SR-9009 (Stenabolic)

Chemical identity

The name on the label is not the name in the database. A query for stenabolic alone returns no PubChem record; the string survives only as the depositor synonym Stenabolic (SR9009), attached to CID 57394020. Under the code name SR9009 the compound resolves cleanly to that single record, which carries CAS 1379686-30-2, the formula C20H24ClN3O4S, a molecular weight of 437.9, InChIKey MMJJNHOIVCGAAP-UHFFFAOYSA-N, UNII X5DCA09N30, ChEMBL identifier CHEMBL1961796 and DrugBank accession DB14013. The IUPAC name is ethyl 3-[[(4-chlorophenyl)methyl-[(5-nitrothiophen-2-yl)methyl]amino]methyl]pyrrolidine-1-carboxylate. The FDA Global Substance Registration System holds the same UNII under the preferred name SR-9009, with the identical formula and a mass of 437.942, and classes the substance as a chemical rather than a protein or peptide.

Structurally this is a pyrrolidine ethyl carbamate joined through a tertiary amine to a 4-chlorobenzyl group and a 5-nitrothiophene. It is neither a peptide nor a steroid, and it has no reported androgen-receptor pharmacology. The molecule is nonetheless sold and discussed inside the selective androgen receptor modulator category, and the 2017 JAMA analysis of products sold under that heading names it among the compounds found in that market. The nitrothiophene ring is the part of the structure its own developers flagged: Kojetin and Burris wrote in 2014 that this group carries a potential toxicological liability, and that they had pursued a separate chemical series specifically to remove it.

Every published in vivo result came from an injection

Solt and colleagues dosed intraperitoneally at 100 mg per kg twice daily. Woldt and colleagues dosed intraperitoneally at 100 mg per kg for thirty days. Sulli and colleagues dosed intraperitoneally. Ni and colleagues dosed intraperitoneally for fourteen days. Stujanna and colleagues dosed intraperitoneally after coronary ligation. Slomnicki and colleagues dosed intraperitoneally after spinal contusion. Across the retrievable in vivo literature the route does not change. No published study administering this compound by mouth to any species was located, and none of the papers describing systemic administration reports an oral arm, an oral comparator or an absorption measurement.

The pharmacokinetic record for the molecule itself is thinner than the volume of citation suggests. Solt and colleagues state only that the compounds displayed reasonable plasma exposure, referring to a supplementary figure. Kojetin and Burris describe four GlaxoSmithKline analogues as orally bioavailable and note that three of them have pharmacokinetic properties similar to this compound, without printing its parameters; the tetrahydroisoquinoline series they developed afterwards is described as not orally bioavailable, with half-lives of roughly two hours by the intraperitoneal route. Wang and colleagues, reviewing the field in 2020, write that current synthetic REV-ERB ligands are cleared rapidly with half-lives under three hours, and call the resulting regimen of repeated daily injection unsatisfactory.

Material bearing the name reaches consumers as capsules and oral solutions. Nothing in the published record measures what happens when the compound is swallowed rather than injected into the peritoneal cavity, and no source located here reconciles the two. That gap sits underneath every quantitative claim carried over from a rodent experiment, because each of those numbers was produced by a route nobody is using.

The endurance result is a single panel in a single paper

Woldt and colleagues published in Nature Medicine in 2013. The paper is mostly about the receptor rather than the ligand: Rev-erb-alpha deficient mice, muscle-specific overexpression by adeno-associated virus, isolated fibres and C2C12 myotubes. One panel, Figure 6a, carries the pharmacological arm. Mice received 100 mg per kg intraperitoneally for thirty days, six animals per group, and were run to exhaustion on a treadmill after two days of acclimatisation and a separate maximal stress test one week earlier. The result sentence reads that mice treated with the compound ran significantly longer both in time and distance than vehicle-treated mice.

