Role of Calcium and EPAC in Norepinephrine-Induced Ghrelin Secretion

Abstract

Ghrelin is an orexigenic hormone secreted principally from a distinct population of gastric endocrine cells. Molecular mechanisms regulating ghrelin secretion are mostly unknown. Recently, norepinephrine (NE) was shown to enhance ghrelin release by binding to β1-adrenergic receptors on ghrelin cells. Here, we use an immortalized stomach-derived ghrelin cell line to further characterize the intracellular signaling pathways involved in NE-induced ghrelin secretion, with a focus on the roles of Ca²⁺ and cAMP. Several voltage-gated Ca²⁺ channel (VGCC) family members were found by quantitative PCR to be expressed by ghrelin cells. Nifedipine, a selective L-type VGCC blocker, suppressed both basal and NE-stimulated ghrelin secretion. NE induced elevation of cytosolic Ca²⁺ levels both in the presence and absence of extracellular Ca²⁺. Ca²⁺-sensing synaptotagmins Syt7 and Syt9 were also highly expressed in ghrelin cell lines, suggesting that they too help mediate ghrelin secretion. Raising cAMP with the phosphodiesterase inhibitor 3-isobutyl-1-methylxanthine also stimulated ghrelin secretion, although such a cAMP-mediated effect likely does not involve protein kinase A, given the absence of a modulatory response to a highly selective protein kinase A inhibitor. However, pharmacological inhibition of another target of cAMP, exchange protein-activated by cAMP (EPAC), did attenuate both basal and NE-induced ghrelin secretion, whereas an EPAC agonist enhanced basal ghrelin secretion. We conclude that constitutive ghrelin secretion is primarily regulated by Ca²⁺ influx through L-type VGCCs and that NE stimulates ghrelin secretion predominantly through release of intracellular Ca²⁺. Furthermore, cAMP and its downstream activation of EPAC are required for the normal ghrelin secretory response to NE.

Ghrelin is a 28-amino acid peptide synthesized and released principally from a distinct group of enteroendocrine cells in the gastric mucosa (1, 2). The peptide, originally identified as working through the GH secretagogue receptor (ghrelin receptor) to potently stimulate GH release, has since been shown to have a plethora of actions, including many related to metabolism, such as stimulation of appetite and adipogenesis, reduction in energy expenditure, preservation of lean body mass, and maintenance of glucose homeostasis (3–8). The peptide is posttranslationally modified by the addition of an octanoyl group at its third amino acid, and this unique acylation is essential for its GH secretagogue receptor-mediated actions (9, 10). Plasma ghrelin levels are modulated by the nutritive and metabolic status of the individual, with fasting and chronic energy deprivation stimulating its secretion and feeding and nutritional abundance suppressing its secretion (5, 11). In turn, ghrelin relays information about feeding and nutrient status to the brain, where it is thought to have many of its effects on metabolism (6).

Despite several advances in the understanding of ghrelin action, knowledge on the molecular mechanisms regulating ghrelin biosynthesis and secretion lags behind. This deficit stems at least in part from the sparse distribution of ghrelin cells, which constitute less than 1% of gastrointestinal mucosal cells and lie scattered within the mucosa, and the initial lack of suitable models with which to investigate ghrelin secretion (1, 12). However, recently, few types of ghrelin secretion models have been developed. These include genetically engineered mouse models in which green fluorescence protein reports on the location of ghrelin-expressing cells, thus enabling direct visualization of ghrelin cells and fluorescence-activated cell sorting-mediated isolation of ghrelin cells for expression analyses and cell culture (12–14). Also, primary cell cultures of dispersed gastric mucosal cells from adult mice and 8-day-old rat pups have been developed to investigate ghrelin secretion (13, 15, 16). Additionally, ghrelin-secreting immortalized cell lines developed from ghrelinomas occurring in the stomachs (stomach-derived ghrelinoma [SG]-1, Mouse Ghrelinoma 3 [MGN3]-1) and pancreatic islets (pancreas-derived ghrelinoma [PG]-1) of transgenic mice expressing simian virus 40 large T-antigen under the control of preproghrelin promoter are now available (17, 18). These ghrelinoma cell lines retain many key phenotypic features of ghrelin cells, responding to many of the same regulators of ghrelin secretion that have been described in vivo and in primary culture systems (13, 15, 17, 18).

