- Top of page
- Materials and methods
We report the cloning, genome mapping, functional expression, pharmacology and anatomical distribution of three melanocortin (MC) receptors from zebrafish (z). Phylogenetic analysis showed with high bootstrap support that these genes represent one MC4 receptor and two MC5 receptors. Chromosomal mapping showed conserved synteny between regions containing zMC4 and human (h) MC4 receptors, whereas the two zMC5 receptor genes map on chromosome segments in which the zebrafish has several genes with two orthologues of a single mammalian gene. It is likely that the two copies of zMC5 receptors arose through a separate duplication in the teleost lineage. The zMC4, zMC5a, and zMC5b receptors share 70–71% overall amino acid identity with the respective human orthologues and over 90% in three TM regions believed to be most important for ligand binding. All three zebrafish receptors also show pharmacological properties remarkably similar to their human orthologues, with similar affinities and the same potency order, when expressed and characterized in radioligand binding assay for the natural melanocyte stimulating hormone (MSH) peptides α-, β-, and γ-MSH. Stimulation of transfected mammalian cells with α-MSH caused a dose-dependent increase in intracellular cAMP levels for all three zebrafish receptors. All three genes were expressed in the brain, eye, ovaries and gastrointestinal tract, whereas the zMC5b receptor was also found in the heart, as determined by RT-PCR. Our studies, which represent the first characterization of MC receptors in a nonamniote species, indicate that the MC receptor subtypes arose very early in vertebrate evolution. Important pharmacological and functional properties, as well as gene structure and syntenic relationships have been highly conserved over a period of more than 400 million years implying that these receptors participate in vital physiological functions.
The teleost zebrafish (Danio rerio) is an important experimental animal in genomic and developmental research. It has a compact genome with a similar number of genes but is only one-third in size as compared with the human genome, sharing large blocks of conserved synteny (Postlethwait et al. 2000). While the complete sequencing of the zebrafish genome is in progress, currently less than 30% of the genes in the zebrafish genome have been cloned, and these include about 40 genes for G-protein coupled receptors (GPCR). The majority of these belong to the groups of olfactory, chemokine and opsin receptors. The other main families of GPCRs cloned in zebrafish include neuropeptide Y (NPY), opioid, melatonin, endothelin and parathyroid hormone (PTH) receptors. Only a few families of peptide-binding GPCRs from ‘lower’ vertebrates, including fishes, have been expressed and pharmacologically characterized. Although some of these receptors show a high level of similarity to their mammalian counterparts, which relate to sequence identity, ligand selectivity and affinity (Darlison et al. 1997; Palyha et al. 2000), other gene families seem to have undergone major transformation in the mammalian lineage (Rubin et al. 1997; Cerda-Reverter and Larhammar 2000). It is conceivable that the physiological systems that involve GPCRs may be largely conserved between vertebrate classes even though the molecular components may differ significantly.
Our understanding of the mechanisms for central regulation of food intake and metabolism in mammals is growing rapidly. Several neuropeptides are involved that bind to GPCRs in the hypothalamic regions (Kalra et al. 1999). Very little is known however, about the molecular mechanisms of central regulation of energy balance in non-mammalian species, despite the fact that the growth rate of several species, including fishes, has important global ecological and economical impact.
