The Hypothalamic Feeding Circuit: Peptides That Flip Hunger
Gut-Brain Peptide Signaling
2 opposing neuron populations
The arcuate nucleus contains two peptide-producing neuron types with opposite effects: AgRP/NPY neurons drive hunger, while POMC neurons produce alpha-MSH to suppress it.
Vohra et al., Eur J Pharmacol, 2022
Vohra et al., Eur J Pharmacol, 2022
View as imageYour brain decides whether you feel hungry or full based on a peptide tug-of-war happening in a tiny brain region called the arcuate nucleus. Two populations of neurons sit side by side, producing opposite signals. One group releases AgRP and NPY, peptides that say "eat." The other group releases alpha-MSH (from the precursor POMC), a peptide that says "stop eating." Both signals converge on the same downstream target: the melanocortin-4 receptor (MC4R) in the paraventricular nucleus. When AgRP wins, you eat. When alpha-MSH wins, you stop.[1] This circuit is not abstract neuroscience. It is the direct target of the most effective weight loss drugs ever developed, and mutations in its components cause the most severe forms of genetic obesity. For a broader look at how the vagus nerve connects gut peptides to the brain, see the cluster pillar.
Key Takeaways
- The arcuate nucleus contains two opposing peptide neuron types: AgRP/NPY (hunger-promoting) and POMC (satiety-promoting), both converging on melanocortin-4 receptors[1]
- AgRP acts as an inverse agonist at MC4R, actively blocking alpha-MSH's satiety signal rather than simply competing with it
- In diet-induced obesity, ghrelin resistance develops in AgRP/NPY neurons: high-fat feeding suppresses the ghrelin system and reduces NPY/AgRP responsiveness to hunger signals[5]
- A distinct population of AgRP-negative NPY neurons was discovered in 2023 that drives feeding under positive energy balance by altering POMC neuron leptin responsiveness[3]
- Incretin receptor agonists (GLP-1 and GIP drugs) rapidly inhibit AgRP neurons via direct action on the arcuate nucleus, and dual agonism produces stronger inhibition than either alone[2]
- Setmelanotide, an MC4R agonist, produced at least 10% weight loss in 80% of POMC-deficient and 45% of LEPR-deficient patients in phase 3 trials[4]
The Arcuate Nucleus: Where Hunger and Satiety Are Decided
The arcuate nucleus (ARC) sits at the base of the hypothalamus, directly adjacent to the median eminence, a circumventricular organ where the blood-brain barrier is permeable. This positioning is not accidental: it allows ARC neurons to sense circulating hormones (insulin, leptin, ghrelin) that other brain neurons cannot access. The ARC is the brain's metabolic sensor, and it uses peptides to broadcast what it detects.
Two neuron populations in the ARC produce the core feeding signals:[1]
AgRP/NPY neurons co-express three signals: agouti-related peptide (AgRP), neuropeptide Y (NPY), and the inhibitory neurotransmitter GABA. All three promote food intake through different mechanisms. These neurons are activated by fasting, ghrelin, and low leptin levels. They are the brain's hunger neurons.
POMC neurons express pro-opiomelanocortin, a large precursor peptide that is cleaved to produce alpha-melanocyte-stimulating hormone (alpha-MSH). Alpha-MSH activates melanocortin-4 receptors (MC4R) in downstream brain regions to suppress appetite and increase energy expenditure. POMC neurons are activated by leptin, insulin, and serotonin.
The two populations physically inhibit each other. AgRP/NPY neurons send GABAergic projections directly onto POMC neurons, silencing them when hunger signals dominate. The result is a seesaw: activating one population suppresses the other.
MC4R: The Receptor Where It All Converges
Both neuron populations project to the paraventricular nucleus of the hypothalamus (PVH), where melanocortin-4 receptors sit on downstream neurons. MC4R is the decision point.
Alpha-MSH from POMC neurons activates MC4R, which reduces food intake and increases energy expenditure. AgRP from hunger neurons does not simply compete with alpha-MSH for receptor binding. AgRP is an inverse agonist: it binds MC4R and actively suppresses its baseline activity below normal levels, producing a stronger hunger signal than just blocking alpha-MSH would achieve.
This mechanism explains why MC4R mutations produce the most common form of monogenic (single-gene) obesity. Loss-of-function mutations in MC4R remove the satiety brake entirely, regardless of how much alpha-MSH POMC neurons produce. For a detailed analysis of MC4R mutations and childhood obesity, see the dedicated article. The downstream role of AgRP as the peptide that blocks MC4R is covered separately.
How Peripheral Hormones Set the Circuit
The ARC does not operate in isolation. It integrates signals from the gut, pancreas, and fat tissue:
Ghrelin (from the stomach): The only known circulating hormone that stimulates appetite. Ghrelin activates AgRP/NPY neurons via the growth hormone secretagogue receptor (GHS-R). When ghrelin rises before meals, AgRP/NPY neurons fire, producing hunger. For more on how ghrelin stimulates both growth hormone and appetite, see the dedicated article.
Leptin (from fat tissue): Signals energy reserves. Leptin inhibits AgRP/NPY neurons and activates POMC neurons. When fat stores are adequate, leptin suppresses hunger and promotes satiety. When leptin drops (during dieting or fat loss), AgRP/NPY neurons are released from inhibition, driving hunger. The failure of this system in obesity, leptin resistance, is a critical concept.
