US007833971B2

(12) United States Patent (10) Patent No.: US 7,833,971 B2 Grinberg et al. (45) Date of Patent: Nov. 16, 2010

(54) USES OF CERBERUS, COCO AND WO WO-03/O12082 2, 2003 DERVATIVES THEREOF WO WO-03/05.5911 T 2003 WO WO-O3,O72714 9, 2003 (75) Inventors: Asya Grinberg, Boston, MA (US); John WO WO-03/106657 12/2003 Knopf, Carlisle, MA (US); Ravindra WO WO-2004/074460 9, 2004 Kumar, Shrewsbury, MA (US); Jasbir WO WO-2005/003 158 1, 2005 Seehra, Lexington, MA (US) WO WO-2005/115439 A2 12/2005 (73) Assignee: Acceleron Pharma Inc., Cambridge, OTHER PUBLICATIONS MA (US) Avsian-Kretchmer et al., “Comparative geneomic analysis of the eight-membered ring cystine knot-containing bone morphogenetic (*) Notice: Subject to any disclaimer, the term of this protein antagonists.” Molecular Endocrinology 18(1): 1-12 (2004). patent is extended or adjusted under 35 Belo et al., “Cerberus-like is a secreted factor with nerualizing activ U.S.C. 154(b) by 0 days. ity expressed in the anterior primitive endoderm of the mouse gas trula.” Mechanisims of Development 68:45-57 (1997). (21) Appl. No.: 12/001,494 Biben et al., “Murine Cerberus Homologue mCer-1: A canadidate anterior patterning molecule.” Developmental Biology 194:135-151 (22) Filed: Dec. 10, 2007 (1998). Bouwmeester et al., “Cerberus is a head-inducing secreted factor (65) Prior Publication Data expressed in the anterior endoderm of Spemann's organizer. Nature US 2009/OO82270 A1 Mar. 26, 2009 382:595-601 (1996). Esteban et al., “The novel Cer-like protein Caronte mediates the extablishment of embryonic left-right asymmetry.” Nature 401:243 Related U.S. Application Data 251 (1999). (60) Provisional application No. 60/873,933, filed on Dec. Katoh et al., “Identification and characterization of human 8, 2006. CKTSFIB2 and CKTSFIB3 genes in silico.” Oncology Reports 12:423-427 (2004). Kuroda et al., “Neural Induction in Xenopus. requirement for (51) Int. Cl. extodermal and endomesodermal signals via chordin, noggin, C7K I4/0 (2006.01) B-catenin and cerberus.” PLoS Biology 2(5):0623-0634 (2004). (52) U.S. Cl...... 514/2:530/350 Lahet al., “Human cerberus related gene CERI maps to chromosome (58) Field of Classification Search ...... None 9.” Genomics 55:364-366 (1999). See application file for complete search history. Marques et al., “The activity of the nodal antagonist Cerl-2 in the mouse node is required for correct LR body axis.” Genes & Devel (56) References Cited opment 18:2342-2347 (2004). U.S. PATENT DOCUMENTS Pearce et al., “A mouse cerberus/dan-related gene family.” Develop mental Biology 209:98-110 (1999). 5,935,852 A 8, 1999 Follettie et al. Piccolo et al., “The head inducer Cerberus is a multifunctional 6,133,232 A 10, 2000 De Robertis et al. antagonist of Nodal, BMP and Wnt signals.” Nature 397:707-710 6,610,510 B1 8, 2003 Valenzuela et al. (1999). 7,316,998 B2 1/2008 Knopfet al. Silva et al., “Endogenous Cerberus activity is required for anterior 2002fO164682 A1 11/2002 Follettie et al. head specification in Xenopus.” Development 130(20):4943-4953 2003/0134790 A1 7/2003 Langenfeld (2003). 2003. O194704 A1 10/2003 Penn et al. Stanley et al., “Murine cerberus homologue Cer 1 maps to chromo 2003. O199042 A1 10/2003 Valenzuela et al. some 4.” Genomics 49:337-338 (1998). 2004.0005560 A1 1/2004 Isogai et al. Motoko Yanagita. BMP antagonists: Their roles in development and 2004/O181033 A1 9, 2004 Han et al. involvement in pathophysiology. Cytokine and Growth Factor 2005, 0186663 A1 8, 2005 Davies et al. Reviews. vol. 16, No. 3, pp. 309-319. 2005. 2006/0105376 A1 5/2006 Isogai et al. Brecher, A.S., et al., "Acetaldehyde Inhibits Chymotrypsin and 2008, OO32304 A1 2/2008 Isogai et al. Serum Anti-Chymotrypsin Activity.” J. Investig Med., 46(4): 146-152 (1998). Abstract only. FOREIGN PATENT DOCUMENTS Livingston, S.F., et al., “The Significance of Chymotrypsin-Inhibitor EP 1347046 9, 2003 Levels in the Serum of Patients with Carcinoma of the Breast.” WO WO-97/.48275 12/1997 Cancer Research, 17(9):857-861 (1957). WO WO-98,34951 8, 1998 WO WO-98/49296 11, 1998 Primary Examiner Karen Cochrane Carlson WO WO-99/O1553 1, 1999 (74) Attorney, Agent, or Firm Ropes & Gray LLP WO WO-99/40181 8, 1999 WO WO-OO,55193 9, 2000 (57) ABSTRACT WO WO-02/10214 2, 2002 WO WO-O2/32929 4/2002 The disclosure relates to Cerberus/Coco polypeptides or vari WO WO-O2/O54940 T 2002 WO WO-O2/O77204 10, 2002 ants thereof for use in treating a variety of disorders associ WO WO-O2/O78516 10, 2002 ated with myostatin, nodal and GDF-11. WO WO-O2/O90992 11, 2002 WO WO-03/055443 1, 2003 6 Claims, 8 Drawing Sheets U.S. Patent Nov. 16, 2010 Sheet 1 of 8 US 7,833,971 B2

Figure 1 U.S. Patent Nov. 16, 2010 Sheet 2 of 8 US 7,833,971 B2

RU 2000 1700 1400 1100 800 500 200 -100 -100 O 100 200 300 400 500 600 TIME S

Figure 2 U.S. Patent Nov. 16, 2010 Sheet 3 of 8 US 7,833,971 B2

SYNTHETIC poly(A) SGNA / TRANSCRIPTIONAL PAUSE SITE (FOR 3ACKGROUND - - REDUCTION, / 3V1 Amp' f of pG-3BASIC VECTOR ws.xx-x-r-w f : 20052004 Ban H N.Ya P.

poly(A)SIGNALSy40 ATE K, -3. (FOR CREPORTER, Xball 742 Hipg 1902

Figure 3 U.S. Patent Nov. 16, 2010 Sheet 4 of 8 US 7,833,971 B2

U.S. Patent Nov. 16, 2010 Sheet 5 of 8 US 7,833,971 B2

8 8.: 3. 38:

U.S. Patent Nov. 16, 2010 Sheet 6 of 8 US 7,833,971 B2

U.S. Patent Nov. 16, 2010 Sheet 7 of 8 US 7,833,971 B2

S.Po g human serum added to CM (37 deg. Covernight) O 1 1 25 5

U.S. Patent Nov. 16, 2010 Sheet 8 of 8 US 7,833,971 B2

Inhibition of GDF-11 by human Coco mfc (Conditioned Medium)

