We studied the in vitro and in vivo effects of octanoylated goldfish ghrelin peptides (gGRL-19 and gGRL-12) on luteinizing hormone (LH) and growth hormone (GH) release in goldfish. gGRL-19 and gGRL-12 at picomolar doses stimulated LH and GH release from dispersed goldfish pituitary cells in perifusion and static incubation. Incubation of pituitary cells for 2 h with 10 nM gGRL-12 and 1 or 10 nM gGRL-19 increased LH-β mRNA expression, whereas only 10 nM gGRL-19 increased GH mRNA expression. Somatostatin-14 abolished the stimulatory effects of ghrelin on GH release from dispersed pituitary cells in perifusion and static culture. The GH secretagogue receptor antagonist d-Lys3-GHRP-6 inhibited the ghrelin-induced LH release, whereas no effects were found on stimulation of GH release by ghrelin. Intracerebroventricular injection of 1 ng/g body wt of gGRL-19 or intraperitoneal injection of 100 ng/g body wt of gGRL-19 increased serum LH levels at 60 min after injection, whereas significant increases in GH levels were found at 15 and 30 min after these treatments. Our results indicate that, in addition to its potent stimulatory actions on GH release, goldfish ghrelin peptides have the novel function of stimulating LH release in goldfish.
- mRNA expression
growth hormone (GH) secretagogues (GHS) are a group of peptide and nonpeptide synthetic compounds that have potent GH-releasing activity (43). In 1996, a G protein-coupled receptor to which the GHS can bind and elicit a biological response was identified from the pituitary and hypothalamus of humans and swine (16). This receptor has been named the GHS receptor (GHSR) (16). Ghrelin is a recently discovered endogenous ligand of the GHSR (23). Ghrelin has a unique, n-octanoyl modification in the third residue, serine, that is essential for its biological activity (23). In mammals, ghrelin is a multifunctional hormone, mainly involved in regulation of GH secretion (23, 50) and food intake (47, 30). Ghrelin stimulates GH release from rat (23, 38), porcine (11), and bovine (12) somatotropes in vitro. Intracerebroventricular (7) and intraperitoneal (49) injections of ghrelin in rats and intravenous injections in humans (41) and rats (31, 45) increase circulating GH levels.
Recently, we reported the structure of the cDNA encoding ghrelin in the goldfish (48). There are two putative cleavage sites and amidation signals in the mature peptide region, suggesting the presence of two amidated ghrelin peptides, 12 (gGRL-12) and 19 (gGRL-19) amino acids long, in goldfish (48). Although we have not purified the ghrelin peptides from goldfish, in a similar study, it was found that two peptides, orexin-A and orexin-B, originate from the preproorexin mRNA, which has two cleavage sites and amidation signals (41). Strong expression of ghrelin mRNA was found in the gut and spleen and weaker expression in the brain of goldfish (49).
Ghrelin has also been identified from other nonmammalian vertebrates, including chicken (21), bullfrog (17), frog (10), eel (19), tilapia (20, 33), and rainbow trout (18). In bullfrog, chicken, eel, and trout, more than one form of ghrelin was found. In rainbow trout, amidated and nonamidated forms of ghrelin, modified with octanoyl or decanoyl groups, are present (18). The role of the amide group in fish ghrelin is not known (18). Octanoylated ghrelin is present in all species from which ghrelin has been isolated and identified, including the species in which ghrelin with other acyl modifications (e.g., decanoylated ghrelin) was also found (18, 21). Intracerebroventricular injections of octanoylated gGRL-19 stimulated food intake in goldfish (48). Using an assay specific for the n-octanoylated ghrelin, we detected a meal-related change in octanoylated ghrelin in goldfish serum (unpublished observations). Ghrelin stimulates GH and prolactin (PRL) release from incubated whole pituitaries of eel (19) and tilapia (20, 39). Recently, Kaiya et al. (18) found that ghrelin stimulates GH release from incubated whole pituitaries of trout, whereas they found no effect on somatolactin and PRL release. Intracranial injections of ghrelin inhibited water intake in eels (24). Here, we present the effects of ghrelin on luteinizing hormone [LH; recently, a consensus was reached to use the terms LH for fish gonadotropin II (GtH II) and LH-β for fish GtH IIβ (51)] and GH release in vitro from dispersed pituitary cells and in vivo after intracerebroventricular or intraperitoneal injections in the goldfish. We also report the effects of ghrelin on LH-β and GH mRNA expression. Furthermore, we demonstrate the in vitro effects of somatostatin-14 (SS-14) on ghrelin-induced GH release and the in vitro effects of the GHSR antagonist d-Lys3-GHRP-6 on ghrelin-induced LH and GH release in goldfish.