That sentence is the whole published human-readable finding. No percentage appears in the text or the figure legend. No p value is printed for the running result. The methods section names the strain and genotype for the knockout and overexpression arms but does not state the strain, sex or age of the animals in the pharmacological arm. Phrase searches of Europe PMC pairing the compound name with running distance and with time to exhaustion return this paper and, in the first case, one review that cites it. Nothing else in the indexed literature reports either measurement for this molecule.

The mitochondrial claim comes from the same paper by a different route. C2C12 cells were incubated with 5 micromolar of this compound, 5 micromolar of SR9011 or vehicle, four replicates per condition, and mitochondrial content was read by flow cytometry using green and red mitotracker stains. Handschin, reviewing exercise mimetics in 2016, adds that application of the compound does not alter muscle fibre-type distribution, and observes that the consequences of using it as an exercise mimetic in terms of human circadian rhythms would have to be carefully evaluated. That evaluation has not been published.

The 2019 result that unmade the mechanism

Dierickx and colleagues at the University of Pennsylvania built the experiment the field had not run. Earlier conditional deletion of REV-ERB-alpha had removed the DNA-binding domain in frame, producing a mutant protein rather than an absent one. They floxed exons three to five for an out-of-frame deletion, crossed the line to an existing REV-ERB-beta floxed model, and generated embryonic stem cells and hepatocytes with neither protein. Target-gene derepression confirmed loss of function. The compound was then tested in cells that had nothing for it to act on.

Viability, cell number and the proportion of cells in S phase all fell after two days at 10 micromolar, and they fell to similar extents in the double-knockout cells as in the floxed controls. SR9011 behaved the same way. In hepatocytes treated for eight hours, 431 genes were differentially expressed in controls; roughly 55 per cent of them moved in the same direction in cells lacking both receptors. Of the 193 genes that changed only in controls, 25 moved in the direction a genuine agonist would predict. The authors concluded that the effects of the compound cannot be used solely as a surrogate for REV-ERB activity.

Corroboration arrived from an unrelated group working on a different tissue. Xu and colleagues reported in 2022 that the compound was lethal to prostate cancer cell lines and inhibited the FOXM1 pathway, and that neither knockdown nor knockout of REV-ERB rescued that effect. Their analysis identified LXR-alpha as the receptor actually being activated. Two laboratories, two systems, one conclusion: the pharmacology attributed to this molecule is not the pharmacology its name describes. Dierickx and colleagues also note that both this compound and SR9011 contain known toxicophores, naming the nitrothiophene moiety.

What the metabolic experiments measured

Solt and colleagues started with twenty-week-old C57BL6 mice averaging 41 grams, kept on a diet of 20 per cent carbohydrate and 60 per cent fat for fourteen weeks. Dosing was 100 mg per kg intraperitoneally, twice daily, for thirty days, with six to ten animals per group. Reported changes were a 12 per cent fall in plasma triglycerides, 47 per cent in total cholesterol, 23 per cent in non-esterified fatty acids, 19 per cent in glucose, 80 per cent in leptin and 72 per cent in interleukin-6, with a 35 per cent downward trend in insulin. Food intake did not differ significantly between groups.

Weight loss carries a caveat the paper states plainly and the circulating version drops. Vehicle-treated controls lost weight too, which the authors attribute to the stress of handling and twice-daily injection, and they note that handling itself reduced food intake. The 60 per cent figure is the excess of treated weight loss over that already-falling baseline, not a comparison against untreated animals.

Energy expenditure belongs to the sibling compound. The comprehensive laboratory animal monitoring system experiment in that paper was run on SR9011, not on this molecule: a 5 per cent rise in oxygen consumption over ten days of twice-daily dosing, alongside a 15 per cent decrease in movement. Figure 3 is titled for SR9011 throughout. Results for this compound appear in a supplementary figure described only as similar. In the transcriptional assays it suppressed REV-ERB-driven reporters with a half-maximal inhibitory concentration of 670 nanomolar at REV-ERB-alpha and 800 nanomolar at REV-ERB-beta, and 710 nanomolar against a full-length Bmal1 reporter.

