Skip to main content
NCBI home page
Proceedings of the National Academy of Sciences of the United States of America logoLink to Proceedings of the National Academy of Sciences of the United States of America
. 2008 Feb 13;105(7):2687–2692. doi: 10.1073/pnas.0712062105

Starvation after AgRP neuron ablation is independent of melanocortin signaling

Qi Wu *, Maureen P Howell †, Michael A Cowley ‡, Richard D Palmiter *,§
PMCID: PMC2268197  PMID: 18272480

Abstract

Ablation of inhibitory agouti-related protein (AgRP)-expressing neurons in the arcuate nucleus that also synthesize γ-amino-butyric acid (GABA) and neuropeptide Y in adult mice leads to starvation within 1 week. The removal of inhibition from the AgRP neurons onto neighboring proopiomelanocortin neurons and their common postsynaptic neurons is predicted to stimulate melanocortin signaling, which is known to inhibit appetite. To examine the importance of uncontrolled melanocortin signaling in mediating starvation in this model, we ablated AgRP neurons in Ay/a mice that have chronic blockade of the melanocortin signaling. The blockade of melanocortin signaling did not ameliorate the rate of starvation. On both WT and Ay/a genetic backgrounds, there was a progressive decrease in meal frequency after AgRP neuron ablation. Surprisingly, intraoral feeding also was dramatically reduced after the ablation of AgRP neurons. These results indicate that both the appetitive and consummatory aspects of feeding become impaired in a melanocortin-independent manner after AgRP neuron ablation.

Keywords: agouti-related protein, feeding behavior, neuropeptide Y, Fos expression, diphtheria toxin

The melanocortin signaling pathway in the medial hypothalamus has emerged as a critical mediator of hormonal and neuronal signals that regulate energy balance (1–3). Genetic, pharmacological, anatomical, and electrophysiological techniques have established a pivotal role for proopiomelanocortin (POMC) neurons in the arcuate (ARC) and their signaling via melanocortin (α-MSH) to cells bearing melanocortin-4 receptors (MC4R) in the paraventricular nucleus (PVN) and other brain regions in the control of appetite and metabolism (1, 4, 5). Neighboring cells that coexpress neuropeptide Y (NPY) and agouti-related protein (AgRP) also have been implicated in appetite and energy balance because these peptides stimulate robust feeding when injected into the brain, and their mRNA levels increase in the ARC under starvation conditions, as well as in obese animals deficient in leptin signaling (2, 6–8). Various experiments suggest that NPY acts by inhibiting the activity of POMC neurons and postsynaptic MC4R-bearing cells by activating Gαi-coupled NPY Y1 and/or Y5 receptors (7, 9–12). The activation of these receptors is predicted to counteract melanocortin activation of Gαs-coupled MC4R. AgRP acts in concert with NPY by blocking the binding of α-MSH to MC4R (8, 13, 14). The neurons that make NPY and AgRP also produce GABA, which inhibits POMC neurons and probably MC4R-bearing cells (15, 16). Thus, these AgRP neurons are poised to inhibit melanocortin signaling by the neighboring POMC cells.

The activation of melanocortin signaling by leptin inhibits appetite while stimulating metabolism (1, 2, 6). Consequently, the inactivation of mouse genes encoding leptin, leptin receptor, POMC, or MC4R leads to obesity (17–20). However, the inactivation of genes encoding NPY, AgRP, or both has little effect on energy balance (21–23). These genetic results suggest that compensatory mechanisms in the KO mice mask the normal role of these peptides. Nevertheless, the AgRP neurons are essential for normal body weight regulation because ablating them in adult mice inhibits feeding and results in starvation (24, 25). However, the critical signaling molecules that they make and the neural networks that they regulate are unknown. Excessive activation of the melanocortin pathway after ablation of AgRP neurons represents an attractive explanation for the starvation phenotype because AgRP neurons normally inhibit nearby POMC neurons and project to nearly all of the same brain regions as POMC neurons where GABA, NPY, and/or AgRP would individually or in combination counteract melanocortin signaling (1, 26–31).

Feeding behavior can be subdivided into at least two phases: an appetitive phase, characterized by food-seeking behavior that involves motivation to initiate feeding, followed by a consummatory phase that includes ingestion, chewing, and swallowing the food. The appetitive phase involves many brain regions, including the dopaminergic reward pathways (32), whereas the consummatory phase involves the hindbrain that can function independently of forebrain structures (33). We present evidence that both the appetitive and consummatory aspects of feeding are disrupted by the ablation of AgRP neurons, and these effects appear to be independent of melanocortin signaling.

Results

Ablation of AgRP Neurons Results in Neuronal Activation.

The cell bodies of all AgRP-expressing neurons reside exclusively in the ARC. We targeted expression of a human diphtheria toxin receptor (hDTR) cDNA to the Agrp locus of mice (AgrpDTR) to allow selective killing of AgRP-expressing neurons in adult mice by the administration of diphtheria toxin (DT). The standard protocol (two injections of DT, 2 days apart) results in progressive loss of food consumption and body weight, such that the mice lose 20% of their initial body weight after 6–7 days and would starve without intervention (24). In situ hybridization revealed that this procedure results in the loss of ≈99% of the Agrp hybridization signal in the ARC of AgrpDTR mice by 6 days, compared with mice that were pair-fed to lose the same amount of weight (Fig. 1 A–C). DT treatment of control mice had no effect on Agrp expression, food intake, or body weight (ref. 24 and data not shown).

