Neuroimmunology of Eating Disorders
Jan Pieter Konsman and Robert Dantzer
Eating is a behaviour familiar to all of us. Despite this familiarity, regulation of eating in humans is complex and involves biological, psychological, social as well as cultural factors (Fischler, 2001). Most of our current knowledge concerning the biological factors involved in the regulation of food intake has been gained from investigations of eating behaviour in animals. The results obtained have recently been incorporated into a neurobiological model of the regulation of food intake (see for review Schwartz et al., 2000) that most probably bears important implications for a better understanding of eating behaviour and disorders in humans.
In this paper, we will, therefore, first review the neurobiological basis of the regulation of food intake before addressing two of the most common causes of anorexia in the absence of lesions or dysfunction of the gastro-intestinal tract, namely anorexia associated with infectious diseases and anorexia nervosa. Factors inhibiting food intake will be discussed in more depth than signals promoting eating since the focus of this paper is on anorexia. Besides, our present day understanding of how feeding is initiated is limited and the prevailing model of the regulation of food intake is based on the assumption that ingestion of food generates signals that subsequently inhibit eating.
Food intake needs to be regulated to assure the supply of amino acids and energy to cells throughout the body. All macronutrients, carbohydrates, lipids and proteins, can provide usable cellular energy, albeit with different efficiencies. Most cells are, however, capable of using carbohydrates in the form of glucose or lipids in the form of free fatty acids as energy substrates.
The early so-called depletion–repletion model of food intake postulated that organisms start eating when energy substrates, for example glucose levels, are low (depleted) and stop ingesting food as soon as energy substrate levels are replenished. According to this model a relationship should exist between the size of a meal and the time passed since the preceding meal. However, no such relationship is found when measuring meal size and the interval between meals in rats with free access to food (Le Magnen and Tallon, 1966). Instead, a relationship exists between the size of a meal and the time lag before the rat eats its next meal (Le Magnen and Tallon, 1966). These findings indicate that the ingestion of a meal generates satiety signals that suppress food intake.
During the 1970s and 1980s it became clear that the gut peptide cholecystokinin (CCK) constitutes a meal-generated satiety signal. CCK is synthesized in endocrine cells within the mucosa of the proximal small intestine (Buchan et al., 1978) and secreted upon ingestion of food (Liddle et al., 1985). Intraperitoneal administration of a synthetic peptide corresponding to the eight amino acids at the C-terminal portion of CCK (CCK-8) in rats just prior to food presentation causes a dose-dependent decrease in meal size, but not in water intake (Gibbs et al., 1973). CCK-8 administered intravenously to humans also induces earlier satiation without reports or signs of sickness (Geary et al., 1992).
Low concentrations of CCK-8 act on afferent fibres of the vagus nerves to reduce food intake in rats (Smith et al., 1981). Vagal afferent fibres express CCK receptors (Lin and Miller, 1992) and terminate in the brain stem at the level of the nucleus of the solitary tract. Lesioning this brain structure attenuates the satiety effects of CCK-8 (Edwards et al., 1986) indicating that it plays a role in the CCK-induced reduction of meal size. Although forebrain structures most probably influence brain stem circuits activated by CCK–8, it is important to note that brain stem circuits are sufficient to mediate the inhibitory effects of peripheral CCK-8 administration on sucrose intake (Grill and Smith, 1988).
More recently, with the development of CCK receptor antagonists, it has become possible to test the hypothesis that abdominal release of endogenous CCK constitutes a satiety signal. The expected effect of a CCK receptor antagonist would be to increase meal size. Indeed, administration of a CCKA (CCK type A) receptor antagonist increases meal size by about 25% in rats (Brenner and Ritter, 1995). It is important to note here that these results were obtained with an antagonist that did not enter the brain, since CCK receptors are also present in the CNS. Altogether, these findings indicate that the peripheral release of CCK after ingestion of food constitutes a satiety signal suppressing eating by acting on the vagus nerve and activating hind brain circuits.
Long-Term Meal Size
Although CCK clearly inhibits the ingestion of food during a meal, systematic administration of CCK-8 to rats at the start of each spontaneous meal turns out to have no effect on body weight, since animals eat meals more frequently (West et al., 1984). This indicates that other factors regulate eating over a longer time span to maintain energy stores. The hypothesis that the regulation of food intake is linked to the amount of energy stocked, for example in the form of fat, was first formulated in the 1950s by Kennedy (1953). This so-called lipostatic model of food intake and energy balance postulates that the organism eats to maintain a set point level of body adiposity. This stock of energy in the form of fat is used to meet the energetic demands of the organism. In this model the existence of adiposity signals reflecting the amount of energy stocked in the form of fat was postulated and proposed to act on