Endocrinology and Metabolism

Unsaturated fatty acids prevent desensitization of the human growth hormone secretagogue receptor by blocking its internalization

Patric J. D. Delhanty, Anke van Kerkwijk, Martin Huisman, Bedette van de Zande, Miriam Verhoef-Post, Carlotta Gauna, Leo Hofland, Axel P. N. Themmen, A. J. van der Lely


The composition of the plasma membrane affects the responsiveness of cells to metabolically important hormones such as insulin and vasoactive intestinal peptide. Ghrelin is a metabolically regulated hormone that activates the G protein-coupled receptor GH secretagogue receptor type 1a (GHSR) not only in the pituitary gland but also in peripheral tissues such as the pancreas, stomach, and T cells in the circulation. We have investigated the effects of lipids and altered plasma membrane composition on GHSR activation. Oligounsaturated fatty acids (OFAs) disrupt the structure of membranes and make them more fluid. Prolonged (96 h), but not acute, treatment of the GHSR cells with the 18C OFAs oleic and linoleic acid caused a significant increase in sensitivity of the receptor to ghrelin (EC50 reduced by a factor of 2.4 and 2.9 at 60 and 120 μM OFAs, respectively). OFAs were found to block the inhibitory effects of ghrelin pretreatment on subsequent ghrelin responsiveness, suggesting that OFAs suppress desensitization of GHSR. Radioligand displacement studies did not show a significant shift in receptor binding after incubation with OFAs. However, it was found that OFA treatment suppressed GHSR internalization, likely explaining OFA-induced refractoriness to ligand-induced desensitization. The involvement of lipid rafts in this process was indicated by the altered responsiveness of GHSR under conditions that alter membrane cholesterol. In conclusion, our findings demonstrate the importance of membrane composition for GHSR activation and desensitization and indicate at least part of the mechanism through which OFAs and cholesterol could affect ghrelin's activity in vivo.

  • ghrelin receptor fatty acids

clinical studies suggest that the growth hormone secretagogue response mediated through activation of the G protein-coupled growth hormone (GH) secretagogue receptor (GHSR) by ghrelin is modified by alterations in circulating lipid profile (3, 11). Treatment of humans with lipid infusions suppresses the GH secretagogue effect of ghrelin (11). This modulation of the GH secretagogue response in vivo may well be complicated by additional effects of lipids in the pituitary, for example, on GH-releasing hormone receptor function (15). However, the metabolic role of ghrelin is more important than its GH regulatory function, as demonstrated by studies in genetically modified mice in which neither Ghrl- nor Ghsr-deficient animals are dwarves but have altered metabolic phenotypes (53, 54, 63). Levels of lipids are significantly altered in obesity, anorexia, and lipodystrophy, but the composition of lipids may also be altered by nutritional intake (1, 17, 34, 40, 41). Since ghrelin has an important role in metabolic regulation, it is necessary to understand how alteration in metabolic and nutrional profiles could affect ghrelin signaling not only centrally but also in peripheral tissues such as the pancreas, where it induces an inhibitory tone on insulin secretion, and in the stomach (32, 33, 52, 53, 58). Ghrelin binds to the GHSR and elicits a rapid increase in PLC activity via Gq/11α proteins (27). PLC hydrolyzes phosphatidylinositol 4,5-bisphosphate, increasing intracellular inositol 1,4,5-trisphosphate, which leads to a rapid transient rise in intracellular calcium concentration ([Ca2+]i) (26). These signaling pathways are intimately associated with and influenced by cell membrane structure and dynamics (19, 20, 39, 42). It has been shown in numerous studies that growing cells in medium with an altered balance of specific lipids, particularly oligounsaturated fatty acids (OFAs), and sterols such as cholesterol results in an alteration of plasma membrane structure (8, 12, 22, 38). Such compounds have also been shown to have beneficial effects in vivo, for example, in cardiovascular and inflammatory disease and atherosclerosis, at least in part through modification of cell surface receptor function (16).