The family of melanocortin (MC) receptors consists of five subtypes cloned from human, rat and mouse and also one non-mammalian species, namely chicken (for review see Schiöth 2001). The primary structures of the MC receptors show no major similarity to other peptide-binding GPCRs and evidence indicates that the ligand-binding pocket of these receptor mainly involves transmembrane regions 2, 3 and 7 but not 4 and 5, like many other rhodopsin-like GPCRs (Yang et al. 2000; Schiöth et al. 1998c; Haskell-Luevano et al. 2001). The mammalian MC receptors have been extensively characterized and lately received great attention due to their important role in the regulation of the energy balance, involving the centrally expressed MC3 and MC4 receptors (Huszar et al. 1997; Chen et al. 2000). MC4 receptor antagonists are among the most effective agents available to induce both acute and long-term food intake in mammals (Kask et al. 1998a,b; Skuladottir et al. 1999). Agonists for the MC4 receptor are also highly effective in reducing feeding (Fan et al. 1997) and affecting metabolic rate (Jonsson et al. 2001). These findings have attracted great interest from researchers studying central regulation of body weight homeostasis, including the pharmaceutical industry. The MC5 receptor subtype is primarily expressed in a variety of peripheral tissues but it has also been detected in the brains of mammals (Barret et al. 1994; Griffon et al. 1994; Labbéet al. 1994; Chhajlani 1996; Mountjoy and Wong 1997). This receptor is believed to have a role in exocrine gland functions in mice (Chen et al. 1997). Both the MC4 and MC5 receptor genes have been linked with human obesity-related phenotypes (Chagnon et al. 1997; Hinney et al. 1999). The two other MC receptor subtypes are found in the periphery. The MC1 receptor has a role in pigmentation and it also mediates the broad anti-inflammatory actions of the melanocortins. The MC2 receptor is exclusively found in the adrenal gland where it mediates the effect of adrenocorticotropic hormone (ACTH) on steroid production (Schiöth 2001).
The MC receptor system is unique in that it has both natural agonists and antagonists. The agonists are the pro-opiomelanocortin (POMC) cleavage products α-melanocyte stimulating hormone (MSH), β-MSH, γ-MSH and ACTH, whereas separate precursors generate natural antagonists named agouti and agouti-related peptide (Agrp). POMC has been extensively used as a model for studies of the evolution of neuropeptides. There are several examples of two copies of the POMC gene in fishes (Dores et al. 1990). It seems to have arisen in early chordate evolution and has undergone a number of lineage-specific modifications but seems always to include both α- and β-MSH. The sequence of α-MSH has been cloned from at least seven species of fish, showing high evolutionary conservation of this 13 amino acid peptide (Danielson and Dores 1999). The principal source of the mammalian melanocortins is the pituitary, but the POMC gene is also expressed in a variety of other brain regions and also in a number of peripheral tissues, particularly in the skin. Besides the physiological functions mentioned above, the melanocortins have been implicated in a broad array of other processes, as they also effect sexual functions in both sexes, memory, pyretic control, pain perception, blood pressure, nerve growth and regeneration, and several events surrounding parturition (Eberle 1988; Schiöth 2001).
In this paper we report the cloning and characterization of the first MC receptors from a ‘lower’ (nonamniote) vertebrate. We describe three MC receptors in zebrafish with respect to structure, chromosomal localization, pharmacological profile and anatomical distribution and discuss their evolution through gene duplications.
- Top of page
- Materials and methods
Degenerate primers corresponding to highly conserved regions in the cloned MC receptors were made and used for PCR on genomic zebrafish DNA. About 40 PCR products of expected size, approximately 400–600 bp, were cloned and sequenced. Three clones of 600 bp showed high sequence identity to the MC4 and MC5 receptors. These sequences were used as probes for high stringency screening of a gridded genomic zebrafish cosmid library. Three clones that hybridized strongly were identified and subsequently sequenced with specific primers. Each of the three clones had a unique MC receptor-like sequence. One of the sequences had highest identity to the mammalian MC4 receptors. This clone was designated as the zMC4 receptor. The other two clones had highest sequence identity to each other. Among the mammalian and chicken receptors they had highest identity to the MC5 receptors (70–77%). Hence, they were designated as the zMC5a and zMC5b receptors. Alignment of the protein sequences of the three zebrafish receptors with the previously cloned human MC1, MC2, MC3, MC4 and MC5 receptors and the chicken and mouse MC4 and MC5 receptors is shown in Fig. 1. The percentage identities of the complete sequences and the TM regions are shown in Table 1.
