The Hyperphagic Effect of Ghrelin Is Inhibited in Mice by a Diet High in Fat

Transcription

1 GASTROENTEROLOGY 2010;138: The Hyperphagic Effect of Ghrelin Is Inhibited in Mice by a Diet High in Fat JAMES V. GARDINER, DANIEL CAMPBELL, MICHAEL PATTERSON, AYSHA KENT, MOHAMMED A. GHATEI, STEPHEN R. BLOOM, and GAVIN A. BEWICK Department of Investigative Medicine, Hammersmith Campus, Imperial College London, London, United Kingdom BACKGROUND & AIMS: Ghrelin is the only peripheral hormone known to increase food intake. It is released from the stomach and is thought to function as a signal of energy deficit and a meal initiator. We generated transgenic mice in which levels of bioactive ghrelin are increased in the stomach and circulation. These mice, as expected, are hyperphagic and glucose intolerant. We investigated whether exposure to a high-fat diet (HFD) would exacerbate this phenotype. METHODS: We investigated the effect of HFD on energy and glucose homeostasis in ghrelin transgenic mice. We determined dietary preference; expression of hypothalamic neuropeptides that control food intake; and, using fast-performance liquid chromatography, the circulating forms of ghrelin. We measured food intake during continuous administration of ghrelin in wild-type mice fed either regular chow or an HFD. RESULTS: Ghrelin transgenic mice were resistant to diet-induced obesity because of their reduced food intake. This was not caused by alterations to food preference, hypothalamic signaling of neuropeptides that control food intake, or the form of circulating acylated ghrelin. Long-term administration of ghrelin to wild-type mice failed to increase ingestion of an HFD but, as expected, increased intake of regular chow. CONCLUSIONS: This is the first report that diets high in fat inhibit the hyperphagic effect of ghrelin; these findings indicate that features of the diet are important determinants of ghrelin s function. This information is important for the development of anti-obesity drugs that target ghrelin signaling. Keywords: Transgenic; Appetite Regulation; Diabetes; Leptin; Hypothalamus. Ghrelin is the endogenous ligand for the growth hormone secretagogoue receptor and is found in high concentrations within the stomach. 1 Initial data suggested that ghrelin plays a physiologic role in meal initiation and signalling states of energy deficit. For example, it increases food intake when administered exogenously. In addition, plasma ghrelin levels rise just prior to a meal and also following food deprivation. 2 4 Despite the strong pharmacologic data, the physiologic role of ghrelin is still controversial. Particularly because transgenic approaches have not provided definitive answers. Mice with targeted deletion of either ghrelin or growth hormone secretagogoue receptor exhibit an essentially normal metabolic phenotype when fed a regular chow (RC) diet, suggesting that ghrelin may have a redundant role in the regulation of food intake. 5,6 However, when on a high-fat diet (HFD), these mice are resistant to diet-induced obesity (DIO), exhibiting reduced adiposity and increased energy expenditure. 7,8 More recent data suggest that this effect is lost on a pure C57BLK/6 genetic background implying that other susceptibility factors interact with ghrelin. 9 The conflicting food intake and body weight data from transgenic models have made defining a key role for endogenous ghrelin in the control of appetite difficult. We recently generated a model of ghrelin overexpression in which bioactive ghrelin concentrations were increased in the stomach and circulation. 10 Initial characterization of this model indicated that, as expected, ghrelin overexpression caused hyperphagia and inhibited glucose-stimulated insulin release. Our data suggested that ghrelin regulates appetite under normal feeding conditions and has an important role in regulating -cell function. Ghrelin s role in glucose homeostasis is reviewed elsewhere. 11 Here, we present data from feeding this mouse an HFD. We found that transgenic mice overexpressing ghrelin (Tg) were unexpectedly protected against DIO because of a profound reduction in food intake compared with wild-type (Wt) littermates. In support of this, long-term administration of ghrelin via osmotic minipump to Wt mice did not increase food intake when mice were fed an HFD, but, as expected, ghrelin significantly increased cumulative food intake when mice were fed RC. These data suggest that dietary constituents, in particular fat, may have an important impact on the function of ghrelin. Abbreviations used in this paper: AUC, area under the curve; BAT, brown adipose tissue; DIO, diet-induced obesity; FPLC, fast-performance liquid chromatography; HFD, high-fat diet; NYP, neuropeptide Y; POMC, proopiomelanocortin; RC, regular chow; SEM, standard error of mean; UCP-1, uncoupling protein-1; Wt, wild type by the AGA Institute /$36.00 doi: /j.gastro

