ACE-031

ACE-031, also called ramatercept, is a recombinant fusion protein that joins the extracellular domain of human activin receptor type IIB (ActRIIB) to the Fc region of human IgG1. It acts as a ligand trap that binds myostatin and related ligands (Campbell et al., 2017). Acceleron Pharma developed it with Shire (Smith and Lin, 2013).

PubChem holds only substance records for this protein: SID 135293965 supplies the names and CAS number, and SID 172232441 links KEGG DRUG D10347, whose formula and weight are calculated from the sequence.

Published work on ACE-031 and related ActRIIB-Fc proteins falls into six areas, each naming its model.

A trap for myostatin, GDF-11 and activins

Using a soluble ActRIIB-Fc, Acceleron scientists found that myostatin (GDF-8) and GDF-11 bind the extracellular domain with affinities comparable to activin A, while BMP-2 and BMP-7 bind at least 100-fold more weakly (Sako et al., 2010).

In wild-type mice, a soluble ActRIIB raised muscle mass by up to 60 percent in two weeks, and a smaller effect in myostatin-knockout mice suggested that other ligands also limit muscle mass (Lee et al., 2005). Unlike homodimeric ActRIIB-Fc, a heterodimeric ActRIIB:ALK4-Fc did not trap BMP9, a vascular regulatory ligand (Li et al., 2021). Models: binding assays and mice.

Skeletal muscle in mice

In C57BL/6 mice given ACE-031 for 28 days, four hindlimb muscles weighed 26 to 46 percent more than in controls, with larger type I and type II fibers and no change in fiber-type distribution (Cadena et al., 2010).

In mdx mice, a Duchenne model, a mouse-Fc version at 1.0 or 10.0 mg/kg for 12 weeks raised lean muscle mass by 33 percent at the higher level and lowered serum creatine kinase (Pistilli et al., 2011). In an academic study, four months of ActRIIB-Fc caused extreme muscle fatigability and fewer capillaries, plus raised serum lactate and a severe metabolic myopathy in mdx mice (Relizani et al., 2014). Models: wild-type and mdx mice.

Bone in rodent models

In castrated mice, a mouse-Fc ActRIIB fusion prevented the loss of bone mass measured by DXA and micro-CT (Koncarevic et al., 2010). In young adult mice, a Pfizer ActRIIB-Fc raised bone volume fraction by 132 percent in the distal femur and 27 percent in the lumbar vertebrae, through thicker, more numerous trabeculae. A myostatin antibody did not raise bone mass, while the decoy receptor did so even in myostatin-knockout mice (Bialek et al., 2014). Models: castrated, wild-type and knockout mice.

A nonhuman primate study

Eight common marmosets received ACE-031 at 3.0 mg/kg once a week for 14 weeks, and four received vehicle. Lean body mass rose from baseline only in the ACE-031 group, biceps brachii fibers of both types grew in cross-sectional area, and the authors reported it was generally well tolerated (Cadena et al., 2026). Model: common marmoset.

Human studies and the end of development

In a randomized, double-blind, placebo-controlled phase 1 study, 48 healthy postmenopausal women received placebo or one of several single ascending amounts. It was generally well tolerated, with a mean half-life of 10 to 15 days. At day 29 the highest-amount group showed significant increases in total body lean mass (3.3 percent, DXA) and thigh muscle volume (5.1 percent, MRI), and serum biomarker changes the authors linked to bone and fat metabolism (Attie et al., 2013).

In a similarly designed trial in ambulatory boys with Duchenne muscular dystrophy, ACE-031 was not associated with serious or severe adverse events, but the study was stopped after its second regimen because of potential safety concerns of epistaxis (nosebleeds) and telangiectasias (small dilated blood vessels).

Trends, none statistically significant, pointed to higher lean body mass and bone mineral density and a steady six-minute walk distance against a decline on placebo (Campbell et al., 2017). A 2013 review lists a multiple ascending study in healthy women as also terminated, with nosebleeds common, and flags BMP9 because mutations in its receptor ALK1 are linked to hereditary hemorrhagic telangiectasia (Smith and Lin, 2013). Models: randomized human trials and a review.

Doping control and products sold online

Of 14 products sold online as ACE-031, 12 held an ActRIIB-reactive protein that mass spectrometry and IdeS digestion identified as full-length human ActRIIB, not an Fc fusion; 11 of the 12 were His-tagged. One held follistatin 344, and one held no protein but ipamorelin. In rats given 10 mg/kg of one product, it was detectable in serum for up to 48 hours (Reichel et al., 2025). Models: purchased products and rats.