2.1. The discovery of leptin
One major breakthrough in understanding body weight regulation was the discovery of leptin. This discovery was the result of decades of research into mouse strains with heritable forms of obesity. Five such spontaneous mouse mutants displaying obesity have been described until now: ob, db, tub, fat and Ay. Surprisingly, the genes mutated in these mice, can all be placed into the same anatomical and signalling route.
Ob/ob and db/db mice have identical phenotypes, each weighing three times more than normal mice with a fivefold increase in body fat content. Data from cross-circulation (parabiosis) experiments suggested that the ob gene was responsible for the generation of a circulating factor that regulated energy balance and that the db gene encoded the receptor for this factor (Friedman and Halaas, 1998).
The cloning and characterization of the ob gene showed that it encodes a hormone, named leptin, that is expressed abundantly in adipose tissue. The db gene encodes the receptor for leptin. Levels of leptin are proportional to adipose tissue mass. The discovery of leptin was received with great enthusiasm, since it was thought that a drug mimicking leptin could treat the obesity epidemic. The clinical phenotype of human congenital leptin deficiency is very similar to that seen in the ob/ob mouse. Both leptin-deficient humans and mice have early-onset obesity, increased food intake, hypo gonadism, hyper-insulinaemia and defective function of the hypothalamo-pituitary thyroidal axis. Thus, leptin plays a similar role in mice and humans. Indeed treatment with leptin of these rare cases of obese individuals carrying mutations in the leptin gene, is very successful to normalise body weight and neuroendocrine functioning (Farooqi et al., 2002). However, most obese people have plenty of leptin, and it is therefore thought that obese people are leptin-resistant. Consequently, research has focused on the downstream effector pathway of leptin, which was expected to malfunction in obese individuals.
2.2. The neural circuitry mediating leptin's effect
Besides some peripheral tissues, such as adrenal cortex, liver and pancreas, leptin receptors have been found in several hypothalamic nuclei that are involved in the regulation of energy balance, which include the arcuate nucleus, the ventromedial hypothalamus (VMH), the lateral hypothalamus (LH), the dorsomedial hypothalamus (DMH) and the paraventricular nucleus (PVN). Leptin receptors have also been found in brainstem nuclei such as the nucleus of the tractus solitarius, the dorsal motor nucleus of the vagus nerve, the lateral parabrachial nucleus, and the central gray.
The leptin-responsive hypothalamic nuclei express one or more neuropeptides and neurotransmitters that regulate food intake and/or body weight. The arcuate nucleus has a large density of leptin receptors. In the arcuate nucleus there are at least two different populations of neurons that are oppositely regulated by leptin. Neurons expressing pro-opiomelanocortin (POMC) and cocaine-and-amphetamine-related transcript (CART) are activated by high plasma leptin levels. POMC is the precursor of the melanocortins a-, P- and y-melanocyte stimulating hormone (MSH) and P-endorphin. a-MSH, when injected into the brain, decreases food intake and body weight. Neurons expressing Agouti-related protein (AgRP) and Neuropeptide Y (NPY) are activated when plasma levels are low, such as during starvation. The importance of the medial hypothalamus, which includes the arcuate nucleus, was already known for decades, since lesions of this area result in obesity, and electrical stimulation inhibits food intake. With the discovery of leptin and its direct effector pathways in the arcuate nucleus, the neuropeptides that probably mediated these effects were identified. Now that important parts of the neural circuitry underlying regulation of energy balance have been discovered, the validation of new drug targets within these neural circuits can be explored for the treatment of eating disorders and obesity.
Major identified pathways involved in the regulation of energy balance are depicted. Note: 1. the central position of the arcuate nucleus, where signals from the periphery (such as leptin) reach the brain; 2. serotonergic and dopaminergic pathways super-imposed on the hypothalamic centers involved in regulation of energy balance.
One of the neuropeptides acting downstream of leptin, NPY, is the most potent orexigenic agent known when administered in the brain. Further analysis of which aspect of feeding behavior is influenced by NPY revealed that NPY stimulates energy intake when rats can go to food, but not when food is administered intra-orally (Ammar et al., 2000). This suggest that NPY is involved in the appetitive phase, but not in the consumatory phase of food intake. Detailed analysis of how NPY regulates food intake in animal models is important to understand the role of NPY in feeding behavior in humans.