Using these models, the direct effects on ghrelin release of various peptide hormones, monoaminergic neurotransmitters, glucose, and fatty acids and of second messengers, potential downstream effector enzymes and channels have now been investigated in a handful of studies. Indeed, insulin, glucagon, oxytocin, somatostatin, dopamine, glucose, and long-chain fatty acids all have been shown to regulate ghrelin secretion through their direct interaction with ghrelin cells (13–18). In addition, all of the above models as well as confirmatory in vivo studies have been used to implicate the catecholamines norepinephrine (NE) and epinephrine as direct ghrelin secretatagogues (13, 15, 17, 19, 20). Fasting-induced elevation of plasma ghrelin levels in mice are inhibited by administration of atenolol, a β1-adrenergic receptor blocker, and by reserpine, an alkaloid that depletes sympathetic nerve terminals of catecholamines (17). These data are supported by high levels of β1-adrenergic receptor expression in ghrelin cells enriched from the stomach of ghrelin-humanized Renilla green fluorescent protein reporter mice as well as in the SG-1 and PG-1 ghrelin cell lines (17). Forskolin, a potent activator of adenlyl cyclase, mimics the effect of nonepinephrine (17), suggesting that activation of adenylyl cyclase and an ensuing elevation of cAMP occurs after engagement of β1-adrenergic receptors, as has been shown in other cell systems (21, 22). Altogether, these findings link fasting-induced stimulation of the sympathetic nervous system and ensuing release of NE locally in the stomach wall to the release of ghrelin (20, 23).

In the present study, we further characterize the cellular events mediating ghrelin secretion, using the SG-1 ghrelin cell line as a model system. We show that constitutive ghrelin secretion from these cells is mediated by Ca²⁺ influx through voltage-gated Ca²⁺ channels (VGCCs) and that NE induces enhancement of Ca²⁺ influx as well as release of Ca²⁺ from intracellular stores. We also identify the presence of Ca²⁺-sensing synaptotagmins (Syt) that may be involved in ghrelin release. Finally, our findings suggest that the actions of NE are mediated by exchange protein-activated by cAMP (EPAC), independently of protein kinase A (PKA).

Materials and Methods

Chemicals

All chemicals were purchased form Sigma-Aldrich, except for brefeldin A and 8-(p-chlorophenylthio)-2′-O-methyladenosine-3′,5′-cAMP (8CPT-2Me-cAMP) (Tocris Bioscience) and myristoylated PKA inhibitor (14–22) amide (PKI) (EMD Millipore).

Cell culture and ghrelin secretion experiments

Establishment of SG (SG-1) cells was described previously (17). Cells were maintained in DMEM/F-12 (1:1) medium (Mediatech, Inc) containing 10% (vol/vol) Fetal Bovine Serum (Atlanta Biologicals), supplemented with 100 U/mL penicillin and 100 μg/mL streptomycin sulfate in a 37°C incubator with 5% CO₂ and 5% O₂. For secretion experiments, cells were plated at a density of 5 × 10⁴ cells/mL per well onto 24-well plates precoated with poly-D-lysine (d 0). On day 2, sodium octanoate-BSA was added to the medium to achieve a final concentration of 50μM. On day 3, the medium was aspirated, and the cells were treated with the test compounds in 500 μL serum-free DMEM (Life Technologies) supplemented with 5mM glucose and 50μM sodium octanoate. After a 6-hour incubation, the medium was collected, placed on ice, and immediately centrifuged at 3000 rpm for 5 minutes. Hydrochloric acid was added to the supernatant to achieve a final concentration of 0.1N (for stabilization of acyl-ghrelin) and stored at −80°C until analysis. Assay for acyl-ghrelin was performed using a commercial ELISA kit (Cayman Chemical) according to the manufacturer's instructions, in a PowerWave XS Microplate spectrophotometer (BioTek Instruments, Inc).