Table 1. Sequence comparison of melanocortin receptor subtypes in different species
Phylogenetic analyses of the full-length amino acid sequences of the three zebrafish receptors, together with the amniote (human, mouse and chicken) MC receptors, using maximum parsimony and neighbour joining methods are shown in Fig. 2. The topology of the trees were the same with or without an out group (data not shown). The parsimony tree (right) had length of 1421, consistency index 0.78, homoplasy index 0.22. The topologies of the trees are identical except for two minor differences. The neighbour joining tree places the MC2 receptors as the sister group of the MC3/4/5 and MC1 receptors as the earliest branch, whereas the parsimony tree shows both the MC1 and MC2 receptors at the base of the tree. Similarly, the neighbour joining tree has the MC3 receptors as the sister group of the MC4 and MC5 receptors, whereas the parsimony tree places the MC3, MC4, and MC5 receptor groups equidistantly from each other. For each of the five receptor subtypes, human and mouse are more closely related than either is to chicken, as expected. The zMC4 receptor is basal to the three amniote MC4 receptor sequences corresponding to the relationships of the species. The zMC5 receptor sequences are most similar to each other and branch off basal to the three amniote MC5 receptor sequences, suggesting that they arose by gene duplication after the divergence of amniote and teleost lineages. Neighbour joining and maximum parsimony analyses of only the TM regions give trees with the same topology as for the full-length receptors with respect to the positioning of the zebrafish receptors with one exception (data not shown), namely that the TM maximum parsimony tree interchanges the branches for the MC2 and the MC1 receptor groups. These phylogenetic trees can be obtained from the correspondence author upon request. Molecular clock calculations indicate that the duplication that generated the two zMC5 receptors took place in the teleost lineage later than the divergence of mammals and birds which is usually dated to 300 million years ago.
The zMC4, zMC5a, and MC5b receptor genes mapped to chromosomes [linkage group (LG)], LG2, LG19, LG16, respectively. Figure 3 shows a map of locations of the zMC4, zMC5a, zMC5b receptor genes, together with a map of HSA18 with the location of the hMC4, hMC5 and hMC2 receptor genes. The zMC4 receptor gene is found close to another locus, MYOM1 found on HSA18p11.32. A third HSA18 gene, NPC1, lies distantly on LG2. The zMC5a and zMC5b receptor genes show no known conserved synteny to HSA18. LG16 and LG19 have however, closely related genes that correspond to the same mammalian orthologue, suggesting that they are duplicated chromosome segments (Amores et al. 1998; Woods et al. 2000).
Figure 3. Map showing chromosomal positions of the MC receptors and conserved syntenies on zebrafish linkage groups LG2, LG16, LG19 and human chromosome HSA18. The abbreviations for the genes are listed below with the full names and the zebrafish and human NCBI accession numbers (or Locus ID): NPC (Niemann–Pick C1, Protein precursor), AI722583, O15118; MYOM1 (myomesin 1), AI658181, Q62234; atp1a3 (Na+/K+ ATPase alpha subunit isoform), AF308599, AY008374, AAA53499; rxrd (retinoid X receptor delta), U29941, NP_068811; hoxa11, Y14528, NP_005514; apoe (apolipoprotein E precursor protein), Y13652, AAB59397; gdf6 (growth and differentiation factor 6), AAB34226 and Locus ID 9571.
