Variant ActRIIB proteins and uses thereof

Abstract

Variant ActRIIB proteins and uses thereof are provided. In certain embodiments, the present invention provides compositions and methods for modulating (promoting or inhibiting) the growth of red blood cells or tissues such as bone, cartilage, muscle, fat, and / or nerve tissue. The present invention also provides methods for screening for compounds that modulate the activity of ActRIIB protein and / or ActRIIB ligand. The compositions and methods provided herein are useful in treating diseases associated with abnormal activity of ActRIIB protein and / or ActRIIB ligand.

Description

[Technical Field]

[0001] This application claims the benefit of priority from U.S. Provisional Application No. 62 / 404,718, filed October 5, 2016, the specification of which is incorporated herein by reference in its entirety. [Background technology]

[0002] The transforming growth factor-beta (TGF-beta) superfamily contains a variety of growth factors that share common sequence elements and structural motifs. These proteins are known to exert biological effects on a wide variety of cell types in both vertebrates and invertebrates. Members of the superfamily play important functions in pattern formation and tissue specification during embryonic development and can influence various differentiation processes, including adipogenesis, myogenesis, chondrogenesis, cardiogenesis, hematopoiesis, neurogenesis, and epithelial cell differentiation. The family is divided into two general branches: the BMP / GDF and the TGF-beta / activin / BMP10 branches, and these members have diverse, often complementary, effects. Manipulating the activity of TGF-beta family members can often induce significant physiological changes in organisms. For example, Piedmontese and Belgian Blue cattle breeds carry loss-of-function mutations in the GDF8 (also known as myostatin) gene that cause a significant increase in muscle mass. Grobet et al., Nat Genet. (1997) 17(1):71-4. Furthermore, in humans, inactive alleles of GDF8 are associated with increased muscle mass and reportedly exceptional strength. Schuelke et al., N Engl J Med (2004), 350:2682-8. [Prior art documents] [Non-patent literature]

[0003] [Non-Patent Document 1] Grobet et al., Nat Genet. (1997) 17(1):71-4 [Non-patent document 2] Schuelke et al., N Engl J Med (2004), 350:2682-8 Summary of the Invention [Means for solving the problem]

[0004] Changes in red blood cell levels, bone, cartilage, and other tissues can be achieved by agonizing or antagonizing signaling mediated by appropriate TGF-beta family members. Thus, there is a need for agents that function as potent regulators of TGF-beta signaling. In certain aspects, the present disclosure provides variant ActRIIB polypeptides, particularly variant ActRIIB homomultimeric proteins and variant ActRIIB heteromultimeric proteins. As demonstrated by the examples, several variant ActRIIB polypeptides have been identified that exhibit altered binding affinity to one or more ActRIIB-binding ligands. ActRIIB variants with reduced and increased ligand binding activity have been identified. Such variants may be particularly useful for selectively increasing or decreasing ligand selectivity compared to the corresponding unmodified ActRIIB polypeptide in various applications. For example, the examples further demonstrate that some variant ActRIIB polypeptides have various in vivo effects, including, for example, the ability to increase body mass (e.g., muscle mass) and increase red blood cell and hemoglobin levels. Thus, variant ActRIIB polypeptides should be useful in various therapeutic applications, including, for example, those described herein.

[0005] In certain aspects, the present disclosure provides a method for identifying a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, 100%, 101%, 102%, 103%, 104%, 105%, 106%, 107%, 108%, 109%, 110%, 111%, 112%, 113%, 114%, 115%, 116%, 117%, 118%, 119%, 120%, 121%, 122%, 123%, 124%, 125%, 126%, 127%, 128%, 129%, 130%, 131%, 132%, 133%, or 134 identical to an amino acid sequence beginning with any one of amino acids 20-29 (e.g., 20, 21, 22, 23, 24, 25, 26, 27, 28, or 29) of SEQ ID NO:2 and ending with any one of amino acids 109-134 (e.g., 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, or 134) of SEQ ID NO:2.

[0014] The present invention relates to a variant ActRIIB polypeptide comprising an amino acid sequence that is 1%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to amino acids 29-109 of SEQ ID NO:2, wherein the variant ActRIIB polypeptide comprises one or more amino acid substitutions at positions selected from the group consisting of K55, F82, L79, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78, D80, and F82, as well as heteromultimeric complexes comprising one or more such ActRIIB polypeptides. In some embodiments, the ActRIIB polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to amino acids 29-109 of SEQ ID NO:2. In some embodiments, the ActRIIB polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 25-131 of SEQ ID NO: 2. In some embodiments, the ActRIIB polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 20-134 of SEQ ID NO: 2. In some embodiments, the ActRIIB polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO:53.In some embodiments, the ActRIIB polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 12. In some embodiments, the ActRIIB polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 5. In some embodiments, the polypeptide comprises an amino acid substitution at the amino acid position corresponding to K55 of SEQ ID NO: 2. For example, in some embodiments, the substitution is K55A. In some embodiments, the substitution is K55E. In some embodiments, the polypeptide comprises an amino acid substitution at the amino acid position corresponding to L79 of SEQ ID NO: 2. For example, in some embodiments, the substitution is L79D. In some embodiments, the substitution is L79E. In some embodiments, the substitution is L79P. In some embodiments, the substitution is L79A. In some embodiments, the polypeptide comprises an amino acid substitution at an amino acid position corresponding to F82 in SEQ ID NO:2. For example, in some embodiments, the substitution is F82I. In some embodiments, the substitution is F82K. In some embodiments, the substitution is F82A. In some embodiments, the polypeptide comprises an amino acid substitution at an amino acid position corresponding to A24 in SEQ ID NO:2. For example, in some embodiments, the substitution is A24N. In some embodiments, the polypeptide comprises an amino acid substitution at an amino acid position corresponding to K74 in SEQ ID NO:2. For example, in some embodiments, the substitution is K74A. In some embodiments, the substitution is K74F. In some embodiments, the substitution is K74A. In some embodiments, the substitution is K74I. In some embodiments, the substitution is K74A. In some embodiments, the substitution is K74Y. In some embodiments, the polypeptide comprises an amino acid substitution at the amino acid position corresponding to D80 in SEQ ID NO: 2. For example, in some embodiments, the substitution is D80A. In some embodiments, the substitution is D80F.In some embodiments, the substitution is D80K. In some embodiments, the substitution is D80G. In some embodiments, the substitution is D80M. In some embodiments, the substitution is D80I. In some embodiments, the substitution is D80N. In some embodiments, the substitution is D80R. In some embodiments, the polypeptide comprises an amino acid substitution at an amino acid position corresponding to R64 of SEQ ID NO:2. For example, in some embodiments, the substitution is R64K. In some embodiments, the substitution is R64N. In some embodiments, the polypeptide comprises an amino acid substitution at an amino acid position corresponding to P129 of SEQ ID NO:2. For example, in some embodiments, the substitution is P129S. In some embodiments, the polypeptide comprises an amino acid substitution at an amino acid position corresponding to P130 of SEQ ID NO:2. For example, in some embodiments, the substitution is P130A. In some embodiments, the substitution is P130R. In some embodiments, the polypeptide comprises an amino acid substitution at an amino acid position corresponding to E37 of SEQ ID NO:2. For example, in some embodiments, the substitution is E37A. In some embodiments, the polypeptide comprises an amino acid substitution at an amino acid position corresponding to R40 in SEQ ID NO:2. For example, in some embodiments, the substitution is R40A. In some embodiments, the polypeptide comprises an amino acid substitution at an amino acid position corresponding to D54 in SEQ ID NO:2. For example, in some embodiments, the substitution is D54A. In some embodiments, the polypeptide comprises an amino acid substitution at an amino acid position corresponding to R56 in SEQ ID NO:2. For example, in some embodiments, the substitution is R56A. In some embodiments, the polypeptide comprises an amino acid substitution at an amino acid position corresponding to W78 in SEQ ID NO:2. For example, in some embodiments, the substitution is W78A.

[0006] In certain aspects, the present disclosure relates to a variant ActRIIB polypeptide comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 31. In some embodiments, the variant ActRIIB polypeptide comprises an alanine at a position corresponding to K55 of SEQ ID NO: 2.

[0007] In certain aspects, the present disclosure relates to a variant ActRIIB polypeptide comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 33. In some embodiments, the variant ActRIIB polypeptide comprises an alanine at a position corresponding to K55 of SEQ ID NO: 2.

[0008] In certain aspects, the present disclosure relates to a variant ActRIIB polypeptide comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 34. In some embodiments, the variant ActRIIB polypeptide comprises a glutamic acid at a position corresponding to K55 of SEQ ID NO: 2.

