The discovery of leptin changed the view of adipose tissue from that of a passive vessel that stores fat to that of a dynamic endocrine organ that actively regulates behaviour and metabolism. Secreted by adipose tissue, leptin functions as an afferent signal in a negative feedback loop, acting primarily on neurons in the hypothalamus and regulating feeding and many other functions. The leptin endocrine system serves a critical evolutionary function by maintaining the relative constancy of adipose tissue mass, thereby protecting individuals from the risks associated with being too thin (starvation and infertility) or too obese (predation). In this Review, the biology of leptin is summarized, and a conceptual framework is established for studying the pathogenesis of obesity, which, analogously to diabetes, can result from either leptin hyposecretion or leptin resistance. Herein, these two states are distinguished with the terms ‘type 1 obesity’ and ‘type 2 obesity’: type 1 obesity describes a subset of obese individuals with low endogenous plasma leptin levels who respond to leptin therapy, whereas type 2 obesity describes most obese individuals, who are leptin resistant but might respond to leptin therapy in combination with other drugs, such as leptin sensitizers.
A large body of evidence dating to Lavoisier has shown that the same laws that govern chemical and physical processes in inanimate systems also apply to bioenergetics in living organisms1. The second law of thermodynamics states that energy within a system must be conserved and that any change must result from either a change in the energy input to, or output from, that system. Thus, among mammals, in which energy is stored primarily as energy-rich triglycerides (9 kilocalories per gram) in adipose tissue, any change in weight (fat mass) must necessarily be a result of changes in energy input (food intake) and/or energy expenditure. For weight to remain stable in a living organism, energy input thus must be precisely balanced against energy output over long periods of time.
When food is readily available, adult humans maintain a remarkably stable weight while consuming approximately 1 million kilocalories each year2. Collectively, these findings have long suggested the existence of a biologic mechanism indexing food intake to energy expenditure to maintain the stability of weight and adipose tissue mass3,4. Maintaining the stability of energy stored in adipose tissue is also essential for survival, because lower levels of body fat have deleterious consequences by increasing the risk of starvation during periods of food insecurity or famine, as well as decreasing fertility and immune function; in contrast, a higher adipose tissue mass increases the risk of predation5,6,7,8.
It had thus been postulated that nutritional state, probably from a fat-derived signal, is sensed to maintain homeostasis of adipose tissue mass4. However, the identification of the molecular components of this putative system has proven elusive, owing in part to the intrinsic difficulty in applying biochemical methods to identify signals that physiologically regulate food intake or energy expenditure in vivo. This difficulty is further compounded by the ability of many factors to decrease food intake by non-specifically eliciting an aversive (non-physiologic) response, including nausea and other unpleasant sensations.
Clues from genetic models of obesity
A clue as to the identity of the factors that regulate energy balance was provided by the identification of the recessive mouse obese (ob) mutation in 1950 (ref. 9) and the subsequent identification of the diabetes (db) spontaneous mutation, both of which cause mice to develop massive obesity and show identical phenotypes on the same genetic background10,11. A similar phenotype also develops in genetically obese rats carrying the fatty (fa) mutation12. These fully penetrant recessive mutations cause extreme obesity and insulin-resistant diabetes as part of a complex syndrome including a broad set of unusual abnormalities not generally observed among obese humans, including infertility, immune alterations and hypothermia13. The explanation for this unusual constellation of abnormalities is explained in further detail below.
The extreme obesity of these mutants resembles that in animals with lesions of the ventromedial hypothalamus (VMH), thus suggesting that the encoded gene products might interact with this brain region, although whether the lesion also affects neurons in adjacent nuclei or fibbers of passage was unclear14. A possible function of this brain region was first suggested in the late 1950s in studies using parabiosis (a surgical union of two animals, resulting in chronic blood exchange) between wild-type and VMH-lesioned rats. In these experiments, the paired wild-type animals showed decreased food intake and a substantial loss of fat mass, whereas the phenotype of the lesioned animals was unchanged15. This observation suggested that animals with hypothalamic lesions overproduce a blood-borne signal that normally decreases weight, but, because of the lesion, they cannot respond. The lesion was further suggested to lead to a compensatory increase in the level of humoral factor, thus resulting in weight loss in the paired animal with a normal hypothalamus.
Later studies showed that, similarly to the results after parabiosis of animals with hypothalamic lesions, wild-type mice parabiosed to db mice, and wild type rats parabiosed to fa/fa rats also lost large amounts of weight, thus suggesting that these gene products might be expressed in the hypothalamus11,16,17. Finally, similarly to wild-type mice, ob mice parabiosed to db mice lost copious amounts of weight, thus indicating that they respond normally to the elevated levels of the putative humoral factor circulating in db mice. This finding suggested that ob mice carry a mutation in the humoral factor11,18. In aggregate, these results, derived from several different laboratories over the course of multiple decades, led to an internally consistent hypothesis that ob mice are obese because they do not produce a hormone that regulates food intake and body weight; db mice and fa rats are obese because they do not express the receptor for that factor; and finally, the receptor should be expressed in the hypothalamus. However, despite intensive research, the identity of the putative hormone and its site of expression were unknown until the ob gene was identified19. As described below, subsequent studies confirmed that all elements of this hypothesis are largely correct.
Identification of leptin and its receptor
The ob gene was identified in 1994 through positional cloning and shown to be expressed in adipose tissue as a secreted ~14-kDa polypeptide19. In agreement with the aforementioned hypothesis, the gene product circulates in plasma in normal animals, and its levels are elevated in db mice and decrease after weight loss20,21. Administration of the hormone, named leptin, was found to significantly decrease body weight and adipose tissue mass in ob mice and wild-type animals, but to have no discernible effect in db animals22,23,24. In contrast to weight loss after dieting, leptin had no effect on decreasing lean mass. These and subsequent data from animals have established that leptin levels increase when fat mass increases, thereby suppressing food intake, whereas weight loss leads to a decrease in leptin levels and a consequent increase in food intake. This mechanism maintains homeostatic control of adipose tissue mass within a relatively narrow range, thus serving an important evolutionary function. Although the effects of leptin treatment in humans have been less well studied, the available evidence suggests that leptin serves the same function in humans (described below).
Subsequent studies have shown that this new hormone acts primarily on a cytokine family receptor encoded by the db gene expressed in discrete neural populations in the hypothalamus and elsewhere in the brain25,26,27,28 (Fig. 1). Several splice variants of the leptin receptor exist. All but one of the different mutant strains of db mice and fa rats carry null mutations that affect all of the splice variants26,27,29. However, the original db strain, maintained on the C57BLKS genetic background, has an unusual mutation affecting only one of the splice variants, ObRb, while leaving the function of the other forms unaltered26,27. The ObRb isoform has a long cytoplasmic region containing multiple motifs required for signal transduction, whereas the other isoforms do not. In addition, whereas the other splice variants are expressed in many tissues, ObRb is highly enriched in the hypothalamus26,30. Because the phenotype of this mutant is identical to those of animals with null mutations, these findings provided initial evidence that leptin acts primarily in the hypothalamus. In agreement with this hypothesis, plasma levels of leptin increase after hypothalamic lesions, and low doses of intracerebroventricular leptin replicate the effects of much higher doses of peripheral leptin21,31. A brain-specific knockout of the leptin receptor leads to obesity, whereas brain-specific expression of the leptin receptor suppresses obesity in db mice32,33. Leptin is thus unusual in that it is the only peptide hormone whose principal site of action is in the brain. The only site outside the central nervous system for which a direct effect of leptin has been confirmed is on immune cells34,35,36.
Leptin signal transduction
The leptin receptor is a member of the JAK–Stat family, and leptin-receptor signalling activates the JAK2 kinase, which phosphorylates and activates the Stat3 transcription factor in the hypothalamus as well as in immune cells, but has much weaker effects in other tissues30,37. A point mutation in the Stat3-binding site of the leptin receptor results in recapitulation of much of the obesity of db mice, as does brain-specific knockout of JAK2 and Stat3, thus establishing the importance of this signal-transduction pathway38,39,40, although roles for several other signal transduction pathways have also been suggested41,42,43. Leptin also activates the regulatory proteins SOCS3 and PTP1b, both of which diminish leptin signaling44,45,46,47 (Fig. 2). A mutation of the SOCS3-binding site on the leptin receptor potentiates leptin action, as does a SOCS3 knockout44,47.