Fig. 1.

Fig. 1.

Open in a new tab

DT treatment of AgrpDTR/+ mice results in the loss of Agrp transcripts in the ARC and induction of Fos in the ARC, PVN, and NTS. AgrpDTR/+ mice were injected twice with DT, and brain sections were prepared from them and from pair-fed (PF) controls 6 days later for in situ hybridization. (A and B) Representative sections through the ARC showing the loss of Agrp expression. (C) Quantification of the hybridization pixels ± SEM from six to eight sections from three or four different mice in each group. (D and E) Representative sections through the ARC showing induction of Fos expression. (F) Quantification as in C. (G and H) Representative sections thought the PVN showing the induction of Fos expression. (I) Quantification as in C. (J and K) Representative sections through the NTS showing the induction of Fos expression. (L) Quantification as in C. The dotted lines in A, G, and J outline regions in the ARC, PVN, and NTS, respectively, that were quantified. *, P < 0.01, ANOVA.

The ablation of inhibitory AgRP neurons could result in unopposed excitation that might manifest as an increase in Fos gene expression. Sections from DT-treated AgrpDTR mice and pair-fed controls were hybridized with a Fos probe 6 days after initiation of DT treatment. Fig. 1 D–F shows robust induction of Fos expression in the ARC that could represent expression in POMC neurons that are known to be inhibited by AgRP neurons (9). In addition, there was robust induction of Fos in brain regions, such as the PVN and the nucleus tractus solitarius (NTS), that are innervated by POMC and AgRP neurons (Fig. 1 G–L). There was essentially no Fos expression in untreated control mice in any of these brain regions, and the DT treatment of control mice had little effect on Fos expression (data not shown).

POMC neurons receive inhibitory GABA signaling from NPY neurons that coexpress AgRP (9). Thus, the ablation of AgRP neurons should remove this inhibitory tone, resulting in hyperactivity of POMC neurons. To test this hypothesis, the AgrpDTR allele was bred into a Pomc-EGFP genetic background so that POMC neurons could be identified in slices to allow electrophysiological recording. AgrpDTR/+; Pomc-EGFP mice were treated with DT as above, and brain slices were prepared for electrophysiology when the mice had lost ≈20% of their body weight. Control Pomc-EGFP mice also were treated with DT, but pair-fed to lose ≈20% of their body weight. Miniature inhibitory postsynaptic currents (mIPSCs) were recorded by a whole-cell, patch-clamp technique in POMC neurons within hypothalamic brain slices (Fig. 2). The frequency of mIPSCs from fasted, control Pomc-EGFP mice (3.3 ± 0.77 Hz; n = 9 cells) was higher than the 2-Hz frequency observed in fed mice (9). In contrast, the frequency of mIPSCs onto POMC cells from DT-treated, AgrpDTR/+; Pomc-EGFP mice was 6-fold lower (0.52 ± 0.12 Hz; n = 6 cells; Mann–Whitney U test, P < 0.05). These results confirm that the ablation of AgRP neurons removes most of the inhibitory afferents onto POMC neurons, which may account for the excessive activation of POMC neurons, as well as postsynaptic neurons that receive melanocortin signaling.

Fig. 2.

Fig. 2.

Open in a new tab

Ablation of AgRP neurons reduces the frequency of GABA release onto POMC neurons. The frequency of mIPSCs recorded from POMC neurons of Pomc-EGFP; AgrpDTR/+ mice 4 days after treatment with DT or Pomc-EGFP mice that had been treated with DT and pair-fed to lose ≈20% of body weight. *, P < 0.05, Mann–Whitney U test.

Starvation Persists in Adult Ay/a Mice After Ablation of AgRP Neurons.

To ask whether the blockade of melanocortin signaling would blunt the hyperactivity of postsynaptic neurons and thereby prevent starvation, we analyzed the feeding behavior of control mice and mice with chronic expression of agouti protein, which blocks the action of α-MSH on melanocortin receptors (1). Homozygous AgrpDTR/DTR mice were bred with heterozygous lethal yellow agouti Ay/a mice to generate a/a; AgrpDTR/+ mice (which we will refer to as AgrpDTR/+ mice) and Ay/a; AgrpDTR/+ (Ay; AgrpDTR/+) mice. The two groups of mice could be distinguished because the Ay/a; AgrpDTR/+ mice had yellow fur and developed early-onset obesity (1). The 6-week-old AgrpDTR/+ and Ay; AgrpDTR/+ mice were treated with DT as before, which effectively ablated the AgRP neurons in the Ay genetic background (data not shown).

