Mednotes

L12 Metabolism Neuroendocrine control of appetite

Properties2
tagsmetabolism, lecture, neuroendocrinology, obesity
aliasesNeuroendocrine control of appetite

19 min read

Overview

This lecture argues that body weight and food intake are controlled primarily by unconscious, physiological (homeostatic/allostatic) mechanisms rather than by free will alone, and traces the evidence for this from twin/animal genetics, the parabiosis experiments that led to the discovery of leptin, and the central circuit in the arcuate nucleus (AgRP/NPY and POMC neurons) that integrates peripheral hormonal signals into feeding behaviour. It then covers why leptin fails to prevent human obesity (leptin resistance and its proposed mechanisms), the intracellular JAK2/STAT3/SOCS-3 signalling pathway leptin uses, and finishes with applied content: current and emerging pharmacological therapies (especially GLP-1/GIP agonists), bariatric surgery types and outcomes, and the case for reframing obesity as a biological rather than a moral/willpower problem.

Body weight as a tightly regulated physiological variable

  • Two competing framings of why people become obese are presented: free will (conscious control over diet, exercise, sleep) vs physiology (food intake is unconsciously regulated by homeostatic mechanisms to match energy expenditure, involving hormone imbalance and genetics/gene-environment interaction).
  • Evidence body weight is under precise physiological control: consuming roughly 1 million calories per year while maintaining body weight to within ±1 kg requires intake and expenditure to be matched to within ±6000 calories — a 0.6% error, equivalent to about 16 calories/day (illustrated by a balance between “¼ of an apple” and “4 minutes of walking”).
  • Only about 20% of energy expenditure is under voluntary control.
  • Weight loss is difficult to sustain: the chance of returning to a normal weight after becoming obese is only about 1 in 210 for men and 1 in 124 for women per year (BBC News, 17 July 2015).
  • The upper education/socioeconomic bracket has markedly lower obesity than other groups, attributed to greater consciousness, motivation, and access to tools (time, money) needed to remain lean.

Genetic and evolutionary basis of obesity

  • Twin/adoptee studies indicate 40–85% of obesity-related phenotypes are heritable, similar to the heritability of height.
  • Single-gene mutations can cause obesity in animal models: the ob/ob mouse, the Agouti (Ay/a) mouse, the db/db mouse, and the Zucker fa/fa rat.
  • Thrifty genotype hypothesis (Neel, 1962): in environments where food is scarce, genes that predispose to increased energy storage during periods of food availability would confer a survival advantage. The lecture notes that the current wide availability of energy-rich food may be unprecedented in human history, so these genes are now a liability.

The homeostatic/allostatic signalling loop

The lecture presents a circulating signalling loop linking gut, brain, and fat:

  1. Food intake feeds into the GI tract.
  2. The GI tract sends neural signals, nutrients, and gut hormones to the brain, and energy absorbed to adipose tissue.
  3. Adipose tissue releases leptin and adipokines, which signal (inhibit) both the brain and energy expenditure.
  4. The brain integrates all these inputs to generate hunger, which drives further food intake; energy expended (only ~20% voluntary) also feeds back into the loop.

Falling adipose mass lowers leptin, and low leptin acts as a signal to prevent weight loss — it raises hunger and activity in brain reward centres, because the brain "wants" to store fuel for energy-demanding activities (e.g. reproduction). This is why sustained voluntary eating-less is described as "fighting against biology."

Discovery of leptin: parabiosis experiments

  • Parabiosis experiments (1950s–1970s) surgically joined the circulation of an obese animal and a lean/wild-type partner and observed the outcome in the lean partner (Hervey 1959; Coleman 1973):
Obese animalPaired withOutcome
MBH (mediobasal hypothalamus) lesionNormalLean partner starved
db/dbWild-typeLean partner starved
fa/faWild-typeLean partner starved
ob/obWild-typeob/ob (obese) animal lost body weight
  • Interpretation: MBH-lesioned, db/db, and fa/fa obese animals still produce a circulating satiety signal but cannot themselves respond to it (implying a receptor defect), so the signal crossed over and starved their lean partners. The ob/ob animal instead lacked the signal itself; receiving it from its wild-type partner via shared circulation caused it to lose weight — identifying the missing factor in ob/ob mice as leptin.

Slide 13 flag: the transcribed slide shows the leptin-deficient (ob/ob) mouse labelled "Lean/Fertile" in the rendered image, which conflicts with the phenotype stated elsewhere in the deck (obese/infertile). This is most likely a build/animation artifact (a post-treatment overlay caught mid-animation) rather than the intended base phenotype, but is recorded here exactly as it appeared on the captured slide, not corrected.

