| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Laboratory of Molecular and Cellular Endocrinology (R.G., F.S., L.T., S.D., M.A.), Division of Endocrinology and Metabolism (R.G., F.S., L.T., S.D., M.A., E.G.), and Departments of Internal Medicine (R.G., F.S., L.T., S.D., M.A., E.G., L.B.), Anatomy, Pharmacology, and Forensic Medicine (F.C., M.M., C.G., G.M.), and Clinical and Biological Sciences and San Luigi Hospital (M.P.), University of Turin, 10126 Turin, Italy; Department of Medicine (R.N., F.B.), Transplant Unit, Scientific Institute San Raffaele, Vita-Salute University, 20132 Milan, Italy; and Department of Internal Medicine (J.I.), Sahlgrenska Academy, University of Göteborg, SE-40530 Göteborg, Sweden
Address all correspondence and requests for reprints to: Riccarda Granata, Ph.D., Laboratory of Molecular and Cellular Endocrinology, Division of Endocrinology and Metabolism, Department of Internal Medicine, University of Turin, Corso Dogliotti 14, 10126 Turin, Italy. E-mail: riccarda.granata{at}unito.it.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
/TNF-
, whose synergism is a major cause for ß-cell destruction in type I diabetes. HIT-T15 ß-cells expressed ghrelin but not ghrelin receptor (GRLN-R), which binds acylated ghrelin (AG) only. However, both unacylated ghrelin (UAG) and AG recognized common high-affinity binding sites on these cells. Either AG or UAG stimulated cell proliferation through Gs protein and prevented serum starvation- and IFN-/TNF--induced apoptosis. Antighrelin antibody enhanced apoptosis in either the presence or absence of serum but not cytokines. AG and UAG even up-regulated intracellular cAMP. Blockade of adenylyl cyclase/cAMP/protein kinase A signaling prevented the ghrelin cytoprotective effect. AG and UAG also activated phosphatidyl inositol 3-kinase (PI3K)/Akt and ERK1/2, whereas PI3K and MAPK inhibitors counteracted the ghrelin antiapoptotic effect. Furthermore, AG and UAG stimulated insulin secretion from HIT-T15 cells. In INS-1E ß-cells, which express GRLN-R, AG and UAG caused proliferation and protection against apoptosis through identical signaling pathways. Noteworthy, both peptides inhibited cytokine-induced NO increase in either HIT-T15 or INS-1E cells. Finally, they induced cell survival and protection against apoptosis in human islets of Langerhans. These expressed GRLN-R but showed also UAG and AG binding sites. Our data demonstrate that AG and UAG promote survival of both ß-cells and human islets. These effects are independent of GRLN-R, are likely mediated by AG/UAG binding sites, and involve cAMP/PKA, ERK1/2, and PI3K/Akt. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Unacylated ghrelin (UAG), the major circulating form of ghrelin, does not bind GRLN-R, is devoid of GH-releasing activity as well as of any effect on other anterior pituitary functions but is not an inactive peptide (3, 11, 12). Indeed, it was initially found able to exert some nonendocrine actions, sharing with AG the same protective effect on cardiac and endothelial cells (13) or antiproliferative effects on neoplastic cell lines (26). Therefore, these actions were supposed to be mediated by ghrelin receptor subtype(s) different from the type 1a. Later on, the bioactivity of UAG has been demonstrated also at the metabolic level. Indeed, UAG has been shown able to 1) counteract the inhibitory effect on insulin secretion and the hyperglycemic effect of AG in humans (27), 2) directly modulate glucose metabolism at the hepatic level as indicated by its ability to block both basal and AG-stimulated glucose output from pig hepatocytes (28), 3) negatively affect insulin secretion in ghrelin transgenic mice that overexpressed ghrelin in the pancreas (29), and 4) directly modulate adipocyte differentiation and proliferation as well as lipolysis (30, 31).
The hypothesis that the ghrelin system exerts a relevant role in metabolic balance and namely in the modulation of pancreatic ß-cell function has been further supported by more recent data. Indeed, ghrelin-producing cells have been discovered as a new islet 
-cell population (4, 18, 32). Interestingly, ghrelin expression in the endocrine pancreas of rodents and humans has been demonstrated to be even more abundant than in the stomach during fetal life (18, 33). This evidence suggested that the pancreas, more than the stomach, is the major source of ghrelin during fetal life and that the ghrelin system would have a paracrine/autocrine role in the regulation of hormonal secretion and/or islet growth and differentiation (4, 18, 32). Moreover, it has been recently shown that AG would prevent the development of diabetes in streptozotocin-treated newborn rats (25).
In the present study, we investigated the effect of either AG or UAG on proliferation and survival of pancreatic ß-cells. Apoptosis is the most critical final step in the development of autoimmune diabetes and an important factor contributing to ß-cell loss and the onset of type 2 diabetes (34, 35). Moreover, interferon (IFN)-/TNF- synergism is implicated in islet ß-cell destruction, both in vitro and in vivo, which results in type 1 diabetes (36) and also seems to be involved in early loss of islet mass in islet transplantation (37).
