Introduction

The hypothalamic-pituitary-thyroid (HPT) axis and the hypothalamic-pituitary-adrenal (HPA) axis represent major endocrine systems that participate in the regulation of energy balance (1, 2). Because leptin signaled these endocrine axes (3) that a positive energy balance exists, we speculated that ghrelin signals those axes when a negative energy balance is experienced (4). Ghrelin, a secreted 28-residue peptide hormone from the stomach, is an endogenous growth hormone secretagogue (GHS) (5). Its orexigenic and anabolic actions, however, are independent from its ability to stimulate growth hormone (GH) secretion from the pituitary (6). Ghrelin induces adiposity in rodents by increasing food intake and decreasing fat use (6). More recently, ghrelin has also been shown to induce appetite and increase food intake in humans (7). Although ghrelin receptors are expressed in peripheral organs, such as adipose tissue and the pancreas, ghrelin regulation of energy balance is believed to be mediated by the hypothalamus (4). Ghrelin regulates hypothalamic neuropeptide expression that is also influenced by leptin, although in an opposite manner (4). These neurons that are well known for maintaining energy homeostasis are also known to be located in close proximity to or projecting to hypophysiotropic neurons, such as those producing corticotropin-releasing hormone (CRH) or thyreotrophin-releasing hormone (TRH) in the paraventricular nucleus (8, 9). In fact, ghrelin and other GHSs have been shown to stimulate the HPA axis and to decrease circulating thyroid-stimulating hormone (TSH) levels in rats (10). Both mechanisms can induce a positive energy balance (11, 12). There is also evidence for CRH influencing energy balance, which can occur through activation of either the HPA axis or the sympathetic nervous system, depending on the specific projection of the paraventricular neurons producing CRH (3, 8). To investigate if ghrelin-induced adiposity is mediated by changes in HPT or HPA, we studied the effect of ghrelin administration on determinants of energy balance in hypophysectomized, adrenalectomized, and thyroidectomized rats. To investigate if gastric ghrelin secretion is regulated by pituitary hormones, we studied the effect of hypophysectomy on circulating ghrelin levels. We also studied the effect of administration of GH, insulin-like growth factor I (IGF-I), and GH-releasing peptide-6 (GHRP-6) on circulating rat ghrelin levels to further specify the mechanisms of a possible gastro-hypophyseal feedback loop.

Research Methods and Procedures

Twelve-week-old male Sprague-Dawley rats (Taconic, Germantown, NY) were used for all experiments. Fourteen-week-old growth hormone–deficient dwarf rats (6) were purchased at Harlan UK (Shaw's Farm Blackthorn, Bicester, Oxon, England). For all surgical procedures, rats were anesthetized with sodium pentobarbital, hair around the skin area where the incision was planned was shaved, and skin was scrubbed with 70% alcohol. Adrenalectomy of rats was performed on both sides, while bilateral sham-adrenalectomy was performed exactly the same way without removing the adrenal glands. Complete absence of adrenal function was confirmed by showing complete absence of plasma corticosterone immunoreactivity (Linco Research Assay Services, St. Charles, MO) at the end of the study. The drinking water of adrenalectomized rats was replaced with a 0.9% sodium chloride solution instead to prevent hypoaldosteronism-induced hyponatremia.

For thyroidectomy (with parathyroid transplant), special care was taken not to disturb the recurrent laryngeal nerve. The thyroid glands were placed in a Petri dish where the parathyroids were gently dissected away and placed back into the thoracic cavity alongside the trachea. The muscle was approximated and salivary and lymphatic tissue was replaced. Sham-thyroidectomies were performed identically, but the thyroid was not disturbed.

For hypophysectomy, the animal was mounted on a Hoffman-Reiter hypophysectomy instrument, which had the ear bar angle and needle length calibrated to the animal's weight. The needle, fitted to a glass syringe, was inserted through the hollow right ear bar, and the pituitary was removed with gentle aspiration. The aspirate was discharged into a Petri dish and inspected to insure completeness of removal. Complete removal of the pituitary was additionally confirmed by demonstration of plasma GH levels lower than the detection limit of an established radioimmunoassay (Linco Research Assay Services). Sham-hypophysectomy was performed in an identical manner, with the exception that no aspiration took place. All hypophysectomized and sham-hypophysectomy animals received a 5% sucrose solution ad libitum in addition to drinking water.

