original article © The American Society of Gene & Cell Therapy Soluble activin type II receptors (ActRIIA/ActRIIB), via binding to diverse TGF-β proteins, can increase muscle and bone mass, correct anemia or protect against diet- induced obesity. While exciting, these multiple actions of soluble ActRIIA/IIB limit their therapeutic potential and highlight the need for new reagents that target spe- cific ActRIIA/IIB ligands. Here, we modified the activin A and activin B prodomains, regions required for mature growth factor synthesis, to generate specific activin antagonists. Initially, the prodomains were fused to the Fc region of mouse IgG2A antibody and, subsequently, “fastener” residues (Lys 45 , Tyr 96 , His 97 , and Ala 98 ; activin A numbering) that confer latency to other TGF-β proteins were incorporated. For the activin A prodomain, these modifications generated a reagent that potently (IC 50 5 nmol/l) and specifically inhibited activin A signaling in vitro, and activin A-induced muscle wasting in vivo. Interestingly, the modified activin B prodomain inhibited both activin A and B signaling in vitro (IC 50 ~2 nmol/l) and in vivo, suggesting it could serve as a general activin antagonist. Importantly, unlike soluble ActRIIA/IIB, the modified prodomains did not inhibit myostatin or GDF- 11 activity. To underscore the therapeutic utility of spe- cifically antagonising activin signaling, we demonstrate that the modified activin prodomains promote signifi- cant increases in muscle mass. Received 5 August 2014; accepted 9 November 2014; advance online publication 9 December 2014. doi:10.1038/mt.2014.221 INTRODUCTION Activin type II receptors (ActRIIA/ActRIIB) mediate the signal- ing of a subset of transforming growth factor-β (TGF-β) ligands, including activin A, activin B, myostatin, and GDF-11. 1 Ligand binding to ActRIIA/IIB leads to the activation of type I receptors (ALK4, 5, or 7), which initiate an intracellular signaling cascade centred on Smad2/3 transcription factors. 2 Smad2/3 activation drives the expression of genes involved in cellular proliferation, diferentiation, apoptosis and extracellular matrix deposition, 3 and is critical for the maintenance of adult tissue homeostasis. Accordingly, pharmacological blockade of the ActRIIA/IIB sig- naling pathway ofers great potential to restore homeostasis in disease-afected tissues. To date, ligand traps consisting of the extracellular domains of human ActRIIA or ActRIIB fused to IgG Fc regions have proven the most eicacious therapeutic reagents. 4–9 Soluble ActRIIA increases bone mass and strength and prevents cancer-induced bone destruction in models of myeloma and breast cancer, 5,8 by antagonising local activin A signaling. Interestingly, by inhibit- ing GDF-11, soluble ActRIIA also improves inefective erythro- poiesis and corrects anemia in a mouse model of β-thalassemia. 10 Soluble ActRIIB has even broader therapeutic potential, as it not only mimics the efects of soluble ActRIIA on bone growth and erythropoiesis, 7,9 but also dramatically increases muscle mass, predominantly by antagonizing myostatin activity. 4 As such, soluble ActRIIB has been used variously to improve muscle mass and function in the mdx mouse model of Duchenne muscular dystrophy, 11 and to reverse muscle wasting and prolong survival in murine models of cancer cachexia. 12 Recently, the potential of targeting the ActRIIA/IIB pathway to induce skeletal muscle hypertrophy has been conirmed using a human anti-ActRIIA/IIB antibody. 13 Surprisingly, this reagent also increased the mass and thermogenic activity of brown adipose tissue. 14 Although, individually, these studies demonstrate the thera- peutic potential of inhibiting the ActRIIA/IIB pathway, collec- tively they highlight problems associated with using ligand traps that target multiple TGF-β proteins. hus, there is a growing acceptance that interventions that target either one, or a small subset, of ActRIIA/IIB ligands will be the most efective way to achieve a desired outcome (e.g., muscle growth) with minimum risk of incurring signiicant of-target efects. In support, develop- ment of a modiied soluble ActRIIB with negligible activin bind- ing was recently shown to inhibit GDF-11 activity and promote erythropoiesis in mice. 9 The first two and the last two authors contributed equally. Correspondence: Paul Gregorevic, Baker IDI Heart and Diabetes Institute, P.O. Box 6492, St. Kilda Rd. Central, Melbourne 8008, Australia. E-mail: paul.gregorevic@bakeridi.edu.au Development of Novel Activin-Targeted Therapeutics Justin L Chen 1,2,6 , Kelly L Walton 1 , Sara L Al-Musawi 1 , Emily K Kelly 1 , Hongwei Qian 2 , Mylinh La 3 , Louis Lu 3 , George Lovrecz 3 , Mark Ziemann 2 , Ross Lazarus 2 , Assam El-Osta 2 , Paul Gregorevic 2,4,5,6 and Craig A Harrison 1,7,8 1 MIMR-PHI Institute of Medical Research, Clayton, Australia; 2 Baker IDI Heart and Diabetes Institute, Melbourne, Australia; 3 NCRIS Facility, CSIRO Material Sciences and Engineering, Clayton, Australia; 4 Department of Neurology, The University of Washington School of Medicine, Seattle, WA, USA; 5 Department of Physiology, The University of Melbourne, Melbourne, Australia; 6 Department of Biochemistry and Molecular Biology, Monash University, Clayton, Australia; 7 Department of Physiology, Monash University, Clayton, Australia; 8 Department of Molecular and Translational Sciences, Monash University, Clayton, Australia. Molecular Therapy 1