[0009] In certain aspects, the present disclosure relates to a variant ActRIIB polypeptide comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 36. In some embodiments, the variant ActRIIB polypeptide comprises a glutamic acid at a position corresponding to K55 of SEQ ID NO: 2.

[0010] In certain aspects, the present disclosure relates to a variant ActRIIB polypeptide comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of SEQ ID NO: 37. In some embodiments, the variant ActRIIB polypeptide comprises an isoleucine at a position corresponding to F82 of SEQ ID NO: 2.

[0011] In certain aspects, the present disclosure relates to a variant ActRIIB polypeptide comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 39. In some embodiments, the variant ActRIIB polypeptide comprises an isoleucine at a position corresponding to F82 of SEQ ID NO: 2.

[0012] In certain aspects, the present disclosure relates to a variant ActRIIB polypeptide comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 40. In some embodiments, the variant ActRIIB polypeptide comprises a lysine at a position corresponding to F82 of SEQ ID NO: 2.

[0013] In certain aspects, the present disclosure relates to a variant ActRIIB polypeptide comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 42. In some embodiments, the variant ActRIIB polypeptide comprises a lysine at a position corresponding to F82 of SEQ ID NO: 2.

[0014] In certain aspects, the present disclosure relates to a variant ActRIIB polypeptide comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of SEQ ID NO: 43. In some embodiments, the variant ActRIIB polypeptide comprises a glutamic acid at a position corresponding to L79 of SEQ ID NO: 2.

[0015] In certain aspects, the present disclosure relates to a variant ActRIIB polypeptide comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 45. In some embodiments, the variant ActRIIB polypeptide comprises a glutamic acid at a position corresponding to L79 of SEQ ID NO: 2.

[0016] In certain aspects, the present disclosure relates to a variant ActRIIB polypeptide comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 46. In some embodiments, the variant ActRIIB polypeptide comprises a glutamic acid at a position corresponding to L79 of SEQ ID NO: 2.

[0017] In certain aspects, the present disclosure relates to a variant ActRIIB polypeptide comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 48. In some embodiments, the variant ActRIIB polypeptide comprises a glutamic acid at a position corresponding to L79 of SEQ ID NO: 2.

[0018] In certain aspects, the present disclosure relates to a variant ActRIIB polypeptide comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 49. In some embodiments, the variant ActRIIB polypeptide comprises a glutamic acid at a position corresponding to L79 of SEQ ID NO: 2.

[0019] In certain aspects, the present disclosure relates to a variant ActRIIB polypeptide comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 50. In some embodiments, the variant ActRIIB polypeptide comprises a glutamic acid at a position corresponding to L79 of SEQ ID NO: 2.

[0020] In certain aspects, the present disclosure relates to a variant ActRIIB polypeptide comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 51. In some embodiments, the variant ActRIIB polypeptide comprises a glutamic acid at a position corresponding to L79 of SEQ ID NO: 2.

[0021] In certain aspects, the present disclosure relates to a variant ActRIIB polypeptide comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of SEQ ID NO: 52. In some embodiments, the variant ActRIIB polypeptide comprises a glutamic acid at a position corresponding to L79 of SEQ ID NO: 2.

[0022] In certain aspects, the variant ActRIIB polypeptides of the present disclosure form homodimers. In some embodiments, the variant ActRIIB polypeptides can form heterodimers through covalent interactions. In some embodiments, the variant ActRIIB polypeptides can form heterodimers through non-covalent interactions. In some embodiments, the variant ActRIIB polypeptides can form heterodimers through both covalent and non-covalent interactions.

[0023] In certain aspects, the variant ActRIIB polypeptide, including its homomultimer (e.g., homodimer), binds to one or more TGF-beta superfamily ligands. In some embodiments, the variant ActRIIB polypeptide, including its homomultimer, binds to at least 1×10 -7 K of M D and bind to one or more TGF-beta superfamily ligands. In some embodiments, the one or more TGF-beta superfamily ligands are selected from the group consisting of BMP6, BMP7, BMP9, BMP10, GDF3, GDF7, GDF8, GDF11, GDF15, activin A, activin B, activin C, activin E, activin AB, activin AC, activin AE, activin BC, and activin BE.

[0024] In certain aspects, the variant ActRIIB polypeptide comprising its homomultimer (e.g., homodimer) inhibits one or more TGF-beta superfamily ligands. In some embodiments, the variant ActRIIB polypeptide comprising its homomultimer inhibits signaling of one or more TGF-beta superfamily ligands. In some embodiments, the variant ActRIIB polypeptide comprising its homomultimer inhibits Smad signaling of one or more TGF-beta superfamily ligands. In some embodiments, the variant ActRIIB polypeptide comprising its homomultimer inhibits signaling of one or more TGF-beta superfamily ligands in a cell-based assay. In some embodiments, the variant ActRIIB polypeptide, including its homomultimers, inhibits one or more TGF-beta superfamily ligands selected from the group consisting of BMP6, BMP7, BMP9, BMP10, GDF3, GDF7, GDF8, GDF11, GDF15, activin A, activin B, activin C, activin E, activin AB, activin AC, activin AE, activin BC and activin BE.

[0025] In certain aspects, the present disclosure relates to heteromultimers comprising at least one variant ActRIIB polypeptide (e.g., one or more variant ActRIIB polypeptides described herein). For example, in some embodiments, a heteromultimeric protein of the present disclosure comprises a first ActRIIB polypeptide and a second ActRIIB polypeptide, wherein the first ActRIIB polypeptide begins with any one of amino acids 20-29 (e.g., 20, 21, 22, 23, 24, 25, 26, 27, 28, or 29) of SEQ ID NO:2 and begins with any one of amino acids 109-134 (e.g., 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 220, 221, 222, 223, 224, 225, 226, 227, 228, or 229) of SEQ ID NO:2 and begins with any one of amino acids 109-134 (e.g., 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, or 129) of SEQ ID NO:2. 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, or 134), and , D80, and F82, and the second ActRIIB polypeptide begins with any one of amino acids 20-29 (e.g., 20, 21, 22, 23, 24, 25, 26, 27, 28, or 29) of SEQ ID NO:2 and continues with any one of amino acids 109-134 (e.g., 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133 or 134), and the first ActRIIB polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to an amino acid sequence ending in any one of the following: 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133 or 134), and the first ActRIIB polypeptide comprises a different amino acid sequence compared to the second ActRIIB polypeptide.In some embodiments, the first polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 29-109 of SEQ ID NO: 2. In some embodiments, the second polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 29-109 of SEQ ID NO: 2. In some embodiments, the first polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 25-131 of SEQ ID NO: 2. In some embodiments, the second polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 25-131 of SEQ ID NO: 2. In some embodiments, the first polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 20-134 of SEQ ID NO: 2. In some embodiments, the second polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 25-131 of SEQ ID NO: 2. In some embodiments, the first polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 53. In some embodiments, the second polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO:53.In some embodiments, the first polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 5. In some embodiments, the second polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 5. In some embodiments, the first polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 12. In some embodiments, the second polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 12. In some embodiments, the first polypeptide comprises an amino acid substitution at an amino acid position corresponding to K55 of SEQ ID NO: 2. For example, in some embodiments, the substitution is K55A. In some embodiments, the substitution is K55E. In some embodiments, the first polypeptide comprises an amino acid substitution at an amino acid position corresponding to L79 of SEQ ID NO: 2. For example, in some embodiments, the substitution is L79D. In some embodiments, the substitution is L79E. In some embodiments, the substitution is L79P. In some embodiments, the substitution is L79A. In some embodiments, the first polypeptide comprises an amino acid substitution at an amino acid position corresponding to F82 of SEQ ID NO: 2. For example, in some embodiments, the substitution is F82I. In some embodiments, the substitution is F82K. In some embodiments, the substitution is F82A. In some embodiments, the first polypeptide comprises an amino acid substitution at the amino acid position corresponding to A24 of SEQ ID NO:2. For example, in some embodiments, the substitution is A24N. In some embodiments, the polypeptide comprises an amino acid substitution at the amino acid position corresponding to K74 of SEQ ID NO:2. For example, in some embodiments, the substitution is K74A.In some embodiments, the substitution is K74A. In some embodiments, the substitution is K74F. In some embodiments, the substitution is K74A. In some embodiments, the substitution is K74I. In some embodiments, the substitution is K74A. In some embodiments, the substitution is K74Y. In some embodiments, the first polypeptide comprises an amino acid substitution at an amino acid position corresponding to D80 of SEQ ID NO:2. For example, in some embodiments, the substitution is D80A. In some embodiments, the substitution is D80F. In some embodiments, the substitution is D80K. In some embodiments, the substitution is D80G. In some embodiments, the substitution is D80M. In some embodiments, the substitution is D80I. In some embodiments, the substitution is D80N. In some embodiments, the substitution is D80R. In some embodiments, the first polypeptide comprises an amino acid substitution at an amino acid position corresponding to R64 of SEQ ID NO:2. For example, in some embodiments, the substitution is R64K. In some embodiments, the substitution is R64N. In some embodiments, the first polypeptide comprises an amino acid substitution at an amino acid position corresponding to P129 in SEQ ID NO:2. For example, in some embodiments, the substitution is P129S. In some embodiments, the first polypeptide comprises an amino acid substitution at an amino acid position corresponding to P130 in SEQ ID NO:2. For example, in some embodiments, the substitution is P130A. In some embodiments, the substitution is P130R. In some embodiments, the first polypeptide comprises an amino acid substitution at an amino acid position corresponding to E37 in SEQ ID NO:2. For example, in some embodiments, the substitution is E37A. In some embodiments, the first polypeptide comprises an amino acid substitution at an amino acid position corresponding to R40 in SEQ ID NO:2. For example, in some embodiments, the substitution is R40A. In some embodiments, the first polypeptide comprises an amino acid substitution at an amino acid position corresponding to D54 in SEQ ID NO:2. For example, in some embodiments, the substitution is D54A. In some embodiments, the first polypeptide comprises an amino acid substitution at an amino acid position corresponding to R56 in SEQ ID NO:2. For example, in some embodiments, the substitution is R56A.In some embodiments, the first polypeptide comprises an amino acid substitution at an amino acid position corresponding to W78 of SEQ ID NO: 2. For example, in some embodiments, the substitution is W78A. In some embodiments, the second ActRIIB polypeptide comprises one or more amino acid substitutions at positions of SEQ ID NO: 2 selected from the group consisting of A24, K74, R64, P129, P130, E37, R40, D54, R56, W78, D80, and F82. For example, in some embodiments, the second ActRIIB polypeptide comprises one or more amino acid substitutions relative to the amino acid sequence of SEQ ID NO: 2 selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79A, L79D, L79E, L79P, P129S, P130A, P130R, E37A, R40A, D54A, K55A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N, D80R, and F82A. In some embodiments, the first ActRIIB polypeptide and / or the second ActRIIB polypeptide comprises one or more amino acid modifications that promote heteromultimer formation. In some embodiments, the first ActRIIB polypeptide and / or the second ActRIIB polypeptide comprises one or more amino acid modifications that inhibit heteromultimer formation. In some embodiments, the first ActRIIB polypeptide and / or the second ActRIIB polypeptide comprises one or more amino acid modifications that promote heteromultimer formation and one or more amino acid modifications that inhibit heteromultimer formation. In some embodiments, the heteromultimer of the present disclosure is a heterodimer.