2 June 2010 HIGH-FAT DIETS INHIBIT GHRELIN HYPERPHAGIA 2469 Materials and Methods Generation of Ghrelin Tg Mice The generation of ghrelin Tg mice was described previously. 10 Briefly, we identified a bacterial artificial chromosome (BAC) containing the ghrelin gene RP23-441K11 (Invitrogen, Huntsville, AL). Tg mice were created using standard pronuclear injection techniques. F0 mice were mated with CBA/C57Bl6 mice, and transgenic lines were maintained separately. Mice were housed in cages under controlled temperature (21 C 23 C) and light (11 hours light/13 hours dark) with ad libitum access to food (RM1 diet; SDS UK Ltd, Witham, UK, or HFD 60% kcal as fat, Research Diets, Inc, New Brunswick, NJ) and water. Animal procedures performed were approved under the British Home Office Animals Scientific Procedures Act Body Weight, Food Intake, Indirect Calorimetry, and Body Composition Mice (n 6) were singly housed from weaning and fed either RC or HFD, and food intake and body weight were measured. Body composition of 16-week-old mice was measured as previously described. 10 Metabolic parameters were obtained using the open-circuit Oxymax comprehensive laboratory animal monitoring system (Columbus Instruments, Columbus, OH) as previously described. 12 For the food preference studies, male mice (n 8) were offered a choice between RC and HFD from 5 weeks of age. Consumption of both diets was measured separately. Glucose Tolerance Test and Insulin Tolerance Test Intraperitoneal glucose tolerance tests and insulin tolerance tests were performed in conscious 16-week-old mice as previously described. 10 Measurement of Circulating Hormones Whole blood was collected by cardiac puncture from fasted 16-week-old mice for transgenic studies and at the end of the study for the exogenous ghrelin administration study. Plasma insulin and leptin concentrations were determined by enzyme-linked immunosorbent assay (Chem Inc, Downers Grove, IL). Octanylated ghrelin concentrations were analyzed using an enzyme-linked immunosorbent assay (LINCO Research, St. Charles, MO). Northern Blot Analysis Tissues from 16-week-old mice were snap frozen, and RNA was extracted using TRI reagent. Northern blot analysis was used to determine uncoupling protein-1 (UCP-1) messenger RNA expression in brown adipose tissue (BAT) and neuropeptide Y (NPY) and proopiomelanocortin (POMC) levels from hypothalamic extracts as previously described. 10,13 Briefly, 50 g total RNA was size separated on a denaturing gel and transferred to a Hybond-N membrane (Amersham Biosciences, Little Chalfont, UK). The RNA was fixed before probing with riboprobes corresponding to either nucleotides of the mouse UCP-1 (accession No. BC012701) or nucleotides of the mouse NPY gene (accession No. NM_023456) or nucleotides of the mouse POMC gene (accession No. NM_008895). These were generated from either total BAT RNA or hypothalamic RNA by reverse-transcription polymerase chain reaction. Hybridization was carried out overnight at 55 C before nonspecific hybridization was removed by increasingly stringent washes, the final one being in 0.1X saline sodium citrate/0.1% (wt/vol) sodium dodecyl sulfate at 70 C for 30 minutes. The membranes were exposed to storage phosphor screens (Molecular Dynamics, Amersham, UK) and quantified using a Storm imaging system (Molecular Dynamics, Amersham, UK). The ratio of specific message to 28S ribosomal RNA was used to quantify expression for each tissue sample. Fast-Performance Liquid Chromatography Ghrelin was extracted from plasma using Sep-Pak C18 cartridges from either Wt or Tg male mice fed either RC or HFD (n 10) (Waters, Hertfordshire, UK) as described previously. 14 Pooled peptide extracts from 3 ml plasma were dissolved in 1 ml distilled water plus trifluoroacetic acid (TFA) 0.05% (vol/vol). Of this volume, 0.9 ml was fractionated by fast-performance liquid chromatography (FPLC) on a high-resolution reverse-phase (Pep RPC 1-mL high resolution) column (GE Healthcare, Little Chalfont, UK). The column was eluted with a 20% 27% gradient of AcN/water 0.05% (vol/vol) TFA over 50 minutes, and 1 ml fractions were collected at 1-minute intervals. The ghrelin-immunoreactivity in all fractions was determined by radioimmunoassay. The elution of the ghrelin-immunoreactivity was compared with the elution positions of octanoylated ghrelin, decanoylated ghrelin, and des-acylated ghrelin. Exogenous Administration of Ghrelin Four groups of 10-week-old male C57BLK/6 mice (n 8) were implanted with subcutaneous minipumps (Alzet model 2002; Charles River, Margate, UK) containing either saline or ghrelin (0.6 nmol/g/day). Mice were fed either RC or HFD, and food intake and body weight were measured daily. Statistical Analysis Values are the mean standard error of mean (SEM) unless otherwise stated. Differences in cumulative food intake through time were compared across experimental groups using generalized estimating equation curve analysis (Stata 9.1; Statacorp, College Station, TX). All other comparisons were made using an un-paired Student t test. P values.05 were considered significant.

