Synaptic Plasticity in Energy Balance Regulation


Section of Comparative Medicine, Yale University School of Medicine, 375 Congress Ave. LSOG 117, PO Box 208016, New Haven, CT 06510-8016. E-mail:


Leptin regulates energy balance, in part, by modulating the activity of neuropeptide Y (NPY) and proopiomelanocortin (POMC) neurons in the hypothalamic arcuate nucleus. Leptin-deficient (ob/ob) mice differ from wild-type mice in the number of excitatory and inhibitory post-synaptic densities and currents onto NPY and POMC neurons. When leptin was delivered to ob/ob mice, the synaptic density rapidly normalized, an effect detectable within 6 hours, several hours before leptin's effect on food intake. Synaptic currents were also shifted toward wild-type values in leptin-replaced ob/ob mice. These data suggest that leptin-mediated plasticity in the ob/ob hypothalamus may underlie some of the hormone's behavioral effects. In an effort to determine whether the observed synaptic plasticity is leptin specific, we analyzed the effects of an orexigenic hormone, ghrelin, and anorexigenic hormone, estradiol. Ghrelin rearranged synapses in wild type animals to support suppressed POMC tone, whereas the estradiol triggered a robust increase in the number of excitatory, glutamate inputs of POMC neurons. The rearrangement of synapses by estradiol was leptin independent, because it was also evident in leptin- (ob/ob) and leptin receptor–deficient (db/db) mice and was paralleled with decreased food intake and increased energy expenditure in these mutant, obese animals. Such plasticity was also observed in other hypothalamic regions and extrahypothalamic sites. These observations raise the notion that synaptic plasticity is a major way through which peripheral metabolic hormones influence brain functions.


Increasingly sophisticated methods have been brought to bear on the problem of brain involvement in the physiology of energy homeostasis and the pathogenesis of obesity. The combination of novel genetic with sophisticated physiological techniques has allowed for great progress in the identification of metabolic hormones and their relationship to key peptidergic systems in the hypothalamus. While the central integration of afferent signals reflecting acute and chronic energy requirements becomes clearer (Figure 1), the neuronal pathways that actually initiate changes in ingestive behavior or energy expenditure are still largely unknown, as is our understanding of the fine signaling modality of central body weight regulation. Our recent discovery that there is a rapid synaptic remodeling involving hypothalamic peptidergic systems may shed new light on the mechanism of the central regulation of metabolism and offer explanations as to why, despite the breadth of knowledge gained in the past decades, no cure has been developed for metabolic disorders.

Figure 1.

(A) Schematic overview of central nervous system circuitry, neuropeptides, and neurotransmitters that are involved in the control of appetite and body fat. It should be noted that this schematic overview is an extremely simplified version. (B) Schematic overview of the distribution of receptors and sensors in the hypothalamus and brainstem that are involved in the control of appetite and body fat. It should be noted that this schematic overview is an extremely simplified version. CB1R, cannabinoid receptor-1; CCKAR, cholecystokinin-A receptors; D1R, D2R, D3R, D4R, and D5R, dopamine receptors 1–5; EstrogenR, estrogen receptor; GLP-1R, glucagon-like peptide-1 receptors; GLP-2R, glucagon-like peptide-2 receptors; Glucocort. R, glucocorticoid receptors; Glucose-sensitive, glucose-sensing neurons; GHR, growth hormone receptor; GhsR, ghrelin receptor; InsulinR, insulin receptors; LeptinR, leptin receptors; MC3R, melanocortin receptor 3; MC4R, melanocortin receptor 4; Orexin1R, orexin/hypocretin receptor-1; Orexin2R, orexin/hypocretin receptor-2; MCHR, melanin-concentrating hormone; Y1R, Y2R, Y4R, and Y5R, NPY/PYY/PP receptors 1, 2, 4, and 5.