All studies were performed between days 8 and 18 after at least 1 week of recovery after surgery. This time frame was chosen to study the lack of pituitary, adrenal, or thyroid function in a period where it is acute and not yet fully adapted, but the animals have completely recovered from surgery and hormonal deficiencies have not caused chronic changes, sickness, or altered behavior. Animals were closely monitored, and only rats exhibiting normal feeding behavior and no signs of disease were used in experiments. The study protocols used in these experiments were approved by the Animal Care Committee of Eli Lilly and Co. All animal experiments were conducted in accordance with the principles and procedures outlined in the National Institute of Health Guide for the care and use of laboratory animals.

Subcutaneous drug administration was performed shortly before the start of the dark/light cycle (12-hour dark period, 12-hour light period), and animals were housed in a stress-free environment with constant temperature and humidity conditions. All rats were fed a normal rodent diet (Nutrition International 5001 Rodent Diet; Purina Mills Inc., St. Louis, MO).

Ghrelin synthesis and purification was performed at Lilly Research Laboratories. The ghrelin (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28) sequence was synthesized on 0.25-mmol scale with an ABI 433 peptide synthesizer (Applied Biosystems, Foster City, CA) using the standard 90-minute N,N'-dicyclohexylcarbodiimide/N-hydroxybenzotriazole single coupling cycles. All prepackaged reagents for the instrument were purchased from Applied Biosystems, the protected amino acids from Midwest Biotech (Fishers, IN), and the Rink amide resin with indicated substitution of 0.53 mmol/g from Novabiochem (San Diego, CA). The protecting group scheme was as follows: fluorenylmethoxycarbonyl (Fmoc)-Arg(2,2,4,6,7-pentamethyldihydrobenzofuran-5-sulfonyl), Fmoc-Gln(trityl), FmocGlu(O-tert butyl), Fmoc-His(trityl), Fmoc-Lys(tert butyloxycarbonyl), Fmoc-Ser(O-tert butyl) (positions 2,6,18) Fmoc-Ser (position 3) with the N-terminal glycine incorporated in its butyloxycarbonyl (Boc) protected form. Esterification of the free hydroxyl side-chain of serine at position 3 was accomplished by treatment with 12.0 equivalents each of octanoic acid and N,N,'-diisopropylcarbodiimide in the presence of 0.3 equivalents of 4-(dimethylamino)pyridine in N-methylpyrrolidone. The reaction seemed complete after 4.0 hours. Cleavage from solid support and simultaneous side-chain deprotection was accomplished by treatment with 95/5% (vol/vol) trifluoroacetic acid (TFA)/triisopropylsilane for 2 hours at ambient temperature. Analytical and preparative high-performance liquid chromatography conditions were as follows: a Zorbax C18 column (buffer A = 0.1% TFA/H₂O, buffer B = 0.1% TFA/CH₃CN); gradient, 0% to 65% B over 4 minutes (analytical mode) and 60 minutes (preparative mode); and detection set to 214 nm. Final characterization was provided by high-performance liquid chromatography, electrospray/mass spectrometry, and liquid chromatography/mass spectrometry.

GH was synthesized at Lilly Research Laboratories as described earlier (13), and IGF-I was purchased at GroPep Limited (Thebarton, Australia). GHRP-6 has been purchased at Bachem/Peninsula Laboratories, Inc. (San Carlos, CA). Plasma rat ghrelin was measured using a commercial radioimmunoassay (Phoenix Pharmaceuticals Inc., Belmont, CA) as described earlier (14).

Body composition was analyzed by chemical methods at Covance Laboratories (Covance Inc., Princeton, NJ) at the end of treatment on frozen carcasses. All values are the mean SEM, or percentage of mean change and SEM relative to controls. Means were compared by ANOVA and Tukey's post hoc test using SIGMA-stat software.

Results

Neither Administration of GHRP-2 nor Administration of Ghrelin Increases Body Weight of Hypophysectomized Rats

Nine days of treatment with the ghrelin receptor agonist GHRP-2 (8 mg/kg) increased body weight by 21.2 g in sham-hypophysectomized rats (n = 6 each group, p = 0.001; 229% more weight gain than vehicle injected controls, +9.1 g; Figure 1). This was expected because we have shown induction of a adiposity by GHRP-2 in mice (15). It also had been shown that ghrelin and other GHSs stimulate body weight gain in rats and mice (4, 6, 10, 16). Therefore, we did not repeat this arm of the experiment with expensive bioactive ghrelin peptide. In hypophysectomized animals, no increase in body weight was found during treatment with GHRP-2 (n = 6 each group, p = 0.77, 8 mg/kg) or ghrelin (8 mg/kg, n = 6 each group, p = 0.88) after 9 days of daily injections compared with vehicle-injected hypophysectomized control rats (Figure 1).

