the duplication event that originated the POMC and POMC? paralogues happened in the teleost lineage, after the branching of the longnose gar lineage. Within euteleosts, the duplication seems to have happened before the separation of the superorders Ostariophysii (zebrafish, goldfish, and carp), Protacanthopterygii (salmon and trout), and Acanthopterygii (Tetraodon, Fugu, barfin flounder, and halibut). POMC sequences from eels, which are not considered euteleleosts, are grouped in the same branch as euteleost POMC and POMC?, suggesting that the POMC gene duplication happened before the branching of elopomorphs (eels and relatives) from the lineage that led to euteleosts (see Discussion).
Apart from the duplication that originated POMC and POMC?, there have been other duplications involving the POMC gene in the history of teleosts. As evidenced in fig. 2B, there are duplicates of POMC (POMC-I and -II) in the halibut and carp. In the case of barfin flounder, there are three POMC sequences reported: two belonging to the POMC group (POMC-A and -B) and another to the POMC? group (POMC-C). In addition, there are two similar POMC duplicates in the primitive chondrostean fishes sturgeons and paddlefish (POMC-A and -B). These very similar pairs of POMC genes in teleosts and chondrosteans may have arisen in recent, independent genome duplications in these lineages (see Discussion).
Neighborhood of POMC Genes
To determine whether POMC and POMC? originated in a single-gene duplication or in a higher order event (segmental, chromosomal, or whole-genome duplication), we compared the gene content around POMC loci from several species. POMC and POMC? genes from fish, as well as the POMC genes from mammals, chicken, and the frog Xenopus tropicalis are all flanked at the 3' end by a 21-exon gene of unknown function, dubbed KIAA0953 (fig. 3). KIAA0953 encodes a protein similar to a Drosophila transmembrane lipase (Huang et al. 2004). Phylogenetic analyses show that the KIAA0953 genes downstream of all POMC genes are homologues (not shown). At the 5' side, both zebrafish POMC genes and the unique chicken POMC gene in chromosome 3 are flanked by a gene encoding an angiopoietin-like factor (ANGPL, fig. 3). Tetraodon and Fugu POMC? genes also have an ANGPL gene located at the 5' side, although separated from POMC? by a few intervening genes. Phylogenetic analyses show that the angiopoietin-like factors near the POMC genes from zebrafish, tetraodontids, and chicken are true homologues (data not shown). In addition, the gene encoding sortin nexin 9 (SNX9) is conserved at the 5' side of teleost POMC genes (fig. 3).
FIG. 3.— Analysis of POMC neighborhood in different vertebrate genomes. Genomic scaffolds and chromosomes where the loci are located in each species are indicated. Gene neighborhoods around Fugu POMC loci (not shown) are identical to that of Tetraodon. Genes were named according to the human orthologs. Only genes showing homology to others are shown (black bars). White square denotes omitted intervening genes. The dotted line denotes change in gene order. Arrowheads show the direction of transcription of POMC genes. Hypothetical reconstructions of POMC gene neighborhood in the ancestor of Tetraodon and zebrafish (teleost ancestor) and teleosts and tetrapods (bony fish ancestor) are shown.
We conclude that the gene order ANGPL-POMC-KIAA0953 was likely found in the last common ancestor of ray-finned and lobe-finned fishes (fig. 3), which lived 400–450 MYA (Kumar and Hedges 1998). In addition, the gene order SNX9-ANGPL-POMC-KIAA0953 was probably present in the ancestor of higher teleosts. The conservation of gene content around duplicated POMC loci in teleosts indicates that these paralogues are not the result of a single-gene duplication, but rather of a higher order duplication event. The possibility of a chromosomal or genome duplication is further supported by the fact that the zebrafish POMC loci are located in different chromosomes.
Partitioning of Expression Domains of POMC in Tetraodon
The persistence of two ancient POMC duplicates in teleosts suggests that each of them performs a unique function. Because it has been observed that gene duplicates have frequently partitioned their expression domains, we decided to check this possibility for POMC paralogues. At first, to verify which regions express the POMC protein, we performed immunohistochemistry on slices of brain and pituitary of adult Tetraodon using an anti-ACTH antibody that also recognizes the whole POMC prohormone. Because the ACTH region is very similar in both Tetraodon POMC paralogues (fig. 1B), it is expected that the antibody will recognize the products of both genes.