We provide evidence that both AG and UAG promote cell growth of pancreatic ß-cells. Moreover, they potently counteract apoptosis induced by serum starvation and/or cytokine synergism in either ß-cells or human islets of Langerhans. These effects are mediated by the cAMP/protein kinase A (PKA) pathway and by ERK1/2 and phosphatidyl inositol 3-kinase (PI3K)/Akt signaling. Indeed, this study shows that even endogenous ghrelin exerts an antiapoptotic effect in HIT-T15 cells and suggests that, besides its metabolic action in the pancreas, ghrelin directly regulates ß-cell fate in physiological and/or pathological conditions.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, NF-449, PD-98059, MDL-12330A, Hoechst-33258, and Mammalian Cell Lysis Kit for protein extraction and pertussis toxin were from Sigma-Aldrich (Milan, Italy). IFN- was provided by Euroclone (Celbio, Milan, Italy). Wortmannin and 3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyltetrazolium bromide (MTT) were provided from ICN Biomedicals (Aurora, OH). KT-5720 and 8-Br-cAMP were from Biomol Research Laboratory Inc. (Diagnostic Brokers Associated, Srl, Segrate, Milan, Italy). F12 (Ham) medium, horse serum, fetal bovine serum (FBS), penicillin streptomycin, fungizone, and trypsin were from Life Technologies, Inc. (Invitrogen, Milan, Italy). ECL reagent was from Perkin-Elmer Life Sciences Inc. (Boston, MA). P-Akt and P-ERK antibodies were from Cell Signaling Technology (Celbio, Milan, Italy). ERK1/2 and Akt antibodies were from Santa Cruz Biotechnology Inc. (Diagnostic Brokers Associated, Srl). Quantikine immunoassay caspase-3 kit and cAMP assay kit were provided from R&D Systems (Space Import-Export, Milan, Italy). ELISA kit for AG or UAG were provided by Linco Research (St. Charles, MO). Reagents for RT-PCR were from PE Applied Biosystem (Milan, Italy). Primers for RT-PCR were from TibMol Biol (Genoa, Italy).
Cell culture
Hamster HIT-T15 insulin-secreting cells, a widely used pancreatic ß-cell line (38), were obtained from the Istituto Zooprofilattico Sperimentale della Lombardia e dell’Emilia (Brescia, Italy). Cells were cultured under 5% CO2 at 37 C in F12 medium supplemented with 15% horse serum (HS), 2.5% FBS, 1% penicillin-streptomycin, 0.5% fungizone, and 10 µg/ml reduced glutathione. The glucose-sensitive clonal ß-cell line INS-1E (39) was kindly provided by Dr. Claes Wolheim (University of Geneva Medical Center, Geneva, Switzerland). Cells were maintained in RPMI 1640 medium (Sigma) supplemented with 10% fetal calf serum, 10 mM HEPES (pH 7.6), 1 mM sodium pyruvate, 2 mM glutamine, 50 µM ß-mercaptoethanol, 100 U/ml penicillin, and 100 µg/ml streptomycin and incubated at 37 C in a 5% CO2 and 95% air atmosphere. For experiments, cells were rinsed three times with serum-free medium and then plated in serum-free conditions with the various stimuli for the indicated times.
Islet isolation
Human islets of Langerhans were obtained from the pancreas of heart-beating multiorgan donors as described previously (40). Islet preparations with a purity of more than 70%, which could not be used for transplantation because of low total islet number, were used after approval by the local ethical committee. After isolation, islets (10,000) were cultured at 37 C in a humidified atmosphere (5% CO2), in CMRL medium (Invitrogen) supplemented with 10% FBS, 100 U/ml penicillin, 100 µg/ml streptomycin sulfate (Euroclone, Celbio), and 2 mM glutamine.
RT-PCR
Total RNA extraction from HIT-T15 cells and reversed transcription to cDNA from 3 µg RNA were performed as described (41). The primer sequences were the following: rat ghrelin, forward 5'-TTGAGCCCAGAGCACCAGAAA-3', and reverse 5'-AGTTGCAGAGGAGGCAGAAGCT-3' (NM_021 669) (42); rat GRLN-R, forward 5'-GTCGAGCGCTACTTCGC-3', and reverse 5'-GTACTGGCTGATCTGAGC-3' (NM_032075) (31); rat 18S rRNA, forward 5'-GTGGAGCGATTTGTCTGGTT-3', and reverse 5'-CGCTGAGCCAGTTCAGTGTA-3' (X01117). cDNA (9 µl) was amplified by PCR in a 50-µl volume with AmpliTaq Gold polymerase in a GeneAmp PCR System (Perkin-Elmer Italia, Applied Biosystems, Monza, Milan, Italy) under the following conditions: one cycle of 94 C for 30 sec, annealing for 30 sec, 72 C for 30 sec, and 72 C for 7 min for elongation. The annealing temperature was adjusted for each target: 61 C for ghrelin and 59 C for GRLN-R. The PCR products (347 bp for ghrelin, 492 bp for GRLN-R, and 201 bp for 18S RNA) were separated by 1.5% agarose gel electrophoresis and visualized by ethidium bromide staining. Positive control included hamster gastric mucosa for ghrelin and hamster pituitary cells for GRLN-R, whereas a negative control was carried out by using buffer alone with no RNA. Amplification for 18S rRNA subunit served as internal control for the RNA samples.
ELISA for AG and UAG detection
AG and UAG were measured in duplicate by ELISA (active ghrelin ELISA kit and des-acyl-ghrelin ELISA kit; Linco Research), using a mouse monoclonal antibody against the N terminus of either the acylated or des-acylated forms of the peptide. Ghrelin concentrations were assessed in culture medium of HIT-T15 cells after 48 h in the presence or absence of serum or IFN-/TNF- (100 and 200 ng/ml, respectively), according to the manufacturer’s instructions.
Immunohistochemistry for ghrelin and GRLN-R
HIT-T15 cells were grown on glass coverslips, fixed using methanol (5 min at –20 C) and acetone (5 sec at –20 C), and stained for ghrelin or GRLN-R using a standard immunoperoxidase procedure with streptavidin peroxidase (StrAviGen MultiLink kit, BioGenex Laboratories Inc., San Ramon, CA). For ghrelin detection, slides were incubated for 1 h at room temperature with a 1/600 dilution of a rabbit antirat polyclonal antibody that recognizes the C-terminal portion of both AG and UAG (Phoenix Pharmaceuticals). Slides were then incubated with a biotinylated antibody and subsequently with a solution of strepatavidin-peroxidase conjugate, as previously described (43). Finally, 3,3'-diaminobenzidine chromogen solution (LiquidDAB Substrate Pack; Biogenex) was used to reveal the final reaction product, and slides were counterstained in Mayer’s hematoxylin (Biogenex). For GRLN-R detection, slides were incubated for 1 h at room temperature with a 1/2000 dilution of a rabbit antirat GRLN-R antibody (Phoenix Pharmaceuticals). The immunoreactive proteins were visualized using streptavidin-peroxidase and diaminobenzidine. Hamster stomach and pituitary cells served as positive control for ghrelin and its receptor. Negative controls were prepared by replacing the first antibody solution with buffer.