MATERIALS AND METHODS
Goldfish (Carassius auratus) of the common or comet variety, ∼35–45 g body wt and in sexually regressed stage, were purchased from Mount Parnell Fisheries (Mercersburg, PA) and maintained under a simulated natural photoperiod of Edmonton, AB, Canada. Fish used for the in vitro studies were maintained in a 300-liter flow-through aquarium that received a constant flow of aerated and dechlorinated city water at 17°C; the fish used for the in vivo experiments were kept in 65-liter glass aquaria receiving aerated water at 20°C. Fish were fed commercially prepared fish pellets (Corey Aquafeeds, Corey Fish Mills, Fredericton, NB, Canada) daily at a specific time. Fish were acclimatized to the respective tanks for ≥14 days. All protocols were approved by the Faculty of Sciences Animal Care Committee of the University of Alberta, based on the guidelines of the Canadian Council for Animal Care.
Hormones and Receptor Antagonist
gGRL-12 and gGRL-19 with n-octanoyl modification (48) were synthesized in the laboratory of Dr. Jean E. Rivier (Clayton Foundation Laboratories for Peptide Biology, The Salk Institute of Biological Sciences, San Diego, CA). Although goldfish ghrelins have not been isolated and characterized, we have used the octanoylated forms of both putative peptides (gGRL-12 and gGRL-19) in the present study. Ghrelin stocks were made by dissolving the peptide in sterile, double-distilled water. Stocks were kept at −20°C. Peptide stocks were thawed and then diluted in fish physiological saline (3) for in vivo studies or in testing medium for in vitro studies. d-Lys3-GHRP-6 was purchased from Phoenix Pharmaceuticals (Belmont, CA). SS-14 was purchased from Peninsula (Bachem) Laboratories (San Carlos, CA). SS-14 and d-Lys3-GHRP-6 stocks were made by dissolving in sterile double-distilled water and frozen. The stocks were diluted in testing media for experiments. Thawed peptide stocks were never refrozen.
Tissue Culture Reagents
Medium 199 was purchased from Sigma-Aldrich (Oakville, ON, Canada). Bovine serum albumin (BSA, fraction V) was purchased from Calbiochem (EMD Biosciences, San Diego, CA). All other solutions and media used for pituitary dispersion, perifusion, and static incubation experiments and primers for amplifying the GH and LH-β probes were purchased from GIBCO-Invitrogen Life Technologies (Burlington, ON, Canada).
In Vitro Experiments
Dispersion of pituitary cells.
Fish were anesthetized by submersion in 0.05% tricaine methane sulfonate (MS-222, Syndel Laboratories, Vancouver, BC, Canada) and euthanized, and pituitaries were removed and placed in dispersion medium (medium 199 with Hanks' salts, 25 mM HEPES, 2.2 g/l NaHCO3, 0.3% BSA, 100,000 U/l penicillin, and 1,000 mg/l streptomycin, with pH adjusted to 7.18 with NaOH). Pituitary cells were dispersed by trypsin-DNase treatment (4), and cells were resuspended in plating medium (which has the same composition as the dispersion medium, except 0.3% BSA was substituted with 1% horse serum).
Column perifusion of pituitary cells.
Perifusion experiments were conducted as described elsewhere (5). Briefly, dispersed cells were cultured overnight on preswollen Cytodex beads at 28°C and 5% CO2 humidity. Before the experiment, ∼1.5 × 106 dispersed cells were loaded onto each column and perifused with testing medium (medium 199 with Hanks' salts, 25 mM HEPES, 2.2 g/l NaHCO3, 0.3% BSA, 100,000 U/l penicillin, and 1,000 mg/l streptomycin, with pH adjusted to 7.18 with NaOH) for 4 h at a rate of 15 ml/h to stabilize the basal secretion rate of GH and LH. After the initial 4 h of perifusion, six 5-min fractions were collected. The average GH concentration in these fractions was considered the basal hormone release level. On the seventh fraction, a 5-min pulse of testing medium containing the specified concentration (0.001, 0.01, 0.1, 1, and 10 nM) of gGRL-12 or gGRL-19 was administered, and the perifusion medium was switched back to the normal testing medium and 5–13 more 5-min fractions were collected (a total of 12–25 fractions from each column). The perifusate fractions were stored at −20°C until the GH and LH radioimmunoassay (RIA). Only a single dose of gGRL-12 or gGRL-19 was applied to each column.
