12.1.5 Weight regulation: physiology and pathophysiology

Body weight in humans is regulated by highly complex interacting neuronal and endocrine pathways that serve to stimulate food intake and reduce energy expenditure during food deficiency and to inhibit feeding when nutrition is replete. These mechanisms are highly conserved between mammalian species and promote the storage of sufficient quantities of energy-dense triglycerides in adipose tissue, thereby permitting survival during the frequent periods of food deprivation that were encountered during evolution. However, in modern times the ready availability of energy-dense food and the reduced necessity for energy expenditure has resulted in the excess storage of adipose tissue and a prevalence of overweight and obesity that the WHO considers to be an ‘epidemic’ (1). Overweight and obesity cause major morbidity and mortality, which are greatly attenuated when even modest amounts of weight are lost. Over the past two decades molecular biology and genetic studies have delineated many of the signals and pathways of appetite and body weight regulation. The therapeutic manipulation of these targets in the treatment of obesity is currently underway.

It is now known that homoeostatic mechanisms are in place in order to maintain body weight within a narrow range specific to each individual. For example, rats will lose weight following a period of caloric restriction but when allowed to ad libitum feed, the animals increase their food intake until they return to their previous body weight (2). Likewise, rats with diet-induced obesity will return to their previous body weight following the cessation of high-fat feeding. In humans, despite marked changes in food intake and physical activity day to day, over 1 year, body weight remains remarkably constant in both lean and obese individuals. The proclivity toward an apparent ‘set-point’ of body weight is underscored by the difficulty reported by individuals who attempt to lose weight and the propensity to regain the lost weight in a high proportion of cases.

The importance of the hypothalamus in body weight regulation came to the fore in the middle of the 20th, century when researchers reported that the lesioning of discrete hypothalamic nuclei could induce hyperphagia and obesity or aphagia depending on the targeted nucleus (3). In 1950, Kennedy hypothesized that the hypothalamus may be ‘directly sensitive to changes in the blood brought about by the ingestion of food’ (4). It was postulated that this signal to the hypothalamus might originate from adipocytes, a hypothesis known as the ‘lipostatic theory’ of weight control. The idea of a bloodborne signal was further explored by the parabiosis experiments of Hervey in which the circulation of a rat rendered obese by lesioning the ventromedial nucleus (VMN) of the hypothalamus was united with that of a lean rat. Hervey found that the control animal of the parabiotic pair became aphagic and died but this could be prevented by lesioning the VMN of the control which led to hyperphagia and weight gain (5). In recent years, several of the key components of the signalling system that regulates body weight have been identified (6). The systems that regulate feeding behaviour and energy balance appear to be composed of both short-term and long-term aspects. The short-term system, comprising changes in plasma gut hormones (see below), plasma glucose, and insulin concentration, body temperature, and plasma amino acids, can modulate meal patterns and feeding throughout the day. The long-term system, which includes the adipocyte-derived hormone leptin (7), balances food intake and energy expenditure and thus plays a dominant role in ultimately regulating the size of the body’s energy stores (Table 12.1.5.1).

Recessive mutations in the mouse ob and db genes result in obesity and diabetes (8). Obese (ob/ob) and diabetes (db/db) mice have identical phenotypes, each mutant weighing three times that of normal mice with a fivefold increase in body fat content. In addition the mice are hyperphagic and hypothermic with reduced energy expenditure. In the 1970s, parabiosis experiments conjoining the circulation of ob/ob mice with db/db mice resulted in hypoglycaemia, aphagia, and death by starvation in the ob mouse while its db partner was unaffected (9). This work seemed to imply that the ob/ob mouse lacked a circulating satiety factor while the db/db mouse appeared to be resistant to this same signal. It was not, however, until 1994 that the ob gene was cloned and shown to encode an adipocyte-derived hormone (7). As the wild type ob gene is required to prevent obesity, its protein product was named ‘leptin’, from the Greek word leptos meaning thin. With this came confirmation of the earlier idea that ob/ob mice lack functional leptin within the circulation due to a single recessive nonsense mutation whilst db/db mice lack a functional leptin receptor so that the phenotype of the former but not the latter can be rescued by the central or peripheral administration of leptin.