3 2470 GARDINER ET AL GASTROENTEROLOGY Vol. 138, No. 7 Figure 1. Excess ghrelin protects against diet-induced obesity by reducing food intake. Male mice (n 6) were fed either RC or an HFD from 5 weeks of age. (A) Cumulative energy intake; (B) body weight curves; (C) body composition measured at 16 weeks of age; (D) plasma leptin concentration measured at 16 weeks of age; (E) cumulative food intake measured by the comprehensive laboratory animal monitoring system (CLAMS); (F) average light and dark phase food intake measured by the CLAMS. Data represent mean standard error of mean. *P.05 compared with TG HF. Results Under High-Fat Feeding Conditions, Excess Ghrelin Protects Against DIO by Inhibiting Food Intake Data from studies using pharmacologic administration of ghrelin and from our ghrelin overexpression model suggest that excess ghrelin increases food intake under normal feeding conditions. Extrapolation of these data would suggest an equivalent or exaggerated response to be displayed under high-fat feeding conditions. Mice were singly housed at weaning, and, from 5 weeks of age, mice were fed either an HFD (60% kcal as fat) or RC. Food intake and body weight were measured for 11 weeks. As we have previously shown, Tg mice exhibited increased daily and cumulative energy intake compared with Wt mice when fed RC (Wt kj vs Tg kj, n 6, P.05). Surprisingly, on an HFD, Tg mice consumed significantly less food than controls (Wt 12, kj vs Tg 11, kj, n 6, P.05) (Figure 1A). As a consequence, Tg mice gained significantly less weight than Wt mice and were 14% lighter by 16 weeks of age (HFD: Wt g vs Tg g, n 6, P.05). Tg mice were not completely protected against DIO because they gained more weight than mice fed on RC (Figure 1B). On an HFD, Tg mice were leaner than Wt controls, having approximately 40% less total body fat, which accounted for the majority of the difference in weight between the 2 groups (Wt g vs Tg g, n 6, P.05) (Figure 1C). In line with this, plasma leptin was decreased by 70% in Tg mice compared with Wt controls (Wt ng/ml vs Tg ng/ml, n 6, P.05) (Figure 1D). We have previously shown that body composition and circulating leptin levels were equivalent between ge-

4 June 2010 HIGH-FAT DIETS INHIBIT GHRELIN HYPERPHAGIA 2471 notypes when fed RC. 10 Tg mice retained a significantly elevated plasma active ghrelin concentration when fed an HFD, suggesting that high-fat feeding did not inhibit the overexpression of ghrelin (Wt fmol/ml vs Tg fmol/ml, n 9, P.05) (Supplementary Figure 1A). Under High-Fat Feeding Conditions, Excess Ghrelin Reduces Dark Phase Food Intake but Does Not Alter Energy Expenditure or Locomotor Activity To investigate further these unexpected findings, we measured food intake, locomotor activity, and oxygen consumption using the comprehensive laboratory animal monitoring system. At 5 weeks of age, mice were individually housed in the comprehensive laboratory animal monitoring system cages and fed an HFD for 3 days. In agreement with our previous studies, Tg mice ate significantly less than Wt mice (Wt gvsTg g, n 6, P.05) (Figure 1E). Tg mice ate 48% less food during the dark phase when mice consume most of their food (Wt g/day vs Tg g/day, n 6, P.05). During the light phase, food intake tended to be lower in Tg compared with Wt mice (Wt g/day vs Tg g/day, n 6) (Figure 1F). The reduction in food intake was not associated with changes in locomotor activity (Supplementary Figure 1B), oxygen consumption a measure of energy expenditure (hourly average VO 2, Wt ml/h/kg vs Tg ml/h/kg, n 6, P.68, Supplementary Figure 1C), or respiratory quotient (hourly average VCO 2 / VO 2, Wt vs Tg , n 6, P.29, Supplementary Figure 1D). UCP-1 expression in BAT is a surrogate