As shown in figure 4, ACTH-positive perykaria were found in three regions: (1) the pituitary gland, (2) the nucleus lateralis tuberis (NLT) in the ventral hypothalamus, and (3) the preoptic nucleus area (POA). In the pituitary, POMC is found in the rostral pars distalis (RPD), presumably in corticotrophs, as well as in melanotrophs of the PI. No staining was observed in the pars nervosa (PN) or in the posterior pars distalis (PPD). In the ventral hypothalamus, we detected ACTH-like immuoreactivity in a few cells of the NLT. This pattern of expression in the pituitary and NLT of Tetraodon is similar to that of other fish like the barfin flounder (Amano et al. 2005), zebrafish (Herzog et al. 2003; Liu et al. 2003), and the cichlid Cichlastoma dimerus (Pandolfi et al. 2003). In addition to expression in these two areas, we observed ACTH immunoreactivity in the preoptic area, both in the anterior periventricular area as well as in the more dorsoposterior magnocellular region (POm). ACTH immunoreactivity in the preoptic area has already been observed in the carp (Metz et al. 2004) and other nonteleost vertebrates (see Discussion).
FIG. 4.— ACTH immunoreactivity in Tetraodon brain and pituitary. The ACTH antibody recognizes the whole POMC prohormone. Top drawing illustrates the location of POMC-expressing cell groups. Anterior is to the left in the drawing and panels (C) and (D). Arrowheads indicate the planes of the transversal cuts in panels (A) and (B). (A) Transversal cut at the level of the anterior preoptic area (POa) showing ACTH-immunoreactivity cell groups (arrows) around the diencephalic ventricle (V). (B) Transversal cut showing ACTH-positive cells in the magnocellular part of the preoptic area (POm, arrows). (C) Sagital cut showing ACTH immunoreactivity in the rostral pars distalis (RPD) and pars intermedia (PI) of the pituitary. (D) ACTH-positive cells (arrows) in a sagital cut of the nucleus lateralis tuberis (NLT). Other abbreviations: Cb, cerebellum; Ep, epiphysis; OfB, olfactory bulb; OpN, optic nerve; OpTec, optic tectum; MO, medulla oblongata; PPD, posterior pars distalis; PN, pars nervosa; Tel, telencephalon; and Thal, thalamus.
To check whether the expression patterns of POMC might have been partitioned between the paralogues, we perfomed radioactive in situ hybridization with specific POMC and POMC? antisense riboprobes on slices of adult Tetraodon brains and pituitaries. As shown in figure 5, POMC is strongly expressed in pituitary RPD and PI, as well as in a few cells of the NLT. No expression was observed in preoptic nuclei or in other brain areas. POMC?, on the other hand, is expressed in the POA and in the pituitary PI, while no expression was detected in the pituitary RPD, NLT, or other brain area. In the pituitary PI, expression of POMC is much higher than that of POMC? (compare fig. 5A and B). The combined expression patterns of POMC and POMC? correspond to the distribution of ACTH-positive perikarya in Tetraodon pituitary and brain. Because there is strong evidence that POMC was expressed in the POA, NLT, and pituitary RPD and PI of ancestral ray-finned fishes (see Discussion), these results indicate that the expression domains of the POMC gene were partitioned between POMC and POMC? during the evolution of Tetraodon.
FIG. 5.— Radioactive in situ hybridization of POMC paralogues in the brain and pituitary of Tetraodon. Riboprobes are complementary to the exon 2 of POMC (left panels) and the exon 3 of POMC? (right panels). All panels show sagital cuts, anterior to the left, dorsal is up. (A, B) Pituitary cuts. RPD, rostral pars distalis; PI, pars intermedia. (C, D) Nucleus lateralis tuberis (NLT) cuts. (E, F) Preoptic area (POA) cuts; OpN, optic nerve. Note that POMC is expressed in the RPD, PI, and NLT, while POMC? is expressed in the POA and weakly in the PI.
Discussion
In this work, we show that: (1) Tetraodon and Fugu have two POMC genes, POMC and POMC?; (2) these two genes are the product of an ancient duplication and are present in all higher teleosts; (3) the expression domains of the primitive POMC gene have been partitioned, with POMC being expressed in the NLT and pituitary RPD and PI, and POMC? in the POA and pituitary PI; and (4) the peptide repertoire of POMC has also been subfunctionalised because a bona fide ?-endorphin is lacking in POMC? in tetraodontids and other lineages.