Calcium imaging

To measure changes in cytosolic Ca²⁺ concentrations ([Ca²⁺]_c), SG-1 cells were treated with Accutase (Innovative Cell Technologies) and plated as dispersed single cells onto 96-well plates precoated with collagen I and poly-L-Lysine (Greiner Bio-One). After sitting overnight in a 37°C incubator with 5% CO₂, SG-1 cells were loaded with 5μM Fura-2 AM in Krebs-Ringer-HEPES (KRH) buffer (119mM NaCl, 4.7mM KCl, 2.5mM CaCl₂, 1.2mM MgCl₂, 1.2mM KH₂PO₄, 10mM HEPES, and 5mM D-glucose) or Ca²⁺-free KRH buffer (119mM NaCl, 4.7mM KCl, 1.2mM MgCl₂, 1.2mM KH₂PO₄, 10mM HEPES, and 5mM D-glucose). [Ca²⁺]_c in individual cells were continuously monitored at 37°C using a BD pathway 855 Bioimaging system (BD Biosciences). Baseline [Ca²⁺]_c was recorded for 1.5 minutes before addition of test substance. Kinetic measurements were made to monitor changes in [Ca²⁺]_c with treatment.

RNA extraction and quantitative PCR (qPCR)

Total RNA was isolated from cultured cells or tissues using the guanidium thiocyanate-phenol-chloroform extraction method by addition of RNA STAT-60 (Tel-Test, Inc), as described previously (13). Total RNA (2 μg) was treated with ribonuclease-free deoxyribonuclease (Roche), and cDNA was synthesized by reverse transcription using SuperScript II (Invitrogen). Our gene-specific primers were as used in previous publications (13, 24, 25) or as designed de novo using Primer Express software (PerkinElmer Life Sciences). Primer sequences are available upon request. All primers were validated by analysis of template titration and dissociation curves. qPCR was performed using an ABI Prism 7900HT sequence detection system (Life Technologies, formerly Applied Biosystems) and SYBR Green chemistry. The reaction mixture for qPCR contained 25 ng of reverse-transcribed RNA, 150nM each primer, and 5 μL of 2× SYBR Green PCR master mix (Applied Biosystems). The mRNA levels are represented relative to the invariant control gene 36B4 or cyclophilin B and were calculated by the comparative threshold cycle method as described previously (13, 24). Data are presented as a percentage to the highest expressed gene.

Animal housing

All animal experiments were approved by the University of Texas Southwestern Medical Center Institutional Animal Care and Use Committee. Mice were fed ad libitum with standard chow diet (Teklad Global Diet 16% protein diet [2016]; Harlan Teklad) and housed under a 12-hour light, 12-hour dark cycle in standard environmentally controlled conditions until the day of experiment.

Immunohistochemistry

Mice were deeply anesthetized by ip administration of chloral hydrate (500 mg/kg body weight) and transcardially perfused with diethyl pyrocarbonate-treated 0.9% PBS followed by 10% neutral buffered formalin. Stomachs were removed, rinsed in PBS, fixed in formalin for 4–6 hours at 4°C, and then stored overnight in 20% sucrose in diethyl pyrocarbonate-treated PBS (pH 7.0), at 4°C. The day after, 15 μm-thick cryosections of stomach tissue were prepared and mounted on SuperFrost slides (Fisher Scientific), air dried, and stored at −80°C until further processing. Thereafter, slides were washed in PBS 3 times, treated with 0.3% hydrogen peroxide in PBS for 30 minutes, and then incubated with 3% normal donkey serum (Equitech-Bio) in 0.25% Triton X-100 PBS at room temperature. Immunolabeling was performed overnight with goat polyclonal antighrelin serum (1:200 dilution, catalog no. sc-10368; Santa Cruz Biotechnology, Inc) and rabbit polyclonal anti-Synaptotagmin (Syt)7 serum (1:200 dilution, catalog no. 105172; Synaptic Systems) in 0.25% Triton X-100 PBS at room temperature. After washes with PBS, the slides were incubated in Alexa Fluor 594 donkey antigoat IgG (1:250; Molecular Probes, Life Technologies) and Alexa Fluor 488 donkey antirabbit IgG (1:250; Molecular Probes) for 1 hour at room temperature. The sections were then mounted with Fluromount G (Electron Microscopy Sciences). Immunoreactivity to ghrelin and Syt7 was examined using an inverted microscope (Model IX51; Olympus USA), and photomicrographs were taken using a camera (Model DP21; Olympus USA) fitted to the microscope. Specificity of the labeling was tested by omitting the primary antibody; no labeling was detected in these control sections.