Information Processing Is through Neurons in the Brain

The signal flow within the brain that underlies the regulation of metabolism is a highly complex process and is based on neuronal interactions. Neurons interact with each other by synapses that are established between axon terminals and dendritic or perikaryal membranes. The information moves from axon terminal to the dendritic or perikaryal membranes. The information transmission in synapses occurs either electrically, chemically through the release of neurotransmitters or modulators from synaptic vesicles of the axon terminal (neurotransmitters and neuromodulators), or by the release of gaseous substances such as NO or carbon monoxide. In most cases, neurons are capable of signaling by all three mechanisms, whereas one mode of transmission might dominate depending on the action potential and on substances that may directly signal to the axon terminal from the extracellular space (i.e., other neuromodulators released by nearby axons or available by volume transmission, extracellular anion and cation concentrations, or substances released to the extracellular space by paracrine or endocrine processes). The specificity of signal transduction is ensured by the appropriate connectivity within a given network and by the availability of receptors at the right sites for released neurotransmitters or neuromodulators and for peripheral metabolic hormones. It is reasonable to suggest that neurons and their interplay with each other hold a key to the understanding of metabolism regulation. Thus, the establishment of the connectivity and hierarchical relationship between hypothalamic neuromodulator systems is critical to understanding the blueprint for the hypothalamic regulation of daily energy homeostasis. Much progress has been made in the past decade to accomplish this goal. The schematic illustration in Figure 1 summarizes some of the critical findings of several laboratories, including our own.

We have been studying the relationship between key peptidergic circuits and their inputs in the hypothalamus. We have also examined whether alterations in the wiring of hypothalamic metabolic circuits occur in response to the changing metabolic state. Strikingly, our studies conducted in the non-human primate hypothalamus revealed a robust and rapid rearrangement of synaptic inputs of orexigenic circuits and their respective interneuronal controllers in response to short-term fasting (unpublished observations). Fasting resulted in a balance of stimulatory inhibitory synapses on orexin and neuropeptide Y (NPY)1 neurons that favored increased activity of these neurons. These cells have been implicated as key orexigenic neurons in the hypothalamus (1, 2). On the other hand, putative inhibitory interneurons of the same regions (neurons that would inhibit either orexin or NPY neuronal activity) exhibited a synaptic balance during fasting that support neuronal inactivation, thereby further enhancing the activity level of orexin and NPY perikarya. These observations raised the intriguing possibility that metabolic signals, leptin in particular, may have an acute effect on synaptic plasticity within the appetite center. The fact that the hypothalamus is not hard-wired, i.e., it goes through continuous synaptic reorganization, is not novel. Rapid rearrangement of synapses have been shown to occur in the magnocellular system during changes in water homeostasis (3, 4, 5), the arcuate nucleus interneuronal system during changes in the gonadal steroid milieu (6, 7), and on the perikarya of luteinizing hormone-releasing hormone neurons during changes in the gonadal steroid milieu (8) or during changes in photoperiod lengths (9). However, such synaptic plasticity has never been considered as a critical component in the regulation of daily energy homeostasis. Our observations now suggest that synaptic plasticity is a key component in the physiological regulation of energy homeostasis and that, under pathological conditions, the synaptic constellation and its plasticity is impaired.

Leptin Is a Key Metabolic Signal Associated with the Rapid Rewiring of Hypothalamic Pathways

