AOD-9604 10mg Peptide

AOD-9604 is a 16 amino acid peptide built from the C-terminal end of human growth hormone. The name is an internal development code from Metabolic Pharmaceuticals in Melbourne, where the letters stand for anti-obesity drug. Chemically it is Tyr-hGH177-191: residues 177 to 191 of the mature 191 amino acid hormone, with a tyrosine attached at the front end. The tyrosine was added so the peptide could be radioiodinated for tracking studies, and it stayed in the final molecule.

The idea behind it came from Frank Ng's group at Monash University, who had spent two decades arguing that growth hormone behaves like a pro-hormone whose separate biological activities live in separate parts of the chain. If the lipolytic activity sits in the C-terminal loop and the growth-promoting, IGF-1 raising, glucose-disturbing activity sits elsewhere, then a short C-terminal fragment should move fat without moving anything else. AOD-9604 was the molecule built to test that proposition.

It is one of very few research peptides with a genuine human trial record: six randomised, placebo-controlled studies enrolling roughly 900 adults between 2001 and 2007, all of them using oral capsules rather than injection. The obesity programme ended in 2007 when the largest of those studies failed its primary endpoint. The compound then had two afterlives, one as a self-affirmed food ingredient and one as Lateral Pharma's clinical candidate LAT8881. AOD-9604 is supplied here as a lyophilised reference material for in vitro laboratory research only.

AOD-9604 does not work through the growth hormone receptor. Heffernan and colleagues tested this directly in 2001 using BaF-B03 cells transfected with the human GH receptor: AOD-9604 did not compete with radiolabelled hGH for receptor binding and did not drive the cell proliferation that hGH drives in the same assay. That single result is the reason the peptide leaves IGF-1 alone in every study that has measured it, and it is the mechanistic difference that separates the fragment from the parent hormone.

What it does do, in obese rodents, is shift substrate use toward fat. Chronic administration raised whole-body fat oxidation and plasma glycerol, an index of lipolysis, without raising plasma glucose or suppressing insulin. In adipose tissue from treated Zucker rats, lipolytic activity was measurably higher than in controls. The peptide also appears to suppress lipogenesis, so the net effect in these models is less new triglyceride laid down and more of the existing store mobilised.

The beta-3 adrenergic receptor is where most descriptions of this peptide go wrong. AOD-9604 does raise beta-3 adrenergic receptor messenger RNA in obese mouse adipose tissue, restoring it toward the levels seen in lean mice, and that upregulation plausibly increases how readily the tissue responds to catecholamines. But the receptor is not the target. In beta-3 adrenergic receptor knockout mice, chronic treatment with either hGH or AOD-9604 failed to produce the weight and lipolysis changes seen in wild-type animals, while an acute dose still raised energy expenditure and fat oxidation in those same knockout animals. The authors concluded that the lipolytic action is not mediated directly through the beta-3 receptor. A receptor for AOD-9604 has never been identified.

The work began at Monash University, where Frank Ng had been dissecting growth hormone since the 1970s. His group's position was that the intact hormone is a pro-hormone carrying several activities in separate regions of the chain, and that the lipolytic activity lives at the C-terminus. Synthetic fragments of that region reproduced the fat-mobilising effect of the whole hormone in adipose tissue assays, which suggested the fat effect could be taken without the growth effect.

AOD-9604 was the lead molecule out of that programme. The first published metabolic study appeared in Hormone Research in 2000: obese Zucker rats given 500 micrograms per kilogram body weight orally for 19 days gained 15.8 grams against 35.6 grams in controls, adipose lipolytic activity rose, and euglycaemic clamp testing showed no loss of insulin sensitivity. Two 2001 papers filled in the mechanism, one in the International Journal of Obesity establishing that the effect runs through fat oxidation and lipolysis without hGH receptor engagement, one in Endocrinology testing the beta-3 adrenergic hypothesis in knockout mice.

Metabolic Pharmaceuticals took it into humans from 2001. By 2004 the company was reporting a positive 12 week result, and by 2007 the programme was over. The intellectual property moved to Calzada, then to PolyNovo Biomaterials, and eventually to Lateral Pharma, which renamed it LAT8881 and pointed it at pain rather than fat.

Six randomised, double-blind, placebo-controlled studies were run, and Stier and colleagues published a consolidated safety analysis of all six in 2013. METAOD001 was an intravenous dose escalation from 25 to 400 micrograms per kilogram in 15 healthy men. METAOD002 and METAOD003 were Latin square crossover studies in obese men, the first intravenous at 25, 50 and 100 micrograms per kilogram in 23 subjects, the second oral at 9, 27 and 54 mg in 17 subjects. METAOD004 gave the same three oral strengths to 36 obese men for seven days.