Double immunostaining of human pancreatic islets for insulin and GRLN-R
The 4-µm-thick slides of formalin-fixed, paraffin-embedded human pancreatic islets were deparaffinized. After antigen retrieval, obtained with three 5-min passages in a microwave oven at 650 W in pH 6.0 citrate buffer, endogenous peroxidase activity was blocked with hydrogen peroxide and endogenous biotin activity was inhibited with avidin from egg albumen. The primary antibody to GHS-R (GRLN-R) was applied overnight at 4 C (purified rabbit antibody from Phoenix Pharmaceuticals; diluted 1/100). The immune reaction was amplified with biotinylated tyramide (diluted 1/5 in triphosphate saline for 15 min at room temperature) and revealed with red fluorochrome. The double staining was completed by antiinsulin antibody (monoclonal from Biogenex; diluted 1/40), incubated overnight at 4 C. The reaction was revealed with fluorescein-labeled goat antimouse green IgG. After mounting with 4',6-diamidino-2-phenylindole, the slides were observed under a fluorescence microscope (Olympus, type BX41TF) and pictures taken with different filters for green light at 524 nm and red light at 612 nm emission peaks.
AG and UAG radioiodination
Rat 125I-labeled [Tyr4] AG and 125I-labeled [Tyr4] UAG were radioiodinated (125I-labeled [Tyr4] AG with a specific activity of 1700–2000 Ci/mmol; 125I-labeled [Tyr4] UAG with a specific activity of 1800–2100 Ci/mmol) using a lactoperoxidase method by Amersham Biosciences (Milan, Italy) and used as ligands in the binding studies. Both ghrelin analogs have been reported to be reliable probes for labeling GHS-R in tissue or cell membranes (13, 26, 44) and to have in vivo and in vitro the same biological activities of the native peptides. 125I-labeled [Tyr4] AG retained the same GH-releasing activity of native AG in neonate rats (45) and 125I-labeled [Tyr4] UAG retained more than 90% of the antilipolytic activity as measured in rat adipocytes stimulated by isoproterenol (31). Both 125I-labeled [Tyr4] AG and 125I-labeled [Tyr4] UAG were used within 2 wk of iodination.
AG and UAG binding assay
Binding of radiolabeled AG and UAG to HIT-T15 membranes was carried out as previously described (13, 44). In preliminary experiments, the optimal binding conditions for HIT-T15 cells were found to be similar to those previously reported for other tissues or cells (cardiomyocytes and adipocytes), which specifically bound AG and UAG (13, 28, 39). For saturation binding studies, cell membranes (corresponding to 100 µg protein) were incubated in triplicate at 23 C for 2 h under constant shaking with increasing concentrations (0.035–6 nM) of 125I-labeled [Tyr4] AG or 125I-labeled [Tyr4] UAG in a final volume of 0.5 ml assay buffer (50 mM Tris-HCl, 2.5 mM EGTA, 0.002% bacitracin, 0.1% BSA, pH 7.4). Parallel incubations, where 2.0 µM unlabeled AG or UAG was also present, were used to determine nonspecific binding, which was subtracted from total binding to yield specific binding values. The binding reaction was terminated by rapid vacuum filtration over Whatman GF/B filters (Maidstone, Kent, UK) and individual filter units counted for membrane-bound radioactivity in a Packard -counter A5003. Saturation binding data were analyzed by Scatchard analysis, and the Hill equation and maximum binding capacity (Bmax) and dissociation constant (Kd) values were calculated using GraphPad Prism version 4 (GraphPad Software, San Diego, CA). Receptor binding competition studies were performed by incubating in triplicate HIT-T15 cell membranes (150 µg) with a fixed concentration (1 nM) of 125I-labeled [Tyr4] AG or 125I-labeled [Tyr4] UAG in the absence and in the presence of increasing concentrations (0.01 nM to 1 µM) of different competitors. Nonspecific ligand binding was determined by incubation of radioiodinated AG or UAG and membranes in the presence of 2 µM unlabeled AG or UAG, respectively. Specific binding of AG or UAG constituted 80–85% of total binding at 1 nM concentration of each radioligand. Data were plotted and curves fit using the GraphPad Prism software as described above, assuming that the binding was due to one binding site, thus allowing determination of the concentration of a competitor causing 50% inhibition of specific radioligand binding (IC50). Receptor binding competition studies were also performed by incubating membranes (corresponding to 100 µg protein) from human islets of Langherans with 1 nM human 125I-labeled [Tyr4] UAG in the absence and in the presence of a fixed concentration (100 nM) of unlabeled human UAG, AG, or hexarelin. The specificity of binding was also tested in the presence of a single concentration (100 nM) of insulin, glucagon, or somatostatin.
Cell viability assay
Cell viability was assessed by MTT as described previously (13). Cells were seeded on 96-well plates at a density of 5 x 103 cells per well. After treatments, cells were incubated with 1 mg/ml MTT for approximately 1 h. The medium was aspirated and the formazan product solubilized with 100 µl dimethylsulfoxide. Viability was assessed by spectrophotometry at 570 nm absorbance using a 96-well plate reader.