Figure 1:.

Daily administration of the ghrelin receptor agonist GH-releasing peptide-2 (GHRP-2) or ghrelin does not increase body weight of hypophysectomized rats. GHRP-2–treated normal rats, however, gain body weight (n = 5 each group, p = 0.001).

Full figure and legend (95K)

Administration of Ghrelin Increases Body Weight, Food Intake, and Fat Mass but Does Not Change Protein Mass in Normal Sprague-Dawley Rats after Undergoing Sham-Adrenalectomy Surgery

Daily administration of two different ghrelin doses for 7 days dose-dependently increased food intake [0.8 mg/kg ghrelin, 9 7% more food intake than vehicle-injected controls (not significant), 8 mg/kg ghrelin: 15 2% more food intake than vehicle-injected controls; p = 0.04; n = 5 each group; Figure 2A], body weight (0.8 mg/kg ghrelin, 78 26% more body weight gain than vehicle-injected controls, p = 0.1; 8 mg/kg ghrelin, 149 23% more body weight gain than vehicle-injected controls, p = 0.003; n = 5 each group; Figure 2B), and body fat mass [0.8 mg/kg ghrelin, 18 10% more total fat mass than vehicle-injected controls (not significant); 8 mg/kg ghrelin, 33 4% more total fat mass than vehicle-injected controls, p = 0.02; n = 5 each group; Figure 2C]. Daily administration of either ghrelin dose did not alter body protein content [0.8 mg/kg ghrelin, -0.7 2% change in protein content compared with vehicle-injected controls (not significant); 8 mg/kg ghrelin: -3 0.9% change in protein content compared with vehicle-injected controls (not significant), p = 0.4; n = 5 each group].

Figure 2:.

Ghrelin induces a dose-dependent increase in food intake, body weight, and body fat in sham-adrenalectomized controls but not in adrenalectomized or thyroidectomized rats when compared with vehicle-injected controls (nine groups of n = 5). Data are presented as percentage increase relative to changes in matched vehicle-treated control rats. One week of daily subcutaneous ghrelin injections in sham-operated rats causes a dose-dependent increase of (A) average daily food intake (p = 0.04), (B) body weight gain (p = 0.003), and (C) total fat mass (p = 0.02) compared with matched vehicle-treated sham-surgery control rats. In adrenalectomized rats, ghrelin only tended to increase (D) average daily food intake, (E) body weight gain, and (F) total-body fat mass (p > 0.07). In thyroidectomized rats, ghrelin did not change (G) average daily food intake, (H) body weight gain, and (I) total fat mass, (p > 0.07). *Represents statistically significant difference.

Full figure and legend (103K)

Administration of Ghrelin Does Not Increase Body Weight, Food Intake, or Fat Mass in Adrenalectomized Sprague-Dawley Rats

Neither dose of synthetic ghrelin for 7 days significantly increased total food intake [0.8 mg/kg ghrelin, 10 5% more total food intake than vehicle-injected controls; 8 mg/kg ghrelin, 5 6% more total food intake than vehicle-injected controls (not significant), p = 0.47; n = 5 each group; Figure 2D], body weight [0.8 mg/kg ghrelin, 55 15% more body weight gain than vehicle-injected controls; 8 mg/kg ghrelin, 55 20% more body weight gain than vehicle-injected controls (not significant), p = 0.07; n = 5 each group; Figure 2E], or body fat mass [0.8 mg/kg ghrelin, 23 8% more total fat mass than vehicle-injected controls (not significant); 8 mg/kg ghrelin, 20 12% more total fat mass than vehicle-injected controls (not significant), p = 0.3; n = 5 each group; Figure 2F] in adrenalectomized rats compared with vehicle-injected controls. Daily administration of two different doses of synthetic ghrelin in adrenalectomized rats for 7 days did not significantly reduce body protein content [0.8 mg/kg ghrelin, -3 1% change in protein content compared with vehicle-injected controls (not significant); 8 mg/kg ghrelin, -2 2% change in protein content compared with vehicle-injected controls, p = 0.4; n = 5 each group). No significant difference in food intake, body weight, or body fat mass was observed between vehicle-treated sham-adrenalectomized and vehicle-treated adrenalectomized rats (p > 0.05). Carcass analysis did not reveal any differences for content of water between treated groups and vehicle-injected controls in adrenalectomized or sham rats.

