Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Nat Neurosci. Author manuscript; available in PMC 2009 Jul 14.
Published in final edited form as:
PMCID: PMC2710100

Brain-derived neurotrophic factor regulates energy balance downstream of melanocortin-4 receptor


The melanocortin-4 receptor (MC4R) is critically involved in regulating energy balance, and obesity has been observed in mice with mutations in the gene for brain-derived neurotrophic factor (BDNF). Here we report that BDNF is expressed at high levels in the ventromedial hypothalamus (VMH) where its expression is regulated by nutritional state and by MC4R signaling. In addition, similar to MC4R mutants, mouse mutants that expresses the BDNF receptor TrkB at a quarter of the normal amount showed hyperphagia and excessive weight gain on higher-fat diets. Furthermore, BDNF infusion into the brain suppressed the hyperphagia and excessive weight gain observed on higher-fat diets in mice with deficient MC4R signaling. These results show that MC4R signaling controls BDNF expression in the VMH and support the hypothesis that BDNF is an important effector through which MC4R signaling controls energy balance.

Neurotrophins are a family of structurally related growth factors, including brain-derived neurotrophic factor (BDNF), that exert many of their effects on neurons through Trk receptor tyrosine kinases. Among these, BDNF and its receptor TrkB are the most widely and abundantly expressed in the brain. BDNF has been shown to regulate neuronal development and to modulate synaptic plasticity1. Obesity phenotypes have been observed in BDNF heterozygous mice and in mice in which the BDNF gene has been deleted in excitatory neurons in the brain24. These mutants also show hyperactivity, hyperleptinaemia, hyperinsulinaemia, hyperglycemia and increased linear growth3,4. Moreover, both central and peripheral administration of BDNF decrease food intake, increase energy expenditure and ameliorate hyperinsulinaemia and hyperglycemia in diabetic db/db mice58. However, the means by which BDNF alters energy balance and the relationship of TrkB signaling to the major pathways previously shown to be involved in the regulation of energy balance remain unknown.

The status of energy balance is communicated to the hypothalamus through neuronal and hormonal signals. Several hypothalamic nuclei are involved in the regulation of energy balance, including the ventromedial (VMH), paraventricular (PVN) and arcuate (ARC) hypothalamic nuclei and the lateral hypothalamic area (LH). Although the observation that bilateral VMH lesions produce hyperphagia and obesity has implicated this region in the regulation of energy balance911, the neural mechanisms through which the VMH functions to influence energy balance are not clear.

One of the major signals that serves as a monitor of energy balance is leptin, a polypeptide generated in adipocytes12. One of the main targets of leptin in the hypothalamus is the ARC, a region containing at least two distinct populations of neurons1315. One population of neurons expresses two orexigenic polypeptides, neuropeptide Y (NPY) and agouti-related protein (AgRP), whereas the other expresses the anorexigenic polypeptides proopiome-lanocortin (POMC) and cocaine– and amphetamine–regulated transcript (CART). Leptin suppresses food intake and increases energy expenditure at least partially by positively regulating the subset of neurons expressing POMC and CART and by negatively regulating the subset of those expressing NPY and AgRP. Other metabolic signals apart from leptin regulate the POMC-expressing neurons in the ARC; these include glucose, insulin, the gut peptide PYY3–36 and serotonin1620. These melanocortin-expressing neurons thus seem to serve as one of the conduits for signals relevant to energy balance21.

Two melanocortin receptors, MC4R and melanocortin–3 receptor (MC3R), are expressed in the brain and play a critical role in the control of energy balance, especially the MC4R. In the agouti lethal yellow mutant (Ay), ectopic expression of the agouti protein (a homologue of AgRP) causes hyperphagia and obesity22,23. The phenotype of the MC4R−/− null mutant is almost identical to that of the Ay mutant, and consistent with this, the agouti protein has been shown to be an antagonist of MC4R, but not MC3R2426. Hyperphagia resulting from abnormal feeding response to dietary fat probably contributes significantly to obesity in the MC4R−/− mice27.

Here we report that MC4R- and TrkB-mediated signaling are coupled in the regulation of energy balance. Our data show that signaling through the MC4R controls BDNF expression in the VMH, supporting a model in which BDNF functions as an important effector through which MC4R signaling regulates energy balance.


BDNF in the VMH is an anorexigenic factor

Expression of BDNF was examined in the hypothalamus by using a BDNFlacZ mouse strain in which the reporter gene for β-galactosidase, lacZ, is targeted to the BDNF locus28. BDNF is modestly expressed in some cells of the dorsomedial hypothalamus (DMH), but not in the ARC (Fig. 1a). Only a small number of cells express BDNF in the LH. Consistent with in situ data3, BDNF is highly expressed in the VMH. Since bilateral VMH lesions produce hyperphagia and obesity911, this expression pattern suggests that BDNF in the VMH may suppress food intake and therefore may be regulated by feeding status. To test this, we examined BDNF expression in food-deprived wild-type (WT) C57BL/6 mice with in situ hybridization. Food deprivation has been shown to increase expression of orexigenic neuropeptides while decreasing expression of anorexigenic neuropeptides13,15,29. Food deprivation specifically reduced the level of BDNF mRNA by 60% in the VMH (Fig. 1b–d). In contrast, expression was not affected in the neocortex (Fig. 1d).