The effect of SS-14 on ghrelin's stimulatory role on GH release was studied using column perifusion. In this experiment, after the collection of six 5-min fractions, in four perifusion columns the perifusion medium was changed to testing medium containing 100 nM SS-14, while two perifusion columns were continuously perifused with plain testing medium (control columns). After 30 min (6 fractions), a 5-min pulse of 10 nM gGRL-19 was administered to all columns, and perifusion was continued with medium containing SS-14 or plain testing medium. After another 25 min (5 fractions), the medium was switched back to plain testing medium in the four columns perifused with SS-14. A total of 24 fractions was collected from each column. The perifusate fractions were stored at −20°C until the GH and LH RIA.
Static incubation of dispersed pituitary cells.
For the 2-h static incubation experiments on GH release, 0.25 × 106 cells·ml−1·well−1 were cultured overnight in plating medium on 24-well Falcon Primaria culture plates (BD Biosciences, Mississauga, ON, Canada) under 5% CO2 humidity and 28°C as described previously (5). Before the experiment, the wells were emptied, 1 ml of fresh testing medium was added, and the plates were again incubated under the same conditions. After 1 h of incubation, plates were removed from the testing medium, and depending on the experiment, 1 ml of plain testing medium or testing medium containing 1 or 10 nM gGRL-12 or gGRL-19, 10 nM gGRL-19 and 1,000 nM SS-14, or 10 nM gGRL-19 and 1,000 nM d-Lys3-GHRP-6 was added, and the plates were incubated again under the same conditions. After 2 h, 800 μl of the medium were collected from each well and stored at −20°C until the RIA for GH and LH. All doses were tested in a minimum of four wells in each experiment, and the experiments were repeated at least three times.
mRNA Expression Studies
Static culture and total RNA extraction.
In this experiment, ∼2 × 106 cells·ml−1·well−1 were plated. On the day of the experiment, the wells were emptied and 1 ml of new testing medium was added, and the plates were incubated for 1 h. The testing medium was removed, and plain testing medium or testing medium containing 1 or 10 nM gGRL-12 or gGRL-19 was added, and the plates were incubated under the same culture conditions described above. After 2 h, 800 μl of the medium were removed, 1 ml of TRIzol reagent (GIBCO-Invitrogen) was added and mixed by repeated pipetting, and the entire contents of each well were transferred to an RNase-free microfuge tube kept on wet ice. Total RNA was extracted immediately following the manufacturer's (TRIzol, GIBCO-Invitrogen) protocol. The concentration of the RNA was determined using a spectrophotometer at 260/280 nm. The RNA samples were kept at −80°C until the slot-blot analysis was carried out.
Slot-blot analysis of GH and LH mRNA expression.
