Hagfish, Genome Duplications, and RFamide Neuropeptide Evolution

As a phylum, the chordates are remarkably similar in terms of body plan design compared with many of the other major phyla of metazoans. The minimal features for placement within the phylum are a dorsal nerve cord and a notochord and pharyngeal gills slits at some point in the developmental life history of the organism (1). As shown in Fig. 1A, chordates can be initially segregated based on whether the organism lacks an internal cartilaginous or bony skeleton (protochordates; such as Amphioxus or the tunicates) or have an internal skeleton (subphylum Vertebrata). The vertebrates can be further separated into those organisms that lack a hinged jaw (superclass Agnatha; such as the hagfishes or lampreys) and those organisms that have a hinged jaw (superclass Gnathostoma; such as the cartilaginous fishes, bony fishes, amphibians, reptiles, birds, and mammals). As indicated from the fossil record, the ancestral protochordates most likely emerged during the Cambrian Period, some 550 million years ago (MYA) (2), and were followed in relatively quick succession by the ancestral agnathan vertebrates (late Ordovician Period; 450 MYA), and later the ancestral gnathostomes (Silurian Period 420 MYA) (3). Although the external and internal morphological design of the vertebrates (agnathans and gnathostomes) is based on common themes, there are features of the chordates that clearly have undergone considerable change during the evolution of the phylum. Two of these features, the diversity of gene families, and the interactions between the brain and pituitary, are considered in the article by Osugi et al. (4) in this issue of Endocrinology on the evolution, characterization, and localization of RFamide peptides in the central nervous system (CNS) of a hagfish. Why study hagfish, and what information on neuropeptide gene family evolution and brain/pituitary interactions can be gleaned from these organisms?

Fig. 1.

The phylogeny of the chordates. Panel A, The chordates are a monophyletic assemblage that can be divided into three subphyla: protochordates, (Amphioxus and the tunicates), Agnatha (hagfish and lamprey), and Gnathostoma (cartilaginous fishes, bony fishes, and the tetrapods). The common names for each of these groups are presented. Collectively, the Agnatha and Gnathostoma can be grouped together as the vertebrates (chordates with an internal skeleton). The position of three genome duplication events 1R, 2R, and 3R are shown. Panel B, Hypothetical scheme for duplication of an ancestral gene after two genome duplication events. [Modified from R. M. Dores et al.: Mass Spectrometry Reviews 21:220–243, 2002 (10), with permission. © John Wiley and Sons.]

First, when considering the evolution of gene families within chordates, the complexity of these gene families varies considerably between protochordates and gnathostomes. This complexity appears to be the result, in part, of genome duplication events which form the basis for the 2R hypothesis (5, 6). In this hypothesis, R refers to the complete replication of the genome of some ancestral species that in turn served as the initiation point for a new round of diversification within the chordates. It appears that over the past 500 million years, the evolution of phylum Chordata has been punctuated by at least two genome duplication events (Fig. 1A). The first genome duplication event occurred in a lineage of ancestral protochordates and most likely precipitated the emergence of the jawless vertebrates (Agnatha). It also appears that the second genome duplication event occurred in an ancestral agnathan lineage and may have paralleled or led to the emergence of the gnathostomes (Fig. 1A). In this scenario, as shown in Fig. 1B, a single gene in an ancestral protochordate could potentially give rise to four paralogous genes in gnathostomes as a result of the two genome duplication events. Support for the 2R hypothesis comes from studies on several gnathostome gene families with some of the most extensive analyses having been done for the HOX gene clusters (7, 8). After homologous chromosomes have been duplicated after a genome duplication event, the paralogous genes may retain the same function as the ancestral gene (subfunctionalization), evolve novel functions (neofunctionalization), or in some cases become pseudogenes (9). This array of options can be seen for the opioid/orphanin gene family (10). However, the scheme presented in Fig. 1B can be altered as a result of local gene duplication events on chromosomes, gene loss or chromosome loss, and additional genome duplication events within a lineage. An example of the latter phenomenon is the 3R event that occurred during the radiation of the ray-finned fishes (Fig. 1A) (11, 12). An example of how these various mechanisms can affect the evolutionary history of a gene family can be seen from the analysis of the neuropeptide Y receptor gene family (13). Finally, because all of the gnathostome lineages presented in Fig. 1A emerged after the 2R event, it would be unreasonable to assume that the gene sequences within a family have remained unchanged. All genes change over time as a result of point mutations, albeit at different rates, and the presence of conserved amino acid positions within members of a gene family is a reflection of natural selection pressures.

Given these caveats, a perception of the 1R state (outcome of the first predicted chordate genome duplication event) of a gene family may be possible from the characterization of members of the gene family that are expressed in an agnathan species. However, the variety of agnathans to sample from is limited. Although there were several orders of agnathans in the Ordovician Period, only two lineages have persisted to the present: order Myxiniformes (hagfishes; 43 extant species) and order Petromyzontiformes (lampreys; 41 extant species) (14). That said, in the Osugi et al. (4) article, the authors conducted their study on the brown hagfish, Paramyxine atami, an agnathan native to the Sea of Japan.