NPY mRNA is increased in ob/ob mice and decreased after leptin treatment or following starvation. Therefore, it was expected that increased NPY signalling was responsible for the obesity observed in ob/ob mice. Since in knockouts for NPY and leptin (NPY-/-; ob/ob) the obesity of ob/ob mice was only partially inhibited, it was realised that also other neuropeptides might play a role (Erickson et al., 1996).
In viable yellow mice (Ay) such another pathway, acting in parallel to NPY, was identified. Ay mice have ectopic overexpression of Agouti protein. Agouti is normally only expressed in skin where it acts as an antagonist at the melanocortin MC1 receptor, resulting in the switch from eumelanin (dark pigment) to phaeomelanin (pale pigment) synthesis. Expressed in the brain as in Ay mice, Agouti antagonises the melanocortin MC4 receptor. The activity of the MC4 receptor is activated by MSH, which is an agonist, and inhibited by AgRP (the homolog of Agouti expressed normally in brain), an inverse agonist (Adan and Kas, 2003). Thus, Agouti disrupts the leptin down-stream effector pathway by blocking the effect of POMC neurons releasing MSH. Ay mice therefore have a similar phenotype as melanocortin-MC4-knockout mice: obesity and leptin resistance.
Thus, ob, db and Ay mice have deficits in the same genetic pathway. How about the other two spontaneous obese mouse mutants fat and tubby? The gene defect underlying obesity in the fat mouse is a mutation in the carboxypeptidase E gene (Naggert et al., 1995). Carboxypeptidase E is involved in the processing of neuropeptides like POMC but also of insulin in secretory granules. Impaired processing of POMC could explain the obesity observed in the fat mouse (Berman et al., 2001). The tubby gene, a transcription factor which is mutated in the tub mouse, is expressed in the hypothalamus. In tubby mice levels of expression of POMC and NPY are altered (Guan et al., 1998). Thus, all five naturally occurring obese mouse mutants have defects in leptin or its downstream effector pathways. The primary leptin-responsive neurons are located in the arcuate nucleus. From here, POMC/CART and AgRP/NPY neurons have wide-spread projections to other areas of the brain involved in regulation of food intake and energy balance. Important regions include the ventromedial hypothalamus (VMH), the lateral hypothalamus (LH), the paraventricular nucleus (PVN) and the mesolimbic system. In these areas so-called second order leptin responsive neurons are present. The PVN is an important projection area for leptin-responsive arcuate nucleus neurons. Thyroid-releasing hormone (TRH) and corticotrope-releasing hormone (CRH) neurons have been implicated in the leptin downstream signalling cascade (Legradi et al., 1997; Masaki et al., 2003). TRH influences energy balance via the pituitary-thyroid axis, which is involved in regulating metabolic rate. CRH, when injected in the brain, suppresses food intake.
The VMH has been implicated in the regulation of food intake, it senses blood glucose levels, controls digestive system functions and regulates glucagon and insulin levels. The VMH expresses amongst other neuropeptides TRH and cholecystokinin (CCK), and the neurotrophin brain-derived neurotrophic factor (BDNF). CCK is a satiety factor released from duodenal mucosa, stimulated by lipids from digestion (Noble and Roques, 2002). Injected into the brain CCK induces satiety and anxiety (Moran and Schwartz, 1994). Therefore, the CCK system is interesting with regard to eating disorders, since in anorexia nervosa anxiety for ingesting food is a major characteristic.
Mice carrying mutations in BDNF signalling are hyperphagic and hyperactive. It was shown that BDNF acts downstream of the MC4 receptor. BDNF suppressed food intake when injected in the brain, and MC4 receptor knockouts have lower VMH BDNF levels (Xu et al., 2003). Thus, also BDNF has been implicated in the downstream signalling pathway of leptin.
The lateral hypothalamus contains at least of two interesting populations of neurons that produce either orexin or melanin-concentrating hormone (MCH). Orexins are involved in the sleep-wake cycling and stimulate food intake during the night in nocturnal species (Hara et al., 2001). MCH also stimulates food intake. Transgenic mice overexpressing MCH in the lateral hypothalamus are mildly obese and display insulin-resistance (Ludwig et al., 2001).