The feeding response of DT-treated mice was analyzed by using “lickometer” cages that recorded every contact of the mice with the liquid diet dispensers (34). The mice were allowed to acclimate to the lickometer cages for 3 days before measuring baseline food intake. DT treatment had no effect on body weight, intake of liquid diet, or licking activity by WT or Ay/a mice (Fig. 3 A and B) (data not shown). However, starting at 2 days after the second DT injection, AgrpDTR/+ and Ay; AgrpDTR/+ mice displayed a gradual decrease in food intake and a progressive loss of body weight. By 6 days after the first DT treatment, consumption of the liquid diet of both experimental groups dropped to <5% of the untreated level, and their body weights fell to ≈80% of their initial weights, the limit established for intervention by our animal care protocol (ANOVA, P < 0.001) (Fig. 3 A and B). Because we have never observed recovery, it is likely that the mice would have starved to death. These results indicate that chronic blockade of the melanocortin signaling in the Ay mice did not prevent starvation caused by AgRP neuron ablation.

Fig. 3.

Fig. 3.

Open in a new tab

Starvation persists in adult Ay; AgrpDTR/+ mice after the ablation of AgRP neurons. (A) Percentage of initial body weight of AgrpDTR/+ (n = 12), Ay; AgrpDTR/+ (n = 12), and WT (n = 6) mice after injection with 50 μg/kg DT. (B) Liquid diet intake of the mice described in A. (C) Licking activity of Ay; AgrpDTR/+ (n = 6) and AgrpDTR/+ (n = 6) mice before and after two DT injections. The average number of licks in 2-h bins is plotted in both groups. Body weight (day 0) of Ay; AgrpDTR/+ mice was 32.1 ± 1.48 g, compared with 25.6 ± 0.9 g for AgrpDTR/+ mice. Error bars represent ± SEM.

The diurnal feeding pattern persisted, but the average number of licks by both AgrpDTR/+ and Ay; AgrpDTR/+ mice progressively declined, and no significant difference was identified between these two groups (ANOVA, P > 0.05) (Fig. 3C). Further analysis revealed that the number of meals (clusters of licks) decreased dramatically in both groups as they lost weight, suggesting that the mice became less motivated to initiate feeding as they lost weight. However, meal size actually increased 2- to 3-fold (Fig. 4 A–D).

Fig. 4.

Fig. 4.

Open in a new tab

Licking activity in Ay; AgrpDTR/+ and AgrpDTR/+ mice before and after two DT injections. (A and C) The total number of meals (A) and the average meal size (as shown by licks per meal) (C) were plotted in 2-h bins for both groups as the animals were progressively losing their body weight. (B and D) Daily meal counts (B) and meal size (D) were measured in both groups. Meals are defined as the number of consecutive 10-s bins that contain five or more licks, and each meal must be separated by 12 or more 10-s bins in which no significant licking activity was recorded (i.e., four or less licks per 10-s bin). * and #, P < 0.05, ANOVA.

Food Consumption Measured by Intraoral Feeding Also Is Disrupted by AgRP Neuron Ablation.

To ascertain whether consummatory control of feeding was intact, we measured the intake of palatable liquid diets during delivery by intraoral fistula. When sucrose or milk is slowly pumped into the mouth, rodents manifest characteristic behaviors (taste reactivity), including paw wiping, face washing, and bipedal rearing, which ends with tongue protrusions, head shaking, and food dripping from the mouth when they become satiated. By contrast, intraoral delivery of bitter substances, such as quinine, elicits distinctive gaping, chin-rubbing, and paw-pushing behaviors (35). Control and AgrpDTR/DTR mice were fitted with intraoral fistulas and then treated with DT. The control mice were pair-fed to match the weight loss of the AgrpDTR/DTR mice. As the control mice lost body weight, their consumption of either sucrose solution or milk increased, in agreement with previous studies with rats (36). In contrast, the DT-treated AgrpDTR/DTR mice did not increase their intake at 10% loss of body weight, and their intake decreased dramatically as their weight loss approached 20% (Fig. 5). Compared with pair-fed controls, the DT-treated AgrpDTR/DTR mice displayed normal hedonic behaviors, but they quickly transitioned to food-rejection behaviors indicative of satiety. The DT-treated AgrpDTR/DTR mice displayed normal aversive responses to quinine (data not shown).

Fig. 5.

Fig. 5.

Open in a new tab

Consumption of sucrose or milk solution is disrupted in AgRP neuron-ablated mice. Consumption of 0.1 M sucrose solution (A) and diluted condensed milk (B) infused through cheek fistulas was measured in DT-treated AgrpDTR/DTR and pair-fed WT mice when at 100%, ≈90%, and ≈80% of original body weight. At least six mice in each group were tested. * and #, P < 0.05, ANOVA; ** and ##, P < 0.001, ANOVA.

Discussion

Mice with targeted ablation of AgRP-expressing neurons provide an opportunity to study the neural circuits that mediate the starvation phenotype. Here we determine whether the starvation phenotype is a result of excessive activation of the melanocortin pathway, which is a well established target of AgRP neurons. AgRP neurons not only directly inhibit POMC neurons, but they also send projections to most of the same target neurons as POMC neurons (1, 13, 27, 28). Direct GABA-mediated inhibition of POMC neurons has been documented electrophysiologically (9). The frequency of mIPSCs onto POMC neurons increased in pair-fed control mice as they lost weight, as would be expected due to decreased leptin signaling onto GABAergic AgRP neurons (37, 38). The ablation of AgRP neurons suppressed this inhibitory input onto POMC neurons 6-fold. We also observed robust activation of Fos gene expression in the ARC, PVN, NTS, and many of the other targets of AgRP neurons after the ablation of AgRP neurons (Fig. 1 and Q.W., M.P.H., and R.D.P, unpublished work). These observations are consistent with our hypothesis that the ablation of AgRP neurons removes the inhibitory tone onto postsynaptic cells, resulting in excessive activation by unopposed excitatory inputs.

