The origin and function of the pituitary adenylate cyclase-activating polypeptide (PACAP)/glucagon superfamily.

The pituitary adenylate cyclase-activating polypeptide (PACAP)/ glucagon superfamily includes nine hormones in humans that are related by structure, distribution (especially the brain and gut), function (often by activation of cAMP), and receptors (a subset of seven-transmembrane receptors). The nine hormones include glucagon, glucagon-like peptide-1 (GLP-1), GLP-2, glucose-dependent insulinotropic polypeptide (GIP), GH-releasing hormone (GRF), peptide histidine-methionine (PHM), PACAP, secretin, and vasoactive intestinal polypeptide (VIP). The origin of the ancestral superfamily members is at least as old as the invertebrates; the most ancient and tightly conserved members are PACAP and glucagon. Evidence to date suggests the superfamily began with a gene or exon duplication and then continued to diverge with some gene duplications in vertebrates. The function of PACAP is considered in detail because it is newly (1989) discovered; it is tightly conserved (96% over 700 million years); and it is probably the ancestral molecule. The diverse functions of PACAP include regulation of proliferation, differentiation, and apoptosis in some cell populations. In addition, PACAP regulates metabolism and the cardiovascular, endocrine, and immune systems, although the physiological event(s) that coordinates PACAP responses remains to be identified.

Targeted disruption of the pituitary adenylate cyclase-activating polypeptide gene results in early postnatal death associated with dysfunction of lipid and carbohydrate metabolism.

Pituitary adenylate cyclase-activating polypeptide (PACAP) is a hormone belonging to the glucagon superfamily of hormones. These hormones are known to play important roles in metabolism and growth. PACAP is a neuropeptide that causes accumulation of cAMP in a number of tissues and affects the secretion of other hormones, vasodilation, neural and immune functions, as well as the cell cycle. To determine whether PACAP is essential for survival and to evaluate its function(s), we have generated mice lacking the PACAP gene via homologous recombination. We found that most PACAP null mice died in the second postnatal week in a wasted state with microvesicular fat accumulation in liver, skeletal muscle, and heart. Gas chromatography-mass spectrometry showed that fatty acid beta-oxidation in liver mitochondria of PACAP(-/-) mice was not blocked based on the distribution of 3-hydroxy-fatty acids (C6-16) in the plasma. Instead, increased metabolic flux through the beta-oxidation pathway was suggested by the presence of ketosis. Also, serum triglycerides and cholesterol were significantly higher (2- to 3-fold) in PACAP null mice than littermates. In the fed state, both serum insulin and blood glucose were normal in 5-d-old null mice compared with their littermates. In contrast, fasted PACAP null pups had a significant increase in insulin, but a decrease in blood glucose compared with littermates. Glycogen in the liver was reduced. These results suggest PACAP is a critical hormonal regulator of lipid and carbohydrate metabolism.

Estrous cycles after electrical stimulation of the brain in conscious rats: effect of current strength, estradiol benzoate and progesterone.

Conscious unrestrained rats were stimulated through chronically-implanted electrodes in the median eminence-arcuate (ME-ARC), medial preoptic area (MPOA), amygdala (AMYG) or caudateputamen area of the brain on diestrus-2 of a 4-day estrous cycle or diestrus-3 of a 5-day cycle. Each rat was repeatedly tested after returning to normal cycles according to the following procedure: stimulation at current levels of 25, 50 or 100 muA (biphasic pulses), injection of 1 or 3 mug estradiol benzoate or 1 mg progesterone, and injection of the same hormones 24 h before stimulation at 50 muA. Indirect evidence of advanced ovulation, judged by the pattern of vaginal smears, was obtained depending on the current and site of stimulation: 25 muA was subthreshold for all brain areas, 50 muA was threshold for the ME-ARC and AMYG, and 75-100 muA was very effective in the ME-ARC, but could not be tested in the MPOA or AMYG due to abnormal behavior. Histological studies of the ovary revealed premature luteinization of some follicles and occasional advancement of ovulation. Pseudopregnancy-length diestrus often followed ovulation or advanced ovulation. This event was produced by a lower threshold current of 25 muA in the ME-ARC and MPOA. A current of 50 muA was maximally effective in the ME-ARC, but less so in the MPOA and AMYG; 75-100 muA caused successive periods of pseudoprengancy-length diestrus in the ME-ARC group. It is concluded that specificity of neural circuits from the AMYG to LHRH neurons is questionable since stimuli which led to reproductive changes also produced seizure activity. But stimuli producing no abnormal behavior in conscious rats clearly altered reproductive cycles when applied to the ME-ARC, and in the MPOA produced mino changes in cycles. Estradiol facilitated the effect of stimulation on early appearance of leucocytes in the vaginal smear in the estrous cycle and pseudopregnancy-length diestrus; contrarily, progesterone, in a few cases, inhibited both effects.