marker of energy expenditure. As we have previously shown, Tg mice fed an RC diet have increased BAT UCP-1 expression compared with controls. When fed an HFD, Wt mice increase their BAT UCP-1 expression compared with those fed RC, a function of dietinduced thermogenesis, but Tg mice did not 15 (Supplementary Figure 1E). These data are consistent with the oxygen consumption data suggesting that energy expenditure was not altered between genotypes under high-fat feeding conditions. Together, this suggests that resistance to DIO obesity in TG mice was due to a specific reduction in food intake and not to alterations in energy expenditure or to alterations in the substrate used for fuel, carbohydrate versus fat, by Tg mice. High-Fat Feeding Does Not Alter Glucose Sensitivity or Insulin Tolerance in Ghrelin Overexpressing Mice Pharmacologic administration of ghrelin inhibits glucose stimulated insulin release. 16 In agreement with this, genetic deletion of ghrelin augments glucose-stimulated insulin release, whereas overexpression inhibits it. 10,17 19 Despite similar fasting blood glucose concentrations between genotypes irrespective of diet (RC: Wt mmol/l vs Tg mmol/l; HFD: mmol/l vs mmol/l, n 6) (Figure 2A), Wt mice fed an HFD exhibited markedly increased fasting plasma insulin concentrations compared with Wt mice fed RC or Tg mice fed either an HFD or RC (RC: Wt ng/ml vs Tg ng/ml; HFD: Wt ng/ml vs Tg ng/ml, n 6, P.05, Wt RC vs Wt HFD, P.05, Wt HFD vs Tg HFD) (Figure 2B), suggesting they were comparatively insulin resistant because of their obesity. To explore further the interaction of high-fat feeding with glucose homeostasis, we performed glucose and insulin tolerance tests. In agreement with our previous findings, Tg mice were glucose intolerant but remained equally insulin sensitive when compared with Wt mice fed RC (RC: Wt area under the curve (AUC) vs Tg AUC, n 6, P.05) (Figure 2C F). As expected under high-fat feeding conditions, Wt mice became glucose intolerant and insulin resistant consistent with their obese phenotype. High-fat feeding did not exacerbate the impaired glucose tolerance exhibited by Tg mice and had no effect on their insulin sensitivity despite their increased total body fat mass compared with mice fed on RC (HFD: Wt AUC vs Tg AUC, n 6) (Figure 2C F). These results suggest that the relative leanness of Tg mice compared with Wt mice protected them from the impairment of glucose tolerance associated with DIO. Under High-Fat Feeding Conditions, Food Preference, Hypothalamic Neuropeptide Expression, and Ghrelin Acylation Are Unaltered in Ghrelin Overexpressing Mice To investigate whether excess ghrelin altered food preference, we gave the mice a choice between RC and HFD and monitored food intake for 1 month. Both genotypes ate equivalent amounts of RC and exhibited a marked preference for HFD (average daily food intake, 0 2 weeks RC: Wt g/day vs Tg g/day; HFD: g/day vs g/day, n 6, P.05). As in the previous study, Tg mice consumed significantly less HFD compared with Wt mice (Figure 3A). This suggested that the reduction in food intake was not due to a dislike of the HFD. Ghrelin is thought to increase food intake by activating orexigenic NPY neurones found within the arcuate nucleus of the hypothalamus. 20,21 Alterations in this and other hypothalamic pathways implicated in the control of food intake, for example, the anorectic POMC-derived peptides, could account for the unexpected reduction in food intake following high-fat feeding. However, we found that messenger RNA expression of hypothalamic NPY and POMC were unaltered in Tg mice when fed HFD (POMC: Wt arbitrary units vs Tg arbitrary units; NPY: arbitrary units vs arbitrary units, n 6) (Figure 3B).