A relatively complete profile of AgRP- and POMC-immunoreactive fibers projected from the ARC has been characterized in rodents (27–31). However, the functional significance of the overlap in POMC- and AgRP-positive fibers in target areas is not established. POMC neurons provide excitatory input to postsynaptic cells, as revealed by treatment with melanocortin analogs, which activates Fos gene expression in target neurons (39–41). Thus, we expected that chronic blockade of melanocortin signaling would reduce the activation of postsynaptic cells when AgRP neurons were ablated and either prevent or retard the rate of starvation. However, blockade of melanocortin signaling by chronic expression of agouti protein had no discernable effect on the starvation phenotype caused by AgRP neuron ablation, which suggests that excessive activation of the melanocortin signaling is not responsible for starvation.

One explanation for why blockade melanocortin signaling does not ameliorate starvation could be that POMC neurons express transmitters other than melanocortin that are responsible for hyperactivity in downstream neurons. For example, POMC neurons also express cocaine- and amphetamine-regulated transcript (CART), and these neurons have been reported to be either GABAergic or glutamatergic (42–44). Thus, in addition to α-MSH, these or as yet unidentified neurotransmitters could influence Fos expression in postsynaptic cells. Our results from comparing Fos expression after the ablation of AgRP neurons in either WT or Ay genetic background showed that Fos activation was reduced in many targets of POMC and AgRP neurons in the Ay genetic background, but not to basal levels (Q.W., M.P.H., and R.D.P., unpublished data), suggesting that excitatory inputs from other neurons contribute to the neuronal activation. Because of the residual activation of Fos gene expression in postsynaptic cells after AgRP neuron ablation in Ay mice, we suggest that the dysregulation of neuron activity in some or all of those brain regions contributes to the starvation phenotype.

As a first step in trying to understand which neuronal populations are most important in mediating starvation, we sought to distinguish whether appetitive or consummatory aspects of feeding were perturbed. Our data indicate that both processes become disrupted after the ablation of AgRP neurons. The decrease in the number of meals as the mice lost weight (Fig. 4 A and B) reflects a decrease in the initiation of feeding bouts, an appetitive deficit. In addition, the amount of food consumed by means of direct delivery of milk or sugar water into the mouth also decreased as the ablated mice lost weight (Fig. 5). Thus, even if the mice were motivated to initiate meals, they would not consume an adequate amount of food to maintain body weight.

The consummatory deficit is one of the most striking findings of these studies. Previous studies, primarily in rats, have shown that the food consumption delivered by intraoral fistula responds to both hormonal and vagal satiety signals. However, the magnitude of the reduction of food consumption in these paradigms is small relative to that described here (36). Pair-fed control mice that had lost ≈20% of their body weight would consume ≈2.5 ml of milk before letting it spill from their mouth, whereas the ablated mice consumed only ≈0.2 ml. These results suggest that AgRP neuron ablation results in the dysregulation of consummatory circuits, such that the mice behave as if they are satiated. If AgRP neurons normally inhibit neural circuits, thereby facilitating consumption, then the death of AgRP neurons could inappropriately activate those circuits and inhibit consumption. Another situation in which consummatory responses are grossly impaired is a rat preparation lacking a thalamus (45). However, those rats showed nonspecific aversive responses to both quinine and sucrose, unlike the ablated mice, which displayed normal taste reactivity to sucrose and quinine. The observation that the size of the meals consumed in the free-feeding situation actually increased as the ablated mice lost weight (Fig. 4 C and D) seems to contradict the intraoral feeding data. However, the actual meal size only increased to 200–300 licks, which, at 1.07 μl per lick, is equivalent to 0.2–0.3 ml, similar to the volume that the ablated mice consumed via intraoral fistula. Thus, as the free-feeding, AgRP-ablated mice lost body weight, they drank much less per meal than their capacity of ≈2.5 ml.

Classical experiments using decerebrate rats revealed that consummatory responses are largely intact, indicating that the hindbrain circuits can function independently of forebrain inputs, including those from the hypothalamus (33, 36). Those experiments do not eliminate the possibility that the forebrain influences hindbrain functions in a normal animal. There is general agreement that the activation of caudal brainstem is necessary for feeding (33), but the various afferents to the caudal brainstem and neural circuits within the brainstem that regulate feeding remain to be elucidated. AgRP neurons innervate caudal brainstem regions, such as the periaqueductal gray, parabrachial nuclei, and NTS (27), and some of these nuclei receive afferents from forebrain regions, such as the PVN and lateral hypothalamus, that receive input from AgRP neurons (46). The experiments described here suggest that the dysregulation within the caudal brainstem is responsible for consummatory deficit that occurs after ablation of AgRP neurons in the adult mouse, but it is unclear whether the dysregulation is due to the loss of direct AgRP neuron inputs to brainstem nuclei or to indirect effects mediated by the dysregulation in forebrain regions. Neuroanatomical studies indicate that hypothalamic POMC neurons directly innervate a few brainstem regions, including the NTS (47), and there is a local population of POMC-expressing neurons in the NTS (1, 48). Despite these indications that the NTS is an important site of melanocortin signaling, our results suggest that blockade of melanocortin signaling with agouti protein is insufficient to overcome the effects of AgRP neuron ablation.