The rapid rearrangement of synapses during fasting in monkeys coincided with diminishing circulating levels of leptin (10). This, together with the fact that both NPY and orexin neurons express leptin receptors in the non-human primate (11), suggested that leptin may be an important contributor to the observed synaptic plasticity during changing metabolic states. We carried out a series of studies to directly test that proposition. To test our hypothesis, we turned our attention to ob/ob mice and their wild-type littermates. These mice lack the leptin gene, and, thus, the phenotype of these animals resembles morbid human obesity (12, 13, 14, 15). Replacement of leptin to ob/ob animals rapidly decreases food intake and triggers weight loss (12, 13, 14, 15). Thus, ob/ob mice and their wild-type littermates presented a great model in which we could determine whether the presence of leptin predicts a different wiring pattern of hypothalamic peptidergic circuits. We aimed to analyze two distinct populations of neurons in the arcuate nucleus: one that produces NPY/agouti-related peptide (AgRP) and the other that expresses proopiomelanocortin (POMC). Arcuate nucleus neurons that coproduce NPY and AgRP (15) are key orexigenic cells that interact with those local cells that express the POMC-derived peptide, α-melanocyte stimulating hormone (α-MSH), the most potent anorexigenic peptide, and are considered a primum movens of metabolism regulation by the brain (16, 17). The leptin receptor, LRb, is coexpressed with both neuronal subtypes (18, 19, 20). Increased NPY/AgRP activity and suppressed POMC tone is thought to underlie feeding and fat deposition. In contrast, increased POMC tone and suppressed NPY/AgRP activity support decreased feeding and lean body mass (21, 22). In line with this, the ob/ob mouse expresses increased NPY and decreased POMC (23, 24). Thus, the perikarya of these neurons in ob/ob mice and their wild-type littermates represented a unique model in which we could test our hypothesis regarding synaptic plasticity and the effect of leptin on it. While our studies focused on the arcuate nucleus melanocortin system, it is important to emphasize that energy metabolism regulation from the brain is organized from various sites, which is not limited to the hypothalamus, but also includes other regions, most notably the brain stem (25, 26).

Leptin-deficient ob/ob Animals Have Altered Synaptology and Electrophysiological Properties in the Arcuate Nucleus

We analyzed transgenic animals generated by the laboratory of Dr. Jeffrey Friedman at Rockefeller University, in which τ-sapphire green fluorescent protein (GFP) is expressed under the transcriptional control of the NPY genomic sequence or τ-topaz GFP is expressed under the transcriptional control of POMC genomic sequence (27). We first examined the afferent inputs to POMC and NPY neurons in the arcuate nucleus of ob/ob and wild-type animals using patch-clamp electrophysiology recording in slice preparations. NPY-GFP or POMC-GFP cells were held in the whole cell voltage-clamp configuration, and the number of excitatory and inhibitory post-synaptic currents (EPSCs/IPSCs) was determined. In wild-type animals, NPY neurons had similar levels of spontaneous EPSCs and IPSCs. The inputs onto POMC neurons were principally IPSCs. We next compared the afferent inputs onto NPY and POMC neurons in the ob/ob mouse relative to those seen in the wild-type mouse. The electrophysiological recordings revealed a large shift in the ratio of EPSCs to IPSCs onto both the NPY and POMC neurons in the ob/ob slices. NPY neurons from ob/ob mice showed a significant increase in the frequency of EPSCs combined with a significant decrease in the frequency of IPSCs. There was a similarly large shift in the ratio of inputs onto POMC neurons from ob/ob mice, with a robust increase in the frequency of IPSCs. While there were no alterations in the frequency of EPSCs onto the POMC neurons, there was a marked net increase in inhibitory tone onto these neurons. Thus, in hypothalamic slices from ob/ob mice, there were reciprocal effects on NPY and POMC neurons, with a marked increase in excitatory tone onto the NPY neurons and a net decrease in excitatory tone onto the POMC neurons. To assess whether the number of synaptic inputs to the NPY and POMC neurons correspond to the above electrophysiological properties, we used stereology to quantify the synaptic density on NPY and POMC perikarya from ob/ob and wild-type animals. In ob/ob animals, there was significantly higher total number of synapses on the perikarya of NPY neurons compared with wild-type littermates. This value was fully accounted for by an increase in the number of excitatory synapses onto these cells that was more numerous than inhibitory ones in ob/ob mice. In contrast, in wild-type animals, inhibitory synapses onto the NPY neurons were more numerous than excitatory ones. This altered synaptic profile of NPY cells in the ob/ob animals is entirely consistent with their electrophysiological profile. In the POMC cells of wild-type mice, excitatory synapses dominated over inhibitory contacts, whereas in ob/ob cells, the opposite was observed. In addition, the total number of synapses on the POMC neurons was lower in ob/ob mice compared with wild-type littermates. Here too, the excitatory/inhibitory synaptic balance is in agreement with the increased inhibitory tone seen on these neurons in slices from ob/ob hypothalami described above.