Phylogeny of POMC Duplicates in Teleosts
Molecular phylogenetic studies indicate that, among higher teleosts, one of the first groups to diverge from the lineage leading to the Acanthopterygii (which includes Tetraodon, Fugu, halibut, and barfin flounder) was the Ostariophysii (zebrafish, carps, catfishes, and their relatives), followed by the Protacanthopterygii (salmons and trouts) (fig. 2A). In our phylogenetic analyses, POMC? sequences from fishes as divergent as zebrafish, salmon, and Tetraodon are grouped together in a branch clearly separated from their paralogues, which also tend to group together. Thus, the duplication that originated the two POMC genes likely occurred in a common ancestor to all higher teleosts and, therefore, POMC and POMC? genes will probably be present in the genomes of all 22,000 higher teleost species. A recent inventory of the melanocortin system in Fugu identified only one POMC gene (Klovins et al. 2004), which corresponds to POMC. The low level of nucleotide identity of Fugu POMC? to other POMCs has probably prevented the identification of this gene in that study.
Recently, the hypothesis of an ancient, teleost-specific tetraploidization followed by rediploidization has gained much support (reviewed in Postlethwait et al. 2004; Volff 2005). In Fugu and zebrafish, phylogenetic analyses of duplicate paralogues suggest that an extensive gene duplication event, most likely a tetraploidization, happened around 320 MYA (Taylor et al. 2001; Christoffels et al. 2004; Vandepoele et al. 2004). In addition, analyses of Tetraodon and zebrafish genomes show that many syntenic segments (paralogons) occurring once in mammalian genomes are found duplicated in teleost genomes (Woods et al. 2000; Jaillon et al. 2004). Because POMC and POMC? seem to be common to all higher teleosts, it is likely that these genes are derived from the ancient teleost tetraploidization. This possibility is supported by the conservation of gene content around POMC paralogues in tetraodontids and zebrafish, which rules out the possibility of a single-gene or retrovirus-mediated duplication of POMC. In addition, the fact that the zebrafish POMC loci are found in different LGs, i.e., in chromosomes 17 and 20, is also compatible with a whole-genome duplication. Importantly, zebrafish LGs 17 and 20 have other genetic markers in common (snap25, bmp2, sox11), consistent with their being, to a great extent, sister chromosomes originated in the teleost genome duplication (fig. 4 in Woods et al. 2000; Taylor et al. 2003).
The phylogenetic timing of the genome duplication in the evolution of teleosts is still not known. Recently, Hoegg et al. (2004) analyzed duplicated genes from basal ray-finned fishes and basal teleosts (bony tongues and eels) and concluded that the genome duplication happened after the separation of the bichir, sturgeon, and gar, but before the separation of bony tongues (Osteoglossomorpha) and eels (Elopomorpha) from the higher teleost lineage (see fig. 2A), even though this hypothesis and the timing of 320 MYA for the genome duplication do not agree with the available actinopterygian fossil record (Benton 2000; see Discussion in Christoffels et al. 2004 and Hoegg et al. 2004). Our results are compatible with the phylogenetic dating proposed by Hoegg et al. (2004) because the POMC of eels (Anguilla) are grouped tightly within the branch of POMC and POMC? from higher teleosts (fig. 2B). In addition, there were duplications of the POMC gene which occurred independently from the duplication that originated POMC and POMC?. It is likely that many of these extra POMC genes are the result of more recent whole-genome duplications, which are known to have happened in sturgeons (Ludwig et al. 2001), salmonids (Palti et al. 2004), and some cyprinids like carp and goldfish (David et al. 2003).