Treatment of humans with lipid infusions suppresses the GH secretagogue effect of ghrelin (11). This modulation of the GH secretagogue response in vivo may well be complicated by additional effects of lipids in the pituitary, for example, on GH-releasing hormone receptor function (15). We have used an in vitro approach to directly examine the modulatory effects of OFAs (oleic and linoleic acids) and cholesterol on GHSR function.


Cell Culture

Chinese hamster ovary (CHO) cells stably transfected with a human GHSR1a expression construct with a Neo cassette and a pIRES-puro (Clontech, Mountain View, CA) construct containing mitochondrially targeted apoaequorin were used in this study for luminometric determination of ghrelin induced [Ca2+]i (Euroscreen, Gosselies, Belgium). These GHSR/aequorin-expressing cells were maintained in Ham's F-12 medium (Invitrogen, Carlsbad, CA) containing 10% fetal bovine serum, 400 μg/ml G418, and 5 μg/ml puromycin (23). Oleic and linoleic acids were obtained as complexes with BSA in phosphate-buffered saline (Sigma, St. Louis, MO) and added directly to cells in culture. 25-Hydroxy cholesterol (Sigma) was dissolved in ethanol and then diluted 1,000-fold in medium. Vehicle control cells were incubated in 0.1% ethanol. β-Cyclodextrin was dissolved in BSA assay buffer (DMEM-Ham's F-12, 15 mM HEPES, 0.1% BSA, 100 U/ml penicillin, 100 μg/ml streptomycin, and 2.5 μg/ml fungizone, without phenol red) to the required concentration.

Reporter Gene Analyses

GHSR/aequorin-expressing cells in midlog phase were detached from plates with trypsin, recovered by centrifugation, and resuspended in BSA assay buffer. Cells were counted and adjusted to a density of 5 × 106 cells/ml. Coelenterazine h (Sigma) was then added to a concentration of 5 μM, and the cells were mixed for 3 h at room temperature. The cells were then diluted 10-fold with BSA assay buffer, and mixing was continued for 30 min. Ghrelin (NeoMPS, Strasbourg, France) was diluted from aqueous solution into BSA assay buffer to the required concentrations, and 100 μl was pipetted into the wells of a black 96-well plate. The plate was placed into a Victor2 multiplate reader, 100 μl of GHSR/aequorin cell suspension (50,000 cells) was automatically injected into the wells, and luminescence was measured over an integration period of 20 s. One percent Triton X-100 was subsequently injected to measure residual receptor response. The aequorin data are expressed as the fractional response of the cells to ghrelin relative to their total capacity to elicit a response. This is calculated by dividing the luminescent response to ghrelin by the combined luminescence generated by ghrelin and Triton X-100. This corrects for cell number and run-to-run variation in luminescence output.

To analyze leuteinizing hormone receptor (LHR) activity, 2 μg of LHR cDNA in the mammalian expression plasmid pSG5, 2 μg of pCRE6lux, and 1 μg of pRL-SV40 (Promega) were transfected into CHO-K1 cells using Fugene according to the manufacturer's instructions (Roche) in the presence or absence of 120 μM oleic acid. To analyze ACTH receptor [melanocortin 2 receptor (MC2R)] activity, HEK-293 cells stably transfected with MC2R and melanocortin receptor accessory protein were transfected with 2 μg of pCRE6lux and 1 μg of pRL-SV40 (Promega) using Fugene (Roche) in the presence or absence of 120 μM oleic acid. Cells were then detached using trypsin-EDTA at 72 h posttransfection, transferred to white μClear TC-treated plates (Greiner Bio-One, Alphen aan den Rijn, The Netherlands), and maintained for a further 24 h in the presence or absence of oleic acid. At 96 h, the medium was replaced by serum-free medium supplemented with 0.1% BSA and 25 mM HEPES and a threefold dilution series of human chorionic gonadotropin (hCG) or ACTH-(1–24). After 6 h of stimulation, wells were aspirated and cells were lysed with 25 mM Tris phosphate, pH 7.8, 8 mM MgCl, 1 mM dithiothreitol, 15% glycerol, and 1% Triton X-100 and analyzed for Photinus luciferase (cAMP-responsive element driven) and Renilla luciferase (transfection control) activities using the Dual-Luciferase Reporter Assay System according to the manufacturer's instructions (Promega, Madison, WI). The LHR assay was run three times in duplicate, and the MC2R assay was run four times in duplicate.