5 2472 GARDINER ET AL GASTROENTEROLOGY Vol. 138, No. 7 Figure 2. High-fat feeding does not exacerbate ghrelin-induced impairment of glucose tolerance. Male mice (n 6) were fed either RC or HFD from 5 weeks of age. (A) Fasting plasma glucose; (B) fasting insulin levels measured at 16 weeks of age; (C) plasma glucose disposal during a glucose tolerance test at 14 weeks of age; (D) area under the curve for C *P.05 compared with RC Wt; (E) plasma glucose during an insulin tolerance test at 15 weeks of age; (F) area under the curve for E. Data represent mean standard error of mean. *P.05. Ghrelin s bioactivity is dependent on acylation of the serine 3 residue with the medium chained fatty acid octanoic acid (C8:0). This reaction is catalyzed by the recently identified enzyme Ghrelin O-Acyltransferase. 22,23 The acyl group added to ghrelin has been shown to be dependent on the fatty acid content of the diet. For example, addition of heptanoic acid (C7:0) to the diet causes ghrelin to be preferentially acylated with heptanoate rather than with octanoic acid. 24,25 The biologic significance of acylation with different medium chain fatty acids is currently unclear. However, the reduced food intake exhibited by our Tg mice fed an HFD could be explained by an alternate acylation of ghrelin because of exposure to increased levels of dietary fat. Using FPLC, we found that the ratio of plasma acylated to des-acylated ghrelin was unaffected by high-fat feeding compared with RC feeding. Perhaps more importantly, we did not detect alternate species of acylated ghrelin following exposure to increased dietary fat (Figure 3C F). We concluded that, in the context of excess energy availability in the form of increased dietary fat, the ghrelin system protects against DIO because of an unexpected inhibition of food intake without altering energy expenditure. The reduction in food intake did not appear

6 June 2010 HIGH-FAT DIETS INHIBIT GHRELIN HYPERPHAGIA 2473 Figure 3. Food preference, hypothalamic neuropeptide expression, and acylation substrate are unchanged by high-fat feeding. (A) Average daily food intake in male mice (n 8) fed a choice of RC and HFD from 5 weeks of age; (B) proopiomelanocortin (POMC) and neuropeptide Y (NPY) messenger RNA expression in whole hypothalamus extracts from male mice (n 6) fed an HFD. Acylated species of ghrelin in Wt (C and D) and Tg (E and F) plasma from mice fed RC HFD resolved using FPLC. Data represent mean standard error of mean. *P.05. to be a function of altered food preference or altered signalling in hypothalamic neuropeptide systems implicated in the control of food intake or because of changes in the substrate used by Ghrelin O-Acyltransferase to acylate ghrelin. Under High-Fat Feeding Conditions, Exogenous Administration of Ghrelin Does Not Increase Energy Intake in Wt Mice As another model of chronically elevated ghrelin, we used osmotic minipumps to deliver ghrelin to Wt mice fed either an RC or HFD. Intraperitoneal delivery of ghrelin via osmotic minipump caused a 7-fold increase in circulating active ghrelin (RC: saline fmol/ ml vs ghrelin fmol/ml, n 8, P.05; HFD: fmol/ml vs fmol/ml, n 8, P.05) (Figure 4A). Ghrelin administration did not alter the body weight of mice fed either diet during the study (Figure 4B). As expected, we found that chronic administration of ghrelin significantly increased cumulative food intake versus saline in mice fed RC (food intake, RC: saline g vs ghrelin g, n 8, P.05; energy intake, RC: kj vs kj, n 8, P.05) 26 (Figure 4C and D). However, in accordance with our Tg model, long-term excess ghrelin tended to reduce food intake in mice fed an HFD

7 2474 GARDINER ET AL GASTROENTEROLOGY Vol. 138, No. 7 Figure 4. Chronic administration of ghrelin does not increase food intake under high-fat feeding conditions. Male mice (n 8) were implanted with subcutaneous osmotic minipumps, which delivered 0.6 nmol/g/day of ghrelin or saline intraperotoneally. Mice were fed either regular or HFD. (A) Plasma active ghrelin (*P.05); (B) body weight; (C) cumulative food intake; (D) cumulative energy intake. Data represent mean standard error of mean. (*P.05 between saline and ghrelin-treated mice fed regular chow.) (food intake, HFD: saline g vs ghrelin g, n 8, P.09; energy intake, HFD: kj vs kj, n 8, P.09) (Figure 4C and D). These data confirm that dietary fat may inhibit ghrelin s hyperphagic action. Discussion We have recently developed a mouse