Subfunctionalization of POMC Duplicates
The POMC gene plays several roles in vertebrate physiology. Among its functions are the regulation of serum glucocorticoid levels (mediated by pituitary ACTH), pain sensitivity (mediated by brain ?-endorphin), food intake (mediated by brain melanocortin peptides -MSH, ?-MSH, and -MSH), and the control of skin pigmentation (mediated by pituitary -MSH). This diversity of functions is, to a great extent, a consequence of the modular structure of the POMC prohormone, which can give rise to several bioactive peptides in different cell types. The POMC peptide repertoire has undergone different specializations in some vertebrate groups. For example, cartilaginous fishes have an extra melanocortin, -MSH, not present in other lineages (Dores and Lecaude 2005). In another case, the segment encoding -MSH, which is found in all vertebrate groups, has been lost from teleost POMC before the divergence of eels (Takahashi et al. 2001; Dores and Lecaude 2005). Both POMC and POMC? lack -MSH, suggesting that the degeneration of this melanocortin took place before the whole-genome duplication in the teleost lineage. Interestingly, mammalian -MSH binds preferentially to the MC3R, a receptor that is absent from the genome of tetraodontids (Klovins et al. 2004).
More importantly, we observed a high degree of degeneration of the ?-endorphin–coding sequence in teleost POMC? prohormones (fig 1B). ?-endorphin is characterized by the N-terminal sequence YGGFM (or close variants), which is necessary for opioid receptor binding (Reinscheid et al. 1995; Waldhoer, Bartlett, and Whistler 2004). The absence of this canonical motif suggests that tetraodontid POMC? peptides cannot stimulate opioid receptors. The missing opioid signal and the fact that these regions seem to be evolving neutrally (Ka/Ks around 1) suggest that the ?-endorphin regions of tetraodontid POMC? sequences are presently nonfunctional, in contrast to the -MSH/ACTH region, which has strong sequence conservation and low evolving rate (Ka/Ks << 1). In other teleosts, ?-endorphin peptides from POMC? show varying degrees of degeneracy: opioid signatures of POMC? are mutated in zebrafish and barfin flounder but are conserved in salmonids. POMC, on the other hand, has retained a conserved ?-endorphin peptide in all teleosts. Given that salmonids occupy an intermediate phylogenetic position among euteleosts (fig. 2A), it seems that the ?-endorphin region of POMC? started degenerating independently in several lineages, probably as a consequence of compensation provided by POMC and/or the partitioning of POMC expression domains between the two paralogues. The partitioning of the peptide repertoire between ancient POMC paralogues is one of the few examples of subfunctionalization of peptide functional domains, as suggested by Force et al. (1999). The unusual modular structure of the POMC prohormone, with its distinct functional peptides, makes it an ideal target for this sort of subfunctionalization. The evolution of POMC? ?-endorphin is reminiscent to that of -MSH, which also shows distinct levels of degeneration in different lineages of ray-finned fishes: sturgeons and gar have a recognizable but degenerating -MSH sequence, while teleosts lack this melanocortin completely (Dores and Lecaude 2005).
In addition to the subfunctionalization of peptide domains, the expression patterns of POMC and POMC? in the brain and pituitary of Tetraodon also indicate subfunctionalization. POMC mRNA is detected in the PI and rostral pars distalis in the pituitary and in the NLT in the hypothalamus. POMC? mRNA, on the other hand, is detected in the preoptic area (POA) in the brain and, weakly, in the PI of the pituitary. What about other teleosts? In the case of zebrafish, in situ hybridization using a POMC riboprobe detected expression in the pituitary and ventral hypothalamus but not in the POA (Herzog et al. 2003; Liu et al. 2003). However, ACTH- and ?-endorphin–like immunoreactivity is detected in the POA of a close species, the common carp (Metz et al. 2004). Thus, the partitioning of POMC expression domains might be a feature shared by other teleosts.
In mammals, POMC mRNA is mainly detected in the pituitary and in the arcuate nucleus of the hypothalamus. In teleosts, the NLT is considered to be the topological and functional equivalent to the arcuate nucleus. In contrast, the teleost POA is considered to be the functional and topological equivalent to the paraventricular and supraoptic nuclei of the hypothalamus, which do not express POMC in mammals. Is POMC expression in the POA a teleost novelty? It seems not. In the lungfish Protopterus annectens, the pituitary, NLT, and POA show -MSH immunoreactivity (Vallarino, Tranchand Bunel, and Vaudry 1992). Similarly, studies on the frog Xenopus laevis show POMC mRNA expression in the same regions (Tuinhof et al. 1998). These data suggest that POMC was expressed in the pituitary, NLT, and POA of the last common ancestor of lobe- and ray-finned fishes. After the genome duplication in teleosts, the expression domains of the original POMC gene were partitioned between the two newly arisen paralogues so that each of them performs an essential subfunction and, therefore, is kept under selective pressure. In addition, perhaps after POMC? expression was restricted to the POA, its ?-endorphin module started degenerating in different lineages. Thus, it is likely that ?-endorphin does not play an essential role in POA neurons of tetraodontids and, possibly, other teleosts as well.