Whole Cell Binding Assays

GHSR/aequorin-expressing cells (23) were grown in monolayer to 80–90% confluence. Competition binding experiments were performed for 2.5 h at 4°C using 25 pM 125I-ghrelin (NEN, Boston, MA) in 0.2 ml of BSA assay buffer. The medium was then aspirated, and the cells were washed twice with 0.25 ml of ice-cold Ham's F-12. The cells were then lysed in 0.25 ml of 0.1 N NaOH and frozen. Lysates were counted in a γ-counter. Nonspecific binding was determined in the presence of 1 μM unlabeled ghrelin.

Confocal Microscopy

A GHSR cDNA was amplified by PCR from a pcDNA3.1 expression construct (UMR cDNA Resource Center; www.cdna.org) using a 3′ primer that subtended the EcoRI site and a primer that generated a BamHI site at the 5′ end. This was digested with EcoRI and BamHI and ligated into pEGFP-N1 (Clontech) to produce a COOH-terminal enhanced green fluorescent protein (EGFP)-tagged receptor construct. To enable us to control more precisely the level of expression of the GFP construct, we made use of the T-Rex tetracycline-inducible system (Invitrogen) (49). Therefore, we excised (EcoRI, NotI) the chimeric GHSR-GFP cDNA construct and ligated it into the tetracycline-inducible expression plasmid pcDNA4/TO (Invitrogen). This final construct was transiently transfected into T-Rex CHO cells stably transfected with a tetracycline repressor construct (pcDNA6/TR; Invitrogen). The cells were treated with OFA for the prescribed period and then stimulated with tetracycline to express the GHSR-GFP construct (pcDNA5/TO; Invitrogen) on the day of vehicle or ghrelin treatment. Fixed cells were examined using an LSM 510 META confocal microscope (Zeiss, Sliedrecht, The Netherlands). Image analysis was performed using ImageJ software (http://rsbweb.nih.gov/ij/download.html). Native Zeiss .lsm files were processed using the “smooth” function to lessen noise, and then endosome number was quantified using the “Find Maxima” function. This function identifies maxima of luminance in the image. Images were individually assessed for accuracy of counting. Differences between control and OFA-treated samples were assessed by Student's t-test.

Analysis of Data

Dose-response curves were fitted by nonlinear regression analysis using a single-site model and used to calculate EC50s of ghrelin following the various cell treatment protocols (Prism v.5; GraphPad Software, San Diego, CA). Prism was also used for statistical testing of differences between dose-response curves (Student's t-test and ANOVA with Dunnett's post hoc test).


Oligounsaturated Free Fatty Acids

Effect of OFA treatment on GHSR sensitivity.

Initial experiments showed that oleic and linoleic acid had no acute effects on the response of the GHSR to ghrelin, as assessed using the aequorin assay (120 μM OFA; Fig. 1A). In this case, GHSR/aequorin cells were exposed to OFAs only during the 20-s integration period required to measure aequorin's fluorescence response to ghrelin. A range of OFA concentrations ≤1 mM were tested, none of which had acute effects on the sensitivity of the receptor to ghrelin (data not shown). However, prolonged 96-h treatment of the GHSR/aequorin cells with OFA caused a significant increase in sensitivity of the cells to ghrelin relative to untreated controls. At 60 μM OFA the EC50 was significantly decreased, 2.4- (P < 0.001) and 2.9-fold (P < 0.001) for oleic and linoleic acid, respectively, relative to controls (Fig. 1B). At 120 μM OFA the EC50 was decreased 3.7-fold (P < 0.001) for both oleic and linoleic acid relative to controls (Fig. 1C). A possible mechanism for the improved response to ghrelin in OFA-treated cells is that receptor affinity for its ligand has been increased. Therefore, we next measured the ability of unlabeled ghrelin to displace radiolabeled ghrelin from GHSR/aequorin-expressing cells. We found that treatment for 96 h with either 60 or 120 μM oleic acid had a slight but nonsignificant effect on binding (P > 0.05, t-test; Fig. 2, A and B).

