Search tips
Search criteria 


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 2013 August 28.
Published in final edited form as:
PMCID: PMC3755601

Lighting up the hypothalamus: coordinated control of feeding behavior


A new optogenetics study finds that stimulation of pro-opiomelanocortin (POMC) and agouti-related peptide (AGRP) neurons acutely regulates feeding behavior. AGRP-induced hyperphagia is independent of melanocortin signaling.

It could be argued that the hypothalamus was initially implicated in the control of food intake and obesity nearly 170 years ago with the first clinical description, by Bernhard Mohr, of hypothalamic-pituitary injury resulting in obesity1. At the turn of the twentieth century, Joseph Babinski and Alfred Frolich also reported a condition of obesity and retarded sexual development was attributed to hypothalamic-pituitary injury1. In the past century, several reports have advanced the concept that the hypothalamus is required for the regulation of energy balance1,2. Despite these seminal observations, throughout the decades contemporary methodologies have been a limiting factor in determining the physiological effects of perturbing signaling in vivo at the cellular and/or circuit level. In this issue of Nature Neuroscience, Aponte and colleagues3 use mouse optogenetics to probe neural circuits in the arcuate nucleus of the hypothalamus and find that POMC and AGRP neurons have counter-regulatory roles on the regulation of food intake.

Aponte et al.3 took advantage of channelrhodopsin-2 (ChR2) to alter neuronal activity and extend previous observations in an attempt to determine whether neurons that produce AGRP and neuropeptide Y (NPY) can acutely regulate food intake. ChR2, isolated from Chlamydomonas reinhardtii, was used previously4 to rapidly and robustly control neuronal activity. ChR2 absorbs blue (480 nM) light, which induces a conformational change that opens the channel pore, permitting ions to flow, causing a depolarization and ultimately resulting in increased action potentials. In the brief time since the introduction of optogenetics, this technique has found widespread use in neuroscience, including studies that range from expanding our understanding of basic neuroscience principles to investigating neuropsychiatric disorders such as Parkinson’s disease4,5. Aponte et al.3 used a Cre recombinase-dependent virus expressing the light-activated ChR2 fused to the fluorescent protein tdtomato. Injection of the virus into either Pomc-cre or Agrp-cre mice enabled identification of neurons expressing the ChR2 and allowed light-mediated activation of either POMC or AGRP neurons, respectively. Light activation of POMC neurons resulted in a frequency-dependent reduction of food intake that required downstream melanocortin receptor activity (Fig. 1). Activation of AGRP neurons resulted in an acute, frequency-dependent increase in food intake; however, the acute hyperphagia was independent of melanocortin signaling, supporting a previously identified role of GABA in mediating the acute starvation (Fig. 1)6.

Figure 1
Optogenetics defines feeding behavior in the arcuate nucleus of the hypothalamus. Photostimulation of arcuate POMC neurons results in the activation of melanocortin receptors on second-order neurons that ultimately results in decreased food intake. Light ...

A key advance of the work of Aponte et al.3 is that it highlights the importance of the temporal regulation of AGRP and POMC neuronal activity. This temporal regulation is manifest in at least two mechanisms, one long term and the other acute. The long-term issue relates to the potential for developmental ‘compensation’ when genes are inactivated using standard or conditional knockout approaches. For example, recent reports have generated some confusion over the role of NPY and AGRP neurons in the regulation of food intake6. Although a conventional knockout in the mouse of either POMC, which is the precursor of α-melanocyte stimulating hormone (α-MSH), or its downstream receptors (MC3 or MC4) causes hyperphagia and obesity, a conventional knockout of either NPY or AGRP results in normal body weight, adiposity and food intake7-9. In addition, toxin-mediated ablation of NPY and AGRP neurons early in development results in only modest changes in food intake6. However, ablation of NPY and AGRP neurons in adult mice causes acute hypophagia and a reduction in body weight6,10. Notably, the acute ablation kills the target cells entirely, whereas, in the conventional knockout models, the mice lack only the specified peptide and the neurons are therefore still present in the CNS. This suggests that there is the potential for compensatory mechanisms independent of NPY and AGRP in the conventional knockout models that may account for the regulation of food intake. Some recent work suggests that the neurotransmitter GABA, which is colocalized in NPY and AGRP neurons, may be critical for regulating food intake and/or energy balance6. Collectively, these observations further demonstrate that there are inherent limitations in approaches using mice that develop while either lacking or overexpressing key proteins of interest. Moving forward, approaches such as that used by Aponte et al.3, in combination with newer inducible Cre models, will help us understand how motivated behaviors such as feeding are regulated.