The subfunctionalization of POMC paralogues in teleosts has some similarities to what happened in a lower vertebrate, the marine lamprey Petromyzon marinus (a jawless fish). This lamprey has two POMC genes with an unique organization regarding their peptide content: proopiomelanotropin (POM) gives rise to two different MSHs and ?-endorphin but no ACTH, whereas proopiocortin (POC) produces ACTH, MSH, and a different ?-endorphin (Takahashi et al. 2001; Dores and Lecaude 2005). POM is expressed in the pituitary PI, while POC is expressed in the RPD and weakly in the PPD, with none of the paralogues being expressed in the lamprey brain (Ficele et al. 1998). Thus, there has also been expression and peptide subfunctionalization between the lamprey paralogues, underlining the tendency of the modular POMC prohormone for subfunction partitioning. The duplication that originated these paralogues is very ancient, being probably specific to the lamprey lineage (Dores and Lecaude 2005). In any event, the POMC duplication in lamprey occurred independently of the duplication we observed in teleosts.
What could be the mechanism for the partitioning of POMC expression domains? As stated in the DDC model (Force et al. 1999), differential degenerative mutations in the regulatory regions of each duplicate gene might have led to the loss of expression of the POMC paralogue in the POA and loss of POMC? expression from the pituitary RPD and NLT, provided that POMC regulatory regions are modular in nature. Figure 6 summarizes the most likely scenario for POMC evolution in the Tetraodon lineage, starting with the loss of -MSH and the origin of two POMC paralogues in the whole-genome duplication until the differential loss of enhancers and ?-endorphin. POMC enhancers are shown as modular blocks that work independently of each other. We have previously shown that the expression of the mammalian POMC gene is controlled by two independent regulatory regions: one is located near the promoter and directs expression to the pituitary (Rubinstein et al. 1993) and the other, composed of two enhancers located around 10 kb upstream of the gene, directs expression to the arcuate nucleus (de Souza et al. 2005). Further studies on the regulatory regions of tetraodontid POMC paralogues, taking advantage of the reduced amount of intergenic DNA in this species, might shed light on the mechanisms of subfunction partitioning of duplicate genes, a crucial process in the evolution of vertebrates.
FIG. 6.— Hypothetical scheme showing the evolution of the POMC gene in teleosts. The genes are shown with the peptide-coding domains and putative enhancers for expression in the nucleus lateralis tuberis (NLT), preoptic area (PO), and pituitary rostral pars distalis (RPD) and pars intermedia (PI). After -MSH is lost from the POMC gene, a whole-genome duplication gives rise to two identical POMC copies, which subsequently undergo enhancer partitioning and, in some teleost lineages including Tetraodon, peptide partitioning. The PI enhancer maintains residual activity in Tetraodon POMC? (denoted by white triangle). Enhancer and peptide partitioning did not occur necessarily in the order presented in the scheme.
Supplementary Material
Supplementary Figure 1 is available at Molecular Biology and Evolution online (http://www.mbe.oxfordjournals.org/).
Acknowledgements
We thank L. F. Franchini for thoughtful comments and suggestions, M. Pandolfi for help with fish anatomy, and A. E. Lima for help with histology. This work was supported in part by a Fogarty International Research Collaborative Award TW01233 (M.J.L., M.R.), NIH grant DK68400 (M.J.L., M.R.), International Research Scholar Grant of the Howard Hughes Medical Institute (M.R.), Agencia Nacional de Promoción Científica y Tecnológica (M.R.), and Universidad de Buenos Aires (M.R.). F.S.J.S. and V.F.B. are recipients of fellowships from the Consejo Nacional de Investigaciones Cientificas y Tecnicas (Argentina).
References
Altschmied, J., J. Delfgaauw, B. Wilde, J. Duschl, L. Bouneau, J. N. Volff, and M. Schartl. 2002. Subfunctionalization of duplicate mitf genes associated with differential degeneration of alternative exons in fish. Genetics 161:259–267.