Leptin, appetite, and fertility

  • Leptin is required for appetite regulation and fertility: wild-type mice are lean and fertile; leptin-deficient (ob/ob) mice are obese and infertile (see flag above regarding the exact slide 13 labelling).
  • Leptin suppresses food intake, decreases body weight, and restores fertility in leptin-deficient subjects.
  • Case example: a leptin-deficient boy weighed 42 kg at age 3.5; after leptin treatment he weighed 32 kg at age 7. This was drawn as a parallel to an insulin-deficient girl treated with insulin (before/after), framing leptin therapy for leptin deficiency as analogous to insulin therapy for diabetes (Flier, Endocr Rev 2019).

Peripheral signals to the brain

Leptin is one of several peripheral hormones that influence food intake and basal metabolic rate (BMR), signalling to the brain via the vagus and spinal nerves:

  • Leptin (adipose tissue): ↓ intake, ↑ BMR
  • Insulin (pancreas): ↓ intake, ↑ BMR
  • Adiponectin (adipose tissue): ↑ BMR
  • Ghrelin (stomach): ↑ intake
  • PYY (small intestine): ↓ intake
  • CCK (small intestine, near pancreas): ↓ intake
  • GLP-1 (gut): ↓ intake

GLP-1 and related gut hormones also slow digestive emptying (the “ileal brake”) and have additional roles in bile release and insulin release.

Central integration: the arcuate nucleus (ARC) circuit

Slide 16 flag: this slide has no visible title bar (unlike every other slide in the deck) — the diagram fills the whole slide. Content is transcribed as shown; the missing title may be an original slide omission.

In the arcuate nucleus of the hypothalamus, two first-order neuron populations integrate peripheral signals:

  • AgRP/NPY neurons (agouti-related peptide / neuropeptide Y) — activation increases food intake.
  • POMC neurons (proopiomelanocortin) — activation decreases food intake.

Peripheral inputs to these first-order neurons:

  • Ghrelin (+) stimulates AgRP/NPY.
  • PYY (−) inhibits AgRP/NPY.
  • Leptin/insulin inhibit AgRP/NPY (−) and stimulate POMC (+).
  • AgRP/NPY neurons directly inhibit POMC neurons (−).

Downstream:

  • AgRP/NPY neurons project to second-order neurons that increase food intake.
  • POMC neurons project to second-order neurons that decrease food intake.
  • Both AgRP/NPY and POMC neurons also send satiety signals to the NTS (nucleus tractus solitarius) satiety centre in the brainstem.
  • The NTS separately receives CCK signals via the vagus nerve, and other afferents from the GI tract and liver via spinal nerves — both converge on satiety at the NTS.

Leptin resistance in human obesity

  • Complete leptin deficiency is extremely rare in humans.
  • Leptin levels are actually elevated in most obese humans, since leptin is secreted in proportion to the number and size of fat cells.
  • Human obesity is therefore thought to reflect a state of leptin resistance, not leptin deficiency.
  • Mechanisms of leptin resistance are not well understood, but three are proposed:
    1. Down-regulation of hypothalamic leptin receptors (illustrated with the Zucker fa/fa rat).
    2. Impaired leptin transport across the blood–brain barrier.
    3. Increased presence of negative regulators of leptin signalling, such as suppressor of cytokine signalling-3 (SOCS-3).

Leptin intracellular signalling pathway

  1. Leptin binds a cytokine-linked receptor on the cell membrane.
  2. This activates (phosphorylates) JAK2 associated with the receptor.
  3. Phosphorylated JAK2 activates STAT3, which becomes phosphorylated and dimerises.
  4. STAT3 translocates to the nucleus to regulate gene transcription.
  5. Chronically elevated leptin upregulates SOCS-3 (itself one of the genes induced by STAT3 activity), and SOCS-3 protein then feeds back to inhibit further JAK2/STAT3 activation.

This SOCS-3 negative-feedback loop is the proposed mechanism (mechanism 3 above) by which chronically high leptin, as seen in obesity, can itself induce leptin resistance.

Pharmacological management of obesity

Approved medications (mechanism, approval, mean placebo-subtracted weight loss %, main adverse effects):

MedicationMechanismApprovalWeight loss (%)Adverse effects
PhentermineCentrally acting sympathomimeticUSA, short-term adjunct (max 12 weeks; no long-term data)NADry mouth, insomnia, dizziness, palpitations, flushing, fatigue, constipation
Phentermine/topiramateSympathomimetic + GABA agonist/glutamate antagonist/CA inhibitorUSA8.6–9.3Paresthesias, dry mouth, constipation, insomnia, dysgeusia, anxiety, depression
OrlistatGI and pancreatic lipase inhibitorUSA, EU3Bloating, flatulence, mild steatorrhea, faecal urgency
Naltrexone/bupropionOpioid receptor antagonist + dopamine/norepinephrine reuptake inhibitorUSA, EU5.2Nausea, dizziness, dry mouth, insomnia, constipation
LiraglutideGLP-1 receptor agonistUSA, EU5.7 (1 yr), 4.4 (3 yr)Nausea, diarrhea
SemaglutideGLP-1 receptor agonistUSA10.3–12.4 (with lifestyle intervention)Nausea, diarrhea

A 2018 market prediction (cited on-slide) projected 6 new obesity drugs by 2026 (5 GLP-1 receptor agonists, 1 “first in class”).

