Ghrelin, the only known peripherally produced and centrally acting peptide that stimulates food intake, is synthesized primarily in the stomach and acts through the growth hormone secretagogue receptor (GHS-R1a). In addition to its orexigenic effect, ghrelin stimulates the release of growth hormone (GH). In this study, we investigated the biological properties of full-length and shortened ghrelin analogs in which octanoylated Ser3 is replaced with an octanoic acid moiety coupled to diaminopropionic acid (Dpr). Ghrelin analogs stabilized with Dpr(N-octanoyl) in position 3 and noncoded amino acids in position 1 (sarcosine) and/or position 4 (naphthylalanine or cyclohexylalanine) were found to possess affinities similar to those of ghrelin for cell membranes with transfected GHS-R1a. In vivo, the prolonged orexigenic effects of analogs containing Dpr(N-octanoyl)3 compared with that of ghrelin in adult mice and a similar impact on GH secretion in young mice were found. Full-length [Dpr(N-octanoyl)3]ghrelin and its analogs with a noncoded amino acid in position 1 and/or 4 showed significantly prolonged stability in blood plasma compared with that of ghrelin. Ghrelin analogs with a prolonged orexigenic effect are potential treatments for GH deficiency or cachexia that accompanies chronic diseases. Desoctanoylated ghrelin analogs and N-terminal penta- and octapeptides of ghrelin did not show any biological activity.
Ghrelin is the only known hormone synthesized in the gut and acting centrally that possesses an orexigenic effect (for reviews, see Depoortere, 2009; Castañeda et al., 2010; Wisser et al., 2010; Briggs and Andrews, 2011; Nass et al., 2011; Scerif et al., 2011). In the adenohypophysis, ghrelin releases Ca2+ from intracellular stores that stimulates and amplifies pulsatile secretion of growth hormone (GH) (Smith et al., 1997). Therefore, the term growth hormone secretagogue receptor (GHS-R1a) is used for the ghrelin receptor. In the arcuate nucleus of the hypothalamus, ghrelin implements its orexigenic effect in neurons that express the most powerful neuropeptide, orexigenic neuropeptide Y (Kohno et al., 2003).
Ghrelin is the only hormone containing a serine acylated with n-octanoic acid, which is necessary for its biological activity (Kojima et al., 1999). The ester bond of Ser3 is hydrolyzed easily by esterases in the blood; therefore, only 10 to 20% of circulating ghrelin is octanoylated (Depoortere, 2009). Although octanoylation is necessary for ghrelin's biological activity, the link between octanoic acid and ghrelin can be modified. When octanoic acid was connected to the peptide backbone of ghrelin by an amide (Bednarek et al., 2000), thioether, or ether bond (Matsumoto et al., 2001a), neither the binding potency nor the calcium release in the cells with transfected GHS-R1a was affected.
In vitro structure-activity studies in cells that overexpress the ghrelin receptor reveal that an N-terminal tetrapeptide with an octanoyl group in position 3 is the minimal core of ghrelin that is necessary for biological activity (Bednarek et al., 2000; Matsumoto et al., 2001a,b; Torsello et al., 2002). An alanine scan of 14 amino acids at the N terminus of ghrelin showed that, in addition to octanoylation at position 3, an N-terminal positive charge and Phe4 are essential for the biological activity of ghrelin (Van Craenenbroeck et al., 2004). However, the C-terminal part of ghrelin was shown to be important for the stability of ghrelin in in vivo testing (Morozumi et al., 2011).
On the basis of the structure-activity studies mentioned above, ghrelin analogs with potentially enhanced stability were designed in this study. Initially, ghrelin was modified by replacing the serine in position 3 with diaminopropionic acid (Dpr), which can form a stable amide bond with octanoyl acid [Dpr(N-octanoyl)] (according to Bednarek et al., 2000) and protect the molecule from hydrolysis by esterases. Subsequently, N-terminal glycine and Phe4 were replaced with noncoded amino acids to protect the analogs against the activity of aminopeptidases or chymotrypsin-like proteinases, respectively. The designed analogs of ghrelin were examined for their affinities to cell membranes with transfected GHS-R1a (Guerlavais et al., 2003), orexigenic effects in adult mice, GH secretagogue activities in young mice, and stabilities in blood plasma. In addition, desoctanoylated and shorter N-terminal analogs of ghrelin were tested similarly in the hope of obtaining a shorter active ghrelin analog.