Altschul, S. F., W. Gish, W. Miller, E. W. Myers, and D. J. Lipman. 1990. Basic local alignment search tool. J. Mol. Biol. 215:403–410.
Amano, M., A. Takahashi, T. Yamanome, Y. Oka, N. Amyia, H. Kawauchi, and K. Yamamori. 2005. Immunocytochemical localization and ontogenic development of -melanocyte stimulating hormone (-MSH) in the brain of a pleuronectiform fish, barfin flounder. Cell Tissue Res. 320:127–134.
Aparicio, S., J. Chapman, E. Stupka et al. (38 co-authors). 2002. Whole-genome shotgun assembly and analysis of the genome of Fugu rubripes. Science 297:1301–1310.
Benton, M. J. 2000. Vertebrate palaeontogy. 2nd edition. Blackwell Science, Oxford.
Christoffels, A., E. G. Koh, J. M. Chia, S. Brenner, S. Aparicio, and B. Venkatesh. 2004. Fugu genome analysis provides evidence for a whole-genome duplication early during the evolution of ray-finned fishes. Mol. Biol. Evol. 21:1146–1151.
David, L., S. Blum, M. W. Feldman, U. Lavi, and J. Hillel. 2003. Recent duplication of the common carp (Cyprinus carpio L.) genome as revealed by analyses of microsatellite loci. Mol. Biol. Evol. 20:1425–1434.
de Souza, F. S., A. M. Santangelo, V. Bumaschny, M. E. Avale, J. L. Smart, M. J. Low, and M. Rubinstein. 2005. Identification of neuronal enhancers of the proopiomelanocortin gene by transgenic mouse analysis and phylogenetic footprinting. Mol. Cell. Biol. 25:3076–3086.
Dores, R. M., and S. Lecaude. 2005. Trends in the evolution of the proopiomelanocortin gene. Gen. Comp. Endocrinol. 142:81–93.
Dores, R. M., C. Sollars, P. Danielson, J. Lee, J. Alrubaian, and J. M. Joss. 1999. Cloning of a proopiomelanocortin cDNA from the pituitary of the Australian lungfish, Neoceratodus forsteri: analyzing trends in the organization of this prohormone precursor. Gen. Comp. Endocrinol. 116:433–444.
Felsenstein, J. 2004. PHYLIP (phylogeny inference package). Version 3.6. Distributed by the author, Department of Genome Sciences, University of Washington, Seattle.
Ficele, G., J. A. Heinig, H. Kawauchi, J. H. Youson, F. W. Keeley, and G. M. Wright. 1998. Spatial and temporal distribution of proopiomelanotropin and proopiocortin mRNA during the life cycle of the sea lamprey: a qualitative and quantitative in situ hybridization study. Gen. Comp. Endocrinol. 110:212–225.
Force, A., M. Lynch, F. B. Pickett, A. Amores, Y. L. Yan, and J. H. Postlethwait. 1999. Preservation of duplicate genes by complementary, degenerative mutations. Genetics 151:1531–1545.
González-Nu?ez, V., R. González-Sarmiento, and R. E. Rodríguez. 2003. Identification of two proopiomelanocortin genes in zebrafish (Danio rerio). Brain Res. Mol. Brain Res. 120:1–8.
Hadley, M. E., and C. Haskell-Luevano. 1999. The proopiomelanocortin system. Ann. N. Y. Acad. Sci. 885:1–21.
Hansen, I. A., T. T. To, S. Wortmann, T. Burmester, C. Winkler, S. R. Meyer, C. Neuner, M. Fassnacht, and B. Allolio. 2003. The pro-opiomelanocortin gene of the zebrafish (Danio rerio). Biochem. Biophys. Res. Commun. 303:1121–1128.
Herzog, W., X. Zeng, Z. Lele, C. Sonntag, J. W. Ting, C. Y. Chang, and M. Hammerschmidt. 2003. Adenohypophysis formation in the zebrafish and its dependence on sonic hedgehog. Dev. Biol. 254:36–49.
Hoegg, S., H. Brinkmann, J. S. Taylor, and A. Meyer. 2004. Phylogenetic timing of the fish-specific genome duplication correlates with the diversification of teleost fish. J. Mol. Evol. 59:190–203.
