Ghrelin and des-acyl ghrelin inhibit cell death in cardiomyocytes and endothelial cells through ERK1/2 and PI 3-kinase/AKT

Ghrelin is an acyl-peptide gastric hormone acting on the pituitary and hypothalamus to stimulate growth hormone (GH) release, adiposity, and appetite. Ghrelin endocrine activities are entirely dependent on its acylation and are mediated by GH secretagogue (GHS) receptor (GHSR)-1a, a G protein–coupled receptor mostly expressed in the pituitary and hypothalamus, previously identified as the receptor for a group of synthetic molecules featuring GH secretagogue (GHS) activity. Des-acyl ghrelin, which is far more abundant than ghrelin, does not bind GHSR-1a, is devoid of any endocrine activity, and its function is currently unknown. Ghrelin, which is expressed in heart, albeit at a much lower level than in the stomach, also exerts a cardio protective effect through an unknown mechanism, independent of GH release. Here we show that both ghrelin and des-acyl ghrelin inhibit apoptosis of primary adult and H9c2 cardiomyocytes and endothelial cells in vitro through activation of extracellular signal–regulated kinase-1/2 and Akt serine kinases. In addition, ghrelin and des-acyl ghrelin recognize common high affinity binding sites on H9c2 cardiomyocytes, which do not express GHSR-1a. Finally, both MK-0677 and hexarelin, a nonpeptidyl and a peptidyl synthetic GHS, respectively, recognize the common ghrelin and des-acyl ghrelin binding sites, inhibit cell death, and activate MAPK and Akt. These findings provide the first evidence that, independent of its acylation, ghrelin gene product may act as a survival factor directly on the cardiovascular system through binding to a novel, yet to be identified receptor, which is distinct from GHSR-1a.

Ghrelin inhibits doxorubicin-induced cell death in vitro To investigate whether ghrelin may act as survival factor for cardiovascular cells, we have treated H9c2 cardiomyocytes with doxorubicin, which induces apoptotic cell death of cardiomyocytes both in vitro and in vivo ( Arola et al., 2000 ; Wu et al., 2000 ). After 24 h of treatment, 1 μM doxorubicin induced cell death of H9c2 cardiomyocytes as observed by phase–contrast microscopy ( Fig. 1 A). Dying cells became light refractive and detached off the plate. However, in the presence of a saturating concentration of ghrelin doxorubicin-treated H9c2 cells maintained their morphology and were still attached to the plate. In the absence of doxorubicin, ghrelin did not affect cell morphology. Ghrelin cytoprotective activity was also assayed by measuring cell viability of H9c2 cardiomyocytes. After 24 h of treatment with doxorubicin, 50% of the cells were not viable as measured by the 3-(4,5-dimethylthiazole-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. However, doxorubicin-induced cell death was almost completely abolished in the presence of a saturating concentration of ghrelin ( Fig. 1 B). The cytoprotective activity of ghrelin was concentration dependent, 0.1 nM being the lowest active concentration ( Fig. 1 C). Similar results were obtained when cell viability was measured as loss of membrane impermeability. Upon 48 h of treatment, doxorubicin induced lactate dehydrogenase (LDH) release in the medium, which was completely prevented in the presence of a saturating concentration of ghrelin ( Fig. 1 D). Similar results were obtained in PAE endothelial cells (unpublished data). Ghrelin alone did not affect cell survival of H9c2 cardiomyocytes as measured by either MTT or the release of LDH. Thus, these data provide the first indication that ghrelin acts directly on cardiomyocytes to inhibit experimentally induced cell death. Figure 1. Ghrelin protects H9c2 cardiomyocytes from doxorubicin-induced cell death. H9c2 cardiomyocytes were treated with 1 μM doxorubicin (grey bar) in the presence or absence of 1 μM ghrelin. (A) Phase–contrast images (200×; bar, 0.1 mm) after 24 h of treatment: untreated (a), ghrelin (b), doxorubicin (c), and doxorubicin + ghrelin (d). (B) Cell death measured by MTT assay after 24 h of treatment is expressed as the percentage of cell death. Values are the mean ± SE of eight samples (* t test, P < 0.01). (C) Dose response of ghrelin in MTT cell death assay after 24 h of treatment. Values are the mean ± SE of eight samples. (D) Cell death measured by LDH release after 48 h of treatment. Values are the mean ± SE of eight samples (* t test, P < 0.01).