Total RNA (2 μg) was blotted onto a Hybond nylon membrane (Amersham Biosciences, Baie d'Urfe, PQ, Canada) using the Bio-Dot SF blotting apparatus (Bio-Rad Laboratories, Hercules, CA). Specific cDNA probes covering the full length of goldfish LH-β cDNA (GenBank accession no. D88024) and 716 bp of goldfish GH cDNA (GenBank accession no. AF401272) were amplified using specific primers: GtH forward (5′ACTTTTAACAGCCTGCTGAG3′) and GtH reverse (5′ACATTTACACAACATTTATT3′) for LH-β and GH forward (5′ATGGCTAGAGCATTAGTGCTG3′) and GH reverse (5′GATCATAATAACTTAAGGAAG3′) for GH. The PCR conditions for amplifying LH-β cDNA probes were denaturation at 94°C for 5 min followed by 35 cycles of denaturation for 1 min at 95°C, annealing at 54°C for 1 min, and extension at 73°C for 1 min. The PCR conditions for amplifying the GH probe were denaturation at 95°C for 5 min followed by 25 cycles of denaturation for 1 min at 95°C, annealing for 1 min at 50°C, and extension for 1 min at 72°C and a final extension at 72°C for 5 min. The probes were labeled with [α-32P]dCTP using a Rediprime II-Quickprime kit (Qiagen, Mississauga, ON, Canada) following the manufacturer's protocol. The labeled probes were then purified using a QIAquick nucleotide removal kit (Qiagen) following the manufacturer's instructions, and 106 cpm/ml of probe were used for hybridization. First, the membrane was hybridized with the GH probe using previously described methods (48). After hybridization overnight at 65°C, the membranes were washed twice with washing solution (0.04 M NaHPO4, 1 mM EDTA, and 1% SDS) at 65°C. The membranes were then exposed to a PhosphorImager screen (Molecular Dynamics, Sunnyvale, CA) for 3 h, and the screen was scanned using a PhosphorImager 445 SI (Molecular Dynamics) and analyzed using IMAGEQUANT software (Molecular Dynamics). The membrane was then stripped and reprobed with the LH-β probe as described previously and exposed to a phosphor screen for 4 h. The membrane was stripped again and reprobed with an [α-32P]dCTP-labeled goldfish β-actin (GenBank accession no. AF079831) probe as an internal control. To reduce variability, the control and experimental samples were blotted on the same nylon membrane.
In Vivo Experiments
Our previous studies found that intracerebroventricular injections of 1 ng/g body wt of gGRL-19 stimulated food intake in the goldfish (48). Hence, we selected this dose for intracerebroventricular injections in the present study. Four groups (for sampling at 15, 30, 45, or 60 min) of 12 fish (6 control and 6 experimental fish) each were acclimatized to the aquarium conditions for ≥14 days. Intracerebroventricular injections were conducted as described previously (48). Briefly, fish were anesthetized, and a circular saw attached to a dentist drill was used to cut a three-sided flap in the roof of the skull. The flap was folded to the side, and the brain was exposed. Fish were then placed in a stereotaxic apparatus, and the needle of a 5-μl microsyringe (Hamilton, Reno, NV) was placed in the preoptic region of the brain according to the coordinates (+1.0, M, D 1.2) from the brain atlas of Peter and Gill (36). At time 0, fish were injected with 2 μl of fish physiological saline (control fish) or 2 μl of fish physiological saline containing 1 ng/g body wt of gGRL-19 (experimental fish). The cranial cavity was then filled with fish physiological saline, and the skull flap was replaced and sutured using surgical thread. Fish were returned to their respective aquaria to recover from anesthesia (which normally occurs within 5 min). Six control and six experimental fish were anesthetized, and blood samples were collected by caudal puncture at 15 min after injection. Similarly, six control and six experimental fish were sampled at 30, 45, or 60 min after injection. The experiments were conducted in 2 days. Injections were administered in a staggered manner to enable timely sampling. Each fish underwent intracerebroventricular injections and blood sampling only once. Samples were kept at 4°C for several hours to allow clotting and then spun at 8,000 rpm for 5 min for collection of serum. Serum samples were stored at −20°C until the RIA for GH and LH.
Previously, we found that intraperitoneal injections of 100 ng/g body wt of gGRL-19 stimulated food intake in goldfish (unpublished observations). Hence, we used 100 ng/g body wt of gGRL-19 for intraperitoneal injections in the present study. Four groups (for sampling at 15, 30, 45, or 60 min) of 16 fish (8 control and 8 experimental fish) each were acclimatized to the aquarium conditions for ≥14 days. Intraperitoneal injections were conducted as previously described (15). Briefly, fish were anesthetized and placed in a wet sponge holder. At time 0, a 26-gauge needle (Becton Dickinson, Franklin Lakes, NJ) attached to a 250-μl Hamilton syringe was used to inject one group of fish (n = 2) with 100 μl of fish physiological saline (control fish), and a second group (n = 2) was injected with 100 μl of fish physiological saline containing 100 ng/g of gGRL-19 (experimental fish) in the peritoneal cavity. Eight control and eight experimental fish were anesthetized, and blood samples were collected by caudal puncture at 15 min after injection. Similarly, eight control or eight experimental fish were sampled at 30, 45, or 60 min after injection. The experiments were conducted in 2 days. Injections were administered in a staggered manner to enable timely sampling. Each fish underwent intraperitoneal injections and blood sampling only once. Samples were kept at 4°C for several hours to allow clotting and then spun at 8,000 rpm for 5 min to collect serum. Serum samples were stored at −20°C until the RIA for GH and LH.