Fig. 1.

Effects of unsaturated fatty acids on the growth hormone secretagogue receptor (GHSR) response to ghrelin treatment. Oleic and linoleic acids [oligounsaturated fatty acids (OFAs)] have no acute effect on responsiveness of GHSR to ghrelin, whereas long-term (96 h) treatment with OFAs increases the sensitivity of GHSR to ghrelin. A: response of cells to a dose-response curve of ghrelin in control or 120 μM OFA without preincubation. B: control vs. 96-h treatment with 60 μM OFA. C: control vs. 96-h treatment with 120 μM OFA. In all experiments, data were integrated over 20 s. Data are representative of at least 3 independent experiments run in duplicate.

Fig. 2.

OFAs do not significantly improve the affinity of ghrelin binding to the GHSR. Displacement of 50 pM radioiodinated ghrelin from membranes generated from GHSR/aequorin cells treated with control medium or medium containing 60 (A) or 120 μM (B) oleic acid for 96 h. Data points were run in triplicate (P > 0.05).

Effect of OFA treatment on desensitization of the GHSR.

An additional possible mechanism for the increased responsiveness to ghrelin in the cells is that intrinsic mechanisms for receptor deactivation following ligand binding are downregulated. We tested this in GHSR/aequorin cells that had been grown for 96 h in the absence (controls) or presence of 120 μM OFAs. Cells treated in this way were incubated for 15 min with 0, 1, 10, or 100 nM ghrelin and washed extensively, and then their response to a dose curve of ghrelin was measured. We found that cells incubated under control conditions without free fatty acid were rapidly desensitized by ghrelin (representative data shown in Fig. 3, AD). Figure 3A shows the response of cells that had not been preexposed to ghrelin to a dose curve of ghrelin. These experiments reconfirmed the increased sensitivity of OFA-treated cells. Pretreatment of the cells with 1, 10, and 100 nM ghrelin (Fig. 3, B, C, and D) significantly reduced the response of control cells (dashed lines) to a subsequent ghrelin dose curve. Not only are the response curves shifted to the right, but the maximal response is also reduced. Furthermore, the degree of change in these parameters was dose dependent. In contrast, cells that had been grown in OFAs for 96 h showed a significantly less pronounced desensitization by pretreatment with ghrelin. Particularly with 1 and 10 nM ghrelin pretreatment (Fig. 3, B and C), the right shift and drop in maximal response of the ghrelin dose curve was significantly less than that observed in control cells. Cumulative data from three independent experiments (duplicate samples) are shown in Fig. 3E. Significant differences in sensitivity to ghrelin were found between cells grown in control (dotted line) and OFA-containing medium following pretreatment with 1 and 10 nM ghrelin (P < 0.001). No significant difference between controls and OFA-treated cells was observed with 100 nM ghrelin pretreatment, although OFA-treated cells were still not completely desensitized. The increased sensitivity to ghrelin of cells grown in OFA was independently confirmed in cells that had not been pretreated with ghrelin (EC50 reduced from 5 to 3.2 and 2.0 nM in oleic and linoleic acid-treated cells, respectively, P < 0.001; Fig. 3, A and E). Our data also suggest a slight difference in effect or potency between oleic and linoleic acid (Fig. 3, A and E).

Fig. 3.

Oleic and linoleic acids retard the desensitizing effects of ghrelin pretreatment. AD: representative ghrelin dose-response curves of GHSR/aequorin cells grown for 96 h in medium that lacked or contained 120 μM OFA, followed by 15 min pretreatment with 0, 1, 10, and 100 nM ghrelin (indicated above each graph). The luminescence response was integrated over 15 s immediately following exposure to ghrelin. E: cumulative data from 3 independent experiments. The x-axis represents the pretreatment concentration of ghrelin, and the y-axis represents the EC50 achieved in the subsequent ghrelin dose response (#P < 0.01, oleic vs. linoleic; ***P < 0.001, control vs. OFA).