Figure 1
Energy status regulates levels of BDNF in the VMH. (a) Expression of BDNF in the hypothalamus was examined by using an adult BDNFlacZ/+ mouse. Brain sections were stained for β-galactosidase with X-gal staining and counterstained with nuclear ...

Coexpression of TrkB and neuropeptides in the hypothalamus

BDNF might regulate energy balance through direct regulation of neuropeptides known to control energy balance. If so, the TrkB receptor should be coexpressed with these neuropeptides. We examined TrkB expression in the hypothalamus by using adult trkBlacZ/+ mice in which a taulacZ reporter gene is inserted into a trkB locus30, recapitulating the expression pattern of TrkB. Since the ARC and LH are two important brain regions in the control of feeding, we examined coexpression of TrkB and neuropeptides in these two nuclei. There was little coexpression of TrkB with melanin-concentrating hormone (MCH) or orexin in the LH and almost no expression of TrkB in neurons that express CART or NPY in the ARC (Fig. 2a–d). Therefore, it is unlikely that BDNF directly modulates CART/POMC- and NPY/AgRP-expressing neurons in the ARC or MCH- and orexin-expressing neurons in the LH.

Figure 2
Coexpression of TrkB and neuropeptides in the hypothalamus. Expression of TrkB was examined by using an adult trkBlacZ/+ mouse. Most TrkB-expressing neurons do not express MCH (a) or orexins (b) in the lateral hypothalamus and CART (c) or NPY (d) in the ...

Reduction in TrkB expression leads to obesity

We previously described the generation—by insertion of a trkB cDNA into the first coding exon of the trkB locus—of a trkB hypomorph, fBZ/fBZ, that only expresses the full-length TrkB receptor tyrosine kinase at approximately a quarter of the normal amount in a normal pattern throughout the brain30,31. The trkB hypomorph becomes severely obese (Fig. 3a). The fBZ allele was backcrossed to the C57BL/6 genetic background for at least four generations before it was used to generate fBZ homozygous mice for characterization of the obesity syndrome. At 5 weeks of age, the trkB mutant started to show higher body weight compared to its wild-type littermate (Fig. 3c,d). By 12 weeks, both female and male trkB mutants were markedly heavier than their wild-type littermates. The body weight of fBZ heterozygous mice was similar to that of the wild-type littermates, indicating that fBZ is a recessive mutation. To determine whether weight gain in the trkB mutant is due to increased adiposity, we dissected and weighed four different fat pads from three female fBZ/fBZ mice and three female wild-type litter-mates at 12 weeks of age. The mass of each was greatly increased in the trkB hypomorph (brown fat, 0.92 ± 0.10 g in the mutant vs. 0.12 ± 0.06 g in the WT control, P < 0.01; subcutaneous fat, 5.08 ± 1.43 g vs. 0.52 ± 0.16 g, P < 0.01; uterine fat, 5.38 ± 2.63 g vs. 0.31 ± 0.12 g, P < 0.05; perirenal fat, 1.14 ± 0.11 g vs. 0.17 ± 0.02 g, P < 0.01).

Figure 3
The trkB hypomorphic mutant shows maturity-onset obesity and increased linear growth. (a) The fBZ/fBZ mutant is obese. The two mice are female littermates at 3 months of age. (b) Both female and male trkB mutant mice show increased linear growth. Measurements ...

The trkB mutant showed altered linear growth when body length (naso–anal) was measured at approximately 3 months of age. The mean lengths of male and female fBZ/fBZ mice were increased approximately 8% and 13%, respectively, compared to controls (Fig. 3b).

Feeding behavior of the trkB hypomorph

The development of obesity in the trkB mutants indicates that these animals have elevated energy intake relative to energy expenditure. Phenotypic effects on energy consumption were first examined by determining the daily food intake of the trkB mutants at 5 and 9 weeks of age. The mutants showed a marked hyperphagia at both ages, consuming 75% more food than controls (Fig. 4a). This likely contributes substantially to their development of obesity. To determine the impact of the trkB mutation on patterns of ingestive behavior, mice were habituated to single housing for 6 d in cages equipped to continuously monitor feeding and drinking activity. Mice were then monitored for an additional 8 d, and data from these days were averaged for examination of intake patterns. The intake patterns are shown for a mutant and wildtype littermate pair (Fig. 4b,c). Similar to the patterns observed for this pair, mutants frequently engaged in prolonged periods of feeding and drinking in comparison to wild-type controls. These episodes were most frequent during the dark cycle but also occurred during the light cycle. In addition, prolonged periods of quiescence were commonly observed during the light cycle. Accompanying alterations in the circadian distributions of feeding and drinking were also seen, such that the percentages of total food and water intake during the dark cycle were decreased in the mutant mice (chow: WT 78 ± 2%, fBZ/fBZ 64 ± 3%, P = 0.001; water: WT 84 ± 2%, fBZ/fBZ 66 ± 3%, P < 0.001). Overall, the trkB mutants were found to be significantly hyperphagic and hyperdipsic during both the light and dark cycles (Fig. 4d,e).