Download figure to PowerPoint
For functional expression, the coding regions of the three zebrafish receptors were transferred to an expression vector containing the cytomegalovirus (CMV) promotor. These clones were readily expressed in HEK-EBNA cells and the receptors were tested in radioligand binding assay to intact cells using methods similar to those employed earlier for the human MC receptor subtypes (Schiöth et al. 1995). The results of saturation analysis (Kd values) were obtained by incubating varying concentrations of the high-affinity ligand [125I]NDP-MSH in the absence or presence of 2 µmaup unlabelled NDP-MSH. We also performed competition binding analysis (Ki values) using a constant concentration of the labelled ligand and varying concentrations of NDP-MSH, α-MSH, β-MSH, γ-MSH and HS014. The results show that [125I]NDP-MSH bound to a single saturable site for the zebrafish receptors in a fashion similar to that of the human MC4 and MC5 receptors, which were used for comparison. Figure 4 shows saturation and competition curves for the zebrafish receptors. Table 2 shows the Kd and the Ki values obtained from saturation and competition analysis, respectively. The hMC4 and hMC5 receptor proteins showed binding values similar to those we reported earlier (Schiöth et al. 1995, 1996). The zMC4 receptor had binding values indistinguishable from the hMC4 receptor for the high potency synthetic ligand NDP-MSH and also for the endogeneous peptides α-, β-, and γ1-MSH. The potency order of the peptides was the same for the two MC4 receptors. The only compound that had a significantly different affinity for the zMC4 receptor as compared with the human receptor was the peptide HS014, which had 88-fold lower affinity for the zMC4 receptor. The pattern is the same for the MC5 receptors. Both the zebrafish receptors had similar affinity as the human receptor for all the peptides except for HS014. The characteristic potency order found for the human and mouse MC5 receptor (NDP-MSH > α-MSH > β-MSH > γ-MSH) was conserved for the zebrafish receptors (Schiöth et al. 1998a). HS014 had however, about 7–8-fold lower affinity for both the zMC5 receptors as compared with the human orthologue.
Figure 4. Saturation and competition curves for the zMC4, zMC5a and zMC5b receptors expressed in intact transfected cells. Saturation curves (left) were obtained with [125I](Nle4,maup-Phe7)α-MSH and the figure shows total binding (▮) and binding in the presence of 3 µmaup cold (Nle4,D-Phe7) α-MSH (▴) for each receptor. Lines represent the computer-modelled best fit of the data assuming that ligands were bound to one-site. Competition curves (right) for (Nle4,maup-Phe7)α-MSH (▴), α-MSH (▮), and β-MSH (□) were obtained by using a fixed concentration of 2 nmaup[125I](Nle4,maup-Phe7) α-MSH and varying concentrations of the non-labelled competing peptide.
Download figure to PowerPoint
Table 2. Ki and Kd values (mean ± SEM) obtained from competition and saturation curves, respectively, for melanocortin peptides analogues on zMC4, hMC4, zMC5 and hMC5 receptor transfected EBNA cells
|Ligand||zMC4 (nmol/L)||hMC4 (nmol/L)||zMC5a (nmol/L)||zMC5b (nmol/L)||hMC5 (nmol/L)|
|[125I]NDP-MSH (Kd)||2.39 ± 0.96||2.35 ± 1.18||2.72 ± 0.82||2.49 ± 0.99||2.64 ± 1.05|
|NDP-MSH (Ki)||3.35 ± 0.31||3.57 ± 0.30||19.2 ± 1.5||5.56 ± 0.54||3.71 ± 0.72|
|α-MSH (Ki)||243 ± 27||289 ± 29||2580 ± 570||2730 ± 230||2950 ± 510|
|β-MSH (Ki)||163 ± 14||126 ± 15||3750 ± 410||3280 ± 480||7660 ± 640|
|γ1-MSH (Ki)||2200 ± 550||3690 ± 260||4150 ± 340||3750 ± 830||5180 ± 830|
|HS014 (Ki)||493 ± 40||5.60 ± 0.22||4220 ± 760||5100 ± 410||627 ± 570|
We also tested zebrafish receptor transfected cells in a cAMP assay with the intent to investigate if the receptors were able to couple to G-proteins after α-MSH stimulation. We used the same vector and cells as for the binding experiments. The results are shown in Fig. 5. α-MSH induced a dose dependent accumulation of intracellular cAMP through all three receptors. The three receptors reached similar levels of maximum response, which is in line with what we have observed for the human MC4 and MC5 receptors in response to α-MSH (Schiöth et al. 1998b).
Figure 5. Generation of cAMP in response to α-MSH for zMC4 (○), zMC5a (▮) and zMC5b (▴) receptors in intact transfected cells. Each point represents the average ± SEM. Untransfected HEK293-EBNA cells showed no cAMP response to α-MSH.