0.6 O.5 0.4 0.3 -0 - GDF-11 + COCO CM 0.2 0.1 O O O1 0.1 1 10 100 Media Dilutions

Figure 8 US 7,833,971 B2 1. 2 USES OF CERBERUS, COCO AND statin reduced-function mutation in a human child is associ DERVATIVES THEREOF ated with gross muscle hypertrophy and a family history of exceptional strength (Schuelke et al. 2004 Jun. 24; 350(26): CROSS-REFERENCE TO RELATED 2682-8). An antibody against myostatin is reported to have APPLICATIONS 5 beneficial effects in animal models of muscle disorders, including amyotrophic lateral Sclerosis (Holzbauer et al. This application claims the benefit of the filing date of U.S. Neurobiol Dis. 2006 September; 23(3):697–707). Application No. 60/873,933, filed Dec. 8, 2006, which is Additionally, it should be noted that the serum and intra incorporated by reference in its entirety. muscular concentrations of immunoreactive myostatin are 10 increased in HIV-infected men with muscle wasting com BACKGROUND OF THE INVENTION pared with healthy men, and correlate inversely with the fat-free mass index. These data Support the hypothesis that Transforming growth factor-B Superfamily proteins repre myostatin is a negative regulator of skeletal muscle growth in sent a large family of cytokines that includes the TGF-Bs, adult men and contributes to muscle wasting in HIV-infected activins, inhibins, bone morphogenetic proteins (BMPs) and 15 men (Nestor et al., Supra). Mullerian-inhibiting substance (MIS) (for review, see Mas In view of the above findings, a need exists for a manner of sague et al., Trends Cell Biol. 7:187-192, 1997). These pro regulating myostatin activity, particularly in individuals who teins contain a conserved C-terminal cysteine-knot motif, and experience muscle wasting as a result of a condition or dis serve as ligands for diverse families of plasma membrane ease state such as, for example, aging related frailty, cachexia receptors. Members of the TGF-B family exert a wide range 20 in Autoimmune Deficiency Syndrome (AIDS), Multiple of biological effects on a large variety of cell types. Many Sclerosis, muscular dystrophy, ALS and cancer-cachexia. members of this family have important functions during The present invention provides methods and compositions embryonal development in pattern formation and tissue which may be utilized to help individuals with such muscle specification; in the adult, these factors are involved in pro wasting conditions and provides further insight into the regu cesses such as tissue repair and modulation of the immune 25 lation of myostatin gene expression. system. Activities of the TGF-B superfamily proteins are regulated SUMMARY OF THE INVENTION through various means. One of the negative regulations for the BMP subfamily of proteins is through a relatively large In part, the disclosure relates to the discovery that two family of so-called Bone Morphogenetic Protein (BMP) 30 human proteins, Cerberus and Coco, that belong to a group of antagonists/repressors. These BMP repressors represent a GDF/BMP antagonists, bind to and antagonize myostatin, subgroup of proteins that bind BMPs, and interfere with BMP GDF11 and Nodal, and furthermore, that the myostatin/ binding to their membrane receptors, thereby antagonizing GDF11 binding domain resides in the cysteine knot domain their actions during development and morphogenesis. of these proteins. Furthermore, with respect to Cerberus, The BMP repressors can be further divided into three 35 myostatin/GDF11 binding and antagonist activity can be groups of proteins based on structural analysis, especially the separated from the BMP4/2 binding and antagonist activity. number of structurally conserved Cys residues in their C-ter Therefore, the disclosure provides, in part, methods for minal characteristic “Cys-knot' structures: the 8-, 9-, or antagonizing myostatin and GDF11 in vivo by administering 10-member ring Cys-knot BMP repressors. The 8-member polypeptides comprising a myostatin binding portion of Cer ring (CAN subfamily) repressors can be divided further into 40 berus or Coco, or variants thereof. One aspect of the invention four Subgroups based on a conserved arrangement of addi provides polypeptides, and pharmaceutical preparations tional cysteine residues gremlin and PRDC, Cerberus and thereof, of Cerberus, Coco (from human or non-human ani coco, and DAN, together with USAG-1 and sclerostin. mals) or a derivative thereof (collectively herein “Cerberus/ Orthologs of these human BMP antagonists in the genomes of Coco proteins”) for inhibiting the function/signaling of several model organisms have also been identified, and their 45 Nodal, myostatin, GDF11 and, in certain forms, BMP4 and/ phylogenetic relationship has been analyzed (AVsian-Kretch or BMP2. In certain embodiments, preparations of the subject mer and Hsueh, Mol Endocrinol. 18(1): 1-12, 2004, incorpo Cerberus/Coco polypeptides may include variant Cerberus or rated herein by reference). Coco proteins that retain all or a substantial portion of the Myostatin, or growth/differentiation factor 8 (GDF-8), binding affinity of the parent protein to Nodal, myostatin, also belongs to the transforming growth factor-fi (TGF-3) 50 GDF11 and/or another BMP (such as BMP4). In certain superfamily (McPherron et al., Nature 387:83-90 (1997)). embodiments, preparations of the subject Cerberus/Coco The human myostatin gene has been cloned (Nestor et al. polypeptides include variant Cerberus or Coco proteins that Proc. Natl. Acad. Sci. 95:14938-43 (1998)), and it has been retain all or a substantial portion of the binding affinity of the reported that myostatin immunoreactivity is detectable in parent protein to myostatin and/or GDF11 while eliminating human skeletal muscle in both type 1 and type 2 fibers. With 55 or reducing binding to BMP4 and/or BMP2. In certain respect to function, myostatin may play a role in negatively embodiments, the disclosure provides the observation that regulating the growth and development of skeletal muscle full-length human Cerberus is unstable in the presence of (Nestor et al., Supra). human serum, and thus altered forms of Cerberus (both A study with myostatin knock-out mice provided the first BMP4 binding forms and selective myostatin/GDF1 1/Nodal evidence that myostatin is a key negative regulator of muscle 60 binding forms) may be prepared that are stable in the serum development (McPherronet al., Nature 387:83-90 (1997)). In for a period of at least 24 hours, and optionally 2, 3, 5, 7, 14 the myostatin null mice, the animals were significantly larger or 21 days or longer. This observation may be extrapolated to than wild-type mice and had a large and widespread increase Coco, and thus altered forms of Coco may be prepared that are in skeletal muscle mass. Furthermore, two breeds of cattle, stable in the serum for a period of at least 24 hours, and characterized by increased muscle mass, have mutations in 65 optionally 2, 3, 5, 7, 14 or 21 days or longer. In certain the myostatin coding sequence (McPherronet al., Proc. Natl. embodiments, the disclosure provides pharmaceutical prepa Acad. Sci.94:12457-61 (1997)). A naturally occurring myo rations for inhibiting myostatin, comprising a myostatin US 7,833,971 B2 3 4 antagonist protein that includes (at least) a myostatin binding 95%, or 99% or more sequence similarity or identity with the domain of a Cerberus/Coco polypeptide or variant thereof. human or mouse Cerberus protein, and Substantially retain The myostatin antagonist protein binds to and neutralizes one the binding affinity of wild-type Cerberus for myostatin. or more of nodal and/or myostatin. Preferably, the pharma Likewise, the Coco protein variant can be derived from a ceutical preparation is Substantially free of pyrogenic mate human, mouse, or other species of Coco, including a human rials so as to be Suitable for administration as a human or or mouse Coco variant sequence sharing at least about 50%, Veterinarian therapeutic. 60%, 70%, 80%, 90%, 95%, or 99% or more sequence simi Myostatin is widely recognized as an antagonist of muscle larity or identity with the human or mouse Coco protein, and growth. Furthermore, myostatin null mice have shown resis substantially retain the binding affinity of wild-type Coco for tance to obesity and diabetes under certain conditions. There 10 myostatin. fore, the Cerberus/Coco proteins and pharmaceutical prepa In certain related embodiments, the mysotatin inhibitor is a rations described herein can be used to reduce the severity of polypeptide that includes a myostatin binding domain of a a pathologic condition, which is characterized, at least in part, Cerberus/Coco protein, which polypeptide does not substan by an abnormal amount, development or metabolic activity of tially bind BMP-4 or BMP-2. For instance, the myostatin muscle or adipose tissue in a Subject. For instance, the phar 15 binding domain can be derived from a human, mouse, or other maceutical preparations of the present invention can be species of N-terminally truncated Cerberus, including a administered in an amount effective to prevent, ameliorate or human or mouse Cerberus derivative, with amino acid resi reduce the severity of a wasting disorder, Such as age-related dues starting from any one of residues 106-119 of SEQID No. wasting, age-related frailty, cachexia, anorexia, Duchenne 1 or 2, and ending at any residue after residue 241 of SEQID Muscular Dystrophy (DMD) syndrome, Becker's Muscular No. 1 or 2, preferably ending at any residue between residues Dystrophy (BMD) syndrome, facio-scapular-humeral (FSH) 241 and 267 of SEQ ID No. 1 or 2 (all residue numbers muscular dystrophy, other muscular dystrophies, AIDS wast inclusive). ing syndrome, neuromuscular diseases, motor neuron dis For example, residues 106-119 of human Cerberus are: eases, diseases of the neuromuscular junction, and inflamma tory myopathies. Excessive BMP4 activity has been 25 associated with pathological ossification of various connec PPGTOSLIQPIDGM (SEO ID NO: 7) tive tissues. Therefore, the Cerberus/Coco proteins and phar Residues 241-267 of human Cerberus are: maceutical preparations that retain anti-BMP4 activity can be used to reduce the severity of a pathologic condition, which is characterized, at least in part, by an abnormal ossification in 30 CKWKTEHEDGHILHAGSODSFIPGVSA (SEQ ID NO: 8) tissues such as muscles, tendons, and ligaments. BMP4 is also associated with Osteoarthritis (OA), including the develop Also included are Cerberus derived variant sequences, e.g., ment of osteophytes and synovial thickening; Fibrodysplasia an N-terminally truncated myostatin binding domain of Cer ossificans progressiva (FOP); and atherosclerosis, especially berus that retains myostatin binding activity but loses other inflammatory response in early steps of atherogenesis in 35 BMP binding activity. Variant sequences may be desirable as lesion-prone areas; and craniosynostoses. Nodal signaling a way to alter selectivity of the inhibitor (e.g., relative to has been associated with certain cancers, particularly mela GDF-8, 11 or nodal binding), alter other binding characteris noma. Accordingly, Cerberus/Coco proteins and pharmaceu tics with respect to myostatin (such as K, and/or K, or K. tical preparations that retain anti-Nodal activity can be used to rates), or improve biodistribution or half life in vivo or on the treat tumors, particularly tumors such as melanomas in which 40 shelf. Nodal participates in tumor growth and development. In certain preferred embodiments, the Cerberus polypep Another aspect of the invention provides a pharmaceutical tide (full-length or N-terminally truncated) comprising the preparation of Cerberus/Coco protein derivative for specifi myostatin binding domain binds myostatin with a K of 1 LM cally inhibiting Nodal and/or myostatin function without sub or less, and more preferably a K of 100 nM, 10 nM or even 1 stantially compromising BMP (such as BMP-4) signaling 45 nM or less. (e.g., does not substantially bind BMP-4 or other BMPs). In certain related embodiments, the mysotatin inhibitor is a Exemplary preparations of this aspect of the invention polypeptide that includes a myostatin binding domain of a include polypeptides including the N-terminal truncated ver Coco protein, such as the human Coco protein shown in SEQ sions of Cerberus or Coco, or other fragments that include the ID NO:5 or in GenBank Accession number 22749329. cysteine-core. These so-called “N-terminally truncated Cer 50 In certain preferred embodiments, the Coco polypeptide berus/Coco derivatives' can be used to reduce the severity of (full-length or N-terminally truncated) comprising the myo a pathologic condition, which is characterized, at least in part, statin binding domain binds myostatin with a K of 1 M or by an abnormal amount, development or metabolic activity of less, and more preferably a Kof 100 nM, 10 nM or even 1 nM muscle or adipose tissue in a Subject. For instance, the phar or less. maceutical preparations of the present invention can be 55 In certain embodiments, the Cerberus/Coco polypeptide administered in an amount effective to prevent, ameliorate or (e.g., a myostatin binding domain thereof) is part of a fusion reduce the severity of a wasting disorder. Such as cachexia, protein including, in addition to the myostatin binding anorexia, DMD syndrome, BMD syndrome, AIDS wasting domain, one or more polypeptide portions that enhance one or syndrome, muscular dystrophies, neuromuscular diseases, more of in vivo stability, in vivo half life, uptake/administra motor neuron diseases, diseases of the neuromuscular junc 60 tion, tissue localization or distribution, formation of protein tion, and inflammatory myopathies. complexes, and/or purification. For instance, the fusion pro In certain embodiments, the mySotatin inhibitor is a tein can include an immunoglobulin Fc domain. The fusion polypeptide that includes a myostatin binding domain of a protein may include a purification Subsequence, such as an Cerberus/Coco protein. For instance, the Cerberus protein epitope tag, a FLAG tag, a polyhistidine sequence, or as a variant can be derived from a human, mouse, or other species 65 GST fusion. of Cerberus, including a human or mouse Cerberus variant In certain embodiments, the Cerberus/Coco polypeptide sequence sharing at least about 50%, 60%, 70%, 80%, 90%, (e.g., myostatin binding domain thereof) is part of a protein US 7,833,971 B2 5 6 that includes one or more modified amino acid residues. Such tering a pharmaceutical preparation of one or more of the as a glycosylated amino acid, a PEGylated amino acid, a subject variant Cerberus/Coco polypeptides. The subject farnesylated amino acid, an acetylated amino acid, a biotiny method can be used to promote muscle growth, promote lated amino acid, an amino acid conjugated to a lipid moiety, adipogenic differentiation, and/or promote bone growth or or an amino acid conjugated to an organic derivatizing agent. mineralization in human patients or in non-human animals. In certain embodiments, a subject variant Cerberus/Coco In certain embodiments, the treatment methods of the polypeptide is selective for binding and inhibition of myosta present disclosure can be used to reduce the severity of a tin, e.g., relative to 11 and/or nodal. For instance, the variant pathologic condition, which is characterized, at least in part, Cerberus/Coco polypeptide can be one which has a dissocia by an abnormal amount, development or metabolic activity of tion constant(K) for myostatin binding that is at least 2 times 10 muscle or adipose tissue in a Subject. For instance, the phar less than its K for binding 11 and/or nodal, and even more maceutical preparations of the present disclosure can be preferably at least 5, 10, 100 or even 1000 times less. Whether administered in an amount effective to prevent, ameliorate or by virtue of binding kinetics or biodistribution, the subject reduce the severity of a wasting disorder, Such as cachexia, variant Cerberus/Coco polypeptide can also be selected based anorexia, DMD syndrome, BMD syndrome, AIDS wasting on relative in vivo potency, such as an inhibitor that has an 15 syndrome, muscular dystrophies, neuromuscular diseases, ECs for inhibiting myostatin activity, or a particular physi motor neuron diseases, diseases of the neuromuscular junc ological consequence (Such as promoting muscle growth) tion, and inflammatory myopathies. that is at least 2 times less than its ECs for inhibiting 11 Exemplary muscular dystrophies that can be treated with a and/or nodal activities, and even more preferably at least 5, regimen including the Subject myostatin include: Duchenne 10, 100 or even 1000 times less. Muscular Dystrophy (DMD). Becker Muscular Dystrophy In certain embodiments, the subject variant Cerberus/Coco (BMD), Emery-Dreifuss Muscular Dystrophy (EDMD), polypeptide is selective for binding and inhibition of myosta Limb-Girdle Muscular Dystrophy (LGMD), Facioscapulo tin, e.g., relative to other BMP proteins such as BMP4. For humeral Muscular Dystrophy (FSH or FSHD) (Also known instance, the variant Cerberus/Coco polypeptide can be one as Landouzy-Dejerine), Myotonic Dystrophy (MMD) (Also which has a dissociation constant (K) for myostatin binding 25 known as Steinert's Disease), Oculopharyngeal Muscular that is at least 2 times less than its K for binding BMP4, and Dystrophy (OPMD), Distal Muscular Dystrophy (DD), and even more preferably at least 5, 10, 100 or even 1000 times Congenital Muscular Dystrophy (CMD). less. Whether by virtue of binding kinetics or biodistribution, Exemplary motor neuron diseases that can be treated with the subject variant Cerberus/Coco polypeptide can also be a regimen including the Subject myostatin include: Amyo selected based on relative in vivo potency, Such as an inhibitor 30 trophic Lateral Sclerosis (ALS) (Also known as Lou Gehrig's that has an ECso for inhibiting myostatin activity, or a par Disease), Infantile Progressive Spinal Muscular Atrophy ticular physiological consequence (such as promoting muscle (SMA, SMA1 or WH) (Also known as SMA Type 1, Werd growth) that is at least 2 times less than its ECs for inhibiting nig-Hoffman), Intermediate Spinal Muscular Atrophy (SMA BMP4 activities, and even more preferably at least 5, 10, 100 or SMA2) (Also known as SMA Type 2), Juvenile Spinal or even 1000 times less. 35 Muscular Atrophy (SMA, SMA3 or KW) (Also known as In certain preferred embodiments, the variant Cerberus/ SMA Type 3, Kugelberg-Welander), Spinal Bulbar Muscular Coco polypeptide binding domain binds myostatin with a K. Atrophy (SBMA) (Also known as Kennedy's Disease and of 1 uM or less, and more preferably a K of 100 nM, 10 nM X-Linked SBMA), and Adult Spinal Muscular Atrophy or even 1 nM or less. (SMA). In general, the Subject myostatin inhibitor preparations are 40 Exemplary inflammatory myopathies that can be treated suitable for use in a human patients. In preferred embodi with a regimen including the Subject myostatin include: Der ments, the subject preparations of variant Cerberus/Coco matomyositis (PM/DM), Polymyositis (PM/DM), and Inclu polypeptides will be substantially free of pyrogenic materials sion Body Myositis (IBM). So as to be suitable for administration to a human patient. Exemplary diseases of the neuromuscular junction that can In other embodiments, the subject variant Cerberus/Coco 45 be treated with a regimen including the Subject myostatin polypeptides can be administered to non-human animals, par include: Myasthenia Gravis (MG), Lambert-Eaton Syndrome ticularly other mammals. For example, the compounds of the (LES), and Congenital Myasthenic Syndrome (CMS). present disclosure can be given to chickens, turkeys, livestock Exemplary myopathies due to endocrine abnormalities that animals (such as sheep, pigs, horses, cattle, etc.), companion can be treated with a regimen including the Subject myostatin animals (e.g., cats and dogs) or may have utility in aquacul 50 include: Hyperthyroid Myopathy (HYPTM) and Hypothy ture to accelerate growth and improve the protein/fat ratio. To roid Myopathy (HYPOTM). further illustrate, the subject variant Cerberus polypeptides Exemplary diseases of peripheral nerve that can be treated can be used to stimulate growth or enhance feed efficiency of with a regimen including the Subject myostatin include: Char animals raised for meat production to improve carcass qual cot-Marie-Tooth Disease (CMT), Dejerine-Sottas Disease ity, or to increase milk production in dairy cattle. 55 (DS), and Friedreich's Ataxia (FA). Another aspect of the disclosure relates to packaged phar Other exemplary myopathies that can be treated with a maceuticals comprising a pharmaceutical preparation of a regimen including the Subject myostatin include: Myotonia variant Cerberus/Coco polypeptide, as described herein, and Congenita (MC), Paramyotonia Congenita (PC), Central a label or instructions for use in promoting growth of muscle Core Disease (CCD), Nemaline Myopathy (NM), Myotubu tissue in a human patient. 60 lar Myopathy (MTM or MM), and Periodic Paralysis (PP). Still another aspect of the disclosure relates to packaged Exemplary metabolic diseases of muscle that can be pharmaceuticals comprising a pharmaceutical preparation of treated with a regimen including the Subject myostatin a variant Cerberus/Coco polypeptide, as described herein, include: Phosphorylase Deficiency (MPD or PYGM), Acid and a label or instructions for veterinarian use in promoting Maltase Deficiency (AMD), Phosphofructokinase Defi growth of muscle tissue in a non-human mammal. 65 ciency (PFKM), Debrancher Enzyme Deficiency (DBD), Another aspect of the disclosure relates to a method for Mitochondrial Myopathy (MITO), Carnitine Deficiency inhibiting myostatin signal transduction in vivo by adminis (CD), Carnitine Palmityl Transferase Deficiency (CPT), US 7,833,971 B2 7 8 Phosphoglycerate Kinase Deficiency (PGK), Phosphoglyc exposure to human serum for 24 hours at 37°C. The myosta erate Mutase Deficiency (PGAM or PGAMM), Lactate tin antagonist activity may be assessed, for example, in an Dehydrogenase Deficiency (LDHA), and Myoadenylate A204 cell based assay. Deaminase Deficiency (MAD). In certain embodiments, the myostatin antagonist protein The subject method can also be used to prevent, ameliorate 5 comprises a modification with respect to the amino acid or reduce the severity of a metabolic disorder, such as in the sequence of SEQID NO:5 such that cleavage in human serum treatment of obesity or type II diabetes. To further illustrate, is reduced or eliminated. The modification with respect to the the subject variant Cerberus/Coco polypeptide preparations amino acid sequence of SEQID NO:5 may reduce or elimi can be used to decrease body fat proportion in a subject. nate cleavage within one or both of the following sequences: In still other embodiments, the variant Cerberus/Coco 10 PARKRW or SRRRVK of human Coco. polypeptide preparations can be used as part of such methods In certain embodiments, the myostatin antagonist protein as reducing frailty associated with aging. may be a fusion protein including one additional polypeptide The Subject pharmaceutical composition can also be used portion that enhances one or more of in vivo stability, in vivo as myostatin antagonist to treat a number of neuronal system half life, uptake? administration, tissue localization or distri disease conditions, including CNS injuries/disease Such as 15 bution, formation of protein complexes, and/or purification. spinal cord injury and stroke, and PNS injuries/diseases. In certain embodiments, the fusion protein includes a portion of an immunoglobulin heavy chain constant domain. In cer In one aspect, the disclosure provides a myostatin antago tain embodiments, the fusion protein comprises an Fc domain nist protein comprising an amino acid sequence that is at least of an immunoglobulin. In certain embodiments, the myosta 90% identical to the sequence of amino acids 162-241 of 20 tin antagonist protein includes one or more modified amino human Cerberus (SEQID NO:2), and wherein said protein is acid residues selected from: a glycosylated amino acid, a substantially serum stable for a period of at least 24 hours. PEGylated amino acid, a farnesylated amino acid, an acety In certain embodiments, the myostatin antagonist protein lated amino acid, a biotinylated amino acid, an amino acid comprises an amino acid sequence that is at least 90%. 95%, conjugated to a lipid moiety, and an amino acid conjugated to 96%, 97%, 98%, 99%, or 100% identical to one or more of the 25 an organic derivatizing agent. following: the sequence of amino acids 156-241 of human In certain embodiments, the myostatin antagonist protein Cerberus, the sequence of amino acids 156-267 of human is a fusion protein that further comprises a second myostatin Cerberus, the sequence of amino acids 141-241 of human inhibitor domain, which is a polypeptide affinity reagent that Cerberus, the sequence of amino acids 141-267 of human selectively binds to myostatin and competes with the binding Cerberus, the sequence of amino acids 119-241 of human 30 of an ALK7 or ALK4 receptor. In certain embodiments, the Cerberus, the sequence of amino acids 41-241 of human affinity reagent is one or more of the following: (i) an anti Cerberus, the sequence of amino acids 41-267 of human body agent, (ii) a peptide or scaffolded peptide that selec Cerberus, the sequence of amino acids 18-241 of human tively binds to myostatin and competes with the binding of an Cerberus, or the sequence of amino acids 18-267 of human ALK7 or ALK4 receptor, (iii) a myostatin binding domain of Cerberus. 35 ALK7 or ALK4, or (iv) small organic molecule that selec In certain embodiments, the myostatin antagonist protein tively binds to myostatin and competes with the binding of an retains at least 50% of the myostatin antagonist activity after ALK7 or ALK4 receptor. Examples of suitable antibody exposure to human serum for 24 hours at 37°C. The myosta agents include, for example, a recombinant antibody; a tin antagonist activity may be assessed, for example, in an monoclonal antibody; a V. domain; a V, domain; an scFV; an A204 cell based assay. 40 Fab fragment; an Fab' fragment; an F(ab'); an Fv, or a dis In certain embodiments, the myostatin antagonist protein ulfide linked Fv. In certain embodiments, the antibody agent comprises a modification with respect to the amino acid is a fully human antibody or a humanized chimeric antibody, sequence of SEQID NO:2 such that cleavage in human serum or an antigen binding fragment thereof. is reduced or eliminated. The modification with respect to the In another aspect, the disclosure provides a pharmaceutical amino acid sequence of SEQID NO:2 may reduce or elimi 45 preparation comprising one or more of the myostatin antago nate cleavage within one or more of the following sequences: nist proteins described herein. the sequence SHCLPAK (SEQ ID NO: 22) of human Cer In another aspect, the disclosure provides a method for berus, the sequence MFRKTP (SEQ ID NO. 23) of human inhibiting myostatin and/or GDF11 and/or Nodal in a patient, Cerberus, or the sequence NQRELP (SEQ ID NO: 24) of 50 the method comprising administering to the patient an effec human Cerberus. tive amount of one or more of the myostatin antagonist pro In another aspect, the disclosure provides a myostatin teins described herein. In certain embodiments, inhibiting antagonist protein, the protein comprising an amino acid myostatin and/or GDF11 and/or Nodal in a patient causes a sequence that is at least 90% identical to the sequence of detectable change in the expression of a gene that is regulated amino acids 101-185 of human Coco (SEQ ID NO:5), and 55 by myostatin and/or GDF11 and/or Nodal. wherein said protein is substantially serum stable for a period In another aspect, the disclosure provides a method for of at least 24 hours. increasing muscle mass in a patient, the method comprising In certain embodiments, the myostatin antagonist protein administering to the patient an effective amount of one or comprises a modification with respect to the amino acid more of the myostatin antagonist proteins described herein. sequence of SEQIDNO:5 such that cleavage in human serum 60 In another aspect, the disclosure provides a pharmaceutical is reduced or eliminated. The modification with respect to the preparation Substantially free of pyrogenic materials, com amino acid sequence of SEQID NO:5 may reduce or elimi prising a myostatin antagonist protein including a myostatin nate cleavage within one or both of the following sequences: binding domain of a Cerberus or Coco polypeptide or variant PARKRW (SEQ ID NO:25) or SRRRVK (SEQ ID NO: 26) thereof, which myostatin antagonist protein: (a) binds to and of human Coco. 65 inhibits the signaling activity of one or more of Nodal, In certain embodiments, the myostatin antagonist protein GDF11 and/or myostatin; and (b) does not substantially bind retains at least 50% of the myostatin antagonist activity after to BMP4. US 7,833,971 B2 9 10 In certain embodiments, the myostatin antagonist protein patient causes a detectable change in the expression of a gene promotes growth of muscle tissue. that is regulated by myostatin and/or GDF11. In certain In certain embodiments, the myostatin binding domain has embodiments, the myostatin antagonist does not substantially an amino acid sequence that is at least 90%. 95%, 96%, 97%, bind to BMP4. 98%, 99%, or 100% identical to one or both of the following: In another aspect, the disclosure provides a method for amino acids 162-241 of SEQID NO: 2, amino acids 101-189 increasing skeletal muscle mass in a patient in need thereof, of SEQ ID NO:5. In certain embodiments, the myostatin the method comprising administering to the patient an effec binding domain has an amino acid sequence that is at identical tive amount of a myostatin antagonist protein including a to an amino acid sequence selected from the group consisting myostatin binding domain of a Cerberus or Coco polypeptide of amino acids 162-241 of SEQID NO: 2 and amino acids 10 or variant thereof, which myostatin antagonist protein binds 101-189 of SEQID NO:5, or any naturally occurring human to and inhibits the signaling activity of one or more of nodal, allelic variant thereof. GDF11 and/or myostatin. In certain embodiments, the myo In certain embodiments, the myostatin antagonist protein statin binding domain has an amino acid sequence that is at does not include a full-length mature human Cerberus pro least 80% identical to an amino acid sequence selected from tein. 15 the group consisting of amino acids 162-241 of SEQID NO: In certain embodiments, the myostatin antagonist protein 2 and amino acids 101-189 of SEQ ID NO:5. In certain is a fusion protein including one additional polypeptide por embodiments, the myostatin binding domain has an amino tion that enhance one or more of in vivo stability, in vivo half acid sequence that is at identical to an amino acid sequence life, uptake? administration, tissue localization or distribution, selected from the group consisting of amino acids 162-241 of formation of protein complexes, and/or purification. In cer SEQID NO: 2 and amino acids 101-189 of SEQID NO:5, or tain embodiments, the fusion protein includes a portion of an any naturally occurring human allelic variant thereof. In cer immunoglobulin heavy chain constant domain. In certain tain embodiments, the myostatin antagonist does not substan embodiments, the fusion protein comprises an Fc domain of tially bind to BMP4. an immunoglobulin. In another aspect, the disclosure provides use of a myosta In certain embodiments, the myostatin antagonist protein 25 tin antagonist protein including a myostatin antagonist pro includes one or more modified amino acid residues selected tein including a myostatin binding domain of a Cerberus or from: a glycosylated amino acid, a PEGylated amino acid, a Coco polypeptide or variant thereof, which myostatin antago farnesylated amino acid, an acetylated amino acid, a biotiny nist protein binds to and inhibits the signaling activity of one lated amino acid, an amino acid conjugated to a lipid moiety, or more ofnodal. GDF11 and/or myostatin for the preparation and an amino acid conjugated to an organic derivatizing 30 of a medicament for promoting growth of muscle tissue in a agent. mammal. In certain embodiments, the myostatin antagonist protein is a fusion protein that further comprises a second myostatin BRIEF DESCRIPTION OF THE DRAWINGS inhibitor domain, which is a polypeptide affinity reagent that selectively binds to myostatin and competes with the binding 35 FIG. 1 shows a schematic drawing of where Wnt, Nodal of an ALK7 or ALK4 receptor. In certain embodiments, the and BMP bind to Cerberus. BMP-2 and the highly related affinity reagent is one or more of the following: (i) an anti BMP4 competitively bind Cerberus, likely in the same body agent, (ii) a peptide or scaffolded peptide that selec region. Other more distantly related or unrelated proteins, tively binds to myostatin and competes with the binding of an such as TGF-beta1, EGF, and PDGF, do not compete with ALK7 or ALK4 receptor, (iii) a myostatin binding domain of 40 BMP-4. The N-terminally truncated version of Cerberus still ALK7 or ALK4, or (iv) a small organic molecule that selec binds Xnr-1 (Xenopus homolog of mouse Nodal). (Adapted tively binds to myostatin and competes with the binding of an from Piccolo et al., Nature 397: 707-710, 1999). ALK7 or ALK4 receptor. Examples of suitable antibody FIG.2 Binding of Caronte-Fc to 11. The tracing shows that agents include, for example, a recombinant antibody; a Caronte-Fe binds to 11 on a BiaCore chip. 11 was immobi monoclonal antibody; a V. domain; a V, domain; an Schv; an 45 lized on a BiaCore CM5 chip using standard amine coupling Fab fragment; an Fab' fragment; an F(ab'); an Fv, or a dis procedure. Trace: Caronte-Fc (200 ug/ml, R&D Systems) ulfide linked Fv. In certain embodiments, the antibody agent was injected on the 11 coupled chip. is a fully human antibody or a humanized chimeric antibody, FIG. 3 A-204 Reporter Gene Assay. The figure shows the or an antigen binding fragment thereof. Reporter vector: pGL3(CAGA)12 (described in Dennileretal, In another aspect, the disclosure provides a method for 50 1998, EMBO 17:3091-3100.) The CAGA12 motif (SEQ ID inhibiting myostatin and/or GDF11 in a patient, the method NO: 30) is present in TGF-Beta responsive genes (PAI-1 comprising administering to the patient an effective amount gene), so this vector is of general use for factors signaling of a myostatin antagonist protein including a myostatin bind through Smad2 and 3. ing domain of a Cerberus or Coco polypeptide or variant FIG. 4 Caronte-Fe inhibits 11 signaling in the A-204 thereof, which myostatin antagonist protein binds to and 55 Reporter Gene Assay. An ActRIIA-Fc (IIA muG2a) fusion inhibits the signaling activity of one or more ofnodal. GDF11 also inhibits 11 signaling. and/or myostatin. In certain embodiments, the myostatin FIG. 5 Caronte-Fe does not inhibit Activin A in the A-204 binding domain has an amino acid sequence that is at least Reporter Gene Assay. An ActRIIA-Fc fusion (“IIA muG2a), 90% identical to an amino acid sequence selected from the as expected, does inhibit Activin A signaling. group consisting of amino acids 162-241 of SEQID NO: 2 60 FIG. 6 Cerberus-FC and Caronte-Fo both inhibit GDF-8 and amino acids 101-189 of SEQID NO:5. In certain embodi signaling in the A-204 Reporter Gene Assay. ments, the myostatin binding domain has an amino acid FIG. 7 Human Cerberus-Fc is degraded in human serum. sequence that is at identical to an amino acid sequence Conditioned medium from cells expressing human Cerberus selected from the group consisting of amino acids 162-241 of Fc was incubated overnight at 37 deg. C. with varying SEQID NO: 2 and amino acids 101-189 of SEQID NO:5, or 65 amounts of human serum (percentages of serum added are any naturally occurring human allelic variant thereof. In cer shown at top), and resolved by SDS-PAGE. Cerberus was tain embodiments, inhibiting myostatin and/or GDF11 in a detected by Western blot with a primary antibody: biotiny US 7,833,971 B2 11 12 lated polyclonal anti-cerberus, and a secondary antibody: closure and in the specific context where each term is used. avidin-HRP. The left lane is molecular weight standards. The Certain terms are discussed below or elsewhere in the speci major band, at roughly 70 kD is Cerberus-Fc, which is com fication, to provide additional guidance to the practitioner in pletely degraded when incubated with 5% human serum. describing the compositions and methods of the disclosure FIG. 8 Human Coco-Fc (murine Fc) inhibits 11 signaling and how to make and use them. The Scopean meaning of any in a cell based assay. Conditioned medium from cells express use of a term will be apparent from the specific context in ing human Coco-mFc was tested for effects on A-204 reporter which the term is used. gene expression in the presence of 11. About' and “approximately' shall generally mean an DETAILED DESCRIPTION acceptable degree of error for the quantity measured given the 10 nature or precision of the measurements. Typically, exem I. Overview plary degrees of error are within 20 percent (%), preferably within 10%, and more preferably within 5% of a given value Cerberus is expressed in the anterior endomesoderm (Bou or range of values. wmeester et al., Nature 382: 595-601, 1996: Piccolo et al., 15 Alternatively, and particularly in biological systems, the Nature 397: 707-10, 1999; Rodriguez et al., Nature 401: terms “about and “approximately” may mean values that are 243-51, 1999) during development. Caronte, a chick within an order of magnitude, preferably within 5-fold and ortholog, is involved in left-right asymmetry in the chick more preferably within 2-fold of a given value. Numerical embryo (Rodriguez, Supra). Cerberus functions as a multiva quantities given herein are approximate unless stated other lent growth factor antagonist in the extracellular space and wise, meaning that the term “about' or “approximately” can inhibits signaling by BMP-4, nodal, and Wnt (Belo et al., be inferred when not expressly stated. Genesis 26:265-70, 2000). Mouse Cerberus binds to BMP proteins and nodal via independent sites (Piccolo, Supra), The methods of the disclosure may include steps of com whereas the Xenopus Cerberus also binds Wnt proteins and paring sequences to each other, including wild-type sequence inhibits their actions (Belo, supra). Cerberus has the unique to one or more mutants/sequence variants Such comparisons property of inducing ectopic heads in the absence of trunk 25 typically comprise alignments of polymer sequences, e.g., structures (Piccolo, supra). The expression of Cerberus dur using sequence alignment programs and/or algorithms that ing gastrulation is activated by nodal-related signals in endo are well known in the art (for example, BLAST. FASTA and derm and by Spemann-organizer factors (Yamamoto et al., MEGALIGN, to name a few). The skilled artisan can readily Dev Biol 257: 190-204, 2003). appreciate that, in Such alignments, where a mutation con Orthologs for Cerberus can be found in Xenopus tropicalis 30 tains a residue insertion or deletion, the sequence alignment and Fugu rubripes, but are missing in invertebrates. In Fugu will introduce a 'gap' (typically represented by a dash, or rubripes, there is only one ortholog for Cerberus. All ortholo 'A') in the polymer sequence not containing the inserted or gous genes for Cerberus have two exons; the first eight amino deleted residue. acids of the cystine-knot domain are encoded by the 3' end of “Homologous.” in all its grammatical forms and spelling the first exon and the remainder of the motif by the second variations, refers to the relationship between two proteins that exon. In some orthologs, a predicted proteolytic cleavage site possess a "common evolutionary origin, including proteins can be found upstream of the beginning of the cystine-knot from Superfamilies in the same species of organism, as well as domain. homologous proteins from different species of organism. Coco is another member of the Cerberus/Dan family of Such proteins (and their encoding nucleic acids) have proteins that inhibits Nodal signaling. 40 sequence homology, as reflected by their sequence similarity, In part, the present disclosure provides Coco or Cerberus whether in terms of percent identity or by the presence of derivatives for inhibiting Nodal, 11 and/or myostatin func specific residues or motifs and conserved positions. tion. In certain embodiments, the Coco and Cerberus deriva The term "sequence similarity.” in all its grammatical tives inhibit Nodal, 11 and/or myostatin function without forms, refers to the degree of identity or correspondence substantially compromising BMP (such as BMP-4) signaling 45 between nucleic acid or amino acid sequences that may or (e.g., does not substantially bind BMP-4 or other BMPs). The may not share a common evolutionary origin. subject Cerberus derivatives may also be used to inhibit BMP However, in common usage and in the instant application, (such as BMP-4) signaling. the term “homologous,” when modified with an adverb such Exemplary preparations of the Subject disclosure include as “highly, may refer to sequence similarity and may or may Cerberus polypeptide derivatives, including the N-terminal 50 not relate to a common evolutionary origin. truncated versions of Cerberus or Coco. These so-called A nucleic acid molecule is “hybridizable' to another “Cerberus derivatives” or “Coco derivatives' can be used to nucleic acid molecule. Such as a cDNA, genomic DNA, or reduce the severity of a pathologic condition, which is char RNA, when a single stranded form of the nucleic acid mol acterized, at least in part, by an abnormal amount, develop 55 ecule can anneal to the other micleic acid molecule under the ment or metabolic activity of muscle or adipose tissue in a appropriate conditions of temperature and solution ionic Subject. For instance, the pharmaceutical preparations of the strength (see Sambrook et al. Molecular Cloning. A Labora present disclosure can be administered in an amount effective tory Manual, Second Edition (1989) Cold Spring Harbor to prevent, ameliorate or reduce the severity of a wasting Laboratory Press, Cold Spring Harbor, N.Y.). The conditions disorder, such as cachexia, anorexia, DMD syndrome, BMD 60 oftemperature and ionic strength determine the 'stringency’ syndrome, AIDS wasting syndrome, muscular dystrophies, of the hybridization. For preliminary screening for homolo neuromuscular diseases, motor neuron diseases, diseases of gous nucleic acids, low Stringency hybridization conditions, the neuromuscular junction, and inflammatory myopathies. corresponding to a T (melting temperature) of 55° C. can be II. Definitions used, e.g., 5xSSC, 0.1% SDS, 0.25% milk, and no forma 65 mide; or 30% formamide, 5xSSC, 0.5% SDS). The terms used in this specification generally have their Moderate stringency hybridization conditions correspond ordinary meanings in the art, within the context of this dis to a higher T, e.g., 40% formamide, with 5x or 6xSSC. High US 7,833,971 B2 13 14 stringency hybridization conditions correspond to the highest in a more preferred embodiment, the T is 65°C. In a specific T. e.g., 50% formamide, 5x or 6xSSC. SSC is 0.15 MNaCl, embodiment, “high stringency” refers to hybridization and/or 0.015 MNa-citrate. washing conditions at 68°C. in 0.2xSSC, at 42°C. in 50% “High stringency condition' is well understood in the art to formamide, 4xSSC, or under conditions that afford levels of encompass conditions of hybridization which allow hybrid hybridization equivalent to those observed under either of ization of structurally related, but not structurally dissimilar, these two conditions. nucleic acids. The term “stringent' is a term of art which is Suitable hybridization conditions for oligonucleotides understood by the skilled artisan to describe any of a number (e.g., for oligonucleotide probes or primers) are typically ofalternative hybridization and wash conditions which allow Somewhat different than for full-length nucleic acids (e.g., annealing of only highly complementary nucleic acids. 10 full-length cDNA), because of the oligonucleotides lower Exemplary high Stringent hybridization conditions is melting temperature. Because the melting temperature of oli equivalent to about 20-27°C. below the melting temperature gonucleotides will depend on the length of the oligonucle (T) of the DNA duplex formed in about 1 M salt. Many otide sequences involved, Suitable hybridization tempera equivalent procedures exist and several popular molecular tures will vary depending upon the oligonucleotide molecules cloning manuals describe Suitable conditions for stringent 15 used. Exemplary temperatures may be 37° C. (for 14-base hybridization and, furthermore, provide formulas for calcu oligonucleotides), 48°C. (for 17-base oligonucleotides), 55° lating the length of hybrids expected to be stable under these C. (for 20-base oligonucleotides) and 60° C. (for 23-base conditions (see e.g. Current Protocols in Molecular Biology, oligonucleotides). Exemplary Suitable hybridization condi John Wiley & Sons, N.Y. (1989), 6.3.1-6 or 13.3.6; or pages tions for oligonucleotides include washing in 6xSSC, 0.05% 9.47-9.57 of Sambrook, et al. (1989) Molecular Cloning. 2" Sodium pyrophosphate, or other conditions that afford ed., Cold Spring Harbor Press). equivalent levels of hybridization. Hybridization requires that the two nucleic acids contain “Polypeptide.” “peptide' or “protein’ are used inter complementary sequences, although depending on the strin changeably to describe a chain of amino acids that are linked gency of the hybridization, mismatches between bases are together by chemical bonds called "peptide bonds.” A protein possible. The appropriate stringency for hybridizing nucleic 25 or polypeptide, including an enzyme, may be a “native' or acids depends on the length of the nucleic acids and the “wild-type.” meaning that it occurs in nature; or it may be a degree of complementation, variables well known in the art. “mutant.” “variant, or “modified, meaning that it has been The greater the degree of similarity or homology between two made, altered, derived, or is in Some way different or changed nucleotide sequences, the greater the value of T, for hybrids from a native protein or from another mutant. of nucleic acids having those sequences. The relative stability 30 As used herein, the term “Cerberus/Coco protein' is used (corresponding to higher T) of micleic acid hybridizations to signify the human Cerberus and Coco proteins, as well as decreases in the following order: RNA:RNA, DNA:RNA, homologs from other species (e.g., Caronte is the chicken DNA:DNA. For hybrids of greater than 100 nucleotides in Cerberus homolog) and derivatives (including forms with length, equations for calculating T, have been derived (see altered sequences and truncated forms) that retain a biologi Sambrook et al., Supra, 9.51). For hybridization with shorter 35 cal activity of the naturally occurring form. nucleic acids, i.e., oligonucleotides, the position of mis “Cerberus or Cerberus-like protein’ refers to mammalian matches becomes more important, and the length of the oli Cerberus and Cerberus-like proteins, such as the murine gonucleotide determines its specificity (see Sambrook et al., (NCBI RefSeq, ID NP 034017) or human (NCBI RefSeq, ID supra, 11.8). A minimum length for a hybridizable nucleic NP 005445) Cerberus proteins (also see SEQID Nos. 2 and acid is at least about 10 nucleotides; preferably at least about 40 8, respectively, of US 2002/0164682 A1, the entire contents 15 nucleotides; and more preferably the length is at least of which is incorporated herein by reference), and other pro about 20 nucleotides. teins which share sequence homology to the highly conserved Unless specified, the term “standard hybridization condi cysteine pattern of the C-terminal portion of the mammalian tions’ refers to a T of about 55° C., and utilizes conditions as Cerberus proteins. Exemplary amino acid sequences for Cer set forth above. In a preferred embodiment, the T is 60°C.; berus proteins include