The arcuate nucleus in the hypothalamus is one of the brain regions that lack a clear blood brain barrier. Besides receptors for leptin, the arcuate nucleus senses changes in other plasma derived factors that have been identified in regulation of energy balance, such as insulin, amylin, PYY and Ghrelin. Insulin acts on most of the brain centers that also have leptin receptors, and inhibits food intake. Amylin, co-released with insulin from P cells of the pancreas, also inhibits food intake (Rushing, 2003). PYY is post-prandially released from intestine acts at pre-synaptic, autoinhibitory NPY-2 receptors expressed on NPY/AgRP neurons in the arcuate nucleus. PYY inhibits the activity of these orexigenic neurons and thus decreases food intake (Batterham et al., 2002). Ghrelin, mainly produced in the stomach in particular during starvation, but also in the hypothalamus itself, activates growth-hormone-secretagogue receptors (GHS-R) expressed on AgRP/NPY neurons in the arcuate nucleus, resulting in increased activity of these neurons, which results in increased food intake (Cowley et al., 2003).
The hypothalamus also receives input from other brain areas. Serotonin has been implicated in regulation of food intake. Serotonin 2C receptor knockout mice are hyper-phagic and obese. Fenfluramine, a serotonin reuptake inhibitor which has been on the market to treat obesity, requires serotonin 2C receptors in order to inhibit food intake (Heisler et al., 2002), expressed on arcuate nucleus neurons expressing POMC. Atypical antipsychotics such as olanzapin, may thus increase food intake via antagonism of the serotonin 2C receptor.
The mesolimbic dopamine system is innervated by both primary and secondary leptin responsive neurons. Dopamine deficient mice display decreased motor activity and reduced food intake. Restoration of dopamine signaling in the caudate putamen, but not in the nucleus accumbens, normalizes food intake (Szczypka et al., 2001).
Thus, genes (in particular receptor genes) constituting the serotonergic and dopaminergic systems are expressed in primary and secondary leptin responsive brain nuclei that affect energy balance, and may via this route be involved in eating disorders and obesity and its treatments.
Although five spontaneous mutations have been identified resulting in obesity, there is only one spontaneous mouse mutant identified for anorexia: the anx mutant. Anx/anx mice suffer from a autosomal recessive mutation resulting in poor appetite, reduced body weight, emaciated appearance, body tremors, head weaving, hyperactivity, and uncoordinated way of walking. These mice have significantly reduced serum leptin levels. By performing immunohistochemical studies it has been demonstrated that neuropeptide Y (NPY) and agouti-related peptide (AGRP) accumulate in the cell bodies rather than in the dendritic extensions in the arcuate nucleus (Broberger et al., 1999). These studies also showed decreased levels of pro-opiomelanocortin (POMC), neuropeptide receptors and cocaine and amphetamine-related transcript (CART) mRNAs in the arcuate nucleus. Anx/anx mice have been shown to have an increase in serotonergic fibers in the forebrain and arcuate nucleus. Serotonin appears to play a role both in abnormal behaviour and in appetite. The anx gene has been localized on mouse chromosome 2 and cloning and elucidation of the biochemical function of this gene will help to understand more about the regulation of food intake and the alterations that occur in eating disorders (Siegfried et al., 2003).
Dopamine-deficient (DD) mice were generated by deletion of the tyrosine hydroxylase gene specifically in dopaminergic neurons. DD mice are born normal but gradually become hypoactive and hypophagic and die at 3-4 weeks of age. DD mice can be rescued by daily treatment with L-DOPA or by introducing a liquid diet directly into the mouth, which indicated that these mice die due to lack of nutrients. Szczypka et al. (Szczypka et al., 2001) believe that the anorexia of DD mice is not exclusively a consequence of motor deficits because they do move as much or more than wild-type mice under certain conditions. Also their ingestive behaviour in response to novel food is initially indistinguishable from that of normal mice.
M3 muscarinic receptor knockout mice (M3R -/-) display a significant decrease in food intake, reduced body weight and peripheral fat deposits, and very low levels of leptin and insulin (Yamada et al., 2001). Both decreased MCH expression and reduced responsiveness to AGRP are major contributing factors to the hypophagic phenotype in M3R "" mice.