[0026] In certain aspects, the ActRIIB polypeptides of the present disclosure, including variant ActRIIB polypeptides, are fusion proteins comprising an ActRIIB polypeptide domain and one or more heterologous domains. In some embodiments, the ActRIIB polypeptide is an ActRIIB-Fc fusion protein. In some embodiments, the ActRIIB-Fc fusion protein further comprises a linker domain located between the ActRIIB polypeptide domain and one or more heterologous domains or the Fc domain. In some embodiments, the linker domain is selected from TGGG, TGGGG, SGGGG, GGGGS, GGG, GGGG, SGGG, and GGGGS.

[0027] In certain aspects, the present disclosure relates to an ActRIIB heteromultimeric protein comprising a first ActRIIB-Fc fusion protein and a second ActRIIB-Fc fusion protein, wherein the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 13, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 13.

[0028] In certain aspects, the present disclosure relates to an ActRIIB heteromultimeric protein comprising a first ActRIIB-Fc fusion protein and a second ActRIIB-Fc fusion protein, wherein the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 14, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 14.

[0029] In certain aspects, the present disclosure relates to an ActRIIB heteromultimeric protein comprising a first ActRIIB-Fc fusion protein and a second ActRIIB-Fc fusion protein, wherein the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 15, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 15.

[0030] In certain aspects, the present disclosure relates to an ActRIIB heteromultimeric protein comprising a first ActRIIB-Fc fusion protein and a second ActRIIB-Fc fusion protein, wherein the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 16, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 16.

[0031] In certain aspects, the present disclosure relates to an ActRIIB heteromultimeric protein comprising a first ActRIIB-Fc fusion protein and a second ActRIIB-Fc fusion protein, wherein the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 17, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 17.

[0032] In certain aspects, the present disclosure relates to an ActRIIB heteromultimeric protein comprising a first ActRIIB-Fc fusion protein and a second ActRIIB-Fc fusion protein, wherein the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 18, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 19.

[0033] In certain aspects, the present disclosure relates to an ActRIIB heteromultimeric protein comprising a first ActRIIB-Fc fusion protein and a second ActRIIB-Fc fusion protein, wherein the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 19, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 18.

[0034] In certain aspects, the present disclosure relates to an ActRIIB heteromultimeric protein comprising a first ActRIIB-Fc fusion protein and a second ActRIIB-Fc fusion protein, wherein the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 20, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 21.

[0035] In certain aspects, the present disclosure relates to an ActRIIB heteromultimeric protein comprising a first ActRIIB-Fc fusion protein and a second ActRIIB-Fc fusion protein, wherein the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 21, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 20.

[0036] In certain aspects, the present disclosure relates to an ActRIIB heteromultimeric protein comprising a first ActRIIB-Fc fusion protein and a second ActRIIB-Fc fusion protein, wherein the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 22, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 23.

[0037] In certain aspects, the present disclosure relates to an ActRIIB heteromultimeric protein comprising a first ActRIIB-Fc fusion protein and a second ActRIIB-Fc fusion protein, wherein the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 23, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 22.

[0038] In certain aspects, the present disclosure relates to an ActRIIB heteromultimeric protein comprising a first ActRIIB-Fc fusion protein and a second ActRIIB-Fc fusion protein, wherein the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 22, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 25.

[0039] In certain aspects, the present disclosure relates to an ActRIIB heteromultimeric protein comprising a first ActRIIB-Fc fusion protein and a second ActRIIB-Fc fusion protein, wherein the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 25, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 24.

[0040] In certain aspects, the present disclosure relates to an ActRIIB heteromultimeric protein comprising a first ActRIIB-Fc fusion protein and a second ActRIIB-Fc fusion protein, wherein the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 26, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 27.

[0041] In certain aspects, the present disclosure relates to an ActRIIB heteromultimeric protein comprising a first ActRIIB-Fc fusion protein and a second ActRIIB-Fc fusion protein, wherein the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 27, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 26.

[0042] In certain aspects, the present disclosure relates to an ActRIIB heteromultimeric protein comprising a first ActRIIB-Fc fusion protein and a second ActRIIB-Fc fusion protein, wherein the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 28, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 29. In some embodiments, the first ActRIIB-Fc fusion protein Fc domain comprises a cysteine ​​at amino acid position 132, a glutamic acid at amino acid position 138, a tryptophan at amino acid position 144, and an aspartic acid at amino acid position 217. In some embodiments, the second ActRIIB-Fc fusion protein Fc domain comprises a cysteine ​​at amino acid position 127, a serine at amino acid position 144, an alanine at position 146, an arginine at amino acid position 162, an arginine at amino acid position 179, and a valine at amino acid position 185.

[0043] In certain aspects, the present disclosure relates to an ActRIIB heteromultimeric protein comprising a first ActRIIB-Fc fusion protein and a second ActRIIB-Fc fusion protein, wherein the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 28, and the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 29. In some embodiments, the second ActRIIB-Fc fusion protein Fc domain comprises a cysteine ​​at amino acid position 132, a glutamic acid at amino acid position 138, a tryptophan at amino acid position 144, and an aspartic acid at amino acid position 217. In some embodiments, the first ActRIIB-Fc fusion protein Fc domain comprises a cysteine ​​at amino acid position 127, a serine at amino acid position 144, an alanine at position 146, an arginine at amino acid position 162, an arginine at amino acid position 179, and a valine at amino acid position 185.