Two protochordate genes encode pituitary adenylate cyclase-activating polypeptide and related family members.

To address the origin of the glucagon superfamily, we isolated and sequenced the complementary DNA and partial gene that encode pituitary adenylate cyclase-activating polypeptide (PACAP) from a protochordate (tunicate), a sister group of the amphioxus and vertebrates, but one that evolved before the amphioxus. This is the first report of any superfamily member sequenced from an invertebrate. Transcription of the tunicate pacap1 gene results in a messenger RNA that is 507 bp. The gene contains 3 exons that encode a signal peptide, GRF-like peptide(1-27), and PACAP(1-27). The tunicate GRF-like peptide has 59% identity with human GRF, whereas the deduced amino acids of tunicate PACAP(1-27) have 96% identity with the ovine, human, and salmon PACAP(1-27) forms. Another complementary DNA clone pacap2 was isolated and shown to contain 4 exons that encode a signal peptide, a cryptic peptide, and two peptides that are clearly members of the glucagon superfamily. One of the peptides has 89% sequence identity to the tunicate PACAP encoded in pacap1. A comparison of the two structurally related PACAP clones, each encoding two peptides on separate exons, shows high inter- and intraexon nucleotide sequence identity. Sequence analysis suggests that an exon duplication followed by a gene duplication was responsible for the origin of the two genes. It is argued that the PACAP gene is derived from the protochordate ancestral genes that led to the vertebrate forms of GRF and PACAP.

Regulation and expression of gonadotropin-releasing hormone gene differs in brain and gonads in rainbow trout.

The GnRH gene is transcribed in both the brain and gonads. GnRH in the brain is critical for reproduction, but the function and importance of GnRH in the ovary and testis is not clear. In this study we examine whether regulation of the GnRH gene is distinct in the brain and gonads, whether the regulation of the GnRH gene in the gonads is altered after genome duplication, and whether the regulatory region of the GnRH gene is tightly conserved in vertebrates. From ovary and testis, we isolated and sequenced for the first time two different genes and their complementary DNAs that encode the identical peptide known as salmon GnRH. Rainbow trout were selected because they are tetraploid due to genome duplication. A downstream promoter is used in the brain and gonads by salmon GnRH messenger RNA1 (mRNA1) and mRNA2, but mRNA2 also uses an upstream promoter only in the gonads. Two types of long mRNA2 transcripts in ovary and testis both use an alternative start site at position 323; one of these types also retains intron 1. This long 5'-untranslated region is a likely site for distinct regulation of mRNA in the gonad. Additional evidence for separate regulation is that a different expression pattern exists in brain and gonads for GnRH mRNAs during development and maturation. Gene duplication did not alter the encoded peptide, but changed the expression pattern and resulted in complete divergence of the promoter sequence from position -215. A comparison of the mammalian and trout GnRH genes reveals that the promoters are without sequence identity except for a few consensus sites in both regulatory regions. The duplicated trout genes provide a model to study a critical gene whose product controls reproduction in all vertebrates.

Early expression of pituitary adenylate cyclase-activating polypeptide and activation of its receptor in chick neuroblasts.

To investigate the involvement of pituitary adenylate cyclase- activating polypeptide (PACAP) and GH-releasing factor (GRF) during early chick brain development, we established neuroblast- enriched primary cell cultures derived from embryonic day 3.5 chick brain. We measured increases in cAMP generated by several species-specific forms of the peptides. Dose-dependent increases up to 5-fold of control values were measured in response to physiological concentrations of human/salmon, chicken, and tunicate PACAP27. Responses to PACAP38 were more variable, ranging from 5-fold for human PACAP38 to 4-fold for chicken PACAP38, to no significant response for salmon PACAP38, compared with control values. The responses to PACAP38 may reflect a greater difference in peptide structure compared with PACAP27 among species. Increases in cAMP generated by human, chicken, and salmon/carp GRF were not statistically significant, whereas increases in response to lower-range doses of tunicate GRF27-like peptide were significant, but small. We also used immunocytochemistry and Western blot to show synthesis of the PACAP38 peptide. RT-PCR was used to demonstrate that messenger RNAs for PACAP and GRF and a PACAP-specific receptor were present in the cells. This is a first report suggesting an autocrine/paracrine system for PACAP in early chick brain development, based on the presence of the ligand, messages for the ligand and receptor, and activation of the receptor in neuroblast-enriched cultures.