All samples were assayed for GH and LH (except the SS experiment samples, which were assayed only for GH). GH and LH levels in the perifusate fractions, static culture samples, and serum samples were measured using previously validated RIAs for GH (28) and LH (37). All samples were assayed in duplicate in the same RIA.
Data from the perifusion experiments are presented as the percentage of the prepulse GH or LH levels (means ± SE). Prepulse GH or LH levels are the averages of the GH or LH content in the first six fractions of perifusate collected before the administration of the hormone. Because the basal levels of GH in different columns were different, data were first converted to the percent basal of individual columns and then pooled. For static incubation experiments, the data are presented as percentage of GH or LH levels in control wells (means ± SE). The LH and GH mRNA levels are expressed as a ratio of the GH or LH-β mRNA to β-actin mRNA and then normalized as a percentage of its expression levels in the cells in control wells. Statistical analysis of data from the perifusion and static incubation experiments was conducted by ANOVA followed by least significant difference test. Data from the in vivo experiments (intracerebroventricular and intraperitoneal injections) were analyzed using independent t-test. Significance was considered at P < 0.05.
Cell Column Perifusion
Effects of gGRL-12 and gGRL-19 on LH release.
gGRL-12 at 0.1, 1, and 10 nM caused a significant stimulation of LH release (Fig. 1A); however, there were no significant differences in the LH release responses between these doses of gGRL-12. No significant effects on LH release were found for 0.001 and 0.01 nM gGRL-12 (Fig. 1A).
gGRL-19 at 0.01, 0.1, 1, and 10 nM caused a significant stimulation of LH release (Fig. 1B); however, there were no significant differences in the LH release responses between these doses of gGRL-19. No effects on LH release were found for 0.001 nM gGRL-19 (Fig. 1B).
Effects of gGRL-12 and gGRL-19 on GH release.
gGRL-12 at 0.01, 0.1, 1, and 10 nM caused a significant increase in GH release (Fig. 1C). The effects of 10 nM gGRL-12 in stimulating GH release were significantly greater than the other doses tested. No significant effects on GH were found for 0.001 nM gGRL-12 (Fig. 1C).
gGRL-19 at 0.01, 0.1, 1, and 10 nM caused a significant stimulation of GH release (Fig. 1D); however, there were no significant differences in the GH release responses elicited by 0.01, 0.1, 1, and 10 nM gGRL-19. No significant stimulatory effects on GH release were found for 0.001 nM gGRL-19 (Fig. 1D).
Effects of SS on gGRL-19-induced GH release.
SS-14 at 100 nM inhibited the stimulatory effects of 10 nM gGRL-19 on GH release from perifused pituitary cells of goldfish (Fig. 2).
Static Incubation Experiments
gGRL-12 at 10 nM and gGRL-19 at 10 nM stimulated LH release from dispersed pituitary cells in static culture (Fig. 3, A and B). No effects on LH release were found for 1 nM gGRL-12 or 1 nM gGRL-19.
gGRL-12 at 10 nM stimulated GH release from dispersed pituitary cells in static culture (Fig. 3C). gGRL-19 at 1 and 10 nM stimulated GH release (Fig. 3D). No effects on GH release were found for 1 nM gGRL-12.
SS-14 inhibited 10 nM gGRL-19-induced GH release from dispersed pituitary cells in static culture (Fig. 4). d-Lys3-GHRP-6 inhibited the stimulatory effect of ghrelin on LH release from dispersed pituitary cells in static culture (Fig. 5A). The GHSR antagonist d-Lys3-GHRP-6 had no effect on the stimulatory effects of 10 nM gGRL-19 on GH release from dispersed pituitary cells in static culture (Fig. 5B).
LH-β and GH mRNA Expression
gGRL-12 at 10 nM and gGRL-19 at 1 and 10 nM stimulated LH-β mRNA expression in dispersed pituitary cells in static culture (Fig. 6A). Stimulatory effects of 10 nM gGRL-19 on LH-β mRNA expression were greater than those of 10 nM gGRL-12 and 1 nM gGRL-19. No significant stimulatory effects on LH-β mRNA expression were found for 1 nM gGRL-12 (Fig. 6A).
gGRL-19 at 10 nM increased GH mRNA expression in dispersed pituitary cells in static culture (Fig. 6B). No stimulatory effects on GH mRNA expression were found for 1 and 10 nM gGRL-12 and 1 nM gGRL-19 (Fig. 6B).