Figure 4
Hyperphagia and hyperdipsia of the trkB hypomorphic mutant mice. (a) The food intake of individually housed mice was measured every day over a one-week period. The bars represent mean of five measurements for each of five fBZ/fBZ mice (2 male, 3 female) ...

MC4R regulates BDNF expression in the VMH

These phenotypic analyses indicate that the trkB hypomorph is remarkably similar to the MC4R−/− null mutant in development of obesity. The trkB hypomorph and MC4R−/− mutant show maturity-onset obesity and increased linear growth. In addition, they have similar sex differences in the obesity syndrome. For example, female and male fBZ/fBZ mice, on average, are 107% and 50% heavier than their sex-matched wild-type littermates at 12 weeks of age (Fig. 3c,d). Similarly, female and male MC4R−/− null mice, on average, are 100% and 50% heavier than their sex-matched wild-type littermates at 15 weeks of age24. Moreover, linear growth in these two mutants is increased over wild-type mice to a similar extent (13% in fBZ/fBZ vs. 11% in MC4R−/− in females; 8% in fBZ/fBZ vs. 7% in MC4R−/− in males).

To examine whether MC4R signaling regulates BDNF expression in the VMH, we first determined whether POMC- or AgRP-expressing neurons project to the VMH in mice. We used antibodies to gammamelanocyte-stimulating hormone (γMSH), a POMC-derived peptide, to reveal POMC fibers (Supplementary Fig. 1 online). Consistent with what has been reported in rats32, the number of POMC-immunoreactive fibers in the VMH is fewer than the number in the DMH, but is comparable to the number observed in the LH. As the POMC fibers in the VMH contain boutons, these fibers must innervate VMH neurons. Similarly, VMH neurons also receive innervation from AgRP-immunoreactive fibers (Supplementary Fig. 1 online). In addition, there are substantial numbers of POMC and AgRP fibers surrounding the VMH, which may innervate dendrites extending from neurons within the VMH.

Next we quantified expression of BDNF by using in situ hybridization in the Ay mutant in which ectopically expressed agouti blocks the function of the MC4R22,23,26. Compared to sex-matched wild-type littermates, BDNF expression in the VMH was reduced by 30–40% in the Ay mutant (Fig. 5a). The data suggest that reduction in MC4R signaling results in downregulation of BDNF expression in the VMH. We also used in situ hybridization to quantify BDNF expression in the VMH of MC4R−/− mice, demonstrating that it is also significantly reduced in the VMH (Fig. 5b). To determine whether MC4R signaling acutely regulates the level of BDNF mRNA in the VMH, we used wild-type mice to examine the possibility that administration of a melanocortin receptor agonist reverses the inhibitory effect of food deprivation on BDNF expression in the VMH. Similar to the results shown in Fig. 1, 47 h of fasting resulted in an approximately 65% reduction in the level of BDNF mRNA in the VMH that was not prevented by intracerebroventricular injection of artificial cerebrospinal fluid (ACSF; Fig. 5c). Compared to ACSF administration alone, however, administration of an MC4R agonist, MTII, significantly increased (by ∼70%) the level of BDNF mRNA in the VMH of food-deprived mice within 3 h of treatment (Fig. 5c). Although the level of BDNF mRNA in the VMH of MTII-treated food-deprived mice was still significantly lower than that in the VMH of vehicle-treated fed mice, these results indicate that the melanocortin pathway is at least partially responsible for downregulation of VMH BDNF during food deprivation.

Figure 5
Melanocortins regulate the levels of BDNF mRNAs in the VMH. (a) Expression of BDNF in the VMH is significantly reduced in the Ay mutant. Expression levels of BDNF in the VMH were determined with in situ hybridization on C57BL/6J–Ay/a mice and ...

MC4R regulates BDNF expression in selective VMH neurons

Neurons in hypothalamic nuclei are extremely heterogeneous. To determine whether ectopic expression of agouti reduces expression of BDNF in all BDNF-expressing VMH neurons or selectively abolishes BDNF expression in only some, we crossed the BDNFlacZ/+ mouse to the Ay mutant and used β-galactosidase as a marker to examine the effect of agouti on expression of BDNF. Compared with BDNFlacZ/+ mice, fewer neurons expressed BDNF in the VMH of Ay/a;BDNFlacZ/+ mice (Fig. 6a–d). Furthermore, it appeared that the number of neurons expressing BDNF in the caudal part of the VMH was more severely reduced in Ay mice (Supplementary Fig. 2 online). The effect of agouti on BDNF expression seems specific to the VMH. BDNF expression in the Ay mutant remains normal in the DMH, LH (Fig. 6a,b), PVN (Fig. 6e,f), hippocampus and cerebral cortex (data not shown). Reduction in the number of BDNF–expressing neurons in the VMH is unlikely to be due to cell death because cresyl violet staining shows a normal VMH in the Ay/a;BDNFlacZ/+ mutant (data not shown).