Murine Cerberus protein (NCBI RefSeq, ID NP_034017) (SEQ ID NO: 1) : MHLLLVOLLW LLPLGKADLC WDGCOSOGSL SFPLLERGRR DLHVANHEEA. EDKPDLFWAV

6 PHLMGTSLAG EGORORGKML SRLGRFWKKP ETEFYPPRDV ESDHVSSGMO AVTOPADGRK

12 VERSPLOEEA KRFWERFMFR KGPAFOGVIL PIKSHEVHWE TCRTVPFNOT IAHEDCOKVV

18 VONNLCFGKC SSIRFPGEGA DAHSFCSHCS PTKFTTWHLM LNCTSPTPVV KMVMOVEECO

24 CMVKTERGEE RLLLAGSOGS FIPGLPASKT NP Human Cerberus protein (NCBI RefSeq, ID NP_005445) (SEQ ID NO: 2) : MHLLLFOLLW LGKTTRH ODGRONOSSL SPWLLPRNOR ELPTGNHEEA EEKPDLFVAV

6 PHL WATSPAG EGO ROREKML SRFGRFWKKP EREMHPSRDS DSEPFPPGTO SLIOPIDGMK

12 MEKSPLREEA KK PASOGVIL PIKSHEVHWE TCRTVPFSQT ITHEGCEKVV

18 VONNLCFGKC HFPGAAO HSHTSCSHCL PAKFTTMHLP LNCTELSSVI KVVMLVEECO

24 CKWKTEHEDG HI HAGSQDS FIPGVSA

US 7,833,971 B2 17 18

- Continued NCBI RefSeq, ID NM_005454. 1 (human Cerberus mRNA) (SEQ ID NO : 4). atgcatc toc tottatttca gotgctggta ct cotgcct c taggaaagac cacacggcac 61 caggatggcc gcc agaatca gagttct citt toccc.cgitac toctogccaag gaatcaaaga 121 gagct tcc.ca caggcaacca taggaagct gaggaga agc Cagatctgtt ttcgcagtg 181 ccacaccittg tagccaccag ccctgcaggg galaggccaga ggcagagaga galagatgctg 241 t ccagatttg gCaggttctg. galagaagcct gagagagaaa to atcCat C Cagggactica 301 gat agtgagc cct tcccacc tdggacc cag tocct catcc agc.cgataga tiggaatgaaa 361 atggagaaat citcct citt.cg ggaagaa.gcc aagaaattct ggcaccactt catgttcaga 421 aaaactic cqg Cttct caggg ggit Catcttg cc catcaaaa gC catgaagt acattgggag 481 acctgcagga cagtgcc ctt cagccagact at aacccacg aaggctgttga aaaagtagtt 541 gttcagaaca acctittgctt togaaatgc gggtctgttc atttitcc tig agcc.gc.gcag 601 cactic ccata cct cotgctic toactgtttg cc toccaagt to accacgat go acttgc.ca 661 ctgaactgca citgaactitt C Ctc.cgtgat C aaggtggtga tigctggtgga ggagtgc.ca.g 721 to alaggtga agacggagca talagatgga cacat Cotac atgctggct C cc aggattic C 781 tittatcc cag gag titt cagc titga

It is also expected that Cerberus related proteins also exist tein of GenBank Accession 22749329, and other proteins in other species, including family members in Xenopus, and which share sequence homology to the highly conserved Drosophila, C. elegans, Zebrafish, as well as in all mammals, 30 cysteine pattern of the C-terminal portion of the mammalian for example, rats, mice and humans. “Cerberus or Cerberus Coco proteins. An exemplary amino acid sequences for like proteins” also includes variants of the Cerberus proteins, human Coco protein is

(SEO ID NO; 5) 1 MLLGOLSTLL CLLSGALPTG SGRPEPOSPR POSWAAANOT WALGPGALPP LVPASALGSW

61 KAFLGLOKAR OLGMGRLORG ODEWAAVTLP LNPOEVIOGM. CKAVPFWOWF SRPGCSAIRL

121 RNHLCFGNCS SLYIPGSDPT PLWLCNSCMP ARKRWAPWWL WCLTGSSASR RRWKISTMLI

181 EGCHCSPKA

Such as allelic variants or variants induced by mutagenesis or Amino acids 1-21 of SEQID NO:5 correspond to a signal deletions, and fragments of Cerberus proteins which variants 45 peptide that may be replaced with an alternative leader and fragments retain myostatin binding activity. “Cerberus sequence. A mature secreted Coco polypeptide is expected to like' proteins is also used to signify the family of proteins correspond to amino acids 22-189 of SEQID NO:5, although sharing structural and/or functional similarity, including imprecisions in the signal peptide processing enzymes may those proteins which are described further herein. Such pro 50 lead to alternative or additional cleavage at positions ranging teins may have amino acid sequences sharing significant from one to five amino acids towards the amino terminus or sequence identity (e.g., at least about 50%. 60%, 70%, 80%, the carboxy terminus from the glycine at position 22. As 90%. 95%, 99% or more) with the human or mouse Cerberus proteins, over the full-length, or at least within the myostatin disclosed herein, atPA leader sequence or other heterologous leader sequence may be used in place of the native leader binding domain of the human or mouse Cerberus. Cerberus 55 like proteins also include proteins that have amino acid sequence. Proposed leader sequences are as follows: sequences that are encoded by nucleic acid sequences that hybridize under stringent conditions with the coding sequences for human or mouse Cerberus, particularly that (i) Honey bee mellitin (HBML) : portion of the coding sequence for the myostatin binding 60 MKFLWNWALWFMWWYISYIYA (SEQ ID NO: 14) domain. A Cerberus derivative or variant sequence may or may not lack the N-terminal BMP binding domain. A variety (ii) Tissue Plasminogen Activator (TPA) : of allelic variants of human Cerberus are known, including MDAMKRGLCCWLLLCGAWFWSP (SEQ ID NO: 15) A65G (alanine 65 to glycine), V179I and L221 V. 65 The human Coco coding sequence is disclosed in GenBank “Coco or Coco-like protein’ refers to mammalian Coco Accession 22749328 (incorporated herein by reference) proteins and related homologs, such as the human Coco pro (SEQID NO:6). US 7,833,971 B2 19 20

agt ccggaca gacagaCagg cagacagacg cacggacaag Cagatgctcc ttggc.cagct 61 atccact Ctt Ctgtgcctgc titagcggggc cctgcctaca ggct Caggga ggcctgaac C 121 C cagt ct cot Cacct cagt cctgggctgc agccaat cag acctgggctic tiggccCagg 181 ggCCCtgcc C C Cactggtgc cagct tctgc ccttgggagc tiggaaggcct tcttgggcct 241 goagaaagcc aggcagctgg ggatgggcag gCtgcagcgt gggcaagacg aggtggctgc 301 ttgactctg. cc.gctgaac C. Ctcaggaagt gatcCagggg atgtgta agg Ctgtgc cctt 361 cqttcaggtg ttct c ccggc ccggctgctic agccatacgc ct cogaaatc atctgtgctt 421 togg to attgc ticcitct citct acatc cctdg ct cqgacccc acco cactag to ctgtgcaa. 481 cagctgt atg cct gctcgca agcgttgggc accc.gtggit C ctgtggtgtc. tcactggcag 541 ct cagcctcc cqtcgacggg taagat at C Caccatgctg atcgaggggt gt cactgcag 601 cccaaaagca taactgagc atcgtggatg ggtgcacgga gacacgcacc ttggagaaat 661 gaggggagat ggaccaagaa agacgtggac Ctggatgatg tact ctgggit caagagacca 721 gggatgcagg gttaggcaga Caggit cocca gagt cct cac cctgctic cc C agacagtaga 781 cacagtgc.cc gtc.ctggagt togcaccact g at agt cacag cacacaatga ttgacaactic 841 acttitttittt tttitttittga gatggagt ct cqctctgtcg cc caggctgg agtgcagtgg 901 cqcaatcto a gct cactgca agctic cacct cocqqgttta togcc attctic ct gt ct cagc 961 ctic cc.gagta gctgggacta caggcacccg ccaac acgcc cqgctaattt tt cqtattitt O21 tagtaaagac agggittt cac cqtgttagcc aggatggtct ct at CtcCtg acct cqtgat 081 ctgcctgcct tdgccttatt atttitttittt tttaaggaca gagtictotct ctogtoaccca 141 ggctggagtg caatggcgcg atcttggctic actgtaactt CC acttgcca ggct calagca 2O1 gttct cotgc ct cagcct cc tagtagctg ggact acagg caccc.gc.cac catgcc cagc 261 taattitttgt atttittagta gagacagagt tt caccatat tagcctggct gg.tctcaaac 321 t cctggcct C aggtgatctg. cccacct cq9 cc toccaaag tectgggat C aaatcc actg 381 ttaat catta ggctgaactg. tct cittatagaatgaggtoa aaga cactcc cagttgcagg 441 gaggg tagat ggc.cccaccc agaccgagag acacagtgat gaccticago C taggga cacc 5O1 aaaaaaaaaa aaaaaaaaaa cccaaac caa aaacgcaaac caaag.caggc aggcagacag 561 ctgctggggg aaatcctggg gtcCttgaga cagaggcagg accctctgt toccagctgc 621 ct cittgcctt gatagtggtg ctgtgtc. cct ct cagacccc ccacctgagt ct coacagag 681 ccc cacgcct ggcatgg cat to cacagaaa ccatalaaggit toggctgagtic c