Leptin is synthesized and secreted by adipocytes signalling information about energy stores and nutritional status. Serum leptin levels are proportional to fat mass and fall in both humans and mice after weight loss. Administration of leptin to wild-type mice results in a dose-dependent decrease in body weight at incremental increases of plasma leptin within the physiological range.

Quantitative changes in plasma leptin concentration elicit a potent biological response. Decreases in plasma leptin levels activate what can be termed a ‘response to starvation’, while increasing leptin levels elicit a ‘response to obesity’. Several clues concerning the ‘response to starvation’ are provided by the phenotype of ob mice. Leptin-deficient (ob/ob) mice manifest a myriad of endocrine and metabolic abnormalities. Many of these derangements, which include decreased body temperature, hyperphagia, decreased energy expenditure (including activity) and infertility, are also observed in starved animals. This suggests that in the absence of leptin, ob/ob mice exist in a state of perceived starvation and thus exhibit a constellation of signs that are characteristic of the starved state. Indeed, in circumstances where food is readily available, this biological response would be expected to lead to the massive obesity evident in ob/ob mice. As would be predicted by such a model, replacement of leptin corrects all of the aforementioned abnormalities of mutant ob/ob mice (10, 11).

The available evidence suggests that the metabolic response to leptin is markedly different from the response to reduced food intake. While food restriction leads to the loss of both lean body mass and adipose tissue mass, leptin-induced weight loss is specific for the adipose tissue mass (12). Leptin also prevents the reduced energy expenditure normally associated with a decreased food intake. Finally, hyperleptinaemic animals undergoing a rapid period of weight loss fail to show any rise in serum free fatty acids or ketones (13). This is in contrast to food-restricted (pair fed) animals, which show a marked rise in serum free fatty acids. Indeed, despite the fact that the respiratory quotient falls after leptin treatment (indicative of fatty acid oxidation), the metabolic fate of stored triglycerides in adipose tissue is unknown.

Leptin also has effects on glucose metabolism. The possibility that leptin modulates glucose metabolism was first suggested in studies of ob/ob mice treated with leptin. Ob/ob mice are diabetic and the severity of the diabetes is dependent on the background strain carrying the mutation. In one study, leptin normalized the hyperglycaemia and hyperinsulinaemia evident in C57BL/6J ob/ob mice at doses that did not decrease weight. The antidiabetic effects of leptin have also been observed in insulin-deficient rats. Furthermore, leptin administration corrects the insulin resistance and hyperglycaemia of a lipodystrophic transgenic mouse line and of human subjects with lipodystrophy (14, 15). The antidiabetic effects of leptin appear to result from leptin’s ability to clear lipid from peripheral sites.

The possibility that falling plasma leptin levels signal nutrient deprivation is further suggested by the observation that exogenous leptin attenuates the neuroendocrine response to food restriction (16). Starvation is associated with decreased immune function which leptin corrects by the stimulation of CD4⁺ T-cell proliferation and increased production of cytokines by T-helper 1 cells (17). These findings indicate that leptin may be a key link between nutritional state and the immune system.

Hypogonadotropic hypogonadism is seen ob/ob mice and humans with genetic leptin deficiency or hypoleptinaemia as a consequence of low body weight and this can be corrected by the administration of leptin (18, 19). Treatment of mice with leptin accelerates the maturation of the female reproductive tract and leads to an earlier onset of the oestrous cycle and reproductive capacity. In humans, a surge in plasma leptin concentration is seen in prepubertal males. Gonadotropin-releasing hormone (GnRH) neurons of the mediobasal hypothalamus do not express the leptin receptor. The influence of leptin on the hypogonadotropic–pituitary–gonadal (HPG) axis may be mediated through the release of the peptide hormone kisspeptin. The intracerebroventricular (ICV) and peripheral administration of kisspeptin to male rats leads to a marked rise in plasma luteinizing hormone (LH), follicle-stimulating hormone (FSH), and total testosterone, and the application of kisspeptin to hypothalamic explants stimulates GnRH release. This stimulatory effect on the HPG axis is also seen following the peripheral administration of kisspeptin to humans (20, 21). Hypothalamic kisspeptin neurons express the leptin receptor and the expression of kisspeptin mRNA is reduced in ob/ob mice but can be increased by leptin administration. These studies suggest that leptin modulates reproductive function and provides a direct link between reproduction and the nutritional status of an animal.