[0044] In certain aspects, the present disclosure relates to an ActRIIB heteromultimeric protein comprising a first ActRIIB-Fc fusion protein and a second ActRIIB-Fc fusion protein, wherein the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 30, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 23. In some embodiments, the first ActRIIB-Fc fusion protein Fc domain comprises a cysteine ​​at amino acid position 132, a tryptophan at amino acid position 144, and an arginine at amino acid position 435. In some embodiments, the second ActRIIB-Fc fusion protein Fc domain comprises a cysteine ​​at amino acid position 127, a serine at amino acid position 144, an alanine at amino acid position 146, and a valine at amino acid position 185.

[0045] In certain aspects, the present disclosure relates to an ActRIIB heteromultimeric protein comprising a first ActRIIB-Fc fusion protein and a second ActRIIB-Fc fusion protein, wherein the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 30, and the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 23. In some embodiments, the second ActRIIB-Fc fusion protein Fc domain comprises a cysteine ​​at amino acid position 132, a tryptophan at amino acid position 144, and an arginine at amino acid position 435. In some embodiments, the first ActRIIB-Fc fusion protein Fc domain comprises a cysteine ​​at amino acid position 127, a serine at amino acid position 144, an alanine at amino acid position 146, and a valine at amino acid position 185.

[0046] In certain aspects, the present disclosure relates to a heteromultimer comprising a first ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 33, and a second ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 5, wherein the first ActRIIB polypeptide does not comprise the amino acid sequence of the second ActRIIB polypeptide. In some embodiments, the first ActRIIB polypeptide comprises an alanine at the amino acid position corresponding to 55 of SEQ ID NO: 2. In some embodiments, the second ActRIIB polypeptide does not comprise an alanine at the amino acid position corresponding to 55 of SEQ ID NO: 2. In some embodiments, the second ActRIIB polypeptide comprises a lysine at an amino acid position corresponding to 55 of SEQ ID NO: 2. In some embodiments, the first ActRIIB polypeptide comprises one or more amino acid substitutions at amino acid positions corresponding to any one of F82, L79, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78, D80 and F82 of SEQ ID NO: 2. In some embodiments, the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79A, L79D, L79E, L79P, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N, D80R and F82A. In some embodiments, the second ActRIIB polypeptide includes one or more amino acid substitutions at amino acid positions corresponding to any one of F82, L79, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78, D80 and F82 of SEQ ID NO: 2.In some embodiments, the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79A, L79D, L79E, L79P, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N, D80R, and F82A. In some embodiments, the first ActRIIB polypeptide and / or the second ActRIIB polypeptide comprise one or more amino acid modifications that promote heteromultimer formation. In some embodiments, the first ActRIIB polypeptide and / or the second ActRIIB polypeptide comprise one or more amino acid modifications that inhibit heteromultimer formation. In some embodiments, the heteromultimer is a heterodimer.

[0047] In certain aspects, the present disclosure relates to a heteromultimer comprising a first ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 36, and a second ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 5, wherein the first ActRIIB polypeptide does not comprise the amino acid sequence of the second ActRIIB polypeptide. In some embodiments, the first ActRIIB polypeptide comprises a glutamic acid at the amino acid position corresponding to 55 of SEQ ID NO: 2. In some embodiments, the second ActRIIB polypeptide does not comprise a glutamic acid at the amino acid position corresponding to 55 of SEQ ID NO: 2. In some embodiments, the second ActRIIB polypeptide comprises a lysine at an amino acid position corresponding to 55 of SEQ ID NO: 2. In some embodiments, the first ActRIIB polypeptide comprises one or more amino acid substitutions at amino acid positions corresponding to any one of F82, L79, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78, D80 and F82 of SEQ ID NO: 2. In some embodiments, the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79A, L79D, L79E, L79P, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N, D80R and F82A. In some embodiments, the second ActRIIB polypeptide includes one or more amino acid substitutions at amino acid positions corresponding to any one of F82, L79, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78, D80 and F82 of SEQ ID NO: 2.In some embodiments, the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79A, L79D, L79E, L79P, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N, D80R, and F82A. In some embodiments, the first ActRIIB polypeptide and / or the second ActRIIB polypeptide comprise one or more amino acid modifications that promote heteromultimer formation. In some embodiments, the first ActRIIB polypeptide and / or the second ActRIIB polypeptide comprise one or more amino acid modifications that inhibit heteromultimer formation. In some embodiments, the heteromultimer is a heterodimer.

[0048] In certain aspects, the present disclosure relates to a heteromultimer comprising a first ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 39, and a second ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 5, wherein the first ActRIIB polypeptide does not comprise the amino acid sequence of the second ActRIIB polypeptide. In some embodiments, the first ActRIIB polypeptide comprises an isoleucine at the amino acid position corresponding to 82 of SEQ ID NO: 2. In some embodiments, the second ActRIIB polypeptide does not contain an isoleucine acid at the amino acid position corresponding to 82 of SEQ ID NO: 2. In some embodiments, the second ActRIIB polypeptide contains a phenylalanine at the amino acid position corresponding to 82 of SEQ ID NO: 2. In some embodiments, the first ActRIIB polypeptide contains one or more amino acid substitutions at amino acid positions corresponding to any one of L79, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78 and D80 of SEQ ID NO: 2. In some embodiments, the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79A, L79D, L79E, L79P, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N and D80R. In some embodiments, the second ActRIIB polypeptide comprises one or more amino acid substitutions at amino acid positions corresponding to any one of L79, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78 and D80 of SEQ ID NO:2.In some embodiments, the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79A, L79D, L79E, L79P, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N, and D80R. In some embodiments, the first ActRIIB polypeptide and / or the second ActRIIB polypeptide comprise one or more amino acid modifications that promote heteromultimer formation. In some embodiments, the first ActRIIB polypeptide and / or the second ActRIIB polypeptide comprise one or more amino acid modifications that inhibit heteromultimer formation. In some embodiments, the heteromultimer is a heterodimer.

[0049] In certain aspects, the present disclosure relates to a heteromultimer comprising a first ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 42, and a second ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 5, wherein the first ActRIIB polypeptide does not comprise the amino acid sequence of the second ActRIIB polypeptide. In some embodiments, the first ActRIIB polypeptide comprises a lysine at the amino acid position corresponding to 82 of SEQ ID NO: 2. In some embodiments, the second ActRIIB polypeptide does not comprise a lysine at the amino acid position corresponding to 82 of SEQ ID NO: 2. In some embodiments, the second ActRIIB polypeptide comprises a phenylalanine at an amino acid position corresponding to 82 of SEQ ID NO: 2. In some embodiments, the first ActRIIB polypeptide comprises one or more amino acid substitutions at amino acid positions corresponding to any one of L79, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78 and D80 of SEQ ID NO: 2. In some embodiments, the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79A, L79D, L79E, L79P, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N and D80R. In some embodiments, the second ActRIIB polypeptide comprises one or more amino acid substitutions at amino acid positions corresponding to any one of L79, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78 and D80 of SEQ ID NO:2.In some embodiments, the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79A, L79D, L79E, L79P, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N, and D80R. In some embodiments, the first ActRIIB polypeptide and / or the second ActRIIB polypeptide comprise one or more amino acid modifications that promote heteromultimer formation. In some embodiments, the first ActRIIB polypeptide and / or the second ActRIIB polypeptide comprise one or more amino acid modifications that inhibit heteromultimer formation. In some embodiments, the heteromultimer is a heterodimer.

[0050] In certain aspects, the present disclosure relates to a heteromultimer comprising a first ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 45, and a second ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 48, wherein the first ActRIIB polypeptide does not comprise the amino acid sequence of the second ActRIIB polypeptide. In some embodiments, the first ActRIIB polypeptide comprises an acidic amino acid position corresponding to 79 of SEQ ID NO: 2. In some embodiments, the acidic amino acid is aspartic acid. In some embodiments, the acidic amino acid is glutamic acid. In some embodiments, the second ActRIIB polypeptide does not contain an acidic amino acid (e.g., aspartic acid or glutamic acid) at the amino acid position corresponding to 79 of SEQ ID NO: 2. In some embodiments, the second ActRIIB polypeptide contains a leucine at the amino acid position corresponding to 79 of SEQ ID NO: 2. In some embodiments, the first ActRIIB polypeptide contains one or more amino acid substitutions at amino acid positions corresponding to any one of F82, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78, D80 and F82 of SEQ ID NO: 2. In some embodiments, the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79P, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N, D80R, and F82A. In some embodiments, the second ActRIIB polypeptide comprises one or more amino acid substitutions at amino acid positions corresponding to any one of F82, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78, D80, and F82 of SEQ ID NO:2.In some embodiments, the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N, D80R, and F82A. In some embodiments, the first ActRIIB polypeptide and / or the second ActRIIB polypeptide comprise one or more amino acid modifications that promote heteromultimer formation. In some embodiments, the first ActRIIB polypeptide and / or the second ActRIIB polypeptide comprise one or more amino acid modifications that inhibit heteromultimer formation. In some embodiments, the heteromultimer is a heterodimer.

