ACE-031 Research: A Soluble ActRIIB Decoy and a Clinical Programme Stopped on Safety Grounds

What is ACE-031?

What is ACE-031? It is a soluble receptor decoy: the extracellular, ligand-binding domain of the human activin receptor type IIB (ACVR2B, UniProt Q13705) fused to the Fc region of human IgG1, forming a dimeric protein. The Fc half does two jobs — it dimerises the construct, matching the way the receptor engages its dimeric ligands, and it gives the molecule the long circulating half-life characteristic of antibody-derived therapeutics. Its non-proprietary name is ramatercept; the developer was Acceleron Pharma.

It is not a peptide, and it is not a myostatin inhibitor in the narrow sense. It is a protein that pretends to be a receptor. Ligands that would otherwise dock at cell-surface ActRIIB bind the decoy instead and are removed from circulation. Everything below describes that mechanism and what published studies measured; none of it is a statement about an effect in a reader.

Discovery and origin in the literature

The design follows directly from a single mechanistic finding. Lee et al. (2005, PNAS, PMID 16330774) reported that muscle growth is regulated by multiple ligands signalling through activin type II receptors, not by myostatin alone, and that a soluble form of the receptor produced larger effects than blocking myostatin by itself. That result made the receptor, rather than the ligand, the obvious intervention point — and it is also the reason the safety picture later turned out to be broader than the muscle picture.

Preclinical characterisation followed. Cadena et al. (2010, J Appl Physiol, PMID 20466801) reported that administration of a soluble activin type IIB receptor promoted skeletal muscle growth independent of fibre type; Pistilli et al. (2011, Am J Pathol, PMID 21356379) targeted the receptor in the mdx mouse model of Duchenne muscular dystrophy; and Zhou et al. (2010, Cell, PMID 20723755) reported that ActRIIB antagonism reversed cancer cachexia and muscle wasting with prolonged survival in animal models. A primate study of ACE-031 in the common marmoset was published much later (Cadena et al. 2026, PLoS One, PMID 41686840). On that basis the compound entered human trials as a candidate for neuromuscular disease.

Reference data

Receptor data are from the UniProt human ACVR2B record (Q13705). Several fields that a peptide certificate would carry cannot be given for this molecule, and the reasons are stated rather than papered over: a glycosylated homodimeric Fc-fusion protein does not have a single meaningful molecular weight, and no CAS number is listed for it.

Molecule and analytical facts only. No reconstitution volume, concentration, route or handling instruction is given or implied.

Mechanisms researchers have examined

The ACE-031 mechanism is ligand trapping at the receptor level, and its consequences follow logically:

• Decoy binding — the soluble ActRIIB extracellular domain presents the same binding surface as the membrane receptor. Ligands bind it in the extracellular space and never reach cells, so signalling downstream through ALK4/ALK5 and Smad2/3 is not initiated.
• Multi-ligand trapping is the design, not a side effect — ActRIIB is a promiscuous receptor. Lee et al. (2005) established that trapping at this receptor captures a family of ligands, which is why the approach outperformed myostatin-specific inhibition preclinically. The same promiscuity means that any biology downstream of activins or GDF-11, in any tissue, is also affected.
• Fc-mediated persistence — the IgG1 Fc region confers the long circulating half-life reported in humans (10–15 days, Attie et al. 2013). It also means exposure cannot be withdrawn quickly.
• Effects reported outside muscle — Attie et al. (2013) reported changes in serum biomarkers that the authors interpreted as involving bone and fat metabolism, and Campbell et al. (2017) reported trends in bone mineral density and fat mass alongside lean mass. Whatever else it is, this is not a muscle-selective intervention.

Research findings by area

The healthy-volunteer study

Attie et al. (2013, Muscle Nerve, PMID 23169607) conducted a double-blind, placebo-controlled single-ascending-dose study in 48 healthy postmenopausal women randomised 3:1 to ACE-031 (0.02–3 mg/kg subcutaneous) or placebo. The compound was reported as generally well tolerated; adverse events included injection-site erythema. Exposure increased linearly with dose and mean half-life was 10–15 days. At day 29 in the 3 mg/kg group the authors reported statistically significant increases in mean total body lean mass (3.3%, P = 0.03, by DXA) and thigh muscle volume (5.1%, P = 0.03, by MRI). Reported as the paper reports it: these are measured body-composition endpoints in a single-dose study of 48 volunteers, not an outcome in anyone reading this page.