Exon skipping in the gene encoding pituitary adenylate cyclase-activating polypeptide in salmon alters the expression of two hormones that stimulate growth hormone release.

In mammals, GRF and pituitary adenylate cyclase-activating polypeptide (PACAP) are encoded in separate genes. We report here that in the salmon a 4.5-kilobase gene contains five exons that encode the biologically active part of the GRF-like peptide (amino acids 1-32) on exon 4 and PACAP on exon 5. Analysis of two fish messenger RNAs reveals that a long precursor containing GRF and PACAP and a short precursor containing only PACAP are both expressed in the brain of at least five species of salmon, whereas mice express only the long precursor encoded by the PACAP gene. Synthetic salmon PACAP-38 and salmon GRF-like peptide-45 both stimulated GH release from cultured salmon pituitary cells; PACAP stimulated a concentration-dependent release of GH at both 4 and 24 h of incubation, whereas GRF-like peptide did not. Alternative splicing, resulting in the short precursor in which GRF-32 is excised, may provide a means for differential control of GH secretion with higher production of the more potent PACAP. A duplication of the GRF-like/PACAP gene in evolution after the divergence of fish and tetrapods would explain separate genes and regulation for GRF and PACAP in mammals.

Comparative biological properties of lamprey gonadotropin-releasing hormone in vertebrates.

The biological activities of lamprey GnRH were determined in the lamprey, chicken, sheep, and rat. Lamprey GnRH elevated plasma steroid levels and stimulated ovulation in the lamprey, but had little or no LH-releasing activity in chicken and sheep pituitary bioassays or effect on GnRH receptor-binding activity in the rat pituitary. The lamprey GnRH molecule is structurally distinct from other known vertebrate GnRHs and is the first identified molecule in this family to have different amino acids in the third and sixth positions. These data suggest that the presence of Tyr3 or Glu6 in lamprey GnRH may account for the lack of biological activity in the representatives of the two different vertebrate classes investigated in this study.

Primary structure of a novel gonadotropin-releasing hormone in the brain of a teleost, Pejerrey.

The neuropeptide GnRH is the major regulator of reproduction in vertebrates acting as a first signal from the hypothalamus to pituitary gonadotropes. Three GnRH molecular variants were detected in the brain of a fish, pejerrey (Odontesthes bonariensis), using chromatographic and immunological methods. The present study shows that one form is identical to chicken GnRH-II (sequence analysis and mass spectrometry) and the second one is immunologically and chromatographically similar to salmon GnRH. The third form was proven to be a novel form of GnRH by isolating the peptide from the brain and determining its primary structure by chemical sequencing and mass spectrometry. The sequence of the novel pejerrey GnRH is pGlu-His-Trp-Ser-Phe-Gly-Leu-Ser-Pro-Gly-NH(2), which is different from the known forms of the vertebrate and protochordate GnRH family. The new form of GnRH is biologically active in releasing gonadotropin and GH from pituitary cells in an in vitro assay.

Primary structure and function of three gonadotropin-releasing hormones, including a novel form, from an ancient teleost, herring.

The evolution of GnRH and the role of multiple forms within the brain are examined. Three forms of GnRH were purified from the brain of Pacific herring (Clupea harengus pallasi) and characterized using Edman degradation and mass spectrometry. Two forms correspond with the known structures of chicken GnRH-II and salmon GnRH that are found in many vertebrate species. The third form, designated herring GnRH (hrGnRH), has a primary structure of pGlu-His-Trp-Ser-His-Gly-Leu-Ser-Pro-Gly-NH2. This novel peptide is a potent stimulator of gonadotropin II and GH release from dispersed fish pituitary cells. The content of hrGnRH in the pituitary was 8-fold that of salmon GnRH and 43-fold that of chicken GnRH-II, which provides supporting evidence that hrGnRH is involved in the release of gonadotropin. Herring is the most phylogenetically ancient animal in which three forms of GnRH have been isolated and sequenced. Our evidence suggests that the existence of three GnRHs in the brain of one species 1) is an ancestral condition for teleosts, 2) has the potential for separate regulation of the distinct GnRHs, and 3) may be an evolutionary advantage for refined control of reproduction in different environments.