A negative feedback loop regulating adipose tissue mass
Leptin modulates food intake and body weight largely, although not exclusively, by stimulating pro-opiomelanocortin (POMC) neurons and inhibiting neuropeptide Y and Agouti-related protein (NPY/AGRP) neurons in the arcuate nucleus of the hypothalamus48,49,50,51. This brain region is adjacent to the median eminence, a circumventricular organ with a porous blood–brain barrier. An earlier study of mice with leptin-receptor knockout in POMC neurons provided evidence that these neurons diminish food intake and body weight after leptin treatment52. However, a recent report showing that a CRISPR knockout of the leptin receptor in AGRP neurons causes massive obesity suggests that this neural population is the principal target mediating leptin’s effects on energy balance52. However, mutations that disrupt melanocortin signalling in animals and humans cause leptin resistance and obesity, thus suggesting that they also mediate some of leptin’s effects (described below and in refs. 53,54,55,56). POMC neurons also regulate glucose metabolism in response to leptin independently of effects on body weight52.
Leptin has been shown to acutely affect the firing rate of POMC and NPY/AGRP neurons in slice preparations via effects on specific ion channels, although it is not known how the leptin signal-transduction pathway alters the function of these channels or how leptin activates firing of some cells (POMC neurons) while inhibiting others (such as NPY/AGRP neurons)57,58,59. Leptin treatment of ob mice also elicits rapid and substantial plasticity of the synaptic inputs to NPY/AGRP and POMC neurons, with opposite effects on each, and is also necessary for the development of a specific hypothalamic projection60,61. These neurons modulate food intake via numerous projections to other brain regions including the parabrachial nucleus, which conveys satiety signals and nausea, and the paraventricular nucleus of the hypothalamus62,63. When active, AGRP neurons convey a negative valence (that is, the unpleasant sensation of hunger), thereby augmenting the drive to eat. Ingestion of food alleviates this negative sensation by suppressing the activity of the NPY/AGRP neurons64. Leptin also modulates the activity of reward pathways and decreases food intake partially by decreasing the hedonic value of food65. Direct effects of leptin on higher centres have also been reported, including the hippocampus, where leptin appears to exert an antidepressant effect66. Numerous other neural populations also respond to leptin and mediate various aspects of leptin’s diverse effects67,68,69,70,71,72,73. Overall, tracing of the neural pathways regulated by leptin has led to a detailed circuit map of the neural pathways that regulate feeding, which now encompasses the most detailed wiring diagram of any complex behavior74.
One of the remaining questions is the mechanism by which leptin crosses the blood–brain barrier and subsequently acts on deeper structures. Approximately half of leptin’s action is mediated by direct effects on dendrites close to the median eminence, which has a permeable blood–brain barrier75. Leptin can also cross the blood–brain barrier and is found in the cerebrospinal fluid76. Recent evidence suggests that leptin may be transported into the cerebrospinal fluid by tanycytes, although leptin might possibly also be transported across the vascular endothelium via a yet-unknown mechanism77. Although a recent report did not find the leptin receptor in tanycytes, the data do not exclude the possibility that other molecules might transport leptin, such as megalin, which transports leptin out of the renal tubules78,79,80
Perhaps the key unanswered question is how and where complex sensory and interoceptive information is processed by the brain to initiate feeding rather than other competing behaviours. In other words, how and where is the decision to eat made? In a classic set of papers, Tinbergen provided a conceptual basis for considering the hierarchy of the motor behaviours needed to complete a goal-directed behavior81. In 1906, Sherrington referred to the anatomic site at which the integrated movements of a reflex are controlled as the final common pathway82,83. Advances in neuroscience now provide an opportunity to overlay specific circuits and neural populations on the ethologic framework provided by Tinbergen and the neurobiologic framework provided by Sherrington. Although the neural basis of most reflexes has been delineated, the sites at which the coordinated movements of feeding are controlled are only recently emerging84. An important advance has been the identification of neurons in the reticular formation and periaqueductal gray that are required for the control of the motor outputs required for feeding. On the basis of this information, together with findings from other studies identifying the primary nodes processing relevant sensory and interoceptive information48, it should now be possible to link the relevant inputs and outputs and to establish the anatomic sites and neural mechanisms that control the initiation of feeding behaviour. Indeed, a circuit map for feeding in Drosophila linking sites of sensory input to sites of motor output has been constructed85, and analogous studies aiming to link inputs and outputs in the mammalian brain are underway. The critical neural population responsible for the decision to eat should be the final node in the circuit that can activate the behaviour and satisfy the following criteria: (i) activation of the key neurons should lead to the initiation of the complete behaviour; (ii) these neurons should directly or indirectly receive relevant inputs and connect to the relevant motor outputs; (iii) these neurons should activate feeding even when other sites that inhibit feeding are also modulated; (iv) inhibiting these neurons should prevent feeding even if other sites that induce feeding are modulated; and (v) activating these neurons should extinguish competing behaviours. Broad advances in neuroscience have made the question of how and where behavioural decisions are made tractable and within reach.
Leptin and the adaptive response to starvation
In addition to increasing appetite, the absence of leptin in ob mice is associated with many physiologic changes including hypothermia, infertility, immune alterations, the insulin resistance of starvation and a euthyroid sick state, among many other changes not generally associated with obesity13. Instead, this set of physiologic changes is generally associated with starvation, thus suggesting that ob mice are obese because their brains interpret a low leptin level as a signal that the adipose tissue mass is dangerously low. For this reason, ob mice, while eating voraciously and showing a massive increase in weight, manifest a syndrome distinct from that generally associated with obesity. This observation initially suggested that leptin levels, in addition to inducing a state of positive energy balance to restore the lost weight, activates an adaptive ‘starvation’ response whose net effect is to conserve energy during times of privation. Consequently, leptin treatment corrects all the aforementioned abnormalities, including the reproductive defects in ob mice86,87. Leptin is also required for the normal onset of puberty88,89. In addition, leptin ameliorates the immune and neuroendocrine alterations that accompany food restriction, although the precise thresholds for these different responses appear to be subtly different34,90. Thus, in leptin-deficient animals, leptin has pleiotropic effects and provides a link between changes in nutritional state and physiologic responses in many (perhaps all) other physiologic systems.
Increases in leptin levels in mice or rats have the opposite effect on energy balance and lead to weight loss by decreasing food intake and suppressing the compensatory decrease in energy expenditure that is typically associated with dieting22,23,24. Overall, there is a graded decrease in fat mass in mice with incremental increases in leptin concentrations within the physiologic range, thus providing evidence that hyperleptinaemia in normally responsive animals restricts weight gain31. However, in contrast to treatment of leptin-deficient animals, increasing the physiologic leptin concentrations has few other discernible effects on other physiologic systems. The weight loss induced by leptin treatment is a result of lipolysis via activation of sympathetic efferent signalling and subsequent decreases in adipose tissue91. However, although analysis of the respiratory quotient indicates that lipids are being oxidized, the sites at which fat is burned have not been precisely determined92.
In the leptin system, similarly to other endocrine systems, increasing leptin concentrations above the physiologic range in rodents decreases weight, and the dose–response curve of the effects of exogenous leptin on energy balance is not linear; in addition, a proportionately greater decrease in food intake and body weight is observed after treatment of animals with absent or low leptin levels compared with treatment of normal, lean animals. Thus the potency of leptin is greater when leptin levels are low, and a diminished response occurs as the levels approach the physiologic range or higher. In aggregate, the data indicate that leptin is a long-term signal that maintains homeostatic control of adipose tissue mass. Abundant evidence has also indicated that hormonal and neural signals from the gastrointestinal tract comprise a short-term system regulating hunger and satiety93,94,95. The brainstem is a primary site of action of these short-term signals, which interact extensively with leptin-activated circuits originating in the hypothalamus. Other signals might possibly also restrain weight gain when obesity develops, although their precise identity has not been determined96,97. Overall, these findings establish that leptin is a key means through which nutritional changes are sensed and elicit adaptive physiologic responses.
Leptin and the pathogenesis of obesity
The identification of the leptin endocrine system has provided a framework for understanding the pathogenesis of nutritional disorders including obesity and other forms of metabolic disease. In general, endocrine disorders can result from hormone deficiency (complete or partial) or hormone resistance. Rodents with a complete deficiency as a consequence of leptin mutations, as well as those with a partial deficiency, lose copious amounts of weight with leptin treatment21,22,23,98. In addition, animals with lipodystrophy—which have diminished leptin levels as a result of defects in adipose tissue differentiation that prevent fat-tissue formation—show marked metabolic improvements on leptin therapy as well as weight loss99.
