| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
National Institutes of Health National Institute of Diabetes and Digestive and Kidney Diseases Bethesda, Maryland 20892
Address all correspondence and requests for reprints to: Karen Teff, Ph.D., Monell Chemical Senses Center, 3500 Market Street, Philadelphia, Pennsylvania 19104. E-mail: kteff{at}pobox.upenn.edu.
It is not accidental that all phenomena of human life are dominated by the search for daily bread ... . Precise knowledge of what happens to food entering the organism must be the subject of ideal physiology, the physiology of the future.
—Ivan Pavlov, Nobel Lecture, 12 December 1904
At the turn of the last century, the Russian physiologist Pavlov demonstrated that the release of enzymes and secretions necessary for the digestion and metabolism of ingested food could be conditioned, i.e. learned (1). In the now classic experiment, Pavlov paired the administration of food to a fasted dog with various signals or cues, including the sound of a metronome. After repeated presentations of the food paired with the metronome, he demonstrated that the sound alone, independent of food presentation, could elicit the release of physiological responses required for nutrient digestion. Thus, the dog "learned" that the bell signaled the onset of food and responded in a physiologically adaptive manner by anticipating the metabolic responses required for nutrient digestion and metabolism. In 1904, the identified learned metabolic responses included the secretion of saliva, gastric acid, and pancreatic enzymes, and by the end of the century, it was recognized that insulin release could also be conditioned (2). Now, over 100 yr later, Drazen et al. report in the current issue of Endocrinology (3) that the release of a newly identified hormone, ghrelin, thought to be involved in meal initiation and energy homeostasis, can also be entrained or learned. This finding not only broadens the current conceptualization of the physiological role of ghrelin but also highlights an important but often ignored factor contributing to the regulation of food intake and body adiposity: learning.
Ghrelin, the endogenous ligand for the GH secretagogue receptor, is synthesized primarily in the gastrointestinal tract (4). Accumulating evidence supports a role for ghrelin in the regulation of energy balance (5, 6). Exogenous administration of the peptide increases food intake in animals (7) and humans (8), whereas demonstrated rises in ghrelin before meal consumption suggest that increases in ghrelin are indicative of hunger and may be involved in the initiation of meals (9). Subsequent declines during the postprandial period indicate that ghrelin is inhibited during energy replete states, and whereas nutrient-specific suppression has been postulated (10), the magnitude of suppression appears to be related to caloric load (11). However, in the study by Drazen et al. (3), the premeal rise in ghrelin is shown not to be solely a consequence of energy depletion, but instead to have a learned component related to the timing of meal availability. To demonstrate the anticipatory nature of the ghrelin response, the authors trained rats to ingest all of their daily food within a 4-h period (meal-trained rats). By the end of 14 d, the meal-trained animals were shown to ingest more food during the 4-h period compared with ad libitum-fed animals fasted for an identical amount of time. Plasma ghrelin levels were then measured in the meal-trained rats and compared with ad libitum-fed animals. The authors demonstrated that the meal-trained rats exhibited a significant rise in ghrelin that started 2 h before the time of the expected meal and peaked 30 min before food availability. In contrast, the ad libitum-fed rats that were not expecting a meal during the middle of the day, but were equally food deprived, exhibited no significant rise in plasma ghrelin levels. Thus, the meal-trained animals learned the time at which food was going to be made available and were physiologically prepared to ingest that food as demonstrated by the anticipatory release of ghrelin. In a separate experiment, the authors demonstrated that the ad libitum-fed animals were also capable of exhibiting anticipatory ghrelin release, but this occurred before the onset of the dark cycle, the time at which these animals normally ate. Interestingly, under these circumstances, the magnitude of ghrelin release was substantially less than that occurring in the 4-h meal-fed animals. These data suggest that the meal-trained animals who had learned to ingest large meals over the 4-h period also released greater amounts of ghrelin in proportion to the size of the meal they expected to ingest. One interpretation of these results is that the meal-fed animals learned to calibrate their physiological response to the size of the expected meal. Alternatively, it is possible that, despite equivalent lengths of food deprivation, the meal-fed animals may be chronically unsatiated due to their decreased body weight. In humans, elevated plasma ghrelin levels are correlated with decreased satiety (8), In fact, elevated ghrelin levels were also observed in a subsequent experiment where, in contrast to the ad libitum-fed animals, ghrelin levels in the meal-trained animals were not suppressed during the postprandial period. Again, although the results suggest learned differences in postprandial ghrelin levels, the contribution of reduced body weight and the possibility that these animals are functioning below their normal body weight set point must be considered.