Methods

Animal Maintenance.

Mice were housed in a temperature- and humidity-controlled facility with a 12-h light/dark cycle. All animal care and experimental procedures were approved by the Institutional Animal Care and Use Committee at the Universities of Washington and Oregon. AgrpDTR/+ mice were generated as described previously (24). Ay; AgrpDTR/+ mice were generated by mating homozygous AgrpDTR/DTR with heterozygous lethal yellow Ay/a mice (The Jackson Laboratory). The yellow mice derived from this cross comprise the Ay; AgrpDTR/+ group, whereas the black littermates served as AgrpDTR/+ controls. AgrpDTR/+ mice were bred with Pomc-EGFP mice (9) to generate Pomc-EGFP; AgrpDTR/+ mice and controls (Pomc-EGFP) for the slice electrophysiology experiments. Mice were group-housed with standard chow diet (5053; LabDiet) and water ad libitum until the beginning of the experiments.

AgRP Neuron Ablation and Feeding Experiments.

To ablate AgRP neurons, i.m. injection of DT (two injections of 50 μg/kg, 2 days apart; List Biological Laboratories) in 6-week-old mice was performed (24). For all feeding assays, mice were transferred into lickometer cages (Columbus Instruments) supplied with water and a rodent liquid diet (5LD-101; TestDiet). The mice were allowed to acclimate for 3 days before initiating DT treatment and data collection. Body weight and intake of liquid diet were recorded every 24 h. Licking activity at food and water dispensers was monitored continuously by computer as described (24, 49). Licking data were processed and grouped into 2-h bins by using the MatLab software (Mathworks). Meal number and meal size were analyzed as described (50), and the average lick volume was 1.07 μl per lick.

Before the implantation of cheek fistulas, the mice were anesthetized by i.p. injection (10 ml/kg body weight) of a KXA mixture (PBS buffer consisting of 10 mg/ml ketamine, 2 mg/ml xylazine, and 0.1 mg/ml acepromazine). Each fistula assembly had an internal washer on one end of a 5-cm-long PE50 polyethylene tubing (Becton Dickinson). A 23-gauge needle mounted on the other end of the fistula tubing was used to pierce the cheek just rostral and lateral to the first molar where damage to the mastication muscle was minimized. The cheek fistula was then carefully pulled out until the internal washer was located snugly against the oral cavity wall and locked by fixing an external washer against the skin surface. Mice were allowed to recover from surgery for 1 week, during which 25 mg/kg Baytril (Bayer) was supplemented in drinking water to prevent inflammation. Animals were maintained on liquid diet (5LD-101), and fistulas were flushed with PBS buffer on a daily basis.

After recovery, mice were familiarized to the intraoral infusion protocol by infusing liquid food once per day for 3 or 4 days. The diet for intraoral feeding was 0.1 M sucrose or diluted sweetened condensed milk (1:5 dilution in water), which have been used previously (51). The fistula was connected with polyethylene tubing (PE50; Becton Dickinson) to a 5-ml syringe mounted on a single-speed syringe pump (PHM-100-.5; Med Associates). Intraoral feeding tests were conducted with direct observation by using an open-topped, 9 × 12 × 6-inch acrylic chamber, in which the mouse was clearly visible. A paper towel was placed under the chamber to help visualize spillage. During the test, liquid food was infused at a speed of 0.1 ml/min in successive 2-min periods, separated by 20-s resting periods until the liquid started dripping out of the animal's mouth. Infusion was suspended for 20 s and then restarted. If the mouse refused the food again within 10 s on three successive restarts, the mouse was determined to be satiated, and the test was terminated. The mice were videotaped during the feeding sessions to record their behaviors.

Electrophysiology.

Hypothalamic brain slices were prepared for electrophysiological recordings as described (9, 26, 52). Briefly, under isoflurane anesthesia, male mice were decapitated, and the brain was removed. Coronal hypothalamic slices containing the ARC were cut at 185 μm on a vibratome under ice-cold, gassed (95%O2/5%CO2) artificial cerebrospinal fluid and then maintained at 22°C for 1 h before recordings. POMC cells were located under fluorescent illumination, and then high-resistance seals (≈1GΩ) onto cells were obtained under infrared illumination. To measure mIPSCs, voltage-clamp recordings were obtained by using a Axopatch 200B patch-clamp amplifier and analyzed by using pClamp 8.0 (Axon Instruments) and MiniAnalysis 5.0 (Synaptosoft). Ten minutes of recording were obtained from each cell. Excitatory currents were blocked with 10 μM CNQX and 25 μM APV in the bath solution to isolate GABAergic mIPSCs. The extracellular solution had the following composition (mM): NaCI 126, KCI 2.5, MgCI2 1.2, CaCI2 2.4, NaH2PO4 1.2, NaHCO3 21.4, glucose 11.1; pH 7.4. The cells were filled with a cesium-based solution with the following composition (mM): CsCI 140, MgCI2 5, (Mg)ATP 5, (Na)GTP 0.3, BAPTA 1, Hepes 10; pH 7.3 (with CsOH), 290 mOsm; the holding potential was −60 mV.