Huang, F. D., H. J. Matthies, S. D. Speese, M. A. Smith, and K. Broadie. 2004. Rolling blackout, a newly identified PIP2-DAG pathway lipase required for Drosophila phototransduction. Nat. Neurosci. 7:1070–1078.
Hurst, L. D. 2002. The Ka/Ks ratio: diagnosing the form of sequence evolution. Trends Genet. 18:486.
Inoue, J. G., M. Miya, K. Tsukamoto, and M. Nishida. 2003. Basal actinopterygian relationships: a mitogenomic perspective on the phylogeny of the "ancient fish". Mol. Phylogenet. Evol. 26:110–120.
———. 2004. Mitogenomic evidence for the monophyly of elopomorph fishes (Teleostei) and the evolutionary origin of the leptocephalus larva. Mol. Phylogenet. Evol. 32:274–286.
Jaillon, O., J. M. Aury, F. Brunet et al. (47 co-authors). 2004. Genome duplication in the teleost fish Tetraodon nigroviridis reveals the early vertebrate proto-karyotype. Nature 431:946–957.
Jones, D. T., W. R. Taylor, and J. M. Thornton. 1992. The rapid generation of mutation data matrices from protein sequences. CABIOS 8:275–282.
Kikugawa, K., K. Katoh, S. Kuraku, H. Sakurai, O. Ishida, N. Iwabe, and T. Miyata. 2004. Basal jawed vertebrate phylogeny inferred from multiple nuclear DNA-coded genes. BMC Biol. 2:3.
Klovins, J., T. Haitina, D. Fridmanis, Z. Kilianova, I. Kapa, R. Fredriksson, N. Gallo-Payet, and H. B. Schi?th. 2004. The melanocortin system in Fugu: determination of POMC/AGRP/MCR gene repertoire and synteny, as well as pharmacology and anatomical distribution of the MCRs. Mol. Biol. Evol. 21:563–579.
Kumar, S., and S. B. Hedges. 1998. A molecular timescale for vertebrate evolution. Nature. 392:917–920.
Liu, N. A., H. Huang, Z. Yang, W. Herzog, M. Hammerschmidt, S. Lin, and S. Melmed. 2003. Pituitary corticotroph ontogeny and regulation in transgenic zebrafish. Mol. Endocrinol. 17:959–966.
Liu, R. Z., Q. Sun, C. Thisse, B. Thisse, J. M. Wright, and E. M. Denovan-Wright. 2005. The cellular retinol-binding protein genes are duplicated and differentially transcribed in the developing and adult zebrafish (Danio rerio). Mol. Biol. Evol. 22:469–477.
Low, M. J. 2003. Somatostatin. Pp. 379–388 in H. L. Henry and A. W. Norman, eds. Encyclopedia of hormones. Elsevier Science, San Diego, Calif.
Ludwig, A., N. M. Belfiore, C. Pitra, V. Svirsky, and I. Jenneckens. 2001. Genome duplication events and functional reduction of ploidy levels in sturgeon (Acipenser, Huso and Scaphirhynchus). Genetics 158:1203–1215.
McClintock, J. M., M. A. Kheirbek, and V. E. Prince. 2002. Knockdown of duplicated zebrafish hoxb1 genes reveals distinct roles in hindbrain patterning and a novel mechanism of duplicate gene retention. Development 129:2339–2354.
Metz, J. R., M. O. Huising, J. Meek, A. J. Taverne-Thiele, S. E. Wendelaar-Bonga, and G. Flik. 2004. Localization, expression and control of adrenocorticotropic hormone in the nucleus preopticus and pituitary gland of the common carp (Cyprinus carpio L.). J. Endocrinol. 182:23–31.
Palti, Y., S. A. Gahr, J. D. Hansen, and C. E. Rexroad III. 2004. Characterization of a new BAC library for rainbow trout: evidence for multi-locus duplication. Anim. Genet. 35:130–133.
Pandolfi, M., M. M. Canepa, M. A. Ravaglia, M. C. Maggese, D. A. Paz, and P. G. Vissio. 2003. Melanin-concentrating hormone system in the brain and skin of the cichlid fish Cichlasoma dimerus: anatomical localization, ontogeny and distribution in comparison to -melanocyte-stimulating hormone-expressing cells. Cell Tissue Res. 311:61–69.