Figure 6
Melanocortins regulate expression of BDNF in selective populations of neurons in the VMH. (a,b) Expression of BDNF in the hypothalamus of BDNFlacZ/+ and Ay/a;BDNFlacZ/+ mice. Expression of BDNF is revealed with anti-β-galactosidase immunohistochemistry ...

To further confirm that melanocortins acutely regulate BDNF expression in the VMH, we injected MTII into the dorsal third ventricle of Ay/a;BDNFlacZ/+ mice. MTII injection increased the number of intensely stained neurons in the area of the VMH where expression of BDNF is reduced in the Ay mice (Fig. 6h,i). Compared to saline-injected brains, the number of neurons expressing β-galactosidase was increased by 39% in the caudal VMH of MTII-treated brains (P = 0.017) (Fig. 6g). Thus, results indicate that an MC4R agonist acutely increases expression of BDNF in a set of VMH neurons, even in the presence of the MC4R antagonist agouti. This is consistent with the previous observation that MTII injection significantly inhibits food intake in Ay mice33.

TrkB regulates feeding response to dietary fat

On the basis of these findings, we reasoned that BDNF may be a downstream component in the MC4R-mediated control of energy balance. MC4R is required for the feeding response to dietary fat27. Transition from a low-fat diet (Purina diet 5001, 12.8% fat calories) to a moderatefat diet (Purina diet 5015, 25.1% fat calories) leads to a dramatic increase in energy intake and rate of weight gain in the MC4R−/− mice, but does not increase the rate of weight gain in wild-type mice. In contrast, leptin-deficient ob/ob mice are similarly hyperphagic on both the lowfat and moderate-fat diets and do not exhibit an increase in the rate of weight gain on the moderate-fat diet27. To examine whether mice with deficient BDNF-mediated signaling exhibit an abnormal response to the moderate-fat diet similar to that of MC4R−/− mice, the weight and food intake of the animals were monitored for 3 d on the low-fat diet and then for 7 d on the moderate-fat diet, followed by another 4 d on the low-fat diet. Female fBZ/fBZ mice gained weight to a similar extent as control mice (WT or fBZ/+ mice) on the low-fat diet, but gained weight at a markedly increased rate on the moderate-fat diet (Fig. 7a). Male fBZ/fBZ mice showed a similar abnormal response in weight gain to the diet transition (data not shown).

Figure 7
TrkB acts downstream of MC4R to regulate feeding response to dietary fat. (a,b) Body weight and energy intake of female fBZ/fBZ (n = 4) and control littermates (+/+ or fBZ/+, n = 4). The average weights of the fBZ/fBZ and control mice at the start of ...

The weight gain in the fBZ/fBZ mice was associated with a sharp increase in energy intake on the moderate-fat diet (Fig. 7b). The trkB mutant mice and their littermate controls consumed a similar amount of the low-fat diet (Fig. 7c; 4.9 ± 0.2 g for controls vs. 5.4 ± 0.4 g for fBZ/fBZ, P = 0.307). Transition from the low-fat to the moderate-fat diet led control mice to significantly reduce food intake (4.9 ± 0.2 g on the low-fat diet vs. 3.9 ± 0.2 g on the moderate-fat diet, P = 0.003), thereby maintaining a similar energy intake level (Fig. 7b,c). In contrast, the fBZ/fBZ mice greatly increased food intake after the diet was switched (Fig. 7b). On average, the fBZ/fBZ mice ingested 71% more calories on the moderate-fat than on the low-fat diet (Fig. 7c). These results indicate that activation of the TrkB receptor is required for the normal feeding response to elevated dietary fat.

We then determined whether BDNF is downstream of MC4R by attempting to suppress with infusion of BDNF into the brain hyperphagia and the increased rate of weight gain observed in Ay mice on the moderate-fat diet. For two weeks, PBS or BDNF (50 ng/h) were infused into the lateral ventricles of female Ay and WT C57BL/6 littermates using osmotic pumps. Transition from the low-fat to the moderate-fat diet led to an increase in energy intake in the PBS-treated Ay mice (Fig. 7d; 11.2 ± 0.4 kcal on the low-fat diet vs. 14.0 ± 0.5 kcal on the moderate-fat diet, P < 0.001). BDNF infusion significantly suppressed hyperphagia in the Ay mice on the moderate-fat diet without altering energy intake on the low-fat diet (Fig. 7d). Transition from the low-fat to the moderate-fat diet led to a faster weight gain in the PBS-treated Ay mice than in the PBS-treated, wild-type mice. BDNF infusion reduced the rate of weight gain in these Ay mice on the moderate-fat diet to a level similar that observed in wild-type mice (Fig. 7e). Over the 9-d period on the moderate-fat diet, BDNF infusion did not significantly reduce weight gain in wild-type mice (P = 0.244), but greatly reduced it in the Ay mice from 15.3% to 7.6% (P = 0.0027) (Fig. 7f). The data support a model in which BDNF at least partially mediates the effect of MC4R on the feeding response to dietary fat.