A second form of gonadotropin-releasing hormone (GnRH) with characteristics of chicken GnRH-II is present in the primate brain.

The primate brain was thought to contain only the GnRH known as mammalian GnRH (mGnRH). This study investigates whether a second form of GnRH exists within the primate brain. We found that brain extracts from adult stumptail and rhesus monkeys contained two forms of GnRH that were similar to mGnRH and chicken GnRH-II (cGnRH-II) based on the elution position of the peptides from HPLC and on cross-reactivity with antisera that are specific to mammalian or chicken GnRH-II in RIAs. The fetal brain of rhesus monkeys also contained mGnRH and a cGnRH-II-like peptide by the same criteria. Immunocytochemistry with a cGnRH-II-specific antiserum in adult and fetal rhesus monkeys showed immunopositive neurons generally scattered in the periaqueductal region of the midbrain, with a few positive cells in the posterior basal hypothalamus. Neurons immunopositive for cGnRH-II were fewer in number and smaller in size, with less defined nuclei and thinner neurites compared with those for mGnRH. Administration of synthetic cGnRH-II to adult rhesus monkeys resulted in a significant increase in the plasma LH concentration during the luteal phase of the menstrual cycle, but not during the midfollicular phase. We conclude that the primate brain contains mGnRH and a cGnRH-II-like molecule, although the function of the latter is unknown.

Molecular cloning of two distinct vasotocin precursor cDNAs from chum salmon (Oncorhynchus keta) suggests an ancient gene duplication.

The structures of two different vasotocin precursors from chum salmon brain have been elucidated through the molecular cloning of their corresponding cDNAs. Although the predicted precursors, consisting respectively of 153 and 158 amino acids, have the same structural organisation, they show 35% amino acid sequence divergence, of which only approximately half are isofunctional substitutions. Remarkably, while the C terminal segments of both precursors resemble the glycopeptide moiety of the related mammalian vasopressin precursor, both salmon precursors lack consensus sequences for N-glycosylation.

Sequence of two gonadotropin releasing hormones from tunicate suggest an important role of conformation in receptor activation.

The primary structure of two forms of gonadotropin releasing hormone (GnRH) from tunicate (Chelyosoma productum) have been determined based on mass spectrometric and chemical sequence analyses. The peptides, tunicate GnRH-I and -II, contain features unprecedented in vertebrate GnRH. Tunicate GnRH-I contains a putative salt bridge between Asp5 and Lys8. A GnRH analog containing a lactam bridge between Asp5 and Lys8 was found to increase release of estradiol compared with that of the native tunicate GnRH-I and -II. Tunicate GnRH-II contains a cysteine residue and was isolated as a dimeric peptide. These motifs suggest that the conformation plays an important role in receptor activation.

Two populations of luteinizing hormone-releasing hormone neurons in the forebrain of the rhesus macaque during embryonic development.

To investigate the possibility that a second luteinizing hormone-releasing hormone (LHRH) population appears during development in primates, embryos and fetal brains of rhesus monkeys were immunostained with antisera specific to different LHRH forms. Two LHRH cell populations were discernible by immunoreactivity to antisera LR-1 and GF-6. Because one LHRH cell type migrated out from the olfactory placode several days earlier than the other, they were referred to as "early" and "late" LHRH cells, respectively. Although late LHRH neurons were immunoreactive to all anti-mammalian LHRH antisera tested, early LHRH neurons were only detected by antiserum GF-6. Early LHRH neurons (approximately 10 x 7 microm) were smaller than late LHRH neurons (approximately 18 x 7 microm). Early LHRH neurons were first found around the olfactory placode, in the nasal mesenchyme, and in the rostroventral forebrain on embryonic day 30 (E30), whereas late LHRH neurons were first seen in the olfactory pit on E32. Early LHRH cells were located throughout the basal forebrain on E32-E42, whereas late LHRH cells were found in the olfactory pit and along the terminal nerve on E34-E36 and were not seen in the forebrain until E38. By E51-E62, late LHRH neurons reached into the basal hypothalamus in a distribution resembling that in the older brain, while early LHRH neurons were found in the septum, preoptic region, stria terminalis, medial amygdala, claustrum, internal capsule, and globus pallidus. Based on the distribution pattern of immunopositive cells with antiserum LR-1, late LHRH cells are bona fide LHRH neurons that regulate the pituitary-gonadal axis. In contrast, the molecular form of early LHRH cells is unclear, although it is plausible that early LHRH cells may contain the molecule in which the C-terminal epitope of LHRH is modified or absent. It is concluded that in primates there is a second population of LHRH neurons that originates from the embryonic olfactory placode before the origin of mammalian LHRH-like neurons, and that these two populations of LHRH-immunopositive neurons have different morphologic features and different final distributions in the brain.