The authors hypothesize that the anticipatory increase in ghrelin is a preparatory response that potentially facilitates the subsequent digestion, absorption, and metabolism of the large meal to which the meal-trained rats were exposed. The hypothesis is consistent with that put forward in his seminal article, "The Eating Paradox," where Woods (12) draws the analogy between drug tolerance to that of food ingestion such that the disruption of homeostatic mechanisms caused by food intake leads to the development of adaptive or anticipatory responses that allows the organism to better tolerate the physiological challenge of eating. In that article, he postulates that consistent overeating of large meals would lead to larger anticipatory responses to allow the animal (or person) to metabolically compensate for the large influx of nutrients. The current data support the hypothesis of a learned increase in an anticipatory response proportional to the size of the expected meal. What remains unanswered is the physiological significance of the increase in anticipatory ghrelin. To date, the premeal rise in ghrelin is considered a signal to the central nervous system of depleted energy stores. However, if the premeal ghrelin rise is a learned response, the question arises as to whether the increase in preprandial ghrelin improves postprandial nutrient metabolism to facilitate homeostatic mechanisms.
The prototypic example of the importance of an anticipatory response in the optimization of nutrient metabolism is the relationship between cephalic phase and postprandial insulin release. Cephalic phase insulin refers to the vagally mediated, preabsorptive release of insulin that occurs before or at the onset of food intake, independent of coincident increases in blood glucose (13, 14, 15). Verification of mediation by vagal efferent activity is evident by complete blockade of cephalic phase insulin release (CPIR) with the muscarinic antagonist, atropine (16). Although the magnitude of CPIR is small relative to postprandial insulin release (only 5%), its presence is critical for optimal postprandial glucose metabolism (17). In the absence of cephalic phase insulin, postprandial glucose levels are 30% higher in nondiabetic subjects, and postprandial insulin is delayed and blunted (18, 19). These data suggest that the presence of insulin during this very early time period preempts sustained elevations in plasma glucose. With regard to ghrelin, the relationship between preprandial and postprandial ghrelin is not known, and the mechanisms mediating both the release and suppression of the hormone have not been fully elucidated. However, similar to known mechanisms for CPIR, vagal mediation of ghrelin physiology and function has been hypothesized.
Vagal afferents may mediate the ghrelin-initiated food intake response because vagotomy inhibits the stimulatory effect of ghrelin on meal initiation in both animals (20, 21) and humans (22). Vagotomy has also been shown to attenuate the gradual increase in ghrelin after a prolonged 48-h fast (23). Because similar results occur after atropine administration (23), this suggests that vagal efferent activity may mediate a slow rise in ghrelin. Thus, slow activation of tonic vagal efferent activity at the level of the stomach may contribute to increased ghrelin levels. Because sympathetic nervous system activity is gradually suppressed during long-term food deprivation (24), it is also possible that disinhibition by sympathetic activity facilitates the release of ghrelin over this period of time. However, the mechanisms mediating increases in ghrelin during a long-term fast may be completely different than those mediating the meal-to-meal rises typically observed during the course of a day. In the paper by Drazen et al. (3), the anticipatory rise in ghrelin occurs very slowly over a 1.5-h period. These data contrast sharply to the temporal patterning of CPIR that occurs very rapidly, typically peaking within 5 min of meal onset and returning to baseline by 10 min (15). Even under comparable conditioning paradigms as described in the Drazen paper (3), CPIR exhibits a rapid onset (25), suggesting that the mechanisms mediating the release of the two anticipatory responses are very different. However, some studies using an experimental paradigm that is typically used to elicit CPIR demonstrate acute vagal efferent mediation of both ghrelin release and suppression. Thus, using a modified sham-feed that involves human subjects tasting and expectorating food, two studies report modest increases in ghrelin (26, 27). Conversely, other studies demonstrate decreases in plasma ghrelin (27, 28). None of the studies administered atropine to verify vagal mediation of these responses. The contradictory results are mostly likely a function of learned responses; for example, the reported decrease in ghrelin may in fact be learned as well as combined parasympathetic and sympathetic activation. Bidirectional responses have also been reported for the conditioned glucose response (29), and studies on this phenomenon reveal the complexity of autonomic control on hormonal patterning and of learning paradigms, in general.
The demonstration of a learned component to ghrelin responses suggests neural mediation of ghrelin release and possibly suppression. However, as has been the case with CPIR elucidation of the required stimuli, the mechanisms mediating the response and verification of the physiological significance will be gradual due to the complexity of the systems involved. The new data presented by Drazen et al. (3) provide an experimental framework for investigating and uncoupling the differential mechanisms involved in ghrelin release and suppression. Furthermore, the work underscores the importance of learning in food intake behavior and the necessity for incorporating learning as a variable in our investigation of energy homeostasis, what Pavlov termed "the physiology of the future."
 |
Footnotes |
|---|
Received October 24, 2005.
Accepted for publication October 24, 2005.
|
References |
|---|
 Top References |
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