It is also expected that Coco-related proteins also exist in 50 encoded by nucleic acid sequences that hybridize under strin other species, including family members in Xenopus, and gent conditions with the coding sequences for human Coco, Drosophila, C. elegans, Zebrafish, as well as in all mammals, particularly that portion of the coding sequence for the myo for example, rats, mice and non-human primates. "Coco or statin binding domain. A Coco derivative or variant sequence Coco-like proteins also includes variants of the naturally may or may not lack the N-terminal BMP binding domain. occurring Coco proteins, such as allelic variants or variants 55 Unless specifically stated otherwise, “Cerberus (deriva induced by mutagenesis or deletions, and fragments of Coco tive) therapeutics' or its grammatical variations include the proteins which variants and fragments retain myostatin bind full-length or the N-terminally truncated versions of Cerberus ing activity. "Coco-like' proteins is also used to signify the therapeutics. family of proteins sharing structural and/or functional simi 60 As used herein, the term “Cerberus or Cerberus-like activ larity, including those proteins which are described further ity” refers to one or more of the activities which are exhibited herein. Such proteins may have amino acid sequences sharing by the mammalian Cerberus-like proteins of the present dis significant sequence identity (e.g., at least about 50%. 60%, closure. In particular, “Cerberus or Cerberus-like activity” 70%, 80%, 90%, 95%, 99% or more) with the human Coco includes the ability to induce, enhance and/or inhibit the protein, over the full-length, or at least within the myostatin 65 formation, growth, proliferation, differentiation, mainte binding domain of the human Coco. Coco-like proteins also nance of neurons and/or related neural cells and tissues such include proteins that have amino acid sequences that are as brain cells, Schwann cells, glial cells and astrocytes. "Cer US 7,833,971 B2 21 22 berus or Cerberus-like activity also includes the ability to myostatin under Such conditions requires an antibody that is induce molecular markers of neuroendocrine or ectoderm selected for its specificity for a particular protein. The affinity tissue, such as OTX2, N-CAM, MASH, chromagranin, and constant (Ka, as opposed to Kd) of the antibody binding site AP2, as well as the ability to induce the formation of neurons for its cognate monovalent antigen is at least 10", usually at and/or related neural cells and tissues such as brain cells, least 10, preferably at least 10, more preferably at least Schwann cells, glial cells and astrocytes. “Cerberus or Cer 10", and most preferably at least 10'M. A variety of immu berus-like activity” may also include the ability to regulate noassay formats are appropriate for selecting antibodies spe the interaction of ligands and their protein receptors. “Cer cifically reactive with myostatin. For example, Solid-phase berus or Cerberus-like activity” may further include the abil ELISA immunoassays are routinely used to select mono ity to regulate the formation, differentiation, proliferation 10 clonal antibodies specifically reactive with a protein. See and/or maintenance of other cells and/or tissue, for example Harlow and Lane (1988) Antibodies, A Laboratory Manual, connective tissue, organs and wound healing. In particular, Cold Spring Harbor Publications, New York, for a description “Cerberus or Cerberus-like activity” may include the ability of immunoassay formats and conditions that can be used to to enhance and/or inhibit the formation, growth, proliferation, determine specific reactivity. differentiation and/or maintenance of cardiac, spleen, liver, 15 Immunoassays in the competitive binding format can be pancreas, stomach, kidney, lung and brain cells and tissue, as used to determine cross-reactivity of antibodies with myosta well as osteoblasts and bone, chondrocytes and cartilage, tin, e.g., to identify whether a test antibody is a myostatin tendon, epidermis and muscle. “Cerberus and Cerberus-like neutralizing antibody. For example, the myostatin protein, or activity” also includes the activities of Cerberus and Cer a fragment thereof is immobilized to a solid support. Test berus-like protein in the assays described in the examples and antibodies are added to the assay compete with the binding of specification herein. a TGF receptor, such as ActRII or ALK7, to the immobilized Cerberus and Cerberus-like nucleotide sequences in mouse antigen. The ability of the testantibodies to compete with the and human areas disclosed in US 2002/0164682 A1, as SEQ binding of a TGF receptor to the immobilized myostatin ID NOS. 1 and 7 (incorporated herein by reference). Also see antigen is compared. NCBI RefSeq, ID NM 005454.1 (human) and 25 Similarly, immunoassays in the competitive binding for NM 009887.1 (mouse). mat can be used to determine cross-reactivity determinations, In certain related embodiments, the mysotatin inhibitor is a e.g., to determine the specificity of a myostatin neutralizing polypeptide that includes a myostatin binding domain of a antibody. For example, the myostatin protein, or the myosta Coco protein, such as the human Coco protein. tin epitope thereof is immobilized to a solid support. Epitopes The terms “antibody' and “antibody agent” are used inter 30 from other proteins, such as 11, Nodal or BMP-4 or other changeably herein, and refer to an immunoglobulin molecule proteins having sequence homology with myostatin are obtained by in vitro or in vivo generation of the humoral added to the assay to compete with the binding of a potential response, and includes both polyclonal and monoclonal anti myostatin neutralizing antibody to the immobilized antigen. bodies. The term also includes genetically engineered forms The ability of the test peptides to compete with the binding of Such as chimericantibodies (e.g., humanized murine antibod 35 potential myostatin neutralizing antibody with the immobi ies), heteroconjugate antibodies (e.g., bispecific antibodies), lized myostatin antigen is compared. The percent cross-reac and recombinant single chain Fv fragments (scFV). The term tivity of the potential myostatin neutralizing antibody for the “antibody' also includes antigenbinding forms of antibodies other antigens is calculated, using standard calculations. In (e.g., Fab'. F(ab'), Fab, Fv, rigG, and, inverted IgG). certain preferred embodiments, the Subject myostatin neu The term “antigen binding fragment' includes any portion 40 tralizing antibodies have less than 10% cross-reactivity with of an antibody that binds to a target epitope. An antigen 11. In other preferred embodiments, the subject myostatin binding fragment may be, for example, a polypeptide includ neutralizing antibodies have less than 1%. 5%, or 10% cross ing a CDR3 region, or other fragment of an immunoglobulin reactivity with BMP-4. molecule which retains the affinity and specificity of the myostatin epitope. 45 III. Exemplary Cerberus and Coco Derivatives “Specifically binds' includes reference to the preferential association of a ligand, in whole or part, with a particular In certain embodiments, the mysotatin inhibitor is a Cer target molecule (i.e., “binding partner or “binding moiety') berus polypeptide sharing at least about 50%, 60%, 70%, relative to compositions lacking that target molecule. It is, of 80%, 90%, 95%, 99% or more sequence identity over the course, recognized that a certain degree of non-specific inter 50 full-length of the human or mouse Cerberus protein. action may occur between the Subject myostatin neutralizing In certain other embodiments, the mysotatin inhibitor is a antibodies and a other proteins. Nevertheless, specific bind polypeptide that includes a Cerberus sequence obtained from ing, may be distinguished as mediated through specific rec human, mouse, or other species, their variants or derivatives, ognition of the myostatin protein. Typically specific binding including N-terminally truncated versions of Cerberus. The results in a much stronger association between the antibody 55 full-length mouse and human Cerberus proteins, disclosed as and myostatin protein than between the antibody and other SEQID NOS. 2 and 8, respectively, in US 2002/0164682 A1, proteins, e.g., GDF11. Specific binding by an antibody to are also disclosed in NCBI RefSeq format below:

Human Cerberus full length protein (SEQ ID NO: 2): 1 MHLLLFOLLV LLPLGKTTRH QDGRONOSSL SPWLLPRNOR ELPTGNHEEA EEKPDLFWAV

61 PHL WATSPAG EGOROREKML SRFGRFWKKP EREMHPSRDS DSEPFPPGTQ SLIQPIDGMK 121 MEKSPLREEA KKFWHHFMFR KTPASOGVIL PIKSHEVHWE TCRTVPFSOT ITHEGCEKVV 181 VONNLCFGKC GSVHFPGAAQ HSHTSCSHCL PAKFTTMHLP LNCTELSSVI KVVMLVEECO

241 CKWKTEHEDG HILHAGSODS FIPGWSA US 7,833,971 B2 23 24 Residues 106-119 (from any one of which residues the These same conserved regions may be used to generate Subject Cerberus derivatives may begin), and residues 241 probes for screening nucleic acid libraries at moderate to low 267 (to any one of which residues the subject Cerberus deriva stringency hybridization conditions (see definition section). tives may end) are underlined.

Mouse Cerberus full length protein (SEQ ID NO : 1) : 1 MHLLLWOLLW LLPLGKADLC WDGCOSOGSL SFPLLERGRR DLHVANHEEA. EDKPDLFVAV