Download figure to PowerPoint
Tissue distribution was determined by RT-PCR. We isolated total RNA from heart, muscle, eye, the gastrointestinal tract (GI), total brain and ovary. The length of a zebrafish is only about 3 cm, and the brain only weighs a few milligrams. The results of the RT-PCR for each of the receptor genes are shown in Fig. 6. Each probe bound specifically to the designated clone and there were no signs of cross hybridization under our experimental conditions (see Fig. 6d). The zMC4 receptor was expressed in four tissues. There were strong signals in the GI and the ovary whereas there were weaker signals in the brain and the eye. No signals were seen in the heart or the muscle. The zMC5a transcript was also found in the same four tissues with strongest signals in the ovary, brain and GI tract, whereas the signal for the eye was weaker. The zMC5b gene was expressed in five tissues. Strong signals were found in the ovary, GI tract, brain and the eye, whereas a weaker signal was found in the heart. However, it should be noted that the PCR assay was not designed for quantification. The experiment was performed three times and there were no qualitative differences between the runs. cDNA from muscle did not show any positive signal and we confirmed the integrity of this sample by performing RT-PCR with a zebrafish bradykinin receptor probe previously found to show positive signal in this tissue (data not shown, Dunér et al. personal communication).
Figure 6. Expression of zMC4 (a), zMC5a (b) and zMC5b (c) receptor mRNA as determined by RT-PCR on zebrafish tissues. The tissues and the controls are denoted at the top of each figure. The left part shows the agarose gel stained with ethidium bromide. The right part shows a 4-h exposure on an X-ray film after hybridization with the corresponding probe. The predicted lengths of the PCR products were 355 bp (zMC4), 355 bp (zMC5a) and 325 bp (zMC5a). The PCR reactions were performed three times with qualitatively the same result. Cross-hydrization experiment for each of the probes against positive control PCR products is shown in (d).
Download figure to PowerPoint
- Top of page
- Materials and methods
The zMC4, zMC5a, and zMC5b receptors share 69.8%, 71.3% and 69.6% overall amino acid identity with the respective human orthologues (see Table 1). Prior to this study we only had MC receptors from amniote species whose ancestors diverged about 450 million years ago from the lineage leading to zebrafish. The ancient time of divergence makes the amino acid identity percentages alone not entirely conclusive in determining which of the mammalian subtype the new receptors might belong to. Moreover, the mammalian MC4 and MC5 receptors (and MC3 receptors) show high degree of amino acid identity to each other (55–65%) and so do the zebrafish receptors we cloned. The phylogenetic analysis however, indicates clearly that the new receptors are one orthologue of the MC4 and two of the MC5 receptor. Both the maximum parsimony and neighbour joining analyses of the full-length receptor proteins show the same topology with regard to the zebrafish receptors with very high bootstrap values (see Fig. 2). The chromosomal localization also shows that there is some conserved synteny between the zMC4 receptor gene and HSA18, containing the human MC4 receptor gene (see Fig. 3). Moreover, the two zMC5 receptor genes lie in linkage groups in vicinity of several genes where the zebrafish has duplicate or orthologues of several single-copy mammalian genes, suggesting that the entire chromosome was duplicated. The zebrafish, like most teleost fishes, probably underwent a linage specific genome duplication approximately 250 million years ago (Amores et al. 1998; Postlethwait et al. 1998; Meyer and Schartl 1999; Taylor et al. 2001) even though it also has been suggested that the teleost fishes underwent frequent independent duplications (Aparicio 2000; Robinson-Rechavi et al. 2001). It is likely that the two copies of the MC5 receptors that we have found, may have resulted from this tetraploidization. Comparison of the two zMC5 receptor sequences indicates that the duplication is not a very recent event as perhaps could be concluded on the basis of the similar pharmacology of the receptors. Our molecular clock calculations date it to the same time period as has been proposed for the teleost tetraploidization, providing further support for the hypotheses that the teleost linage indeed underwent a major or whole genome duplication at the proposed time interval.