[0051] In certain aspects, the present disclosure relates to a heteromultimer comprising a first ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 50, and a second ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 52, wherein the first ActRIIB polypeptide does not comprise the amino acid sequence of the second ActRIIB polypeptide. In some embodiments, the first ActRIIB polypeptide comprises an acidic amino acid position corresponding to 79 of SEQ ID NO: 2. In some embodiments, the acidic amino acid is aspartic acid. In some embodiments, the acidic amino acid is glutamic acid. In some embodiments, the second ActRIIB polypeptide does not contain an acidic amino acid (e.g., aspartic acid or glutamic acid) at the amino acid position corresponding to 79 of SEQ ID NO: 2. In some embodiments, the second ActRIIB polypeptide contains a leucine at the amino acid position corresponding to 79 of SEQ ID NO: 2. In some embodiments, the first ActRIIB polypeptide contains one or more amino acid substitutions at amino acid positions corresponding to any one of F82, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78, D80 and F82 of SEQ ID NO: 2. In some embodiments, the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79P, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N, D80R, and F82A. In some embodiments, the second ActRIIB polypeptide comprises one or more amino acid substitutions at amino acid positions corresponding to any one of F82, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78, D80, and F82 of SEQ ID NO:2.In some embodiments, the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N, D80R and F82A.

[0052] In certain embodiments, the present disclosure provides a heteromultimeric protein comprising any of the ActRIIB polypeptides disclosed herein and a second polypeptide or functional fragment thereof selected from the group consisting of ALK1, ALK2, ALK3, ALK4, ALK5, ALK6, ALK7, ActRIIA, TGFBRII, BMPRII, and MISRII polypeptides. In some embodiments, the second polypeptide is an ALK1 polypeptide or a functional fragment thereof. In some embodiments, the ALK1 polypeptide or functional fragment thereof comprises an amino acid sequence at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to SEQ ID NO: 54 or a functional fragment thereof. In some embodiments, the ALK1 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of any one of SEQ ID NOs: 54, 55, 56, 57, 60, and 61, or a functional fragment thereof. In some embodiments, the second polypeptide is an ALK2 polypeptide or functional fragment thereof. In some embodiments, the ALK2 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of SEQ ID NO: 64 or 65, or a functional fragment thereof. In some embodiments, the ALK2 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NOs: 64, 65, 66, 67, 70 and 71. In some embodiments, the second polypeptide is an ALK3 polypeptide or functional fragment thereof.In some embodiments, the ALK3 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 74. In some embodiments, the ALK3 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 74, 75, 76, 77, 80 or 81. In some embodiments, the second polypeptide is an ALK4 polypeptide or functional fragment thereof. In some embodiments, the ALK4 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 84 or 85. In some embodiments, the ALK4 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 84, 86, 85, 87, 88, 89, 92 and 93. In some embodiments, the second polypeptide is an ALK5 polypeptide or functional fragment thereof. In some embodiments, the ALK5 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 96 or 97. In some embodiments, the ALK5 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 96, 98, 97, 99, 100, 101, 104 and 105. In some embodiments, the second polypeptide is an ALK6 polypeptide or functional fragment thereof.In some embodiments, the ALK6 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 108 or 110. In some embodiments, the ALK6 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 108, 109, 110, 111, 112, 113, 116 and 117. In some embodiments, the second polypeptide is an ALK7 polypeptide or functional fragment thereof. In some embodiments, the ALK7 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 120, 121 or 122. In some embodiments, the ALK7 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 120, 123, 124, 125, 121, 126, 122, 127, 128, 129, 130, 133 and 134. In some embodiments, the second polypeptide is an ActRIIA polypeptide or functional fragment thereof. In some embodiments, the ActRIIA polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 137. In some embodiments, the ActRIIA polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NOs: 137, 138, 139, 140, 141, 144 and 145. In some embodiments, the second polypeptide is a TGFBRII polypeptide or functional fragment thereof.In some embodiments, the TGFBRII polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 204. In some embodiments, the TGFBRII polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NOs: 161, 162, 160, 163, 164, 165, 166, 167, 172, 173, 174 and 175. In some embodiments, the second polypeptide is a BMPRII polypeptide or functional fragment thereof. In some embodiments, the BMPRII polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 148 or 149. In some embodiments, the BMPRII polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 148, 150, 149, 151, 152, 153, 156 and 157. In some embodiments, the second polypeptide is a MISRII polypeptide or functional fragment thereof. In some embodiments, the MISRII polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 180, 181 or 182. In some embodiments, the MISRII polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 180, 183, 181, 184, 182 and 185, or functional fragment thereof.

[0053] In certain aspects, the heteromultimers of the present disclosure bind to one or more TGF-beta superfamily ligands. In some embodiments, the heteromultimers of the present disclosure bind to at least 1×10 -7 K of M D and bind to one or more TGF-beta superfamily ligands. In some embodiments, the one or more TGF-beta superfamily ligands are selected from the group consisting of BMP6, BMP7, BMP9, BMP10, GDF3, GDF7, GDF8, GDF11, GDF15, activin A, activin B, activin C, activin E, activin AB, activin AC, activin AE, activin BC, and activin BE.

[0054] In certain aspects, the heteromultimers of the present disclosure inhibit one or more TGF-beta superfamily ligands. In some embodiments, the heteromultimers of the present disclosure inhibit signal transduction of one or more TGF-beta superfamily ligands. In some embodiments, the heteromultimers of the present disclosure inhibit Smad signal transduction of one or more TGF-beta superfamily ligands. In some embodiments, the heteromultimers of the present disclosure inhibit signal transduction of one or more TGF-beta superfamily ligands in cell-based assays. In some embodiments, the heteromultimers of the present disclosure inhibit one or more TGF-beta superfamily ligands selected from the group consisting of BMP6, BMP7, BMP9, BMP10, GDF3, GDF7, GDF8, GDF11, GDF15, activin A, activin B, activin C, activin E, activin AB, activin AC, activin AE, activin BC, and activin BE.

[0055] In certain aspects, the present disclosure relates to ActRIIB polypeptides, including variant ActRIIB polypeptides, and homomultimers and heteromultimers comprising the same, including variant ActRIIB polypeptides comprising one or more amino acid modifications selected from the group consisting of glycosylated amino acids, PEGylated amino acids, farnesylated amino acids, acetylated amino acids, biotinylated amino acids, and amino acids conjugated to a lipid moiety. In some embodiments, the ActRIIB polypeptides of the present disclosure are glycosylated and have a glycosylation pattern obtainable from the polypeptide in CHO cells.