Administration of Ghrelin Does Not Increase Body Weight, Food Intake, or Fat Mass in Thyroidectomized Sprague-Dawley Rats

Daily administration of two different doses of synthetic ghrelin in thyroidectomized rats for 7 days did not significantly change food intake (0.8 mg/kg ghrelin, -3 3% less food intake than vehicle-injected controls, p = 0.08; 8 mg/kg ghrelin, +4 3% more food intake than vehicle-injected controls, p = 0.07; n = 5 each group), body weight (0.8 mg/kg ghrelin, -6 44% less body weight gain than vehicle-injected controls, p = 0.2; 8 mg/kg ghrelin, +68 24% more body weight gain than vehicle-injected controls, p = 0.27; n = 5 each group), and body fat mass [0.8 mg/kg ghrelin, +2 2% more total fat mass than vehicle-injected controls (not significant); 8 mg/kg ghrelin, +12 5% more total fat mass than vehicle-injected controls (not significant), p = 0.36; n = 5 each group] compared with vehicle-injected controls. Daily administration of two different doses of synthetic ghrelin in thyroidectomized rats for 7 days did not change body protein content [0.8 mg/kg ghrelin, -0.9 1% change in protein content compared with vehicle-injected controls (not significant); 8 mg/kg ghrelin, -4 1% change in protein content compared with vehicle-injected controls (not significant), p = 0.1; n = 5 each group]. No significant difference in food intake, body weight, or body fat mass was observed between vehicle-treated sham-thyroidectomized and vehicle-treated thyroidectomized rats (p > 0.05). Carcass analysis did not reveal any differences for content of water between treated groups and vehicle-injected controls in thyroidectomized or sham rats.

Plasma Ghrelin Concentrations in Hypophysectomized Rats Are 3-Fold Higher Compared with SHAM Surgery Controls

Four weeks after surgical removal of the pituitary, plasma ghrelin levels in (untreated) hypophysectomized rats (5.92 0.8 ng/mL) were significantly higher than plasma ghrelin levels in sex- and age-matched sham-operated controls (1.92 0.18 ng/mL, p < 0.001, n = 20 each group) or normal healthy controls (1.69 0.13 ng/mL). Plasma ghrelin levels of GH-deficient dwarf rats (n = 6), however, did not differ significantly from those of normal control rats of the same age (Figure 3A).

Figure 3:.

Effect of changes in the concentration of somatotropic hormones growth hormone (GH), insulin-like growth factor I (IGF-I), and GH-releasing peptide-6 (GHRP-6) on rat plasma ghrelin levels. (A) Hypophysectomy acutely increases circulating plasma ghrelin levels in rats compared with plasma ghrelin levels of sham-hypophysectomized or normal rats (p = 0.001, n = 20 each group). Plasma ghrelin levels of dwarf rats are not significantly different from those of normal controls. (B) A single injection of GH or (C) GHRP-6 is followed by a significant decrease of circulating ghrelin levels in healthy Sprague-Dawley rats, whereas a single injection of (D) IGF-I is not followed by changes of plasma ghrelin levels in normal Sprague-Dawley rats. *Represents statistically significant difference.

Full figure and legend (115K)

Administration of GH, but Not Administration of IGF-I, Decreases Circulating Levels of Ghrelin in Normal Sprague-Dawley Rats

In an additional series of experiments, we studied the effect of the administration of GH, the growth hormone secretagogue GHRP-6, and IGF-I on circulating ghrelin levels in Sprague-Dawley rats (age 12 weeks) to determine if a potential feedback signal acts on a GHS, IGF-I, or only GH-level. Thirty normal 12-week-old Sprague-Dawley rats were injected with either 100 g of GHRP-6, GH, or IGF-I. After 30, 60, 90, 120, 150, and 180 minutes, five rats from each group were killed, and trunk blood was collected to quantify plasma ghrelin concentrations (radioimmunoassay). A singular subcutaneous injection of 100 g GH decreased plasma ghrelin levels from 0.81 0.06 to 0.41 0.06 ng/mL (after 60 minutes; p = 0.009; n = 5 each group). Three hours after injection, plasma ghrelin levels returned to 0.78 0.11 ng/mL (Figure 3B). A singular subcutaneous injection of 100 g GHRP-6 decreased plasma ghrelin levels from 0.81 0.06 to 0.54 0.06 ng/mL (after 60 minutes; p = 0.027; n = 5 each group). Three hours after injection, plasma ghrelin levels returned to 0.88 0.06 ng/mL (Figure 3C). A singular injection of IGF-I, however, did not alter circulating plasma ghrelin levels significantly (p > 0.09; n = 5 at each time-point; Figure 3D).

