What it is
AOD-9604 is a synthetic 16-amino-acid cyclic peptide (sequence Tyr-Leu-Arg-Ile-Val-Gln-Cys-Arg-Ser-Val-Glu-Gly-Ser-Cys-Gly-Phe, molecular formula C₇₈H₁₂₃N₂₃O₂₃S₂, CAS 221231-10-3) corresponding to the C-terminal hGH(177-191) lipolytic domain with an added N-terminal tyrosine substituting the native Phe¹⁷⁶. An intramolecular Cys7-Cys14 disulfide bond constrains the molecule into a cyclic conformation that mimics the three-dimensional architecture of the lipolytic region within intact growth hormone. The tyrosine addition improves proteolytic stability and is the basis for the 'AOD-9604' designation (Anti-Obesity Drug 9604, distinguishing it from the earlier 'AOD-9401' construct). The molecule was designed by Frank Ng and colleagues at Monash University in the 1990s and developed clinically by Metabolic Pharmaceuticals (later Calzada Ltd) in Melbourne. Unlike full-length hGH, AOD-9604 does not bind or activate the classical hGH receptor, does not raise IGF-1, does not activate JAK-STAT signaling, and does not drive linear growth or muscle hypertrophy — the design intent was to isolate the lipolytic domain from the anabolic and diabetogenic domains. After the 2007 Phase 2b efficacy failure the asset was repositioned as a nutraceutical ingredient under a self-affirmed GRAS framework (Moré 2014). It has never been FDA-approved for any human therapeutic use.
In plain English
AOD-9604 is a lab-made chain of 16 amino acids (the building blocks of proteins). It copies the tail end of human growth hormone — the piece scientists think is responsible for burning fat. One small change was added to make it more stable. A chemical bond folds the chain into a loop that mimics the shape of that piece inside the real hormone. Researchers at Monash University in Australia designed it in the 1990s. A company called Metabolic Pharmaceuticals then developed it. Unlike real growth hormone, AOD-9604 does not attach to the main growth-hormone receiver. It does not raise the usual blood growth signal. It does not make muscles or bones grow. Those effects were deliberately designed out. After the big 2007 weight-loss trial failed, the company relabeled it as a food ingredient instead of a drug. It has never been FDA-approved for any medical use.
How it works
- 01
β3-adrenergic receptor upregulation in adipose tissue
The primary mechanism characterized in vivo is restoration and sensitization of β3-adrenergic receptor (β3-AR) expression in adipocytes. Heffernan 2001 (Endocrinology 142(12):5182-5189) showed that both full-length hGH and AOD-9604 increased β3-AR mRNA in obese mouse adipose tissue to levels comparable with those of lean mice after 14 days of chronic intraperitoneal administration, and that β3-AR knockout animals were completely unresponsive to the lipolytic effect of AOD-9604 — a loss-of-function confirmation that is the strongest single piece of mechanistic evidence in the program. Downstream, β3-AR activation couples through Gαs–adenylyl cyclase–cAMP–PKA to phosphorylate hormone-sensitive lipase (HSL) and perilipin, mobilizing triglycerides from the lipid droplet.
In plain English
It turns up a fat-burning receiver on fat cells
Fat cells have a little "release button" on the outside. When something presses it, the cell lets stored fat out so the body can burn it. Obese mice have too few of these buttons. In a 2001 study, 14 days of AOD-9604 brought their button count back to lean-mouse levels. Researchers then tested the drug in mice bred without any of these buttons. It did nothing. That is strong proof the drug works through that one button. It is the most convincing piece of the whole research program.
- 02
Absence of classical hGH receptor engagement
A design-defining feature of AOD-9604 is what it does NOT do. Ng 2000 (Horm Res) and subsequent work confirmed that AOD-9604 does not bind or activate the hGH receptor, does not raise circulating IGF-1, does not activate hepatic JAK2-STAT5 signaling, and does not produce the insulin-resistance signal seen with exogenous GH in obese rodents. This receptor-negative profile distinguishes AOD-9604 from every GHRH analog (sermorelin, tesamorelin, CJC-1295) and from the ghrelin-receptor secretagogues (ipamorelin, MK-677) — all of which work by raising endogenous GH and IGF-1. The trade-off is that AOD-9604 also cannot borrow any of the well-characterized GH/IGF-1 axis efficacy evidence.