In Situ Hybridization.

Brains were sectioned (coronal 25 μm), and every eight sections were used for either Nissl staining or in situ hybridization with Agrp or Fos probes by using an automated procedure for hybridization and image capture. The materials and procedures concerning this high-throughput data-generation process (riboprobe production, tissue processing, in situ hybridization, and image capture and processing) have been described (53) and are available on the Allen Brain Atlas web site (www.brain-map.org).

ACKNOWLEDGMENTS.

We thank Glenda Froelick for help with histology, Aundrea Rainwater for help with mouse breeding, Serge Luquet for initiating this project, and the Allen Institute for Brain Science staff for performing the in situ hybridization studies. This work was supported in part by National Institutes of Health Grants RR0163 and DK62202 (to M.A.C.) and DA024908 (to R.D.P.).

Footnotes

The authors declare no conflict of interest.

References

  • 1.Cone RD. Anatomy and regulation of the central melanocortin system. Nat Neurosci. 2005;8:571–578. doi: 10.1038/nn1455. [DOI] [PubMed] [Google Scholar]
  • 2.Morton GJ, et al. Central nervous system control of food intake and body weight. Nature. 2006;443:289–295. doi: 10.1038/nature05026. [DOI] [PubMed] [Google Scholar]
  • 3.Saper CB, Chou TC, Elmquist JK. The need to feed: Homeostatic and hedonic control of eating. Neuron. 2002;36:199–211. doi: 10.1016/s0896-6273(02)00969-8. [DOI] [PubMed] [Google Scholar]
  • 4.Balthasar N, et al. Divergence of melanocortin pathways in the control of food intake and energy expenditure. Cell. 2005;123:493–505. doi: 10.1016/j.cell.2005.08.035. [DOI] [PubMed] [Google Scholar]
  • 5.Kublaoui BM, et al. SIM1 overexpression partially rescues agouti yellow and diet-induced obesity by normalizing food intake. Endocrinology. 2006;147:4542–4549. doi: 10.1210/en.2006-0453. [DOI] [PubMed] [Google Scholar]
  • 6.Barsh GS, Schwartz MW. Genetic approaches to studying energy balance: Perception and integration. Nat Rev Genet. 2002;3:589–600. doi: 10.1038/nrg862. [DOI] [PubMed] [Google Scholar]
  • 7.Kalra SP, et al. Interacting appetite-regulating pathways in the hypothalamic regulation of body weight. Endocr Rev. 1999;20:68–100. doi: 10.1210/edrv.20.1.0357. [DOI] [PubMed] [Google Scholar]
  • 8.Ollmann MM, et al. Antagonism of central melanocortin receptors in vitro and in vivo by agouti-related protein. Science. 1997;278:135–138. doi: 10.1126/science.278.5335.135. [DOI] [PubMed] [Google Scholar]
  • 9.Cowley MA, et al. Leptin activates anorexigenic POMC neurons through a neural network in the arcuate nucleus. Nature. 2001;411:480–484. doi: 10.1038/35078085. [DOI] [PubMed] [Google Scholar]
  • 10.Marsh DJ, Hollopeter G, Kafer KE, Palmiter RD. Role of the Y5 neuropeptide Y receptor in feeding and obesity. Nat Med. 1998;4:718–721. doi: 10.1038/nm0698-718. [DOI] [PubMed] [Google Scholar]
  • 11.Pedrazzini T, et al. Cardiovascular response, feeding behavior and locomotor activity in mice lacking the NPY Y1 receptor. Nat Med. 1998;4:722–726. doi: 10.1038/nm0698-722. [DOI] [PubMed] [Google Scholar]
  • 12.Beck B. Neuropeptide Y in normal eating and in genetic and dietary-induced obesity. Philos Trans R Soc London Ser B. 2006;361:1159–1185. doi: 10.1098/rstb.2006.1855. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Broberger C, Hokfelt T. Hypothalamic and vagal neuropeptide circuitries regulating food intake. Physiol Behav. 2001;74:669–682. doi: 10.1016/s0031-9384(01)00611-4. [DOI] [PubMed] [Google Scholar]
  • 14.Lu XY, et al. Time course of short-term and long-term orexigenic effects of Agouti-related protein (86–132). NeuroReport. 2001;12:1281–1284. doi: 10.1097/00001756-200105080-00045. [DOI] [PubMed] [Google Scholar]
  • 15.Horvath TL, et al. Heterogeneity in the neuropeptide Y-containing neurons of the rat arcuate nucleus: GABAergic and non-GABAergic subpopulations. Brain Res. 1997;756:283–286. doi: 10.1016/s0006-8993(97)00184-4. [DOI] [PubMed] [Google Scholar]
  • 16.Pu S, et al. Interactions between neuropeptide Y, gamma-aminobutyric acid in stimulation of feeding: A morphological and pharmacological analysis. Endocrinology. 1999;140:933–940. doi: 10.1210/endo.140.2.6495. [DOI] [PubMed] [Google Scholar]