Discussion

Ghrelin activates neurons in the hypothalamic arcuate nucleus that release both neuropeptide Y and Agouti-related protein (4). Hypothalamic neurons containing both neuropeptides are known to express GHSR and act as a regulatory center for appetite and energy expenditure, influencing body weight and body composition (2, 4). Accordingly, ghrelin-induced weight gain and adiposity seems to be based on two principal mechanisms: increased nutritional intake and metabolic changes (4, 6). These effects are independent from ghrelin-induced GH secretion (6, 16). Ghrelin, however, is known to also influence the secretion of other pituitary hormones; it stimulates adrenocorticorticotropic hormone secretion (10, 17) and decreases TSH release (10). We therefore investigated if pituitary-derived or peripheral hormones representing either the HPA or the HPT axis would be involved or even essential for ghrelin's ability to induce an increase in food intake and a decrease in fat use, which ultimately leads to obesity. The data presented here prove that the pituitary, as well as the adrenal and thyroid glands, plays a critical role in this process. Hypophysectomized rats, thyroidectomized rats, and to a lesser degree, adrenalectomized rats seem to be resistant to orexigenic and anabolic actions of ghrelin or GHRP-2, a GHSR agonist with a longer half-life. These findings concur with recent data showing that hypothalamic TRH and CRH neurons interact with hypothalamic centers regulating energy balance (8, 9). The codependence of the adrenal and thyroid neuroendocrine axes perpetuates the concept of ghrelin as an endogenous counterpart to leptin such that, together, they participate in achieving energy homeostasis by counter-regulating the very same regulatory structures (4). Dynamics and the phenotype of ghrelin-induced adiposity in rodents are intriguingly similar to body compositional changes caused by hypercorticosteronism (18). Decreased activity of the HPT axis represents another possible cause for a positive energy balance leading to obesity (19). Leptin has been shown to stimulate the HPT axis through an increase of hypothalamic TRH expression (20), whereas the orexigenic agent NPY decreases hypothalamic TRH expression (21). Therefore, it does not seem surprising that increased activity of the HPA axis and decreased activity of the HPT axis seem to be efferent mediating elements in the mediation of ghrelin-induced adiposity (Figure 4).

Figure 4:.

Schematic overview proposing a novel role for the pituitary in (A) the mediation of ghrelin-induced adiposity and (B) the regulation of gastric ghrelin secretion. Injection of synthetic ghrelin (A) stimulates through hypothalamic, and possibly direct, effects at the pituitary, the somatotropic axis, and the hypothalamic-pituitary-adrenal (HPA) axis, whereas it seems to decrease activity of the hypothalamic-pituitary-thyroid (HPT) axis. That way, the balance of lipolytic effects of growth hormone (GH) and thyroid hormones with antilipolytic effects of glucocorticoids may be shifted toward increased fat mass. Secretion of endogenous ghrelin (B) seems to be under the control of somatotropic hormones. These findings suggest the possible existence of a negative feedback loop between the stomach and the pituitary and possible additional negative (auto-)feedback of activated growth hormone secretagogue receptors (GHSRs) on gastric ghrelin secretion.

Full figure and legend (195K)

Ghrelin has originally been identified as an endogenous ligand of the GHSR and is undoubtfully a powerful stimulator of GH secretion (5, 17). GH, however, is known to reduce fat mass and increase muscle mass (21). It therefore seems that ghrelin action induces complex—even antipodal—endocrine and metabolic changes, which finally result in the generation of a positive energy balance. Extrapolation according to knowledge on leptin physiology also suggests that not only the endocrine pathway, but also additional pathways, might be triggered by ghrelin, stimulating adiposity. These pathways are very likely to include the sympathetic as well as the parasympathetic nervous system (1, 2). Based on earlier experiments (6), it is likely that the higher dose of ghrelin (8 mg/kg) administered in this study, are likely to transitorily generate circulating plasma ghrelin or GHRP-2 concentrations in a supraphysiological or even pharmacological range. During the majority of the treatment period, and especially in the group that was treated with a lower dose (0.8 mg/kg), the average ghrelin or ghrelin receptor agonist levels are not expected to exceed physiological plasma concentrations. Further studies on the pharmacokinetics of ghrelin are necessary to allow for a better differentiation of physiologically relevant effects from pharmacological or even artificial effects.