[0056] In certain embodiments, the present disclosure provides a heteromultimeric protein comprising any of the ActRIIB polypeptides disclosed herein and a second polypeptide or functional fragment thereof selected from the group consisting of ALK1, ALK2, ALK3, ALK4, ALK5, ALK6, ALK7, ActRIIA, TGFBRII, BMPRII, and MISRII polypeptides. In some embodiments, the second polypeptide is an ALK1 polypeptide or a functional fragment thereof. In some embodiments, the ALK1 polypeptide or functional fragment thereof comprises an amino acid sequence at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to SEQ ID NO: 54 or a functional fragment thereof. In some embodiments, the ALK1 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of any one of SEQ ID NOs: 54, 55, 56, 57, 58, 59, 60, 61, 62, and 63. In some embodiments, the second polypeptide is an ALK2 polypeptide or functional fragment thereof. In some embodiments, the ALK2 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of SEQ ID NO: 64 or 65. In some embodiments, the ALK2 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NOs: 64, 65, 66, 67, 68, 69, 70, 71, 72 and 73. In some embodiments, the second polypeptide is an ALK3 polypeptide or functional fragment thereof.In some embodiments, the ALK3 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 74. In some embodiments, the ALK3 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 74, 75, 76, 77, 80 or 81. In some embodiments, the second polypeptide is an ALK4 polypeptide or functional fragment thereof. In some embodiments, the ALK4 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 84 or 85. In some embodiments, the ALK4 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 84, 86, 85, 87, 88, 89, 90, 91, 92, 93, 94 and 95. In some embodiments, the second polypeptide is an ALK5 polypeptide or functional fragment thereof. In some embodiments, the ALK5 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 96 or 97. In some embodiments, the ALK5 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 96, 98, 97, 99, 100, 101, 102, 103, 104, 105, 106 and 107. In some embodiments, the second polypeptide is an ALK6 polypeptide or functional fragment thereof.In some embodiments, the ALK6 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 108 or 110. In some embodiments, the ALK6 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118 and 119. In some embodiments, the second polypeptide is an ALK7 polypeptide or functional fragment thereof. In some embodiments, the ALK7 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 120, 121 or 122. In some embodiments, the ALK7 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 120, 123, 124, 125, 121, 126, 122, 127, 128, 129, 130, 131, 132, 133, 134, 135 and 136. In some embodiments, the second polypeptide is an ActRIIA polypeptide or functional fragment thereof. In some embodiments, the ActRIIA polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 137. In some embodiments, the ActRIIA polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 137, 138, 139, 140, 141, 142, 143, 144, 145, 146 and 147, or functional fragment thereof.In some embodiments, the second polypeptide is a TGFBRII polypeptide or a functional fragment thereof. In some embodiments, the TGFBRII polypeptide or a functional fragment thereof comprises an amino acid sequence or a functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of SEQ ID NO: 204. In some embodiments, the TGFBRII polypeptide or a functional fragment thereof comprises an amino acid sequence or a functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 161, 162, 160, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, and 179, or a functional fragment thereof. In some embodiments, the second polypeptide is a BMPRII polypeptide or a functional fragment thereof. In some embodiments, the BMPRII polypeptide or a functional fragment thereof comprises an amino acid sequence or a functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 148 or 149. In some embodiments, the BMPRII polypeptide or a functional fragment thereof comprises an amino acid sequence or a functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 148, 150, 149, 151, 152, 153, 154, 155, 156, 157, 158 and 159. In some embodiments, the second polypeptide is a MISRII polypeptide or a functional fragment thereof. In some embodiments, the MISRII polypeptide or functional fragment thereof comprises an amino acid sequence that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 180, 181 or 182, or a functional fragment thereof.In some embodiments, the MISRII polypeptide or functional fragment thereof comprises an amino acid sequence that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NOs: 180, 183, 181, 184, 182, 185, 186, 187, 188, 189, 190, 191, 192 and 193, or a functional fragment thereof.

[0057] In certain aspects, the present disclosure relates to pharmaceutical preparations comprising ActRIIB polypeptides, including variant ActRIIB polypeptides, and homomultimers and heteromultimers comprising the same, and a pharmaceutically acceptable carrier. In some embodiments, the pharmaceutical preparations comprising one or more ActRIIB heteromultimers contain less than about 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, or less than about 1% homomultimers.

[0058] In certain aspects, the present disclosure relates to isolated and / or recombinant nucleic acids comprising a coding sequence for one or more of the ActRIIB polypeptides described herein. For example, in some embodiments, the present disclosure relates to isolated and / or recombinant nucleic acids that are at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to a nucleic acid sequence corresponding to any one of SEQ ID NOs: 3, 10, 31, 35, 38, 41, 44, or 47. In some embodiments, the isolated and / or recombinant polynucleotide sequences of the present disclosure comprise a promoter sequence operably linked to a coding sequence described herein (e.g., a nucleic acid that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to a nucleic acid sequence corresponding to any one of SEQ ID NOs: 3, 10, 31, 35, 38, 41, 44 or 47). In some embodiments, the present disclosure relates to a vector comprising an isolated and / or recombinant nucleic acid described herein (e.g., a nucleic acid that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to a nucleic acid sequence corresponding to any one of SEQ ID NOs: 3, 10, 31, 35, 38, 41, 44 or 47). In some embodiments, the disclosure relates to a cell comprising an isolated and / or recombinant polynucleotide sequence described herein (e.g., a nucleic acid that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to a nucleic acid sequence corresponding to any one of SEQ ID NOs: 3, 10, 31, 35, 38, 41, 44, or 47). In some embodiments, the cell is a CHO cell. In some embodiments, the cell is a COS cell.

[0059] In certain aspects, the present disclosure relates to methods for producing ActRIIB polypeptides, including variant ActRIIB polypeptides described herein, and homomultimers and heteromultimers comprising the same. Such methods can include expressing any of the nucleic acids disclosed herein in suitable cells (e.g., CHO cells or COS cells). Such methods can include a) culturing the cells under conditions suitable for expression of a soluble ActRIIB polypeptide, wherein the cells contain an ActRIIB polypeptide expression construct. In some embodiments, the methods further include recovering the expressed ActRIIB polypeptide. The ActRIIB polypeptide can be recovered as a crude fraction, a partially purified fraction, or a highly purified fraction using any of the well-known techniques for obtaining proteins from cell culture.

[0060] In some embodiments, the present disclosure relates to a method for increasing red blood cell or hemoglobin levels in a patient, the method comprising administering to a patient in need thereof an ActRIIB polypeptide, including a variant ActRIIB polypeptide described herein, and homomultimers and heteromultimers comprising the same.

[0061] In some embodiments, the present disclosure relates to a method for treating anemia or an anemia-related disorder (e.g., a disorder described herein) in a patient, the method comprising administering to a patient in need thereof an ActRIIB polypeptide, including a variant ActRIIB polypeptide described herein, and homomultimers and heteromultimers comprising the same.

[0062] In some embodiments, the present disclosure relates to a method for increasing muscle mass and / or strength in a patient, the method comprising administering to a patient in need thereof an ActRIIB polypeptide, including a variant ActRIIB polypeptide described herein, and homomultimers and heteromultimers comprising the same.

[0063] In some embodiments, the present disclosure relates to a method for treating a muscle-related disorder in a patient, the method comprising administering ActRIIB, including a variant ActRIIB polypeptide described herein, and homomultimers and heteromultimers comprising the same, to a patient in need thereof. In some embodiments, the disorder is associated with undesirably low muscle growth and / or muscle weakness. Such disorders include muscle atrophy, muscular dystrophy, amyotrophic lateral sclerosis (ALS), and muscle wasting disorders (e.g., cachexia, anorexia, DMD syndrome, BMD syndrome, AIDS wasting syndrome, muscular dystrophies, neuromuscular diseases, motor neuron diseases, diseases of the neuromuscular junction, and inflammatory myopathies).