Bromodeoxyuridine (BrdU) incorporation assay
Cell proliferation was assessed using the BrdU incorporation ELISA (Roche). Cells were seeded on 96-well plates at a density of 5 x 103 cells per well in serum-containing medium until 60–70% confluent and subsequently serum starved for 24 h. Cells were then treated with the various stimuli and incubated for another 24 h. Briefly, cells were incubated with BrdU labeling solution for 2 h at 37 C. After removal of the labeling solution, cells were fixed and denatured and incubated for 90 min with anti-BrdU antibody conjugate, which was subsequently removed by rinsing three times. Finally, cells were incubated in substrate solution at room temperature and proliferation assessed by colorimetric detection.
Hoechst staining of apoptotic cells
Morphological changes in the nuclear chromatin of apoptotic cells were detected by Hoechst 33258 staining. HIT-T15 or INS-1E cells, harvested by PBS-EDTA were pooled with the cells from conditioned medium, fixed with 4% formaldehyde in PBS for 15 min at 4 C, washed, resuspended in 70% ethanol, and stored at –20 C until use. Cells were then washed twice in PBS and stained in 50 µl PBS containing 10 µg/ml Hoechst 33258. After 15 min incubation at room temperature, a 15-µl aliquot was placed on a glass slide and 500 stained nuclei were double counted under a fluorescence microscope (4',6-diamidino-2-phenylindole filter).
Caspase-3 activity assay
The caspase-3 amount was measured in duplicate by Quantikine immunoassay (ELISA) active caspase-3 kit (R&D Systems, Minneapolis, MN). Briefly, harvested cells from incubations with or without serum, AG, and UAG were centrifuged and washed twice with PBS 1x and then incubated with biotin-ZVKD-fluoromethylketone inhibitor, which covalently binds and makes detectable the active caspase-3 only. A monoclonal antibody specific for caspase-3 was precoated onto a microplate, and active protease concentrations were assessed in cell lysates, according to the manufacturer’s instructions.
Western blot analysis
For Akt and ERK immunoblots, equal amounts of proteins (50 µg) from control and treated cells were resolved in 12% SDS-PAGE, as described previously (41). Proteins were transferred to a nitrocellulose membrane after blocking with 1% BSA in Tris-buffered saline with 0.1% Tween for 2 h at room temperature. Membranes were incubated overnight at 4 C with one of the specific antibodies (P-Akt, Akt, P-ERK, and ERK) (1:1000 dilution). The immunoreactive proteins were visualized using horseradish-peroxidase-conjugated goat antimouse or goat antirabbit (1:5000) by enhanced chemiluminescence.
Measurement of intracellular cAMP levels
HIT-T15 and INS-1E cells were seeded at a density of 5 x 105 cells into 100-mm dishes, serum starved for 24 h, and subsequently incubated for another 24 h in serum-free medium with or without the appropriate stimuli in the presence of 100 µM 3-isobutyl-1-methylxanthine (IBMX). After incubations, medium was removed and cells were lysed with ice-cold 0.1 N HCl. Cell lysates were centrifuged for 10 min at 800 rpm, and cAMP in the cell lysates was measured using a Quantikine immunoassay cAMP kit (R&D Systems), according to the manufacturer’s instructions.
Insulin secretion
HIT-T15 cells were plated at density of 5 x 105 cells into 100-mm dishes and serum starved for 24 h before incubation for 1 h at 37 C in HEPES-buffered Krebs-Ringer bicarbonate buffer containing 0.5% BSA with 1.25 mM glucose. The medium was changed, and the cells were incubated again for 1 h in Krebs-Ringer bicarbonate buffer/0.5% BSA containing 1.25, 7.5, or 15 mM glucose. After acid ethanol extraction of the hormone, secreted insulin was quantitated by a RIA kit (Linco Research, Labodia, Yens, Switzerland) that recognizes human insulin and cross-reacts with rat insulin.
Nitrite production
HIT-T15 and INS-1E cells, plated at density of 5 x 105 cells into 100-mm dishes, were starved for 24 h before incubation in serum-free medium with or without AG, UAG, or cytokines for 24 h. Nitrite production was measured from concentrated (18-fold) conditioned medium, using a nitric oxide quantitation kit (total nitric oxide assay kit; Assay Designs, Ann Arbor, MI) (46). The principle is based on the enzymatic conversion of nitrate to nitrite by the enzyme nitrate reductase. Nitrite is then quantified by the addition of Griess reagent, which converts it into a purple-colored azo compound. This absorbs visible light at 540 nm and can be detected by photometric measurement and referred to a standard curve.
Statistical analysis
The statistical significance in the data was assessed by ANOVA, and P values were calculated using unpaired Student’s t test or Newman-Keuls multiple-range test. Data are presented as mean ± SEM. Significance was established when P < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
shows that ghrelin immunoreactivity was detected in HIT-T15 ß-cells. Moreover, these cells were analyzed for their capacity of releasing ghrelin. ELISA experiments demonstrated that either AG or UAG were secreted by HIT-T15 cells (185 and 242 pg/ml, respectively) after 48 h incubation in complete medium (Fig. 1B). Moreover, RT-PCR analysis showed that ghrelin mRNA was also expressed (Fig. 1C). With respect to GRLN-R, we failed to identify its expression in HIT-T15 cells, either at the protein or at the mRNA level (Fig. 1, D and E). These data suggest that HIT-T15 cells express ghrelin mRNA and peptide but not GRLN-R.
|
). Scatchard analysis of the specific binding data (Fig. 2B) and Hill equation, with a slope close to 1, suggested the existence of a single class of binding sites, with a Kd of 0.49 ± 0.07 nM and a Bmax of 15.9 ± 1.6 fmol/mg protein (mean ± SEM of four independent experiments). Unlabeled AG, [Tyr4] AG, hexarelin, a synthetic peptidyl GHS, and even UAG competed with 125I-labeled [Tyr4] AG for HIT-T15 binding sites (Fig. 2C). The IC50 values, all expressed as nanomolar concentrations, were 3.1 ± 0.3 for AG, 3.8 ± 0.6 for [Tyr4] AG, 2.3 ± 0.4 for UAG, and 20.8 ± 2.3 for hexarelin (mean ± SEM of three independent experiments). The results of this competition binding study also revealed that the binding of 125I-labeled [Tyr4] AG to HIT-T15 cell membranes was specific and was not inhibited by some peptides structurally unrelated to AG, such as somatostatin, insulin, or glucagon.