At this point, we cannot completely rule out the possibility that mechanisms other than impaired pituitary, HPA, or HPT activity contribute to the prevention of ghrelin-induced adiposity. One mechanism that might contribute to the decreased responsiveness of energy balance to ghrelin is the fact that hypophysectomized rats exhibit about 3-fold increased ghrelin plasma levels (Figure 3). This could reflect increased gastric ghrelin secretion and therefore indicate a gastric-hypophyseal feedback loop. This is in contrast to a recent clinical study that failed to demonstrate increased circulating levels of ghrelin of patients with hypopituitarism (22). However, apart from possible species differences, lack of increased plasma ghrelin levels could be because of the fact that these patients have not been acutely deprived of their pituitary function, and their gastric ghrelin secretion might already have adapted to a new balance. Also, chronic GH deficiency changes body composition and gastric ghrelin secretion in these patients and (22) might have adjusted according to increased fat mass in addition to decreased GH feedback (21). Chronic exposure to increased plasma ghrelin levels might have triggered induction of ghrelin resistance. Such resistance has been postulated for leptin in obese individuals and rodent models with chronic exposure to high levels of leptin (23, 24). Adaptive changes could also be the reason why we did not find increased plasma ghrelin levels in rats with GH deficiency because of a spontaneous mutation ("dwarf rats," Figure 3A). Ghrelin secretion in these rats might have normalized after adaptation to the lack of negative GH feedback during early developmental stages. Regulatory feedback loops are a core principle of endocrine control mechanisms. All four classic endocrine axes, corticotrophic, thyreotrophic, gonadotrophic, and somatotrophic, are carefully balanced by multiple feedback mechanisms on the hypothalamic as well as the pituitary level (25, 26). GH secretion from pituitary cells is under the control of growth hormone-releasing hormone (GHRH) and somatostatin (SRIF) (25, 26). Hypophysectomy causes an increase in the expression and secretion of both GHRH and SRIF because of the removal of the negative feedback signal by GH and its effector-protein IGF-I (26, 27). It seems logical that gastric ghrelin release as an additional endogenous stimulus of GH secretion would undergo an increase. Still, the data shown here (Figure 3) represent to our knowledge the first evidence for a feedback loop connecting a centrally located hormone with a gastrointestinal-releasing hormone. The concept of a regulatory gastro-hypophyseal feedback loop, which may additionally involve hypothalamic circuits, gains further support by our results demonstrating that subcutaneous injection of GH or its secretagogue GHRP-6 acutely decreases circulating rat ghrelin levels (Figure 3B and 3C). Plasma ghrelin levels did not change after injection of IGF-I. We therefore speculate that GH, but not its effector protein IGF-I, represents a crucial feedback signal for ghrelin. Decreased plasma ghrelin levels after administration of GHRP-6 might reflect a secondary effect mediated by a GHRP-triggered increase of pituitary-derived GH. Another explanation would be if the GHSR mediates negative feedback on ghrelin-secreting endocrine cells at the level of the oxyntic glands. Theoretically this would allow ghrelin receptor agonists to decrease ghrelin secretion as well as suggest the possibility of negative auto-feedback of gastric ghrelin on its own secretion (Figure 4B). Once well-characterized, epitope-specific, monoclonal antibodies against ghrelin are available to distinguish bioactive synthetic ghrelin from endogenous ghrelin, this theory can be tested. Additional complexity of the interactions between the classic somatotropic axis and ghrelin derives from the fact that a small amount of ghrelin is produced in pituitary cells (28). However, it remains to be shown if pituitary-derived ghrelin is (patho-)physiologically relevant.

Apart from the findings described here, the only known mechanisms to date to influence ghrelin secretion are acute or chronic changes in energy balance, such as food intake or fasting, and states of anorexia or obesity (4, 14, 15, 29, 30, 31). Because ghrelin seems to be a key player secreted by the stomach into circulation to connect nutritional state with endocrine changes, multiple links seem possible. Increased release of GH and adrenocorticorticotropic hormone from the pituitary during fasting may be triggered by enhanced ghrelin secretion as a response to an empty stomach. Most published studies, as well as the data presented here on ghrelin secretion, are based on the quantification of circulating hormone concentrations using a commercial radioimmunoassay (Phoenix Pharmaceuticals, Inc.). This immunoassay relies on an antiserum that recognizes N-terminal epitopes, and therefore, measures des-octanoylated as well as bioactive ghrelin. It remains unclear which fraction of circulating ghrelin reflects physiologically relevant changes of its bioactivity best. Earlier results produced with this method, such as postprandial decrease of circulating ghrelin levels, have recently been confirmed by alternative immunoassays, as well as at RNA-expression level.