Intracerebroventricular administration of 1 ng/g body wt of gGRL-19 significantly increased serum LH levels at 60 min after injection (Fig. 7A). No significant changes in serum LH levels were observed at 15, 30, or 45 min after injection (Fig. 7A). Intracerebroventricular administration of 1 ng/g body wt of gGRL-19 caused a significant increase in serum GH levels at 15 and 30 min after injection (Fig. 7B). Serum GH levels at 45 and 60 min after injection were not significantly different from the saline-injected controls (Fig. 7B).
Intraperitoneal administration of 100 ng/g body wt of gGRL-19 significantly increased serum LH levels at 60 min after injection (Fig. 7C). No significant differences in serum LH levels were found at 15, 30, and 45 min after injection (Fig. 7C). Intraperitoneal administration of 100 ng/g body wt of gGRL-19 caused a significant increase in serum GH levels in goldfish at 15 and 30 min after injection (Fig. 7D). No significant changes in serum GH levels were found at 45 and 60 min after injection (Fig. 7D).
Effects of Ghrelin on LH and GH Release
A novel finding of this study is the in vitro and in vivo stimulation of LH release in goldfish by gGRL-12 and gGRL-19. Both peptides dose dependently stimulated LH release from perifused pituitary cells and pituitary cells in static culture. These results indicate that ghrelin acts directly on the goldfish gonadotropes. The lowest dose that stimulates LH release from the goldfish pituitary cells is 0.01 nM gGRL-19. The minimal effective dose of ghrelin in this study is 0.001–0.01 nM gGRL-19 or 0.01–0.1 nM gGRL-12, indicating that ghrelin is a potent LH-releasing hormone in goldfish. The minimal effective doses of ghrelin that stimulate LH release are comparable to the minimal effective doses of other LH-releasing hormones in goldfish, such as the gonadotropin-releasing hormone (GnRH) (27) and neuropeptide Y (34). Our results indicate that ghrelin acts directly on the goldfish gonadotropes. In a preliminary study, we found that human ghrelin also stimulates LH release from dispersed pituitary cells in vitro (unpublished observations), adding more supportive evidence for our present findings. Our results contradict a previous report of no effects of ghrelin on LH and follicle-stimulating hormone release in vitro from the pituitary cells of rats (23).
As our in vitro studies indicated that both gGRL-12 and gGRL-19 are biologically active in stimulating LH release, gGRL-19 was used for our in vivo studies. Intracerebroventricular (1 ng/g) and intraperitoneal (100 ng/g) injections of gGRL-19 significantly increased serum LH levels at 60 min after injection. These results agree with the findings of our in vitro studies. However, no effects of ghrelin on LH release in vivo in rats were found (23, 46). In sheep, intracerebroventricular injection of ghrelin decreases LH pulse frequency, but not amplitude (8). Our results indicating stimulation of LH release in vitro and in vivo by ghrelin are unique. The results of the present study indicate that ghrelin may have divergent functions (e.g., LH-releasing activity) in certain animal groups or species. For example, ghrelin is an orexigen in mammals (47, 30) and in goldfish (48), whereas it is a potent anorexigen in chicken (40). Similarly, ghrelin stimulates PRL release in tilapia (39) and eel (19), whereas it has no effects on PRL release in rainbow trout (18).
In the present study, we found that ghrelin dose dependently stimulates GH release from dispersed pituitary cells of goldfish in perifusion and static incubation. These results indicate that ghrelin acts directly on the goldfish somatotropes. The lowest dose that stimulates GH release from goldfish pituitary cells is 0.01 nM. The minimal effective dose of ghrelin in this study is 0.001–0.01 nM, indicating that ghrelin is a potent GH-releasing hormone in goldfish. A recent study found that 1 pM ghrelin is effective in stimulating GH release from porcine somatotropes in vitro (26). In goldfish, zebrafish pituitary adenylate cyclase-activating peptide-38 (zPACAP-38) stimulates GH release from dispersed pituitary cells at doses as low as 0.01 nM (49). The minimal effective doses of ghrelin we found in this study are comparable to the minimal effective doses of ghrelin in mammals and of zPACAP-38 in goldfish. In a preliminary study, we found that human ghrelin also stimulates GH release from dispersed pituitary cells in vitro (unpublished observations), adding more support for our present findings. Ghrelin stimulates GH release from the dispersed pituitary cells of rats (23, 38), pigs (11), and cattle (12). Stimulatory effects of ghrelin on GH release were also reported from studies using the incubated whole pituitaries of tilapia (20, 39), eel (19), and trout (18). Our results agree with these studies that found a stimulatory effect for ghrelin on GH release in mammals and fish.