Data analysis and statistics

Statistics were performed using GraphPad Prism 5.0 (GraphPad Software, Inc), and all data are expressed as mean ± SEM. P values less than or equal to .05 were considered statistically significant.

Results

NE elevates cytosolic calcium in SG-1 cell line

NE directly induces ghrelin secretion from ghrelin cells through activation of β1-adrenergic receptors (13, 15, 17). However, the intracellular signaling cascade that couples activation of β1-adrenergic receptors to exocytosis of ghrelin has not been fully addressed. Because elevation of [Ca²⁺]_c is a key trigger for exocytosis in other endocrine cell types (26), we examined its role in ghrelin secretion, using Fura-2 ratiometric Ca²⁺ indicator dye. Addition of 10μM NE to SG-1 cells incubated with 2.5mM Ca²⁺-containing KRH buffer enhanced [Ca²⁺]_c within a few seconds and plateaued after the peak response (Figure 1A). Subsequent treatment with a depolarizing stimulus (30mM KCl) confirmed cell viability and response to membrane depolarization. The further enhancement of [Ca²⁺]_c in response to KCl also indicated that the depolarization induced by 10μM NE was not maximal (Figure 1A). When this experiment was repeated in the presence of Ca²⁺-free KRH buffer, NE (10μM) again induced elevation of [Ca²⁺]_c, with a peak response similar to that observed in Ca²⁺-containing medium (Figure 1B). However, unlike the sustained [Ca²⁺]_c response occurring in the presence of Ca²⁺-containing buffer, that observed in Ca²⁺-free buffer was more transient and started to wane after achieving a peak elevation (Figure 1B). Also, the absence of extracellular Ca²⁺ prevented further elevation of [Ca²⁺]_c in response to the 30mM KCl depolarizing stimulus, as had been achieved in Ca²⁺-containing buffer. These results suggest that NE (10μM) induces elevation of [Ca²⁺]_c by release of Ca²⁺ from intracellular stores, with a sustained response requiring extracellular Ca²⁺.

Figure 1.

NE induces elevation of free cytosolic calcium in SG-1 cells. Traces of NE-induced changes in [Ca²⁺]_c in the presence (A, n = 8 cells) or absence of extracellular calcium (B, n = 4 cells). Change in [Ca²⁺]_c is indicated by the change in the ratio of emission from Fura-2 dye, when excited with 340 and 380 nm of light, and continuously monitored using live cell imaging. Responses to 30mM KCl also are indicated. Values are expressed as mean ± SEM.

Effects of interfering with calcium influx on ghrelin secretion

We next tested the effect of blockade of Ca²⁺ influx on ghrelin secretion, using nifedipine (50μM; a selective L-type dihydropyridine VGCC blocker), EGTA (2mM; a Ca²⁺ chelator), and CdCl₂ (100μM; a nonselective VGCC blocker). All 3 treatments independently reduced constitutive acyl-ghrelin secretion from the SG-1 cells (nifedipine by 82.6 ± 1.4%, EGTA by 79.1 ± 2.0%, and CdCl₂ by 50.3 ± 1.9%) (Figure 2A). Addition of NE (10μM) in the absence of Ca²⁺ entry blockers enhanced ghrelin secretion by 106.4 ± 10.9%. That enhancement was significantly reduced when incubated in the presence of the VGCC blockers or Ca²⁺ chelator (Figure 2A). We also generated dose-response curves of NE-induced ghrelin secretion in the presence and absence of 50μM nifedipine (Figure 2B). Nifedipine induced a significant reduction in ghrelin secretion at all employed concentrations of NE (0μM–10μM), completely blocking NE-induced enhancement of ghrelin secretion at NE concentrations less than 1μM (Figure 2B).

Figure 2.