  • 17.Zhang Y, et al. Positional cloning of the mouse obese gene and its human homologue. Nature. 1994;372:425–432. doi: 10.1038/372425a0. [DOI] [PubMed] [Google Scholar]
  • 18.Tartaglia LA, et al. Identification and expression cloning of a leptin receptor, OB-R. Cell. 1995;83:1263–1271. doi: 10.1016/0092-8674(95)90151-5. [DOI] [PubMed] [Google Scholar]
  • 19.Huszar D, et al. Targeted disruption of the melanocortin-4 receptor results in obesity in mice. Cell. 1997;88:131–141. doi: 10.1016/s0092-8674(00)81865-6. [DOI] [PubMed] [Google Scholar]
  • 20.Krude H, et al. Severe early-onset obesity, adrenal insufficiency and red hair pigmentation caused by POMC mutations in humans. Nat Genet. 1998;19:155–157. doi: 10.1038/509. [DOI] [PubMed] [Google Scholar]
  • 21.Erickson JC, Clegg KE, Palmiter RD. Sensitivity to leptin and susceptibility to seizures of mice lacking neuropeptide Y. Nature. 1996;381:415–421. doi: 10.1038/381415a0. [DOI] [PubMed] [Google Scholar]
  • 22.Erickson JC, Hollopeter G, Palmiter RD. Attenuation of the obesity syndrome of ob/ob mice by the loss of neuropeptide Y. Science. 1996;274:1704–1707. doi: 10.1126/science.274.5293.1704. [DOI] [PubMed] [Google Scholar]
  • 23.Qian S, et al. Neither agouti-related protein nor neuropeptide Y is critically required for the regulation of energy homeostasis in mice. Mol Cell Biol. 2002;22:5027–5035. doi: 10.1128/MCB.22.14.5027-5035.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Luquet S, et al. NPY/AgRP neurons are essential for feeding in adult mice but can be ablated in neonates. Science. 2005;310:683–685. doi: 10.1126/science.1115524. [DOI] [PubMed] [Google Scholar]
  • 25.Gropp E, et al. Agouti-related peptide-expressing neurons are mandatory for feeding. Nat Neurosci. 2005;8:1289–1291. doi: 10.1038/nn1548. [DOI] [PubMed] [Google Scholar]
  • 26.Cowley MA, et al. The distribution and mechanism of action of ghrelin in the CNS demonstrates a novel hypothalamic circuit regulating energy homeostasis. Neuron. 2003;37:649–661. doi: 10.1016/s0896-6273(03)00063-1. [DOI] [PubMed] [Google Scholar]
  • 27.Broberger C, et al. The neuropeptide Y/agouti gene-related protein (AGRP) brain circuitry in normal, anorectic, and monosodium glutamate-treated mice. Proc Natl Acad Sci USA. 1998;95:15043–15048. doi: 10.1073/pnas.95.25.15043. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Haskell-Luevano C, et al. Characterization of the neuroanatomical distribution of agouti-related protein immunoreactivity in the rhesus monkey and the rat. Endocrinology. 1999;140:1408–1415. doi: 10.1210/endo.140.3.6544. [DOI] [PubMed] [Google Scholar]
  • 29.Jacobowitz DM, O'Donohue TL. alpha-Melanocyte stimulating hormone: Immunohistochemical identification and mapping in neurons of rat brain. Proc Natl Acad Sci USA. 1978;75:6300–6304. doi: 10.1073/pnas.75.12.6300. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Nilaver G, et al. Adrenocorticotropin and beta-lipotropin in the hypothalamus. Localization in the same arcuate neurons by sequential immunocytochemical procedures. J Cell Biol. 1979;81:50–58. doi: 10.1083/jcb.81.1.50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Watson SJ, Akil H, Richard CW, Barchas JD. Evidence for two separate opiate peptide neuronal systems. Nature. 1978;275:226–228. doi: 10.1038/275226a0. [DOI] [PubMed] [Google Scholar]
  • 32.Berthoud HR. Multiple neural systems controlling food intake and body weight. Neurosci Biobehav Rev. 2002;26:393–428. doi: 10.1016/s0149-7634(02)00014-3. [DOI] [PubMed] [Google Scholar]
  • 33.Grill HJ. Distributed neural control of energy balance: Contributions from hindbrain and hypothalamus. Obesity. 2006;14(Suppl 5):216S–221S. doi: 10.1038/oby.2006.312. [DOI] [PubMed] [Google Scholar]
  • 34.Cannon CM, Palmiter RD. Reward without dopamine. J Neurosci. 2003;23:10827–10831. doi: 10.1523/JNEUROSCI.23-34-10827.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Grill HJ, Norgren R. The taste reactivity test: I. Mimetic responses to gustatory stimuli in neurologically normal rats. Brain Res. 1978;143:263–279. doi: 10.1016/0006-8993(78)90568-1. [DOI] [PubMed] [Google Scholar]