Insulin (from the pancreas): Like leptin, insulin inhibits AgRP/NPY neurons and activates POMC neurons. Rising insulin after a meal shifts the circuit toward satiety.
Incretins (GLP-1 and GIP from the gut): These peptides, traditionally known for their role in insulin secretion, also directly inhibit AgRP neurons.[2][6]
Ghrelin Resistance: When Obesity Breaks the Hunger Circuit
In lean animals, ghrelin powerfully activates AgRP/NPY neurons and stimulates food intake. But in diet-induced obesity, this response breaks down. Briggs and colleagues demonstrated in 2010 that 12 weeks of high-fat feeding in mice caused ghrelin resistance in arcuate NPY/AgRP neurons.[5]
The mechanism was multi-layered: plasma ghrelin levels dropped, ghrelin and GOAT (ghrelin O-acyltransferase) mRNA decreased in the stomach, and hypothalamic expression of the ghrelin receptor (GHS-R) fell. When ghrelin was injected directly into the brain of obese mice, it failed to induce food intake or activate arcuate neurons, even though downstream NPY/AgRP targets remained functional. The problem was not in the appetite circuits downstream; it was that ghrelin could no longer reach or activate the AgRP/NPY neurons that initiate the signal.
This finding has implications for understanding why obesity is self-perpetuating. As the hunger circuit becomes desensitized to ghrelin, the normal meal-initiation signal degrades. But the leptin resistance that also develops in obesity means the satiety brake weakens simultaneously. The result is a dysregulated circuit that neither properly signals hunger nor properly signals fullness.
The Third Population: AgRP-Negative NPY Neurons
For decades, the feeding circuit was described as a simple two-neuron model: AgRP/NPY versus POMC. A 2023 Cell Metabolism paper by Qi and colleagues revealed a third player: a population of NPY-expressing neurons in the arcuate nucleus that do not co-express AgRP.[3]
These AgRP-negative NPY neurons drive feeding under conditions of positive energy balance (when animals are already well-fed), a situation where classical AgRP/NPY neurons are relatively quiet. The mechanism involves altering leptin responsiveness in neighboring POMC neurons: by reducing POMC neurons' sensitivity to leptin, these AgRP-negative NPY neurons weaken the satiety signal, enabling continued feeding even when energy stores are sufficient.
This discovery complicates the simple hunger-versus-satiety model and may help explain why overeating persists even when the classical AgRP hunger system is not strongly activated.
How GLP-1 Drugs Target This Circuit
The connection between incretin drugs and the hypothalamic feeding circuit was directly demonstrated by McMorrow and colleagues in 2025.[2] Using fiber photometry (which measures neural activity in real time in living mice), they showed that both GLP-1 and GIP receptor agonists rapidly inhibited AgRP neuron firing.
Key findings:
- GIP, not GLP-1, was required for normal nutrient-mediated inhibition of AgRP neurons during a meal
- Both GIP and GLP-1 analogs at pharmacological doses were sufficient to inhibit AgRP neurons
- Dual GIP+GLP-1 agonism more potently inhibited AgRP neurons and suppressed food intake than either agonist alone
- The magnitude of AgRP inhibition correlated directly with the degree of appetite suppression
This study provides a mechanistic explanation for why tirzepatide (dual GIP/GLP-1 agonist) produces greater weight loss than semaglutide (GLP-1 only): it more strongly silences the hunger neurons. For a broader view of how the gut-brain axis controls blood sugar and satiety, and how brain versus gut peptides coordinate appetite, see the sibling articles.
Setmelanotide: Proof That the Circuit Works as a Drug Target
The most direct pharmacological validation of the melanocortin feeding circuit came from setmelanotide (Imcivree), an MC4R agonist approved by the FDA in 2020 for genetic obesity caused by deficiencies in the POMC, PCSK1, or LEPR genes.
In phase 3 trials, Clement and colleagues treated patients with severe obesity due to POMC or LEPR deficiency with setmelanotide for approximately one year.[4] Eight of ten (80%) patients with POMC deficiency and five of eleven (45%) patients with LEPR deficiency achieved at least 10% weight loss. Hunger scores dropped in both groups.
The drug works by substituting for the missing alpha-MSH signal. In POMC-deficient patients, POMC neurons cannot produce alpha-MSH, so MC4R receives no satiety signal. Setmelanotide directly activates MC4R, restoring the downstream effects of the satiety pathway. In LEPR-deficient patients, POMC neurons cannot detect leptin and therefore remain inactive; again, setmelanotide bypasses the broken sensor by activating the downstream receptor directly.
Setmelanotide does not work for common obesity. Its efficacy is limited to patients with specific genetic defects in the melanocortin pathway. But it proves the principle: activating MC4R reduces appetite and body weight in humans, confirming decades of animal research on this circuit.
The Bottom Line
The hypothalamic feeding circuit centers on two opposing peptide neuron populations in the arcuate nucleus. AgRP/NPY neurons drive hunger; POMC neurons produce alpha-MSH to signal satiety. Both converge on MC4R, where the competition determines whether you eat or stop. Peripheral hormones (ghrelin, leptin, insulin, incretins) set the balance. GLP-1 and GIP drugs work in part by directly inhibiting AgRP neurons, while setmelanotide validates MC4R as a drug target for genetic obesity. The 2023 discovery of AgRP-negative NPY neurons adds complexity to a circuit previously thought to be a simple binary switch.