Conducting a study on a hagfish is also of interest in terms of the evolution of the interaction between the brain and the pituitary. It is clear from the extensive studies that have been done on a broad array of gnathostomes over the past fifty yr that neurosecretory hypophysiotropic neurons originating in or near the hypothalamus release chemical signals (e.g. GnRH, CRH, TRH, GHRH, somatostatin, dopamine) at the median eminence (a capillary plexus linking the superior pituitary artery and the pituitary portal vein). The pituitary portal vein in turn delivers the hypophysiotropic factors to the anterior pituitary, and these factors regulate the release of the various anterior pituitary hormones (15). This elegant neurosecretory/vascular network is found in mammals, birds, reptiles, amphibians, and some of the older lineages of the bony fishes such as the sturgeons and gars. Among the gnathostomes, the major deviation from this pathway is found in the teleosts, the 20,000 plus species of modern bony fishes (14). In a teleost like the zebrafish, the median eminence capillary network has been eliminated and instead the hypophysiotropic neurons directly innervate the anterior pituitary and release their chemical signals in close proximity to the distinct hormone-producing cells that are segregated by cell type in the anterior pituitary of teleosts. By contrast, in the extant agnathan species (hagfishes and lampreys), although these organisms have a pituitary gland, there is neither a median eminence capillary network nor evidence for direct innervation of the pituitary by putative hypophysiotropic neurons (15). This issue begs the question of how the CNS of a hagfish can communicate with the hagfish pituitary.

In this issue of Endocrinology, Osugi and colleagues (4) address the issues via a study on the functional evolution of RFamide-related peptides (neuropeptides with a C-terminal arginine-phenylalanine-amide motif) characterized from the CNS of the brown hagfish. The RFamide peptides are derived from a family of neuropeptide-coding genes that have been identified in both invertebrates and vertebrates (16, 17). From studies on gnathostomes, the chordate branch of this gene family consists of five distinct genes: 1) the PQRFamide gene, which encodes the neuropeptides NPFF, NPAF, and NPSF; 2) the LPXRF gene, which includes the gonadotropin-inhibiting hormone (GnIH); 3) the prolactin-releasing peptide gene; 4) the kisspeptin gene; and 5) the 26RFA gene (18, 19). The phylogenetic relationships of these genes within the family can be seen in Fig. 4 of the article by Osugi et al. (4) in this issue.

As described in the article, biochemical analysis of hagfish CNS extracts resulted in the isolation and characterization of four RF-amide-related neuropeptides that belong to the PQRFamide gene family. Subsequent cloning of PQRFamide-related cDNA from CNS extracts yield two nearly identical cDNA sequences that encode the four peptides. Given the level of nucleotide sequence identity in the 5′ and 3′ untranslated regions of these cDNA, it appears there may be a single hagfish PQRFamide gene, and the authors propose that the two cDNA are the result of a polymorphism. Immunocytochemical and in situ hybridization analyses indicate that the PQRFamide peptide mRNA and the peptides derived from these mRNA can be localized in the infundibular nucleus of the hagfish hypothalamus. In addition, the PQRFamide immunoreactive nerve fibers terminated at blood vessels in the infundibular nucleus rather than at synapses. Therefore, the authors propose that the PQRFamide-secreting neurons are neurosecretory in nature and may have the hagfish pituitary as one of their targets. The latter argument is based on the observation that incubation of hagfish pituitaries with the hagfish PQRFamide-related peptides resulted in an increase in the expression of hagfish gonadotropin β mRNA. It is noteworthy that there is no evidence for direct innervation of the hagfish pituitary by hypophysiotropic neurons originating from the hypothalamus. Hence, these experiments provide support for the hypothesis that hagfish neurosecretory neurons release their chemical signals into the open circulatory system of the organism, and these chemical signals reach the pituitary via simple diffusion (15). These observations are in agreement with an earlier study on the characterization of PQRFamide peptides from a lamprey CNS (18).

With respect to the evolution of the RFamide gene family (see Fig. 4 in Ref. 4, this issue), although the presence of five gnathostome genes in the family is not a perfect match for the hypothetical model of chordate genome duplication events (Fig. 1B), the chordate branch of this gene family could have resulted from two genome duplications of an ancestral RFamide gene present in protochordates to yield four paralogous RFamide genes in the ancestral gnathostomes. A local gene duplication event could have occurred early in the radiation of the ancestral gnathostomes to yield the fifth member of the family. This scenario has been proposed to account for the five members of the melanocortin receptor gene family (19), although other scenarios are possible. However, because the hagfishes and the lampreys represent 1R lineages, and only one RFamide coding gene has been detected in the genomes of the brown hagfish (this article) and the marine lamprey (18), the prediction would be that at least one other RFamide gene should also be present in these 1R species. The hunt for the other proposed RFamide gene would be a reasonable step in understanding the evolution of the chordate RFamide gene family.