MCH knockouts have a hypophagic and lean phenotype and an increased metabolic rate (Shimada et al., 1998). After 24h starvation, weight loss was greater in MCH " " than in controls and after 48h of fasting, and 75% of the MCH "" mice died, while no control mice died.
Corticotropin releasing hormone receptor 2 (Crhr2) knockouts have anxietylike behaviour and are hypersensitive to stress (Bale et al., 2000). After 24 h of food deprivation mutant mice consumed 75% of wild-type food levels in the 24 h period following food deprivation. However their body weights did not differ from those of wild type mice. It is not clear whether this is a direct effect of metabolism, or if the stress of fasting alters the anxiety state of the animals, and thus affects appetite or metabolism.
Cannabinoid receptor 1 (CB1) knockouts have significantly lower food intake than wild-type mice after 18h of fasting, although no difference was found in weight and food consumption (Di Marzo et al., 2001). Studies using a selective CB1 antagonist indicated that endogenous cannabinoids acting at CB1 receptor may be involved in maintaining food intake in mice made hyperphagic by food deprivation
3. EATING DISORDERS AND OBESITY 3.1. The obesity phenotype
Obesity is defined by the WHO as a chronic disease characterised by an excess of body fat in such a degree that it leads to significant health risks (2000). The body mass index (BMI), calculated as body weight (in kg) divided by height2 (in m), is widely used as a measure of the degree of overweight and is excellently correlated to body fat mass. In Caucasian populations a BMI between 18.5 and 25 is considered to be normal; overweight is defined as a BMI between 25 and 30 and obesity as a BMI above 30. Not only the total amount of body fat is important to predict the health risks, but also fat distribution: intra-abdominal (visceral) fat is much more associated with co-morbidity than subcutaneous, femorogluteal fat. The intra-abdominal fat mass can easily be estimated by measuring the waist circumference: a waist circumference > 94 cm (males) or >88 cm (females) carries a significantly increased health risk.
Co-morbidity, associated with obesity, consists primarily of the metabolic syndrome (insulin resistance, eventually resulting in type 2 diabetes, dyslipidemia and hypertension among others), leading to an increased prevalence of cardiovascular disease. In addition, certain types of cancer, osteoarthritis, gout, gallstones, liver steatosis, obstructive sleep-apnea syndrome, gastro-esophageal reflux, disturbances in gonadal hormonal function, depression, psychosocial dysfunctioning and considerable loss of quality of life are clearly associated with obesity. It has also been established that obesity, through its comorbidity, definitely leads to increased mortality (Calle et al., 1999).
The increasing prevalence of overweight and obesity in the Western world is truly alarming and fighting this epidemic poses one of the greatest challenges for health caretakers. Presently, 10-25% of the adult populations in Western countries are obese and 40-65% are overweight. Although numerous research has pointed to the existence of a genetic predisposition to obesity, clearly the enormous increase in obesity prevalence during the last decades has to be caused by our obesogenic environment, characterised by abundance of easily available, energy-dense food on one hand and our sedentary lifestyle with lack of physical activity on the other hand.
According to basal thermodynamic laws, excessive fat deposition, as is the case in obesity, is the consequence of a longstanding positive energy balance in which energy intake has been greater than energy expenditure. It is important to realise that only a small, but structural, energy excess is needed to cause obesity: a structural daily excess of not more than 100 kcal will lead to an average weight gain of more than 4 kg after one year! Considering the wide daily variation in both our energy intake and energy expenditure, it is in fact more remarkable that most people are able to keep their weight stable. The neural and hormonal mechanisms which are involved in this energy homeostasis are only partly unravelled and it is very well conceivable that these mechanisms are somewhat less finely tuned in people predisposed to obesity, possibly leading to altered hunger and satiety signals and/or metabolic rate. In addition, an altered functioning of hormones, neurotransmitters and receptors in the parts of the central nervous system involved in eating behaviour may also be related to some different phenotypes in obese subjects which can be encountered (for instance "stress eaters" versus people who primarily respond to external food cues in an exaggerate way). Clarification of these mechanisms could lead to more effective and more individualised pharmacological intervention in obesity.