The second important advance of the Aponte et al.’s study3 relates to a much smaller temporal window. Specifically, their results highlight the importance of acute neuronal activity in regulating feeding behaviors, especially those behaviors mediated by classic neurotransmitters. Consistent with this observation, recent work found that AGRP neurons regulate food intake and energy balance independent of melanocortin signaling6,11. Specifically, selective deletion of the vesicular GABA transporter (Vgat) from AGRP neurons using Cre-loxP resulted in mice that are lean and resistant to high-fat diet (HFD)-induced obesity, but with no major effect on food intake, similar to mice with neonatal ablation of AGRP neurons11. The lean phenotype is a result of increased locomotor activity, whereas an increased HFD-induced thermogenesis contributes to the resistance of HFD-induced obesity. Notably, these mice do show an attenuated hyperphagic response to ghrelin, which was attributed to an inability of AGRP neurons to antagonize POMC neuronal activity. Moreover, it was also reported that AGRP neurons modify food intake at least in part through a GABAergic regulation of the parabrachial nucleus (Fig. 1)6. The current study by Aponte et al.3 is also quite notable in that it provides a bona fide functional correlate (that is, feeding behavior) with altered neuronal firing. These results can now be integrated with several in vitro slice studies demonstrating that AGRP and NPY neurons alter their firing rates when metabolically relevant signals are applied to the slice (for example, ghrelin, leptin or serotonin)2.

Not surprisingly, these results also raise several questions for future investigation. In addition to regulating energy balance, the mediobasal hypothalamus has been of intense interest in the regulation of glucose homeostasis12,13. In 1849, Claude Bernard suggested that the brain regulates blood glucose levels and suggested that this effect was dependent on stimulation of sympathetic innervation of the liver. However, few have been able to further elucidate this finding until recently available molecular techniques have narrowed possible cell populations in the brain that may be responsible for these effects on blood glucose. For instance, an early report suggested overexpression of the Pomc gene lowered blood glucose in mice2,14. Opposing these reports of POMC-induced regulation of glucose homeostasis, several studies suggest little if any involvement of NPY and AGRP neurons on glucose homeostasis2. In addition, both POMC and NPY and AGRP neurons have been implicated in the regulation of processes that include reward, addiction, locomotor activity and energy expenditure2,15. However, owing to earlier technological limitations, their roles have not been completely defined. Thus, the current optogenetic strategy has great potential to delineate the dependence of many modalities on both POMC and NPY/AGRP neurons.

In summary, from the time of Mohr, Bernard, Babinski and Frolich, the field of obesity and diabetes research has seen substantial progress in understanding the neural mechanisms involved in the regulation of energy and glucose homeostasis. Through the decades, our understanding has been restricted as a result of inherent limitations in available techniques. Aponte et al.3 extend previous observations in the field with the use of a state-of-the-art optogenetics strategy to functionally probe neural circuits in the medial basal hypo thalamus. Clearly, the authors have expanded our current understanding of the arcuate-based network of POMC and NPY and AGRP neurons. Although much is left to be learned, with this new technology in hand, the future of hypothalamic research is clearly bright.



The authors declare no competing financial interests.

Contributor Information

Kevin W Williams, Division of Hypothalamic Research, Department of Internal Medicine, University of Texas Southwestern Medical Center, Dallas, Texas, USA.

Joel K Elmquist, Division of Hypothalamic Research, Department of Internal Medicine, University of Texas Southwestern Medical Center, Dallas, Texas, USA; Department of Pharmacology, University of Texas Southwestern Medical Center, Dallas, Texas, USA.


1. Brobeck JR. Physiol. Rev. 1946;26:541–559. [PubMed]
2. Williams KW, Scott MM, Elmquist JK. Eur. J. Pharmacol. 2011 Jan 3; published online, doi:10.1016/j.ejphar.2010.11.042.
3. Aponte Y, Atasoy D, Sternson SM. Nat. Neurosci. 2011;14:351–356. [PMC free article] [PubMed]
4. Boyden ES, Zhang F, Bamberg E, Nagel G, Deisseroth K. Nat. Neurosci. 2005;8:1263–1268. [PubMed]
5. Kravitz AV, et al. Nature. 2010;466:622–626. [PMC free article] [PubMed]
6. Wu Q, Palmiter RD. Eur. J. Pharmacol. 2011 Jan 3; published online, doi:10.1016/j.ejphar.2010.10.110.
7. Yaswen L, Diehl N, Brennan MB, Hochgeschwender U. Nat. Med. 1999;5:1066–1070. [PubMed]
8. Qian S, et al. Mol. Cell. Biol. 2002;22:5027–5035. [PMC free article] [PubMed]
9. Butler AA, Cone RD. Neuropeptides. 2002;36:77–84. [PubMed]
10. Gropp E, et al. Nat. Neurosci. 2005;8:1289–1291. [PubMed]
11. Tong Q, Ye CP, Jones JE, Elmquist JK, Lowell BB. Nat. Neurosci. 2008;11:998–1000. [PMC free article] [PubMed]
12. Schwartz MW, Porte D., Jr. Science. 2005;307:375–379. [PubMed]
13. Woods SC, Seeley RJ, Porte D, Jr., Schwartz MW. Science. 1998;280:1378–1383. [PubMed]
14. Mizuno TM, Kelley KA, Pasinetti GM, Roberts JL, Mobbs CV. Diabetes. 2003;52:2675–2683. [PubMed]
15. Koob G. Neurobiol. Dis. 2000;7:543–545. [PubMed]