Newer drug targets based on current understanding of bodyweight regulation:

  • Melanocortin receptor agonists, e.g. setmelanotide (MC4R agonist) — good results specifically in patients with POMC mutations.
  • GLP-1 and GIP analogues: gut-derived peptide analogues that slow digestive passage, stimulate insulin secretion and improve blood glucose control, and are anorexigenic. Use among adults and children without diabetes has risen sharply since 2020; cost is about NZ$400/month.
    • Semaglutide (Wegovy, Ozempic) — GLP-1 agonist, 5–10% body weight loss.
    • Tirzepatide (Mounjaro, Zepbound) — dual GLP-1/GIP agonist, >20% body weight loss.
    • Retatrutide (in clinical trials) — GLP-1, GIP + glucagon agonist.
    • Also reported to reduce smoking, drinking, shopping, and libido.

As of a 3 May 2025 news report, the WHO backs global use of weight-loss drugs for adults, while also raising the issue of cost.

How GLP-1/GIP agonists work (tirzepatide example)

  • Tirzepatide binds both GLP-1-secreting and GIP-secreting enteroendocrine cells in the gut lining.
  • Downstream actions on three targets:
    • Brain: decreased appetite, decreased food intake, increased weight loss.
    • Pancreas: increased insulin secretion, increased insulin synthesis, increased beta-cell survival.
    • White adipose tissue: increased lipolysis (via GLP-1) and increased lipogenesis (via GIP) together, which avoids ectopic fat deposition.
  • GLP-1 and its receptors are expressed in both gut and brain; systemically administered GLP-1 agonists can act directly in the arcuate nucleus (e.g. on POMC neurons) — shown by GLP-1 receptor immunofluorescent staining around the third ventricle/ARC.
  • GLP-1’s central actions are required for its weight-loss effect, but not for its glucose-lowering effect.
  • GLP-1 analogues likely act non-physiologically: they act centrally in a way that normal gut-derived GLP-1 does not, because endogenous GLP-1 does not normally reach the brain.

Personalised drug response

  • Semaglutide is effective for about 3/4 of participants, but only if treatment is maintained — weight is regained during an off-treatment extension phase after the ~68-week treatment phase.
  • Setmelanotide (MC4R agonist) is effective for patients whose obesity is due to POMC deficiency, but not effective for patients with PCSK1 or LEPR deficiency — response is genotype-specific.

Surgical management of obesity

Four bariatric surgery types:

  • A. Vertical sleeve gastrectomy — stomach reshaped into a narrow tube (“gastric sleeve”); the larger lateral portion is resected.
  • B. Roux-en-Y gastric bypass (RYGB) — a small gastric pouch is created at the top of the stomach and connected directly to an “intestinal Roux limb”; most of the stomach and proximal intestine is bypassed.
  • C. One-anastomosis gastric bypass — a longer gastric pouch connects to intestine via an efferent limb; an afferent limb carries digestive juices around the bypassed stomach portion.
  • D. Single anastomosis duodeno-ileal bypass — the gastric sleeve/pouch retains a connection near the gallbladder; a biliopancreatic limb carries digestive juices and joins food flow at a “common channel” further along the intestine; the stomach is resected.

Outcomes (Swedish weight loss study; weight change % vs years of follow-up):

  • Control: stays near 0%, slight decline to about −2% by 20 years.
  • Banding: drops to about −20% at 1 year, rises back to about −14% by 20 years.
  • Vertical banded gastroplasty (VBG): drops to about −25% at 1 year, rises to about −18% by 20 years.
  • Gastric bypass (GBP): drops to about −32% at 1–2 years, stays lowest throughout, about −26% at 20 years.

Gastric bypass (GBP) produces the greatest and most durable weight loss of the surgical options, though all surgical groups show some regain from their nadir over time.

Gut peptide hormone correlate of RYGB (rat model, meal-induced release, Shin et al 2010):

  • Active GLP-1: RYGB rats show a sharp post-meal rise to ~220 pg/ml at 10–20 min, remaining markedly elevated (~110 pg/ml) through 90 min; sham-operated and lean controls stay low and flat (~30–50 pg/ml).
  • PYY: RYGB rats show an elevated baseline (~120–130 pg/ml, significant even pre-meal), rising further to ~230 pg/ml at 20 min and remaining elevated (~150 pg/ml) at 90 min; sham and lean controls stay low and flat.
  • This demonstrates that RYGB greatly increases meal-induced release of the satiety gut hormones GLP-1 and PYY compared to sham-operated or lean controls.