In conclusion, the pituitary seems to be important for the mediation of ghrelin action as previously assumed, even regarding GH-independent effects such as ghrelin-induced adiposity. Such ghrelin-induced adiposity is mediated in part through changes in the activity of the HPA and HPT axis (Figure 4A). In addition, we found evidence for a role of the pituitary in the regulation of ghrelin secretion, supporting a novel concept of a gastro-hypophyseal feedback loop (Figure 4B). The existence of such a feedback loop once again suggests a fascinating role for ghrelin as an endocrine interface between regulatory centers of nutritional intake, metabolic control, and growth regulation.

References

1. Webber, J., Macdonald, I. A. (2000) Signalling in body-weight homeostasis: neuroendocrine efferent signals. Proc Nutr Soc. 59: 397–404. | PubMed | ChemPort |
2. Spiegelman, B. M., Flier, J. S. (2001) Obesity and the regulation of energy balance. Cell. 104: 531–543. | Article | PubMed | ISI | ChemPort |
3. Heiman, M. L., Sloop, K. W., Chen, Y., Caro, J. F. (1999) Extension of neuroendocrine axes to include leptin. J Anim Sci. 77(Suppl 3): 33–42.
4. Horvath, T. L., Diano, S., Sotonyi, P., Heiman, M., Tschöp, M. (2001) Ghrelin and the regulation of energy balance—a hypothalamic perspective. Endocrinology. 142: 4163–4169. | Article | PubMed | ISI | ChemPort |
5. Kojima, M., Hosoda, H., Date, Y., Nakazato, M., Matsuo, H., Kangawa, K. (1999) Ghrelin is a growth-hormone-releasing acylated peptide from stomach. Nature. 402: 656–660. | Article | PubMed | ISI | ChemPort |
6. Tschöp, M., Smiley, D. L., Heiman, M. L. (2000) Ghrelin induces adiposity in rodents. Nature. 407: 908–913. | Article | PubMed | ISI | ChemPort |
7. Wren, A. M., Seal, L. J., Cohen, M. A., et al. (2001) Ghrelin enhances appetite and increases food intake in humans. J Clin Endocrinol Metab. 86: 5992. | Article | PubMed | ISI | ChemPort |
8. Li, C., Chen, P., Smith, M. S. (2000) Corticotropin releasing hormone neurons in the paraventricular nucleus are direct targets for neuropeptide Y neurons in the arcuate nucleus: an anterograde tracing study. Brain Res. 854: 122–129.
9. Mihaly, E., Fekete, C., Tatro, J. B., Liposits, Z., Stopa, E. G., Lechan, R. M. (2000) Hypophysiotropic thyrotropin-releasing hormone-synthesizing neurons in the human hypothalamus are innervated by neuropeptide Y, agouti-related protein, and alpha-melanocyte-stimulating hormone. J Clin Endocrinol Metab. 85: 2596–2603.
10. Wren, A. M., Small, C. J., Ward, H. L., et al. (2000) The novel hypothalamic peptide ghrelin stimulates food intake and growth hormone secretion. Endocrinology. 141: 4325–4328. | Article | PubMed | ISI | ChemPort |
11. Rothwell, N. J. (1990) Central effects of CRF on metabolism and energy balance. Neurosci Biobehav Rev. 14: 263–271. | PubMed | ISI | ChemPort |
12. Silva, J. E. (1995) Thyroid hormone control of thermogenesis and energy balance. Thyroid. 5: 481–492. | PubMed | ISI | ChemPort |
13. Hsiung, H. M., MacKellar, W. C. (1987) Expression of bovine growth hormone derivatives in Escherichia coli and the use of the derivatives to produce natural sequence growth hormone by cathepsin C cleavage. Methods Enzymol. 153: 390–401. | Article | PubMed | ISI | ChemPort |
14. Tschöp, M., Weyer, C., Tataranni, P. A., Devanarayan, V., Ravussin, E., Heiman, M. L. (2001) Circulating ghrelin levels are decreased in human obesity. Diabetes. 50: 707–709. | Article | PubMed | ISI | ChemPort |
15. Tschöp, M., Statnick, M. A., Suter, T., Heiman, M. L. (2002) Growth hormone-releasing peptide-2 (GHRP-2) increases fat mass in mice lacking neuropeptide Y (NPY): indication for a crucial mediating role of hypothalamic Agouti-related protein (AGRP). Endocrinology. 143: 558–568. | Article | PubMed | ISI | ChemPort |
