AOD-9604: Structure, Lipolytic Mechanism and Analytical Detection

AOD-9604 differs from the growth hormone fragment hGH 176-191 by one residue. Phenylalanine 176 is replaced by tyrosine, so the analogue carries a single extra oxygen atom and the formula C78H123N23O23S2 against C78H123N23O22S2 for the unmodified fragment. Everything from leucine 177 to phenylalanine 191, including the disulfide loop, is unchanged.

That one atom is a useful integrity check. PubChem CID 71300630 returns C78H123N23O23S2 at 1815.1 g/mol with a monoisotopic mass of 1813.860 Da, and a preparation whose mass spectrum centres 16 Da lower is the unmodified fragment rather than the tyrosine-bearing analogue. The two are sold under overlapping names, and mass alone distinguishes them.

Conformationally the molecule is small and constrained. Ogru and colleagues solved the solution structure of the cyclic domain by nuclear magnetic resonance in 2000 and reported defined beta-turn geometry in the disulfide-closed region, closely matching the corresponding segment of the published growth hormone crystal structure. The isolated domain therefore holds a shape it also adopts inside the intact hormone.

AOD-9604 traces to a mapping effort that located the lipolytic activity of human growth hormone in its carboxy-terminal tail rather than the receptor-binding helices. Natera, Jiang and Ng reported in 1994 that a synthetic hGH 177-191 peptide inhibited lipogenesis in adipose tissue of C57BL/6J ob/ob mice, isolating the domain from the intact hormone.

The mapping was then extended across species and preparations. Ogru and colleagues found that the cyclic domain stimulated lipolysis and inhibited lipogenesis in vitro in rodent, porcine and human adipose tissue, which established that the activity was not a rodent artefact. Heffernan and colleagues reported in 2000 that the closely related analogue AOD-9401, given orally to ob/ob mice for 30 days, reduced lipogenic activity and raised lipolytic activity in adipose tissue examined ex vivo, with the same directional effects reproduced in isolated human adipose tissue.

AOD-9604 is the version of that domain carried forward. Ng and colleagues reported in Hormone Research in 2000 that Zucker rats given the peptide orally at 500 micrograms per kilogram over 19 days showed increased adipose lipolytic activity with no adverse change in insulin sensitivity, which is the point of divergence from chronic exposure to the intact hormone.

AOD-9604 and its close congener AOD9401 act on two enzymes that sit on opposite sides of adipocyte lipid handling. Ng and colleagues reported in the Journal of Molecular Endocrinology in 2000 that the synthesised domain stimulated hormone-sensitive lipase and inhibited acetyl-CoA carboxylase in isolated rat adipose tissue, matching the intact hormone.

The same work reported a morphological readout alongside the enzyme data. Mean adipocyte diameter in Zucker rats exposed for 20 days fell from 110 to 80 micrometres, and the peptide reproduced the effect of the intact hormone on diacylglycerol release in adipocytes. Acetyl-CoA carboxylase is the committed step of fatty acid synthesis and hormone-sensitive lipase hydrolyses stored triacylglycerol, so inhibition of one with stimulation of the other describes a coordinated shift rather than a single-enzyme effect.

A receptor requirement was established genetically. Heffernan and colleagues reported in Endocrinology in 2001 that both the intact hormone and AOD9604 raised beta-3 adrenergic receptor RNA in adipose tissue over 14 days, restoring the repressed levels seen in ob/ob animals toward those of lean controls, and that the chronic lipid effects were absent in beta-3 adrenergic receptor knockout mice. Knockout dependency is the strongest form of that claim available in the AOD literature, since it removes the candidate pathway rather than blocking it pharmacologically.

AOD-9604 separates from intact growth hormone at the receptor itself. Heffernan and colleagues reported in 2001 that the fragment does not compete for the growth hormone receptor, and that chronic exposure produced neither hyperglycaemia nor reduced insulin secretion, whereas the intact hormone produced both in the same experimental animals.

Human data point the same way. Stier, Vos and Kenley pooled six randomised, double-blind, placebo-controlled trials in the Journal of Endocrinology and Metabolism in 2013 and reported no effect on serum IGF-1, oral glucose tolerance results showing no negative change in carbohydrate metabolism, and no anti-AOD9604 antibodies in the samples assayed for immunogenicity. Tolerability in those trials was described as indistinguishable from placebo.