|
). Scatchard analysis (Fig. 2E) demonstrated the existence of a single class of binding sites that showed values of Bmax (13.9 ± 0.8 fmol/mg protein) and Kd (0.68 ± 0.10 nM, mean ± SEM of four independent experiments), very close to those detected by radiolabeled AG. Unlabeled UAG and AG, as well as [Tyr4] AG and hexarelin, but not somatostatin, insulin, or glucagon competed with 125I-labeled [Tyr4] UAG for binding sites (Fig. 2F). IC50 values, all expressed as nanomolar concentrations (mean ± SEM of three independent experiments) were 2.8 ± 0.3 for UAG, 4.3 ± 0.5 for AG, 3.7 ± 0.4 for [Tyr4] AG, and 18.0 ± 2.1 for hexarelin. These results suggest that UAG and AG recognize common high-affinity binding sites on HIT-T15 cell membranes.
AG and UAG induce HIT-T15 cell proliferation through the Gs signaling pathway
Based on evidence of UAG- and AG-specific binding sites, we investigated the effect of ghrelin on HIT-T15 cell proliferation. Cells were incubated in serum-free medium in the presence or absence of increasing concentrations, ranging from 1 pM to 1 µM (10–12 to 10–6 M) of either AG or UAG for 24 h. BrdU incorporation assay showed that both peptides significantly and dose-dependently induced cell proliferation (Fig. 3A). The efficacy of cell growth stimulation was within 1 nM to 1 µM, equal to the one that was found effective in displacing radiolabeled AG or UAG from HIT-T15 binding sites. This effect was similar to that observed in cells cultured in normal conditions, i.e. in the presence of serum (15% HS, 2.5% FBS).
|
s protein-coupled receptor antagonist (47). This resulted in complete blockade of AG- and UAG-induced cell proliferation, whereas pretreatment with pertussis toxin (PTX; 50 ng/ml), an inhibitor of Gi/o protein-coupled receptor, had no effect (Fig. 3B).
Fig. 3C is a representative phase contrast image showing that both AG and UAG counteract HIT-T15 ß cell loss in serum deprived conditions by increasing the number and size of islet-like structures, with respect to untreated cells. Taken together, these results suggest that AG and UAG promote ß cell proliferation, likely involving the Gs signaling pathway.
AG and UAG prevent apoptosis induced by both serum starvation and IFN-/TNF- synergism
Apoptosis is the main form of pancreatic ß-cell death in animal models of type 1 diabetes mellitus (48). IFN-/TNF- synergism has been shown to play an important role in autoimmune diabetes in vivo as well as ß-cell apoptosis in vitro (36). On the basis of the results showing that ghrelin promotes HIT-T15 cell proliferation, we aimed to examine whether either AG or UAG inhibited apoptosis induced by serum deprivation or by IFN-/TNF- synergism. Hoechst 33258 staining showed that after 48 h, cells cultured in the presence of serum were round shaped, formed islet-like structures, and had very low apoptotic rate (
2%) (Fig. 4A inset, upper panel, and 4B). In serum-deprived medium, apoptosis increased up to approximately 20% and cells displayed typical chromatin condensation and nuclear fragmentation. Moreover, they partially lost their capacity to form islet-like structures (Fig. 4A, upper panel, and 4B). Cytokines further increased apoptosis (27%), cells appearing smaller and unable to form islets (Fig. 4A, lower panel, and 4B).
|
AG and UAG, at 100 nM, almost completely prevented serum-starvation-induced apoptosis and restored islet-like structures (Fig. 4A, upper panel, and 4B). Similarly, AG and UAG equally and significantly reduced apoptosis (12%) triggered by the IFN-/TNF- combination and induced cell enlargement and small islet formation (Fig. 4A, lower panel, and 4B).
The ghrelin antiapoptotic effect was dose dependent, 1 nM (10–9 M) being the lowest significantly active concentration of both peptides (data not shown). Furthermore, caspase-3 activation in both serum-starved and cytokine-treated cells was significantly reduced by either AG or UAG (Fig. 4C), providing additional evidence of their antiapoptotic effect in HIT-T15 pancreatic ß-cells.
Indeed, cytokines strongly decreased cell proliferation, and unexpectedly, both AG and UAG dramatically restored cell proliferation up to rates that were even higher than those observed in the presence of serum (Fig. 4D). Ghrelin effect on cell survival was also investigated by MTT assay in both serum-free conditions and in the presence of cytokines. The results of these experiments indicated that either AG or UAG equally and significantly increased cell viability under both experimental conditions (data not shown).
Herein, we previously showed that HIT-T15 cells express and release AG and UAG, suggesting that they may act through autocrine/paracrine mechanisms. To investigate this possibility, AG and UAG secretion was measured in cells cultured in the presence of serum and in serum-free medium alone or with addition of IFN-/TNF-. Figure 4E shows that both AG and UAG levels were significantly reduced in serum-starved cells and even more after exposure to cytokines. Surprisingly, addition of a specific antighrelin antibody with equal specificity for AG and UAG not only increased serum-starvation-induced apoptosis but also induced apoptosis in cells cultured in the presence of serum, suggesting that endogenous ghrelin may exert autocrine/paracrine action on cell survival. As expected, no effect was observed in cytokine-induced apoptosis, where ghrelin secretion is likely too low to counteract such a strong cell death increase (Fig. 4F).