To provide further biochemical evidence supporting the hypothesis that ghrelin may act as a survival factor, we have investigated ghrelin's ability to stimulate intracellular signaling pathways leading to inhibition of apoptosis. Survival factors, such as insulin growth factor-1 (IGF-1), inhibit apoptosis by stimulating several intracellular signaling pathways, including protein-tyrosine phosphorylation and activation of extracellular signal–regulated kinase (ERK)-1/2 and Akt ( Parrizas et al., 1997 ). In H9c2 cardiomyocytes, ghrelin induced tyrosine phosphorylation of two intracellular proteins of ∼100 and 60 kD, suggesting that it may activate a protein-tyrosine kinase ( Fig. 3 A). Similar results were obtained in PAE cells (unpublished data). Ghrelin activated both ERK1/2 and Akt in H9c2 cardiomyocytes ( Fig. 3, B and C ) and in endothelial cells (unpublished data). Activation of ERK1/2 and Akt, measured as serine residue-specific phosphorylation, occurred as early as 2 min after ghrelin stimulation, peaked between 5 and 10 min, and lasted at least 30 min. These data suggest that early activation of ERK1/2 and Akt may be involved in the transduction of the antiapoptotic signaling triggered by ghrelin. To verify such hypothesis, we have investigated whether pharmacological inhibition of either ERK1/2 or Akt impairs ghrelin cytoprotective activity after doxorubicin treatment. Activation of Akt was prevented by inhibition of phosphatidylinositol (PI) 3-kinase, whose enzymatic product PI(3,4,5)P 3 mediates growth factor–induced activation of Akt ( Alessi et al., 1996 ; unpublished data). Cell treatment of both H9c2 cardiomyocytes and PAE cells either with 30 μM PD98059, a highly specific ERK1/2 inhibitor, or 100 nM wortmannin, a highly specific inhibitor of PI 3-kinase, abolished ghrelin's cytoprotective activity after 20 h of treatment with 1 μM doxorubicin ( Fig. 3, D and E ). Under these conditions, neither PD98059 nor wortmannin affected doxorubicin-induced cell death. Thus, these data suggest that ghrelin inhibits cell death by activating at least two prosurvival signaling pathways conveyed by ERK1/2 and PI 3-kinase/Akt. Figure 3. Ghrelin stimulates protein-tyrosine phosphorylation and activates ERK1/2 and Akt. Overnight serum-starved H9c2 cells were stimulated with 1 μM ghrelin and lysed at different times. Total lysates were analyzed by Western blot with antiphosphotyrosine antibodies (A), specific anti-phosphoERK1/2 antibodies (B, top) and anti-ERK1/2 antibodies (B, bottom), and specific anti-phosphoAkt antibodies (C, top) and anti-Akt antibodies (C, bottom). H9c2 cardiomyocytes (D) and PAE cells (E) were treated with 1 μM doxorubicin for 20 h (black bars) in the presence or absence of 1 μM ghrelin and in the presence or absence of 100 nM wortmannin or 30 μM PD98059. Cell death was measured by MTT assay. Values are the mean ± SE of eight samples (* t test, P < 0.01).