Another aetiology for diminished leptin expression has recently been identified in animals with a mutation in a fat-specific long non-coding RNA, LncOb91. On both chow and high-fat diets, the knockout mice show increased fat mass with decreased leptin levels, despite becoming more obese. The diet-induced-obese (DIO) LncOb mice also show significant weight loss after leptin treatment, thus suggesting that leptin can decrease weight in animals with a relative leptin deficiency100. Together, these findings indicate that leptin is an effective treatment in obese animals with complete or partial hormone deficiency, and, as illustrated by the phenotype of the LncOb-knockout mice, mutations that dysregulate leptin gene expression can result in a hypoleptinaemic, leptin-responsive form of obesity.
Leptin resistance in obesity
Most forms of obesity in animals are associated with high endogenous plasma leptin levels and a diminished response to exogenous hormone20,31. The normal plasma leptin concentration in animals (and humans) is ~5 ng ml–1. DIO yellow agouti (Ay) and New Zealand obese mice are all highly hyperleptinaemic (levels range between approximately 25 and 100 ng ml–1, and can be even higher) and show little or no response to peripheral leptin infusion31. This diminished response is in contrast to the aforementioned response of normal animals, in which leptin treatment at physiologic levels induces a dose-dependent decrease in food intake, adipose tissue mass and body weight. The finding that lean wild-type animals have lower leptin levels and lose weight on leptin therapy, whereas obese animals have high endogenous levels and do not, indicates that these obese animals are leptin resistant31,101.
Several independent lines of evidence support the conclusion that obesity can result from leptin resistance in animals and humans (described below), and that leptin signalling restrains weight gain in lean and obese animals. This evidence includes the finding that mutations in melanocortin pathways normally activated by leptin lead to obesity, hyperleptinaemia and a decreased response to leptin treatment102 (Table 1). Several other lines of evidence suggest that leptin resistance leads to obesity (Table 1). First, mutations in SOCS3 and PTP1b, which normally shut off leptin signal transduction, enhance leptin signalling and prevent diet-induced obesity in animals fed a high-fat diet44,45,47. Second, in agreement with the previous observation, mice with a mutation in Y935 of the leptin receptor, a site of tyrosine phosphorylation in the intracellular domain and SOCS3 binding, are also resistant to diet-induced obesity and maintain leptin sensitivity. In contrast, elevated levels of SOCS3 are associated with decreased Stat3 phosphorylation and result in obesity, and SOCS3 overexpression in leptin-receptor neurons in transgenic mice leads to leptin resistance and obesity46,103,104,105. Third, extremely obese DIO mice whose weight has stabilized at ~50 g show significant weight gain after treatment with a leptin antagonist106. These data unequivocally establish that obese animals are still dependent on endogenous leptin for maintaining a stable (obese) weight and further suggest that leptin resistance resets the animals’ weights at new higher levels. Fourth, cellular resistance to leptin has been confirmed in electrophysiologic recordings in normal and leptin-resistant cells. Leptin can induce electrophysiologic responses in POMC and AGRP neurons, and these responses are diminished in neurons in slice preparations derived from DIO (leptin-resistant) animals57,107. Fifth, the aforementioned LncOb-knockout mice, with diminished plasma leptin levels, respond to leptin, whereas wild-type DIO mice do not, thus indicating that the latter group (which have higher leptin levels) are leptin resistant, and the knockouts (with lower plasma leptin levels) are not100. However, the LncOb mice still lose weight on leptin therapy despite having higher plasma leptin levels than those wild-type mice fed a chow diet (but lower levels than those of wild-type DIO mice). This finding suggests that an inappropriately low plasma leptin level relative to adipose tissue mass determines the response to leptin therapy and that absolute plasma leptin levels may not be the most useful predictor of the leptin response. Sixth, although leptin may no longer activate Stat3 in obese animals, treatment with amylin, a pancreatic peptide, or celastrol, a small molecule that alleviates endoplasmic reticulum stress, restores leptin-mediated Stat3 phosphorylation. As a consequence, treatment with amylin or celastrol can lead to resensitizaton to leptin and can, alone or in combination with leptin, promote weight loss in DIO mice108,109. Combinations of leptin and other gut hormones show similar weight-reducing effects in DIO animals110,111. Seventh, although some of leptin’s effects are lost in the leptin-resistant state, other effects of leptin are retained, thus suggesting that, similarly to insulin resistance, leptin resistance is selective112. Finally, ob/ob mice receiving a chronic leptin infusion that maintains plasma leptin levels in the physiologic range of 5 ng ml–1 still maintain leptin sensitivity when fed a high-fat diet. Thus, obese animals that cannot become hyperleptinaemic do not develop leptin resistance. This observation suggests that in some instances, as has also been observed for insulin, leptin resistance can be caused by compensatory down-regulation of the response to the endogenous hormone (that is, tachyphylaxis).
Mechanistic basis of leptin resistance
The precise mechanism through which hyperleptinaemia leads to leptin resistance is not known. However, induction of PTP1b and/or SOCS3 in cells expressing the leptin receptor, perhaps secondarily to hyperleptinaemia, is likely to contribute44,45. The causes of leptin resistance appear to be heterogeneous and, as mentioned, also include constitutive defects in the neural circuit downstream of leptin, such as in mice lacking the leptin receptor or mice with defects in melanocortin signalling, such as Ay or melanocortin 4 receptor–knockout mice22,31. Indeed, to date, all genes identified as Mendelian causes of obesity in humans are expressed in the central nervous system, and most are components of the neural circuit modulated by leptin. Decreased transport of leptin across the blood–brain barrier has also been suggested to contribute to leptin resistance both in DIO mice and in New Zealand obese mice, which lose weight after intracerebroventricular but not subcutaneous administration of the hormone31,76. These findings suggest the possibility that decreased transport across the blood–brain barrier, or a lowered Vmax, could cause obesity, although definitive evidence supporting this possibility is lacking. Thus, the aetiology of leptin resistance is known in some individuals, including humans and animals with mutations in the neural circuit that responds to leptin to regulate feeding, whereas in other cases the aetiology is less clear. This scenario is analogous to insulin resistance, whose pathogenesis in the general population is unclear but is known to be caused by discrete mutations in components of the insulin signal-transduction pathway113,114.
The cause of leptin resistance in DIO is less clear, but in this case, hyperleptinaemia contributes115. Whereas C57BL/6J mice become obese on a high-fat diet, other mouse strains do not52. One possibility explaining these findings is that genetic differences among mouse strains and possibly humans may determine whether or not leptin resistance develops as weight is gained and leptin levels increase during times of surfeit. Leptin resistance could thus confer a survival advantage in environments in which starvation is the prevailing threat, whereas maintenance of leptin sensitivity would be selected for in circumstances in which predation is the greater danger (further discussion in ‘Evolutionary considerations’ below). Nonetheless, although the molecular pathogenesis of leptin resistance is known in some cases, the precise cause of leptin resistance in diet-induced obesity or obesity in the human population is largely unknown.
A more comprehensive understanding of the molecular mechanism responsible for the development of leptin resistance could provide a basis for new anti-obesity treatments, although this task will be challenging. It is worth remembering in this context that insulin resistance was first defined in the 1950s, yet the nature of the precise molecular defect in most people with type 2 diabetes is still largely unknown even though in vitro assays (insulin effects on liver, fat and muscle cells) and in vivo assays (insulin clamps) of insulin action are readily available. In contrast, similarly robust means for assaying leptin resistance are lacking. Moreover, leptin acts at a diverse set of neural targets, which display varying degrees of leptin resistance, thus amplifying the difficulty of establishing the identity of the leptin-resistant cell types and the molecular nature of the cellular ‘block’67,68.
Mapping of the sites of leptin resistance and the nature of the molecular defect could provide new approaches for treating metabolic disease by reversing leptin resistance through leptin sensitizers such as celastrol, or amylin and other gut peptides, which show potent weight-reducing effects in DIO mice when co-administered with leptin. Pre-treatment with either agent can restore leptin-mediated Stat3 phosphorylation, although the cellular and molecular mechanisms have not been determined108,109. An alternative approach is to modulate the activity of neurons downstream of the ‘block’. Inhibition of GABAergic neurons in the dorsal raphe nucleus has recently been shown to significantly decrease food intake and body weight in ob and DIO mice, thus showing that this neural population is ‘downstream’ of the site of leptin resistance116. The development of pharmacologic agents that modulate the activity of these and other distal neurons could thus be of therapeutic value, providing a further rationale for intense efforts by many laboratories to map feeding circuits.