METAOD005 is the study everything else is built on. Three hundred obese adults with a body mass index at or above 35, roughly evenly split by sex, took a placebo run-in and then 12 weeks of oral capsules at 1, 5, 10, 20 or 30 mg, 50 subjects per arm. The reported outcome was counterintuitive: the lowest arm did best. The 1 mg group lost about 2.8 kilograms over 12 weeks against about 0.8 kilograms on placebo, with a small improvement in cholesterol profile and fewer subjects meeting criteria for impaired glucose tolerance. The higher strengths did not do better, which is a pattern worth noting whenever a dose-response curve is assumed to be monotonic.

METAOD006, called the OPTIONS study, was designed to confirm it and did not. It recruited 536 subjects and analysed 502 obese adults with body mass index between 30 and 45, gave a four week single-blind placebo run-in, then 24 weeks of oral capsules at 0.25, 0.5 or 1 mg against placebo, then a four week follow-up. No arm separated from placebo on weight at either the 12 week primary endpoint or at 24 weeks. Metabolic Pharmaceuticals stopped development for obesity in 2007. Nothing about the safety database changed: IGF-1 did not move (p = 0.51 at 12 weeks, p = 0.76 at 24 weeks), oral glucose tolerance testing showed no trend in any arm, no anti-AOD9604 antibodies were found, and no serious adverse event was attributed to the compound. The trials predate mandatory registration, so none of them has a ClinicalTrials.gov record.

Two details are routinely dropped when this history is retold. Every human efficacy datapoint came from an oral capsule, not an injection, so there is no human efficacy evidence at all for the parenteral route the compound is now associated with. And the failed study was a Phase 2b, not a Phase 3: the compound never reached a registrational trial.

The most misrepresented fact about this peptide is its United States food status. In 2014 an independent panel of qualified experts, assembled on behalf of the sponsor, reviewed the toxicology and human safety package and concluded that AOD-9604 is generally recognised as safe under its intended conditions of use as a food ingredient, at intakes up to 1 mg per person daily. That is a self-affirmed GRAS determination. The conclusion belongs to the expert panel, not to any regulator.

A self-affirmed determination is not the same thing as a GRAS notice. Under the voluntary notification programme a sponsor may submit its determination to FDA and receive a letter saying the agency has no questions. FDA's public GRAS Notice Inventory carries no notice for this peptide, which means no agency review took place and no no-questions letter exists. AOD-9604 has never been approved by FDA as a drug for any indication, and GRAS is a safety framework for food ingredients that says nothing about efficacy.

The safety package the panel relied on is real and unusually large for a compound in this catalogue: six human trials with roughly 900 subjects, chronic oral toxicology in rats at up to 100 mg per kilogram daily for 26 weeks, and in cynomolgus monkeys at up to 50 mg per kilogram daily for 39 weeks, with no genotoxic or toxicological findings reported and no immunogenic response. What that package cannot do is establish that the compound produces an effect, which is precisely what METAOD006 tested and did not find.

These two are sold as different products and described as different molecules, and the actual difference is one hydroxyl group. The mature human growth hormone sequence, UniProt P01241 after the 26 residue signal peptide, ends VETFLRIVQCRSVEGSCGF. Counting back, residue 176 is phenylalanine and residue 177 is leucine. So the true native fragment 176-191 reads Phe-Leu-Arg-Ile-Val-Gln-Cys-Arg-Ser-Val-Glu-Gly-Ser-Cys-Gly-Phe. AOD-9604 takes residues 177 to 191 and puts a tyrosine in front instead, giving Tyr-Leu-Arg-Ile-Val-Gln-Cys-Arg-Ser-Val-Glu-Gly-Ser-Cys-Gly-Phe. Tyrosine is phenylalanine with a para hydroxyl. Position 1 is the only difference between the two chains.

This matters for identity checks. Much of the material sold as HGH Fragment 176-191 is specified with a tyrosine at position 1 and a molecular weight near 1,815, which is AOD-9604, not the native fragment. The native fragment is 1,799.1 g/mol in its oxidised form because it lacks that oxygen. A certificate of analysis showing 1,815 g/mol and a tyrosine N-terminus describes AOD-9604 whatever the label says. Mass spectrometry separates them by 16 daltons.

A second common claim is that AOD-9604 is the stabilised, disulfide-bridged version while fragment 176-191 is a fragile linear peptide. Both chains contain the same two cysteines, at positions 7 and 14 of the fragment, corresponding to Cys182 and Cys189 of the hormone, and both form the same intramolecular loop. Neither is linear. The tyrosine was introduced to allow radioiodination for pharmacokinetic tracking, and whatever stability advantage it confers is modest next to that.