[0064] In some embodiments, the present disclosure relates to a method for reducing body fat content or decreasing the rate of increase in body fat content and treating disorders associated with unwanted weight gain, such as obesity, non-insulin-dependent diabetes mellitus (NIDDM), cardiovascular disease, cancer, high blood pressure, osteoarthritis, stroke, respiratory problems and gallbladder disease, comprising administering ActRIIB, including variant ActRIIB polypeptides described herein, and homomultimers and heteromultimers comprising the same, to a patient in need thereof. The present invention provides, for example, the following items. (Item 1) an amino acid sequence beginning with any one of amino acids 20 to 29 (e.g., 20, 21, 22, 23, 24, 25, 26, 27, 28, or 29) of SEQ ID NO: 2 and ending with any one of amino acids 109 to 134 (e.g., 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, or 134) of SEQ ID NO: 2 and at least 75 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to SEQ ID NO: 2, wherein the variant ActRIIB polypeptide comprises one or more amino acid substitutions at positions selected from the group consisting of K55, F82, L79, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78, D80 and F82. (Item 2) 2. The variant ActRIIB polypeptide of item 1, comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 29 to 109 of SEQ ID NO:2. (Item 3) 2. The variant ActRIIB polypeptide of item 1, comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 25 to 131 of SEQ ID NO:2. (Item 4) 2. The variant ActRIIB polypeptide of item 1, comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 20 to 134 of SEQ ID NO:2. (Item 5) 2. The variant ActRIIB polypeptide of item 1, comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 53. (Item 6) 6. A variant ActRIIB polypeptide according to any one of items 1 to 5, which is a fusion protein comprising an ActRIIB polypeptide domain and one or more heterologous domains. (Item 7) 7. The variant ActRIIB polypeptide of item 6, which is an ActRIIB-Fc fusion protein. (Item 8) The variant ActRIIB polypeptide of item 6 or 7, wherein the fusion protein further comprises a linker domain positioned between the ActRIIB polypeptide domain and the one or more heterologous domains or Fc domain. (Item 9) 9. The variant of item 8, wherein the linker domain is selected from TGGG, TGGGG, SGGGG, GGGGS, GGG, GGGG, SGGG and GGGGS. (Item 10) 9. The variant ActRIIB polypeptide of item 8, comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO:5. (Item 11) 9. The variant ActRIIB polypeptide of item 8, comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 12. (Item 12) 12. The variant ActRIIB polypeptide of any one of items 1 to 11, comprising one or more amino acid substitutions with respect to the amino acid sequence of SEQ ID NO: 2 selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79A, L79D, L79E, L79P, P129S, P130A, P130R, E37A, R40A, D54A, K55A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N, D80R, and F82A. (Item 13) 13. A variant ActRIIB polypeptide according to item 12, comprising an A substitution at the position corresponding to K55 of SEQ ID NO:2. (Item 14) 13. A variant ActRIIB polypeptide according to item 12, comprising an E substitution at the position corresponding to K55 of SEQ ID NO:2. (Item 15) 13. A variant ActRIIB polypeptide according to item 12, comprising an I substitution at a position corresponding to F82 of SEQ ID NO:2. (Item 16) 13. A variant ActRIIB polypeptide according to item 12, comprising a K substitution at a position corresponding to F82 in SEQ ID NO:2. (Item 17) 13. A variant ActRIIB polypeptide according to item 12, comprising an E substitution at a position corresponding to L79 of SEQ ID NO:2. (Item 18) 13. The variant ActRIIB polypeptide of item 12, comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 31. (Item 19) 13. The variant ActRIIB polypeptide of item 12, comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 33. (Item 20) 20. The variant ActRIIB polypeptide of item 18 or 19, comprising an A substitution at the position corresponding to K55 of SEQ ID NO:2. (Item 21) 13. The variant ActRIIB polypeptide of item 12, comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 34. (Item 22) 13. The variant ActRIIB polypeptide of item 12, comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 36. (Item 23) 23. The variant ActRIIB polypeptide of item 21 or 22, comprising an E substitution at the position corresponding to K55 of SEQ ID NO:2. (Item 24) 13. The variant ActRIIB polypeptide of item 12, comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 37. (Item 25) 13. The variant ActRIIB polypeptide of item 12, comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 39. (Item 26) 26. A variant ActRIIB polypeptide according to item 24 or 25, comprising an I substitution at a position corresponding to F82 of SEQ ID NO:2. (Item 27) 13. The variant ActRIIB polypeptide of item 12, comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 40. (Item 28) 13. The variant ActRIIB polypeptide of item 12, comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 42. (Item 29) 29. A variant ActRIIB polypeptide according to item 27 or 28, comprising a K substitution at the position corresponding to F82 of SEQ ID NO:2. (Item 30) 13. The variant ActRIIB polypeptide of item 12, comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 43. (Item 31) 13. The variant ActRIIB polypeptide of item 12, comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 45. (Item 32) 32. A variant ActRIIB polypeptide according to item 30 or 31, comprising an E substitution at a position corresponding to L79 of SEQ ID NO:2. (Item 33) 13. The variant ActRIIB polypeptide of item 12, comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 46. (Item 34) 13. The variant ActRIIB polypeptide of item 12, comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 48. (Item 35) 35. A variant ActRIIB polypeptide according to item 33 or 34, comprising an E substitution at a position corresponding to L79 of SEQ ID NO:2. (Item 36) 13. The variant ActRIIB polypeptide of item 12, comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 49. (Item 37) 13. The variant ActRIIB polypeptide of item 12, comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 50. (Item 38) 38. A variant ActRIIB polypeptide according to item 36 or 37, comprising an E substitution at a position corresponding to L79 of SEQ ID NO:2. (Item 39) 13. The variant ActRIIB polypeptide of item 12, comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 51. (Item 40) 13. The variant ActRIIB polypeptide of item 12, comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 52. (Item 41) 41. A variant ActRIIB polypeptide according to item 39 or 40, comprising an E substitution at a position corresponding to L79 of SEQ ID NO:2. (Item 42) 10. The variant ActRIIB polypeptide of item 8 or 9, wherein the Fc domain comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of any one of SEQ ID NOs: 13 to 30. (Item 43) 43. The variant ActRIIB polypeptide of any one of items 1 to 42, which is a homodimeric protein. (Item 44) A heteromultimeric protein comprising a first ActRIIB polypeptide and a second ActRIIB polypeptide, wherein the first ActRIIB polypeptide starts with any one of amino acids 20 to 29 (e.g., 20, 21, 22, 23, 24, 25, 26, 27, 28, or 29) of SEQ ID NO: 2 and includes any one of amino acids 109 to 134 (e.g., 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 18 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133 or 134), and A24, K74, R64, P129, P130, E37, R40, D54, R56, W78, D80 and F82. and wherein the second ActRIIB polypeptide comprises one or more amino acid substitutions at positions of SEQ ID NO:2 selected from the group consisting of: A heteromultimeric protein comprising an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to an amino acid sequence ending in any one of the following: 124, 125, 126, 127, 128, 129, 130, 131, 132, 133 or 134), and wherein the first ActRIIB polypeptide comprises a different amino acid sequence compared to the second ActRIIB polypeptide. (Item 45) 45. The heteromultimer of item 44, wherein the first ActRIIB polypeptide and / or the second ActRIIB polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 29 to 109 of SEQ ID NO: 2. (Item 46) 45. The heteromultimer of item 44, wherein the first ActRIIB polypeptide and / or the second ActRIIB polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 25 to 131 of SEQ ID NO: 2. (Item 47) 45. The heteromultimer of item 44, wherein the first ActRIIB polypeptide and / or the second ActRIIB polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 20 to 134 of SEQ ID NO: 2. (Item 48) The heteromultimer described in Item 44, wherein the first ActRIIB polypeptide and / or the second ActRIIB polypeptide comprises an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 53. (Item 49) 49. The heteromultimer of any one of items 44 to 48, wherein the first ActRIIB polypeptide comprises one or more amino acid substitutions with respect to the amino acid sequence of SEQ ID NO: 2 selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79A, L79D, L79E, L79P, P129S, P130A, P130R, E37A, R40A, D54A, K55A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N, D80R and F82A. (Item 50) 50. The heteromultimer of any one of items 44 to 49, wherein the second ActRIIB polypeptide comprises one or more amino acid substitutions at positions of SEQ ID NO: 2 selected from the group consisting of A24, K74, R64, P129, P130, E37, R40, D54, R56, W78, D80 and F82. (Item 51) 50. The heteromultimer of item 49, wherein the second ActRIIB polypeptide comprises one or more amino acid substitutions with respect to the amino acid sequence of SEQ ID NO: 2 selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79A, L79D, L79E, L79P, P129S, P130A, P130R, E37A, R40A, D54A, K55A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N, D80R and F82A. (Item 52) A heteromultimer described in any one of items 44 to 51, wherein the first ActRIIB polypeptide and / or the second ActRIIB polypeptide is a fusion protein comprising an ActRIIB polypeptide domain and one or more heterologous domains. (Item 53) The heteromultimer described in Item 52, wherein the first ActRIIB polypeptide and / or the second ActRIIB polypeptide is an ActRIIB-Fc fusion protein. (Item 54) 54. The heteromultimer of item 52 or 53, wherein the first ActRIIB-Fc fusion protein and / or the second ActRIIB-Fc fusion protein further comprises a linker domain positioned between the ActRIIB polypeptide domain and the one or more heterologous domains or Fc domain. (Item 55) 55. The heteromultimer of item 54, wherein the linker domain is selected from TGGG, TGGGG, SGGGG, GGGGS, GGG, GGGG, SGGG and GGGGS. (Item 56) a. the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 13, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 13; b. the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 14, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 14; c. the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 15, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 15; d. the first ActRIIB-Fc fusion protein comprises an Fc domain at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 16, and the second ActRIIB-Fc fusion protein comprises an Fc domain at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 16; and e. the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 17, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 17. 