Effects of calcium entry blockade on constitutive and NE-induced ghrelin secretion. (A) Acyl-ghrelin concentration in the culture medium after 6 hours of incubation with (black bars) or without (open bars) NE (10μM) in the presence of nifedipine (50μM), EGTA (2mM), or CdCl₂ (100μM). *, P < .05, significant difference in acyl-ghrelin concentrations with NE treatment when compared with acyl-ghrelin concentrations in the absence of NE but in the presence of same Ca²⁺ entry modulator or vehicle. #, P < .05, significant change in acyl-ghrelin concentrations due to treatment with Ca²⁺ entry modulators compared with vehicle control, n = 4 wells in each group. Values are expressed as mean ± SEM. Data were analyzed by two-way ANOVA followed by Bonferroni post hoc analysis. (B) Dose response of NE-induced acyl-ghrelin secretion in absence (control) or presence of 50μM nifedipine (n = 5 wells in each group). Values are expressed as mean ± SEM. Data were analyzed by two-way ANOVA followed by Bonferroni post hoc analysis. ***, P < .001, comparing NE-induced ghrelin secretion in the presence of nifedipine with NE-induced ghrelin secretion in the absence of nifedipine. #, P < .05, significant change in subsequent NE-induced acyl-ghrelin secretion when compared with the baseline (acyl-ghrelin concentrations in response to the cumulative average of the first 3 concentrations of NE) of the respective curve. Data were analyzed by one-way ANOVA followed by Tukey's post hoc analysis.

Expression of calcium channels in ghrelin cells

Ca²⁺ flux into ghrelin cells could be occurring through 1 or more members of several different families of Ca²⁺ channels. Above, we used pharmacology to pinpoint a role for L-type VGCCs in ghrelin secretion. In order to gain both a broader and more specific understanding of the types of Ca²⁺ channels that may be involved in ghrelin release, we performed qPCR using mRNA isolated from SG-1 cells, quantifying the transcript levels of VGCC subunits (α₁, α₂δ, β, and γ). The expression profile revealed diverse expression of Ca²⁺ channel subunits (Figure 3A). The high voltage-activated L-type α₁ subunits (Ca_v1.2 and Ca_v1.3) were the predominant pore-forming subunit expressed. The high voltage-activated α₁ subunits P/Q type (Ca_v2.1), R type (Ca_v2.3), and N type (Ca_v2.2) and the low voltage-activated α₁ subunit T type (Ca_v3.2) were also detected. Of the α₂δ accessory subunits, α₂δ-1 (Cacna2d1) followed by α₂δ-3 (Cacna2d3) predominated. β2 (Cacnb2) and β3 (Cacnb3) were the predominant β-subunits. γ7 (Cacng7), γ3 (Cacng3), and γ4 (Cacng4) were the predominant γ-subunits (Figure 3A).

Figure 3.

Expression profile of calcium channels in SG-1 cells. (A) Relative mRNA levels of VGCC and ligand-gated Ca²⁺ channel subunits determined by qPCR (n = 3 replicate samples). (B) Relative mRNA levels of SERCA determined by qPCR (n = 3 replicate samples). Values are expressed as mean ± SEM. All values were mathematically adjusted relative to the highest expressed gene, set as 100%.

In addition, we determined relative levels of expression of inositol triphosphate (IP₃) ligand-gated receptors, which are known to release Ca²⁺ from the endoplasmic reticulum. Inositol 1,4,5-trisphosphate receptor, type 1 (ITPR1) mRNA was the most abundant type of IP₃ receptor detected, with little or no expression of other IP₃ receptor subtypes (Figure 3A). Finally, we assessed expression of sarco/endoplasmic reticulum Ca²⁺ ATPase (SERCA) family members. SERCAs are known to sequester Ca²⁺ back into the endoplasmic reticulum. Of the 3 known genes in the family, SERCA2 was most highly expressed in SG-1 cells, followed by SERCA1 and SERCA3 (Figure 3B).

Expression of synaptotagmins in the ghrelin cells

We next sought to determine expression by ghrelin cells of members of the Syt family. Syts are the primary membrane-trafficking proteins enabling fusion of phospholipid membranes, thus resulting in exocytosis of hormones and neurotransmitters from endocrine cells and neurons, respectively, in response to elevation of [Ca²⁺]_c (27). First, qPCR was performed on mRNA from SG-1 cells, PG-1 cells, and isolated C57BL/6 mice gastric mucosal cells, using primer sets specific for each of the 16 known members of the classic Syt family. Syt7, Syt9, and Syt11 were abundantly expressed by both ghrelinoma cell lines (Figure 4, A and B); Syt16 was the Syt most abundantly expressed by gastric mucosal cells, followed by Syt7 and Syt11 (Figure 4C).