Ghrelin is a gastric acyl-peptidic hormone acting on the hypothalamus and pituitary to induce GH release, food intake, and adiposity ( Kojima et al., 1999 ; Tschop et al., 2000 ; Nakazato et al., 2001 ). In addition to its endocrine activities, ghrelin and synthetic GHS protect heart function from experimentally induced cardiac heart failure in vivo ( Locatelli et al., 1999 ; Tivesten et al., 2000 ; King et al., 2001 ; Nagaya et al., 2001b ). However, the cellular and molecular mechanisms underlying ghrelin cardioprotective activity are not known. Based on these observations, we have raised the hypothesis that ghrelin may feature a direct cytoprotective activity toward myocardial cells by inhibiting cell death of cardiomyocytes and endothelial cells. In the heart, antiapoptotic factors, such as IGF-1 and cardiotrophin-1 (CT-1), play a crucial role in maintaining cardiomyocytes survival and myocardial function after ischemia- and pressure overload–induced cardiomyopathies ( Buerke et al., 1995 ; Hirota et al., 1999 ). The data reported here showing the ability of ghrelin to inhibit apoptosis induced by either doxorubicin serum withdrawal or FAS activation suggest that it may act as a survival factor in cardiomyocytes and endothelial cells. The data also supports the hypothesis that ghrelin may play a role in myocardial homeostasis. Survival factors inhibit cell death by activating specific signaling pathways, including stimulation of protein-tyrosine phosphorylation and activation of ERK-1/2 and PI 3-kinase/Akt, which then lead to the inhibition of the apoptotic signaling cascade ( Wu et al., 2000 ). We have shown that ghrelin stimulates tyrosine phosphorylation and activates ERK-1/2 and Akt and that activation of each of the two pathways is required for ghrelin-induced inhibition of apoptosis. Activation of ERK-1/2 by ghrelin was reported previously in HepG2 cells, which express GHSR-1, although in these cells ghrelin inhibits rather than activates Akt ( Murata et al., 2002 ). The receptor-mediating ghrelin intracellular signaling and antiapoptotic activity is not known. Here we provide three pieces of evidence suggesting that ghrelin antiapoptotic activity is not mediated by GHSR-1a. (i) In contrast to the whole heart, no expression of GHSR-1a was detected in H9c2 cardiomyocytes by RT-PCR. (ii) Both ghrelin and its deacylated form, des-acyl ghrelin, recognize a common high affinity binding site, although only ghrelin and not des-acyl ghrelin bind to GHSR-1a ( Kojima et al., 1999 ). Furthermore, affinity constant of ghrelin binding sites on H9c2 cardiomyocytes are ∼10-fold higher than affinity constant of GHSR-1a as measured on pituitary and hypothalamus membranes ( Muccioli et al., 2001 ). (iii) Des-acyl ghrelin, which is devoid of any GH-releasing activity even in vitro ( Kojima et al., 1999 ), is as effective as ghrelin in inhibiting apoptosis and in activating intracellular signaling pathways, i.e., ERK-1/2, Akt, and protein-tyrosine phosphorylation (unpublished data). Thus, these findings suggest that ghrelin and des-acyl ghrelin bind to a novel receptor, distinct from GHSR-1a. The putative novel receptor is expected to be highly similar to GHSR-1a, since it differs only in its lack of ability to discriminate between the esterified and unesterified ghrelin peptide. Moreover, the demonstration that two synthetic ligands of GHSR-1a, hexarelin and MK-0677, bind and activate the novel receptor provides further support to the hypothesis that the two receptors are highly related. Whether such receptor is encoded by alternative splicing of GHSR-1 gene or by a distinct gene still remains to be determined. GHSR-1b, a GHSR-1–truncated splicing variant, although expressed in the heart, does not appear to be functional ( Howard et al., 1996 ). We have reported previously that both ghrelin and des-acyl ghrelin bind to a common receptor in breast cancer cells ( Cassoni et al., 2001 ). In these cells, ghrelin