model in which ghrelin is overexpressed in its physiologically relevant tissues. On a normal diet, Tg mice were hyperphagic but did not exhibit alterations in body weight because of an increase in energy expenditure. This suggested that ghrelin plays an important role in the physiologic regulation of energy homeostasis under normal feeding conditions. Consistent with previous data, we also found that excess ghrelin inhibited glucose-stimulated insulin release, strengthening the importance of ghrelin in regulating -cell function. To investigate further the effect of excess ghrelin in the control of energy homeostasis, we fed our mice an HFD expecting the phenotype to be exacerbated. Surprisingly, we found that our Tg mice were resistant to DIO because of a profound reduction in HFD ingestion, which was not mediated by alterations in food preference, major changes in neuropeptide signalling in the hypothalamus, or in ghrelin acylation. This is in contrast to the current hypothesis regarding ghrelin s function. Ghrelin is the only known peripherally released orexigen. Pharmacologically, the evidence unequivocally supports this conclusion. However, genetic manipulation of the ghrelin system has provided conflicting data as to the physiologic role of ghrelin. Ghrelin null mice fed an RC diet were found to be indistinguishable from their Wt littermates in terms of energy homeostasis. 5 Subsequently, Wortley et al independently confirmed these results and also found that ghrelin deletion reduced the respiratory quotient of mice fed an HFD indicating a potential role for ghrelin in determining the type of metabolic substrate (fat vs carbohydrate) used for the generation of energy. 6 When exposed to HFD from 5 weeks of age, ghrelin-deficient mice were resistant to DIO because of increased metabolic rate and locomotor activity but without alterations in their food intake. 7 Similarly, Zigman et al found ghrelin receptor-deficient mice were also resistant to DIO, but this was associated with reduced food intake. 8 These data suggest that ghrelin does not play a critical role in the control of energy homeostasis under normal feeding conditions, but, in the presence of excess dietary fat, ghrelin signalling may be important in the development of an obese phenotype. More recently, Sun et al have shown that ghrelin and ghrelin receptor-deficient mice backcrossed to a pure C57BL/6J genetic background are not resistant to DIO when fed an HFD from 16 weeks of age, 9 suggesting that

8 June 2010 HIGH-FAT DIETS INHIBIT GHRELIN HYPERPHAGIA 2475 genetic background and age are important factors in determining ghrelin s role in susceptibility to DIO. These are unlikely to be confounding factors in our study because Tg and Wt mice have identical mixed genetic background, and mice were fed HFD from an early age. These data suggest that there are complex interactions between ghrelin and genetic background, diet, and age. This, coupled with our data, indicates that the role of ghrelin in the control of energy homeostasis is much more complex than originally thought, and its function is critically dictated by the prevailing dietary environment. This complexity offers an explanation for the conflicting data regarding ghrelin s physiologic role in energy homeostasis. Given the unexpected nature of our findings, we asked the following question: does exposure to an HFD inhibit the hypophagic effect of ghrelin in Wt mice? To achieve this, we used osmotic minipumps to deliver bioactive ghrelin to normal mice fed either an HFD or RC. Whereas there was no change in body weight following exposure to ghrelin, a likely function of the duration of the study, ghrelin as expected increased cumulative food intake in mice fed RC. In contrast, ghrelin-treated mice fed an HFD tended to consume fewer calories than saline-treated controls. Although this reduction did not reach statistical significance, there was a difference between ghrelin-treated mice fed an RC and those fed HFD in which ghrelin failed to increase energy consumption. The reduction in food intake following high-fat feeding was not as marked as in Tg mice. A potential explanation for this is the method of delivery used. Ghrelin administered by osmotic minipump produces a sustained increase in circulating ghrelin. In contrast, excess ghrelin in our Tg mice is likely to be physiologically regulated because the transgene is under the control of the endogenous ghrelin promoter, which would preserve the pulsatile nature of its release. Our data suggest a direct interaction between the ghrelin system and high-fat feeding resulting in inhibition of food intake. Whereas this is perhaps counterintuitive, the hyperphagic and growth hormone stimulating effects of acute and chronic ghrelin administration have been shown to be inhibited in obese mice fed an HFD for 20 weeks. 