Figure 4.

Expression profile of synatotagmins in ghrelin-secreting cells. Comparison of relative mRNA expression levels of all known Syts in (A) SG-1 cells, (B) PG-1 cells, and (C) mouse gastric mucosa (n = 3 replicate samples each). (D) Relative levels of Syt7 mRNA expression in SG-1 and PG-1 cells in comparison with mouse liver, brain, stomach, pancreatic islets, and islet-derived endocrine cell lines (αTC1, βTC6, and MIN6) (n = 3 replicate samples). Values are expressed as mean ± SEM. All values were mathematically adjusted relative to the highest expressed gene, set as 100%.

We further characterized the expression of Syt7, because it is the predominant Ca²⁺-binding Syt found in large dense-core endocrine secretory vesicles (25, 28–30). This was accomplished first using qPCR, by determining the relative expression of Syt7 in ghrelin cells in comparison with other endocrine cell lines and mouse tissues, several of which are known to express Syt7 (Figure 4D) (29, 31, 32). Syt7 expression in SG-1 and PG-1 cells was higher than that observed in whole stomach. Syt7 was expressed in SG-1 and PG-1 cells at levels comparable with those in the pancreatic β-cell lines βTC6 and MIN6 but less than the levels in the pancreatic α-cell line αTC1, pancreatic islets, and brain. Furthermore, expression of Syt7 in ghrelinoma cells was higher than that observed in the liver. Next, immunohistochemistry was performed to determine the spatial distribution of Syt7-containing cells within the mouse gastric mucosa. Within C57BL/6 stomach sections, Syt7 immunoreactivity was detected within the vast majority of ghrelin-immunoreactive cells as well as some non-ghrelin cells (Figure 5).

Figure 5.

Colocalization of ghrelin and Syt7 in mouse gastric mucosa. Representative photomicrographs demonstrating cells with Syt7-immunoreactivity (green) and/or ghrelin-immunoreactivity (red). Right panel shows the merge of both, with yellow indicating dual-labeled cells. Arrows indicate representative cells with Syt7-immunoreactivity only (left facing), ghrelin-immunoreactivity only (upward facing), and both ghrelin- and Syt7-immunoreactivity (right facing). Scale bars, 50 μm.

NE-induced ghrelin secretion involves elevation of cAMP and activation of EPAC

Although elevation of [Ca²⁺]_c serves as the primary trigger for exocytosis in many endocrine cell types, much evidence suggests that cAMP also serves as an important second messenger modulating the endocrine secretory pathway (33). Given that general knowledge and also the previous finding of roles for β1-adrenergic receptors and adenylyl cyclase in ghrelin release, we sought to determine whether pharmacological elevation of cAMP would stimulate ghrelin secretion. Addition of 3-isobutyl-1-methylxanthine (IBMX) (500μM), an inhibitor of phosphodiesterase enzymes that hydrolyze and inactivate cAMP, enhanced ghrelin secretion in SG-1 cells by 47.0 ± 0.15%; this effect of IBMX was not additive with NE-induced ghrelin secretion (Figure 6A). Next, we tested the involvement of downstream effectors of cAMP (PKA and EPAC) in regulating ghrelin secretion. Pretreatment for 16 hours with myristoylated PKI (10μM), a specific PKA inhibitor (34) that inhibits cAMP-regulated PKA-dependent exocytosis (33), did not inhibit constitutive or NE-induced ghrelin secretion from SG-1 cells (Figure 6B). However, brefeldin A (100 μM), a known inhibitor of EPAC, decreased both constitutive and NE-induced ghrelin secretion (Figure 6, C and D). Conversely, the EPAC agonist 8CPT-2Me-cAMP (10 μM) enhanced basal ghrelin secretion (Figure 6C), an effect that was prevented by simultaneous addition of brefeldin A (Figure 6C).

Figure 6.