and des-acyl ghrelin inhibit cell proliferation, whereas, on the other hand, GHSR-1a appears to mediate a proliferative signal ( Murata et al., 2002 ). Whether the same ghrelin receptor mediates the inhibition of both growth and apoptosis in breast cancer cells and in cardiomyocytes and endothelial cells, respectively, it is not determined. Although most survival factors also display a proliferative activity, members of the transforming growth factor-β family inhibit both cell growth and cell death in myoblasts ( Chen et al., 2001 ; Rios et al., 2001 ). In addition, the finding that GHSR-1a is expressed in the whole heart but not in H9c2 cardiomyocytes provides further support to the hypothesis that multiple ghrelin receptors may be expressed in the cardiovascular system ( Papotti et al., 2000 ; Bodart et al., 2002 ; Gnanapavan et al., 2002 ; Katugampola et al., 2001 ). Each receptor may then contribute independently to mediate the wide array of cardiovascular activities induced by both ghrelin and synthetic GHS

Cells were plated in 96-well slides (∼25,000–30,000 cells/cm 2 ) for MTT and LDH assays and in 10-cm dishes for other studies. Cell death was induced by incubating the cells with 1 μM doxorubicin in 5% FCS (H9c2) or 0% FCS (PAE) for the indicated time. FAS-mediated apoptosis of adult cardiomyocytes was obtained by 30-min cell incubation on ice in the presence of 1 μg/ml anti-FAS agonist antibodies (Bender MedSystem) and 1 μg/ml protein A (Sigma-Aldrich). Cells were then placed for 6 h at RT in tyrode medium as described previously ( Gallo et al., 2001 ). Cell viability was measured by MTT assay (Sigma-Aldrich) as described previously ( Filigheddu et al., 2001 ) and by Cytotoxicity Detection Kit (Roche), which measures LDH released in the medium. Percentage cytotoxicity was calculated according to the manufacturer's instructions. Apoptosis was evaluated in H9c2 and PAE cells by FACS ® analysis (Becton Dickinson) of the proportion of cells displaying shrunken/hypergranular morphology or those stained with Annexin-V Fluos (Alexis) above the untreated controls. Briefly, 0.5 × 10 5 cells were stained with annexin V (1:50) and 1 μg/ml propidium iodide in 10 mM Hepes-NaOH (pH 7.4), 140 mM NaCl, 5 mM CaCl 2 , and incubated for 10–15 min at RT in the dark. Data were analyzed by flow citometry software (Becton Dickinson). Apoptosis in adult cardiomyocytes was detected as nucleosomal DNA fragmentation measured as the percentage of sub-G1 cells. After treatment, cardiomyocytes were fixed with ice-cold 70% ethanol, stored at −20°C overnight, and then incubated for 10 min in phosphate-citric acid buffer at pH 7.8. Pelleted cells were then stained with 1 ml of 25 μg/ml propidium iodide, 400 mg/ml Rnase A, and 1% Triton X-100 at RT for 15 min, and analyzed by FACS ® . The percentage of sub-G1 cells was calculated by CellFit software (Becton Dickinson) and defined as apoptotic. Direct microscopical observation confirmed the presence of nuclear fragmentation and apoptotic bodies.

Total RNA was extracted by Rneasy mini kit (QIAGEN) from either cultured cells or rat heart and pituitary tissue mechanically triturated in liquid nitrogen and retrotranscribed (1 μg) in 20 μl of RT-buffer in the presence of 500 ng oligodT and 1 μl Powerscript reverse transcriptase. Reverse transcriptase reaction was performed at 42°C for 1.5 h and at 72°C for 15 min. One fifth of the reverse transcriptase product (4 μl of cDNA) was amplified in 50 μl PCR buffer (1× buffer, 1.5 mM MgCl 2 , 200 μM dNTP, 1 μM of each primer, 1 U DyNAzyme EXT taq). PCR was performed by a 5-min incubation at 95°C followed by 5 cycles (30 s at 94°C, 30 s at 55°C, and 3 min at 72°C) and 35 cycles (30 s at 94°C, 30 s at 60°C, and 3 min at 72°C). Primers sequences for GHSR-1a were the rat homologues to the human sequences described previously by Nagaya et al. (2001a) : 5′-CGGAGAGATGGGATGTGCTG-3′ (forward) and 5′-CTCTGCTGGCTGCCCTTCCA-3′ (reverse).