61 PHLNGTSLAG EGORQRGKML SRLGRFWKKP ETEFYPPRDV ESDHVSSGMO AVTOPADGRK

121 WERSPLOEEA KRFWHRFMFR KGPAFOGWIL PIKSHEVHWE TCRTVPFNOT IAHEDCOKVV

181 VONNLCFGKC SSIRFPGEGA DAHSFCSHCS PTKFTTVHLM LNCTSPTPVV KMVMOVEECO 241 CMVKTERGEE RLILLAGSQGS FIPGLPASKT NP

Residues 106-119 (from any one of which residues the In certain embodiments, the mysotatin inhibitor is a Cer Subject Cerberus derivatives may begin), and residues 241 berus polypeptide sharing at least about 50%, 60%, 70%, 272 (to any one of which residues the subject Cerberus deriva 80%, 90%, 95%, 99% or more sequence identity over the tives may end) are underlined. Note that the mouse protein is full-length of the human or mouse Cerberus protein. largely homologous to the human protein throughout the In certain other embodiments, the mysotatin inhibitor is a sequences, with the exception of 5 additional residues at the polypeptide that includes a Coco sequence obtained from C-terminus. Therefore, whenever a non-human Cerberus human, mouse, or other species, their variants or derivatives, derivative is used, the residue numbers refers to those corre including N-terminally truncated versions of Coco. The full sponding to the human sequences. 25 length human Coco protein is disclosed above. The various Cerberus and Coco polypeptides may be pre As described above, in certain embodiments, preferred pared as fusion proteins. A fusion protein may include one or fragments of the human Cerberus derivative proteins are ones more additional polypeptideportion that enhance one or more which begins anywhere from residues 106-119 (inclusive) at of in vivo stability, in vivo half life, uptake/administration, the N-terminus, and ends anywhereafter residue 241. A vari 30 tissue localization or distribution, formation of protein com ety of additional Cerberus and Coco derivatives and variants plexes, and/or purification. For example, a fusion protein may are described in the Examples. include a portion of a constant region of a immunoglobulin Also included are Cerberus derived variant sequence, heavy chains, e.g., an immunoglobulin Fc domain, and/or a including mutants or variants of the wild-type myostatin purification Subsequence selected from: an epitope tag, a 35 FLAG tag, a polyhistidine sequence, and a GST fusion. The binding domains that retain myostatin binding activity, myostatin antagonist protein may include one or more modi optionally substantially loses BMP-4 binding. Variant fied amino acid residues selected from: a glycosylated amino sequences without BMP binding affinity may be desirable as acid, a PEGylated amino acid, a farnesylated amino acid, an away to alter selectivity of the inhibitor (e.g., relative to 11 or acetylated amino acid, a biotinylated amino acid, an amino nodal binding, where preferential binding to one of the pro acid conjugated to a lipid moiety, and an amino acid conju teins occur. Also includes more preferential—higher affinity 40 gated to an organic derivatizing agent. than wild-type-binding to myostatin, or more discrimi A fusion protein or coupled protein system (e.g. non-fusion tory—lower affinity than wild-type truncated version bind covalent linkage by crosslinking) may also include a second ing to BMP-4), alter other binding characteristics with respect myostatin inhibitor domain, which is a polypeptide affinity to myostatin (such as K and/or K, or K-rates), or improve 45 reagent that selectively binds to myostatin and competes with biodistribution or half life in vivo or on the shelf. the binding of an ALK7 or ALK4 receptor. The affinity reagent may be an antibody agent. An antibody agent may be, Certain other Cerberus sequences are listed below based on for example, a recombinantantibody; a monoclonal antibody; homology search in databases of identified proteins, and the aVH domain; a VL domain; an schv: an Fab fragment; an Fab' subject variant Cerberus polypeptides can be derived from fragment; an F(ab')2; an Fv, or a disulfide linked Fv, a fully those proteins as well. Since these sequences are retrieved 50 human antibody or a humanized chimeric antibody, or an from public databases available on the internet, additional antigen binding fragment thereof. An affinity reagent is a homologs of the proteins in other species may be obtained as peptide or scaffolded peptide that selectively binds to myo these databases are being updated. Furthermore, other species statin and competes with the binding of an ALK7 or ALK4 of Cerberus proteins, especially those of mammals, can be receptor. An affinity reagent may include a myostatin binding readily obtained by standard molecular biology protocols, 55 domain of ALK7 or ALK4. For example, an extracellular such as PCR, low stringency hybridization, Ab-mediated domain of ALK7 or ALK4 (preferably human ALK7 or screening of expression libraries using antibodies cross-re ALK4) may be used. The affinity reagent may be a small acting with identified Cerberus homologs in target species, organic molecule that selectively binds to myostatin and com etc. petes with the binding of an ALK7 or ALK4 receptor. For example, sequence alignments using Softwares such as 60 An example of a human ALK7 myostatin binding domain DNAStar's MegaAlign (supra) can identify the most con is shown below: served regions in the known members of a protein family. PCR can then be carried out using degenerate oligoes cover (SEO ID NO: 9) ing Such most conserved regions, and templates DNA from LKCVCLLCDSSNFTCOTEGACWASVMLTNGKEQVIKSCVSLPELNAOVFC the target organism. In preferred embodiments, such con 65 served regions include the kinase domain, and/or the ligand HSSNNWTKTECCFTDFCNNITLHLP binding domain. US 7,833,971 B2 25 26 An example of a human ALK4 myostatin binding domain and the course is slower and far less predictable than that of is shown below: DMD. (Though DMD and BMD affect boys almost exclu sively, in rare cases they can affect girls. Until the 1980s, little was known about the cause of any (SEQ ID NO: 10) 5 kind of muscular dystrophy. In 1986, the dystrophin gene ALLCACTSCLOANYTCETDGACMVSIFNLDGMEHHVRTCIPKVELVPAGK deficiency was identified as the cause of DMD. BMD results PFYCLSSEDLRNTHCCYTDY from different mutations in the same gene. BMD patients have some dystrophin, but its either insufficient in quantity As shown herein, Caronte, the chicken ortholog of Cer or poor in quality. Having some dystrophin protects the berus does not substantially inhibit Activin A signaling in an 10 muscles of those with BMD from degenerating as badly or as A204 Reporter Gene Assay. Similarly, we have determined quickly as those of people with DMD. that human Cerberus and Coco do not inhibit Activin A. Thus, Recent researches demonstrate that blocking or eliminat such myostatin antagonists will preferably exhibit little or no ing Myostatin function in vivo can effectively treat at least interaction with Activin A-mediated signaling. certain symptoms in DMD and BMD patients (Bogdanovich 15 et al., Supra; Wagner et al., Supra). Thus, the Subject Cerberus IV. Exemplary Therapeutic Uses derivatives, especially the N-terminally truncated versions thereof, constitute an alternative means of blocking the func The subject Coco and Cerberus polypeptides, such as the tion of Myostatin in vivo in DMD and BMD patients. full-length and the N-terminally truncated Cerberus deriva Similarly, the subject Coco or Cerberus derivatives, espe tives or Coco derivatives, can be used in a number of thera cially the N-terminally truncated versions thereof, provide an peutic settings to treat a number of diseases resulting from or effective means to increase muscle mass in other disease exacerbated by the presence of myostatin. Decreased myo conditions that are in need of muscle growth. For example, statin expression or activity has been shown to be beneficial Gonzalez-Cadavid et al. (supra) reported that that Myostatin for promoting muscle growth, inhibiting fat accumulation expression correlates inversely with fat-free mass in humans and normalizing glucose homeostasis in the context of mod 25 and that increased expression of the Myostatin gene is asso els of diabetes. ciated with weight loss in men with AIDS wasting syndrome. In certain embodiments, the Subject polypeptides and By inhibiting the function of Myostatin in AIDS patients, at derivatives thereof are used as part of a treatment for a mus least certain symptoms of AIDS may be alleviated, if not cular dystrophy. The term “muscular dystrophy' refers to a completely eliminated, thus significantly improving quality group of degenerative muscle diseases characterized by 30 of life in AIDS patients. gradual weakening and deterioration of skeletal muscles and Since loss of Myostatin function is also associated with fat sometimes the heart and respiratory muscles. Muscular dys loss without diminution of nutrient intake (Zimmers et al., trophies are genetic disorders characterized by progressive supra; McPherron and Lee, supra), the subject Coco or Cer muscle wasting and weakness that begin with microscopic berus derivatives, especially the N-terminally truncated ver changes in the muscle. As muscles degenerate over time, the 35 sions thereof, may further be used as a therapeutic agent for person’s muscle strength declines. Exemplary muscular dys slowing or preventing the development of obesity and type II trophies that can be treated with a regimen including the diabetes. subject myostatin include: Duchenne Muscular Dystrophy The cancer anorexia-cachexia syndrome is among the most (DMD), Becker Muscular Dystrophy (BMD), Emery-Dre debilitating and life-threatening aspects of cancer. Progres ifuss Muscular Dystrophy (EDMD), Limb-Girdle Muscular 40 sive weight loss in cancer anorexia-cachexia syndrome is a Dystrophy (LGMD), Facioscapulohumeral Muscular Dys common feature of many types of cancer and is responsible trophy (FSH or FSHD) (Also known as Landouzy-Dejerine), not only for a poor quality of life and poor response to che Myotonic Dystrophy (MMD) (Also known as Steinert's Dis motherapy, but also a shorter Survival time than is found in ease), Oculopharyngeal Muscular Dystrophy (OPMD), Dis patients with comparable tumors without weight loss. Asso tal Muscular Dystrophy (DD), Congenital Muscular Dystro 45 ciated with anorexia, fat and muscle tissue wasting, psycho phy (CMD). logical distress, and a lower quality of life, cachexia arises Duchenne Muscular Dystrophy (DMD) was first described from a complex interaction between the cancer and the host. by the French neurologist Guillaume Benjamin Amand It is one of the most common causes of death among cancer Duchenne in the 1860s. Becker Muscular Dystrophy (BMD) patients and is present in 80% at death. It is a complex is named after the German doctor Peter Emil Becker, who first 50 example of metabolic chaos effecting protein, carbohydrate, described this variant of DMD in the 1950s. DMD is one of and fat metabolism. Tumors produce both direct and indirect the most frequent inherited diseases in males, affecting one in abnormalities, resulting in anorexia and weight loss. Cur 3,500 boys. DMD occurs when the dystrophin gene, located rently, there is no treatment to control or reverse the process. on the short arm of the X chromosome, is broken. Since males Cancer anorexia-cachexia syndrome affects cytokine pro only carry one copy of the X chromosome, they only have one 55 duction, release of lipid-mobilizing and proteolysis-inducing copy of the dystrophin gene. Without the dystrophin protein, factors, and alterations in intermediary metabolism. Although muscle is easily damaged during cycles of contraction and anorexia is common, a decreased food intake alone is unable relaxation. While early in the disease muscle compensates by to account for the changes in body composition seen in cancer regeneration, later on muscle progenitor cells cannot keep up patients, and increasing nutrient intake is unable to reverse the with the ongoing damage and healthy muscle is replaced by 60 wasting syndrome. Cachexia should be suspected in patients non-functional fibro-fatty tissue. with cancer if an involuntary weight loss of greater than five In DMD, boys begin to show signs of muscle weakness as percent of premorbid weight occurs within a six-month early as age 3. The disease gradually weakens the skeletal or period. Voluntary muscles, those in the arms, legs and trunk. By the Since systemic overexpression of Myostatin in adult mice early teens or even earlier, the boy's heart and respiratory 65 was found to induce profound muscle and fat loss analogous muscles may also be affected. BMD is a much milder version to that seen in human cachexia syndromes (Zimmers et al., of DMD. Its onset is usually in the teens or early adulthood, supra), the subject Coco or Cerberus derivatives, especially US 7,833,971 B2 27 28 the N-terminally truncated versions thereofas a pharmaceu limbs begin to atrophy from disuse. This muscle weakness tical composition can be beneficially used as a Myostatin will become debilitating and a person will need a wheelchair antagonist/blockerto prevent, treat, oralleviate the symptoms or become unable to function out of bed. Most ALS patients of the cachexia syndrome, where muscle growth is desired. die from respiratory failure or from complications of ventila In certain embodiments, the subject variant Coco or Cer tor assistance like pneumonia, 3-5 years from disease onset. berus polypeptides, particularly the N-terminally truncated The causes of these neurological diseases has remained Cerberus derivatives, can be used to form pharmaceutical largely unknown. They are conventionally defined as distinct compositions that can be beneficially used to prevent, treat, or diseases, yet clearly show extraordinary similarities in basic alleviate symptoms of a host of diseases involving neurode processes and commonly demonstrate overlapping Symp generation. While not wishing to be bound by any particular 10 toms far greater than would be expected by chance alone. theory, the Subject Cerberus derivatives may antagonize the Current disease definitions fail to properly deal with the issue inhibitory feedback mechanism mediated through the wild of overlap and a new classification of the neurodegenerative type ALK7 receptor, thus allowing new neuronal growth and disorders has been called for. differentiation. The subject Cerberus derivative as a pharma Huntington's disease (HD) is another neurodegenerative ceutical composition can be beneficially used to prevent, 15 disease resulting from genetically programmed degeneration treat, or alleviate symptoms of diseases with neurodegenera of neurons in certain areas of the brain. This degeneration tion, including Alzheimer's Disease (AD), Parkinson's Dis causes uncontrolled movements, loss of intellectual faculties, ease (PD), Amyotrophic Lateral Sclerosis (ALS), Hunting and emotional disturbance. HD is a familial disease, passed ton's disease, etc. from parent to child through a dominant mutation in the Alzheimer's disease (AD) is a chronic, incurable, and wild-type gene. Some early symptoms of HD are mood unstoppable central nervous system (CNS) disorder that Swings, depression, irritability or trouble driving, learning occurs gradually, resulting in memory loss, unusual behavior, new things, remembering a fact, or making a decision. As the personality changes, and a decline in thinking abilities. These disease progresses, concentration on intellectual tasks losses are related to the death of specific types of brain cells becomes increasingly difficult and the patient may have dif and the breakdown of connections between them. 25 ficulty feeding himself or herself and Swallowing. The rate of AD has been described as childhood development in disease progression and the age of onset vary from person to reverse. In most people with AD, symptoms appear after the person. age 60. The earliest symptoms include loss of recent memory, Tay-Sachs disease and Sandhoff disease are glycolipid faulty judgment, and changes in personality. Later in the storage diseases caused by the lack of lysosomal B-hex disease, those with AD may forget how to do simple tasks like 30 osaminidase (Gravel et al., in The Metabolic Basis of Inher washing their hands. Eventually people with AD lose all ited Disease, eds. Scriver et al., McGraw-Hill, New York, pp. reasoning abilities and become dependent on other people for 2839-2879, 1995). In both disorders, G. ganglioside and their everyday care. Finally, the disease becomes so debili related glycolipid Substrates for B-hexosaminidase accumu tating that patients are bedridden and typically develop coex late in the nervous system and trigger acute neurodegenera isting illnesses. AD patients most commonly die from pneu 35 tion. In the most severe forms, the onset of symptoms begins monia, 8 to 20 years from disease onset. in early infancy. A precipitous neurodegenerative course then Parkinson's disease (PD) is a chronic, incurable, and ensues, with affected infants exhibiting motor dysfunction, unstoppable CNS disorder that occurs gradually and results in seizure, visual loss, and deafness. Death usually occurs by 2-5 uncontrolled body movements, rigidity, tremor, and gait dif years of age. Neuronal loss through an apoptotic mechanism ficulties. These motor system problems are related to the 40 has been demonstrated (Huang et al., Hum. Mol. Genet. 6: death of brain cells in an area of the brain that produces 1879-1885, 1997). dopamine—a chemical that helps control muscle activity. It is well-known that apoptosis plays a role in AIDS patho In most people with PD, symptoms appear after age 50. genesis in the immune system. However, HIV-1 also induces The initial symptoms of PD are a pronounced tremor affect neurological disease. Shietal. (J. Clin. Invest. 98: 1979-1990, ing the extremities, notably in the hands or lips. Subsequent 45 1996) examined apoptosis induced by HIV-1 infection of the characteristic symptoms of PD are stiffness or slowness of central nervous system (CNS) in an in vitro model and in movement, a shuffling walk, Stooped posture, and impaired brain tissue from AIDS patients, and found that HIV-1 infec balance. There are wide ranging secondary symptoms such as tion of primary brain cultures induced apoptosis in neurons memory loss, dementia, depression, emotional changes, and astrocytes in vitro. Apoptosis of neurons and astrocytes Swallowing difficulties, abnormal speech, sexual dysfunc 50 was also detected in brain tissue from 10/11 AIDS patients, tion, and bladder and bowel problems. These symptoms will including 5/5 patients with HIV-1 dementia and 4/5 nonde begin to interfere with routine activities, such as holding a mented patients. fork or reading a newspaper. Finally, people with PD become Neuronal loss is a also a salient feature of prion diseases, so profoundly disabled that they are bedridden. People with such as Creutzfeldt-Jakob disease in human, BSE in cattle PD usually die from pneumonia. 55 (mad cow disease), Scrapie Disease in sheep and goats, and Amyotrophic lateral Sclerosis (ALS; Lou Gehrig's disease; feline spongiform encephalopathy (FSE) in cats. motor neuron disease) is a chronic, incurable CNS disorder The subject Cerberus and Coco polypeptides, including the that attacks the motor neurons, components of the CNS that N-terminally truncated Cerberus derivatives are also useful to connect the brain to the skeletal muscles. In ALS, the motor prevent, treat, and alleviate symptoms of various PNS disor neurons deteriorate and eventually die, and though a person’s 60 ders, such as the ones described below. The PNS is composed brain normally remains fully functioning and alert, the com of the nerves that lead to or branch off from the CNS. The mand to move never reaches the muscles. peripheral nerves handle a diverse array of functions in the Most people are diagnosed with ALS between 40 and 70 body, including sensory, motor, and autonomic functions. years of age. The first motor neurons that weaken are those When an individual has a peripheral neuropathy, nerves of the leading to the arms or legs. Those with ALS may have trouble 65 PNS have been damaged. Nerve damage can arise from a walking, they may drop things, fall, slur their speech, and number of causes, such as disease, physical injury, poisoning, laugh or cry uncontrollably. Eventually the muscles in the or malnutrition. These agents may affect either afferent or US 7,833,971 B2 29 30 efferent nerves. Depending on the cause of damage, the nerve condition characterized by multiple isolated nerve injuries. cell axon, its protective myelin sheath, or both may be injured Mononeuropathies may result from a wide variety of causes, or destroyed. including ischemia; traumatic injury; compression; connec The term peripheral neuropathy encompasses a wide range tive tissue diseases; cumulative trauma disorders; and other of disorders in which the nerves outside of the brain and 5 conditions); Neuralgia (Intense or aching pain that occurs spinal cord peripheral nerves—have been damaged. along the course or distribution of a peripheral or cranial Peripheral neuropathy may also be referred to as peripheral nerve); Peripheral Nervous System Neoplasms (Neoplasms neuritis, or if many nerves are involved, the terms polyneur which arise from peripheral nerve tissue. This includes neu opathy or polyneuritis may be used. rofibromas; Schwannomas; granular cell tumors; and malig Peripheral neuropathy is a widespread disorder, and there 10 nant peripheral nerve sheath tumors. See DeVita Jr et al., are many underlying causes. Some of these causes are com Cancer: Principles and Practice of Oncology, 5" ed., pp 1750 mon, such as diabetes, and others are extremely rare, such as 1); Nerve Compression Syndromes (Mechanical compres acrylamide poisoning and certain inherited disorders. The sion of nerves or nerve roots from internal or external causes. most common worldwide cause of peripheral neuropathy is These may resultina conduction block to nerve impulses, due leprosy. Leprosy is caused by the bacterium Mycobacterium 15 to, for example, myelin sheath dysfunction, or axonal loss. leprae, which attacks the peripheral nerves of affected The nerve and nerve sheath injuries may be caused by people. According to statistics gathered by the World Health ischemia; inflammation; or a direct mechanical effect); Neu Organization, an estimated 1.15 million people have leprosy ritis (A general term indicating inflammation of a peripheral worldwide. or cranial nerve. Clinical manifestation may include pain; Leprosy is extremely rare in the United States, where dia paresthesias; paresis; or hyperthesia); Polyneuropathies (Dis betes is the most commonly known cause of peripheral neu eases of multiple peripheral nerves. The various forms are ropathy. It has been estimated that more than 17 million categorized by the type of nerve affected (e.g., sensory, motor, people in the United States and Europe have diabetes-related or autonomic), by the distribution of nerve injury (e.g., distal polyneuropathy. Many neuropathies are idiopathic—no VS. proximal), by nerve component primarily affected (e.g., known cause can be found. The most common of the inherited 25 demyelinating VS. axonal), by etiology, or by pattern of inher peripheral neuropathies in the United States is Charcot itance). Marie-Tooth disease, which affects approximately 125,000 In certain embodiments, the subject full-length Coco or persons. Cerberus polypeptides or variants thereof are used as part of Another of the better known peripheral neuropathies is a treatment for diseases or conditions characterized by exces Guillain-Barré syndrome, which arises from complications 30 sive or undesirable levels of BMP, such as the ones described associated with viral illnesses. Such as cytomegalovirus, below. Epstein-Barr virus, and human immunodeficiency virus The heterotopic ossification of muscles, tendons, and liga (HIV), or bacterial infection, including Campylobacter jejuni ments is a common problem faced by orthopaedic Surgeons. and Lyme disease. The worldwide incidence rate is approxi Hannallah et al. (J Bone Joint Surg. Am. 2004 January: 86-A mately 1.7 cases per 100,000 people annually. Other well 35 (1):80-91) investigated the ability of Noggin (a BMP bone known causes of peripheral neuropathies include chronic morphogenetic protein antagonist) to inhibit heterotopic alcoholism, infection of the varicella-zoster virus, botulism, ossification. Three varying doses of Noggin-expressing and poliomyelitis. Peripheral neuropathy may develop as a muscle-derived stem cells inhibited the heterotopic ossifica primary symptom, or it may be due to another disease. For tion elicited by BMP-4-expressing muscle-derived stem example, peripheral neuropathy is only one symptom of dis 40 cells. Each of three varying doses of Noggin-expressing eases such as amyloid neuropathy, certain cancers, or inher muscle-derived stem cells also significantly inhibited the het ited neurologic disorders. Such diseases may affect the erotopic ossification elicited by demineralized bone matrix. peripheral nervous system (PNS) and the central nervous All eleven animals that underwent Achilles tenotomy devel system (CNS), as well as other body tissues. oped heterotopic ossification at the site of the injury in the Other PNS diseases treatable with the subject Cerberus and 45 control limbs. In contrast, the limbs treated with the Noggin Coco polypeptides include: Brachial Plexus Neuropathies expressing muscle-derived stem cells had a reduction in the (Diseases of the cervical and first thoracic roots, nerve trunks, formation of heterotopic ossification of 83% and eight of the cords, and peripheral nerve components of the brachial eleven animals had no radiographic evidence of heterotopic plexus. Clinical manifestations include regional pain, pares ossification (p<0.05). Thus, delivery of Noggin mediated by thesia; muscle weakness, and decreased sensation in the 50 muscle-derived stem cells can inhibit heterotopic ossification upper extremity. These disorders may be associated with caused by BMP4, demineralized bone matrix, and trauma in trauma, including birth injuries; thoracic outlet syndrome; an animal model, indicating that genetherapy to deliver BMP neoplasms, neuritis, radiotherapy; and other conditions. See inhibitors (Noggin or Cerberus) may become a powerful Adams et al. Principles of Neurology, 6" ed., pp 1351-2); method to inhibit heterotopic ossification in targeted areas of Diabetic Neuropathies (Peripheral, autonomic, and cranial 55 the body. See also Glaser et al. (J Bone Joint Surg. Am. 2003 nerve disorders that are associated with diabetes mellitus. December: 85-A(12):2332–42). These conditions usually result from diabetic microvascular Osteoarthritis (OA) is a joint disease characterized by injury involving Small blood vessels that Supply nerves (vasa osteophyte development, fibrosis, and articular cartilage nervorum). Relatively common conditions which may be damage. Effects of exogenous transforming growth factor associated with diabetic neuropathy include third nerve 60 beta (TGFbeta) isoforms and bone morphogenetic proteins palsy, mononeuropathy; mononeuropathy multiplex; dia (BMPs) suggest a role for these growth factors in the patho betic amyotrophy; a painful polyneuropathy; autonomic neu genesis of OA. Scharstuhl et al. (Arthritis Rheum. 2003 ropathy; and thoracoabdominal neuropathy. See Adams et al., December; 48(12):3442-51) used adenoviral overexpression Principles of Neurology, 6'ed, p 1325); Mononeuropathies of TGF-beta and BMP antagonists to block the signaling of (Disease or trauma involving a single peripheral nerve in 65 TGF-beta and BMP. The inhibitors studied include a secreted, isolation, or out of proportion to evidence of diffuse periph pan-specific TGF-beta antagonist called murine latency-as eral nerve dysfunction. Mononeuropathy multiplex refers to a sociated peptide 1 (mLAP-1), intracellular inhibitory Smado US 7,833,971 B2 31 32 (a BMP antagonist), and Smad7 (a TGF-beta/BMP inhibitor). and mental retardation. Recent studies have demonstrated Intraarticular injection of papain caused increased protein that gain-of-function mutations in fibroblast growth factor expression of several TGF-beta and BMP isoforms in syn receptors (fgfr) are associated with syndromic forms of cran ovium and cartilage. Adenovirus transfection into the joint iosynostosis. Noggin, an antagonist of bone morphogenetic resulted in a strong expression of the transgenes in the Syn proteins (BMPs), is required for embryonic neural tube, ovial lining. Overexpression of mLAP-1, Smado, and Smad? somites and skeleton patterning. Warren et al. (Nature. 2003 led to a significant reduction in osteophyte formation com Apr. 10; 422(6932): 625-9) show that noggin is expressed pared with that in controls. Smado and SmadT overexpression postnatally in the Suture mesenchyme of patent, but not fus also significantly decreased synovial thickening. Further ing, cranial Sutures, and that noggin expression is Suppressed more, the secreted TGF-beta inhibitor mLAP-1 increased 10 by FGF2 and syndromic fgfr signalling. Since noggin misex articular cartilage PG loss. These results indicate a pivotal pression prevents cranial Suture fusion in vitro and in vivo, it role of excessive endogenous TGF-beta and BMP in the is suggested that syndromic fgfr-mediated craniosynostoses development of osteophytes and synovial thickening, impli may be the result of inappropriate downregulation of noggin cating excessive endogenous TGFbeta and BMP in the patho expression, leading to abnormally high BMP activity. Thus genesis of OA. In contrast, the prevention of cartilage damage 15 the subject Cerberus and Coco therapeutics may be used to by endogenous TGF-beta signifies the protective role of TGF down-regulate BMP activity to prevent or treat such condi beta in articular cartilage. Thus the subject Coco or Cerberus tions. pharmaceutical compositions can be used as BMP antago nists to treat OA, including the development of osteophytes V. Exemplary Formulations and synovial thickening. In an analysis of normal ovarian surface epithelium (OSE) The Subject compositions may be used alone, or as part of and ovarian cancer (OC) cells. Shepherd and Nachtigal (En a conjoint therapy with other compounds/pharmaceutical docrinology. 2003 August: 144(8):33.06-14) observed BMP4 compositions. mRNA expression and found that primary OC cells produce The soluble Coco or Cerberus polypeptides, including the mature BMP4. In addition, each member of the downstream 25 N-terminally truncated Cerberus derivative therapeutics for signaling pathway was expressed in primary OSE and OC use in the subject methods may be conveniently formulated cells. Smad1 was phosphorylated and underwent nuclear for administration with a biologically acceptable medium, translocation in normal OSE and OC cells upon treatment Such as water, buffered saline, polyol (for example, glycerol, with BMP4. Interestingly, the BMP target genes ID1 and ID3 propylene glycol, liquid polyethylene glycol and the like) or were up-regulated 10- to 15-fold in primary OC cells, com 30 suitable mixtures thereof. The optimum concentration of the pared with a 2- to 3-fold increase in normal OSE. The growth active ingredient(s) in the chosen medium can be determined of several primary OC cells was relatively unaltered by BMP4 empirically, according to procedures well known to medici treatment; however, long-term BMP4 treatment of primary nal chemists. As used herein, “biologically acceptable OC cells resulted in decreased cell density as well as medium' includes any and all solvents, dispersion media, and increased cell spreading and adherence. These data demon 35 the like which may be appropriate for the desired route of strate the existence and putative function of BMP signaling in administration of the pharmaceutical preparation. The use of normal OSE and OC cells, and thus the subject Cerberus Such media for pharmaceutically active substances is known pharmaceutical preparations can be used to regulate BMP4 in the art. Except insofar as any conventional media or agent signaling in OC pathogenesis. is incompatible with the activity of the therapeutics, its use in Fibrodysplasia ossificans progressiva (FOP), a rare genetic 40 the pharmaceutical preparation of the disclosure is contem disabling disease characterized by heterotopic bone forma plated. Suitable vehicles and their formulation inclusive of tion, is of special interest for general medicine since the bone other proteins are described, for example, in the book Rem morphogenetic proteins (especially BMP4) involved in its ington's Pharmaceutical Sciences (Remington's Pharma pathogenesis are known to play a role in skeletal morphogen ceutical Sciences. Mack Publishing Co., Easton, Pa., USA esis, and the gene antagonist to BMP-4 (Such as noggin) 45 1985). These vehicles include injectable “deposit formula might be useful in preventing lamellar bone formation. See tions.” Blaszczyketal. (EurJ. Dermatol. 2003 May-June; 13(3):234 Pharmaceutical formulations of the present disclosure can 7). Thus the subject Cerberus therapeutics may also be used to also include Veterinary compositions, e.g., pharmaceutical treat FOP. preparations of the Coco or Cerberus derivative therapeutics Atherosclerosis is now viewed as an inflammatory disease 50 suitable for veterinary uses, e.g., for the treatment of live occurring preferentially in arterial regions exposed to dis stock (cow, sheep, goat, pig, and horse, etc.) or domestic turbed flow conditions, including oscillatory shear stress animals, e.g., cats and dogs. (OS), in branched arteries. Sorescu et al. (J Biol Chem. 278 Methods of disclosure may also be provided by recharge (33):31128-35, 2003) suggest that BMP4 is a mechanosensi able or biodegradable devices. Various slow release poly tive, inflammatory factor playing a critical role in early steps 55 meric devices have been developed and tested in vivo in of atherogenesis in the lesion-prone areas. Thus the Subject recent years for the controlled delivery of drugs, including Cerberus therapeutics may be used to control BMP-4 induced proteinacious biopharmaceuticals. A variety of biocompat inflammatory response in early steps of atherogenesis in ible polymers (including hydrogels), including both biode those areas. gradable and non-degradable polymers, can be used to form During skull development, the cranial connective tissue 60 an implant for the Sustained release of a therapeutic at a framework undergoes intramembranous ossification to form particular target site. skull bones (calvaria). As the calvarial bones advance to The pharmaceutical compositions according to the present envelop the brain, fibrous sutures form between the calvarial disclosure may be administered as either a single dose or in plates. Expansion of the brain is coupled with calvarial multiple doses. The pharmaceutical compositions of the growth through a series of tissue interactions within the cra 65 present disclosure may be administered either as individual nial Suture complex. Craniosynostosis, or premature cranial therapeutic agents or in combination with other therapeutic Suture fusion, results in an abnormal skull shape, blindness agents. The treatments of the present disclosure may be com US 7,833,971 B2 33 34 bined with conventional therapies, which may be adminis lowest dose effective to produce atherapeutic effect. Such an tered sequentially or simultaneously. The pharmaceutical effective dose will generally depend upon the factors compositions of the present disclosure may be administered described above. Generally, intravenous, intracerebroven by any means that enables the Coco or Cerberus derivatives to tricular and Subcutaneous doses of the compounds of this reach the targeted cells/tissues/organs. In some embodi disclosure for a patient will range from about 0.0001 to about ments, routes of administration include those selected from 100 mg per kilogram of body weight per day. the group consisting of oral, intravesically, intravenous, If desired, the effective daily dose of the active compound intraarterial, intraperitoneal, local administration into the may be administered as two, three, four, five, six or more blood supply of the organ in which the targeted cells reside or Sub-doses administered separately at appropriate intervals directly into the cells. Intravenous administration is the pre 10 throughout the day, optionally, in unit dosage forms. ferred mode of administration. It may be accomplished with The term “treatment” is intended to encompass also pro the aid of an infusion pump. phylaxis, therapy and cure. The phrases “parenteral administration' and “adminis The patient receiving this treatment is any animal in need, tered parenterally as used herein means modes of adminis including primates, in particular humans, and other non-hu tration other than enteral and topical administration, usually 15 man mammals such as equines, cattle, Swine and sheep; and by injection, and includes, without limitation, intravenous, poultry and pets in general. intramuscular, intraarterial, intrathecal, intracapsular, The compound of the disclosure can be administered as intraorbital, intracardiac, intradermal, intraperitoneal, tran Such or in admixtures with pharmaceutically acceptable car stracheal, Subcutaneous, Subcuticular, intraarticulare, Sub riers and can also be administered in conjunction with other capsular, Subarachnoid, intraspinal and intrastermal injection antimicrobial agents such as penicillins, cephalosporins, ami and infusion. noglycosides and glycopeptides. Conjunctive therapy, thus The phrases “systemic administration.” “administered sys includes sequential, simultaneous and separate administra temically.” “peripheral administration' and “administered tion of the active compound in a way that the therapeutical peripherally as used herein mean the administration of a effects of the first administered one is not entirely disappeared compound, drug or other material other than directly into the 25 when the Subsequent is administered. central nervous system, such that it enters the patient’s system Combined with certain formulations, the subject Coco or and, thus, is Subject to metabolism and other like processes, Cerberus derivatives can be effective soluble agents. The for example, Subcutaneous administration. therapeutic polypeptide can be provided a fusion peptide These compounds may be administered to humans and along with a second peptide which promotes solubility. To other animals for therapy by any Suitable route of adminis 30 illustrate, the Cerberus derivatives of the present disclosure tration, including orally, intravesically, nasally, as by, for can be provided as part of a fusion polypeptide with all or a example, a spray, rectally, intravaginally, parenterally, intra fragment of the hinge or Fc portion of the immunoglobulin, cistemally and topically, as by powders, ointments or drops, which can promote solubility and/or serum stability. including buccally and Sublingually. The present disclosure also contemplates a peptidomi Regardless of the route of administration selected, the 35 metic sequence of the Subject polypeptide derivatives as compounds of the present disclosure, which may be used in a described herein. Suitable hydrated form, and/or the pharmaceutical composi Generally, the nomenclature used herein and the laboratory tions of the present disclosure, are formulated into pharma procedures utilized in the present disclosure include molecu ceutically acceptable dosage forms such as described below lar, biochemical, microbiological and recombinant DNA or by other conventional methods known to those of skill in 40 techniques. Such techniques are thoroughly explained in the the art. literature. See, for example, “Molecular Cloning: A labora Actual dosage levels of the active ingredients in the phar tory Manual” Sambrook et al., (1989): “Current Protocols in maceutical compositions of this disclosure may be varied so Molecular Biology Volumes I-III Ausubel, R. M., ed. as to obtain an amount of the active ingredient which is (1994); Ausubel et al., “Current Protocols in Molecular Biol effective to achieve the desired therapeutic response for a 45 ogy”, John Wiley and Sons, Baltimore, Md. (1989); Perbal, particular patient, composition, and mode of administration, “A Practical Guide to Molecular Cloning”, John Wiley & without being toxic to the patient. Sons, New York (1988); Watson et al., “Recombinant DNA, The selected dosage level will depend upon a variety of Scientific American Books, New York; Birren et al. (eds) factors including the activity of the particular compound of "Genome Analysis: A Laboratory Manual Series”, Vols. 1-4, the present disclosure employed, or the ester, salt or amide 50 Cold Spring Harbor Laboratory Press, New York (1998); thereof, the route of administration, the time of administra methodologies as set forth in U.S. Pat. Nos. 4,666,828; 4,683, tion, the rate of excretion of the particular compound being 202; 4,801,531; 5,192,659 and 5.272,057; “Cell Biology: A employed, the duration of the treatment, other drugs, com Laboratory Handbook'', Volumes I-III Cellis, J. E., ed. pounds and/or materials used in combination with the par (1994): “Current Protocols in Immunology” Volumes I-III ticular therapeutic employed, the age, sex, weight, condition, 55 Coligan J. E., ed. (1994); Stites et al. (eds), “Basic and Clini general health and prior medical history of the patient being cal Immunology” (8' Edition), Appleton & Lange, Norwalk, treated, and like factors well known in the medical arts. Conn. (1994); Mishell and Shiigi (eds), “Selected Methods in A physician or veterinarian having ordinary skill in the art Cellular Immunology”. W.H. Freeman and Co., New York can readily determine and prescribe the effective amount of (1980); available immunoassays are extensively described in the pharmaceutical composition required. For example, the 60 the patent and scientific literature, see, for example, U.S. Pat. physician or veterinarian could start doses of the compounds Nos. 3,791,932; 3,839,1533,850,752; 3,850,578; 3,853.987; of the disclosure employed in the pharmaceutical composi 3,867,517; 3,879,262; 3,901,654; 3,935,074; 3,984,533; tion at levels lower than that required in order to achieve the 3.996,345; 4,034,074; 4,098,876; 4,879,219; 5,011,771 and desired therapeutic effect and gradually increase the dosage 5,281.521; “Oligonucleotide Synthesis' Gait, M. J., ed. until the desired effect is achieved. 65 (1984): “Nucleic Acid Hybridization’ Hames, B. D., and In general, a suitable daily dose of a compound of the Higgins S. J., eds. (1985): “Transcription and Translation' disclosure will be that amount of the compound which is the Hames, B. D., and Higgins S. J., eds. (1984); 'Animal Cell US 7,833,971 B2 35 36 Culture' Freshney, R.I., ed. (1986): “Immobilized Cells and The following constructs were tested: Enzymes' IRL Press, (1986): “A Practical Guide to Molecu lar Cloning Perbal, B., (1984) and “Methods in Enzymol ogy” Vol. 1-317, Academic Press: “PCR Protocols: A Guide Human Cerberus, full length, no Fo (SEQ ID NO : 11) To Methods And Applications”. Academic Press, San Diego, MHLLLFOLLW LLPLGKTTRH ODGRONOSSL SPWLLPRNOR Calif. (1990); Marshak et al., “Strategies for Protein Purifi ELPTGNHEEA EEKPDLFVAV PHLVATSPAG EGOROREKML cation and Characterization—A Laboratory Course Manual CSHL Press (1996); all of which are incorporated by refer SRFGRFWKKP EREMHPSRDS DSEPFPPGTO SLIOPIDGMK ence as if fully set forth herein. Other general references are MEKSPLREEA KKFWHHFMFR KTPASOGVIL PIKSHEVHWE provided throughout this document. The procedures therein 10 are believed to be well known in the art and are provided for TCRTVPFSOT ITHEGCEKVV VONNLCFGKC GSVHFPGAAO the convenience of the reader. All the information contained therein is incorporated herein by reference. HSHTSCSHCL PAKFTTMHLP LNCTELSSVI KVVMLVEECO CKVKTEHEDG HILHAGSODS FIPGVSA EXEMPLIFICATION 15 Human Cerberus, full length, Fc (TGGG linker (SEQ ID The disclosure now being generally described, it will be NO: 27) and Fc, underlined; native signal sequence under more readily understood by reference to the following lined with dotted line) (SEQ ID NO:12)