Effect of OFA treatment on internalization of the GHSR.

To examine the mechanism for prevention of desensitization by OFAs in more detail, we examined the internalization of GFP-tagged GHSR in CHO cells that had been grown in medium containing 120 μM oleic acid for 96 h. To do this we used a tetracycline-inducible expression system (CHO-T-Rex cells were transfected with pcDNA5/TO containing a GHSR-EGFP expression cassette) to control the level of expression of the GFP-tagged receptor. Cells were concurrently treated with 10 nM ghrelin and transferrin-Alexa 633 conjugate and fixed at 0, 10, 30, and 60 min. Cells that had not been treated with oleic acid showed a rapid response to ghrelin, with the formation of GHSR containing endosomes within 10 min (Fig. 4, insets). These were found to partially colocalize with the clathrin-dependent endosomes identified using transferrin-Alexa conjugate. In cells that had been treated with oleic acid, the GHSR showed a markedly delayed entry into the endocytic pathway following treatment with ghrelin (Fig. 4). Normal function of the endocytic pathway in oleic acid-treated cells was indicated by the rapid internalization of the transferrin-Alexa conjugate within 10 min of treatment. These data were verified by quantification of fluorescent endosomes in >1,300 cells from six separate experiments (Fig. 5). At 10 min following ghrelin stimulation, only 9 ± 0.5% of the number of GHSR-EGFP-containing endosomes/cell had been stimulated in OFA-treated cells compared with control cells. By 60 min following ghrelin treatment, the number of endosomes per cell had reached 24 ± 1% of control levels. In contrast, the number of transferrin-Alexa 633-containing endosomes per cell was identical (100 ± 7%) in control and OFA-treated cells at 10 min following treatment, partially diminishing below control levels to ∼65% at 20–30 min and 50 ± 3% at 60 min following treatment. (Fig. 5, inset).

Fig. 4.

Oleic acid delays ghrelin-induced internalization of the GHSR: fluorescence confocal microscopy. Chinese hamster ovary (CHO)-T-Rex cells were transfected with pcDNA5/TO containing a GHSR-enhanced green fluorescent protein (EGFP) expression cassette and then treated for 96 h with or without 120 μM oleic acid. Receptor expression was induced using tetracycline. Cells were treated with ghrelin (10 nM) and transferrin-Alexa 633 conjugate (red) and fixed at 10, 30, and 60 min. GHSR in cells that had been treated with oleic acid show a delayed entry into the endocytic pathway, with GHSR containing endosomes forming only at 60 min following treatment with 10 nM ghrelin (arrows). Normal function of the endocytic pathway is indicated by the rapid internalization of the transferrin-Alexa conjugate within 10 min of treatment. Insets: cells that had not been treated with oleic acid show a rapid response to ghrelin, with the formation of GHSR (green) containing endosomes within 10 min. These partially colocalize (yellow; arrowheads), with the clathrin-dependent endosomes identified using transferrin-Alexa conjugate. Examples illustrate characteristic results from 6 independent experiments.

Fig. 5.

Oleic acid inhibits ghrelin-induced internalization of the GHSR: image analysis. Green fluorescent GHSR-EGFP endosomes were counted at 0, 10, 20, 30, and 60 min following ghrelin stimulation of control (○) and OFA-treated cells (●) from 6 separate experiments (control: 0 min, n = 72 cells; 10 min, n = 87 cells; 20 min, n = 44 cells; 30 min, n = 41 cells; 60 min, n = 48 cells; OFA: 0 min, n = 97 cells; 10 min, n = 186 cells; 20 min, n = 317 cells; 30 min, n = 203 cells; 60 min, n = 286 cells; ***P < 0.0001). Red fluorescent transferrin-Alexa 633-containing endosomes were also counted in a subset of experiments to compare the function of the endocytic pathway in control (▵) and OFA-treated cells (▴) (control: 0 min, n = 9 cells; 10 min, n = 28 cells; 20 min, n = 20 cells; 30 min, n = 18 cells; 60 min, n = 12 cells; OFA: 0 min, n = 18 cells; 10 min, n = 62 cells; 20 min, n = 91 cells; 30 min, n = 77 cells; 60 min, n = 99 cells).