Leptin function in humans
Treatment of leptin deficiencies
The effects of leptin in humans, though not as intensively studied as in rodents, are consistent with the findings in mice. Although rarely seen, humans with leptin mutations have extreme obesity and lose copious amounts of weight on leptin therapy, primarily as a result of decreased food intake117,118,119. This response confirms a physiologic role of leptin in humans. Before leptin treatment, these people show constitutive activation of reward centres in the striatum and a general lack of food preference (similarly to starved individuals, they prefer all sources of calories nearly equally), and both of these responses are normalized by leptin therapy120. Leptin treatment of people with congenital leptin deficiency and obese people who have lost weight, in contrast to mice, does not induce a net increase in energy expenditure but instead blunts the decreased energy expenditure normally associated with weight loss89,121. Leptin also corrects the immunologic abnormalities evident in people with leptin mutations, which are similar89 to the abnormalities associated with severe weight loss in mice and humans (not obesity). People with leptin mutations also appear to have an increased incidence of death from bacterial infections122, thus raising the yet-untested possibility that leptin might support immune function in people with extremely low leptin levels, including in cancer cachexia, or chronic inflammation, both of which are associated with an increased risk of death from infectious disease.
Treatment of lipodystrophies
Leptin treatment also has robust effects in other patient populations with pathologically low leptin levels. Patients with complete or partial lipodystrophy and low endogenous levels of leptin show a marked improvement in several metabolic parameters after leptin therapy, including significant decreases in triglycerides, steatosis and haemoglobin A1c, as well as food intake123. On the basis of these results, leptin been approved as a treatment for lipodystrophy in the United States, the European Union and Japan. Leptin has also shown potential benefit in people with insulin-receptor mutations (Rabson–Mendenhall syndrome) and in animals with type 1 diabetes, although supporting data from human trials in people with type 1 diabetes are lacking124,125. The available data further suggest that the anti-diabetic effects of leptin are not solely a result of decreased food intake but that other mechanisms contribute, perhaps via effects on POMC neurons104,126,127,128. In lipodystrophic animals, the metabolic improvement is a result of actions on the brain, thus suggesting that leptin indirectly modulates the activity of autonomic outputs to pancreatic islets, liver, fat and muscle129.
Treatment of hypothalamic amenorrhoea
In people with hypothalamic amenorrhea, an infertility syndrome associated with low endogenous leptin levels, leptin treatment can restore fertility. Some patients have even become pregnant during the treatment period130,131. In addition, low leptin132 levels are predictive of amenorrhoea in underweight women, including people with eating disorders132,133. In those cases, leptin treatment also corrects or improves several neuroendocrine abnormalities, most of which are characteristic of the ‘starved’ state, including increased levels of insulin-like growth factor 1 and thyroid hormone130. Improvements in these abnormalities as a result of exogenous leptin treatment occur, although these individuals lose weight, thus confirming that leptin, not the fat mass itself, is the key signal that suppresses the starvation response. In a small study, leptin has also been shown to have significant beneficial effects on the premature osteoporosis often associated with this condition134.
Leptin therapy and weight loss
The finding that people with a diverse set of conditions associated with low endogenous leptin levels lose weight on leptin therapy further confirms a physiologic role of the hormone in humans. In addition, low endogenous leptin levels have been found to be predictive of subsequent weight gain among a cohort of Pima Native Americans135. These findings raise the possibility that lean or even obese people with similarly low levels might lose weight on leptin therapy. Although leptin has been reported to induce modest weight loss among lean healthy individuals, the primary data were not shown in that study, and a dataset showing the results of leptin treatment in lean individuals was lacking until recently136. However, recent evidence has confirmed that treatment with leptin decreases food intake and adipose tissue mass in lean people (C. S. Mantzoros (Harvard Medical School), unpublished data). In future studies, it will be important to further delineate the threshold for the endogenous leptin level that is predictive of a leptin response among these individuals.
The recent finding that lean people lose weight on leptin therapy is consistent with several additional lines of evidence. As mentioned above, people with lipodystrophy who are treated with leptin typically lose weight. In one study, many such patients lost significant amounts of weight, including three patients whose endogenous leptin levels were in the normal range (~3–10 ng/ml). However, even people with elevated leptin levels might respond to leptin treatment, given that two hyperleptinaemic patients with partial lipodystrophy have been found to lose weight during leptin therapy (R. Brown (NIDDK, NIH), unpublished data). As mentioned, leptin treatment has also been found to decrease weight in people with hypothalamic amenorrhea, although the starting leptin level in the individual patients was not reported130. Finally, leptin treatment in a small cohort of overweight or mildly obese men with hepatic steatosis selected for low endogenous plasma leptin levels (<10 ng/ml) has been found to result in decreased steatosis and significant weight loss (E. Oral (University of Michigan), unpublished data).
In agreement with the possibility that lower leptin levels might be predictive of weight loss after leptin treatment, a recent report has shown highly significant (~12%) weight loss in obese people in the lowest decile of endogenous leptin concentration (<16 ng/ml in women and <5 ng/ml in men), although the most pronounced effect may have been in individuals with even lower leptin levels137. Different thresholds were defined for each sex because women usually have a higher body fat content and hence higher baseline plasma leptin levels than men. Thus, although a low starting level is predictive of a response to the hormone, the precise leptin level that is predictive of significant weight loss among obese people (and presumably lean people with similar leptin levels) remains to be determined. It will thus be crucial to rigorously define the threshold leptin level that is predictive of a biologic response, as well as to determine the dose–response relationship of leptin therapy among lean and obese people with low and normal endogenous levels.
In most obese individuals, however, plasma leptin levels are significantly elevated and are highly correlated with adipose tissue mass, although each analysed cohort contains a subset of individuals with relatively low leptin levels20,138. Several single-nucleotide polymorphisms have been identified to be associated with low plasma leptin levels in obesity, one of which is in the human homolog of LncOb RNA100,139. Leptin treatment of obese patients with elevated plasma leptin levels, in contrast to patients with low hormone levels, in a patient cohort that was not selected according to starting leptin levels, did not result in a significant overall response to leptin treatment136. However, other data have shown a significant increase in the proportion of patients losing more than 10% or 5% of their weight under leptin versus placebo treatment, and a retrospective analysis has further suggested enrichment in responders among patients with lower starting plasma leptin levels137.
Differing plasma leptin levels in people with type 1 versus type 2 obesity
The aggregate data suggest that a subset of obese people with low or relatively low endogenous leptin levels lose weight on leptin therapy and that the likelihood of a response diminishes at progressively higher hormone levels. The finding that lean people and some obese people with low plasma leptin respond, whereas most do not, further indicates that non-responders are leptin resistant. Although the cause of leptin resistance is known in obese people with mutations that decrease leptin or melanocortin signalling (that is, people with mutations in the leptin receptor, POMC or melanocortin 4 receptor), in most cases, the cause is unknown102. Leptin therapy is of limited value in this setting, but alternative therapeutic approaches remain viable, including using leptin in combination with amylin, which results in significant weight loss in humans (~13.7%)108. Clinical trials with celastrol, a leptin sensitizer that decreases weight in DIO mice, are also underway109.
The finding that people with low plasma hormone levels show a greater response than those with high levels is reminiscent of diabetes (Table 2): individuals with type 1 diabetes have low plasma insulin levels and respond well to insulin treatment, whereas individuals with type 2 diabetes are insulin resistant and require high doses of insulin, often in combination with insulin-sensitizing drugs. This parallel finding suggests that obesity might be viewed in a similar manner: type 1 obesity (or metabolic disease) would be defined as obesity associated with low plasma leptin levels and leptin sensitivity, whereas type 2 obesity would be characterized by obesity in the presence of elevated plasma leptin levels and leptin resistance. Similarly to that in diabetes, the therapeutic approach to each obesity type would differ. Although the precise leptin levels that are predictive of leptin sensitivity have not been clearly defined, the aforementioned evidence suggests that obese people with leptin levels in the same range as those in lean people (<5 ng ml–1 for men and <16 ng ml–1 for women) might respond.