Intracerebroventricular (1 ng/g) and intraperitoneal (100 ng/g) injections of gGRL-19 significantly increased serum GH levels. These results support our in vitro studies in which ghrelin was found to have a stimulatory effect on GH release. In the present study, intracerebroventricular and intraperitoneal injections of gGRL-19 significantly elevated serum GH levels 15–30 min after injection. By 45–60 min after injection, serum GH levels returned to normal levels. Recently, it was reported that intraperitoneal injections of 250 ng/g body wt of trout ghrelin increased serum GH levels in rainbow trout at 30–60 min after injection, whereas no effects on PRL and somatolactin levels were found (18). Ghrelin has been demonstrated to stimulate GH release in vivo in rats (7, 50, 46), chicken (1, 21), and rainbow trout (18). Our results agree with the results of these studies.
The mechanisms of action of intracerebroventricularly administered ghrelin on GH and LH release are not known. In goldfish, the pituitary is directly innervated by hypothalamic neurons (8). It is possible that the intracerebroventricularly injected ghrelin causes differential actions on the hypothalamic neurons expressing hypophysiotropic peptides [e.g., SS and GH-releasing factor (GnRH)] to stimulate GH or LH release in goldfish. It is also possible that intracerebroventricularly injected ghrelin might diffuse to the pituitary or be transported to the pituitary by the blood vasculature that anastomoses the third ventricle and the pituitary (35).
The results of our present studies on the in vitro and in vivo effects of ghrelin peptides indicate that both putative peptides are biologically active in goldfish. Several structure-activity studies have reported that the acylated NH2-terminal region of ghrelin, composed of four to five amino acids, is biologically active in vitro (2, 29). The octanoylated NH2-terminal region is also called the “bioactive core” or the receptor-binding region of ghrelin (29). Although the number of amino acids in ghrelin peptides varies (15, 17, 18, 21, 48), seven amino acids (GSSFLSP) in the NH2-terminal region of ghrelin are totally conserved among species, except in bullfrog and goldfish, where they are GLTFLSP and GTSFLSP, respectively. Multiple ghrelin peptides (15, 17, 21) and ghrelin peptides with many acyl modifications (15, 17, 18, 21) are present in several species. The diversity in the structure of ghrelin peptides suggests species- or tissue-specific actions of ghrelin and, possibly, the presence of multiple ghrelin receptor(s). Nevertheless, chicken ghrelin (21) stimulates GH release in vivo in rats, whereas bullfrog ghrelin causes a weak stimulation of GH release in vivo in rats (17). Rat ghrelin stimulates GH and PRL release in tilapia (39), and human ghrelin stimulates food intake and LH and GH release in goldfish (unpublished observations) and GH release in chicken (1). These studies showing the biological activity of mammalian ghrelin peptides in nonmammals and vice versa suggest that the octanoylated NH2-terminal conserved region of ghrelin peptides, irrespective of the structural variations, can bind to the ghrelin receptor(s) and elicit some biological responses.
Role of SS-14 on Ghrelin-Induced GH Release
The present study demonstrates that SS-14 can inhibit the stimulatory effects of ghrelin on GH secretion in vitro. SS-14 inhibits basal (52) and stimulated GH release in vitro (34) in the goldfish. The inhibitory actions of SS-14 on ghrelin-stimulated GH release provide further evidence for the direct actions of ghrelin on goldfish pituitary cells. Our results agree with the previous reports on the effects of SS on the GH-releasing activity of ghrelin in mammals (3, 44). Several unpublished observations from our laboratory and similar studies conducted in other cyprinid fish (25) found no effects of SS on LH release.
