GDF-8 (myostatin): the muscle brake, the drugs that block it, and what research stores actually sell

Growth differentiation factor 8 is a member of the transforming growth factor beta superfamily, the same family as activins, bone morphogenetic proteins and TGF-beta itself. It is made almost entirely by skeletal muscle and acts on muscle as a brake: it signals through the activin type IIB receptor to hold back the growth and division of muscle cells. Se-Jin Lee’s laboratory at Johns Hopkins discovered it in 1997 and named it myostatin after showing that mice with the gene knocked out had two to three times the normal muscle mass, through a combination of more fibres and larger fibres ^[1].

The human protein is synthesised as a 375-amino-acid precursor. A 23-residue signal peptide is removed, the 243-residue propeptide (residues 24 to 266) is cut off and stays bound to the mature protein as a latent complex, and the 109-residue C-terminal domain (residues 267 to 375) forms the active, disulfide-linked dimer that binds the receptor ^[21]. The precursor has a mass of about 42.8 kDa; the mature dimer is about 25 kDa. GDF-8 is therefore a protein, far larger than the peptides most research stores sell, and the name describes a natural signalling molecule rather than a designed drug.

The GDF-8 story caught the public imagination because of natural mutants. Belgian Blue and Piedmontese cattle, bred for a century for their extreme double-muscled shape, turned out to carry mutations that inactivate myostatin ^[2]. So do racing whippets, in which one copy of a mutation makes the dog faster and two copies make it grotesquely bulky ^[4]. And in 2004 a German child was described with a splice mutation in both copies of the gene, visible muscle bulk at birth and the ability to hold two 3 kg dumbbells at arm’s length at four years old ^[3]. These cases are why myostatin inhibitors are sold to bodybuilders, and why they are on the WADA list.

Myostatin is secreted into the blood and tissue fluid in a latent form, bound to its own propeptide. When released, the mature dimer binds activin type IIB receptors on muscle cells, which recruit a type I receptor and phosphorylate Smad2 and Smad3, switching on genes that hold back protein synthesis and satellite cell activity ^[6]. It circulates at measurable levels: an assay developed in 2009 found serum myostatin averaged 8.0 ng/mL in young men and 7.0 ng/mL in older men, and levels rose when men were given testosterone ^[9].

Because it circulates, it can be caught. In 2001, Lee and McPherron showed that mice engineered to overproduce the myostatin propeptide, the natural inhibitor follistatin, or a dominant-negative form of the activin IIB receptor all developed the same dramatic increase in muscle mass as myostatin-null mice ^[5]. Those three approaches, propeptide, follistatin and receptor decoy, plus neutralising antibodies, are the basis of every myostatin inhibitor tested since, and of everything sold under the GDF-8 label online.

Two animal findings are important for anyone considering these products. First, myostatin given systemically to mice caused profound muscle and fat wasting, a cachexia-like state ^[7], so the intact protein is the opposite of a muscle builder. Second, muscle from myostatin-deficient mice, although much larger, generated no more maximum force than normal muscle and was weaker per unit of size, with fewer mitochondria and a shift towards fast-twitch fibres ^[8]. Bigger, in other words, was not stronger.

The label GDF-8 on a research site can mean several different things. Some listings are recombinant myostatin protein, usually the mature domain expressed in E. coli, which is a laboratory reagent for receptor and antibody studies and, as the mouse cachexia study shows, would reduce muscle if it did anything at all ^[7]. Others are the myostatin propeptide, the inhibitory fragment, sometimes described as GDF-8 propeptide or myostatin inhibitor. Others again are anti-myostatin antibodies, or follistatin 344, or the receptor decoy ACE-031. Only the last three groups are attempts to block myostatin.