We aimed to design peptidic ghrelin agonists with a biological activity similar to and a higher stability than ghrelin. Such substances could have a prolonged effect in the treatment of cachexia and sarcopenia.
Materials and Methods
Synthesis of Peptides.
Mouse ghrelin (ghr) [Gly-Ser-Ser(O-octanoyl)-Phe-Leu-Ser-Pro-Glu-His-Gln-His-Gln-Lys-Ala-Gln-Gln-Arg-Lys-Glu-Ser-Lys-Lys-Pro-Pro-Ala-Lys-Leu-Gln-Pro-Arg], [Dpr(N-octanoyl)3]ghrelin ([Dpr3]ghr), their full-length analogs with a C-terminal COOH, and N-terminal penta- and octapeptides amidated at the C terminus (Table 1) were assembled in a solid-phase ABI433A synthesizer (Applied Biosystems, Foster City, CA) by stepwise coupling of the corresponding fluorenylmethyloxycarbonyl amino acids to the growing chain on Rink amide resin (1% divinylbenzene, 200–400 mesh, 0.65 mmol/g) or Wang resin (1% divinylbenzene, 100–200 mesh, 0.57 mmol/g) (both from IRIS Biotech GmbH, Marktredwitz, Germany). The fully protected peptides were synthesized, and the peptides were purified and analyzed according to a standard procedure (Maixnerová et al., 2007). Octanoylation of Ser3 and Dpr3 was performed using a trityl group to protect the side chain of Ser3 and a methyltrityl group to protect the side chain of Dpr3, as described in the literature (Bednarek et al., 2000).
Ghrelin and [Dpr3]ghr were iodinated at His9 with either Na[125I] or nonradioactive NaI using IODO-GEN (Thermo Fisher Scientific, Waltham, MA)-coated Eppendorf tubes according to a published procedure (Elbert and Veselá, 2010). The purification was accomplished on an Agilent Prep-C18 5μ column (250 × 4.6 mm) (Agilent Technologies, Santa Clara, CA), a gradient elution from 10% B to 20% B over 5 min and next to 30% B in 60 min (A = water with 0.1% trifluoroacetic acid, B = acetonitrile with 0.1% of trifluoroacetic acid). The nonradioactive iodinated peptides were used as the control for molecular weights using MALDI-TOF Reflex IV mass spectrometry (Bruker Daltonics, Billerica, MA). The specific activity of the [125I]ghr or [125I][Dpr3]ghr was approximately 2000 Ci/mmol. Aliquots of the purified radiolabeled peptides were evaporated to dryness and kept at −20°C. Radiolabeled peptides were used for binding studies within 1 month.
Receptor Binding Studies.
Binding studies were performed as described in the literature (Guerlavais et al., 2003). In brief, isolated plasma membranes from LLC PK-1 cells with transfected human GHS-R1a (10 μg of protein per tube) were used. Cell membranes were incubated with 0.5 to 6 nM [125I]ghr or [125I][Dpr3]ghr in saturation experiments and with 0.05 nM [125I]ghr and 1 pM to 0.1 mM nonradioactive ligand in competitive binding experiments. Incubations were performed in a total volume of 0.5 ml of binding buffer (50 mM Tris/HCl, 5 mM EDTA, 2.5 mM MgCl2, and 0.1% bovine serum albumin) for 45 min at 25°C. Nonspecific binding was determined in the presence of 10 μM ghrelin. The binding reaction was stopped by the addition of ice-cold washing buffer [20 mM Tris (pH 7.4), 5 mM EDTA, 2.5 mM MgCl2, and 0.015% Triton 20] followed by rapid filtration over GF/C filters (Whatman, Clifton, NJ) presoaked with 0.5% polyethyleneimine using a Brandel cell harvester (Brandel Inc., Gaithersburg, MD). Filters were rinsed subsequently three times with ice-cold washing buffer. Bound radioactivity was determined by gamma counting (Wizard 1470 Automatic Gamma Counter; PerkinElmer Life and Analytical Sciences, Waltham, MA). Nonspecific binding amounted to <15% of the total binding. Experiments were carried out in duplicate at least three times.