However, there are at least two notable differences in leptin and obesity versus insulin and diabetes. First, there are likely to be fewer obese people with low leptin levels than with diabetes and insulin deficiency (that is, type 1 diabetes). The percentage of obese people in the former category remains to be determined. Second, provided that a sufficiently high dose is administered, insulin can invariably elicit a biologic response, even in the most insulin-resistant people. However, even very high levels of exogenous leptin do not elicit a significant response in leptin-resistant animals or humans. This difference may reflect the potentially lethal consequences of complete loss of insulin action from ketoacidosis, as evidenced by the paucity of null mutations in insulin or the insulin receptor in the general population. In contrast, individuals with null mutations in leptin or its receptor exist, although the associated infertility and risk of predation has probably diminished the allele frequency in the population102,117. Defects in leptin signalling are strongly selected against, on the basis of the very high probability of loss intolerance (pLi) for the leptin receptor (0.99) and the medium score for leptin (0.46) in the gnomAD database, thus indicating that neither gene tolerates mutations140. The slightly lower score for leptin probably reflects that sequence variation has a greater effect in ligands than receptors. Nonetheless, the evolutionary forces that shaped the responses to insulin and leptin are likely to be distinct.
The size of the adipose tissue mass is under evolutionary selection, particularly in mammals, in which adipose tissue has evolved as a highly specialized organ for storing lipids as a source of calories during times of privation5. The thrifty-gene hypothesis posits that in the ‘wild’, where calories are sometimes limiting, the efficient deposition of calories as fat provides protection against starvation during famines. This hypothesis further posits that when calories are readily available, such as in the modern environment, this adaptation leads to diabetes and obesity.
An orthodox interpretation of this hypothesis related to the consequences of famines has recently been questioned on the basis of the suggestion that famines sufficiently severe to exert selective pressure on people with decreased adipose tissue mass are relatively infrequent7,8. However, a minimal level of adiposity is probably necessary to limit the potentially dangerous consequences of the food insecurity that often affects hunter–gatherer populations, which do not farm crops or store large amounts of food. Thus, a minimal level of adiposity is likely to be required to promote survival during ‘lean’ times7,141,142. Sufficient adipose tissue mass can also help to maintain core temperature in colder climates by providing insulation as well as a source of calories for thermogenesis. Adequate fat stores also prevent the infertility and immune suppression resulting from hypoleptinaemia that can develop with nutritional deprivation86,88,90. The deleterious effects of low leptin levels on immune function may have become increasingly important as humans began to congregate in villages and later cities, thus leading to epidemics of infectious disease143. In addition, adequate stores of adipose tissue have been suggested to provide a selective advantage for surviving the anorexia developing after infections8.
The predator-release hypothesis provides an alternative to the thrifty-gene hypothesis and suggests that obesity can be maladaptive by limiting mobility and thus an animal’s ability to escape danger, thus rendering affected individuals more susceptible to predators7. This hypothesis further suggests that the absence of predators among most modern human populations has led to an increased incidence of obesity as a result of genetic drift due to an accumulation of predisposing alleles that are no longer selected against. Additional selective pressure against obesity may also have been present as humans evolved, even after the risk of predators diminished, because the tendency of obese mothers to develop gestational diabetes increases the frequency of miscarriages and also leads to larger babies. An increased foetal size can lead to cephalopelvic disproportion, which can have catastrophic consequences for the mother and child when Caesarean sections are unavailable. However because the negative sequelae of obesity are generally present in older individuals, reproduction has generally been completed. Thus the extent to which obesity is selected against in ‘modern’ human populations has been controversial and has led to questions regarding whether there is a selective advantage for longevity (Box 1). Recent evidence has thus suggested that such a selective advantage does exist and thus would indicate that obesity is still under genetic selection6.
The thrifty-gene and predator-release hypotheses are not mutually exclusive, and together they suggest that the tendency of a population toward obesity is a result of balancing selection between the relative risk of periodic food insecurity (and potentially famine) and the risk of predation in the environment in which that population evolved. These distinct evolutionary pressures may potentially explain the profound predisposition of Pacific Islanders (and other Aboriginal populations) to obesity6. Pacific populations were not typically vulnerable to predators but instead experienced substantial food insecurity while traveling extremely long distances to populate new islands and in the aftermath of frequent typhoons. In contrast, in environments with larger numbers of predators and the development of agriculture to provide constant supplies of nutrients, such as in Eurasia, obesity and diabetes are less prevalent6. Indeed, Caucasians, especially compared with many Aboriginal populations, appear to be relatively resistant to obesity.
The development of leptin resistance could thus be adaptive or maladaptive, depending on the environment in which the founding population evolved. In an environment where predation is the primary risk, maintaining leptin sensitivity over a broad dynamic range would prevent the development of obesity and promote survival, whereas leptin resistance would be maladaptive. In contrast, in environments where the prevailing risk is food insecurity and starvation, the development of acquired leptin resistance and obesity, perhaps secondarily to elevated leptin levels during times of surfeit, would enable survival through intermittent periods of food insecurity or famine. In agreement with this possibility, as mentioned, hyperleptinaemia is required for the development of obesity in C57BL/6J mice fed a high-fat diet115. In addition, although some mouse strains, such as C57BL/6J, become obese when fed a high diet, other mouse strains, such as Akr, do not144. Thus, some of the alleles that predispose individuals to either obesity or obesity resistance may exert their effects by determining whether hyperleptinaemia leads to tachyphylaxis and the development of leptin resistance. The relative risk of starvation as well as predation also differs among species, and the particular evolutionary pressures experienced by each species might be able to explain the likelihood of developing obesity when calories are abundant. Mice, for example, have a higher metabolic rate than humans and can survive for only a few days without food, whereas humans and other large mammals can survive for extended periods. This difference may explain why leptin mutations in mice are associated with a more severe set of abnormalities than null mutations in humans.
Overall, the level of adiposity has a considerable effect on evolutionary fitness. The evolution of leptin and an endocrine system that sets the level of adiposity has provided a means for ensuring the optimal level of adiposity in particular environmental contexts. Mouse strains differ in their susceptibility to obesity when fed a high-fat diet, and genetic crosses between them have identified polygenes that predispose mice to either obesity or obesity resistance, although the causal genes have not been identified144,145. In addition to continuing genetic studies in humans, the identification of allelic variants that predispose people to obesity or obesity resistance could enable the identification of new components of the biologic system that regulates adiposity and lead to a more comprehensive understanding of the evolutionary factors that shape the level of adiposity in different populations.
Summary and outlook
Before the identification of leptin, adipose tissue was considered by many to be a balloon-like vessel for lipids that passively stored energy when food was consumed and disbursed it in times of privation. Largely because of the identification of leptin, adipose tissue is now viewed as a highly dynamic organ playing a crucial role in the systemic control of energy balance; communicating nutritional changes to central-nervous-system centres that control appetite; and linking changes in the levels of energy stores to adaptive changes in the neuroendocrine axis, immune function and the function of arguably all other physiologic systems. Numerous new insights have thus followed the cloning of the ob gene, including (i) the identification of a new hormone, leptin, and endocrine system regulating food intake and body weight; (ii) the delineation of a physiologic mechanism through which changes in nutritional state regulate (all) other physiologic systems; (iii) the identification of a genetic basis for obesity, with the finding that leptin mutations cause severe obesity that can be successfully treated by hormone replacement; (iv) the provision of an entry point for studies of the neural control of food intake, revealing links between the homeostatic and hedonic control of appetite; (v) the demonstration that a substantial fraction of morbid obesity is the result of Mendelian defects in the neural circuit that is modulated by leptin; (vi) the identification of several leptin-deficiency syndromes including lipodystrophy (a cause of severe insulin resistance and diabetes) and hypothalamic amenorrhea, both of which can be treated with leptin; (vii) the realization that the pathogenesis of obesity is heterogeneous (a leptin-sensitive subset of obese individuals express decreased amounts of the hormone (analogously to type 1 diabetes), and most obese people show leptin resistance (analogously to type II diabetes); moreover, combinations of leptin with short-term signals and/or leptin sensitizers show potential for treating leptin-resistant obesity); and (viii) the identification of leptin provides an alternative to the notion that obesity is a result of a lack of willpower that can be treated by merely advising people to eat less and exercise more146,147. Tens of thousands of papers have now been written on these topics, and still more are likely to come.
Bray, G. A. Obesity: historical development of scientific and cultural ideas. Int. J. Obes. 14, 909–926 (1990).
Friedman, J. M. A war on obesity, not the obese. Science 299, 856–858 (2003).
Adolph, E. F. Urges to eat and drink in rats. Am. J. Physiol. 151, 110–125 (1947).
Kennedy, G. C. The role of depot fat in the hypothalamic control of food intake in the rat. Proc. R. Soc. Lond. 140, 578–596 (1953).