Findings that run the other way

Ni and colleagues gave 100 mg per kg intraperitoneally for fourteen days to two-month-old male wild-type and amyloid precursor protein knock-in mice, six per group. Microglia were activated, pro-inflammatory gene expression rose, and performance on the Y-maze and novel object recognition tests fell in the knock-in animals at P below 0.001. Locomotor activity was unchanged, which rules out the simplest confound. The finding is a cognitive decrement produced by the compound in a disease model, in the same dose range and by the same route as every favourable rodent result.

Slomnicki and colleagues published a null in Neuroscience Letters in 2026, with Burris among the authors. After moderate T9 contusion, mice received 100 mg per kg per day intraperitoneally as two 50 mg per kg doses for seven days. At three days there was less haematoma and lower expression of transcripts tied to barrier disruption, inflammation and cellular stress. Over six weeks of follow-up hindlimb recovery was unchanged, and white matter sparing at study completion showed no significant difference. Acute pathology moved; the outcome did not.

One paper has left the record. Huang and colleagues published on intervertebral disc degeneration in iScience in 2024; the journal retracted it in June 2026 after Elsevier's research integrity team identified undisclosed authorship changes between submission and revision, with two authors added and three removed. The editor stated a loss of confidence in the results and conclusions. The authors disagree with the retraction. No data-integrity finding is asserted in the notice, but the paper no longer stands. Set against all of this, Kojetin and Burris wrote in 2014 that no dedicated toxicological studies had been performed for any REV-ERB ligand, and no such study has been located since.

Sport, law and what is in circulation

The 2026 WADA Prohibited List names the molecule explicitly. Section S4.4.1, under metabolic modulators, reads in one bullet: Rev-erb-alpha agonists, e.g. SR9009, SR9011. The section header states that prohibited substances in classes S4.3 and S4.4 are non-Specified Substances, and section S4 is prohibited at all times, in and out of competition. No regulator has approved the compound as a medicine in any jurisdiction, and no marketing application anywhere was located.

ClinicalTrials.gov holds nothing. Queries on SR9009, SR-9009, stenabolic and SR9011 each return a total count of zero. A REV-ERB query returns eight records, all observational studies of clock-gene expression, none of which administers a REV-ERB ligand. The JAMA analysis states the position in one line: preclinical studies have occurred for the compound, but no human trials. Fourteen years after the first publication, that sentence still holds.

What circulates is not always what the label says. Van Wagoner and colleagues bought 44 products marketed as selective androgen receptor modulators; the single product labelled at 20 mg of this compound per capsule contained 1 to 5 mg of androstatrienedione per capsule and none of the labelled substance. Geldof and colleagues confirmed the compound in a product purchased over the internet, characterised eight human liver microsomal metabolites, then screened 1,511 stored doping control urine samples retrospectively and detected neither parent nor metabolite. The first adverse analytical finding anywhere came from horse racing, reported by Cutler and colleagues in 2021.

What is not known

Nothing is known about this compound in a person. No human trial is registered on ClinicalTrials.gov, none has been published, and no human pharmacokinetic, safety, endocrine or performance measurement exists. There is no published half-life, no Cmax, no AUC, no clearance figure and no absorption study in any species. The route question is unresolved in a way that undercuts the entire transfer from literature to use: every in vivo experiment retrieved here injected the compound into the peritoneal cavity at 100 mg per kg, and no study has administered it by mouth to anything. The mechanism is contested at its root, because a genetic double knockout showed that most of the transcriptional and metabolic response persists without either target receptor, and a second group traced the anticancer effect to LXR-alpha instead. No dedicated toxicology has been performed on any REV-ERB ligand, a gap the developers themselves recorded in 2014 and which no later study has filled, while the nitrothiophene group in the molecule is identified by both the originating laboratory and its critics as a known toxicophore. The endurance result on which the compound's reputation rests is one figure panel in one 2013 paper, six animals per group, with no number attached to it.

Questions