A 2022 study by an Austrian anti-doping laboratory bought 12 black-market myostatin propeptide products. Only nine contained the protein at all, and in those nine the main component had a mass of 54 to 55 kDa rather than the expected 33 kDa, because the manufacturers had left a full-length GST purification tag attached. The products were also relatively impure ^[18]. A 2023 study developed detection methods for six myostatin-inhibitory peptides derived from the propeptide and from follistatin, noting that although none has clinical approval, they are readily available for research purposes and on the black market ^[19].

Buyers should therefore read a GDF-8 listing very carefully. A store that cannot say whether it is selling myostatin, its propeptide, an antibody or a follistatin construct, and cannot show a gel or mass spectrum for the product, is selling a name. The muscle growth peptides page sets these products alongside IGF-1 LR3, PEG-MGF and the growth hormone secretagogues.

Pharmaceutical myostatin inhibitors have reached large trials, and the results are instructive. MYO-029, a neutralising antibody from Wyeth, was given to adults with muscular dystrophies in a phase 1/2 trial published in 2008; it was tolerated but produced no improvement in muscle strength or function ^[10]. ACE-031, a soluble activin IIB receptor fused to an antibody fragment, increased total lean mass by 3.3% and thigh muscle volume by 5.1% within a month of a single 3 mg/kg dose in 48 healthy postmenopausal women ^[11], but its trial in boys with Duchenne muscular dystrophy was stopped for nosebleeds and dilated skin blood vessels, side effects of blocking other ligands of the same receptor ^[12].

LY2495655, an antibody from Eli Lilly, was tested in 201 people aged 75 or over who had fallen and had low muscle strength. Six injections of 315 mg over 20 weeks increased appendicular lean mass by 0.43 kg relative to placebo and improved stair-climbing and chair-rise times, with modest effect sizes ^[13]. Bimagrumab, an antibody that blocks the activin IIB receptor rather than myostatin itself, was tested in 251 people with inclusion body myositis in the RESILIENT trial; it increased muscle mass but did not change six-minute walking distance, and caused muscle spasms and diarrhoea far more often than placebo ^[14]. The same antibody reduced total body fat and increased lean mass over 48 weeks in adults with type 2 diabetes and obesity ^[16], a finding now being pursued in combination with GLP-1 drugs such as semaglutide.

In Duchenne muscular dystrophy, domagrozumab, a Pfizer antibody, showed no difference from placebo in a 120-boy phase 2 trial despite non-significant increases in muscle volume ^[15]. The clearest success is in spinal muscular atrophy: apitegromab, an antibody that binds the latent form of myostatin, improved motor function scores by 1.8 points over placebo at 12 months in the 188-patient SAPPHIRE phase 3 trial in children already on SMN-targeted therapy, with adverse events similar to placebo ^[17]. Even here the effect is a slowing of decline in a rare disease, not a transformation.

The pattern across two decades is consistent: blocking myostatin adds muscle mass, usually a few per cent, and strength or function changes are small or absent. The hopes raised by the myostatin-null mouse and the double-muscled bull have not translated into large functional gains in people.

No study has given a myostatin inhibitor to healthy trained adults to see whether they gain muscle or strength. The closest data are the single-dose ACE-031 study in postmenopausal women ^[11] and the antibody trials in older or ill people ^[13][14]. Those show that the concept works, in the sense that lean mass rises, but the gains are small next to what training already produces, the compounds used are antibodies and receptor fusions made to pharmaceutical standards, and the doses were hundreds of milligrams of protein given intravenously or subcutaneously under monitoring. A vial of propeptide of uncertain identity is not comparable to any of them.

Searches for GDF-8 peptide dosage or a myostatin inhibitor cycle will not find trial data, because no trial exists for the products sold. The animal data also point to a trade-off that bodybuilders may not want: myostatin-deficient muscle is bigger but not proportionally stronger, with fewer mitochondria and less endurance ^[8].