The most important targets in obesity management are weight loss followed by longstanding weight maintenance and treatment of co-morbid conditions. It should always and primarily consist of lifestyle changes directed to a healthier eating behaviour and more physical activity. Longstanding maintenance of a lower weight appears to be the most difficult and challenging part of obesity management as contraregulatory mechanisms (such as decrease of basal metabolic rate after weight loss a.o.) are operating in a (frequently successful) attempt to increase body weight again to the original (or even higher) level. Apart from bariatric surgery (gastroplasty, gastric banding or gastric bypass), which is only available for selected patients with morbid obesity (BMI > 40), the results of nearly all lifestyle and pharmacological interventions directed to restoration of a (near-)normal weight on the long-term have been disappointing (success rate < 10%). Nowadays the goal of treatment has changed from weight normalisation towards a moderate, but longstanding weight loss of about 10% of the original weight. This moderate weight loss is associated with a clear decrease in cardiovascular risk factors.
People diagnosed with anorexia nervosa are at the total opposite site of the weight spectrum. One of the diagnostic criteria of anorexia is a BMI lower then 17.5 kg/m2. Anorectic people have a very severe and selective food intake and do not eat sufficient amounts of dietary fat and sugars. Someone is diagnosed anorexic when the following phenotypic observations are made: overevaluation of shape and weight, active maintenance of a very low BMI and amenorrhoea in postmenarcheal women. Anorexia is predominantly seen in adolescent females. Character traits like low self-esteem and perfectionism are also very often observed in these women.
Phenotypic descriptions of anorexia and bulimia nervosa: The first descriptions of anorexia nervosa were done by the Frenchman Lasegue (Laseque, 1873) and the English physician Gull in the 1870's (Gull, 1874). They described 'a morbid mental state' that causes an illness characterized by a reduction of food intake, emaciation, amenorrhea, hyperactivity, hypothermia and lack of insight, that typically strikes young females. 130 years later this picture is still very accurate, over the years many theories about the influence of society, families, early traumatic experience etc. have come and gone and treatment has tried to counter those problems, resulting in a range of psychologically based treatment programs.
Nowadays, eating disorders are viewed as classic examples of complex psychiatric phenotypes with both genetic and environmental determinants (Devlin et al., 2002). Anorexia nervosa is characterized by pathologic eating behaviour and the relentless pursuit of thinness, resulting in extreme emaciation. In adults, a weight that leads to a Body Mass Index below 17.5 kg/m2 is considered to be the cut-off between normal and anorectic weight, in children and adolescents it is the failure to make the expected weight gain during a period of growth, leading to a bodyweight less that 85% of expected (Diagnostic and Statistical Manual of Mental Disorders Fourth Edition (DSM- IV), 1994). Patients exhibit an overevaluation of shape and weight and often a distorted body image and in addition postmenarcheal women with this disorder are amenorrheic. Abnormally high activity levels and overexercising are seen in 30-80% of patients (Hebebrand et al., 2003), mainly the so-called restrictors or RAN (resticting-type anorexia nervosa), (subjective) binges, laxative abuse and vomiting characterize the purging anorectic patients or BAN (bingeing-type anorexia nervosa). General issues as low self-confidence and self-esteem, and co-morbid anxiety and depressive symptomatology are seen in the majority of cases.
Psychometric studies have consistently linked RAN in particular to a stereotypic cluster of (moderately heritable (Heath et al., 1994)) personality traits that are found with great consistency. These include emotional restraint, avoidance of novelty, anxious worry and self-doubt, compliancy, obsessionality, perfectionism and perseverance in the face of non-reward. Many of these traits are exaggerated by starvation but retrospective accounts suggest that they often predate the onset of the eating disorder (Deep et al., 1995) and persist long after the normalization of weight and menses (Srinivasagam et al., 1995). Considered temperamental risk factors they support the possibility that susceptibility is influenced. Increased serotonergic activity that persists after recovery (Kaye et al., 1991) may be one an important cause.
The presentation of Bulimia Nervosa is less consistent but there are certain common personality and behavioural traits that may play an important biologically mediated etiological role in the development of the disorder. Characteristic traits are thrill seeking and excitability and a tendency to dysphoria in response to rejection or non-reward (Bulik et al., 1995).
Was this article helpful?