Recent secondary literature is not consistent on this point, and the discrepancy is worth flagging. The 2026 orthopaedic review by Rahman, Lee and Seeds groups AOD-9604 with growth hormone secretagogues described as acting through IGF-1 signalling, which sits against the primary receptor-competition and IGF-1 measurements above. Where a review assigns a mechanism by class rather than by measurement, the underlying assay data should be checked before the assignment is carried forward.

AOD-9604 is measured by liquid chromatography coupled to mass spectrometry. Cox and colleagues published a validated urine method in Drug Testing and Analysis in 2015 reaching a limit of detection of 50 pg/mL with 62 percent extraction recovery, and identified six candidate metabolites generated by incubation in serum and urine.

One of those fragments carries the analytical work. The nonapeptide CRSVEGSCG, produced by loss of the N-terminal and C-terminal residues while the disulfide loop survives, proved markedly more persistent than either the parent peptide or the other metabolites, so it became the preferred target for extending the detection window. Persistence tracks the constrained cyclic core, which is the part of the molecule least accessible to exopeptidases.

Circulating persistence of the intact peptide is short. Moré and Kenley reported a serum half-life of roughly 4 minutes in vitro, with intact peptide undetectable 56 minutes after spiking, and roughly 3 minutes in pigs following vascular delivery, alongside an oral availability estimate near 40 percent in rats derived from radioactivity distribution. The same report described no genotoxic activity in the bacterial reverse mutation assay up to 2,000 micrograms per plate.

Matrix handling matters as much as instrument sensitivity. Mazzarino and colleagues reported in The Analyst in 2026 that across a 54-compound prohibited panel, analytes held within 15 percent variation for two months at minus 20 degrees Celsius, while several degraded extensively in serum and plasma at raised temperatures. In dried matrices every compound stayed detectable for the full study, which is why dried blood spots are attractive where cold chain cannot be guaranteed.

AOD-9604 appears by name in section S2 of the World Anti-Doping Agency Prohibited List, among growth hormone fragments alongside hGH 176-191. Substances in S2 are prohibited at all times, in and out of competition, which is why anti-doping laboratories have invested in detection chemistry for a peptide with no approved medicinal use.

Detection required purpose-built chemistry because the existing growth hormone test does not see it. Orlovius, Thomas, Schaenzer and Thevis reported in 2013 that AOD-9604 does not influence the World Anti-Doping Agency hGH isoform immunoassay, so a fragment-specific mass spectrometric method was necessary rather than optional. Vanhee and colleagues published a characterisation of AOD9604 in preparations seized by Belgian authorities in 2014, an early record of the material circulating outside regulated supply.

Secondary literature in 2026 places the compound in a wider survey. Mendias and Awan catalogued AOD-9604 among unapproved peptides in Sports Medicine, concluding that animal-model results across the class outrun the human safety record. Dominikowski and colleagues listed it as a growth hormone fragment in a Frontiers in Endocrinology review, and Coutinho and colleagues covered the 176-191 class in a critical review of peptide use in sport, noting that unregulated supply frequently yields mislabelled or contaminated material.

AOD-9604 is supplied as a lyophilized solid and carries one intramolecular disulfide between cysteine 7 and cysteine 14 of the 16-residue chain, corresponding to Cys182 and Cys189 of the parent hormone. That bond, rather than oxidation of a free thiol, is the structural feature that governs handling.

Disulfide chemistry sets the first liability. Reducing agents, residual thiols in glassware and alkaline buffers all favour ring opening, and a linearised chain has lost the beta-turn geometry Ogru and colleagues associated with activity. Purity work should resolve the reduced and scrambled species explicitly instead of reporting a single main-peak percentage, because a linear isomer co-elutes closely and shares the parent mass to within two hydrogen atoms.

Solution behaviour follows from the computed descriptors. An XLogP of minus 4.8 and a topological polar surface area of 815 square angstroms describe a strongly hydrophilic molecule that dissolves readily in aqueous buffer and will not cross lipid bilayers passively, so cell-uptake work needs an explicit uptake control rather than an assumption of free entry.

Stability data come from two directions. Moré and Kenley reported an in vitro serum half-life near 4 minutes with intact peptide undetectable by 56 minutes, so plasma sampling designs built around slower peptides will miss the parent entirely. Mazzarino and colleagues reported that panel analytes held within 15 percent for two months at minus 20 degrees Celsius while degrading in liquid serum and plasma at raised temperatures, which favours frozen aliquots or dried matrices over ambient shipment of liquid samples.