Leptin Resistance and Obesity
Division of Neuroscience, OHSU, 505 NW 185th Avenue, Beaverton, OR 97006. E-mail: email@example.com
The prevalence of obesity, and the human and economic costs of the disease, creates a need for better therapeutics and better understanding of the physiological processes that balance energy intake and energy expenditure. Leptin is the primary signal from energy stores and exerts negative feedback effects on energy intake. In common obesity, leptin loses the ability to inhibit energy intake and increase energy expenditure; this is termed leptin resistance. This review discusses the evidence in support of leptin resistance in mouse models and humans and the possible mechanisms of leptin resistance.
Obesity is defined as an excessive proportion of body fat relative to lean body mass of sufficient magnitude to produce adverse health consequences (1). There has been a dramatic increase in the prevalence of obesity in the world; >60% of American adults are now overweight or obese. Additionally, the prevalence of obesity in children is growing dramatically, predisposing them to a host of chronic diseases (2).
Energy In − Energy Out = Energy Stored
Energy homeostasis is the balance between energy intake, energy expenditure, and energy storage. Maintenance of body weight depends on the balance between energy intake and energy expenditure. Energy intake is food intake; energy expenditure is derived from complex thermogenesis processes that include basal metabolism, adaptive thermogenesis, and physical activity. Adaptive thermogenesis refers to an increase in heat production through futile metabolic cycles in response to environmental or behavioral changes (excess food consumption, change in the composition of diet, modification of ambient temperature, or a variety of pathogenic stimuli) (3). These thermogenic, metabolically futile cycles are facilitated by uncoupling proteins (1, 2, 3, 4), which decouple oxidative phosphorylation from ATP generation by destabilizing mitochondrial proton gradients.
In most adults, body weight is almost constant despite huge variations in daily food intake and energy expenditure (5). Therefore, complex physiological systems equilibrate energy expenditure with energy intake. Energy balance is regulated by peripheral signals (hormones) that are integrated in the brain centers, including the hypothalamus, brainstem, and reward centers, which in turn modulate feeding and energy expenditure (6, 7). Some hormones reflect the long-term nutritional status of the body (including leptin, insulin, and perhaps, adiponectin), whereas other circulating gut hormones act acutely to initiate or terminate a meal (such as ghrelin, peptide YY3–36, pancreatic polypeptide, oxyntomodulin, glucagon-like peptides 1 and 2, and cholecystokinin) and result in appetite stimulation or satiety (7). To date, we can consider leptin to be the most important peripheral signal for the balance of energy homeostasis (8).
Leptin Informs the Brain of Levels of Stored Fat
Leptin is secreted primarily by adipocytes and is present in serum in direct proportion to the amount of adipose tissue (9, 10, 11). The primary role of leptin is to provide the central nervous system with a signal of energy (adipose) stores in the body to enable the brain to make the adjustments necessary to balance energy intake and expenditure (12). One well-accepted role of leptin signaling is to act as a gatekeeper to a range of activities that are not essential for immediate survival, such as reproduction.
Only when leptin concentrations exceed certain thresholds are these non-essential behaviors expressed.
Leptin regulates food intake by binding to central nervous system receptors and modulating the activity of neurons in appetite control centers in the brain (9, 13). In obese leptin-deficient mice (ob/ob), exogenous administration of leptin effectively reduces hyperphagia and obesity (14). Conversely, obese mice that are deficient in the signaling form of the leptin receptor (db/db) do not respond to leptin (14, 15). As will be described later, in “common” obesity leptin does not cause clear effects on energy balance.
Leptin Regulates Energy Intake and Expenditure
In addition to its effects on appetite, leptin also affects energy expenditure in rodents and humans (14, 16). Activation of central leptin receptors increases the activity of the sympathetic nervous system (17, 18) which stimulates energy expenditure in adipose tissue (19, 20, 21).