The MC protein sequences reported here show a relatively high degree of conservation with mammals as compared with other GPCRs or membrane-bound proteins from zebrafish. The predicted receptors not only show high overall identity to their mammalian orthologues, they also display several of the structural characteristics typical for the mammalian MC receptors, for example, a conserved Cys in the first extra cellular loop (EL), lack of Pro in TM5 and short EL and intracellular loops (IL) (see Fig. 1). The high degree of identity is, however, unevenly distributed over the different regions of the receptors. There are two regions in the zebrafish receptors that show low or no sequence identity, namely the N-terminal and the C-terminal regions. This is in line with results that show that the N-termini is not involved in the binding of the four MSH binding MC receptors (Schiöth et al. 1997). Experiments involving chimeric proteins, cassette mutations, and molecular modelling have suggested which regions within the MC receptors make up the ligand pocket (Schiöth et al. 1998c; Yang et al. 2000; Oosterom et al. 2001). The zebrafish receptors show a particularly high degree of conservation in these regions. It is remarkable that there is only one amino acid difference in the entire TM7 (estimated to be 25 amino acids long) between the zMC5a and the hMC5 receptors. Moreover there is more that 90% overall amino acid identity between the zebrafish and the human MC5 receptors in TM2, TM3 and TM7 (see Fig. 1). It is a general conception that peptide ligands have a higher degree of conservation than the corresponding, but much larger receptors. It is interesting to note that fairly large regions believed to participate in the ligand binding unit of the MC receptors have the same degree of evolutionary conservation as found for α-MSH.
The MSH peptides bind to the mammalian MC receptor subtypes without any major discrimination except that the MC2 receptor only binds ACTH. It is still possible to discriminate between the four other subtypes by careful radioligand binding analysis. It is interesting that the zMC4 receptor shows both of the unique characteristics of the human MC4 receptor. These are the relatively low affinity for γ-MSH and a slightly higher or similar affinity for β-MSH as compared with α-MSH. The physiological roles of γ- and β-MSH still remain obscure despite their high degree of evolutionary conservation. It is intriguing that β-MSH also has a higher affinity to the zMC4 receptor, in line with our previous results for the human receptor (Schiöth et al. 1996), now confirmed by another group (Harrold et al. 2001). Our new results support that β-MSH may have a specific and also an evolutionary conserved role for this receptor subtype. The two new zMC5 receptors had indistinguishable binding properties between them selves and to the hMC5 receptor, providing further support for the conclusion that these are indeed two copies of an ancestral MC5 receptor gene. The mouse MC5 receptor has been reported to have higher affinity for MSH peptides as compared with the hMC5 receptor. It has been speculated that this could be due to the lack of Cys in EL3 of the hMC5 receptor that is conserved in all the other MC receptors (Frändberg et al. 1997; Schiöth et al. 1998a). It is worth mentioning that the zMC5 receptor has this Cys in EL3 but still shares the human MC5 receptor pharmacology. HS014 is a synthetic cyclic MC4 receptor selective peptide. This substance has been widely used in physiological studies. It is interesting that this substance is the only one that has lower affinity for all three new zebrafish receptors as compared with the human ones. HS014 was developed by selection of ligands for the human MC receptor subtypes. It seems thus that even though the ability of the receptors to bind the natural peptides is highly conserved, the 3D binding cavity may not be as well conserved to fit this synthetic ligand. The data show that this peptide may not be as suitable for physiological studies in zebrafish as it has been in studies in rodents. We have also shown that the three fish MC receptors can functionally couple to the Gs linked signalling pathway in response to α-MSH. This is in line with all previous results found for expression of the mammalian MC receptors. Expression of the chicken MC receptors has not yet been reported, but our unpublished preliminary findings indicate that they couple in a similar manner.