In plain English
It skips the normal growth-hormone system (on purpose)
A big part of the design is what AOD-9604 does NOT do. It does not attach to the regular growth-hormone receiver. It does not raise the usual growth-hormone signal in the blood. It does not cause the blood-sugar problems that real growth hormone causes in obese animals. That was the whole point — the designers wanted the fat-burning effect without the side effects. The trade-off: this drug cannot borrow any of the well-known benefits of raising growth hormone, because it does not raise growth hormone.
- 03
Enhanced whole-body fat oxidation (rodent)
Heffernan 2001 (Int J Obes Relat Metab Disord 25(10):1442-1449) used indirect calorimetry in obese mice and reported an approximately 216% increase in whole-body fat oxidation on chronic AOD-9604 relative to vehicle, with a corresponding shift in respiratory exchange ratio toward lipid substrate utilization. The mechanism is proposed to involve CPT-1-mediated mitochondrial fatty-acid import downstream of β3-AR-driven lipolysis, but the specific CPT-1, UCP, or PGC-1α findings have not been independently replicated outside the original program. No human indirect-calorimetry RCT exists.
In plain English
In mice, the body burns a lot more fat for energy
Obese mice given AOD-9604 burned about 216% more fat than mice given nothing. Their bodies switched from running on carbs to running on fat. The exact cell-level steps after that have never been confirmed by an outside lab. No one has measured this in people.
- 04
Proposed anti-lipogenic effect via acetyl-CoA carboxylase
Company-sponsored preclinical work described by Moré 2014 (J Endocrinol Metab) and in the sponsor dossiers proposed that AOD-9604 additionally inhibits acetyl-CoA carboxylase (ACC), the rate-limiting enzyme for de novo lipogenesis, yielding a dual lipolytic / anti-lipogenic effect. This mechanism is less well characterized in independent peer-reviewed literature than the β3-AR arm and should be regarded as hypothesis-level rather than established.
In plain English
It might also slow down new fat creation (unconfirmed)
Company researchers also proposed that AOD-9604 blocks a key enzyme the body uses to build new fat from scratch. If true, the drug would both burn old fat and slow new fat from forming. But this claim comes from company reports, not outside studies. Treat it as an idea, not a proven fact.
- 05
Pharmacokinetics and route of administration
Moré 2014 summarized pig PK studies after oral and intravenous administration and supported oral bioavailability as the basis for nutraceutical framing. Human PK from the 2000–2007 clinical program was not published as a standalone manuscript — only sponsor summaries describe it. Esposito 2014 (PMID 25208511) characterized the in-vitro metabolic profile of AOD-9604 for anti-doping LC-HRMS detection, identifying the major proteolytic cleavage products relevant to urine- and plasma-based WADA screening.
In plain English
How it's absorbed and how long it lasts
Pig studies showed AOD-9604 can be absorbed when taken by mouth. That is what the company used to justify relabeling it as a food ingredient. How it behaves in humans was never published in a standalone paper. Only company summaries describe it. A 2014 lab study mapped how the drug breaks apart, so anti-doping labs know what to look for in athletes' blood and urine.
- 06
What is NOT known about the mechanism
The loss-of-function β3-AR data is strong, but the molecular-level binding partner (receptor or surface protein) that AOD-9604 itself engages to upregulate β3-AR mRNA has not been identified — the β3-AR is a downstream effector, not the primary target. Human pharmacokinetics, distribution, and adipose-tissue bioavailability are not published in peer-reviewed form. The proposed ACC inhibition remains poorly characterized. Whether the 2007 Phase 2b efficacy failure reflects a receptor-desensitization / tachyphylaxis phenomenon, a dose-response inversion (the 1 mg > 10 mg paradox in the 12-week trial suggests non-monotonic pharmacology), or simply a mechanism that does not translate from mice to humans at clinically feasible exposures is unresolved.
In plain English
What we still don't know
The mouse work is strong, but it only shows what happens several steps down the chain. No one knows the exact spot on a cell that AOD-9604 touches FIRST to start the chain reaction. How the human body processes it — how much reaches fat tissue, how fast it clears — has never been published in a peer-reviewed paper. The "blocks new fat" claim is poorly tested. No one knows why the 2007 human trial failed. Does the body get used to the drug? Do higher doses really work worse than lower ones (the trial showed 10 mg did worse than 1 mg)? Or does the mouse effect just not translate to people? All still open questions.