The leptin receptor (ObR) is a single-spanning membrane receptor of the cytokine class I family. Several splice variants have been identified, which differ in the length of their cytoplasmic domain. The long form of the ObR (Ob-Rb) is widely expressed within the mediobasal hypothalamus and is also expressed in brainstem areas known to be important in the regulation of satiety. Binding of leptin to Ob-Rb results in the phosphorylation and activation of cytoplasmic Janus tyrosine kinase (JAK) 2. This results in the phosphorylation of tyrosine residues on ObRb (EC 2.7.10.2) and on signal transducers and activators of transcription (STAT) proteins. STAT proteins are inducible transcription factors. When phosphorylated, they dimerize and form a complex with the DNA-binding protein p48. This STAT-p48 aggregate translocates to the nucleus where it activates the transcription of genes bearing the interferon-response element (ISRE) (Fig. 12.1.5.1).

The available data suggest that the concentration of leptin, glucose, and other afferent signals are sensed by groups of neurons in the hypothalamus and other brain regions. During starvation leptin levels fall, thus activating a behavioural, hormonal, and metabolic response that is adaptive when food is unavailable. Weight gain increases plasma leptin concentration and elicits a different response leading to a state of positive energy balance. It is likely that different neurons respond to increasing versus decreasing leptin levels. In addition, the spectrum of leptin’s effects is likely to be complex as studies have indicated that different thresholds exist for several of leptin’s actions. The arcuate nucleus (ARC) of the mediobasal hypothalamus is located near the median eminence where the blood–brain barrier is incomplete, thus rendering it susceptible to the effects of circulating factors which signal nutritional status and the metabolic milieu (22). The ARC contains two populations of first order leptin-responsive neurons. One group of neurons coexpress the orexigenic peptides neuropeptide Y (NPY) and agouti-related protein (AgRP). The other neuronal subpopulation coexpresses the anorectic peptides alpha-melanocyte-stimulating hormone (α-MSH) derived from proopiomelanocortin (POMC) and cocaine- and amphetamine-related transcript (CART).

Large numbers of NPY neurons that express leptin receptor are present in the ARC and the administration of the leptin results in the inhibition of NPY synthesis and release. Mice deficient in both leptin and NPY are less hyperphagic and have increased energy expenditure than mice deficient in leptin alone, leading to a less obese phenotype (23). This confirms NPY’s position as downstream of leptin in the central regulation of food intake and energy expenditure. The expression of the POMC gene is up-regulated by leptin and reduced during food deprivation whereas the mRNA expression of AgRP is regulated in a fashion that is inversely proportional to POMC mRNA in response to fasting and the administration of leptin. Physiologically, α-MSH mediates a basal tonic inhibition of food intake by agonism at the melanocortin 4 receptor (MC4R). Conversely, AgRP, which is an antagonist at the MC4R, stimulates food intake. Leptin receptors are expressed on ARC POMC neurons and the targeted deletion of ObR from POMC neurons of the ARC leads to mild obesity, which is far less marked than the massive obesity of db/db mice.

Both NPY and melanocortin neurons of the ARC project to the paraventricular nucleus (PVN), where MC4R and NPY G_i-protein-coupled Y1 and Y5 receptors are expressed. These ARC to PVN projections are formed in the early postnatal period and their development is regulated by leptin (24). The PVN controls the secretion of peptides from both the anterior and posterior pituitary gland and projects to nuclei with sympathetic or parasympathetic efferents. The leptin receptor is not highly expressed in the PVN and it is likely that the effect of leptin on PVN outflow is via projections from the leptin first order neurons of other hypothalamic nuclei, such as the ARC and the ventromedial nucleus (VMN). Thus, the PVN may act as a final common pathway of the autonomic and endocrine response to circulating leptin with projections to numerous sites outside the hypothalamus including higher centres known to modulate motivational behaviours.