Neel, J. V. Diabetes mellitus: a “thrifty” genotype rendered detrimental by “progress. Am. J. Hum. Genet. 14, 353–362 (1962).
Diamond, J. The double puzzle of diabetes. Nature 423, 599–602 (2003).
Speakman, J. R. A nonadaptive scenario explaining the genetic predisposition to obesity: the “predation release” hypothesis. Cell Metab. 6, 5–12 (2007).
Speakman, J. R. The evolution of body fatness: trading off disease and predation risk. J. Exp. Biol. 221(Suppl. 1), jeb167254 (2018).
Ingalls, A. M., Dickie, M. M. & Snell, G. D. Obese, a new mutation in the house mouse. J. Hered. 41, 317–318 (1950).
Hummel, K. P., Dickie, M. M. & Coleman, D. L. Diabetes, a new mutation in the mouse. Science 153, 1127–1128 (1966).
Coleman, D. L. Obese and diabetes: two mutant genes causing diabetes-obesity syndromes in mice. Diabetologia 14, 141–148 (1978).
Zucker, L. M. & Zucker, T. F. Fatty, a new mutation in the rat. J. Hered. 52, 275–278 (1961).
Friedman, J. M. & Halaas, J. L. Leptin and the regulation of body weight in mammals. Nature 395, 763–770 (1998).
Hetherington, A. W. & Ranson, S. W. The spontaneous activity and food intake of rats with hypothalamic lesions. Am. J. Physiol. 136, 609–617 (1942).
Hervey, G. R. The effects of lesions in the hypothalamus in parabiotic rats. J. Physiol. (Lond.) 145, 336–352 (1959).
Coleman, D. L. & Hummel, K. P. Effects of parabiosis of normal with genetically diabetic mice. Am. J. Physiol. 217, 1298–1304 (1969).
Harris, R. B. S., Hervey, E., Hervey, G. R. & Tobin, G. Body composition of lean and obese Zucker rats in parabiosis. Int. J. Obes. 11, 275–283 (1987).
Coleman, D. L. Effects of parabiosis of obese with diabetes and normal mice. Diabetologia 9, 294–298 (1973).
Zhang, Y. et al. Positional cloning of the mouse obese gene and its human homologue. Nature 372, 425–432 (1994).
Maffei, M. et al. Leptin levels in human and rodent: measurement of plasma leptin and ob RNA in obese and weight-reduced subjects. Nat. Med. 1, 1155–1161 (1995).
Maffei, M. et al. Increased expression in adipocytes of ob RNA in mice with lesions of the hypothalamus and with mutations at the db locus. Proc. Natl Acad. Sci. USA 92, 6957–6960 (1995).
Halaas, J. L. et al. Weight-reducing effects of the plasma protein encoded by the obese gene. Science 269, 543–546 (1995).
Campfield, L. A., Smith, F. J., Guisez, Y., Devos, R. & Burn, P. Recombinant mouse OB protein: evidence for a peripheral signal linking adiposity and central neural networks. Science 269, 546–549 (1995).
Pelleymounter, M. A. et al. Effects of the obese gene product on body weight regulation in ob/ob mice. Science 269, 540–543 (1995).
Tartaglia, L. A. et al. Identification and expression cloning of a leptin receptor, OB-R. Cell 83, 1263–1271 (1995).
Lee, G. H. et al. Abnormal splicing of the leptin receptor in diabetic mice. Nature 379, 632–635 (1996).
Chen, H. et al. Evidence that the diabetes gene encodes the leptin receptor: identification of a mutation in the leptin receptor gene in db/db mice. Cell 84, 491–495 (1996).
Fei, H. et al. Anatomic localization of alternatively spliced leptin receptors (Ob-R) in mouse brain and other tissues. Proc. Natl Acad. Sci. USA 94, 7001–7005 (1997).
Lee, G. et al. Leptin receptor mutations in 129 db 3J /db 3J mice and NIH fa cp/fa cp rats. Mamm. Genome 8, 445–447 (1997).
Vaisse, C. et al. Leptin activation of Stat3 in the hypothalamus of wild-type and ob/ob mice but not db/db mice. Nat. Genet. 14, 95–97 (1996).
Halaas, J. L. et al. Physiological response to long-term peripheral and central leptin infusion in lean and obese mice. Proc. Natl Acad. Sci. USA 94, 8878–8883 (1997).
Cohen, P. et al. Selective deletion of leptin receptor in neurons leads to obesity. J. Clin. Invest. 108, 1113–1121 (2001).
Kowalski, T. J., Liu, S.-M., Leibel, R. L. & Chua, S. C. Jr. Transgenic complementation of leptin-receptor deficiency. I. Rescue of the obesity/diabetes phenotype of LEPR-null mice expressing a LEPR-B transgene. Diabetes 50, 425–435 (2001).
Lord, G. M. et al. Leptin modulates the T-cell immune response and reverses starvation-induced immunosuppression. Nature 394, 897–901 (1998).
Mackey-Lawrence, N. M. & Petri, W. A. Jr. Leptin and mucosal immunity. Mucosal Immunol. 5, 472–479 (2012).
Reis, B. S. et al. Leptin receptor signaling in T cells is required for Th17 differentiation. J. Immunol. 194, 5253–5260 (2015).
Ghilardi, N. et al. Defective STAT signaling by the leptin receptor in diabetic mice. Proc. Natl Acad. Sci. USA 93, 6231–6235 (1996).
Bates, S. H. et al. STAT3 signalling is required for leptin regulation of energy balance but not reproduction. Nature 421, 856–859 (2003).
Gao, Q. et al. Disruption of neural signal transducer and activator of transcription 3 causes obesity, diabetes, infertility, and thermal dysregulation. Proc. Natl Acad. Sci. USA 101, 4661–4666 (2004).
Robertson, S. et al. Insufficiency of Janus kinase 2-autonomous leptin receptor signals for most physiologic leptin actions. Diabetes 59, 782–790 (2010).
Cota, D. et al. Hypothalamic mTOR signaling regulates food intake. Science 312, 927–930 (2006).
Hill, J. W. et al. Acute effects of leptin require PI3K signaling in hypothalamic proopiomelanocortin neurons in mice. J. Clin. Invest. 118, 1796–1805 (2008).
Leshan, R. L., Greenwald-Yarnell, M., Patterson, C. M., Gonzalez, I. E. & Myers, M. G. Jr. Leptin action through hypothalamic nitric oxide synthase-1-expressing neurons controls energy balance. Nat. Med. 18, 820–823 (2012).
Mori, H. et al. Socs3 deficiency in the brain elevates leptin sensitivity and confers resistance to diet-induced obesity. Nat. Med. 10, 739–743 (2004).
Bence, K. K. et al. Neuronal PTP1B regulates body weight, adiposity and leptin action. Nat. Med. 12, 917–924 (2006).
Björnholm, M. et al. Mice lacking inhibitory leptin receptor signals are lean with normal endocrine function. J. Clin. Invest. 117, 1354–1360 (2007).
Bjørbaek, C., Elmquist, J. K., Frantz, J. D., Shoelson, S. E. & Flier, J. S. Identification of SOCS-3 as a potential mediator of central leptin resistance. Mol. Cell 1, 619–625 (1998).
Friedman, J. M. The alphabet of weight control. Nature 385, 119–120 (1997).
Stephens, T. W. et al. The role of neuropeptide Y in the antiobesity action of the obese gene product. Nature 377, 530–532 (1995).
Cowley, M. A. et al. Leptin activates anorexigenic POMC neurons through a neural network in the arcuate nucleus. Nature 411, 480–484 (2001).
Elias, C. F. et al. Leptin differentially regulates NPY and POMC neurons projecting to the lateral hypothalamic area. Neuron 23, 775–786 (1999).
Balthasar, N. et al. Leptin receptor signaling in POMC neurons is required for normal body weight homeostasis. Neuron 42, 983–991 (2004).
Gao, Y. et al. TrpC5 mediates acute leptin and serotonin effects via Pomc neurons. Cell Rep. 18, 583–592 (2017).
Smith, M. A. et al. Calcium channel CaV2.3 subunits regulate hepatic glucose production by modulating leptin-induced excitation of arcuate pro-opiomelanocortin neurons. Cell Rep. 25, 278–287.e4 (2018).
Caron, A. et al. POMC neurons expressing leptin receptors coordinate metabolic responses to fasting via suppression of leptin levels. eLife 7, e33710 (2018).
Xu, J. et al. Genetic identification of leptin neural circuits in energy and glucose homeostases. Nature 556, 505–509 (2018).