The asymmetry that is real is the evidence base. AOD-9604 carries the toxicology programme, the six human trials, the published pharmacokinetics and the analytical characterisation. The native 176-191 fragment has almost no human data of its own, and most claims made for it are borrowed from AOD-9604 studies. Volta lists both, as aod-9604-10mg and hgh-fragment-176-191-5mg, and the certificate of analysis is the only way to tell what is actually in a vial.

The second research line has nothing to do with fat. Kwon and Park reported in 2015 that AOD-9604 improved outcomes in a collagenase-induced knee osteoarthritis model in 32 mature New Zealand white rabbits. Each knee received 2 mg type II collagenase twice, then weekly ultrasound-guided intra-articular treatment: saline, 6 mg hyaluronic acid, 0.25 mg AOD-9604, or the peptide and hyaluronic acid together, for four to seven weeks after the first collagenase dose.

Gross morphological and modified Mankin histopathology scores were worst in the saline group and better in all three treatment groups. The combination group scored better than either single agent, which the authors read as complementary rather than redundant activity: hyaluronic acid protecting existing chondrocytes while the peptide supports matrix regrowth. Lameness ran a mean of 25 days on saline, 15 on hyaluronic acid, 16 on peptide alone and 11 on the combination.

The proposed mechanism is drawn from unpublished sponsor work the authors cite: promotion of proteoglycan and collagen production in isolated bovine chondrocytes, and differentiation of adipose-derived mesenchymal stem cells toward bone. Because AOD-9604 does not raise IGF-1, a cartilage effect without a systemic growth signal is at least coherent. It is a single animal study with an n of 8 per group, no independent replication has been published, and no human joint trial has been run.

The molecule did not stop being developed in 2007. Lateral Pharma, an Australian company, acquired it, renamed it LAT8881 and ran it in registered clinical trials for pain, on the hypothesis that the same C-terminal fragment activity that touches adipose and cartilage biology also modulates nociceptive signalling.

Three studies are on ClinicalTrials.gov, all completed. NCT03865953 was a Phase 2 study of oral LAT8881 in 53 subjects with neuropathic pain from diabetic peripheral neuropathy or post-herpetic neuralgia, running from April 2019 to May 2020. NCT04153409 was a Phase 2 proof-of-concept study of oral LAT8881 in 21 subjects with acute migraine. NCT05298306 was a two-part Phase 1 proof-of-concept study in 26 subjects with lumbar radicular pain, this one intravenous at 0.8, 1.2 and 1.8 mg per kilogram, running from May 2022 to June 2023.

These are the only registered clinical trials of the compound in existence, and they carry a different name from the one on the vial. Anyone searching trial registries for AOD-9604 finds nothing at all, which is why the obesity programme is often described as the whole story.

The molecule has three features that determine how it is handled in a laboratory. The Cys7 to Cys14 disulfide can oxidise further to sulfoxide and sulfone species, which shifts mass and can alter activity. The tyrosine absorbs at approximately 280 nanometres, making the peptide quantifiable by ultraviolet absorbance and also susceptible to radical-mediated photo-oxidation under prolonged ultraviolet exposure. And the peptide backbone hydrolyses in solution, which is why the lyophilised state is so much more stable than any reconstituted one. Standard practice for research material is storage of the dry powder at -20 degrees Celsius, protected from light and moisture with desiccant, with aqueous stocks held at 2 to 8 degrees Celsius, aliquoted for single use, and freeze-thaw cycling minimised.

Proteolysis in biological matrices is fast and directional. Metabolism studies in pigs showed sequential removal of amino acids from the N-terminus, with the species missing two and three residues dominating. The plasma half-life after intravenous dosing in pigs was around 3 minutes, against roughly 21 minutes for intact human growth hormone measured the same way. Cox and colleagues incubated the peptide in human serum and urine in 2014 and found six metabolites, of which one, the nonapeptide CRSVEGSCG, is substantially more stable than either the other fragments or the parent.

That metabolite is what anti-doping laboratories look for. AOD-9604 is named on the WADA Prohibited List among growth hormone fragments, and a validated urine method reaches a limit of detection of 50 picograms per millilitre with 62 percent recovery. It is worth knowing that the standard WADA hGH isoform immunoassay does not respond to AOD-9604 at all, a point Orlovius and colleagues established in 2013, so the routine growth hormone test would not flag it and a dedicated method is required.

Identity confirmation is not academic here. Vanhee and colleagues at the Belgian Scientific Institute of Public Health published a case report in 2014 on unlabelled peptide preparations seized by customs, working out what was in the vials by mass spectrometry and nuclear magnetic resonance. Two cautions apply when cross-checking this compound in public databases. PubChem CID 71300630 carries the synonym Tyr-somatostatin (177-191), which is a database error: somatostatin is 14 or 28 residues long and has no residue 177. And a certificate of analysis reporting a molecular weight near 1,817 rather than 1,815 is describing reduced material with the disulfide bond open.