Societal framing of obesity

  • Not everyone needs clinical treatment: treatment intensity is matched to increasing BMI/obesity severity along a spectrum — behavioural approaches, then drugs, then surgery (Dowsett and Yeo, TIMM 2023).
  • “The new challenge to society: stop blaming the patient”:
    • Individuals differ in genetic predisposition to obesity and may find it extraordinarily difficult to resist biological drives to eat.
    • People have different body weight “set points”, potentially altered by neonatal programming or by the environment during childhood/adolescence; more research is required to clarify this.
    • Framed as “a war on obesity, not on the obese.”
  • Closing quotation (Jeffrey Friedman, Science, 2003): the approach to the obesity epidemic should identify the molecular components of the body-weight regulatory system, define what differs between lean and obese subjects, and elucidate how environmental and developmental factors alter this system’s function — a foundation needed for rational therapies. Obesity is not a personal failing; those with obesity are fighting “a battle against biology.”

Self-test

  1. What two contrasting explanations for obesity are presented, and what factors fall under each?
  2. What calculation is used to demonstrate that body weight is under precise physiological control?
  3. What proportion of energy expenditure is voluntary?
  4. According to the cited BBC News study, what is the annual chance of a man vs a woman returning to normal weight after becoming obese?
  5. State the thrifty genotype hypothesis and who proposed it.
  6. Name four animal models of single-gene obesity mentioned in the lecture.
  7. Describe the homeostatic/allostatic signalling loop linking food intake, the GI tract, adipose tissue, and the brain.
  8. Why is voluntarily eating less described as “fighting against biology”?
  9. Describe the four parabiosis pairings and their outcomes, and explain what each outcome implies about the obese animal’s signal versus its receptor.
  10. Which parabiosis outcome specifically identified the missing factor in ob/ob mice, and what was that factor?
  11. What effect did leptin treatment have on the leptin-deficient boy described in the lecture (age, weight before and after)?
  12. List the seven peripheral hormones covered, their tissue of origin, and their direction of effect on intake or BMR.
  13. What is the “ileal brake”?
  14. In the arcuate nucleus, what are the two first-order neuron populations, and what is the net effect of each on food intake?
  15. For each of ghrelin, PYY, and leptin/insulin, state its effect on AgRP/NPY neurons (and, for leptin/insulin, on POMC neurons).
  16. Besides projecting to second-order neurons, what other signal do both AgRP/NPY and POMC neurons send, and to where?
  17. How do CCK and other GI/liver afferents reach the NTS?
  18. Define POMC, AgRP, and NPY.
  19. Why do most obese humans have elevated, rather than deficient, leptin levels?
  20. List the three proposed mechanisms of leptin resistance.
  21. Describe the five ordered steps of the leptin JAK2/STAT3 signalling pathway, from receptor binding to gene transcription.
  22. How does chronically elevated leptin lead to leptin resistance via SOCS-3?
  23. Which two of the approved obesity medications listed act via the GLP-1 receptor, and what weight loss percentages are given for each?
  24. Why is phentermine’s long-term weight loss data unavailable?
  25. Compare the reported weight-loss efficacy of semaglutide, tirzepatide, and retatrutide, and state their respective hormone targets.
  26. In the tirzepatide mechanism diagram, what are the three downstream target organs, and what effect is produced at each?
  27. Why can systemic GLP-1 agonists act centrally on POMC neurons when normal gut-derived GLP-1 largely does not reach the brain?
  28. Is GLP-1’s central action required for its weight-loss effect, its glucose-lowering effect, or both?
  29. Which patients respond well to setmelanotide, and which do not, despite also having genetic obesity?
  30. Name the four bariatric surgery types described and give one distinguishing anatomical feature of each.
  31. According to the Swedish weight loss study, which surgical/banding option produced the greatest and most durable weight loss at 20 years?
  32. Describe how RYGB alters meal-induced GLP-1 and PYY release compared to sham-operated or lean controls, and what this demonstrates.
  33. According to Dowsett and Yeo, how is treatment intensity for obesity matched to BMI?
  34. What is meant by “stop blaming the patient,” and what two contributing factors to individual variation in obesity are given?
  35. Close the note. From memory, write out the full ARC/NTS appetite circuit (first-order neurons, their peripheral inputs, second-order projections, and satiety pathway to the NTS), then check against Central integration the arcuate nucleus (ARC) circuit.

Answers

Graph View

Backlinks