Integrating the regulation of food intake.

CONTROL OF FOOD INTAKE is complex, involving multiple interconnected pathways and signal elements. A paper by Wang and Kotz (11) in this issue of the American Journal of Physiology-Regulatory, Integrative and Comparative Physiology illustrates some of the complexity and carries implications for how the regulation of food intake is integrated into the daily life of humans and other animals. Wang and Kotz studied the inhibition, by urocortin injected into the lateral septum, of feeding that had been stimulated either by food deprivation or by injection of orexin A into the lateral hypothalamus. Orexin-A and orexin-B are recent additions to the list of signal peptides that contribute to regulation of food intake. First reported four years ago (6, 10), they were so named because they are synthesized only in a small group of neurons in the lateral hypothalamus, a region of the brain long known to be an important contributor to feeding behavior. The initial and subsequent studies showed that these peptides are in fact orexigenic—when injected into the lateral ventricle or the lateral hypothalamus they promote eating. However, it is by no means clear that this is their primary function. In dogs, narcolepsy is inherited in Mendelian fashion as an autosomal dominant trait with complete penetrance. Inactivating mutations of the orexin type 2 receptor causes narcolepsy in dogs (5). Similarly, orexin knockout mice display a narcoleptic phenotype (4), whereas humans with the disease have abnormally low orexin levels in cerebrospinal fluid (13). On the basis of these findings and on the fact that orexin neurons project widely throughout the central nervous system, it has been proposed that the main function of orexin is to lock the organism in the awake state (8). In addition, orexins augment sympathetic vasomotor output, increasing heart rate and blood pressure (1, 2, 7, 9). Such diversity of reported functions is suggestive of an integrative behavioral program that would activate, or at least prime, several components of wakefulness. A program such as this would be expected to involve many brain structures. One such structure is the lateral septum, which is part of the limbic system and receives a large input from the hippocampus. The septum and hippocampus perform high-order integration related to “formation of new, complex cognitive associations” (3). The lateral septum receives input from, and projects monosynaptically to, the lateral hypothalamus. It might thus be expected to provide contextual information to the hypothalamus. Stimulation or lesions of the lateral septum produce reciprocal, although somewhat complex, changes in feeding behavior; stimulation reduces food intake. However, the signaling pathways remain unclear. One candidate is urocortin, a peptide that appears to be the endogenous ligand for the corticotropin-releasing hormone (CRH) type 2 receptor (CRH2R) and is present at high levels in the lateral septum (10). The paper by Wang and Kotz (11) in this issue addresses this problem. This group previously showed that injection of urocortin into the hypothalamic paraventricular nucleus inhibits food intake induced by food deprivation or by neuropeptide Y (12). However, the authors suggested that other sites are probably involved in urocortin-mediated inhibition of food intake. In the present study (11), the authors show that injection of urocortin into the lateral septal area inhibits feeding induced by both food deprivation and injection of orexin A into the lateral hypothalamus. Importantly, this inhibition is blocked by a CRH2R antagonist in the lateral septum, and, as expected (10), urocortin is a more potent agonist than CRH. Equally important, the effect of urocortin is not due to production of a conditioned taste aversion. This makes it more likely that the pathway being examined is truly part of a regulatory circuit. Obviously it is essential to eat to survive and the organism must be awake to eat. Yet being awake is not synonymous with eating as there are other things the animal must also do. It must make choices about what behaviors are or are not appropriate in a given environment. Presumably, it is this sort of contextual information that is transmitted from the lateral septum to the lateral hypothalamus and other control centers.