Although bilateral lesions in the VMH produce hyperphagia and obesity, the mechanism through which VMH neurons control energy balance is not known. Our results show that BDNF is highly expressed in the VMH where its expression is regulated by nutritional state. We are not aware of another molecule in the VMH that is regulated by food deprivation. Furthermore, our results show that MC4R-mediated signaling stimulates expression of BDNF in the VMH and that infusion of BDNF into the brain rescues the abnormal feeding response of Ay mice to dietary fat. These results suggest that BDNF is one of the molecules that mediate the function of the VMH in the control of energy balance.

As BDNF is a key regulator of neuronal development, neurodevelopmental anomalies may contribute to the obesity phenotype in BDNF and trkB mutant mice. Substantial evidence indicates, however, that the defect of energy homeostasis in these mutants does reflect an adult function of BDNF. First, infusion into the adult brain of BDNF, but not NGF or NT-3, reverses hyperphagia and obesity of BDNF+/− mice3. Second, deletion of the BDNF gene in the brain, using a cre transgene under the control of the Ca2+/calmodulindependent kinase II (CaMKII) promoter leads to obesity4. Most developmental processes have been completed before this promoter is first activated during the third week of postnatal development. Third, administration of BDNF increases energy expenditure in db/db mice6. Fourth, food deprivation specifically reduces the level of BDNF mRNA in the VMH (Fig. 1 and Fig. 5). Finally, MC4R signaling regulates BDNF expression in the VMH (Fig. 5 and Fig. 6). Therefore, the obesity phenotypes observed in BDNF and trkB mutants cannot completely result from developmental anomalies.

There are several possible mechanisms through which MC4R signaling could influence BDNF expression in the VMH. One possibility is direct signaling through MC4R-expressing neurons within the VMH. MC4R is expressed within the VMH and the VMH receives innervations from arcuate POMC- or AgRP-expressing neurons32,34. The level of POMC innervation within the VMH is low, but is comparable to that in the LH (Supplementary Fig. 1 online). In addition, there are a substantial number of POMC and AgRP fibers surrounding the VMH. These fibers may innervate the dendrites extending from neurons within the VMH35, thereby regulating BDNF expression within the VMH. Finally, it is also possible that regulation involves one or more additional neurons that may reside outside of the VMH and whose projections control BDNF expression within the VMH.

Food deprivation for 2 d results in a ∼2/3 reduction in the levels of BDNF mRNAs in the VMH, which is substantially but incompletely reversed by administration of an MC4R agonist. Among potential reasons for the incompleteness of reversal are suboptimal activation of MC4R receptors due to poor penetration or rapid turnover of MTII, assay of BDNF mRNA before elevated synthesis has established a new equilibrium level, and possible involvement of additional signaling pathways whose activities are regulated by feeding status, but do not involve the MC4R.

Where are the TrkB-expressing neurons through which BDNF regulates feeding behavior and energy balance? Targeted deletion of a floxed BDNF allele with a CaMKII-driven cre results in obesity4, suggesting that the BDNF-expressing neurons in the VMH are probably a population of excitatory neurons that express endogenous CaMKII. BDNF is anterogradely transported to postsynaptic neurons in an activitydependent manner36, so the TrkB-expressing neurons could be located in any of the brain regions that receive projections from the VMH. Neurons in the VMH project to many targets within and outside the hypothalamus, including the DMH, anterior hypothalamus, bed nuclei of the stria terminalis, amygdalar nuclei, nucleus of the solitary tract and periaqueductal gray3739. Some of these targets such as the DMH, nucleus of the solitary tract and periaqueductal gray have been implicated in the regulation of energy intake4042. To identify the cells through which TrkB regulates feeding behavior will require extensive analyses using region-specific and cell type–specific elimination and overexpression of the TrkB receptor.

It has been shown that BDNF is required for gustatory development43,44 and that TrkB is expressed in the olfactory bulb45. Since the moderate-fat diet is more palatable, these observations raise the possibility that developmental defects and/or malfunction of the gustatory or olfactory system may affect feeding behavior of the trkB mutant on the moderate-fat diet. A defect in the gustatory or olfactory system most likely results in loss of the ability to distinguish the low-fat diet and moderate-fat diet in the trkB mutant. If so, a gustatory or olfactory defect may reduce ingestion of the moderate-fat diet. Therefore, a defect in the gustatory or olfactory system is unlikely to significantly contribute to hyperphagia on higher fat diets in the trkB mutant.