All of the experiments followed the ethical guidelines for animal experiments and Czech Republic law 246/1992 and were approved by the committee for experiments with laboratory animals of the Academy of Sciences of the Czech Republic.
Inbred C57BL/6 male mice (AnLab, Prague, Czech Republic) were housed at a temperature of 23°C under a daily cycle of 12 h of light and dark (light from 6:00 AM) with free access to water and a standard chow diet that contained 25%, 9%, and 66% of calories from protein, fat, and carbohydrate, respectively. The energy content of the diet was 3.4 kcal/g (St-1; Mlýn Kocanda, Jesenice, Czech Republic).
Effect of Ghrelin Analogs on Food Intake.
Twelve-week-old mice were placed into separate cages for 1 week with free access to water and food pellets. For the evaluation of the orexigenic activity of peptides, mice were fed freely before the experiment. Anorexigenic activity was tested in mice that fasted for 17 h. The experiment started at 8:00 AM, and mice were injected subcutaneously with 0.2 ml of saline or the tested compounds (dissolved in saline) at a dose of 0.1 to 10 mg/kg (n = 5–6 mice per group). Fifteen minutes after the injection, mice were given preweighed food pellets. Food intake was followed for 6 h. The pellets were weighed at 30-min intervals and then returned to the cage, and cumulative food intake was registered. Animals had free access to water during the experiments.
The effects of selected ghrelin analogs on GH release was determined in 6-week-old male C57BL/6 mice. At 8:00 AM, fed mice were injected subcutaneously with 0.2 ml of the tested compounds (dissolved in saline) at a dose of 5 mg/kg (n = 4–5 mice per group). Ten or 30 min after the injection, blood was collected, and the plasma was separated and stored at −20°C until being used. GH in the plasma samples was determined with a radioimmunoassay kit (IZOTOP, Budapest, Hungary) according to the protocol recommended by the manufacturer.
Degradation of Ghrelin Analogs by Mouse Plasma.
Ghrelin and selected ghrelin analogs at a concentration of 1 μM in plasma from 12-week-old mice were incubated at 37°C for various time periods. Incubation was stopped by quick freezing to −20°C. After being thawed, samples were kept at 4°C throughout the following procedures. They were filtered through YM-10 ultrafiltration membranes (Millipore Corporation, Billerica, MA) by centrifugation at 4°C for 20 min at 16,000 x g. The filters were washed three times with 0.1% formic acid, and the filtrates were lyophilized subsequently. The samples were reconstituted with 0.1% formic acid up to the original volume of plasma for subsequent liquid chromatography-mass spectrometry (LC-MS) analysis. Chromatographic separation was achieved on an ACQUITY UPLC BEH C18 column (2.1 × 150 mm), packed with 1.7-μm particles (Waters, Milford, MA). The sample injection volume was 10 μl. The separation was performed using a linear gradient of 95% solvent A in solvent B to 100% solvent B over 20 min [solvent A consisted of H2O/acetonitrile, 98:2 (v/v), with 0.1% formic acid; solvent B was acetonitrile]. The flow rate was 100 μl/min and was generated with a Rheos 2200 pump system (Flux Instruments, Reinach, Switzerland) equipped with an HTS-PAL autosampler (CTC Analytics AG, Zwingen, Switzerland). Mass detection was performed using an LTQ Orbitrap mass spectrometer (Thermo Fisher Scientific) using electrospray ionization in positive mode. Electrospray ionization parameters were as follows: sheath gas flow rate, 25 arb; auxiliary gas flow rate, 5 arb; capillary temperature, 275°C; capillary voltage, 40 V; tube lens, 155 V; and ion spray voltage, 4.3 kV. All of the samples were run in duplicate.