Baver, S. B. et al. Leptin modulates the intrinsic excitability of AgRP/NPY neurons in the arcuate nucleus of the hypothalamus. J. Neurosci. 34, 5486–5496 (2014).
Takahashi, K. A. & Cone, R. D. Fasting induces a large, leptin-dependent increase in the intrinsic action potential frequency of orexigenic arcuate nucleus neuropeptide Y/Agouti-related protein neurons. Endocrinology 146, 1043–1047 (2005).
Spanswick, D., Smith, M. A., Groppi, V. E., Logan, S. D. & Ashford, M. L. J. Leptin inhibits hypothalamic neurons by activation of ATP-sensitive potassium channels. Nature 390, 521–525 (1997).
Pinto, S. et al. Rapid rewiring of arcuate nucleus feeding circuits by leptin. Science 304, 110–115 (2004).
Bouret, S. G., Draper, S. J. & Simerly, R. B. Trophic action of leptin on hypothalamic neurons that regulate feeding. Science 304, 108–110 (2004).
Wu, Q., Boyle, M. P. & Palmiter, R. D. Loss of GABAergic signaling by AgRP neurons to the parabrachial nucleus leads to starvation. Cell 137, 1225–1234 (2009).
Atasoy, D., Betley, J. N., Su, H. H. & Sternson, S. M. Deconstruction of a neural circuit for hunger. Nature 488, 172–177 (2012).
Betley, J. N. et al. Neurons for hunger and thirst transmit a negative-valence teaching signal. Nature 521, 180–185 (2015).
Domingos, A. I. et al. Leptin regulates the reward value of nutrient. Nat. Neurosci. 14, 1562–1568 (2011).
Lu, X.-Y., Kim, C. S., Frazer, A. & Zhang, W. Leptin: a potential novel antidepressant. Proc. Natl Acad. Sci. USA 103, 1593–1598 (2006).
Scott, M. M. et al. Leptin targets in the mouse brain. J. Comp. Neurol. 514, 518–532 (2009).
Leshan, R. L., Björnholm, M., Münzberg, H. & Myers, M. G. Jr. Leptin receptor signaling and action in the central nervous system. Obesity (Silver Spring) 14(Suppl. 5), 208S–212S (2006).
Williams, K. W., Zsombok, A. & Smith, B. N. Rapid inhibition of neurons in the dorsal motor nucleus of the vagus by leptin. Endocrinology 148, 1868–1881 (2007).
Williams, K. W. & Smith, B. N. Rapid inhibition of neural excitability in the nucleus tractus solitarii by leptin: implications for ingestive behaviour. J. Physiol. (Lond.) 573, 395–412 (2006).
Dhillon, H. et al. Leptin directly activates SF1 neurons in the VMH, and this action by leptin is required for normal body-weight homeostasis. Neuron 49, 191–203 (2006).
Vong, L. et al. Leptin action on GABAergic neurons prevents obesity and reduces inhibitory tone to POMC neurons. Neuron 71, 142–154 (2011).
Leinninger, G. M. et al. Leptin action via neurotensin neurons controls orexin, the mesolimbic dopamine system and energy balance. Cell Metab. 14, 313–323 (2011).
Andermann, M. L. & Lowell, B. B. Toward a wiring diagram understanding of appetite control. Neuron 95, 757–778 (2017).
Peng, Y. et al. A general method for insertion of functional proteins within proteins via combinational selection of permissive junctions. Chem. Biol. 22, 1134–1143 (2015).
Banks, W. A. & Farrell, C. L. Impaired transport of leptin across the blood-brain barrier in obesity is acquired and reversible. Am. J. Physiol. Endocrinol. Metab. 285, E10–E15 (2003).
Balland, E. et al. Hypothalamic tanycytes are an ERK-gated conduit for leptin into the brain. Cell Metab. 19, 293–301 (2014).
Yoo, S., Cha, D., Kim, D. W., Hoang, T. V. & Blackshaw, S. Tanycyte-independent control of hypothalamic leptin signaling. Front. Neurosci. 13, 240 (2019).
Yoo, S. et al. Ablation of tanycytes of the arcuate nucleus and median eminence increases visceral adiposity and decreases insulin sensitivity in male mice. Preprint at bioRxiv https://doi.org/10.1101/637587 (2019).
Ceccarini, G. et al. PET imaging of leptin biodistribution and metabolism in rodents and primates. Cell Metab. 10, 148–159 (2009).
Tinbergen, N. The Hierarchical Organization of Nervous Mechanisms Underlying Instinctive Behaviour. Symp. Soc. Exp. Biol. 4, 305–312 (1950).
Sherrington, C. S. The Integrative Action of the Nervous System (Yale University Press, 1906).
Burke, R. E. Sir Charles Sherrington’s the integrative action of the nervous system: a centenary appreciation. Brain 130, 887–894 (2007).
Han, W. et al. Integrated control of predatory hunting by the central nucleus of the amygdala. Cell 168, 311–324.e18 (2017).
Miroschnikow, A. et al. Convergence of monosynaptic and polysynaptic sensory paths onto common motor outputs in a Drosophila feeding connectome. eLife 7, e40247 (2018).
Barash, I. A. et al. Leptin is a metabolic signal to the reproductive system. Endocrinology 137, 3144–3147 (1996).
Chehab, F. F., Lim, M. E. & Lu, R. Correction of the sterility defect in homozygous obese female mice by treatment with the human recombinant leptin. Nat. Genet. 12, 318–320 (1996).
Chehab, F. F., Mounzih, K., Lu, R. & Lim, M. E. Early onset of reproductive function in normal female mice treated with leptin. Science 275, 88–90 (1997).
Farooqi, I. S. et al. Beneficial effects of leptin on obesity, T cell hyporesponsiveness, and neuroendocrine/metabolic dysfunction of human congenital leptin deficiency. J. Clin. Invest. 110, 1093–1103 (2002).
Ahima, R. S. et al. Role of leptin in the neuroendocrine response to fasting. Nature 382, 250–252 (1996).
Zeng, W. et al. Sympathetic neuro-adipose connections mediate leptin-driven lipolysis. Cell 163, 84–94 (2015).
Singh, A. et al. Leptin-mediated changes in hepatic mitochondrial metabolism, structure, and protein levels. Proc. Natl Acad. Sci. USA 106, 13100–13105 (2009).
Clemmensen, C. et al. Gut-brain cross-talk in metabolic control. Cell 168, 758–774 (2017).
Mayer, E. A. Gut feelings: the emerging biology of gut-brain communication. Nat. Rev. Neurosci. 12, 453–466 (2011).
Berthoud, H. R. Vagal and hormonal gut-brain communication: from satiation to satisfaction. Neurogastroenterol. Motil. 20(Suppl. 1), 64–72 (2008).
Ravussin, Y. et al. Evidence for a non-leptin system that defends against weight gain in overfeeding. Cell Metab. 28, 289–299.e5 (2018).
Jansson, J. O. et al. Body weight homeostat that regulates fat mass independently of leptin in rats and mice. Proc. Natl Acad. Sci. USA 115, 427–432 (2018).
Ioffe, E., Moon, B., Connolly, E. & Friedman, J. M. Abnormal regulation of the leptin gene in the pathogenesis of obesity. Proc. Natl Acad. Sci. USA 95, 11852–11857 (1998).
Shimomura, I., Hammer, R. E., Ikemoto, S., Brown, M. S. & Goldstein, J. L. Leptin reverses insulin resistance and diabetes mellitus in mice with congenital lipodystrophy. Nature 401, 73–76 (1999).
Dallner, O. S. et al. Dysregulation of a long noncoding RNA reduces leptin leading to a leptin-responsive form of obesity. Nat. Med. 25, 507–516 (2019).
Frederich, R. C. et al. Leptin levels reflect body lipid content in mice: evidence for diet-induced resistance to leptin action. Nat. Med. 1, 1311–1314 (1995).
Barsh, G. S., Farooqi, I. S. & O’Rahilly, S. Genetics of body-weight regulation. Nature 404, 644–651 (2000).
Wunderlich, C. M., Hövelmeyer, N. & Wunderlich, F. T. Mechanisms of chronic JAK-STAT3-SOCS3 signaling in obesity. JAK-STAT 2, e23878 (2013).
Reed, A. S. et al. Functional role of suppressor of cytokine signaling 3 upregulation in hypothalamic leptin resistance and long-term energy homeostasis. Diabetes 59, 894–906 (2010).