  • 36.Grill HJ, Kaplan JM. The neuroanatomical axis for control of energy balance. Front Neuroendocrinol. 2002;23:2–40. doi: 10.1006/frne.2001.0224. [DOI] [PubMed] [Google Scholar]
  • 37.Pinto S, et al. Rapid rewiring of arcuate nucleus feeding circuits by leptin. Science. 2004;304:110–115. doi: 10.1126/science.1089459. [DOI] [PubMed] [Google Scholar]
  • 38.Takahashi KA, Cone RD. Fasting induces a large, leptin-dependent increase in the intrinsic action potential frequency of orexigenic arcuate nucleus neuropeptide Y/Agouti-related protein neurons. Endocrinology. 2005;146:1043–1047. doi: 10.1210/en.2004-1397. [DOI] [PubMed] [Google Scholar]
  • 39.Glavas MM, et al. Melanocortinergic activation by melanotan II inhibits feeding and increases uncoupling protein 1 messenger ribonucleic acid in the developing rat. Endocrinology. 2007;148:3279–3287. doi: 10.1210/en.2007-0184. [DOI] [PubMed] [Google Scholar]
  • 40.Trivedi P, et al. Exploring the site of anorectic action of peripherally administered synthetic melanocortin peptide MT-II in rats. Brain Res. 2003;977:221–230. doi: 10.1016/s0006-8993(03)02683-0. [DOI] [PubMed] [Google Scholar]
  • 41.Thiele TE, et al. Central infusion of melanocortin agonist MTII in rats: Assessment of c-Fos expression and taste aversion. Am J Physiol. 1998;274:R248–R254. doi: 10.1152/ajpregu.1998.274.1.R248. [DOI] [PubMed] [Google Scholar]
  • 42.Elias CF, et al. Leptin activates hypothalamic CART neurons projecting to the spinal cord. Neuron. 1998;21:1375–1385. doi: 10.1016/s0896-6273(00)80656-x. [DOI] [PubMed] [Google Scholar]
  • 43.Hentges ST, et al. GABA release from proopiomelanocortin neurons. J Neurosci. 2004;24:1578–1583. doi: 10.1523/JNEUROSCI.3952-03.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Collin M, et al. Plasma membrane and vesicular glutamate transporter mRNAs/proteins in hypothalamic neurons that regulate body weight. Eur J Neurosci. 2003;18:1265–1278. doi: 10.1046/j.1460-9568.2003.02840.x. [DOI] [PubMed] [Google Scholar]
  • 45.Grill HJ, Norgren R. Chronically decerebrate rats demonstrate satiation but not bait shyness. Science. 1978;201:267–269. doi: 10.1126/science.663655. [DOI] [PubMed] [Google Scholar]
  • 46.Blevins JE, Schwartz MW, Baskin DG. Evidence that paraventricular nucleus oxytocin neurons link hypothalamic leptin action to caudal brain stem nuclei controlling meal size. Am J Physiol. 2004;287:R87–R96. doi: 10.1152/ajpregu.00604.2003. [DOI] [PubMed] [Google Scholar]
  • 47.Zheng H, Patterson LM, Phifer CB, Berthould HR. Brain stem melanocortinergic modulation of meal size and identification of hypothalamic POMC projections. Am J Physiol. 2005;289:R427–R458. doi: 10.1152/ajpregu.00869.2004. [DOI] [PubMed] [Google Scholar]
  • 48.Joseph SA, Pilcher WH, Bennet-Clarke C. Immunocytochemical localization of ACTH parikaryia in the nucleus tractus solitarius: Evidence for a second opiocortin neuronal system. Neurosci Lett. 1983;38:221–225. doi: 10.1016/0304-3940(83)90372-5. [DOI] [PubMed] [Google Scholar]
  • 49.Cannon CM, et al. Dysregulation of striatal dopamine signaling by amphetamine inhibits feeding by hungry mice. Neuron. 2004;44:509–520. doi: 10.1016/j.neuron.2004.10.009. [DOI] [PubMed] [Google Scholar]
  • 50.Hnasko TS, et al. Cre recombinase-mediated restoration of nigrostriatal dopamine in dopamine-deficient mice reverses hypophagia and bradykinesia. Proc Natl Acad Sci USA. 2006;103:8858–8863. doi: 10.1073/pnas.0603081103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Hudson B, Ritter S. Hindbrain catecholamine neurons mediate consummatory responses to glucoprivation. Physiol Behav. 2004;82:241–250. doi: 10.1016/j.physbeh.2004.03.032. [DOI] [PubMed] [Google Scholar]
  • 52.Batterham RL, et al. Gut hormone PYY (3–36) physiologically inhibits food intake. Nature. 2002;418:650–654. doi: 10.1038/nature00887. [DOI] [PubMed] [Google Scholar]
  • 53.Lein ES, et al. Genome-wide atlas of gene expression in the adult mouse brain. Nature. 2007;445:168–176. doi: 10.1038/nature05453. [DOI] [PubMed] [Google Scholar]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences

RESOURCES