The data are presented as the mean ± S.E.M. Saturation and competitive binding curves were plotted using GraphPad software (GraphPad Software Inc., San Diego, CA) while comparing the best fit for single binding site models (Kd, Bmax, and IC50 values were obtained from nonlinear regression analysis). Inhibition constants (Ki) were calculated from IC50 values using the Cheng-Prusoff equation (Chang and Cheng, 1978).
Food intake was measured in grams of chow consumed, and the ED50 values were calculated using GraphPad Software as the dose of the peptide required to elicit half-maximal consumption at 250 min after the injection of the compound (time of maximal effect).
Data from GH release and from food intake experiments were analyzed by one-way analysis of variance followed by Dunnett's post hoc test using GraphPad Software; p < 0.05 was considered statistically significant.
Design and Synthesis of Ghrelin Analogs.
The peptide sequences were assembled on a solid support as described under Materials and Methods. The purity of all of the peptides was >95%.
Structures of the ghrelin analogs used in this study are shown in Table 1. The ester group of the amino acid in position 3 was replaced with an amide group to increase the stability (Bednarek et al., 2000) in the series of [Dpr3]ghr derivatives (compounds 2–7 and 9-12). To further increase the stability of the ghrelin peptide analogs, sarcosine was used to replace glycine at the N-terminal position (analogs 3, 5, 7, 10, and 12), and/or the noncoded amino acids naphthylalanine (Nal) (analogs 4 and 5) and cyclohexylalanine (Cha) (analogs 6 and 7) replaced phenylalanine at position 4. To find the minimal structure that retained biological activity, N-terminal penta- and octapeptide analogs of ghrelin (8–12) were synthesized. Pentapeptides 13 and 14 are shorter versions of full-length ghrelin in which Trp3 or Nal3 replace Ser(O-octanoyl)3. These analogs were found by others (Matsumoto et al., 2001a) to activate calcium release in cells with transfected ghrelin receptors. A similar pentapeptide [Trp3,Arg5]ghr(1–5) was reported previously to stimulate GH secretion in rats and food intake in mice (Ohinata et al., 2006). Peptides 15 and 16 are the desoctanoyl forms of ghrelin and [Dpr3]ghr, respectively.
Binding of Ghrelin Analogs to GHS-R1a.
The saturable, specific binding of both [125I]ghr and [125I][Dpr3]ghr to cell membranes of LLC PK-1 cells with transfected human GHS-R1a exhibited a Kd value < 1 nM and a Bmax value close to 103 fmol/mg protein (Table 2). Nonspecific binding was approximately 20%. Nonlinear regression analysis showed one binding site for both ghrelin and [Dpr3]ghr (Fig. 1).
The results of competitive displacement of [125I]ghr binding by ghrelin analogs are summarized in Table 1. Ghrelin and its full-length analogs 2 to 8 had Ki values in the nanomolar range. The Ki values for analogs 4 and 5, which contained Nal in position 4, were even lower, in the 0.1 nM range. The N-terminal pentapeptides 8 and 9 and octapeptide 12 showed Ki values in the micromolar range (Table 1). Pentapeptides 13 and 14 did not displace [125I]ghr binding even at a concentration of 10 μM (Table 1), and full-length peptides lacking the octanoyl group showed the same result (15 and 16). Competitive binding experiments performed with [125I][Dpr3]ghr gave results similar to those obtained with [125I]ghr (M. Pýchová, unpublished observations).
In Vivo Studies.
The effects of subcutaneously administered ghrelin analogs on food intake were tested in fed mice (orexigenic activity) or fasted mice (anorexigenic activity). None of the tested ghrelin analogs caused a decrease in food intake in fasted mice at a dose of 10 mg/kg compared with that of the saline-treated group (M. Holubová, unpublished observations).