Chelation of cholesterol with β-cyclodextrin.

The activity of many receptors, including G protein-coupled receptors (GPCRs), appears to be dependent on the level of cholesterol in the membrane. Cholesterol acts as an important structural component of the plasma membrane and facilitates the formation of lipid rafts and caveolae (20, 21, 28, 45, 48). Removal of cholesterol disrupts the structure of lipid rafts, allowing receptors to redistribute into surrounding nonraft membrane.

Initially, we tested the effect of removing cholesterol from membranes of the GHSR/aequorin cells using β-cyclodextrin, a potent cholesterol chelator (62). There was a rapid, dose-dependent, suppressive effect of β-cyclodextrin on the ability of ghrelin to stimulate receptor-mediated calcium flux in the cells. The EC50 of ghrelin was increased significantly from 1.5 nM in controls to 8.7 nM in 20 mM β-cyclodextrin-treated cells (P < 0.001; Fig. 6A). This could not be explained in terms of decreased receptor number, since the maximal response was unaffected and the cells remained viable during the assay, as assessed by trypan blue exclusion.

Fig. 6.

Alteration of cholesterol content in cells causes a reduction in GHSR sensitivity to ghrelin. A: cells treated for 3 h with increasing concentrations of β-cyclodextrin showed a dose-dependent decline in sensitivity to ghrelin. B: exposure of the cells to 1 μM 25-hydroxycholesterol for increasing periods of time showed a decline in response to ghrelin. Experiments were repeated 2–5 times in duplicate. Cell viability was unaffected in either group of experiments, as assessed using trypan blue exclusion.

Cholesterol supplementation.

Since chelation of free cholesterol from transfected cells disrupted ghrelin signaling via both Ca2+ and camp pathways, this suggested that the level of cholesterol is important for GHSR function. Therefore, we examined whether cholesterol repletion could affect ghrelin activation of its receptor. Time course experiments showed that, like oleic and linoleic acids, 25-hydroxycholesterol had no acute effect on the ghrelin response. However, increasing periods of treatment suppressed ghrelin-induced [Ca2+]i in the GHSR/aequorin transfectants, with a significant increase in EC50 from 2.1 nM in controls to 14 nM after 96 h of treatment (P < 0.001; Fig. 6B).

Effect of OFAs on Other GPCRs

To begin to examine the generality of our findings with the GHSR, we also examined the effects of longer-term treatment with oleic acid on the function of two other receptors, the LHR and the ACTH receptor MC2R. Both receptors activate adenylate cyclase via Gsα, therefore representing an alternative signaling pathway to the GHSR. We found that 96 h of oleic acid (120 μM) treatment of cells overexpressing the LHR had no significant effect on the EC50 for hCG (control EC50, 0.23 ng/ml; oleic acid-treated EC50, 0.28 ng/ml), as assessed by measuring their cAMP response using a cAMP response element reporter system (Fig. 7A). MC2R showed a minor increase in EC50 for ACTH in oleic acid-treated cells (control EC50, 0.25 nM; oleic acid-treated EC50, 0.48 nM; Fig. 7B).

Fig. 7.

Oleic acid does not affect the function of LH receptor or melanocortin 2 receptor (MC2R): 2 cAMP-coupled GPCRs. A: LHR response to human chorionic gonadotropin (hCG) in cells treated for 96 h with control or 120 μM oleic acid-containing medium. B: MC2R response to ACTH-(1–24) in cells treated for 96 h with control or 120 μM oleic acid-containing medium. The cAMP response was measured using a CRE-luc reporter system.