Within the VMN, brain-derived neurotrophic factor (BDNF) and its receptor are highly expressed. BDNF mRNA is markedly reduced in the VMN by food deprivation and the infusion of BDNF into the lateral ventricle of rats results in a dose-dependent reduction in food intake and body weight. BDNF knockout (BDNF^–/–) mice die in the early postnatal period but BDNF^+/– mice develop massive adult-onset obesity, which is secondary to hyperphagia as BDNF^+/– mice pair fed to the wild-type do not become obese.

CART is a neuropeptide that is highly expressed within the ARC, the PVN, and the lateral hypothalamus (LH). Within the ARC CART is coexpressed with POMC neurons. The ICV administration of CART reduces food intake and the central administration of an antibody to CART stimulates feeding (25). CART mRNA in ARC POMC neurons is reduced in leptin deficient ob/ob mice and in food-deprived animals, and can be increased by the administration of leptin.

The LH highly expresses the neuropeptide melanin-concentrating hormone (MCH). MCH mRNA levels in the LH increase during starvation and return to baseline after refeeding. The injection of MCH into the LH results in an increase in food intake in rats and MCH knockout mice are hypophagic and have a lower body weight than wild-type controls. The LH also highly expresses the orexins, a family of neuropeptides designated A and B. Levels of orexin pre-pro mRNA increase during fasting and the central administration of orexin stimulates food intake. Orexin exerts its effects by binding to two G-protein-coupled receptors named orexin receptor 1 (OX1R) and orexin receptor 2 (OX2R). OX1R is specific for orexin A whereas OX2R binds both orexins with a similar affinity. In rats ICV orexin increases sympathetic outflow as evidenced by increased brown adipose tissue temperature and a rise in heart rate but this effect is attenuated in VMN lesioned rats. While the orexins were originally named for their stimulatory effect on food intake from the Greek orexis meaning appetite, subsequent studies have demonstrated that the orexins play a significant role in behavioural arousal and human narcolepsy, a condition of sudden, irresistible daytime somnolence has been shown to be associated with a deficiency of orexin or OX2R.

The gastrointestinal tract is the largest endocrine organ in the body. It secretes more than 20 different peptide hormones which serve both a local regulatory function and provide a means by which the gut can regulate appetite and satiety (Fig. 12.1.5.2). Circulating levels of the gastric orexigenic gut hormone ghrelin rise during fasting and fall following food intake (26). The central and peripheral administration of ghrelin to rats results in the marked stimulation of feeding and the peripheral administration of ghrelin to lean and obese humans increases food intake and leads to weight gain when chronically administered. Ghrelin binds to the growth hormone secretagogue (GHS) receptor a G-protein-coupled receptor that is highly expressed by the NPY neurons of the ARC, and ghrelin-induced feeding is abolished by the administration of NPY antagonists. The GHS receptor is also expressed in the nucleus of the solitary tract (NTS), which receives afferent innervation from the vagus nerve and sends efferent output to the ARC. Vagotomy abolishes the ghrelin rise induced by food deprivation, and blockade of the gastric vagal afferent in a rodent model abolishes ghrelin-induced feeding.

Peptide YY (PYY) is released from the L cells of the colon and rectum following food ingestion. Most circulating PYY is the N-terminally truncated form of the full length peptide, the 34 amino acid PYY_3-36, which, when administered by acute peripheral injection, reduces food intake in rodents and humans (27, 28). The ARC appears to be an important site of action in the satiety-inducing effects of peripherally administered PYY_3-36, which may directly inhibit NPY neurons causing the disinhibition of POMC neurons.

Pancreatic polypeptide (PP) is synthesized and released by the PP cells of the pancreatic islets of Langerhans and to a lesser extent the colon and rectum. Circulating levels rise in the postprandial period in proportion to the ingested calorie load and decline during fasting. The peripheral administration of PP to mice and humans reduces food intake and transgenic mice which overexpress PP, eat less, and weigh less than wild-type animals (29). In addition to leptin deficiency, ob/ob mice also lack PP cells and the peripheral administration of PP to ob/ob mice reduces food intake and body weight.