MHLIFQIVIIPLSKTTRH QDGRONOSSL SPWLLPRNOR ELPTGNHEEA EEKPDLFVAW PHLVATSPAG EGORQREKML SRFGRFWKKP EREMHPSRDS DSEPFPPGTO SLIOPIDGMK MEKSPLREEA. KKFWHHFMFR KTPASOGVIL PIKSHEVHWE TCRTVPFSOT ITHEGCEKVV VONNLCFGKC GSWHFPGAAQ HSHTSCSHCL PAKFTTMHLP LNCTELSSVI KVVMLVEECO CKVKTEHEDG HILHAGSODS FIPGWSA TGGGTHTCPP CPAPELLGGP SWFLFPPKPK. DTLMISRTPE VTCVVVDVSH EDPEVKENWY VDGVEVHNAKTKPREEOYNS TYRVVSVLTV. LHODWLNGKE YKCKVSNKAL PVPIEKTISK AKGOPREPOV YTLPPSREEM TKNOWSLTCL WKGFYPSDIA WEWESNGOPE NNYKTTPPVL DSDGSFFLY'S KLTVDKSRWO OGNWFSCSWM HEALHNHYTO KSLSLSPGK

examples, which are included merely for purposes of illus Human Cerberus, short form, Fc (TGGG linker (SEQ ID tration of certain embodiments and embodiments of the NO: 27) and Fc, underlined) (SEQID NO:13)

EVHWETCRTV PFSOTITHEG CEKVVVONNL CFGKCGSVHF PGAAOHSHTS CSHCLPAKFT

TMHLPLNCTE LSSVIKVVML WEECOCKWKT EHEDGHILHA GSODSFIPGV SATGGGTHTCPP

CPAPELLGGP SWFLFPPKPK. DTLMISRTPE WTCVWWDWSH EDPEWKFNWY WDGWEWHNAK

TKPREEQYNS TYRVWSWLTV LHQDWLNGKE YKCKVSNKAL PWPIEKTISK AKGOPREPOW YTLPPSREEM TKNOWSLTCL WKGFYPSDIA WEWESNGQPE NNYKTTPPWL DSDGSFFLYS KLTVDKSRWO OGNWFSCSVM HEALHNHYTO KSLSLSPGK

present disclosure, and are not intended to limit the disclo 50 Three different leader sequences were considered: SUC. Example 1 (i) Honey bee mellitin (HBML) : MKFLWNWALWFMWWYISYIYA (SEQ ID NO: 14)

Sources of Caronte and Human Cerberus Protein 55 (ii) Tissue Plasminogen Activator (TPA) : MDAMKRGLCCWLLLCGAWFWSP (SEQ ID NO: 15) Caronte-Fc (Cerberus homolog from Gallus gallus) was ordered from R&D Systems (Minneapolis, Minn.). (iii) Native: Full-length and N-terminally truncated forms of human 60 MHLLLFOLLW LLPLGKT. (SEQ ID NO: 16) Cerberus sequence were cloned into a human CMV derived expression vector, either with or without a C-terminal fusion A heterologous or native leader sequence may be fused to to an Fc portion of IgG1 (both human and murine IgG1 Fc the protein sequence at any position within the first 30 amino fusions were produced). These constructs were transiently acids. Modeling Suggests that the native leader would yield a transfected in HEK293 cells using polyethylenimine (PEI). 65 product beginning with “TRH... 'at position 18. Analysis of After culturing, cells were harvested and conditioned media products expressed herein indicates that the native leader was collected for purification. more typically yields a product beginning “KTT . . . . at US 7,833,971 B2 37 38 position 16. Therefore, heterologous leader sequences may This is followed by a Luciferase assay. In the absence of be fused N-terminal to position 16 or position 18. any inhibitors, Activin A showed 10 fold stimulation of reporter gene expression and an ED50-2 ng/ml. GDF-8: Example 2 ED50: ~5 ng/ml, 15 fold stimulation. 11: 16 fold stimulation, ED50: ~1.5 ng/ml. Caronte Binds 11 As shown in FIG. 4, Caronte inhibits 11 signaling in the A-204 Reporter Gene Assay. An ActRIIA-Fc (IIA muC2a) 11 is a close homolog of myostatin that regulates neuro fusion also inhibits 11 signaling. As shown in FIG. 5, Caronte logical processes. 11 was immobilized on a BiaCore CM5 does not inhibit Activin A in the A-204 Reporter Gene Assay. chip using standard amine coupling procedure. Trace: Caro 10 An ActRIIA-Fc fusion (IIA muC2a), as expected, does inte (200 ug/ml, R&D Systems) was injected on the 11 inhibit Activin A signaling. Thus, Caronte is a selective coupled chip. The tracing in FIG. 2 shows binding of Caronte inhibitor of 11/myostatin while not affecting Activin A sig to 11. naling. This type of selectivity Suggests that Caronte, Cer Example 3 berus and Coco will have relatively few side effects when 15 used as a therapeutic. As expected, Cerberus behaved much Caronte and Human Cerberus Inhibit 11 and like Caronte, and inhibited myostatin signaling. See FIG. 6. Myostatin-Mediated Signaling Similar experiments were conducted to test the binding of human Cerberus and Coco to Activin A, and these experi An A-204 Reporter Gene Assay was used to evaluate the ments confirm that these molecules do not bind to Activin A. effects of Caronte and Cerberus on signaling by 11, myostatin and Activin A. Cell line: Human Rhabdomyosarcoma (de Example 4 rived from muscle). Reporter vector: pGL3(CAGA)12 (De scribed in Denniler et al., 1998, EMBO 17:3091-3100.) See Sources of Human Coco Protein FIG. 3. The CAGA12 motif is present in TGF-Beta respon sive genes (PAI-1 gene), so this vector is of general use for 25 Full-length human Coco was cloned into a human CMV factors signaling through Smad3 and 3. derived expression vector, either with or without a C-terminal Day 1: Split A-204 cells into 48-well plate. fusion to an Fc portion of IgG1 (both human and murine IgG1 Day 2: A-204 cells transfected with 10 ugpCL3(CAGA)12 Fc fusions were produced). These constructs were transiently or pGL3(CAGA)12(10 ug)+pRLCMV (1 lug) and Fugene. transfected in HEK293 cells using polyethylenimine (PEI). Day 3: Add factors (diluted into medium+0.1% BSA). 30 After culturing, cells were harvested and conditioned media Inhibitors need to be preincubated with Factors for 1 hr before was collected for purification. adding to cells. 6 hrs later, cells rinsed with PBS, and lyse The following construct was made, and a murine Fc fusion cells. was made also, and used in the assays presented herein:

Human Coco, full length, Fc (TGGG linker and mFo) (SEO ID NO : 17) MLLGOLSTLL CLLSGALPTG SGRPEPOSPR POSWAAANOT WALGPGALPP LVPASALGSW

KAFLGLOKAR OLGMGRLORG ODEWAAVTLP LNPOEVIOGM. CKAVPFWOWF SRPGCSAIRL

RNHLCFGNCS SLYIPGSDPT PLWLCNSCMP ARKRWAPWWL WCLTGSSASR RRWKISTMLI

EGCHCSPKA, TGGGTHTCPP CPAPELLGGP SWFLFPPKPK. DTLMISRTPE WTCWWWDWSH

EDPEWKFNWY WDGWEWHNAK TKPREEQYNS TYRWWSWLTV LHQDWLNGKE YKCKWSNKAL PWPIEKTISK AKGOPREPOV YTLPPSREEM TKWOWSLTCL VKGFYPSDIA WEWESNGOPE NNYKTTPPWL DSDGSFFLYS KLTVDKSRWQ QGNWFSCSWM HEALHNHYTQ KSLSLSPGK

50 The following construct is analogous to the short form Cerberus-Fc fusion protein:

Human Coco, short form, Foc (TGGG linker and mEc) (SEQ ID NO: 18) LNPOEVIOGM CKAVPFWOVF SRPGCSAIRL

RNHLCFGHCS SLYIPGSDPT PLWLCNSCMP ARKRWAPWWL WCLTGSSASR RRWKISTMLI

EGCNCSPKA, TGGGTHTCPP CPAPELLGGP SWFLFPPKPK. DTLMISRTPE WTCWWWDWSH

EDPEWKFNWY WDGWEWHNAK TKPREEQYNS TYRWWSWLTV LHQDWLNGKE YKCKWSNKAL

PWPIEKTISK AKGOPREPOV YTLPPSREEM TKNOWSLTCL VKGFYPSDIA WEWESNGOPE

NNYKTTPPVL DSDGSFFLYS KLTVDKSRWO OGNWFSCSVM HEALHNHYTO KSLSLSPGK US 7,833,971 B2 39 40 Three different leader sequences were considered: tion that is expected to be proximal to the 212 position in the three-dimensional structure of the protein. Similar mutations may be made at each of the other sites 38 NQRELP43 (SEQ (i) Honey bee mellitin (HBML) : ID NO: 24) and 138 MFR KTP 143 (SEQ ID NO. 23), MKFLWNWALWFMWWYISYIYA (SEQ ID NO: 14) depending on the length of the Cerberus molecule to be (ii) Tissue Plasminogen Activator (TPA) : employed. A particularly desirable mutation with respect to MDAMKRGLCCWLLLCGAWFWSP (SEQ ID NO: 15) the 38 NORELP43 (SEQID NO:24) cleavage site is an R to (iii) Native: S/T mutation to make the sequence 38 NQ(S/T)ELP43 (SEQ MLLGOLSTLL CLLSGALPTG S. (SEQ ID NO: 19) ID NO: 29), simultaneously eliminating the cleavage site and 10 introducing an N-linked glycosylation site. Additionally, Heterologous or native leader sequences may be fused experiments have shown that products cleaved at E41 and anywhere in the first 30 amino acids, and particularly N-ter K141 retain myostatin binding activity. Accordingly, N-ter minal to any of amino acids 16-23. minally truncated forms of Cerberus, beginning at E41 or Example 5 15 K141 will be resistant to cleavage at these sites and retain activity. The activity of the shortform Suggests that a minimal Human Coco Inhibits 11 Signaling myostatin binding domain is the cysteine knot, located at amino acids 162-241 of SEQID NO:2. Conditioned medium from cells expressing human Coco Cerberus constructs with one or more of the alterations mFc was tested for effects on A-204 reporter gene expression (shown in brackets below: e.g., “R(T) means that an argi in the presence of 11. As shown in FIG. 8, conditioned nine normally at the position may be replaced with a threo medium containing Coco-mFc inhibits 11 signaling in the nine) will have N-linked glycosylation sites that will block A-204 Reporter Gene Assay, much like Cerberus. Similar cleavage and are expected to confer improved pharmacoki experiments showed that Coco-mFc inhibits Nodal signaling. netic properties. The constructs below may be expressed, for 25 example, with a tRA leader sequence and an Fc sequence. Example 6

Human Cerberus-Fc is Degraded in Human Serum (SEQ ID NO: 2O) TRHODGRONQSSLSPVLLPRNQR(T) ELPTGNHEEAEEKYDLFVAVPH 30 The stability of Cerberus polypeptides in the presence of LVATSPAGEGOROREKMLSRFGRFWKKPEREMHPSRDSDSEPFPPGTOSL serum was evaluated. Conditioned medium from cells expressing human full-length Cerberus-Fe was incubated IQPIDGMKMEKSPLREEAKKFWHHFMFR (N) KTPASQGVILPIKSHEV overnight at 37 deg. C. with varying amounts of human serum (percentages of serum added are shown at top), and resolved HWETCRTVPFSOTITHEGCEKVVVONNLCFGKCGSVHFPGAAOHSHTSCS by SDS-PAGE. Western blot (FIG. 7) showed that Cerberus 35 HCLPAKFTTMHLPLNCTELSSVIKVVMLVEECQCKVKTEHEDGHILHIA was completely degraded when incubated with 5% human S. (N) GSODSFIPG (N) VSATG N-terminal sequencing of cleavage fragments revealed that proteolysis occurred at the following sites (cleavage shown It is expected that Coco will behave in a manner similar to 40 Cerberus, however, the two likely cleavage sites in Coco by ): occur within the cysteine knot domain at the sequences: 150 PARKRW 155 (SEQ ID NO: 25) and 168 SRRRVK 173 38 NQR ELP 43 (SEQ ID NO: 24) (SEQ ID NO: 26). Amino acids in these positions may be altered to eliminate the cleavage, with alanine and serine 138 MFR KTP 143 (SEQ ID NO: 23) 45 being preferred amino acids. In addition, or in the alternative, an N-linked glycosylation site (NXT/S) may be introduced at 2O7 SHCLPA 212 (SEQ ID NO: 28) or near either of these positions. The activity of the shortform of Cerberus Suggests that a minimal myostatin binding Example 7 domain of Coco is the cysteine knot, located at amino acids 50 101-185 of SEQ ID NO:5. Serum Stable Human Cerberus and Coco Coco constructs with one or more of the alterations (shown Polypeptides in brackets below) will have N-linked glycosylation sites that will block cleavage and are expected to confer improved To produce serum-stable Cerberus polypeptides, a variety pharmacokinetic properties. The constructs below may be of mutations may be introduced at cleavage sites and the 55 expressed, for example, with a tBA leader sequence and an Fc Surrounding sequences. In short forms of Cerberus, only the Sequence. L212 cleavage site remains (amino acid numbering is with reference to the full length, native Cerberus sequence, SEQ ID NO: 2), and so a mutation of any, some or all of the amino (SEQ ID NO: 21) acids in the sequence SHCLPA (SEQ ID NO: 28) may be 60 GRPEPOSPRPOSWAAANOTWALGPGAIPPLVPASALGSWKAFLGLOKARO altered to eliminate this cleavage site. Generally, mutations LGMGR (N) LO (T) RGODEVAAVTLPLNPQEVIQGMCKAVPFVOVFS will be to Small, uncharged groups, such as alanine or serine. Mutations of C211 and/or L212 to serine or alanine are par RPGCSAIRLRNHLCFGHCSSLYIPGSDPTPLVLCNSCMPAR (N) KR ticularly desirable. In addition, or in the alternative, an (T) WAPVVLWCLTGSSASRR (N) IR (A) V (S) KISTMLIEGCHC N-linked glycosylation site (NXT/S) may be introduced at a 65 position within the range of amino acids 202-222. An SPKA N-linked glycosylation site may also be introduced at a posi US 7,833,971 B2 41 42 Example 8 these proteins remained intact. Similar specific variants may be made with respect to Coco (relative to SEQID NO:5): Cysteine Variants of Cerberus and Coco C115G: In some proteins, odd numbers of cysteine residues result in a free Sulfhydryl group that may cause protein aggregation C145G: or otherwise interfere with protein production. Both native C162G. Coco and native Cerberus have odd numbers of cystein resi Therefore, the disclosure provides Cerberus and Coco vari dues. In order to improve the expression of Cerberus, variants ants in which one or more cysteine residues are deleted or with fewer cysteine residues were generated, as well as Vari 10 replaced. If replaced, the replacementamino acid may be any ants with changes in proximity to one or more of the cys of the other 19 canonical amino acids, although G, A, S and T teines. Relative to SEQID NO:2, the following variants were generated: are preferred.