Presence of luteinizing hormone-releasing hormone fragments in the rhesus monkey forebrain.

Previously, we have shown that two types of luteinizing hormone-releasing hormone (LHRH) -like neurons, "early" and "late" cells, were discernible in the forebrain of rhesus monkey fetuses by using antiserum GF-6, which cross-reacts with several forms of LHRH. The "late" cells that arose from the olfactory placode of monkey fetuses at embryonic days (E) 32-E36, are bona fide LHRH neurons. The "early" cells were found in the forebrain at E32-E34 and settled in the extrahypothalamic area. The molecular form of LHRH in "early" cells differs from "late" cells, because "early" cells were not immunopositive with any specific antisera against known forms of LHRH. In this study, we investigated the molecular form of LHRH in the "early" cells in the nasal regions and brains of 13 monkey fetuses at E35 to E78. In situ hybridization studies suggested that both "early" and "late" LHRH cells expressed mammalian LHRH mRNA. Furthermore, "early" cells predominantly contain LHRH1-5-like peptide and its cleavage enzyme, metalloendopeptidase E.C.3.4.24.15 (EP24.15), which cleaves LHRH at the Tyr5-Gly6 position. This conclusion was based on immunocytochemical labeling with various antisera, including those against LHRH1-5, LHRH4-10, or EP24.15, and on preabsorption tests. Therefore, in primates, a group of neurons containing mammalian LHRH mRNA arises at an early embryonic stage before the migration of bona fide LHRH neurons, and is ultimately distributed in the extrahypothalamic region. These extrahypothalamic neurons contain LHRH fragments, rather than fully mature mammalian LHRH. The origin and function of these neurons remain to be determined.

Immunoreactive luteinizing hormone releasing factor in pituitary stalk blood from female rats: sex steroid modulation of response to electrical stimulation of preoptic area or median eminence.

The effects of sex steroid hormones on the responsiveness of the neural mechanism responsible for the secretion of LH-RF have been examined in the female rat. Responsiveness was determined at pro-oestrus by measuring the increments in immunoreactive LH-RF of pituitary stalk blood produced by electrical stimulation of the medial preoptic area or median eminence. Ovariectomy on the morning of dioestrus reduced the LH-RF response to preoptic stimulation while oestradiol benzoate (OB) or testosterone propionate (TP) administered immediately after ovariectomy significantly augmented the response. The facilitatory effect of TP was possibly due to its conversion to an aromatized derivative since 5alpha-dihydrotestosterone monobenzoate was ineffective. Progesterone did not facilitate preoptic responsiveness, and, when administered to animals ovariectomized at 12.00 h of pro-oestrus, reduced the LH-RF response at 18.00 h the same day. Stimulation of the median eminence produced a significantly greater increment in LH-RF than stimulation of the preoptic area. The facilitatory action of OB on the LH-RF response was less marked for median eminence compared with preoptic stimulation. The administration of ICI 46474 at 17.00 h of dioestrus did not reduce preoptic responsiveness on the morning of the next day, suggesting that this compound does not act as an 'antioestrogen' at the level of the preoptic area.

Molecular forms of GnRH in three model fishes: rockfish, medaka and zebrafish.