Clinical work suggests that the responsiveness of GHSR is blunted by lipid treatment (11). Furthermore, unsaturated fatty acids have been shown to have suppressive effects on other receptors involved in GH secretion. For example, the responses of GH-releasing hormone (GHRH), thyroid hormone-releasing hormone, and vasoactive intestinal peptide receptors to their ligands are all acutely downregulated by unsaturated fatty acids not by modulation of ligand binding or desensitization but through blockade of either downstream signaling or modulation of ion channel function (46, 47). Despite these clinical and in vitro findings, it is interesting that obese people demonstrate a marked response to ghrelin despite raised circulating free fatty acid levels (3). However, in vivo GHRH synergizes with ghrelin in affecting GH release; therefore, it is probable that the clinical effects of lipids reflect a more complex situation in the pituitary, whereby GHRH action is also modulated. Moreover, it is unknown whether GHSR density in the hypothalamo-pituitary unit is altered in obesity (36).

Fatty acids affect different receptors in specific ways and have been shown to increase the responsivity of certain GPCRs. For example, linoleic acid has been shown to be linked with improved β-adrenergic function (35), and oleic acid, but not linoleic acid, increases activation of the 5-HT7A receptor (2). Free fatty acids have also been shown to increase gonadotropin-releasing hormone receptor response (5). In other receptor types, including tyrosine kinase and ion channel receptors, ligand binding and signaling are also upregulated by unsaturated fatty acids (4, 13, 43, 44, 57, 60, 61, 64). Similarly, we found that the unsaturated fatty acids oleic and linoleic acid markedly improved the response of GHSR to ghrelin in vitro, in this case via a mechanism involving retardation of cellular internalization, rather than altered receptor affinity for the ligand.

GH secretagogue activity is evidently not ghrelin's major function, since neither hyperghrelinemic (9) nor ghrelin- or GHSR-deficient (51, 54) mice have an altered growth phenotype. Indeed, ghrelin has many other functions, including the regulation of energy homeostasis, adiposity, and feeding behavior (58). Obesity is characterized by suppressed circulating levels of ghrelin, although appetite is not suppressed accordingly. A possible explanation for this might be that raised circulating levels of free fatty acids, also characteristic of obesity, might increase the ability of ghrelin to signal at hypothalamic centers involved specifically in regulating energy balance [e.g., the ventromedial nucleus of the hypothalamus (VMH)] (58) and perhaps other sites such as the pancreas, leading to longer-term effects on energy storage, including increased adiposity. This possible mechanism requires further investigation since, intriguingly, ghrelin modulates fatty acid metabolism specifically in the VMH as part of the physiological mechanism that controls feeding (36).

An interesting correlate between ghrelin and oleic acid action is that both factors ameliorate hypertension (29, 56). Ghrelin has been shown to decrease blood pressure in animal models, and ghrelin levels are correlated with blood pressure in humans. Moreover, ghrelin causes vasorelaxation and improves endothelial function. These effects are intriguingly comparable with the effects of oleic acid and suggest that ghrelin and oleic acid could have synergistic effects on cardiovascular function.

The GHSR is desensitized via the classical clathrin-mediated pathway (14, 25). Although GHSR internalization is almost completely blocked by OFA treatment, transferrin receptor internalization is only partially suppressed. This suggests that there is specificity in the mechanism by which OFAs block entry of the GHSR into the endocytic pathway. Although the GHSR is prevented from joining the endocytic cycle, our signaling experiments with aequorin-expressing cells indicate that the GHSR is retained on the cell surface, where it can continue to signal in response to ghrelin. Retention of the receptor on the cell surface has been shown to be a mechanism for prevention of desensitization in other receptors. For example, blocking internalization of the structurally related, and Gq/11α-coupled, oxytocin receptor by the overexpression of dominant negative dynamin and Eps15 (components of clathrin-coated pits) has also been shown to prevent ligand-mediated desensitization (50). Likewise, a parathyroid hormone (PTH)/PTH-related protein receptor mutant with an impaired ligand-mediated internalization response was resistant to desensitization in vitro, and mutant receptor knock-in mice showed an exaggerated response to PTH (10, 55). We speculate that unsaturated fatty acid treatment may block the ability of membrane-trafficking proteins to target ligand-bound GHSR into the endocytic pathway. This could be mediated by disruption of normal membrane structure and composition, since these targeting proteins are predominantly integral or physically associated with the plasma membrane. The mechanisms by which OFAs could modify membrane structure include direct incorporation into membranes and, in the longer-term, becoming integrated into component phospholipids (22, 24, 38). A factor not examined in this study is the possible influence of metabolites of oleic and linoleic acid on receptor function. It is also important for future studies to determine tissue specificity of this mechanism, since the GHSR is expressed at varying levels in stomach, pancreatic islet cells, immune cells, and endothelial cells as well as the hypothalamo-pituitary unit (58).