Bjorbak, C. et al. SOCS3 mediates feedback inhibition of the leptin receptor via Tyr985. J. Biol. Chem. 275, 40649–40657 (2000).
Ottaway, N. et al. Diet-induced obese mice retain endogenous leptin action. Cell Metab. 21, 877–882 (2015).
Enriori, P. J. et al. Diet-induced obesity causes severe but reversible leptin resistance in arcuate melanocortin neurons. Cell Metab. 5, 181–194 (2007).
Roth, J. D. et al. Leptin responsiveness restored by amylin agonism in diet-induced obesity: evidence from nonclinical and clinical studies. Proc. Natl Acad. Sci. USA 105, 7257–7262 (2008).
Liu, J., Lee, J., Salazar Hernandez, M. A., Mazitschek, R. & Ozcan, U. Treatment of obesity with celastrol. Cell 161, 999–1011 (2015).
Müller, T. D. et al. Restoration of leptin responsiveness in diet-induced obese mice using an optimized leptin analog in combination with exendin-4 or FGF21. J. Pept. Sci. 18, 383–393 (2012).
Clemmensen, C. et al. GLP-1/glucagon coagonism restores leptin responsiveness in obese mice chronically maintained on an obesogenic diet. Diabetes 63, 1422–1427 (2014).
Mark, A. L. Selective leptin resistance revisited. Am. J. Physiol. Regul. Integr. Comp. Physiol. 305, R566–R581 (2013).
Pessin, J. E. & Saltiel, A. R. Signaling pathways in insulin action: molecular targets of insulin resistance. J. Clin. Invest. 106, 165–169 (2000).
Boucher, J., Kleinridders, A. & Kahn, C. R. Insulin receptor signaling in normal and insulin-resistant states. Cold Spring Harb. Perspect. Biol. 6, a009191 (2014).
Knight, Z. A., Hannan, K. S., Greenberg, M. L. & Friedman, J. M. Hyperleptinemia is required for the development of leptin resistance. PLoS One 5, e11376 (2010).
Nectow, A. R. et al. Identification of a brainstem circuit controlling feeding. Cell 170, 429–442.e11 (2017).
Montague, C. T. et al. Congenital leptin deficiency is associated with severe early-onset obesity in humans. Nature 387, 903–908 (1997).
Farooqi, I. S. et al. Effects of recombinant leptin therapy in a child with congenital leptin deficiency. N. Engl. J. Med. 341, 879–884 (1999).
Licinio, J. et al. Phenotypic effects of leptin replacement on morbid obesity, diabetes mellitus, hypogonadism, and behaviour in leptin-deficient adults. Proc. Natl Acad. Sci. USA 101, 4531–4536 (2004).
Farooqi, I. S. et al. Leptin regulates striatal regions and human eating behaviour. Science 317, 1355 (2007).
Rosenbaum, M. et al. Low-dose leptin reverses skeletal muscle, autonomic, and neuroendocrine adaptations to maintenance of reduced weight. J. Clin. Invest. 115, 3579–3586 (2005).
Ozata, M., Ozdemir, I. C. & Licinio, J. Human leptin deficiency caused by a missense mutation: multiple endocrine defects, decreased sympathetic tone, and immune system dysfunction indicate new targets for leptin action, greater central than peripheral resistance to the effects of leptin, and spontaneous correction of leptin-mediated defects. J. Clin. Endocrinol. Metab. 84, 3686–3695 (1999).
Oral, E. A. et al. Leptin-replacement therapy for lipodystrophy. N. Engl. J. Med. 346, 570–578 (2002).
Yu, X., Park, B. H., Wang, M. Y., Wang, Z. V. & Unger, R. H. Making insulin-deficient type 1 diabetic rodents thrive without insulin. Proc. Natl Acad. Sci. USA 105, 14070–14075 (2008).
Cochran, E. et al. Efficacy of recombinant methionyl human leptin therapy for the extreme insulin resistance of the Rabson-Mendenhall syndrome. J. Clin. Endocrinol. Metab. 89, 1548–1554 (2004).
Brown, R. J. et al. Metreleptin-mediated improvements in insulin sensitivity are independent of food intake in humans with lipodystrophy. J. Clin. Invest. 128, 3504–3516 (2018).
Brown, R. J. et al. Effects of metreleptin in pediatric patients with lipodystrophy. J. Clin. Endocrinol. Metab. 102, 1511–1519 (2017).
Lee, H. L. et al. Effects of metreleptin on proteinuria in patients with lipodystrophy. J. Clin. Endocrinol. Metab. https://doi.org/10.1210/jc.2019-00200 (2019).
Asilmaz, E. et al. Site and mechanism of leptin action in a rodent form of congenital lipodystrophy. J. Clin. Invest. 113, 414–424 (2004).
Welt, C. K. et al. Recombinant human leptin in women with hypothalamic amenorrhea. N. Engl. J. Med. 351, 987–997 (2004).
Chou, S. H. et al. Leptin is an effective treatment for hypothalamic amenorrhea. Proc. Natl Acad. Sci. USA 108, 6585–6590 (2011).
Köpp, W. et al. Low leptin levels predict amenorrhea in underweight and eating disordered females. Mol. Psychiatry 2, 335–340 (1997).
Hebebrand, J. et al. Leptin levels in patients with anorexia nervosa are reduced in the acute stage and elevated upon short-term weight restoration. Mol. Psychiatry 2, 330–334 (1997).
Sienkiewicz, E. et al. Long-term metreleptin treatment increases bone mineral density and content at the lumbar spine of lean hypoleptinemic women. Metabolism 60, 1211–1221 (2011).
Ravussin, E. et al. Relatively low plasma leptin concentrations precede weight gain in Pima Indians. Nat. Med. 3, 238–240 (1997).
Heymsfield, S. B. et al. Recombinant leptin for weight loss in obese and lean adults: a randomized, controlled, dose-escalation trial. J. Am. Med. Assoc. 282, 1568–1575 (1999).
Depaoli, A., Long, A., Fine, G.M., Stewart, M. & O’Rahilly, S. Efficacy of metreleptin for weight loss in overweight and obese adults with low leptin levels. Diabetes https://doi.org/10.2337/db18-296-LB (2018).
Considine, R. V. et al. Serum immunoreactive-leptin concentrations in normal-weight and obese humans. N. Engl. J. Med. 334, 292–295 (1996).
Kilpeläinen, T. O. et al. Genome-wide meta-analysis uncovers novel loci influencing circulating leptin levels. Nat. Commun. 7, 10494 (2016).
Lek, M. et al. Analysis of protein-coding genetic variation in 60,706 humans. Nature 536, 285–291 (2016).
Dhurandhar, E. J. The food-insecurity obesity paradox: A resource scarcity hypothesis. Physiol. Behav. 162, 88–92 (2016).
Prentice, A. M., Hennig, B. J. & Fulford, A. J. Evolutionary origins of the obesity epidemic: natural selection of thrifty genes or genetic drift following predation release? Int. J. Obes. (Lond). 32, 1607–1610 (2008).
Diamond, J. M. Guns, Germs, and Steel: the Fates of Human Societies (W. W. Norton, 1997).
West, D. B., Boozer, C. N., Moody, D. L. & Atkinson, R. L. Dietary obesity in nine inbred mouse strains. Am. J. Physiol. 262, R1025–R1032 (1992).
Clee, S. M. & Attie, A. D. The genetic landscape of type 2 diabetes in mice. Endocr. Rev. 28, 48–83 (2007).
Friedman, J. M. Obesity is genetic. Newsweek (9 September 2009).
Friedman, J. M. Modern science versus the stigma of obesity. Nat. Med. 10, 563–569 (2004).
Park, H. K. & Ahima, R. S. Leptin signaling. F1000Prime Rep. 6, 73 (2014).
Diamond, J. M. Diabetes running wild. Nature 357, 362–363 (1992).
Mayer-Davis, E. J. et al. Incidence trends of type 1 and type 2 diabetes among youths, 2002-2012. N. Engl. J. Med. 376, 1419–1429 (2017).
Hawkes, K. Human longevity: the grandmother effect. Nature 428, 128–129 (2004).
We thank the JPB Foundation and the Rockefeller Foundation for supporting this research. The funding sources were not involved in the research or manuscript preparation. We would like to thank D. Wan for creating figures and I. Piscitello for assistance in preparing this manuscript.
Per institutional policy, J.M.F. and the other inventors receive a portion of the royalty payments for the sale of leptin.
Peer review information: Primary Handling Editor: Christoph Schmitt
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article