Role of GHSR Antagonist on Ghrelin-Induced LH and GH Release
GHSR1a is the only known active ghrelin receptor (14). d-Lys3-GHRP-6 is a specific GHSR antagonist used in studies on the effects of ghrelin on GH release in mammals (11, 30). Interestingly, d-Lys3-GHRP-6 inhibited the stimulatory effects of 10 nM gGRL-19 on LH release, whereas no effects were found on the stimulatory effects of 10 nM gGRL-19 on GH release from goldfish pituitary cells in static culture. These results indicate that gonadotropes in goldfish have a GHSR1a-like receptor, whereas somatotropes have a different GHSR. Chen (6) proposed the presence of more than one receptor in mediating the effects of synthetic GHS. Recently, Thompson et al. (45) found that ghrelin stimulates bone marrow adipogenesis in vivo in rats by acting independently of the GHSR1a, suggesting the presence of a receptor other than GHSR1a in mediating the physiological functions of ghrelin in rats. It is also worth mentioning that d-Lys3-GHRP-6 only decreased the ghrelin-induced increase in intracellular calcium concentration in porcine somatotropes (11) and had no effects on ghrelin-stimulated GH release in vivo in rats (31). GHSR has been identified in the pufferfish Spheroides nephelus (32). Pufferfish GHSR has a high structural similarity with the mammalian GHSRs, suggesting that the receptor structure and function are evolutionarily conserved (32). The structure of ghrelin receptor(s) in goldfish is unknown. The identification of the structure and distribution of ghrelin receptor(s) and the binding characteristics of the various ghrelin peptides to these receptors will provide useful information on the species-specific biological actions of ghrelin.
In Vitro Effects of Ghrelin on LH-β and GH mRNA
gGRL-12 at 10 nM and gGRL-19 at 1 and 10 nM stimulated LH-β mRNA expression, whereas only 10 nM gGRL-19 stimulated GH mRNA expression. gGRL-19 elicited a dose-dependent increase in LH-β mRNA expression. These results indicate that, in addition to the stimulation of LH and GH release, ghrelin also stimulates the synthesis of LH and GH in goldfish. Previously, it has been shown that releaser hormones such as GnRH stimulate GH and LH-β mRNA expression in the goldfish pituitary cells in vitro (22). However, in the only study available on the effects of ghrelin on GH mRNA expression in mammals, intracerebroventricular injections of ghrelin had no effects on GH mRNA expression in rat pituitary (7). The basis for the difference in the effects of ghrelin on GH mRNA and LH-β mRNA expression may be due to various factors, including the presence of multiple receptors mediating the actions of ghrelin and the differential effects on gene expression by signaling cascades in the different cell types. In the present study, we found that GHSR antagonist blocks the effects of ghrelin on LH release, whereas no effects were found on GH release, indicating that ghrelin acts on LH and GH release through different receptors. It is also clear that the effects of regulatory peptides on LH subunit mRNA expression are differentially transduced by various intracellular signaling mechanisms (51).
In conclusion, we provide evidence for the LH- and GH-releasing activity of gGRL-12 and gGRL-19 in goldfish. Ghrelin acts directly on goldfish pituitary cells to release LH and GH. Central and peripheral injections of ghrelin stimulate LH and GH release in vivo. Ghrelin also stimulates LH-β and GH mRNA expression, indicating that ghrelin is also involved in the synthesis of LH and GH in goldfish. SS-14 abolishes ghrelin's stimulatory effects on GH secretion. GHSR antagonist did not inhibit the stimulatory effects of ghrelin on GH, whereas it completely abolished the effects of ghrelin on LH release, indicating the presence of more than one receptor in mediating the hypophysiotropic actions of ghrelin. The results of the present study, together with our previous reports on the stimulatory effects of ghrelin on food intake, indicate that ghrelin is a hypophysiotropic orexigen, linking the physiology of food intake, growth, and reproduction in goldfish.
This work is supported by Natural Sciences and Engineering Research Council of Canada Grant A06371 to R. E. Peter.
We gratefully acknowledge Dr. Jean E. Rivier and Laura Cervini for a generous gift of goldfish ghrelin peptides, Drs. John P. Chang and Warren K. Yunker for advice on cell culture and RIA, and Dr. Luis Fabian Canosa for critical comments on the manuscript.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2004 the American Physiological Society