Three species of fish have become important in the study of reproduction and development. Rockfish are a model for developmental studies of live-bearing perch-like fish, whereas medaka and zebrafish are models for developmental and genetic studies. The forms of GnRH are identified in the brains of each of these fish and in the pituitary of the rockfish to investigate the role of GnRH in reproduction. Here, we report that grass rockfish (Sebastes rastrelliger) have three forms of GnRH in brain extracts as determined by HPLC elution position and RIA. These forms are identified as sea bream GnRH, chicken GnRH-II and salmon GnRH. In contrast, only two forms of GnRH were detected in brain extracts of medaka (Oryzias latipes) and zebrafish (Brachydanio rerio): salmon GnRH and chicken GnRH-II. Rockfish is distinct from medaka and zebrafish in that the most abundant form of GnRH in the rockfish pituitary is sea bream GnRH, whereas this form is absent in the other two fishes. The identification of sea bream GnRH in the rockfish brain and pituitary extracts indicates that the phylogenetic emergence of sea bream GnRH is earlier than the order Perciformes.

Evidence that gonadotropin-releasing hormone (GnRH) functions as a prolactin-releasing factor in a teleost fish (Oreochromis mossambicus) and primary structures for three native GnRH molecules.

Three forms of gonadotropin-releasing hormone (GnRH) are isolated and identified here by chemical sequence analysis for one species of tilapia, Oreochromis niloticus, and by HPLC elution position for a second species of tilapia, O. mossambicus. Of the three GnRH forms in O. mossambicus, chicken GnRH-II (cGnRH-II) and sea bream GnRH (sbGnRH) are present in greater abundance in the brain and pituitary than salmon GnRH (sGnRH). These three native forms of GnRH are shown to stimulate the release of prolactin (PRL) from the rostral pars distalis (RPD) of the pituitary of O. mossambicus in vitro with the following order of potency: cGnRH-II > sGnRH > sbGnRH. In addition, a mammalian GnRH analog stimulated the release of PRL from the pituitary RPD incubated in either iso-osmotic (320 mosmol/l) or hyperosmotic (355 mosmol/l) medium, the latter normally inhibiting PRL release. The response of the pituitary RPD to GnRH was augmented by co-incubation with testosterone or 17 beta-estradiol. The effects of GnRH on PRL release appear to be direct effects on PRL cells because the RPD of tilapia contains a nearly homogeneous mass of PRL cells without intermixing of gonadotrophs. Our data suggest that GnRH plays a broad role in fish, depending on the species, by affecting not only gonadotropins and growth hormone, but also PRL.

Developmental expression, alternative splicing and gene copy number for the pituitary adenylate cyclase-activating polypeptide (PACAP) and growth hormone-releasing hormone (GRF) gene in rainbow trout.

Both growth hormone-releasing hormone (GRF) and pituitary adenylate cyclase-activating polypeptide (PACAP) are encoded on the same gene in fish, but not in mammals. Our objective was to examine the onset and pattern of expression for the grf/pacap gene and to determine whether there is more than one gene in rainbow trout. The results show that grf/pacap mRNA is first expressed at 4 days after fertilization and continues through to hatching. Alternative splicing at all developmental stages produces a full-length transcript and one lacking exon four, which encodes GRF. Thus, independent regulation of the hormones occurs throughout development. Southern analysis shows that two grf/pacap genes exist in trout, but only one gene is responsible for the two identified transcripts. Overexpression of the grf/pacap gene in transgenic fish was attempted, but did not succeed. We conclude that the early and continued expression of grf/pacap mRNA in trout embryos and regulation of the neuropeptide ratio suggests they have a role in early brain development apart from their later role in releasing pituitary hormones.

Characterization of the gene encoding both growth hormone-releasing hormone (GRF) and pituitary adenylate cyclase-activating polypeptide (PACAP) in the zebrafish.

The purpose of this research was to isolate and characterize the gene encoding growth hormone-releasing hormone (GRF) and pituitary adenylate cyclase-activating polypeptide (PACAP) from the zebrafish. The gene is comprised of five exons with two distinct peptides encoded on separate exons, GRF on exon 4 and PACAP on exon 5. Our evidence indicates that the zebrafish genome contains a single copy of the GRF-PACAP gene. The tissue distribution pattern of the mRNA transcript shows expression in the brain, eye, gastrointestinal tract, ovary and testis; each transcript was sequenced and found to be identical to the gene. This is the first report of GRF-PACAP mRNA expression in the eye of a non-mammalian species. Evidence that a duplication of the PACAP gene gave rise to the vasoactive intestinal peptide (VIP) gene is supported by the high amino acid sequence identity between PACAP in zebrafish and VIP in other fish species.