16. Nakazato, M., Murakami, N., Date, Y., et al. (2001) A role for ghrelin in the central regulation of feeding. Nature. 409: 194–198. | Article | PubMed | ISI | ChemPort |
17. Arvat, E., Maccario, M., Di Vito, L., et al. (2001) Endocrine activities of ghrelin, a natural growth hormone secretagogue (GHS), in humans: comparison and interactions with hexarelin, a non-natural peptidyl GHS, and GH-releasing hormone. J Clin Endocrinol Metab. 86: 1169–1174. | Article | PubMed | ChemPort |
18. Rohner-Jeanrenaud, F., Jeanrenaud, B. (1991) Aspects of neuroregulation of body composition and insulin secretion. Int J Obes Relat Metab Disord. 15(Suppl 2): 117–122.
19. Legradi, G., Emerson, C. H., Ahima, R. S., Flier, J. S., Lechan, R. M. (1997) Leptin prevents fasting-induced suppression of prothyrotropin-releasing hormone messenger ribonucleic acid in neurons of the hypothalamic paraventricular nucleus. Endocrinology. 138: 2569–2576. | Article | PubMed | ISI | ChemPort |
20. Fekete, C., Kelly, J., Mihaly, E., et al. (2001) Neuropeptide Y has a central inhibitory action on the hypothalamic-pituitary-thyroid axis. Endocrinology. 142: 2606–2613.
21. Salomon, F., Cuneo, R. C., Hesp, R., Sonksen, P. H. (1989) The effects of treatment with recombinant human growth hormone on body composition and metabolism in adults with growth hormone deficiency. N Engl J Med. 321: 1797–1803. | PubMed | ISI | ChemPort |
22. Janssen, J. A., van Der Toorn, F. M., Hofland, L. J., et al. (2001) Systemic ghrelin levels in subjects with growth hormone deficiency are not modified by one year of growth hormone replacement therapy. Eur J Endocrinol. 145: 711–716. | Article | PubMed | ISI | ChemPort |
23. Considine, R. V., Sinha, M. K., Heiman, M. L., et al. (1996) Serum immunoreactive-leptin concentrations in normal-weight and obese humans. N Engl J Med. 334: 292–295. | Article | PubMed | ISI | ChemPort |
24. Frederich, R. C., Hamann, A., Anderson, S., Lollmann, B., Lowell, B. B., Flier, J. S. (1995) Leptin levels reflect body lipid content in mice: evidence for diet-induced resistance to leptin action. Nat Med. 1: 1311–1314. | Article | PubMed | ISI | ChemPort |
25. Widmaier, E. P. (1992) Metabolic feedback in mammalian endocrine systems. Horm Metab Res. 24: 147–153.
26. Vance, M. L. (1994) Hypopituitarism. N Engl J Med. 330: 1651–1662.
27. Pombo, M., Pombo, C. M., Garcia, A., et al. (2001) Hormonal control of growth hormone secretion. Horm Res. 55(Suppl 1): 11–16. | Article | PubMed | ChemPort |
28. Korbonits, M., Kojima, M., Kangawa, K., Grossman, A. B. (2001) Presence of ghrelin in normal and adenomatous human pituitary. Endocrine. 14: 101–104.
29. Tschöp, M., Wawarta, R., Riepl, R. L., et al. (2001) Post-prandial decrease of circulating human ghrelin levels. J Endocrinol Invest. 24: RC19–RC21. | PubMed | ISI | ChemPort |
30. Otto, B., Cuntz, U., Fruehauf, E., et al. (2001) Weight gain decreases elevated plasma ghrelin concentrations of patients with anorexia nervosa. Eur J Endocrinol. 145: 669–673. | Article | PubMed | ChemPort |
31. Cummings, D. E., Purnell, J. Q., Frayo, R. S., Schmidova, K., Wisse, B. E., Weigle, D. S. (2001) A preprandial rise in plasma ghrelin levels suggests a role in meal initiation in humans. Diabetes. 50: 1714–1719. | Article | PubMed | ISI | ChemPort |
32. Cummings, D. E., Weigle, D. S., Frayo, R. S., et al. (2002) Plasma ghrelin levels after diet-induced weight loss or gastric bypass surgery. N Engl J Med. 346: 1623–1630. | Article | PubMed | ISI |

Acknowledgments

No outside funding/support was provided for this study. We thank Andrea Lyn Woodson, Martin Bidlingmaier, Christian J. Strasburger, Michael Statnick, Paul Burn, and Jose Caro for critical review of the manuscript.