Together, these results show that AG and UAG counteract apoptosis induced by serum starvation and IFN-/TNF- combination in HIT-T15 cells. Moreover, they strongly suggest that even endogenous ghrelin exerts cytoprotective effects, likely via autocrine/paracrine mechanisms.
cAMP/PKA pathway mediates ghrelin proliferative and antiapoptotic effects
cAMP and its principal target, the cAMP-dependent PKA, play important roles in mammalian cell proliferation and apoptosis (51). Elevation of intracellular cAMP levels has been shown to promote cell growth and to delay apoptosis in different cell types, including pancreatic ß-cells (52, 53). Our previous results showed that ghrelin-induced HIT-T15 cell proliferation involves the Gs protein-coupled receptor, which, in turn, has been shown able to activate cAMP/PKA signaling (51); therefore, we investigated whether the proliferative and antiapoptotic effects of both AG and UAG are mediated by this pathway. Initially, we examined AG- and UAG-induced cAMP intracellular variation. Figure 5A shows that incubation of HIT-T15 cells with both peptides, in the presence of the phosphodiesterase inhibitor IBMX, resulted in time-dependent changes of cAMP levels. AG and UAG produced a transient increase within 5 min, which was lower but still significantly above basal level at 10 and 30 min, declining thereafter toward the resting level after 60 min incubation. cAMP induction by ghrelin was then evaluated at 5 min in either serum-free medium alone or with addition of IFN-/TNF- in the presence of IBMX. Results showed that AG and UAG significantly up-regulated cAMP not only in serum-free medium alone but also after incubation with cytokines that, per se, reduced cAMP levels (Fig. 5B).
|
) or cytokine-induced (Fig. 5D) apoptosis, whereas no effect was found using inhibitors alone. Accordingly, AG and UAG antiapoptotic action was similar to that exerted by a cell-permeable cAMP analog, 8-Br-cAMP (Fig. 5, C and D). Similar results were obtained for cell proliferation and cell survival assays (BrdU and MTT, respectively) (data not shown). These results strongly suggest that AG and UAG proliferative and antiapoptotic effect in HIT-T15 pancreatic ß-cells involve the cAMP/PKA signaling pathway.
AG and UAG inhibit apoptosis by activating ERK1/2 and PI3K/Akt signaling
It has been recently demonstrated that in cardiomyocytes and endothelial cells, ghrelin inhibits apoptosis and activates Akt and ERK1/2 survival signaling pathways (13). Thus, we investigated the ability of ghrelin to activate Akt and ERK1/2 in HIT-T15 cells. Figure 6A shows that AG and UAG equally caused a rapid activation (5 min) of Akt phosphorylation that lasted at least 30 min, decreasing thereafter (data not shown). Moreover, either AG or UAG activated ERK1/2, peaking at 15 min and decreasing thereafter. Indeed, ERK1/2 phosphorylation appeared stronger with UAG (Fig. 6B). These results suggested that ghrelin antiapoptotic effect in HIT-T15 cells would be mediated by PI3K/Akt and ERK1/2 signaling. To verify this hypothesis, we investigated whether AG and UAG prevented apoptosis in the presence of specific inhibitors of either Akt (wortmannin) or ERK1/2 (PD98059) (13). AG- and UAG-induced Akt and ERK1/2 phosphorylation was inhibited by wortmannin and PD98059, respectively, whereas no effect was observed using these compounds alone (Fig. 6, C and D). Moreover, wortmannin and PD98059 abolished AG and UAG cytoprotective activity against either serum-starvation- or IFN-/TNF--induced apoptosis, whereas, in agreement with previous studies (56), no effect was observed using inhibitors alone (Fig. 6, E and F). Similar results were obtained for cell survival and cell proliferation assays (data not shown).
|
AG and UAG stimulate insulin release from HIT-T15 cells
Besides stimulating mitogenesis in most experimental ß-cell models, cAMP elevation as well as ERK1/2 phosphorylation have been linked to increased insulin secretion (57, 58). On the basis of our results showing that ghrelin proliferative and antiapoptotic effects are mediated by these pathways, we investigated whether ghrelin would stimulate insulin secretion. HIT-T15 insulin-producing cells (38) were incubated for 1 h in the presence of increasing concentrations of glucose (1.25–15 mM) in the absence or presence of 100 nM AG or UAG. Figure 7 shows that UAG but not AG significantly increased insulin secretion at low glucose concentration (1.25 mM). Moreover, both peptides stimulated insulin release at 7.5 mM glucose. Surprisingly, this stimulatory effect was maintained even at high glucose concentrations (15 mM), UAG being more potent than AG at raising insulin levels. These results indicate that ghrelin, either in its acylated or unacylated form, stimulates glucose-induced insulin secretion from HIT-T15 cells.
|
, A and B, shows that both AG and UAG, at the concentration used for HIT-T15 cells (100 nM), stimulated cell proliferation and inhibited apoptosis of INS-1E cells in either serum-starved conditions or after addition of IFN-/TNF-.
|
). Moreover, ghrelin-induced proliferation (Fig. 8D) in serum-starved conditions was blocked by specific inhibitors of PKA as well as of Akt and ERK1/2 phopshorylation, suggesting that besides the cAMP/PKA pathway, PI3K/Akt and MAPK signaling are also involved. Similar results were obtained in the presence of cytokines and also confirmed by Hoechst 33258 apoptotic staining (data not shown). Indeed, AG rapidly stimulated either ERK1/2 or PI3K/Akt phosphorylation (Fig. 8, E and F). UAG showed similar stimulatory effects (data not shown). These results demonstrate that ghrelin effects are not only restricted to the HIT-T15 ß-cell line, which lacks the GRLN-R, but also can be equally observed in other ß-cell lines such as INS-1E that, conversely, do express this receptor.