On the contrary, full-length analogs of ghrelin showed a very significant orexigenic effect in freely fed mice at doses of 1 to 10 mg/kg s.c. compared with that of the control (Fig. 2). In addition, the orexigenic effect of [Dpr3]ghr analogs lasted longer than that of ghrelin and was even more pronounced in analogs containing noncoded amino acids (peptides 3, 4, and 6 with Sar in position 1 and Nal or Cha in position 4; Fig. 2). The ED50 values when the maximal orexigenic effects were achieved at 250 min (Table 1) were in the range of mg/kg for all of the full-length analogs. The ED50 value for subcutaneously administered ghrelin in this study was approximately 2 mg/kg, which is similar to that reported previously (Perreault et al., 2004).
On the contrary, penta- and octapeptide ghrelin analogs, desoctanoyl ghr, and desoctanoyl [Dpr3]ghr did not show any significant orexigenic effect even at the maximum dose of 10 mg/kg (Table 1).
A significant and comparable increase in the release of GH at a dose of 5 mg/kg was observed in 6-week-old male mice after the administration of ghrelin and the selected analogs [Dpr3]ghr, [Dpr3,Nal4]ghr, and [Dpr3,Cha4]ghr (compounds 2, 4, and 6) both 10 and 30 min after treatment (Fig. 3).
Octanoylation of Ser3 in ghrelin is unique in biological systems, and its necessity for the binding of ghrelin to its receptor and for biological activity was recognized soon after the discovery of ghrelin (Kojima et al., 1999; Bednarek et al., 2000). Protection of the octanoyl group against the activity of esterases has been accomplished by replacing Ser3 with Dpr. This change was found earlier to affect neither the binding activity nor calcium release from cells with transfected with GHS-R1a (Bednarek et al., 2000). However, the in vivo properties of [Dpr3]ghr have not been followed previous to our study. All of the full-length ghrelin analogs in this study were based on [Dpr3]ghr.
In the in vitro study, we confirmed similar affinities of ghrelin and [Dpr3]ghr for GHS-R1a by obtaining similar Kd values in saturation binding and equal Ki values in competitive binding to cell membranes with transfected GHS-R1a.
In the in vivo experiments in adult mice, subcutaneously administered [Dpr3]ghr exhibited an augmented and longer-lasting orexigenic effect than that observed for ghrelin. The result corresponded with >4 h of stability of [Dpr3]ghr in blood plasma. Ghrelin was degraded totally in blood plasma in only 40 min, as reported also by De Vriese et al. (2004).
To protect [Dpr3]ghr against degradation by aminopeptidases, glycine in position 1 was replaced with Sar according to our previous experience (Maletínská et al., 1997), where Sar in position 1 was not found to affect the biological activity of angiotensin II. However, replacement of Gly1 with acetyl or N-acetyl-Gly (Bednarek et al., 2000; Van Craenenbroeck et al., 2004) has been shown to result in the loss of in vitro activity.
The Phe4-Leu5 bond of ghrelin or [Dpr3]ghr could be a cleavage site for chymotrypsin-like proteinases. As reported previously, replacement of Phe4 with alanine or tyrosine has resulted in complete loss of in vitro activity (Van Craenenbroeck et al., 2004). Therefore, Phe4 was replaced with the noncoded bulky hydrophobic phenylalanine derivatives Nal or Cha in this study.
Modification of [Dpr3]ghr with Nal or Cha in position 4 (compounds 4 and 6) preserved both the affinity to the receptor and the orexigenic effect; [Dpr3Cha4]ghr (compound 6) had both Ki and ED50 values of only one order of magnitude less than those of [Dpr3]ghr. Both [Dpr3Nal4]ghr and [Dpr3Cha4]ghr showed stabilities of >4 h in blood plasma, which was similar to that of [Dpr3]ghr. The three analogs [Dpr3Nal4]ghr, [Dpr3Cha4]ghr, and [Dpr3]ghr all showed an effect on GH secretion comparable with that of ghrelin.
Replacement of glycine in position 1 with Sar (compound 3) in [Dpr3]ghr resulted in an analog that preserved both its affinity to the receptor and the orexigenic effect, as observed in the analogs with modifications in position 4 (compounds 4 and 6). Interestingly, placing both Sar in position 1 and Nal or Cha in position 4 did not produce a positive additive effect on the affinity to the receptor or the orexigenic effect (compounds 5 and 7).