C176G, 15 EQUIVALENTS C206G, A skilled artisan will recognize, or be able to ascertain C223G using no more than routine experimentation, many equiva N222D. lents to the specific embodiments of the inventions described Each of these proteins were expressible and retained bind- 20 herein. Such equivalents are intended to be encompassed by ing to GDF11, indicating that the biochemical activity of the following claims.

SEQUENCE LISTING

<16 Os NUMBER OF SEO ID NOS: 3 O

SEO ID NO 1 LENGTH: 272 TYPE PRT ORGANISM: Mus musculus

<4 OOs SEQUENCE: 1

Met His Lieu. Lieu. Lieu. Wall Glin Lieu. Luell Wall Lieu. Luell Pro Tell Gly 1. 5 15

Ala Asp Lieu. Cys Wall Asp Gly Glin Ser Glin Gly Ser Tell Ser Phe 25 3 O

Pro Leu. Lieu. Glu Arg Gly Arg Arg Asp Lieu. His Wall Ala ASn His Glu 35 4 O 45

Glu Ala Glu Asp Pro Asp Lieu. Phe Wall Ala Wall Pro His Leul Met SO 55 60

Gly Thir Ser Luell Ala Gly Glu Gly Glin Arg Glin Arg Gly Met Leu. 65 70 8O

Ser Arg Lieu. Gly Arg Phe Trp Pro Glu Thir Glu Phe Tyr Pro 85 90 95

Pro Arg Asp Wall Glu Ser Asp His Wall Ser Ser Gly Met Glin Ala Wall 1OO 105 110

Thir Glin Pro Ala Asp Gly Arg Llys Wall Glu Arg Ser Pro Tell Glin Glu 115 12O 125

Glu Ala Arg Phe Trp His Arg Phe Met Phe Arg Gly Pro Ala 13 O 135 14 O

Phe Glin Gly Wall Ile Leul Pro Ile Ser His Glu Wall His Trp Glu 145 15 O 155 16 O

Thir Arg Thir Wall Pro Phe Asn Gn. Thir Ile Ala His Glu Asp Cys 1.65 17 O 17s

Glin Wall Wall Wall Glin ASn Asn Lell Phe Gly Cys Ser Ser 18O 185 190

Ile Arg Phe Pro Gly Glu Gly Ala Asp Ala His Ser Phe Ser His 195 2 OO 2O5

Se Pro Thr Phe Thir Thir Wal His Lieu Met Luell ASn Thir US 7,833,971 B2 43 44

- Continued

21 O 215 22O Ser Pro Thr Pro Val Val Lys Met Val Met Glin Val Glu Glu. Cys Glin 225 23 O 235 24 O Cys Met Val Lys Thr Glu Arg Gly Glu Glu Arg Lieu Lleu Lieu Ala Gly 245 250 255 Ser Glin Gly Ser Phe Ile Pro Gly Lieu Pro Ala Ser Lys Thr Asn Pro 26 O 265 27 O

<210s, SEQ ID NO 2 &211s LENGTH: 267 212. TYPE: PRT <213> ORGANISM: Homo sapiens <4 OOs, SEQUENCE: 2 Met His Lieu. Lieu Lleu Phe Glin Lieu. Lieu Val Lieu Lleu Pro Lieu. Gly Lys 1. 5 1O 15 Thir Thr Arg His Glin Asp Gly Arg Glin Asn Glin Ser Ser Leu Ser Pro 2O 25 3O Val Lieu. Lieu Pro Arg Asn Glin Arg Glu Lieu Pro Thr Gly Asn His Glu 35 4 O 45 Glu Ala Glu Glu Lys Pro Asp Lieu. Phe Val Ala Val Pro His Lieu Val SO 55 6 O Ala Thir Ser Pro Ala Gly Glu Gly Glin Arg Glin Arg Glu Lys Met Lieu 65 70 7s 8O Ser Arg Phe Gly Arg Phe Trp Llys Llys Pro Glu Arg Glu Met His Pro 85 90 95 Ser Arg Asp Ser Asp Ser Glu Pro Phe Pro Pro Gly Thr Glin Ser Lieu. 1OO 105 11 O Ile Glin Pro Ile Asp Gly Met Lys Met Glu Lys Ser Pro Lieu. Arg Glu 115 12 O 125 Glu Ala Lys Llys Phe Trp His His Phe Met Phe Arg Llys Thr Pro Ala 13 O 135 14 O Ser Glin Gly Val Ile Leu Pro Ile Llys Ser His Glu Val His Trp Glu 145 150 155 160 Thr Cys Arg Thr Val Pro Phe Ser Glin Thr Ile Thr His Glu Gly Cys 1.65 17O 17s Glu Lys Val Val Val Glin Asn. Asn Lieu. Cys Phe Gly Lys Cys Gly Ser 18O 185 19 O Val His Phe Pro Gly Ala Ala Gln His Ser His Thr Ser Cys Ser His 195 2OO 2O5 Cys Lieu Pro Ala Lys Phe Thr Thr Met His Leu Pro Leu. Asn Cys Thr 21 O 215 22O Glu Lieu. Ser Ser Val Ile Llys Val Val Met Lieu Val Glu Glu. Cys Glin 225 23 O 235 24 O Cys Llys Wall Lys Thr Glu. His Glu Asp Gly His Ile Lieu. His Ala Gly 245 250 255 Ser Glin Asp Ser Phe Ile Pro Gly Val Ser Ala 26 O 265

<210s, SEQ ID NO 3 &211s LENGTH: 1752 &212s. TYPE: DNA <213s ORGANISM: Mus musculus 22 Os. FEATURE: <221s NAME/KEY: MOD RES <222s. LOCATION: (1235) . . (1235 <223> OTHER INFORMATION: a, c, g, t, unknown or other

US 7,833,971 B2 47 48

- Continued gaggttcc.ca caggcaiacca tgaggaagct gaggaga agc Cagatctgtt tt.cgcagtg 18O ccacaccittg tagccaccag CCCtgcaggg gaaggcc aga ggcagagaga galagatgctg 24 O tccagatttg gCaggttctg gaagaa.gc.ct gagagagaaa tgcatccatc caggg actica 3OO gatagtgagc cct tcc cacc tgggacccag t cc ct catcc. agc.cgataga tiggaatgaaa 360 atggagaaat citcct citt cq ggaagaa.gc.c aagaaattct ggcaccactt catgttcaga aaaact Cogg Cttcticaggg ggt catcttg CCC at Caaaa. gcc atgaagt acattgggag acctgcagga cagtgc cctt Cagc.ca.gact ata acccacg aaggctgttga aaaagtagtt 54 O gttcagaa.ca acctittgctt tgggaaatgc gggtctgttc attitt Cotgg agcc.gc.gcag cact cocata cct cotgctic t cactgtttg cctgccaagt t caccacgat gcacttgc.ca 660 ctgaactgca citgaactitt c citc.cgtgatc alaggtggtga tgctggtgga ggagtgc.ca.g 72 O tgcaaggtga agacggagca tgaagatgga CaCat CCtac atgctggct C C Caggattic C tittatcc.cag gag titt cago ttga 804

<210s, SEQ ID NO 5 &211s LENGTH: 189 212. TYPE: PRT <213> ORGANISM: Homo sapiens

<4 OOs, SEQUENCE: 5

Met Leu Lieu. Gly Gln Leu Ser Thr Lieu. Lieu. Cys Lell Lel Ser Gly Ala 1. 5 1O 15

Lell Pro Thr Gly Ser Gly Arg Pro Glu Pro Glin Ser Arg Pro Glin 25 3O

Ser Trp Ala Ala Ala Asn. Glin Thr Trp Ala Lieu. Gly Gly Ala Lieu. 35 4 O

Pro Pro Lieu Wall Pro Ala Ser Ala Lieu. Gly Ser Trp Ala Phe Lieu. SO 55 6 O

Gly Luell Gln Lys Ala Arg Glin Lieu. Gly Met Gly Arg Lel Glin Arg Gly 65 70 7s 8O

Glin Asp Glu Wall Ala Ala Wall Thr Luell Pro Luell Asn Pro Glin Glu Wall 85 90 95

Ile Glin Gly Met Cys Lys Ala Val Pro Phe Wall Glin Wall Phe Ser Arg 105 11 O

Pro Gly Cys Ser Ala Ile Arg Lieu. Arg Asn His Lell Cys Phe Gly His 115 12 O 125

Ser Ser Leu Tyr Ile Pro Gly Ser Asp Pro Thir Pro Lieu Wall Lieu. 13 O 135 14 O

Cys Asn Ser Cys Met Pro Ala Arg Ala Pro Wal Wall Lieu. 145 150 155 160

Trp Lieu. Thr Gly Ser Ser Ala Ser Arg Arg Arg Wall Lys Ile Ser 1.65 17O 17s

Thir Met Lieu. Ile Glu Gly Cys His Cys Ser Pro Ala 18O 185

SEQ ID NO 6 LENGTH: 1731 TYPE: DNA ORGANISM: Homo sapiens

< 4 OOs SEQUENCE: 6 agtc.cggaca gacaga Cagg Cagacagacg cacggacaag Cagatgct cc ttggc.ca.gct 6 O atcc actictt Ctgtgcctgc titagcggggc cct gcctaca ggctic aggga ggcctgaac C 12 O

US 7,833,971 B2 51

- Continued

<210s, SEQ ID NO 9 &211s LENGTH: 75 212. TYPE: PRT <213> ORGANISM: Homo sapiens

<4 OOs, SEQUENCE: 9 Lieu Lys Cys Val Cys Lieu. Lieu. Cys Asp Ser Ser Asn. Phe Thir Cys Glin 1. 5 1O 15 Thr Glu Gly Ala Cys Trp Ala Ser Val Met Lieu. Thir Asn Gly Lys Glu 2O 25 3O Glin Val Ile Llys Ser Cys Val Ser Lieu Pro Glu Lieu. Asn Ala Glin Val 35 4 O 45 Phe Cys His Ser Ser Asn Asn Val Thr Lys Thr Glu. Cys Cys Phe Thr SO 55 6 O Asp Phe Cys Asn. Asn. Ile Thr Lieu. His Lieu Pro 65 70 7s

<210s, SEQ ID NO 10 &211s LENGTH: 70 212. TYPE: PRT <213> ORGANISM: Homo sapiens

<4 OOs, SEQUENCE: 10 Ala Lieu. Lieu. Cys Ala Cys Thir Ser Cys Lieu. Glin Ala Asn Tyr Thr Cys 1. 5 1O 15 Glu Thr Asp Gly Ala Cys Met Val Ser Ile Phe Asn Lieu. Asp Gly Met 2O 25 3O Glu. His His Val Arg Thr Cys Ile Pro Llys Val Glu Lieu Val Pro Ala 35 4 O 45 Gly Llys Pro Phe Tyr Cys Lieu. Ser Ser Glu Asp Lieu. Arg Asn. Thir His SO 55 6 O Cys Cys Tyr Thr Asp Tyr 65 70

<210s, SEQ ID NO 11 &211s LENGTH: 267 212. TYPE: PRT <213> ORGANISM: Artificial Sequence 22 Os. FEATURE: <223> OTHER INFORMATION: Description of Artificial Sequence: Synthetic polypeptide

<4 OOs, SEQUENCE: 11 Met His Lieu. Lieu Lleu Phe Glin Lieu. Lieu Val Lieu Lleu Pro Lieu. Gly Lys 1. 5 1O 15 Thir Thr Arg His Glin Asp Gly Arg Glin Asn Glin Ser Ser Leu Ser Pro 2O 25 3O Val Lieu. Lieu Pro Arg Asn Glin Arg Glu Lieu Pro Thr Gly Asn His Glu 35 4 O 45 Glu Ala Glu Glu Lys Pro Asp Lieu. Phe Val Ala Val Pro His Lieu Val SO 55 6 O Ala Thir Ser Pro Ala Gly Glu Gly Glin Arg Glin Arg Glu Lys Met Lieu 65 70 7s 8O Ser Arg Phe Gly Arg Phe Trp Llys Llys Pro Glu Arg Glu Met His Pro 85 90 95 Ser Arg Asp Ser Asp Ser Glu Pro Phe Pro Pro Gly Thr Glin Ser Lieu. 1OO 105 11 O US 7,833,971 B2 53 54

- Continued Ile Glin Pro Ile Asp Gly Met Lys Met Glu Llys Ser Pro Luell Arg Glu 115 12 O 125

Glu Ala Lys Llys Phe Trp His His Phe Met Phe Arg Lys Thir Pro Ala 13 O 135 14 O

Ser Glin Gly Val Ile Leu Pro Ile Llys Ser His Glu. Wall His Trp Glu 145 150 155 160

Thr Cys Arg Thr Val Pro Phe Ser Glin Thr Ile Thr His Glu Gly 1.65 17O 17s

Glu Lys Val Val Val Glin Asn. Asn Lieu. Cys Phe Gly Cys Gly Ser 18O 185 19 O

Val His Phe Pro Gly Ala Ala Gln His Ser His Thir Ser Ser His 195 2OO 2O5

Cys Lieu Pro Ala Lys Phe Thr Thr Met His Leul Pro Lell Asn Thir 21 O 215 22O

Glu Lieu Ser Ser Val Ile Llys Val Val Met Lieu. Wall Glu Glu Glin 225 23 O 235 24 O

Cys Llys Wall Lys Thr Glu. His Glu Asp Gly His Ile Lell His Ala Gly 245 25 O 255

Ser Glin Asp Ser Phe Ile Pro Gly Val Ser Ala 26 O 265

SEQ ID NO 12 LENGTH: 496 TYPE PRT ORGANISM: Artificial Sequence FEATURE: OTHER INFORMATION: Description of Artificial Sequence: Synthetic polypeptide

SEQUENCE: 12

Met His Lieu. Lieu Lleu Phe Glin Lieu. Lieu Val Luell Luell Pro Luell Gly 1. 15

Thir Thr Arg His Glin Asp Gly Arg Glin Asn Glin Ser Ser Luell Ser Pro 2O 25 3O

Val Lieu. Lieu Pro Arg Asn Glin Arg Glu Lieu Pro Thir Gly Asn His Glu 35 4 O 45

Glu Ala Glu Glu Lys Pro Asp Lieu. Phe Val Ala Wall Pro His Luell Wall SO 55 6 O

Ala Thir Ser Pro Ala Gly Glu Gly Glin Arg Glin Arg Glu Met Luell 65 70 7s

Ser Arg Phe Gly Arg Phe Trp Llys Llys Pro Glu Arg Glu Met His Pro 85 90 95

Ser Arg Asp Ser Asp Ser Glu Pro Phe Pro Pro Gly Thir Glin Ser Luell 1OO 105 11 O

Ile Glin Pro Ile Asp Gly Met Lys Met Glu Llys Ser Pro Luell Arg Glu 115 12 O 125

Glu Ala Lys Llys Phe Trp His His Phe Met Phe Arg Thir Pro Ala 13 O 135 14 O

Ser Glin Gly Val Ile Leu Pro Ile Llys Ser His Glu. Wall His Trp Glu 145 150 155 160

Thr Cys Arg Thr Val Pro Phe Ser Glin Thr Ile Thr His Glu Gly 1.65 17O 17s

Glu Lys Val Val Val Glin Asn. Asn Lieu. Cys Phe Gly Cys Gly Ser 18O 185 19 O

Val His Phe Pro Gly Ala Ala Gln His Ser His Thir Ser Ser His 195 2OO 2O5 US 7,833,971 B2 55 56

- Continued

Luell Pro Ala Phe Thir Thir Met His Luell Pro Lieu. Asn. Cys Thir 21 O 215 22O

Glu Luell Ser Ser Wall Ile Wall Wall Met Luell Wall Glu Glu Cys Glin 225 23 O 235 24 O

Wall Thir Glu His Glu Asp Gly His Ile Lell His Ala Gly 245 250 255

Ser Glin Asp Ser Phe Ile Pro Gly Wall Ser Ala Thir Gly Gly Gly Thir 26 O 265 27 O

His Thir Cys Pro Pro Pro Ala Pro Glu Luell Lell Gly Gly Pro Ser 27s 285

Wall Phe Luell Phe Pro Pro Lys Pro Asp Thir Lell Met Ile Ser Arg 29 O 295 3 OO

Thir Pro Glu Wall Thir Cys Wall Wall Wall Asp Wall Ser His Glu Asp Pro 3. OS 310 315

Glu Wall Phe Asn Trp Wall Asp Gly Wall Glu Wall His Asn Ala 3.25 330 335

Thir Pro Arg Glu Glu Glin Tyr Asn Ser Thir Arg Wall Wall 34 O 345 35. O

Ser Wall Luell Thir Wall Lell His Glin Asp Trp Luell Asn Gly Lys Glu Tyr 355 360 365

Cys Wall Ser Asn Lys Ala Luell Pro Wall Pro Ile Glu Thir 37 O 375

Ile Ser Ala Gly Glin Pro Arg Glu Pro Glin Wall Thir Luell 385 390 395 4 OO

Pro Pro Ser Arg Glu Glu Met Thir Asn Glin Wall Ser Luell Thir 4 OS 415

Lell Wall Gly Phe Pro Ser Asp Ile Ala Wall Glu Trp Glu Ser 425 43 O

Asn Gly Glin Pro Glu Asn Asn Tyr Thir Thir Pro Pro Wall Luell Asp 435 44 O 445

Ser Asp Gly Ser Phe Phe Lell Ser Luell Thir Wall Asp Ser 450 45.5 460

Arg Trp Glin Glin Gly Asn Wall Phe Ser Ser Wall Met His Glu Ala 465 470

Lell His Asn His Tyr Thir Glin Ser Luell Ser Lell Ser Pro Gly Lys 485 490 495

SEQ ID NO 13 LENGTH: 341 TYPE : PRT ORGANISM: Artificial Sequence FEATURE: OTHER INFORMATION: Description of Artificial Sequence: Synthetic polypeptide

<4 OOs, SEQUENCE: 13

Glu Val His Trp Glu Thir Arg Thir Wall Pro Phe Ser Glin Thir Ile 1. 5 1O 15

Thir His Glu Gly Cys Glu Wall Wall Wall Glin Asn Asn Luell Cys Phe 2O 25

Gly Lys Cys Gly Ser Wall His Phe Pro Gly Ala Ala Glin His Ser His 35 4 O 45

Thir Ser Ser His Lell Pro Ala Phe Thir Thir Met His Luell SO 55 6 O

Pro Luell Asn Cys Thir Glu Lell Ser Ser Wall Ile Wall Wall Met Luell US 7,833,971 B2 57