AG and UAG reduce cytokine-induced nitric oxide production from both HIT-T15 and INS-1E ß-cells
Proinflammatory cytokines such as IL-1ß, TNF-, and IFN- are known to induce generation of NO and/or reactive oxygen species that ultimately culminate in ß-cell dysfunction and death (59). Therefore, ghrelin effects on cytokine-induced NO production were investigated in both HIT-15 and INS-1E cells. Figure 9 shows that nitrite levels, which are stable end products of NO production, increased in either HIT-15 (Fig. 9A) or INS-1E cells (Fig. 9B) incubated with IFN-/TNF- for 24 h. Addition of AG and UAG significantly reduced nitrite production in both cell lines. Moreover UAG, but not AG, was able to reduce nitrite formation from HIT-T15 cells under basal conditions, i.e. serum-free medium without cytokines for 24 h. These results suggest that either AG or UAG may exert protective effects against cytokine-induced ß-cell death through inhibition of inducible NO synthase (iNOS) activity and NO formation.
|
, TNF-, and IL-1ß). Hoechst staining showed a reduction in islet cell number and increase in apoptotic fragmented nuclei in both serum-starved conditions and after exposure to cytokines. Both AG and UAG treatment increased the number of cells and restored their round-shaped aspect (Fig. 10A). Indeed, AG and UAG significantly increased cell survival (Fig. 10B) and prevented caspase-3 activation (Fig. 10C) in both experimental conditions and, particularly, in islets exposed to cytokines.
|
Either GRLN-R or AG- and UAG-specific binding sites are present in human islets of Langerhans
After the results showing that both AG and UAG displayed protective effects in human islet cells, we aimed to investigate the presence of GRLN-R and specific AG or UAG binding sites in human islets. Figure 11A shows that ghrelin receptor was abundantly and widely expressed in pancreatic endocrine islets that were also positive for insulin staining. Interestingly, double immunostaining for both GRLN-R and insulin showed that GRLN-R-like immunoreactive cells colocalized with some insulin-positive ß-cells, as shown by the overlapping of the two fluorochromes specific for either GRLN-R (red) or insulin (green). Furthermore, competition studies showed that both unlabeled UAG and AG, as well as hexarelin, but not insulin, glucagon, or somatostatin competed with 125I-labeled [Tyr4] UAG for binding sites on human islet membranes (Fig. 11B). These results suggest that GRLN-R as well as specific binding sites for both AG and UAG are expressed on human islets. Moreover, GRLN-R is at least partly located in insulin-positive ß-cells.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Survival of pancreatic ß-cells is obviously of major importance for maintaining normal glucose metabolism. Apoptosis of pancreatic ß-cells is a critical step in the development of type 1 diabetes (34), but ß-cell growth and survival are critical also in type 2 diabetes (35). Inflammatory cytokines, including IFN-, TNF-, and IL-1ß are strongly implicated in pancreatic islet ß-cell death and functional loss during autoimmune diabetes and also seem to be involved in early loss of islet mass in islet transplantation (37). Indeed, recent findings indicated that IFN-/TNF- synergism is responsible for autoimmune diabetes in vivo as well as ß-cell apoptosis in vitro (36). Thus, blockade of apoptosis through inhibition of cytokines could be a novel approach for protecting pancreatic islets from either cellular or noncellular immune-mediated destruction (60). Interestingly, ghrelin was recently shown to exert inhibitory effects on expression and production of inflammatory cytokines (15) as well as to attenuate the development of acute pancreatitis in rats by reducing inflammatory infiltrates of pancreatic tissues (61). Taking into account that ghrelin-secreting -cells have been discovered as a new pancreatic islet cell population and that the pancreas, more than the stomach, seems the major source of ghrelin during fetal life (4, 18, 32), the hypothesis that the ghrelin system would have a paracrine/autocrine role in the regulation of ß-cell survival and function appeared reasonable.
Herein, we show that serum starvation, a widespread working model for apoptosis in different cell types (41, 62, 63) and to a greater extent IFN-/TNF- synergism, induced apoptosis in HIT-T15 pancreatic ß-cells. This is a glucose-responsive and insulin-secreting hamster cell line, which retains most, if not all, the differentiated functions characteristic of ß-cells, representing a good in vitro model for studying ß-cell physiology and regulation (38). Induction of apoptosis in HIT-T15 cells was confirmed by results showing increased activity of caspase-3, an essential effector molecule for carrying out programmed cell death in eukaryotic cells (64).
Exogenous ghrelin, in either its acylated or unacylated form, not only stimulated cell proliferation but also suppressed apoptosis and potently reduced caspase-3 activation.
Ghrelin immunoreactivity in HIT-T15 ß-cells was detected both at the mRNA and the protein level, and either AG or UAG were secreted in serum-free supernatants, suggesting autocrine/paracrine action of endogenous ghrelin in the prevention of apoptosis. Importantly, ghrelin secretion was reduced in serum-starved cells with respect to cells cultured with serum. Moreover, this reduction was further enhanced after incubation with IFN-/TNF-, suggesting the existence of a negative association between ghrelin secretion and cell death.
The possibility of ghrelin autocrine action was confirmed by results showing that antibody against both AG and UAG significantly increased apoptosis in normal culture conditions and even more during serum starvation. In the presence of cytokines, we found little and nonsignificant effect of antighrelin antibody, likely because of the high apoptotic rate and the concomitant reduced AG and UAG secretion, which per se would not be sufficient for counteracting cell death. Notably, in these experiments we could not distinguish between the survival effect of endogenous AG and UAG because the antibody was directed against both. However, our results showing that AG and UAG had similar action on cell growth and survival and even activated identical intracellular signaling pathways strongly suggested that endogenous ghrelin exerts autocrine/paracrine cytoprotection in either its acylated or unacylated form. In keeping with these data, others have shown that ghrelin exerts an autocrine effect in cell types that do not express GRLN-R (65, 66).