Leptin Acts on a Diffuse Neuronal Network, Including Neurons in the Arcuate Nucleus
There is considerable debate about the anatomical loci through which leptin affects energy balance, but recent work from many laboratories suggests that there is not a single locus of activity. Rather, leptin acts on a dispersed network of neurons, such that it can act on many anatomical loci that partially regulate energy balance. It is clear that leptin receptors (ObRb)1 are highly expressed in regions of the hypothalamus that mediate energy homeostasis (22, 23). To access receptors in areas distal to circumventricular organs, peripheral leptin is transported across the blood–brain barrier (24).
The arcuate nucleus (ARH) is a major site of leptin sensing, and the ARH transduces peripheral signals into neuronal responses (25, 26, 27). The ARH contains at least two key populations of neurons that have opposite actions on food intake. One population expresses anorexigenic (appetite-suppressing) peptides, cocaine- and amphetamine-regulated transcript and α-melanocyte-stimulating hormone [α-MSH; derived from the proopiomelanocortin (POMC) precursor]. The other population expresses the orexigenic (appetite-stimulating) peptides, neuropeptide Y (NPY) and agouti-related peptide (AgRP) (28). Neurons in the ARH subsequently innervate various second-order hypothalamic targets that express melanocortin-4 (MC4R) and NPY receptors (29). Leptin can modulate the activity of both POMC and AgRP neurons. Leptin reduces the expression of NPY/AgRP mRNA (22, 30) and can rapidly inhibit the activity of AgRP/NPY neurons (31) (E.E. Jobst and M.A. Cowley, unpublished data). In contrast, leptin activates POMC neurons, as shown by increased expression of POMC mRNA (32), which might promote the release of the α-MSH, a potent anorexigen at central MC4R (33). This is also shown by electrophysiological recordings that show leptin depolarizing (i.e., activating) POMC neurons (26, 34). Leptin also modulates the activity of neurons in the ventromedial hypothalamus (VMH) that express the transcription steroidogenic factor-1 (SF-1) (35). Deletion of ObRb on POMC or on VMH SF-1 neurons produces comparable obesity (27, 35), showing that multiple neuronal systems can participate in leptin sensing and the control of body weight.
Leptin May Modulate Endocannabinoid Signaling
Endocannabinoids are post-synaptic regulators of presynaptic activity; they act retrogradely to inhibit synaptic activity. The significance of retrograde signaling used by cannabinoids for modulation of energy balance is yet to be explored extensively. There is a possible link between the endocannabinoid system and leptin in the modulation of food intake. In young ob/ob mice lacking leptin, hypothalamic 2-AG (an endogenous endocannabinoid) levels are elevated, and these levels are normalized by peripheral intravenous leptin. Thus, hypothalamic endocannabinoids seem to be under negative control by leptin (36), and leptin modulates the activity of lateral hypothalamic nucleus neurons in an endocannabinoid-dependent manner (37).
Rimonabant is a cannabinoid antagonist that is being developed for obesity. NPY is probably not involved in cannabinoid effects on energy balance (36) because the appetite-suppressing effects of rimonabant are preserved in NPY-deficient mice. There is possibly an interaction between the cannabinoid and melanocortin systems, since α-MSH does not block feeding-induced by a cannabinoid agonist, but rimonabant reduces feeding induced by a MC4R antagonist. This suggests that cannabinoid CB1 receptor blockade of feeding occurs downstream of MC4R (38).
Cannabinoids might act through reward pathways to stimulate feeding. Endocannabinoids elevate extracellular levels of dopamine in the shell of the nucleus accumbens (39), and rimonabant reduces food intake induced by injection of a cannabinoid agonist into the nucleus accumbens shell (40). Rimonabant may also act on the mesolimbic dopamine “reward” circuitry (41) that provides dopamine input to the nucleus accumbens. Cannabinoid CB1 receptors located on afferent pathways to the ventral tegmental area have been proposed to contribute to mediating activity of mesolimbic dopamine neurons and thereby dopamine-induced reward behavior (42, 43).
Most obese humans and rodents do not have low circulating leptin. In contrast, they usually have very high plasma leptin concentrations. However, this endogenous hyperleptinemia may not reduce appetite or increase energy expenditure. There are only rare examples of single-gene mutations that are responsible for obesity in humans, and the majority of “common” obesity is thought to be a consequence of polygenic interaction with the environment (44). This state has been termed “leptin resistance.” Several mechanisms may contribute to leptin resistance. The two hypotheses that have received the most attention are that circulating leptin fails to reach its targets in the brain (45) or that there is a failure of components of the intracellular ObRb signaling cascade (46).