In conclusion, our results indicate that BDNF-to-TrkB signaling is an important downstream target of MC4R-mediated signaling that participates in the regulation of energy balance and feeding behavior. The obesity phenotypes observed in animals with compromised MC4R and TrkB signaling thus seem to be caused by deficits in different steps of a common or overlapping signaling pathway. In the future, it will be important to identify the neurons and cellular mechanisms that control each step of this pathway


Mouse strains

Mice were maintained on a normal 12 h/12 h light/dark cycle with regular mouse chow and water ad libitum. Ay mice and wild-type litter-mates were purchased from the Jackson Laboratories. Generation of fBZ/+, trkBlacZ/+ and BDNFLacZ/+ mice were as described previously28,30,31. The MC4R−/− null mice were offspring from crosses between MC4R−/− mice that had been backcrossed to C57BL/6J for five generations. The control C57BL/6J wild-type mice were purchased from Jackson Laboratories. Animal procedures were approved by the University of California San Francisco Committee on Animal Research.

Feeding behavior

Mice were individually housed for at least 3 d before measurement. Over 5 d, regular mouse chow was weighed and provided to 5- or 9-week old mice ad libitum with measurement at noon each day and careful monitoring for spillage. Each group consisted of three females and two males. The 5-week control group contained one fBZ heterozygous and four wild-type mice, whereas the 9-week control group consisted of three fBZ heterozygous and two wild-type mice.

Low-fat Purina diet 5001 (23.0% protein, 4.5% fat, 5.3% crude fiber, 49% carbohydrate, total digestible nutrient 76%, 3.04 kcal/g metabolizable energy) and moderate-fat Purina diet 5015 (17.0% protein, 11.0% fat, 3.0% crude fiber, 53.5% carbohydrate, total digestible nutrient 88%, 3.73 kcal/g metabolizable energy) were used to monitor the feeding response of mice to dietary fat. Mice were isolated and fed with Purina diet 5001 one week before any measurement of food intake and body weight.

Home cage behavioral monitoring

Male mice were acclimated to individual housing in feeding and licking monitoring cages (plexiglass, 45 × 24 × 17 cm with feeders and water bottles mounted at one end; AFL monitors, DiLog Instruments) for 6 d and then monitored for an additional 8 d of data collection. Additional details are provided in Supplementary Methods online.


X-gal staining and immunohistochemistry were performed as described previously31. Anti-β-galactosidase (1:4,000) was from Cappel Research Reagents. Anti-γ-MSH and anti-AgRP were provided by S. Watson (University of Michigan, Ann Arbor, Michigan) and G. Barsh (Stanford University, Stanford, California), respectively.

For in situ hybridization, mice were anesthetized and fixed with sequential transcardial perfusion of PBS and 4% paraformaldehyde. The brains were dissected, immersed in 30% sucrose, and then frozen. Alternatively, mice were killed by neck dislocation and the brains were dissected and frozen immediately. BDNF in situ hybridization was performed on cryostat coronal sections at 30 µm using 35S-labeled antisense RNA probes complementary to the coding region of the mouse BDNF cDNA After hybridization and washes, sections were exposed on Beta-Max Hyperfilm (Amersham). Images from two sections that covered the caudal VMH and were separated by 150 µm in each mouse were scanned at 1,200 d.p.i. and the optical density of in situ signals was determined using NIH Image.

MTII injection

Ten-week-old female wild-type C57BL/6J mice were food-deprived for 44 h and anesthetized with avertin. The dorsal third ventricle (2 mm caudal from bregma, 2.5 mm deep) was injected with 4 nmol MTII (Bachem) or artificial cerebrospinal fluid (ACSF) in a volume of 3 µl with a Hamilton syringe over a 20-s period. ACSF was also injected into a group of normally fed mice. The needle stayed in the brain for another 15 s before it was removed. Three hours later, mice were killed by cervical dislocation, and their brains were immediately dissected out and frozen.

To determine the effect of MTII injection on expression of BDNF in Ay mice, 3 nmol of MTII or saline was intracerebroventricularly injected into Ay/a; BDNFlacZ/+ mice. Six hours later, the brains were fixed and processed for anti-β-galactosidase immunohistochemistry. The VMH of each brain at approximately bregma – 1.8 mm was photographed, and β-galactosidase-expressing neurons were counted (by an experimenter blind to treatment group).

BDNF infusion

Female Ay mice and wild-type littermates were isolated and fed with Purina diet 5001 one week before pump implantation. Alzet osmotic pumps (model 1002,0.25 µl/h for 14 d) were filled with PBS or human recombinant BDNF (Amgen) in PBS at a concentration of 200 ng/µl. The dosage of BDNF was chosen according to a previous study3. Pumps were then attached to an Alzet brain infusion kit and the entire apparatus was primed overnight in sterile saline at 37 °C. The tip of an infusion catheter was positioned in the lateral ventricle (0.5 mm caudal and 1 mm lateral from bregma, 2.5 mm deep) of mice anesthetized with avertin and secured with Loctisite gel. Starting a day after surgery, body weight and food intake were measured daily. Upon completion of the 2-week study, the position of the cannula tip and the volume of remaining solution in the pump were determined. Animals were excluded from analysis if they died (n = 1), appeared sick (n = 1) or had misplaced or malfunctioning pumps (n = 2).