The ghrelin N-terminal penta- and octapeptides were reported not only to bind to cell membranes with transfected GHS-R1a and mobilize their calcium release (Bednarek et al., 2000; Matsumoto et al., 2001b) but also to potentiate GH secretion in vivo (Matsumoto et al., 2001a) and inhibit the secretion of pancreatic juice (Kapica et al., 2006). However, we found that the N-terminal ghrelin pentapeptide (compound 8), [Dpr3]ghr pentapeptide (compound 9), [Dpr3]ghr octapeptide (compound 11), and their Sar1 analogs (compounds 10 and 12) possessed Ki values only in the micromolar range and had no effect on food intake. Similar results were obtained with the N-terminal pentapeptides containing Nal3 and Trp3, which were reported previously to accomplish calcium release from GHS-R1a-transfected cell membranes and to have an orexigenic effect (Matsumoto et al., 2001a; Ohinata et al., 2006). In agreement with our findings, the biological activity of the ghrelin N-terminal pentapeptide was questioned previously because it neither bound to hypothalamus or pituitary membranes nor stimulated GH secretion in vivo (Torsello et al., 2002).
It has been shown in several studies that the desoctanoylation of ghrelin dramatically decreases binding to GHS-R1a (Kojima et al., 1999; Bednarek et al., 2000; Matsumoto et al., 2001a). In our study, we did not find binding to the receptor, an orexigenic effect, or an anorexigenic effect for both desoctanoyl ghrelin and desoctanoyl [Dpr3]ghr at doses up to 10 mg/kg s.c. in mice. We confirmed that desoctanoylation completely destroys the orexigenic effect of ghrelin and that the desoctanoylated peptide does not possess an anorexigenic effect.
With its unique lipopeptide chemical structure, ghrelin is the only known orexigenic hormone that is released peripherally. The diverse effects of ghrelin suggest possible clinical applications for GH deficiency, eating disorders, or gastrointestinal diseases.
In this study, stabilization of the octanoyl group by the introduction of Dpr(N-octanoyl)3 into ghrelin when combined with noncoded phenylalanine derivatives, such as Nal or Cha in position 4, produced stable ghrelin analogs with a high affinity to GHS-R1a and an orexigenic effect as potent as ghrelin, but with a substantially longer duration.
Ghrelin agonists have been tested recently as promising therapeutics in cachexia that accompanies cancer, chronic kidney disease, or heart failure (reviewed in DeBoer, 2008; Krasnow and Marks, 2010). The ghrelin analogs that were investigated in this study could be useful in further studies on the physiological role of ghrelin, and these peptides may be useful for the treatment of cachexia.
Participated in research design: Maletínská and Železná.
Conducted experiments: Maletínská, Pýchová, Holubová, and Železná.
Contributed new reagents or analytic tools: Blechová, Demianová, and Elbert.
Performed data analysis: Maletínská, Pýchová, and Holubová.
Wrote or contributed to the writing of the manuscript: Maletínská, Pýchová, Demianová, Elbert, and Železná.
We gratefully acknowledge the excellent technical assistance of H. Vysušilová. We thank D. Sýkora from the Institute of Chemical Technology, Prague, Czech Republic, for valuable advice in the LC-MS analyses. LLC PK-1 cell membranes transfected with GHS-R were a gift from C. M'Kadmi, Faculté de Pharmacie, Centre National de la Recherche Scientifique, Montpellier, France.
This work was supported by the Grant Agency of the Czech Republic [Grant 303/09/0744]; and the Academy of Sciences of the Czech Republic [Grant Z40550506].
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
- growth hormone
- diaminopropionic acid
- growth hormone secretagogue receptor
- liquid chromatography-mass spectrometry
- naphthylalanine, Sar sarcosine.
- Received June 22, 2011.
- Accepted December 16, 2011.
- Copyright © 2012 by The American Society for Pharmacology and Experimental Therapeutics