Some strains of mice fed a high-fat diet (HFD) exhibit increased body adiposity and many other characteristics of human obesity, and they show leptin resistance in some paradigms. Diet-induced obesity (DIO) has proved to be an excellent system to explore how leptin functions and identify how leptin signaling becomes compromised in obesity.
The development of obesity and leptin resistance in C57BL/6J mice on HFD can be divided into three stages. In the early stage, mice on HFD gain weight (fat tissue) but maintain an adequate response to the anorectic effect of peripheral leptin injection. In the middle stage, mice on HFD show peripheral leptin insensitivity, expressed by changes in food intake and body weight or by lack of activation of signal transducer and activator of transcription (STAT)-3, however these mice retain the capacity to respond to central leptin injection. Finally, there is a late state in which mice develop central leptin resistance (47, 48, 49) and do not show changes in food intake and body weight or activation of STAT-3 in response to icv leptin.
Recently, Munzberg et al. (50) showed that the expression of p-STAT in cells of the ARH was selectively reduced in leptin-treated DIO mice but not in the VMH or dorsomedial hypothalamus, suggesting that the ARH is a major site of leptin resistance. Increased suppressor of cytokine signaling (SOCS)-3 levels, a known inhibitor of leptin signaling, may cause defective leptin signaling in the ARH.
Decreased SOCS-3 seems to cause increased leptin sensitivity and resistance to DIO (46, 51).
Leptin Resistance in Humans?
Recent work from Rosenbaum et al. (16) suggested that leptin resistance may not be critical in humans. This work shows that in dramatically (10%) weight-reduced patients (caused by ingesting just 800 kcal/d over 6 to 10 weeks), infusion of low doses of leptin that replicate the patients’ prereduction levels normalizes appetite and energy expenditure. Thus, there is little evidence for leptin resistance in this population. Conversely, this interpretation is difficult to reconcile with elevated leptin levels in the obese.
We can speculate that “physiological” leptin signaling attempts to limit body weight, even in obese states, but that other factors are opposing this. In modern society, we might suggest that hedonic cues to overconsume are one such factor. The effects of leptin are more obvious in a controlled dietary paradigm, because the background influence of variable diet is absent. This interpretation suggests the question: “Is obesity caused by increased hedonic and cortical cues to over-eat, which overwhelm the normal counter-regulatory responses to increased obesity?” As an additional complication, we might hypothesize that in some situations leptin signaling can cause desensitization, perhaps when levels are elevated for a significant period of time, and that only after leptin signaling is reduced for a period (such as after a long hypocaloric diet) can the system resensitize. A pertinent clinical example of this phenomenon is the use of gonadotropin-releasing hormone agonists to down-regulate the reproductive axis and androgen levels in men undergoing treatment for prostatic tumors (52).
Whether leptin resistance occurs in humans remains open to debate; however, there is very good evidence that it does occur in animal model systems. If we can conclusively identify the specific neuronal pathways that become leptin resistant in obese states, engaging these circuits may represent a therapeutic strategy to bypass insensitivity to endogenous leptin. Such a strategy would produce the same coordinated actions on energy intake and energy expenditure that are normally regulated by leptin which could help obese patients decrease adiposity or comply with a low-calorie diet.
We thank Sonja K. Billes for helpful review of this manuscript. This study was supported by NIH Grants RR0163 and DK 62202.
Nonstandard abbreviations: ObRb, leptin receptor; ARH, arcuate nucleus; αMSH, α-melanocyte-stimulating hormone; POMC, proopiomelanocortin; NPY, neuropeptide Y; AgRP, agouti-related peptide; MC4R, melanocortin 4 receptor; VMH, ventromedial hypothalamus; SF-1, steroidogenic factor-1; HFD, high-fat diet; DIO, diet-induced obesity; STAT, signal transducer and activator of transcription; SOCS, suppressor of cytokine signaling.