Glucagon-like peptide-1 (GLP-1) is synthesized and secreted by the L cells of the small intestine and colon, the alpha cells of the islets of Langerhans, and neurons within the NTS of the brainstem. Its release is stimulated by food intake with levels in the circulation rising after a meal and expression in the small intestine falling with fasting. The direct injection of GLP-1 into the PVN reduces food intake in rats and evidence for the physiological importance of GLP-1 is suggested by the orexigenic effect of the central administration of the specific GLP-1 receptor antagonist exendin_9-39 to satiated rats, which when administered for 10 days significantly increases body weight (30). Furthermore GLP-1 has an incretin effect such that it augments glucose-dependent insulin secretion, which has led to its development as a therapy for type 2 diabetes mellitus.

Oxyntomodulin (OXM) is cosecreted with GLP-1 and PYY_3-36 following food intake. OXM promotes satiety in both rodents and humans and this effect may partly be through the augmentation of ARC α-MSH signalling. In addition, part of OXM’s suppressive effect on food intake may be due to a reduction in plasma ghrelin.

The L cells of the small intestine also synthesize cholecystokinin (CCK). Circulating levels of CCK rise following food intake and CCK has been shown to reduce food intake in a dose-dependent manner following its administration to rats and to humans. The Otsuka Long Evans Tokushima Fatty rat has a null mutation of the CCK receptor, CCK_A, and is hyperphagic and obese (31). The CCK_A receptor is expressed in vagal afferent and efferent neurons and is also found in the brain in the NTS, area postrema, and the dorsomedial nucleus, areas which are known to be important in the control of food intake. Abdominal or gastric vagotomy has been shown to block the satiety effect of peripherally administered CCK, indicating that the vagus nerve may be particularly important in mediating the effect of CCK on food intake.

The importance of thyroid hormone in the control of metabolic rate and food intake is attested by the reduction in basal energy expenditure and weight gain of hypothyroidism and the increased energy expenditure and hyperphagia of hyperthyroidism. When food intake falls there is a rapid down-regulation of the hypothalamo–pituitary–thyroid (HPT) axis in a pattern consistent with central hypothyroidism. This is in part mediated by the fall in plasma leptin during food deprivation. The leptin receptor is expressed on thyrotropin-releasing hormone (TRH) neurons of the PVN and a fall in circulating leptin results in a reduction in pro-TRH mRNA in these neurons. Furthermore TRH neurons of the PVN receive input from the orexigenic NPY/AgRP and the anorectic POMC neurons of the ARC. A fall in food intake results in increased NPY and AgRP stimulation of the TRH neuron and a fall in α-MSH signalling, resulting in a reduction in TRH synthesis and release. This down-regulation of the HPT axis during food deprivation is critical for energy conservation (see next section). Triiodothyronine (T₃) stimulates food intake. In humans one of the characteristics of thyrotoxicosis is hyperphagia and in rodents the administration of T₃ directly into the VMN elicits a hyperphagic response, although the mechanism by which T₃ stimulates food intake is not known.

Changes in weight can result from alterations in energy intake or energy output. Energy expenditure is markedly decreased in lean and obese humans after weight loss which may represent a compensatory adaptive response that serves to maintain weight at a stable level in each individual (32). Obese individuals who lose weight must therefore consume fewer calories to maintain a constant weight relative to weight-matched subjects whose weight has been stable. This finding may in part account for the high failure rate of dieting for the long-term maintenance of weight loss as lean individuals who had previously been obese need to consume fewer calories while experiencing persistent feelings of hunger (33).

Caloric expenditure can be grouped into several categories including those applied to resting metabolic rate (RMR), thermic effect of feeding (TEF), and thermic effect of exercise (TEE). RMR is unchanged in obese individuals who lose weight, while the TEE is apparently reduced. The molecular basis of this change in energy expenditure is unknown but it may be the result of differential activity of uncoupling proteins. These proteins are proton channels that disrupt the mitochondrial protein gradient in brown adipose tissue and possibly other tissues, resulting in the generation of heat rather than ATP.