The Duchenne trial, and why it stopped

Campbell et al. (2017, Muscle Nerve, PMID 27462804) administered ACE-031 subcutaneously every two to four weeks to ambulatory boys with Duchenne muscular dystrophy in a randomised, double-blind, placebo-controlled ascending-dose trial whose primary objective was safety evaluation. The results section states that ACE-031 was not associated with serious or severe adverse events, and then states this:

“The study was stopped after the second dosing regimen due to potential safety concerns of epistaxis and telangiectasias.” The conclusion adds that “non-muscle-related adverse events contributed to the decision to discontinue the study.”

Epistaxis is nosebleeds; telangiectasias are dilated small blood vessels visible at the skin surface. Both are vascular rather than muscular, which is consistent with a molecule that traps a ligand family with roles well beyond skeletal muscle. The published report does not establish the mechanism behind those events, and this page does not speculate about one.

The efficacy signals in the same trial were trends, not significant results: a trend toward maintenance of six-minute-walk distance in the ACE-031 groups against a decline in placebo (explicitly noted as not statistically significant), together with trends for increased lean body mass and bone mineral density and reduced fat mass. The authors’ own conclusion retained optimism about myostatin inhibition as an approach for Duchenne while reporting the discontinuation. Development of ACE-031 did not continue.

What the field concluded

Wagner (2020, Curr Opin Neurol, PMID 32773450) reviewed the wider clinical experience under the title “The elusive promise of myostatin inhibition for muscular dystrophy”. Wetzlich et al. (2025, Mol Cell Biochem, PMID 39340593) survey therapeutic applications and challenges across the myostatin-inhibition field. ACE-031 is one of the clearest cases in that literature: strong preclinical rationale, measurable target engagement in humans, and a programme ended by adverse events that had nothing to do with muscle.

Research models and evidence status

The models run from cell and rodent work through a marmoset study to two human trials — one single-dose study in 48 healthy postmenopausal women, one halted paediatric trial in Duchenne muscular dystrophy. Both human studies were sponsored by the developer, with company affiliations listed on both author lists; that is stated as a fact about the evidence base, not as an accusation.

What is not established

ACE-031 is not approved anywhere, holds no registered indication, and its clinical development was discontinued. The discontinuation was driven by safety findings — epistaxis and telangiectasias — reported in the trial publication itself, and no page describing this compound is complete without that. No efficacy endpoint reached statistical significance in the Duchenne trial; the walk-test result was a non-significant trend. The mechanism of the vascular adverse events was not established in the published report. There are no long-term human safety data, no data in healthy men, and no data at all on use outside a supervised trial setting. Nothing on this page is a benefit, an outcome, a recommendation or a use.

On anti-doping status, reported as fact and not as a use case: Reichel et al. (2025) record that ACE-031 (ramatercept) is prohibited under section S4.3 of the WADA 2024 List of Prohibited Substances and Methods, covering agents that modify myostatin function.

ACE-031 among the myostatin-axis compounds

The three myostatin-axis pages in this catalogue occupy three different points on one pathway. GDF-8 is the ligand — myostatin itself, the endogenous brake, used in research as the reference agonist that inhibition assays measure against. Follistatin-344 is an endogenous antagonist that neutralises myostatin and activins by binding them directly. ACE-031 is a receptor decoy that blocks at the receptor instead, capturing the whole ligand family. Follistatin and ACE-031 therefore share a broad ligand profile, which is exactly why the ACE-031 safety outcome is relevant when reading follistatin’s literature.

The IGF-axis compounds sit on an unrelated receptor and are not part of this pathway at all: IGF-1 LR3 and IGF-1 DES are engineered IGF-1 analogues defined by reduced binding-protein affinity, and MGF and PEG-MGF are E-domain fragments of the IGF-1 gene product with no identified receptor. Mechanism context only; nothing here is a protocol, a stack or a recommendation.

How to verify this compound yourself

This is the compound in the cluster with published evidence on exactly this question, and the finding is stark. Reichel, Filip, Gmeiner and Thevis (2025, Drug Test Anal, PMID 40312924) analysed 14 black-market ACE-031 products by gel electrophoresis, Western blotting, mass spectrometry and IdeS protease treatment. Two contained no ACVR2B-immunoreactive protein at all. The other twelve contained many additional proteins beyond the main band, and the main band was full-length human activin receptor IIB rather than ACE-031 — the absence of an Fc-fusion protein was confirmed because IdeS, which cleaves IgG Fc, could not cleave the products. None of the fourteen was the labelled molecule.