We thank M. Dalman for constructive suggestions, J. Qiu for help in dissection of fat pads, J. McKean for technical assistance and H. Chen for demonstration of implantation of osmotic pumps. E.H.G. is a Howard Hughes Medical Institute (HHMI) Physician Postdoctoral Fellow. This work has been supported by the HHMI and the National Institute of Neurological Disorders and Stroke (L.F.R.).


Note: Supplementary information is available on the Nature Neuroscience website.


The authors declare that they have no competing financial interests.


1. Huang EJ, Reichardt LF. Neurotrophins: roles in neuronal development and function. Annu. Rev. Neurosci. 2001;24:677–736. [PMC free article] [PubMed]
2. Lyons WE, et al. Brain-derived neurotrophic factor-deficient mice develop aggressiveness and hyperphagia in conjunction with brain serotonergic abnormalities. Proc. Natl. Acad. Sci. USA. 1999;96:15239–15244. [PMC free article] [PubMed]
3. Kernie SG, Liebl DJ, Parada LF. BDNF regulates eating behavior and locomotor activity in mice. EMBO J. 2000;19:1290–1300. [PMC free article] [PubMed]
4. Rios M, et al. Conditional deletion of brain-derived neurotrophic factor in the postnatal brain leads to obesity and hyperactivity. Mol. Endocrinol. 2001;15:1748–1757. [PubMed]
5. Ono M, et al. Brain-derived neurotrophic factor reduces blood glucose level in obese diabetic mice but not in normal mice. Biochem. Biophys. Res. Commun. 1997;238:633–637. [PubMed]
6. Nakagawa T, et al. Brain-derived neurotrophic factor regulates glucose metabolism by modulating energy balance in diabetic mice. Diabetes. 2000;49:436–444. [PubMed]
7. Nonomura T, et al. Brain-derived neurotrophic factor regulates energy expenditure through the central nervous system in obese diabetic mice. Int. J. Exp. Diabetes Res. 2001;2:201–209. [PMC free article] [PubMed]
8. Tsuchida A, et al. Acute effects of brain-derived neurotrophic factor on energy expenditure in obese diabetic mice. Int. J. Obes. Relat. Metab. Disord. 2001;25:1286–1293. [PubMed]
9. Brooks CM, Lockwood RA, Wiggins ML. A study of the effect of hypothalamic lesions on the eating habits of the albino rat. Am. J. Physiol. 1946;147:735–741. [PubMed]
10. Hetherington AW, Ranson SW. The relation of various hypothalamic lesions to adiposity in the rat. J. Comp. Neurol. 1942;76:475–499.
11. Shimizu N, Oomura Y, Plata-Salaman CR, Morimoto M. Hyperphagia and obesity in rats with bilateral ibotenic acid-induced lesions of the ventromedial hypothalamic nucleus. Brain Res. 1987;416:153–156. [PubMed]
12. Friedman JM, Halaas JL. Leptin and the regulation of body weight in mammals. Nature. 1998;395:763–770. [PubMed]
13. Flier JS, Maratos-Flier E. Obesity and the hypothalamus: novel peptides for new pathways. Cell. 1998;92:437–440. [PubMed]
14. Schwartz MW, Woods SC, Porte D, Jr, Seeley RJ, Baskin DG. Central nervous system control of food intake. Nature. 2000;404:661–671. [PubMed]
15. Elmquist JK, Elias CF, Saper CB. From lesions to leptin: hypothalamic control of food intake and body weight. Neuron. 1999;22:221–232. [PubMed]
16. Smith AI, Funder JW. Proopiomelanocortin processing in the pituitary, central nervous system, and peripheral tissues. Endocr. Rev. 1988;9:159–179. [PubMed]
17. Baskin DG, et al. Insulin and leptin: dual adiposity signals to the brain for the regulation of food intake and body weight. Brain Res. 1999;848:114–123. [PubMed]
18. Batterham RL, et al. Gut hormone PYY(3–36) physiologically inhibits food intake. Nature. 2002;418:650–654. [PubMed]
19. Heisler LK, et al. Activation of central melanocortin pathways by fenfluramine. Science. 2002;297:609–611. [PubMed]
20. Benoit SC, et al. The catabolic action of insulin in the brain is mediated by melanocortins. J. Neurosci. 2002;22:9048–9052. [PubMed]
21. Cone RD, et al. The arcuate nucleus as a conduit for diverse signals relevant to energy homeostasis. Int. J. Obes. Relat. Metab. Disord. 2001;25:S63–S67. [PubMed]
22. Bultman SJ, Michaud EJ, Woychik RP. Molecular characterization of the mouse agouti locus. Cell. 1992;71:1195–1204. [PubMed]
23. Miller MW, et al. Cloning of the mouse agouti gene predicts a secreted protein ubiquitously expressed in mice carrying the lethal yellow mutation. Genes Dev. 1993;7:454–467. [PubMed]
24. Huszar D, et al. Ta rgeted disruption of the melanocortin-4 receptor results in obesity in mice. Cell. 1997;88:131–141. [PubMed]