NE-induced enhancement of ghrelin secretion is mediated by activated EPAC but not PKA, downstream of cAMP. (A) Acyl-ghrelin concentrations in the culture medium after 6 hours of incubation without treatment, or with vehicle (Dimethyl sulfoxide), IBMX (500μM), combination of NE (10μM) and IBMX (500μM), or with NE (10μM) alone. Each bar represents acyl-ghrelin levels in the medium relative to the untreated control. (B) Acyl-ghrelin concentrations in the culture medium after 6 hours of incubation with PKI (10μM), combination of PKI (10μM) and NE (10μM), or NE (10μM) alone. (C) Acyl-ghrelin concentrations in the culture medium after 6 hours of incubation with 8CPT-2Me-cAMP (10μM), brefeldin A (100μM), or a combination of brefeldin A and 8CPT-2Me-cAMP. (D) Acyl-ghrelin concentrations in the culture medium without treatment or treatment with NE (10μM) or combination of brefeldin A and NE (n = 9 wells in each group). Values are expressed as mean ± SEM. Data were analyzed by one-way ANOVA followed by Bonferroni post hoc analysis. *, P < .05, ***, P < .001. Bars with the same alphabetic letter indicate no significant change.

Discussion

Here, we used the SG-1 ghrelinoma cell line to investigate the role of Ca²⁺ and cAMP-engaged signaling molecules in both baseline (constitutive) ghrelin secretion and NE-induced ghrelin secretion. Regarding constitutive ghrelin release, we showed that blockade of L-type VGCC using nifedipine reduced release of ghrelin by about 80%. This finding indicates that influx of Ca²⁺ through L-type VGCC is an influential physiological trigger for ghrelin secretion in SG-1 cells and presumably also in native ghrelin cells in vivo (Figure 2). That amount of ghrelin secretion persisting in the presence of nifedipine may be triggered instead by Ca²⁺ entry through nifedipine-insensitive VGCCs that also were found to be expressed in these cells (Figure 3). These results also suggest that regulatory modulators that alter membrane potential, thereby altering the activity of VGCCs, are likely to have profound effects on ghrelin release. EGTA, which chelates extracellular Ca²⁺ (35) and CdCl₂, a nonselective VGCC blocker (36), mimicked the effect of nifedipine in significantly inhibiting constitutive release of ghrelin (Figure 2A). Altogether, these data suggest that Ca²⁺ influx through L-type VGCCs plays a predominant role in constitutive ghrelin secretion. These findings are supportive of a previous study that demonstrated inhibition of glucagon-stimulated ghrelin secretion in the presence of a different L-type VGCC blocker, using primary cultures of gastric mucosal cells established from 8-day old pups (16). The dominant role of the L-type VGCC in mediating ghrelin secretion despite the presence of other non-L-type VGCCs, as was observed here, mirrors the situation in pancreatic β-cells, where many VGCCs are expressed but the L-type plays a predominant role in insulin secretion (37–39). To the best of our knowledge, this is the first comprehensive report showing expression and function of VGCCs in ghrelin cells. Further electrophysiological characterization along with genetic approaches might help clarify the functions of the other VGCC subtypes within the ghrelin cell.

Next, we examined NE-induced changes in [Ca²⁺]_c, because they related to ghrelin secretion. Similar to previous observations (15, 17), NE induced a dose-dependent increase in ghrelin secretion with a maximal response observed at a concentration of 1μM (Figure 2B). Our previous study indicated that NE (10μM)-induced ghrelin secretion from ghrelinoma cells was sustained in a time-dependent manner up to a tested time of 6 hours (17). In comparison, NE-induced elevation of [Ca²⁺]_c in the absence extracellular Ca²⁺ (Figure 1), presumably via release of Ca²⁺ from intracellular stores, was relatively transient, declining quickly after the peak level was obtained. Therefore, sustained NE-induced elevation of [Ca²⁺]_c over several hours can be achieved through enhancement of VGCC activity and resulting influx of extracellular Ca²⁺. Indeed, a fraction of NE-induced enhancement of ghrelin secretion was abolished in the presence of nifedipine (Figure 2B). The effect of NE-enhanced VGCC activity is more apparent at low NE concentrations (<1μM). Additionally, NE-induced Ca²⁺ release from intracellular stores in turn may cause further sustained influx of Ca²⁺ from the extracellular space through poorly characterized capacitative or store-operated channels localized to the plasma membrane, as has been observed in other cell types (40). The latter possibility is further supported by a previous study in which depletion of intracellular Ca²⁺ stores with thapsigargin enhanced ghrelin secretion from primary gastric mucosal cell cultures (15). Expression by SG-1 cells of several VGCC family members (predominantly the L-type) as well as the ITPR1 IP₃ receptor and members of the SERCA family (Figure 3) also supports the roles of both extracellular and intracellular Ca²⁺ pools in ghrelin cell function. In summary, the present findings implicate involvement of both Ca²⁺ entry (from the extracellular space) and Ca²⁺ release (from intracellular pools) mechanisms in NE-induced ghrelin secretion, whereby the contribution of Ca²⁺ release to enhanced [Ca²⁺]_c is higher with increasing concentrations of NE.