What that implies for testing a batch of a fusion protein:

• Is the Fc actually present? — this is the check that failed on every product Reichel et al. examined. An ActRIIB fragment without an Fc region is a different molecule with different size, different dimerisation and a completely different half-life. Immunoblotting against the Fc, or protease-sensitivity testing, answers it; a purity percentage does not.
• Is it a dimer? — reducing versus non-reducing SDS-PAGE shows whether the construct is the intended disulfide-linked dimer or a monomeric species.
• How many other proteins are in there? — the same study reported many additional proteins alongside the main compound. SDS-PAGE across the full lane, not just the target band, is what reveals that; a single reported percentage can conceal it.
• Mass and glycosylation — a glycosylated fusion protein gives a broad, heterogeneous mass distribution rather than one sharp peak, so a certificate quoting a single exact mass for this molecule warrants a question about how it was measured.
• Endotoxin and sterility — where tested, separate quality attributes reported in EU/mg or as a sterility result, independent of purity and identity.

See how to read a COA for what each line on a certificate means, how to verify peptide purity for how the methods fit together, and the self-serve verify tool to check the exact batch on a vial in hand.

Research-grade sourcing and verification

For laboratory research use only. ACE-031 is not held in stock; it is listed as available to order in the sourcing catalogue, with a typical lead time of two to three weeks and a quote on request. Material supplied ships with a per-batch Certificate of Analysis reporting purity and identity confirmation at the lot level. Given the published finding that none of fourteen tested unregulated products was the labelled molecule, batch-level documentation is not a nicety on this compound — and no documentation changes the fact, stated above, that the clinical programme was discontinued after safety findings. Sourcing and identity-assurance framing only.

Frequently asked questions

Literature cited

1. Attie KM, Borgstein NG, Yang Y, et al. “A single ascending-dose study of muscle regulator ACE-031 in healthy volunteers.” Muscle Nerve. 2013;47(3):416–23. PMID 23169607. pubmed.ncbi.nlm.nih.gov/23169607.
2. Campbell C, McMillan HJ, Mah JK, et al. “Myostatin inhibitor ACE-031 treatment of ambulatory boys with Duchenne muscular dystrophy: Results of a randomized, placebo-controlled clinical trial.” Muscle Nerve. 2017;55(4):458–464. PMID 27462804. pubmed.ncbi.nlm.nih.gov/27462804.
3. Lee SJ, Reed LA, Davies MV, et al. “Regulation of muscle growth by multiple ligands signaling through activin type II receptors.” Proc Natl Acad Sci U S A. 2005;102(50):18117–22. PMID 16330774. pubmed.ncbi.nlm.nih.gov/16330774.
4. Cadena SM, Tomkinson KN, Monnell TE, et al. “Administration of a soluble activin type IIB receptor promotes skeletal muscle growth independent of fiber type.” J Appl Physiol (1985). 2010;109(3):635–42. PMID 20466801. pubmed.ncbi.nlm.nih.gov/20466801.
5. Pistilli EE, Bogdanovich S, Goncalves MD, et al. “Targeting the activin type IIB receptor to improve muscle mass and function in the mdx mouse model of Duchenne muscular dystrophy.” Am J Pathol. 2011;178(3):1287–97. PMID 21356379. pubmed.ncbi.nlm.nih.gov/21356379.
6. Zhou X, Wang JL, Lu J, et al. “Reversal of cancer cachexia and muscle wasting by ActRIIB antagonism leads to prolonged survival.” Cell. 2010;142(4):531–43. PMID 20723755. pubmed.ncbi.nlm.nih.gov/20723755.
7. Cadena SM, et al. “ACE-031, a soluble activin type IIB receptor, increases muscle mass and strength in the common marmoset (Callithrix jacchus).” PLoS One. 2026;21(2):e0342666. PMID 41686840. pubmed.ncbi.nlm.nih.gov/41686840.
8. Reichel C, Filip T, Gmeiner G, Thevis M. “Gel Electrophoretic Detection of Black Market ACE-031.” Drug Test Anal. 2025;17(10):1934–1946. PMID 40312924. pubmed.ncbi.nlm.nih.gov/40312924.
9. Wagner KR. “The elusive promise of myostatin inhibition for muscular dystrophy.” Curr Opin Neurol. 2020;33(5):621–628. PMID 32773450. pubmed.ncbi.nlm.nih.gov/32773450.
10. Wetzlich B, Nyakundi BB, Yang J. “Therapeutic applications and challenges in myostatin inhibition for enhanced skeletal muscle mass and functions.” Mol Cell Biochem. 2025;480(3):1535–1553. PMID 39340593. pubmed.ncbi.nlm.nih.gov/39340593.
11. UniProt Consortium. “AVR2B_HUMAN (Q13705) — Activin receptor type-2B.” uniprot.org/uniprotkb/Q13705 (precursor length, signal peptide, chain boundaries).