Brown adipose tissue is a key site of adaptive thermogenesis in small mammals and human neonates. Recently [¹⁸F]2-fluoro-2-deoxy-d-glucose positron emission tomography (FDG PET) scanning has identified a significant amount of brown adipose tissue in adult humans, although its distribution and physiological significance are not currently known. This tissue has abundant mitochondria, is highly vascular, and expresses uncoupling protein-1 (UCP1). The principal known function of uncoupling proteins is to generate heat in response to a cold stress or food intake. Brown adipose tissue receives dense innervation from the sympathetic nervous system and when stimulated by noradrenaline via highly expressed β₃-adrenoceptors responds with the increased expression and activation of UCP1. This results in the dissipation of the proton gradient across the inner mitochondrial membrane leading to heat rather than ATP production. This sympathetic stimulation also leads to an increase in T₃ within brown adipose tissue, which further increases the expression of UCP1. This synergism between the sympathetic nervous system and HPT axis serves to augment the metabolic response to a fall in ambient temperature or excess food intake.

In humans a highly significant correlation between body fat content and plasma leptin concentration has been observed and obese humans generally have high leptin levels. These data suggest that in most cases, human obesity is likely to be associated with insensitivity to leptin although it is unclear whether obesity is the result of leptin resistance or whether the raised circulating leptin level simply reflects the increased mass of adipose tissue. The basis for leptin resistance in obese, hyperleptinaemic human subjects is unknown. It has been suggested that entry of leptin into the cerebrospinal fluid (CSF) may be rate-limiting in some obese subjects. The short form of ObR is highly expressed in several tissues, most notably the choroid plexus and cerebral microvasculature, and appears to be important in the transport of leptin across the blood–brain barrier and into the CNS. The transport of leptin from the blood to the CSF is a saturable process, and in addition it has been found that the high plasma leptin levels in the plasma of obese individuals are not reflected in CSF leptin levels. Leptin uptake has been demonstrated in the capillary endothelium of mouse and human brain and is decreased in preobese animals.

Leptin signalling is terminated by suppressor of cytokine signalling 3 (SOCS3). Leptin induces the transcription of SOCS3 protein, which binds to the leptin-induced phosphorylated tyrosine residues on JAK2 and inactivates the enzyme (see Fig. 12.1.5.1). It has been proposed that excessive SOCS3-mediated negative feedback in the face of high leptin levels may play a role in leptin resistance in obesity.

Leptin resistance in humans is likely to be the result of a complex interplay of many factors. In principle, leptin resistance could result from the altered activity of any of the aforementioned components of the leptin signalling pathway. Factors that directly decrease energy expenditure or activate adipogenesis and lipogenesis could also result in apparent leptin resistance. Finally, leptin’s actions are likely to be influenced by psychological factors via connections between the higher cortical centres which modulate an animal’s motivational state and neural circuits within the hypothalamus. The neuroanatomical and functional relationships between these brain regions are currently being elucidated.

Obese subjects have been found to have reduced ghrelin suppression after a meal compared with normal weight controls and ghrelin levels rise after diet-induced weight loss, which could jeopardize weight loss maintenance. In normal weight individuals, PYY_3-36 levels increase rapidly following nutrient ingestion, however, this response has been reported to be attenuated in obese subjects with a higher caloric load needed to stimulate a rise comparable to lean individuals. An abnormality of PP secretion in obese subjects has been demonstrated by some but not all investigators. Obese subjects have also been found to have an attenuated postprandial release of the satiety-promoting peptide GLP-1 and reduced circulating levels of the hormone that increases with weight loss.

In recent years, an important framework for understanding the regulation of body weight has emerged. The data indicate that a robust physiological system acts to preserve the relative constancy of weight and to uphold weight at different levels in different individuals. When at this set point, individuals maintain a state of energy balance; weight gain elicits a biological response characterized in part by a state of positive energy balance, whereas weight loss among both lean and obese subjects results in a response that leads to a state of negative energy balance. Further studies of the molecular components of this system and mechanisms that determine the set point for weight are likely to have a major impact on our understanding and treatment of obesity and other nutritional disorders.