The trial doses on record are for pharmaceutical inhibitors, not for anything sold as GDF-8. ACE-031 was given as a single subcutaneous dose of 0.02 to 3 mg/kg, with a half-life of 10 to 15 days ^[11]. LY2495655 was given as 315 mg subcutaneously every four weeks for six doses ^[13]. Bimagrumab was infused at 1, 3 or 10 mg/kg every four weeks for at least 48 weeks in inclusion body myositis ^[14] and at 10 mg/kg every four weeks for 48 weeks in obesity ^[16]. Domagrozumab was given at 5, 20 and 40 mg/kg ^[15] and apitegromab at 10 or 20 mg/kg intravenously ^[17].

These figures are reported so that readers can see the scale of what was tested; they are not recommendations and they cannot be transferred to a research-store product, which is a different molecule. Our peptide calculator does the arithmetic for reconstituting a vial but cannot tell you what protein, if any, the vial contains.

Every myostatin inhibitor in trials was a large protein given by subcutaneous or intravenous injection ^{[11][13][14][17]}. None is active by mouth, because proteins are digested. Research suppliers sell lyophilised protein, usually 1 mg vials, for reconstitution and subcutaneous injection. The anti-doping analyses show that what is in those vials is often a tagged, impure bacterial expression product ^[18], and antibodies and receptor fusions are far harder to make and are rarely what a store is actually selling.

In the trials, myostatin-pathway inhibitors caused injection-site reactions ^[11][13], muscle spasms and diarrhoea ^[14], and, for the receptor decoy ACE-031, nosebleeds and telangiectasia serious enough to stop a paediatric trial ^[12]. Blocking the activin IIB receptor affects ligands beyond myostatin, including activins and GDF-11, which is why the receptor-based agents have more off-target effects than myostatin-specific antibodies. Apitegromab, which binds only the latent form of myostatin, had an adverse event profile similar to placebo ^[17].

For research-store products, the greater concern is what the vial contains. Bacterial expression products carry endotoxin unless properly purified, a GST tag makes the protein immunogenic, and injecting an impure foreign protein repeatedly invites antibody formation and allergic reactions. If the product is actually myostatin rather than an inhibitor, the animal data suggest it would cause muscle loss ^[7]. No safety data exist for any of these products in people.

No myostatin inhibitor has marketing authorisation from the FDA, MHRA or EMA as of our review; apitegromab is the most advanced and is under regulatory review. The World Anti-Doping Agency lists agents preventing activin receptor IIB activation, including myostatin-neutralising antibodies and myostatin-binding proteins such as follistatin and myostatin propeptide, under section S4 of the Prohibited List, banned at all times ^[20], and anti-doping laboratories have published detection methods for propeptide, follistatin and antibody products ^[18][19]. Myostatin propeptide is named explicitly ^[18].

As unlicensed proteins sold for research, these products fall under the same rules as other research peptides: see our legal status overview and the UK, US and Australian pages.

Recombinant proteins are less stable than short peptides. Lyophilised myostatin, propeptide and follistatin products are normally shipped cold and stored frozen or refrigerated; once reconstituted they should be kept cold, not refrozen repeatedly, and used within days, because proteins aggregate and lose activity with freeze-thaw cycles. Our guides to how to store peptides and bacteriostatic water cover the general principles; for proteins, carrier-free preparations are particularly prone to sticking to vial walls at low concentrations.

For a protein, the useful tests are different from those for a peptide. A credible certificate for a GDF-8 product should state what the product is (myostatin, propeptide, antibody or follistatin), show an SDS-PAGE gel with a band at the expected mass, report purity and endotoxin, and ideally include a bioassay. The anti-doping study that found tagged 54 kDa proteins in vials labelled as 33 kDa propeptide is a reminder of how far real products can be from their labels ^[18]. Compare GDF-8 prices, read how to read a COA, and see our list of third-party tested suppliers.

We track listings in the UK and the US. Our guides to third-party peptide testing and how to spot a fake peptide supplier apply, with the caveat that standard mass spectrometry identity testing is much harder for a 25 kDa dimer than for a 1 kDa peptide.