We also examined the expression profile of members of the Syt family within the ghrelinoma cells. Several members of the family function as Ca²⁺ sensors, interacting with the soluble N-ethylmaleimide-sensitive factor attachment protein complex in the plasma membrane in response to secretagogue-induced elevations in [Ca²⁺]_c to mediate fusion of the secretory vesicles with the plasma membrane and a resulting release of hormone (27). Of the 3 Syts most highly expressed in ghrelinoma cells (Syt11, Syt7, and Syt9) (Figure 4), only Syt7 and Syt9 have been shown to bind Ca²⁺, suggesting their involvement in ghrelin release (27, 41). The abundant expression of Syt7 and Syt9 is similar to that observed in pancreatic β-cells (25), and Syt7 has previously been shown to be a major Ca²⁺ sensor mediating exocytosis in pancreatic β-cells and other similarly well-studied endocrine cell types, such as pancreatic α-cells and chromaffin cells (25, 27, 30, 42, 43). Syt7, but not Syt 9, binds Ca²⁺ with high affinity and is typically present in large-dense core endocrine vesicles (25, 28, 44, 45) as occur in ghrelin cells (46, 47). We propose that Syt7 may be the likely key Ca²⁺ sensor-mediating ghrelin secretion.

As a final point of discussion, the cAMP signaling system has been found in several other systems to be an important modulator of regulated exocytosis, presumably by potentiating the effects of Ca²⁺-dependent exocytosis (33, 48). In endocrine cells, the Ca²⁺-dependent exocytosis facilitated by cAMP has been reported to be mediated by either of the 2 major downstream effectors of cAMP, EPAC and PKA (33). Here, we show that the NE-induced secretion of ghrelin involves activated EPAC and not PKA (Figure 6). These results differ from a previous observation that showed NE-induced ghrelin secretion to be sensitive to inhibition of PKA (15). The discrepancy between these 2 observations is not immediately apparent but could be due to differences in experimental protocol: 1) we used a mouse-derived, immortalized ghrelinoma cell line, whereas the previous study used primary gastric mucosal cell preparations derived from 8 day-old rat pups; 2) we used the highly selective PKA inhibitor PKI (49), whereas the previous study used a less-specific PKA inhibitor, N-[2-(bromocinnamylamino)ethyl]-5-isoquinoline sulfonamide (H89), which inhibits not only PKA but also several additional kinases as well as the β1-adrenergic receptor (50, 51); and 3) numerous differences between the 2 studies' culture conditions existed. Regardless of whether PKI or H89 was used, constitutive ghrelin secretion was not affected (Figure 6B) (15, 16). Genetic approaches and characterization of the spatial localization of PKA and EPAC within the signaling domains and the secretory vesicles of ghrelin cells may help clarify these confounding results.

A schematic model illustrating the participation of extracellular and intracellular Ca²⁺, L-type VGCCs, Syts, cAMP, and EPAC in ghrelin secretion appears as Figure 7.

Figure 7.

Proposed model of NE-induced ghrelin secretion. Basal ghrelin secretion is sustained by entry of Ca²⁺ into the ghrelin cell through VGCCs (A). Engagement of β1-adrenergic receptors (AR) by NE enhances Ca²⁺ entry through VGCCs and releases Ca²⁺ from the endoplasmic reticulum (ER) (B). Elevated cytosolic Ca²⁺ likely binds to synaptotagmin-7, which directs fusion of the ghrelin-containing vesicles with the plasma membrane, resulting in exocytosis and release of ghrelin. Activation of a cAMP/EPAC pathway downstream of β1-adrenergic receptors enhances the Ca²⁺-triggered ghrelin secretory pathway (C).