Postlethwait, J., A. Amores, W. Cresko, A. Singer, and Y. L. Yan. 2004. Subfunction partitioning, the teleost radiation and the annotation of the human genome. Trends Genet. 20:481–490.
Reinscheid, R. K., H. P. Nothacker, A. Bourson, A. Ardati, R. A. Henningsen, J. R. Bunzow, D. K. Grandy, H. Langen, F. J. Monsma Jr., and O. Civelli. 1995. Orphanin FQ: a neuropeptide that activates an opioidlike G protein-coupled receptor. Science 270:792–794.
Rozas, J., and R. Rozas. 1999. DnaSP version 3: an integrated program for molecular population genetics and molecular evolution analysis. Bioinformatics 15:174–175.
Rubinstein, M., M. Mortrud, B. Liu, and M. J. Low. 1993. Rat and mouse proopiomelanocortin gene sequences target tissue-specific expression to the pituitary gland but not to the hypothalamus of transgenic mice. Neuroendocrinology 58:373–380.
Salbert, G., I. Chauveau, G. Bonnec, Y. Valotaire, and P. Jego. 1992. One of the two trout proopiomelanocortin messenger RNAs potentially encodes new peptides. Mol. Endocrinol. 6:1605–1613.
Takahashi, A., M. Amano, T. Itoh, A. Yasuda, T. Yamanome, Y. Amemyia, T. Sasaki, M. Sakai, K. Yamamori, and H. Kawauchi. 2005. Nucleotide sequence and expression of three subtypes of proopiomelanocortion RNA in barfin flounder. Gen. Comp. Endocrinol. 141:291–303.
Takahashi, A., Y. Amemiya, M. Nozaki, S. A. Sower, and H. Kawauchi. 2001. Evolutionary significance of proopiomelanocortin in agnatha and chondrichthyes. Comp. Biochem. Physiol. B Biochem. Mol. Biol. 129:283–289.
Taylor, J. S., I. Braasch, T. Frickey, A. Meyer, and Y. Van de Peer. 2003. Genome duplication, a trait shared by 22,000 species of ray-finned fish. Genome Res. 13:382–390.
Taylor, J. S., Y. Van de Peer, I. Braasch, and A. Meyer. 2001. Comparative genomics provides evidence for an ancient genome duplication event in fish. Philos. Trans. R. Soc. Lond. B Biol. Sci. 356:1661–1679.
Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22:4673–4680.
Tuinhof, R., R. Ubink, S. Tanaka, C. Atzori, F. J. van Strien, and E. W. Roubos. 1998. Distribution of pro-opiomelanocortin and its peptide end products in the brain and hypophysis of the aquatic toad, Xenopus laevis. Cell Tissue Res. 292:251–265.
Vallarino, M., D. Tranchand Bunel, and H. Vaudry. 1992. -melanocyte-stimulating hormone (-MSH) in the brain of the African lungfish, Protopterus annectens: immunohistochemical localization and biochemical characterization. J. Comp. Neurol. 322:266–274.
Vandepoele, K., W. De Vos, J. S. Taylor, A. Meyer, and Y. Van de Peer. 2004. Major events in the genome evolution of vertebrates: paranome age and size differ considerably between ray-finned fishes and land vertebrates. Proc. Natl. Acad. Sci. USA 101:1638–1643.
Volff, J. N. 2005. Genome evolution and biodiversity in teleost fish. Heredity 94:280–294.
Volkoff, H., L. F. Canosa, S. Unniappan, J. M. Cerda-Reverter, N. J. Bernier, S. P. Kelly, and R. E. Peter. 2005. Neuropeptides and the control of food intake in fish. Gen. Comp. Endocrinol. 142:3–19.
Waldhoer, M., S. E. Bartlett, and J. L. Whistler. 2004. Opioid receptors. Annu. Rev. Biochem. 73:953–990.
Woods, I. G., P. D. Kelly, F. Chu, P. Ngo-Hazelett, Y. L. Yan, H. Huang, J. H. Postlethwait, and W. S. Talbot. 2000. A comparative map of the zebrafish genome. Genome Res. 10:1903–1910.(Flávio S. J. de Souza*,1,)