27 This study does not allow the effect of obesity versus high-fat feeding to be distinguished. However, coupled with our data, it suggests that ghrelin s hyperphagic effect is inhibited by an HFD and not obesity. Human data also suggest that ghrelin may not always drive a shift toward a state of positive energy balance. A missense mutation in the ghrelin receptor has been identified in 2 unrelated families that is associated with short stature and late-onset obesity. 28 Another single nucleotide polymorphism, which may lead to increased expression of the ghrelin receptor, has been found in patients with low body weight and in patients who lost the greatest amount of body weight during the Finnish Diabetes Prevention Study. 29 However, a number of other studies have failed to find an association between polymorphisms of the ghrelin gene and obesity, but the functional significance of these mutations has not been detailed. 30,31 Data from these studies are in opposition to the assumption that ghrelin regulates appetite by increasing the drive to eat such that increased ghrelin signalling leads to obesity, whereas reduced ghrelin signalling leads to hypophagia and leanness. From our data, we conclude that, under normal feeding conditions, the ghrelin system has an important physiologic role in the control of food intake. These data are consistent with a role for ghrelin in meal initiation and as a signal of energy deficit. However, in the presence of excess dietary fat, the function of the ghrelin system is altered so that endogenous ghrelin no longer produces a potent stimulatory effect on food intake and may inhibit it. The mechanisms by which excess dietary fats alter the function of ghrelin are unknown and require further investigation. They are likely to be complex given that we found high-fat feeding did not alter food preference, hypothalamic neuropeptide function, or the form of acyl group bonded to ghrelin. The ghrelin system is a therapeutic target for the development of obesity. Our data suggest that the success of these approaches could critically depend on the constitution of the diet patients ingest. Indeed, obese patients usually eat diets rich in high-fat food, precisely the type of diet that we have found may interact with the ghrelin system to alter its function. Supplementary Material Note: To access the supplementary material accompanying this article, visit the online version of Gastroenterology at and at doi: /j.gastro References 1. Kojima M, Hosoda H, Date Y, et al. Ghrelin is a growth-hormonereleasing acylated peptide from stomach. Nature 1999;402: Tschop M, Smiley DL, Heiman ML. Ghrelin induces adiposity in rodents. Nature 2000;407: Asakawa A, Inui A, Kaga T, et al. Ghrelin is an appetite-stimulatory signal from stomach with structural resemblance to motilin. Gastroenterology 2001;120: Wren AM, Small CJ, Abbott CR, et al. Ghrelin causes hyperphagia and obesity in rats. Diabetes 2001;50: Sun Y, Ahmed S, Smith RG. Deletion of ghrelin impairs neither growth nor appetite. Mol Cell Biol 2003;23: Wortley KE, Anderson K, Garcia K, et al. Deletion of ghrelin reveals no effect on food intake, but a primary role in energy balance. Obes Res 2004;12: Wortley KE, del Rincon JP, Murray JD, et al. Absence of ghrelin protects against early-onset obesity. J Clin Invest 2005;115: Zigman JM, Nakano Y, Coppari R, et al. Mice lacking ghrelin receptors resist the development of diet-induced obesity. J Clin Invest 2005;115:

9 2476 GARDINER ET AL GASTROENTEROLOGY Vol. 138, No Sun Y, Butte NF, Garcia JM, et al. Characterization of adult ghrelin and ghrelin receptor knockout mice under positive and negative energy balance. Endocrinology 2008;149: Bewick GA, Kent A, Campbell D, et al. Mice with hyperghrelinemia are hyperphagic, glucose intolerant and have reduced leptin sensitivity. Diabetes 2009:58: Dezaki K, Sone H, Yada T. Ghrelin is a physiological regulator of insulin release in pancreatic islets and glucose homeostasis. Pharmacol Ther 2008;118: Liu YL, Semjonous NM, Murphy KG, et al. The effects of pancreatic polypeptide on locomotor activity and food intake in mice. Int J Obes (Lond) 2008;32: Bewick GA, Dhillo WS, Darch SJ, et al. Hypothalamic cocaine- and amphetamine-regulated transcript (CART) and agouti-related protein (AgRP) neurons coexpress the NOP1 receptor and nociceptin alters CART and AgRP release. Endocrinology 2005;146: Patterson M, Murphy KG, Le Roux CW, et al. Characterization of ghrelin-like immunoreactivity in human plasma. J Clin Endocrinol Metab 2005;90: Sell H, Deshaies Y, Richard D. The brown adipocyte: update on its metabolic role. Int J Biochem Cell Biol 2004;36: Salehi A, Dornonville de la CC, Hakanson R, et al. Effects of ghrelin on insulin and glucagon secretion: a study of isolated pancreatic islets and intact mice. Regul Pept 2004;118: Dezaki K, Hosoda H, Kakei M, et al. Endogenous ghrelin in pancreatic islets restricts insulin release by attenuating Ca2 signaling in -cells: implication in the glycemic control in rodents. Diabetes 2004;53: Reed JA, Benoit SC, Pfluger PT, et al. Mice with chronically increased circulating ghrelin develop age-related glucose intolerance. Am J Physiol Endocrinol Metab 2008;294:E752 E Sun Y, Asnicar M, Saha PK, et al. Ablation of ghrelin improves the diabetic but not obese phenotype of ob/ob mice. Cell Metab 2006;3: Chen HY, Trumbauer ME, Chen AS, et al. Orexigenic action of peripheral ghrelin is mediated by neuropeptide Y and agoutirelated protein. Endocrinology 2004;145: Kamegai J, Tamura H, Shimizu T, et al. Chronic central infusion of ghrelin increases hypothalamic neuropeptide Y and Agouti-related protein mrna levels and body weight in rats. Diabetes 2001;50: Yang J, Brown MS, Liang G, et al. Identification of the acyltransferase that octanoylates ghrelin, an appetite-stimulating peptide hormone. Cell 2008;132: Gutierrez JA, Solenberg PJ, Perkins DR, et al. Ghrelin octanoylation mediated by an orphan lipid transferase. Proc Natl Acad Sci U S A 2008;105: Nishi Y, Hiejima H, Hosoda H, et al. Ingested medium-chain fatty acids are directly utilized for the acyl modification of ghrelin. Endocrinology 2005;146: Kirchner H, Gutierrez JA, Solenberg PJ, et al. GOAT links dietary lipids with the endocrine control of energy balance. Nat Med 2009;15: Wren AM, Small CJ, Abbott CR, et al. Ghrelin causes hyperphagia and obesity in rats. Diabetes 2001;50: Iwakura H, Akamizu T, Ariyasu H, et al. Effects of ghrelin administration on decreased growth hormone status in obese animals. Am J Physiol Endocrinol Metab 2007;293:E819 E Pantel J, Legendre M, Cabrol S, et al. Loss of constitutive activity of the growth hormone secretagogue receptor in familial short stature. J Clin Invest 2006;116: Mager U, Degenhardt T, Pulkkinen L, et al. Variations in the ghrelin receptor gene associate with obesity and glucose metabolism in individuals with impaired glucose tolerance. PLoS One 2008;3:e Garcia EA, Heude B, Petry CJ, et al. Ghrelin receptor gene polymorphisms and body size in children and adults. J Clin Endocrinol Metab 2008;93: Gueorguiev M, Lecoeur C, Meyre D, et al. Association studies on ghrelin and ghrelin receptor gene polymorphisms with obesity. Obesity (Silver Spring) 2009;17: Received August 10, Accepted February 11, Reprint requests Address requests for reprints to: Stephen Bloom, MD, Department of Investigative Medicine, Hammersmith Campus, Imperial College London, Du Cane Rd, London W12 0NN, United Kingdom. s.bloom@imperial.ac.uk; fax: (44) Acknowledgments These authors contributed equally to this manuscript. Conflicts of interest The authors disclose no conflicts. Funding Supported by grants from the MRC (G ), Wellcome Trust (072643/Z/03/Z), and by an EU FP6 Integrated Project Grant LSHM-CT ; from the NIHR Biomedical Research Centre funding scheme and an IMB Capacity building award; from funding by an MRC studentship (to A.K.); and by the Diabetes Research and Wellness foundation (as a nonclinical Fellow to G.A.B.).

10 June 2010 HIGH-FAT DIETS INHIBIT GHRELIN HYPERPHAGIA 2476.e1 Supplementary Figure 1. A circulating acylated ghrelin measured in male mice fed an HFD n 9. Male mice n 6 were fed either RC or an HFD from 5 weeks of age and placed in the CLAMS. The ClAMS measured B Lococmotor activity, C energy expenditure and D respiratory quotient. E Brown adipose tissue uncoupling protein mrna was measured at 16 weeks of age in male mice n 6 fed either RC or HFD from 5 weeks of age. Data are represented mean SEM, (*P 0.05).