Leptin Induces Rapid Rewiring of Arcuate Nucleus Feeding Circuits in ob/ob Mice

A single dose of leptin reduces food intake in ob/ob mice within 12 hours (28). Groups of ob/ob animals were treated with leptin or phosphate-buffered solution for 6 hours, 48 hours, or 12 days. After 6 hours of leptin treatment, there was a significant decrease in the number of excitatory inputs and a significant increase in the number of inhibitory inputs onto the NPY neurons of ob/ob mice. In contrast, there was a significant increase in the number of excitatory inputs onto the POMC neurons of ob/ob mice. The statistical strength in these changes increased after 48 hours and 12 days. At 48 hours after leptin replacement, we also found that the electrophysiological properties of ob/ob NPY/AgRP and POMC neurons shifted toward wild-type values (27). Both the electrophysiological and anatomical changes were positively correlated with changing food intake and body weight gain of the same animals. These data show that leptin has potent and rapid effects on the wiring of key neurons in the hypothalamus. The fact that these synaptic changes that occur before changes in feeding behavior and body weight can be detected statistically (albeit these parameters are changing) suggests the possibility that the rapid leptin-induced rewiring of the synaptic inputs to the NPY and POMC cells in ob/ob mice may be a prerequisite for some portion of its behavioral effects.

Metabolic State–associated Rewiring of Circuits Is Not Limited to Leptin Levels and to the Arcuate Nucleus

Our results on leptin replacement in ob/ob mice raised (27) the question of whether the observed synaptic rearrangements of feeding circuits is part of more general phenomena. Thus, we next assessed the effect of peripheral ghrelin injections on the synaptology of the arcuate nucleus NPY and POMC neurons of mice and found a shift in the synaptic profile of POMC neurons by ghrelin that was the opposite of that induced by leptin, and it is in line with ghrelin's orexigenic action (27) (Figure 2).

Figure 2.

Perikaryal inputs of NPY/AGRP neurons are dominated by inhibitory connections when circulating leptin levels are high. These connections are rearranged (arrows indicate changes and dotted outlines indicate new locations) when leptin levels diminish (and ghrelin levels increase). Under these circumstances, stimulatory synapses (striped) dominate over inhibitory inputs. On POMC perikarya, the changes occur in the opposite direction of that described for the NPY/AgRP inputs. Some of the inhibitory inputs on POMC cells are likely to originate from the NPY/AgRP neurons (dotted axon), and some of the stimulatory inputs on both cell types originate in the lateral hypothalamic hypocretin neurons (dotted axon). Because rapid synaptic changes were observed in wild-type animals, it is reasonable to propose that synaptic rearrangement of feeding circuits is a continuous phenomenon, which will also occur in a circadian fashion responding to the changing daily metabolic environment. In fact, this proposition was confirmed in new pilot studies (39).

The lateral hypothalamic orexin neurons have emerged as instrumental in triggering arousal, but they are also implicated in food intake regulation. In a recent study (29), we determined the synaptic input organization and stability of orexin perikarya using quantitative and qualitative synaptology and targeted patch-clamp electrophysiology. We found an unorthodox synaptic organization on the orexin neurons in which excitatory synaptic currents forming asymmetric synapses exert control on the perikarya of these long-projective neurons with minimal inhibitory input. Overnight food deprivation, which is accompanied by elevated ghrelin and suppressed leptin levels in circulation, promoted the formation of more excitatory synapses and synaptic currents onto hypocretin perikarya. This event could be blocked by exogenous leptin administration during the fast, and it could be reversed by refeeding. This unique wiring and fasting-induced plasticity of the hypocretin neurons correlates well with its being involved in the control of arousal and alertness that are so vital to survival, but it may also be an underlying cause of insomnia. This study also revealed that synaptic changes induced by metabolic alterations are not limited to the arcuate nucleus.