- Continued

Val Glu Glu. Cys Glin Cys Llys Val Lys Thr Glu. His Glu Asp Gly His 85 90 95 Ile Lieu. His Ala Gly Ser Glin Asp Ser Phe Ile Pro Gly Val Ser Ala 1OO 105 11 O Thr Gly Gly Gly Thr His Thr Cys Pro Pro Cys Pro Ala Pro Glu Lieu. 115 12 O 125 Lieu. Gly Gly Pro Ser Val Phe Leu Phe Pro Pro Llys Pro Lys Asp Thr 13 O 135 14 O Lieu Met Ile Ser Arg Thr Pro Glu Val Thr Cys Val Val Val Asp Val 145 150 155 160 Ser His Glu Asp Pro Glu Val Llys Phe Asn Trp Tyr Val Asp Gly Val 1.65 17O 17s Glu Val His Asn Ala Lys Thr Llys Pro Arg Glu Glu Glin Tyr Asn. Ser 18O 185 19 O Thr Tyr Arg Val Val Ser Val Lieu. Thr Val Lieu. His Glin Asp Trp Leu 195 2OO 2O5 Asn Gly Lys Glu Tyr Lys Cys Llys Val Ser Asn Lys Ala Lieu Pro Val 21 O 215 22O Pro Ile Glu Lys Thir Ile Ser Lys Ala Lys Gly Glin Pro Arg Glu Pro 225 23 O 235 24 O Glin Val Tyr Thr Lieu Pro Pro Ser Arg Glu Glu Met Thr Lys Asn Glin 245 250 255 Val Ser Lieu. Thr Cys Lieu Val Lys Gly Phe Tyr Pro Ser Asp Ile Ala 26 O 265 27 O Val Glu Trp Glu Ser Asn Gly Glin Pro Glu Asn Asn Tyr Lys Thr Thr 27s 28O 285 Pro Pro Val Lieu. Asp Ser Asp Gly Ser Phe Phe Lieu. Tyr Ser Lys Lieu. 29 O 295 3 OO Thr Val Asp Llys Ser Arg Trp Glin Glin Gly Asn Val Phe Ser Cys Ser 3. OS 310 315 32O Val Met His Glu Ala Lieu. His Asn His Tyr Thr Gln Lys Ser Leu Ser 3.25 330 335 Lieu. Ser Pro Gly Lys 34 O

<210s, SEQ ID NO 14 &211s LENGTH: 21 212. TYPE: PRT <213> ORGANISM: Apis sp. <4 OOs, SEQUENCE: 14 Met Llys Phe Leu Val Asn Val Ala Leu Val Phe Met Val Val Tyr Ile 1. 5 1O 15 Ser Tyr Ile Tyr Ala 2O

<210s, SEQ ID NO 15 &211s LENGTH: 22 212. TYPE: PRT <213> ORGANISM: Homo sapiens

<4 OOs, SEQUENCE: 15 Met Asp Ala Met Lys Arg Gly Lieu. Cys Cys Val Lieu. Lieu. Lieu. Cys Gly 1. 5 1O 15

Ala Wall Phe Wal Ser Pro US 7,833,971 B2 59

- Continued

<210s, SEQ ID NO 16 &211s LENGTH: 17 212. TYPE: PRT <213> ORGANISM: Homo sapiens

<4 OOs, SEQUENCE: 16 Met His Lieu. Lieu Lleu Phe Glin Lieu. Lieu Val Lieu Lleu Pro Lieu. Gly Lys 1. 5 1O 15

Thir

<210s, SEQ ID NO 17 &211s LENGTH: 418 212. TYPE: PRT <213> ORGANISM: Artificial Sequence 22 Os. FEATURE: <223> OTHER INFORMATION: Description of Artificial Sequence: Synthetic polypeptide

<4 OOs, SEQUENCE: 17 Met Lieu. Lieu. Gly Glin Lieu. Ser Thr Lieu. Lieu. Cys Lieu Lleu Ser Gly Ala 1. 5 1O 15 Lieu Pro Thr Gly Ser Gly Arg Pro Glu Pro Glin Ser Pro Arg Pro Glin 2O 25 3O Ser Trp Ala Ala Ala Asn Glin Thir Trp Ala Lieu. Gly Pro Gly Ala Lieu. 35 4 O 45 Pro Pro Lieu Val Pro Ala Ser Ala Lieu. Gly Ser Trp Lys Ala Phe Lieu. SO 55 6 O Gly Lieu Gln Lys Ala Arg Glin Lieu. Gly Met Gly Arg Lieu. Glin Arg Gly 65 70 7s 8O Glin Asp Glu Val Ala Ala Val Thir Lieu Pro Lieu. ASn Pro Glin Glu Val 85 90 95 Ile Glin Gly Met Cys Lys Ala Val Pro Phe Val Glin Val Phe Ser Arg 1OO 105 11 O Pro Gly Cys Ser Ala Ile Arg Lieu. Arg Asn His Lieu. Cys Phe Gly His 115 12 O 125 Cys Ser Ser Leu Tyr Ile Pro Gly Ser Asp Pro Thr Pro Leu Val Lieu. 13 O 135 14 O Cys Asn. Ser Cys Met Pro Ala Arg Lys Arg Trp Ala Pro Val Val Lieu. 145 150 155 160 Trp. Cys Lieu. Thr Gly Ser Ser Ala Ser Arg Arg Arg Val Lys Ile Ser 1.65 17O 17s Thr Met Lieu. Ile Glu Gly Cys His Cys Ser Pro Lys Ala Thr Gly Gly 18O 185 19 O Gly Thr His Thr Cys Pro Pro Cys Pro Ala Pro Glu Lieu. Leu Gly Gly 195 2OO 2O5 Pro Ser Val Phe Lieu. Phe Pro Pro Llys Pro Lys Asp Thr Lieu Met Ile 21 O 215 22O Ser Arg Thr Pro Glu Val Thr Cys Val Val Val Asp Val Ser His Glu 225 23 O 235 24 O Asp Pro Glu Val Llys Phe Asn Trp Tyr Val Asp Gly Val Glu Val His 245 250 255 Asn Ala Lys Thr Llys Pro Arg Glu Glu Glin Tyr Asn. Ser Thr Tyr Arg 26 O 265 27 O Val Val Ser Val Lieu. Thr Val Lieu. His Glin Asp Trp Lieu. Asn Gly Lys 27s 28O 285 US 7,833,971 B2 61 62

- Continued

Glu Tyr Lys Wall Ser Asn Ala Luell Pro Wall Pro Ile Glu 29 O 295 3 OO

Lys Thir Ile Ser Lys Ala Gly Glin Pro Arg Glu Pro Glin Wall Tyr 3. OS 310 315

Thir Luell Pro Pro Ser Arg Glu Glu Met Thir Asn Glin Wall Ser Luell 3.25 330 335

Thir Luell Wall Lys Gly Phe Pro Ser Ile Ala Wall Glu Trp 34 O 345 35. O

Glu Ser Asn Gly Glin Pro Glu Asn Asn Tyr Thir Thir Pro Pro Wall 355 360 365

Lell Asp Ser Asp Gly Ser Phe Phe Luell Tyr Ser Lys Lell Thir Wall Asp 37 O 375

Lys Ser Arg Trp Glin Glin Gly Asn Wall Phe Ser Ser Wall Met His 385 390 395 4 OO

Glu Ala Luell His Asn His Thir Glin Lys Ser Lell Ser Luell Ser Pro 4 OS 41O 415 Gly Lys

SEQ ID NO 18 LENGTH: 328 TYPE : PRT ORGANISM: Artificial Sequence FEATURE: OTHER INFORMATION: Description of Artificial Sequence: Synthetic polypeptide

<4 OOs, SEQUENCE: 18

Lell Asn. Pro Glin Glu Wall Ile Glin Gly Met Ala Wall Pro Phe 1. 5 1O 15

Wall Glin Wall Phe Ser Arg Pro Gly Cys Ser Ala Ile Arg Luell Arg Asn 25

His Luell Cys Phe Gly His Ser Ser Luell Tyr Ile Pro Gly Ser Asp 35 4 O 45

Pro Thir Pro Luell Wall Lell Cys Asn Ser Met Pro Ala Arg SO 55 6 O

Trp Ala Pro Wall Wall Lell Trp Luell Thir Gly Ser Ser Ala Ser Arg 65 70

Arg Arg Wall Ile Ser Thir Met Luell Ile Glu Gly His Cys Ser 85 90 95

Pro Ala Thir Gly Gly Gly Thir His Thir Pro Pro Cys Pro Ala 105 11 O

Pro Glu Luell Luell Gly Gly Pro Ser Wall Phe Luell Phe Pro Pro Pro 115 12 O 125

Asp Thir Luell Met Ile Ser Arg Thir Pro Glu Wall Thir Wall Wall 13 O 135 14 O

Wall Asp Wall Ser His Glu Asp Pro Glu Wall Lys Phe Asn Trp Wall 145 150 155 160

Asp Gly Wall Glu Wall His Asn Ala Thir Pro Arg Glu Glu Glin 1.65 17s

Asn Ser Thir Tyr Arg Wall Wall Ser Wall Luell Thir Wall Luell His Glin 18O 185 19 O

Asp Trp Luell Asn Gly Glu Tyr Wall Ser Asn Ala 195 2O5

Lell Pro Wall Pro Ile Glu Lys Thir Ile Ser Lys Ala Gly Glin Pro 21 O 215 22O US 7,833,971 B2 63 64

- Continued

Arg Glu Pro Glin Wall Tyr Thir Luell Pro Pro Ser Arg Glu Glu Met Thir 225 23 O 235 24 O

Asn Glin Wall Ser Lell Thir Luell Wall Lys Gly Phe Pro Ser 245 250 255

Ile Ala Wall Glu Trp Glu Ser Asn Gly Glin Pro Glu Asn Asn 26 O 265 27 O

Thir Thir Pro Pro Wall Lell Asp Ser Asp Gly Ser Phe Phe Luell 27s 28O 285

Ser Lys Luell Thir Wall Asp Lys Ser Arg Trp Glin Glin Gly Asn Wall Phe 29 O 295 3 OO

Ser Ser Wall Met His Glu Ala Luell His ASn His Thir Glin Lys 3. OS 310 315 32O

Ser Luell Ser Luell Ser Pro Gly 3.25

<210s, SEQ ID NO 19 &211s LENGTH: 21 212. TYPE : PRT <213> ORGANISM: Homo sapiens

<4 OOs, SEQUENCE: 19 Met Lieu. Lieu. Gly Glin Lieu. Ser Thr Lieu. Lieu. Cys Lieu Lleu Ser Gly Ala 1. 5 15 Lieu Pro Thr Gly Ser

D NO NGT H: 252 PE: PRT ORGANISM: Artificial Sequence ATU RE: HER INFORMATION: Description of Artificial Sequence: Synthetic polypeptide &22 Os. FEATU RE: NAME/ KEY: MOD RES LOCATION: (23) . . (23) HER INFORMATION: Arg o Thir ATU RE: NAME/ KEY: MOD RES LOCATION: (123) . . (123) HER INFORMATION: Arg o Asn ATU RE: NAME/ KEY: MOD RES LOCATION: (238) ... (238) HER INFORMATION: Ala o Asn ATU RE: NAME/ KEY: MOD RES LOCATION: (247) ... (247) ER INFORMATION: Gly or Asn <4 OOs, SEQUENCE:

Thir Arg is Glin Asp Gly Arg Glin Asn Glin Ser Ser Lell Ser Pro Wall 1. 5 15

Lell Luell Arg Asn Glin Xaa Glu Luell Pro Thir Gly Asn His Glu Glu 2O 25

Ala Glu Glu Pro Asp Lell Phe Wall Ala Wall Pro His Luell Wall Ala 35 4 O 45

Thir Ser Pro Ala Gly Glu Gly Glin Arg Glin Arg Glu Met Luell Ser SO 55 6 O

Arg Phe Gly Arg Phe Trp Pro Glu Arg Glu Met His Pro Ser 65 70

Arg Asp Ser Asp Ser Glu Pro Phe Pro Pro Gly Thir Glin Ser Luell Ile US 7,833,971 B2 65

- Continued

85 90 95 Glin Pro Ile Asp Gly Met Lys Met Glu Lys Ser Pro Lieu. Arg Glu Glu 1OO 105 11 O Ala Lys Llys Phe Trp His His Phe Met Phe Xaa Lys Thr Pro Ala Ser 115 12 O 125 Gln Gly Val Ile Leu Pro Ile Llys Ser His Glu Val His Trp Glu Thr 13 O 135 14 O Cys Arg Thr Val Pro Phe Ser Glin Thr Ile Thr His Glu Gly Cys Glu 145 150 155 160 Llys Val Val Val Glin Asn. Asn Lieu. Cys Phe Gly Lys Cys Gly Ser Val 1.65 17O 17s His Phe Pro Gly Ala Ala Gln His Ser His Thir Ser Cys Ser His Cys 18O 185 19 O Lieu Pro Ala Lys Phe Thir Thr Met His Leu Pro Leu. Asn Cys Thr Glu 195 2OO 2O5 Lieu. Ser Ser Val Ile Llys Val Val Met Lieu Val Glu Glu. Cys Glin Cys 21 O 215 22O Llys Val Llys Thr Glu. His Glu Asp Gly His Ile Lieu. His Xaa Gly Ser 225 23 O 235 24 O Gln Asp Ser Phe Ile Pro Xaa Val Ser Ala Thr Gly 245 250

<210s, SEQ ID NO 21 &211s LENGTH: 168 212. TYPE: PRT <213> ORGANISM: Artificial Sequence 22 Os. FEATURE: <223> OTHER INFORMATION: Description of Artificial Sequence: Synthetic polypeptide 22 Os. FEATURE: <221s NAME/KEY: MOD RES <222s. LOCATION: (55) . . (55) <223> OTHER INFORMATION: Arg or Asn 22 Os. FEATURE: <221s NAME/KEY: MOD RES <222s. LOCATION: (57) . . (57) 223 OTHER INFORMATION: Gin or Thr 22 Os. FEATURE: <221s NAME/KEY: MOD RES <222s. LOCATION: (131) ... (131) <223> OTHER INFORMATION: Arg or Asn 22 Os. FEATURE: <221s NAME/KEY: MOD RES <222s. LOCATION: (133) . . (133) <223> OTHER INFORMATION: Arg or Thr 22 Os. FEATURE: <221s NAME/KEY: MOD RES <222s. LOCATION: (150) ... (150) <223> OTHER INFORMATION: Arg or Asn 22 Os. FEATURE: <221s NAME/KEY: MOD RES <222s. LOCATION: (151) ... (151) <223> OTHER INFORMATION: Arg or Ala 22 Os. FEATURE: <221s NAME/KEY: MOD RES <222s. LOCATION: (152) ... (152) 223 OTHER INFORMATION: Wall or Ser

<4 OOs, SEQUENCE: 21 Gly Arg Pro Glu Pro Glin Ser Pro Arg Pro Glin Ser Trp Ala Ala Ala 1. 5 1O 15 Asn Glin Thir Trp Ala Lieu. Gly Pro Gly Ala Leu Pro Pro Leu Val Pro 2O 25 3O Ala Ser Ala Lieu. Gly Ser Trp Lys Ala Phe Lieu. Gly Lieu. Glin Lys Ala 35 4 O 45 US 7,833,971 B2 67

- Continued

Arg Glin Lieu. Gly Met Gly Xaa Lieu. Xaa Arg Gly Glin Asp Glu Val Ala SO 55 6 O Ala Val Thr Lieu Pro Lieu. Asn Pro Glin Glu Val Ile Glin Gly Met Cys 65 70 7s 8O Lys Ala Val Pro Phe Val Glin Val Phe Ser Arg Pro Gly Cys Ser Ala 85 90 95 Ile Arg Lieu. Arg Asn His Lieu. Cys Phe Gly. His Cys Ser Ser Lieu. Tyr 1OO 105 11 O Ile Pro Gly Ser Asp Pro Thr Pro Leu Val Lieu. Cys Asn Ser Cys Met 115 12 O 125 Pro Ala Xala Lys Xaa Trp Ala Pro Val Val Lieu. Trp Cys Lieu. Thr Gly 13 O 135 14 O Ser Ser Ala Ser Arg Xaa Xaa Xala Lys Ile Ser Thr Met Lieu. Ile Glu 145 150 155 160 Gly Cys His Cys Ser Pro Lys Ala 1.65

<210s, SEQ ID NO 22 &211s LENGTH: 7 212. TYPE: PRT <213> ORGANISM: Homo sapiens

<4 OOs, SEQUENCE: 22 Ser His Cys Lieu Pro Ala Lys 1. 5

<210s, SEQ ID NO 23 &211s LENGTH: 6 212. TYPE: PRT <213> ORGANISM: Homo sapiens

<4 OOs, SEQUENCE: 23 Met Phe Arg Llys Thr Pro 1. 5

<210s, SEQ ID NO 24 &211s LENGTH: 6 212. TYPE: PRT <213> ORGANISM: Homo sapiens

<4 OOs, SEQUENCE: 24 Asn Glin Arg Glu Lieu Pro 1. 5

<210s, SEQ ID NO 25 &211s LENGTH: 6 212. TYPE: PRT <213> ORGANISM: Homo sapiens <4 OOs, SEQUENCE: 25 Pro Ala Arg Lys Arg Trp 1. 5

<210s, SEQ ID NO 26 &211s LENGTH: 6 212. TYPE: PRT <213> ORGANISM: Homo sapiens <4 OOs, SEQUENCE: 26 Ser Arg Arg Arg Val Lys 1. 5 US 7,833,971 B2 69 70

- Continued

SEO ID NO 27 LENGTH: 4 TYPE PRT ORGANISM: Artificial Sequence FEATURE: OTHER INFORMATION: Description of Artificial Sequence: Synthetic peptide

<4 OOs, SEQUENCE: 27 Thr Gly Gly Gly 1.

SEQ ID NO 28 LENGTH: 6 TYPE PRT ORGANISM: Homo sapiens <4 OOs, SEQUENCE: 28 Ser His Cys Lieu Pro Ala 1. 5

SEQ ID NO 29 LENGTH: 6 TYPE PRT ORGANISM: Artificial Sequence FEATURE: OTHER INFORMATION: Description of Artificial Sequence: Synthetic peptide <22O > FEATURE: <221 > NAME/KEY: MOD RES <222> LOCATION: (3) ... (3) <223s OTHER INFORMATION: Ser or Thr

<4 OOs, SEQUENCE: 29

ASn Glin Xala Glu Lieu Pro 1. 5

SEQ ID NO 3 O LENGTH: 108 TYPE: DNA ORGANISM: Artificial Sequence FEATURE: OTHER INFORMATION: Description of Artificial Sequence: Synthetic polynucleotide

< 4 OOs SEQUENCE: 3 O agc.ca.gacaa gccagacaag C cagacaa.gc Cagacaa.gcc agacaa.gc.ca gacaa.gc.ca.g 6 O acaa.gc.ca.ga caa.gc.ca.gac aagccagaca agc.ca.gacaa gcc agaca 108

We claim: 3. The myostatin antagonist protein of claim 1, wherein the 1. An isolated myostatin antagonist protein, the protein protein retains at least 50% of the myostatin antagonist activ ity after exposure to human serum for 24 hours at 37°C. comprising an amino acid sequence that has at least 90% 55 4. The myostatin antagonist protein of claim 1, wherein identity to the sequence of amino acids 22-185 of human myostatin antagonist activity is assessed in an A204 cell Coco (SEQ ID NO:5), wherein said protein comprises a based assay. modification with respect to the sequence of amino acids 5. The myostatin antagonist protein of claim 1, wherein the 22-185 of SEQID NO: 5 such that cleavage in human serum 60 protein comprises a modification with respect to the amino acid sequence of SEQ ID NO:5 that reduces or eliminates is reduced or eliminated, and wherein said protein is Substan cleavage within one or both of the following sequences of tially serum stable for a period of at least 24 hours. human Coco: PARKRW (SEQ ID NO: 25) and SRRRVK 2. The myostatin antagonist protein of claim 1, wherein the (SEQID NO: 26). 6. A pharmaceutical preparation comprising a myostatin protein comprises an amino acid sequence that has at least 65 90% identity to the sequence of amino acids 22-189 of human antagonist protein of claim 1. Coco. k k k k k