25. Chen AS, et al. Role of the melanocortin-4 receptor in metabolic rate and food intake in mice. Transgenic Res. 2000;9:145–154. [PubMed]
26. Lu D, et al. Agouti protein is an antagonist of the melanocyte-stimulating hormone receptor. Nature. 1994;371:799–802. [PubMed]
27. Butler AA, et al. Melanocortin-4 receptor is required for acute homeostatic responses to increased dietary fat. Nat. Neurosci. 2001;4:605–611. [PubMed]
28. Bennett JL, Zeiler SR, Jones KR. Patterned expression of BDNF and NT-3 in the retina and anterior segment of the developing mammalian eye. Invest. Ophthalmol. Vis. Sci. 1999;40:2996–3005. [PubMed]
29. Barsh GS, Schwartz MW. Genetic approaches to studying energy balance: perception and integration. Nat. Rev. Genet. 2002;3:589–600. [PubMed]
30. Xu B, et al. Cortical degeneration in the absence of neurotrophin signaling: dendritic retraction and neuronal loss after removal of the receptor TrkB. Neuron. 2000;26:233–245. [PubMed]
31. Xu B, et al. The role of brain-derived neurotrophic factor receptors in the mature hippocampus: modulation of long-term potentiation through a presynaptic mechanism involving TrkB. J. Neurosci. 2000;20:6888–6897. [PMC free article] [PubMed]
32. Bagnol D, et al. Anatomy of an endogenous antagonist: relationship between Agouti-related protein and proopiomelanocortin in brain. J. Neurosci. 1999;19:1–7. [PubMed]
33. Fan W, Boston BA, Kesterson RA, Hruby VJ, Cone RD. Role of melanocortinergic neurons in feeding and the agouti obesity syndrome. Nature. 1997;385:165–168. [PubMed]
34. Kishi T, et al. Expression of melanocortin 4 receptor mRNA in the central nervous system of the rat. J. Comp. Neurol. 2003;457:213–235. [PubMed]
35. Ter Horst GJ, Luiten PG. Phaseolus vulgaris leuco-agglutinin tracing of intrahy-pothalamic connections of the lateral, ventromedial, dorsomedial and paraventricular hypothalamic nuclei in the rat. Brain Res. Bull. 1987;18:191–203. [PubMed]
36. Kohara K, Kitamura A, Morishima M, Tsumoto T. Activity-dependent transfer of brain-derived neurotrophic factor to postsynaptic neurons. Science. 2001;291:2419–2423. [PubMed]
37. Saper CB, Swanson LW, Cowan WM. The efferent connections of the ventro-medial nucleus of the hypothalamus of the rat. J. Comp. Neurol. 1976;169:409–442. [PubMed]
38. Krieger MS, Conrad LC, Pfaff DW. An autoradiographic study of the efferent connections of the ventromedial nucleus of the hypothalamus. J. Comp. Neurol. 1979;183:785–815. [PubMed]
39. Canteras NS, Simerly RB, Swanson LW. Organization of projections from the ventromedial nucleus of the hypothalamus: a Phaseolus vulgaris-leucoagglutinin study in the rat. J. Comp. Neurol. 1994;348:41–79. [PubMed]
40. Williams G, et al. The hypothalamus and the control of energy homeostasis: different circuits, different purposes. Physiol. Behav. 2001;74:683–701. [PubMed]
41. Skultety FM. Changes in caloric intake following brain stem lesions in cats. 3. Effects of lesions of the periaqueductal gray matter and rostral hypothalamus. Arch. Neurol. 1966;14:670–680. [PubMed]
42. Bernardis LL, Bellinger LL. The dorsomedial hypothalamic nucleus revisited: 1998 update. Proc. Soc. Exp. Biol. Med. 1998;218:284–306. [PubMed]
43. Zhang C, Brandemihl A, Lau D, Lawton A, Oakley B. BDNF is required for the normal development of taste neurons in vivo. Neuroreport. 1997;8:1013–1017. [PubMed]
44. Fritzsch B, Sarai PA, Barbacid M, Silos-Santiago I. Mice with a targeted disruption of the neurotrophin receptor trkB lose their gustatory ganglion cells early but do develop taste buds. Int. J. Dev. Neurosci. 1997;15:563–576. [PubMed]
45. Deckner ML, Frisen J, Verge VM, Hokfelt T, Risling M. Localization of neurotrophin receptors in olfactory epithelium and bulb. Neuroreport. 1993;5:301–304. [PubMed]
PubReader format: click here to try


Save items

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • Cited in Books
    Cited in Books
    PubMed Central articles cited in books
  • ClinVar
    Links to ClinVar
  • Compound
    PubChem Compound links
  • Gene
    Gene links
  • GEO Profiles
    GEO Profiles
    Related GEO records
  • HomoloGene
    HomoloGene links
  • MedGen
    Related information in MedGen
  • OMIM
    OMIM record citing PubMed
  • Pathways + GO
    Pathways + GO
    Pathways, annotations and biological systems (BioSystems) that cite the current article.
  • PubMed
    PubMed citations for these articles
  • Substance
    PubChem Substance links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...