Membrane proteins, including receptors and their associated signaling molecules, are often concentrated in patches of plasma membrane and lipid rafts that are considered to consist of a more ordered structure conferred by the presence of densely packed sphingolipids and cholesterol. Many GPCRs require association with lipid rafts for their efficient internalization and desensitization (7). The GHSR is expressed in T cells and has been shown to assemble into lipid rafts upon activation via T cell receptor ligation (18). This finding has led us to examine the effect of β-cyclodextrin treatment in our cells, which modifies the structure of lipid rafts by chelating cholesterol from the plasma membrane. β-Cyclodextrin blocks ghrelin activation of GHSR, emphasizing the importance of membrane composition in GHSR signaling and suggesting the generality of GHSR association with lipid rafts. 25-Hydroxycholesterol is known to stabilize lipid raft structure (6, 37). However, acute experiments in which this might be expected to improve GHSR function did not demonstrate this effect. In contrast, prolonged treatment with 25-hydroxysterol worsened the ghrelin response. This is likely due to the inhibitory effect that this oxysterol has on cholesterol levels over longer time periods by suppressing transcription of genes that encode cholesterol biosynthetic enzymes (59). Modification of membrane cholesterol using these commonly used methods probably has more widespread effects than disruption of membrane rafts, and therefore, our findings related to these experiments should be interpreted with caution. However, unsaturated fatty acids have been demonstrated by others to alter lipid raft structure (16), and this could constitute part of the mechanism by which OFAs suppress the movement of GHSR into the endocytic pathway on ligand binding and warrants further investigation.

Although this study is focused on the regulation of GHSR function by OFAs, it was of interest to us to test the possibility that the function of GPCRs with other modes of intracellular signaling might be modulated by OFA treatment. With this goal in mind, we tested two GPCRs, the LHR and the MC2R, that are unlike the GHSR in being coupled via Gsα to the production of cAMP by activating adenylate cyclase. Both of these receptors show ligand-dependent internalization (30, 31). However, in the case of these receptors, treatment with oleic acid for 96 h had very little effect on the EC50s of their respective ligands. This indicated that OFA treatment had no appreciable effect on function or internalization characteristics of these particular receptors. However, there are hundreds of characterized GPCRs, and therefore, a high-throughput approach testing the responses of a larger number of GPCRs is warranted to examine the generality of effect of OFA treatment on receptor recycling. A more focused approach could be to examine OFA modulation of function of receptors expressed specifically in the brain and/or hypothalamus.

In conclusion, our findings demonstrate the importance of membrane composition for GHSR activation and desensitization. Importantly, we have found that free fatty acids improve the sensitivity of the GHSR by suppressing ligand-induced receptor internalization by uncoupling the GHSR from the endocytic pathway, a mechanism not described before for OFA action on receptor function. It is conceivable that OFAs could produce some of their beneficial physiological effects partly by modulating ghrelin signaling both centrally in the hypothalamus and at peripheral sites, for example, in the cardiovascular system, immune cells, and the pancreas.


This study was supported by the Netherlands Organization for Scientific Research (NWO Grant 912-03022).


No conflicts of interest, financial or otherwise, are declared by the authors.


We would like to thank Prof. Shin-Ichiro Takahashi (University of Tokyo) for the initial impetus to undertake this study and for helpful comments on this article and Dr. Gert van Cappellen for technical assistance with the confocal microscopy and quantitative analyses.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
  54. 54.
  55. 55.
  56. 56.
  57. 57.
  58. 58.
  59. 59.
  60. 60.
  61. 61.
  62. 62.
  63. 63.
  64. 64.
View Abstract