Ghrelin affects energy expenditure by a hypothalamic mode of action that involves induction of synaptic plasticity in the melanocortin system. However, ghrelin binding is also present in other brain areas, including the telencephalon, where its function remained elusive. In a recent study (30), we found that ghrelin binds to neurons of the hippocampal formation, up-regulates dendritic spine density of CA1 pyramidal neurons, and promotes long-term potentiation in parallel with enhancement of spatial learning and memory performance. These observations revealed a novel function of ghrelin connecting metabolic regulation with that of higher brain functions, and these results promote the idea that synaptic plasticity is an important component of the brain action of metabolic hormone, be it their effect in hypothalamic or extrahypothalamic regions.

What Is the Mechanism of Action of Metabolic Hormones in Triggering Synaptic Plasticity?

The molecular basis for leptin- or ghrelin-mediated plasticity in the brain remains elusive. We predict that activation of the long-form LR for leptin and growth hormone secretagogue receptor 1 for ghrelin are important steps in triggering synaptic plasticity by these hormones. Regarding leptin, its action through LR activates signal transducer and activator of transcription 3 (STAT3) (31, 32, 33) and, to lesser extent, the PI3 kinase (34, 35) signaling pathways (Figure 3). It may be that by rapid activation of STAT3, leptin promotes excitatory synapse formation on POMC perykarya, whereas activation of PI3 kinase in response to sustained leptin and/or insulin exposure during a high-fat diet will promote inhibitory synapse formation and transmission (Figure 3). These actions of leptin, and those of ghrelin through growth hormone secretagogue receptor 1, may occur at the level of the post-synaptic (Figure 3) and/or on pre-synaptic sites. Further studies are warranted to determine the cellular underpinning of synaptic plasticity induced by peripheral hormones.

Figure 3.

We suggest that activation of the long-form leptin receptor (LR) for leptin is important steps in triggering synaptic plasticity by these hormones. Leptin's action through LR activates STAT3, and to a lesser extent, the PI3 kinase signaling pathways. It may be that, by rapid activation of Stat3, leptin promotes excitatory (+) synapse formation on POMC perykarya, whereas activation of PI3 kinase in response to sustained leptin and/or insulin exposure during a high-fat diet will promote inhibitory (−) synapse formation and transmission. These actions of leptin may occur at the level of the post-synaptic (as depicted here) and/or on pre-synaptic sites.


Chronically increased energy intake without a respective increase in energy expenditure leads to obesity and diabetes, as well as a variety of life-threatening consequences of such diseases, such as cancer and cardiovascular diseases (36, 37). While it seems intuitively obvious that, in the majority of cases, positive energy balance should be corrected by changes in lifestyle and/or diet, the impressive dynamics of the spreading obesity epidemic (38) certainly suggests that, in modern industrialized civilizations, an efficient and safe pharmacological approach to treat obesity would be useful. Based on emerging data, we believe that synaptic plasticity–induced by peripheral hormones and its impairment during pathological conditions are fundamental in the regulation of energy homeostasis.


  • 1

    Nonstandard abbreviations: NPY, neuropeptide Y; AgRP, agouti-related peptide; POMC, proopiomelanocortin; αMSH, α-melanocyte stimulating hormone; LR, leptin receptor; GFP, green fluorescent protein; EPSC, excitatory post-synaptic current; IPSC, inhibitory post-synaptic current; STAT3, signal transducer and activator of transcription 3.