In contrast to pancreatic islets and other ß-cell lines (1, 18, 21, 24, 67), we found no GRLN-R expression in HIT-T15 cells, suggesting that ghrelin proliferative and antiapoptotic action in these cells is mediated by a distinct and yet unidentified receptor. Anyway, independently of the presence of GRLN-R that is by definition bound by AG only (1), in HIT-T15 cells, AG and UAG recognized a common high-affinity binding site. Evidence that UAG, although devoid of GH-releasing activity (3), is as effective as AG in promoting cell growth, inhibiting apoptosis and activating cell survival signaling pathways, points toward the existence of an unknown receptor distinct from the classical GRLN-R. Accordingly, previous studies have shown that UAG, like AG, not only prevents apoptosis in cardiomyocytes and endothelial cells (13) but also exerts antiproliferative activity in neoplastic cell lines (26). Actually, these variable effects on cell survival, observed in the absence of GRLN-R, suggest that multiple ghrelin receptors are expressed in different cell types, triggering alternative signaling pathways that would lead to different responses. Indeed, the results of our study suggested that either the AG or UAG proliferative effect in HIT-T15 ß cells is mediated by a Gs protein-coupled receptor, because specific Gs but not Gi/o inhibitors prevented ghrelin-induced cell growth.
AG binding to GRLN-R has been shown to activate intracellular second messengers coupled to Gq/11, leading to phospholipase C and PKC activation and calcium mobilization (5). However, GRLN-R also activates distinct systems of second messengers, including cAMP/PKA and extracellular Ca2+, suggesting that this receptor couples to different intracellular signaling pathways, depending on the cell type and on the binding agonist (68, 69). To date, this is the first study to demonstrate that both AG and UAG activate cAMP/PKA signaling via a receptor that is not GRLN-R, even though others have previously shown cAMP stimulation by AG in cell lines that do not express GRLN-R (66).
cAMP is an important second messenger that either promotes cell growth or inhibits apoptosis, depending on the cell type and the triggering stimulus (51, 52, 53). Through its principal target, the cAMP-dependent PKA, cAMP induces cell proliferation by activating the ERK cascade in different cell types, including pancreatic ß-cell lines (51). Moreover, cAMP mediates antiapoptotic effects in pancreatic ß-cells via PKA, MAPK, and PI3K signaling (53, 58, 70). Notably, cAMP-mediated cell proliferation has been found induced by hormonal activation of Gs, which then stimulates effector molecules such as AC (51). These data, together with our results showing that AG and UAG increase intracellular cAMP and activate ERK1/2 phosphorylation, strengthen the hypothesis that in HIT-T15 ß-cells, ghrelin signals through Gs. Indeed, we show here that both AG and UAG up-regulated cAMP, not only during apoptosis induced by serum-free medium alone but also after addition of cytokines that, per se, increased cell death and reduced cAMP levels. Furthermore, AG and UAG effects on cell growth and protection from apoptosis, challenged with serum-starved medium or IFN-/TNF-, were abolished by MDL12330A (a specific inhibitor of AC) and KT5720 (a cAMP-dependent PKA inhibitor) (54, 55). These results provided additional evidence that cAMP is a positive mediator of ghrelin proliferative and antiapoptotic effects in HIT-T15 ß-cells. Accordingly, the cell-permeable cAMP analog 8-Br-cAMP prevented apoptosis to an extent that was similar to that seen with ghrelin. These findings demonstrate that cAMP/PKA pathway mediates AG- and UAG-induced ß-cell survival.
Moreover, inhibition of both AG- and UAG-induced Akt and ERK1/2 phosphorylation with wortmannin and PD98059, respectively, resulted in complete blockade of ghrelin antiapoptotic and proliferative effect, strongly suggesting the involvement of PI3K and MAPK activation in ghrelin signaling. Indeed, in agreement with previous studies showing AG and UAG cytoprotection either in the presence or absence of GRLN-R (13, 71, 72, 73), herein we provide evidence that ghrelin activates ERK1/2 and PI3K/Akt phosphorylation. Depending upon the stimulus and cell type, these pathways can transmit signals that result in the prevention of apoptosis or induction of cell cycle progression; moreover, they can cross-regulate one another to control both apoptosis and cell growth (13, 41, 74). Whether Akt and ERK activation is mediated by ghrelin-induced cAMP elevation or by alternative pathways remains to be elucidated.
Elevation of cAMP has been associated with increased insulin secretion in insulin-producing ß-cells (75); therefore, we suspected that either AG or UAG would increase insulin secretion from HIT-T15 glucose-responsive cells. Both AG and UAG were found equally able to induce insulin secretion in HIT-T15 cell after glucose stimulation. Because of the absence of GRLN-R and the existence of specific binding sites for AG and UAG, we suggest that AG- and UAG-induced insulin secretion in HIT-T15 cells is mediated by a different receptor that likely binds both peptides.
This study was mainly performed on a ß-cell model (HIT-T15) that, in our hands, did not express GRLN-R but that showed high-affinity binding sites for both AG and UAG. This implies that ghrelin effects were mediated by an unknown receptor, different form GRLN-R. The obvious question was whether ghrelin, in both acylated and unacylated form, would promote survival of ß-cells that expressed GRLN-R. To unravel this issue, we performed experiments on INS-1E ß-cells (39). These were an attractive model as they were previously shown to express GRLN-R (21, 24) as well as to positively respond to both AG- and UAG-induced insulin secretion (24), in accordance with our results in HIT-T15 cells. Indeed, as for our previous cell model, we found that both AG and UAG could either promote proliferation or preserve INS-1E cells from apoptosis induced by serum starvation or by cytokine combination. Even the mechanisms underlying ghrelin effects appeared to be the same, i.e. cAMP/PKA and PI3K/AKT, ERK1/2 pathways.
We did not perform binding studies on INS-1E cells. Theoretically AG would have exerted its action via the classical GRLN-R; however, the same acti