Expression of mammalian MC4 receptors has only been found in brain tissue. The zMC4 receptor was also found in the brain but rather surprisingly also in the eye, GI tract and ovaries. Melanocytes from human choroid and iris respond to MSH (Smith-Thomas et al. 2001) through what is believed to be the MC1 receptor, but a search for other MC receptor subtypes in the different cell types in eyes that respond to MSH has, to the best of our knowledge, not yet been done. The GI tract has not got much attention within the MSH or MC receptor research community, despite the fact that the MC3 receptor subtype has been found in the GI tract of humans but not in other species like rat or mouse. It is believed that MSH peptides have several effects on the central regulation of reproductive processes, at least partially involving the MC4 receptor (Schiöth and Watanobe 2002). α-MSH was found to induce steroidogenesis of ovaries in rats (Durando and Celis 1998). POMC is also found to be expressed in the ovary of an African lungfish (Masini et al. 1997) and MSH peptides have been found in the ovaries of two teleost species, the sea bream and sea bass (Mosconi et al. 1994). The MC4 receptor is expressed in a wide variety of peripheral tissues in the chicken, including heart, adrenals, ovaries, testes, spleen and adipose tissues, and eye as well as brain (Takeuchi and Takahashi 1998; Teshigawara et al. 2001). Taken together, it seems that the expression pattern of the MC4 receptor in mammals is much more confined to central regions, as compared with zebrafish and chicken. This may indicate that this receptor has become more specific, and conceivably more important for central functions during mammalian evolution.
The MC5 receptor is expressed in a much wider range in mammalian tissues than any other MC receptor but no physiological function of this receptor is known beyond its role in secretion from exocrine glands (Chen et al. 1997). This receptor has been found in the adrenal gland, stomach, lung, spleen, skeletal muscle, skin, thymus, bone marrow, testis, leukocytes, lymph node, mammary gland, ovary, uterus, pineal gland, liver, fat cells and also in some cases in the brain including the pituitary (Barret 1994; Griffon et al. 1994; Labbéet al. 1994; Chhajlani 1996, see also review Schiöth 2001). The chicken expresses the MC5 receptor in the brain, kidney, liver, adrenals, ovary, testis and adipose tissue but was not found in the heart, spleen and skeletal muscle (Takeuchi and Takahashi 1998). The tissue expression of the two zMC5 receptors does comply with that for the ‘higher’ vertebrates with one exception, as we found expression of the zMC5b receptor in the heart. The only mammalian MC receptor found in the heart is the MC3 receptor but the role is unknown (Chhajlani 1996). These data could prompt the speculation that the zMC5b receptor may either be an ancient MC3 receptor or it may share some of its physiological functions. The phylogenetic, structural, genetic mapping and pharmacology analysis strongly suggests that the MC5b is indeed a duplicate copy of an ancient MC5 receptor gene, rendering the latter suggestion as more likely.
Classical models for evolution of duplicate genes predict that both genes survive only on rare occasions when one duplicate may acquire a new adaptive function and the other retains the old function, resulting in the preservation of both members of the pair (Ohno 1970). The recently developed degeneration complementation (DDC) model predicts however, that the usual mechanism of duplicate gene preservation is the partitioning of ancestral functions rather than the evolution of new functions (Force et al. 1999). The fact that the two copies of zebrafish MC5 receptors show similar pharmacology and are found in multiple tissues with somewhat different distribution pattern may indicate joint preservation of the functions of the parent gene, providing further support for the DDC hypothesis. It has recently been suggested that genes with broad tissue distribution may have a high degree of sequence conservation (Duret and Mouchiroud 2000). The widespread tissue distribution patterns for the MC5 receptors follows this pattern. Because the MC4 receptor is equally well conserved, it would appear likely that its widespread expression in zebrafish and chicken are representative for its role during vertebrate evolution whereas the narrow neuronal expression in mammals may be of more recent origin.
In conclusion, our data show that the MC4 and MC5 receptor subtypes were already present before the divergence of ray-finned fishes and tetrapods, early in vertebrate evolution. Thus, important structural and functional properties of the MC4 and MC5 receptors have been remarkably conserved over a period of at least 400 million years. The receptor clones will facilitate functional characterization of the MC receptors in zebrafish as well as further cloning in other nonamniote species and thereby allow further elucidation of the evolutionary origin and function of the complex MC receptor system. It is conceivable that the MC receptors arose even before the appearance of vertebrates, implying that these receptors participated in an important physiological role that might have to do with regulation of the energy homeostasis.