56. The heteromultimer of any one of items 53 to 55, comprising an Fc domain selected from: (Item 57) a. the first ActRIIB-Fc fusion protein comprises an Fc domain at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 18, and the second ActRIIB-Fc fusion protein comprises an Fc domain at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 19; b. the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 19, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 18. 56. The heteromultimer of any one of items 53 to 55, comprising an Fc domain selected from: (Item 58) a. the first ActRIIB-Fc fusion protein comprises an Fc domain at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 20, and the second ActRIIB-Fc fusion protein comprises an Fc domain at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 21; b. the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 21, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 20. 56. The heteromultimer of any one of items 53 to 55, comprising an Fc domain selected from: (Item 59) a. the first ActRIIB-Fc fusion protein comprises an Fc domain at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 22, and the second ActRIIB-Fc fusion protein comprises an Fc domain at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 23; b. the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 23, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 22. 56. The heteromultimer of any one of items 53 to 55, comprising an Fc domain selected from: (Item 60) a. the first ActRIIB-Fc fusion protein comprises an Fc domain at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 22, and the second ActRIIB-Fc fusion protein comprises an Fc domain at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 25; b. the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 25, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 24. 56. The heteromultimer of any one of items 53 to 55, comprising an Fc domain selected from: (Item 61) a. the first ActRIIB-Fc fusion protein comprises an Fc domain at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 26, and the second ActRIIB-Fc fusion protein comprises an Fc domain at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 27; b. the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO:27, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO:26. 56. The heteromultimer of any one of items 53 to 55, comprising an Fc domain selected from: (Item 62) 56. The heteromultimer of any one of items 53 to 55, wherein the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 28, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 29. (Item 63) 63. The heteromultimer of claim 62, wherein the first ActRIIB-Fc fusion protein Fc domain comprises a cysteine ​​at amino acid position 132, a glutamic acid at amino acid position 138, a tryptophan at amino acid position 144, and an aspartic acid at amino acid position 217, and the second ActRIIB-Fc fusion protein Fc domain comprises a cysteine ​​at amino acid position 127, a serine at amino acid position 144, an alanine at position 146, an arginine at amino acid position 162, an arginine at amino acid position 179, and a valine at amino acid position 185. (Item 64) 56. The heteromultimer of any one of items 53 to 55, wherein the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 28, and the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 29. (Item 65) 63. The heteromultimer of item 62, wherein the second ActRIIB-Fc fusion protein Fc domain comprises a cysteine ​​at amino acid position 132, a glutamic acid at amino acid position 138, a tryptophan at amino acid position 144, and an aspartic acid at amino acid position 217, and the first ActRIIB-Fc fusion protein Fc domain comprises a cysteine ​​at amino acid position 127, a serine at amino acid position 144, an alanine at position 146, an arginine at amino acid position 162, an arginine at amino acid position 179, and a valine at amino acid position 185. (Item 66) 56. The heteromultimer of any one of items 53 to 55, wherein the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 30, and the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 23. (Item 67) 67. The heteromultimer of item 66, wherein the first ActRIIB-Fc fusion protein Fc domain comprises a cysteine ​​at amino acid position 132, a tryptophan at amino acid position 144, and an arginine at amino acid position 435, and the second ActRIIB-Fc fusion protein Fc domain comprises a cysteine ​​at amino acid position 127, a serine at amino acid position 144, an alanine at amino acid position 146, and a valine at amino acid position 185. (Item 68) 56. The heteromultimer of any one of items 53 to 55, wherein the second ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 30, and the first ActRIIB-Fc fusion protein comprises an Fc domain that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 23. (Item 69) 67. The heteromultimer described in Item 66, wherein the second ActRIIB-Fc fusion protein Fc domain comprises a cysteine ​​at amino acid position 132, a tryptophan at amino acid position 144, and an arginine at amino acid position 435, and the first ActRIIB-Fc fusion protein Fc domain comprises a cysteine ​​at amino acid position 127, a serine at amino acid position 144, an alanine at amino acid position 146, and a valine at amino acid position 185. (Item 70) 70. The heteromultimer of any one of items 44 to 69, which is a heterodimer. (Item 71) A heteromultimer comprising a first ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 33, and a second ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 5, wherein the first ActRIIB polypeptide does not contain the amino acid sequence of the second ActRIIB polypeptide. (Item 72) 72. The heteromultimer of item 71, wherein the first ActRIIB polypeptide comprises an alanine at the amino acid position corresponding to 55 of SEQ ID NO:2. (Item 73) 73. The heteromultimer of claim 71 or 72, wherein the second ActRIIB polypeptide does not contain an alanine at the amino acid position corresponding to 55 of SEQ ID NO: 2. (Item 74) 74. The heteromultimer of item 73, wherein the second ActRIIB polypeptide comprises a lysine at the amino acid position corresponding to 55 of SEQ ID NO:2. (Item 75) A heteromultimer described in any one of items 71 to 74, wherein the first ActRIIB polypeptide comprises one or more amino acid substitutions at amino acid positions corresponding to any one of F82, L79, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78, D80 and F82 of SEQ ID NO: 2. (Item 76) 76. The heteromultimer of item 75, wherein the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79A, L79D, L79E, L79P, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N, D80R and F82A. (Item 77) 77. A heteromultimer described in any one of items 71 to 76, wherein the second ActRIIB polypeptide comprises one or more amino acid substitutions at amino acid positions corresponding to any one of F82, L79, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78, D80 and F82 of SEQ ID NO: 2. (Item 78) 78. The heteromultimer of item 77, wherein the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79A, L79D, L79E, L79P, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N, D80R and F82A. (Item 79) A heteromultimer described in any one of items 71 to 78, wherein the first ActRIIB polypeptide and / or the second ActRIIB polypeptide contains one or more amino acid modifications that promote heteromultimer formation. (Item 80) A heteromultimer described in any one of items 71 to 79, wherein the first ActRIIB polypeptide and / or the second ActRIIB polypeptide contains one or more amino acid modifications that inhibit heteromultimer formation. (Item 81) 81. The heteromultimer of any one of items 71 to 80, which is a heterodimer. (Item 82) A heteromultimer comprising a first ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 36, and a second ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 5, wherein the first ActRIIB polypeptide does not contain the amino acid sequence of the second ActRIIB polypeptide. (Item 83) 83. The heteromultimer of item 82, wherein the first ActRIIB polypeptide comprises a glutamic acid at the amino acid position corresponding to 55 of SEQ ID NO:2. (Item 84) 84. The heteromultimer of item 82 or 83, wherein the second ActRIIB polypeptide does not contain a glutamic acid at the amino acid position corresponding to 55 of SEQ ID NO: 2. (Item 85) 85. The heteromultimer of item 84, wherein the second ActRIIB polypeptide comprises a lysine at the amino acid position corresponding to 55 of SEQ ID NO:2. (Item 86) A heteromultimer described in any one of items 82 to 85, wherein the first ActRIIB polypeptide comprises one or more amino acid substitutions at amino acid positions corresponding to any one of F82, L79, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78, D80 and F82 of SEQ ID NO: 2. (Item 87) 87. The heteromultimer of item 86, wherein the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79A, L79D, L79E, L79P, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N, D80R and F82A. (Item 88) 88. A heteromultimer described in any one of items 82 to 87, wherein the second ActRIIB polypeptide comprises one or more amino acid substitutions at amino acid positions corresponding to any one of F82, L79, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78, D80 and F82 of SEQ ID NO: 2. (Item 89) 89. The heteromultimer of item 88, wherein the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79A, L79D, L79E, L79P, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N, D80R and F82A. (Item 90) A heteromultimer described in any one of items 82 to 89, wherein the first ActRIIB polypeptide and / or the second ActRIIB polypeptide contains one or more amino acid modifications that promote heteromultimer formation. (Item 91) A heteromultimer described in any one of items 82 to 90, wherein the first ActRIIB polypeptide and / or the second ActRIIB polypeptide contains one or more amino acid modifications that inhibit heteromultimer formation. (Item 92) 92. The heteromultimer of any one of items 82 to 91, which is a heterodimer. (Item 93) A heteromultimer comprising a first ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 39, and a second ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 5, wherein the first ActRIIB polypeptide does not contain the amino acid sequence of the second ActRIIB polypeptide. (Item 94) 94. The heteromultimer of item 93, wherein the first ActRIIB polypeptide comprises an isoleucine at the amino acid position corresponding to 82 of SEQ ID NO: 2. (Item 95) 95. The heteromultimer of claim 93 or 94, wherein the second ActRIIB polypeptide does not contain an isoleucine at the amino acid position corresponding to 82 of SEQ ID NO: 2. (Item 96) 96. The heteromultimer of item 95, wherein the second ActRIIB polypeptide comprises a phenylalanine at the amino acid position corresponding to 82 of SEQ ID NO: 2. (Item 97) 97. A heteromultimer described in any one of items 93 to 96, wherein the first ActRIIB polypeptide comprises one or more amino acid substitutions at amino acid positions corresponding to any one of L79, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78 and D80 of SEQ ID NO: 2. (Item 98) 98. The heteromultimer of item 97, wherein the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79A, L79D, L79E, L79P, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N and D80R. (Item 99) 99. A heteromultimer described in any one of items 93 to 98, wherein the second ActRIIB polypeptide comprises one or more amino acid substitutions at amino acid positions corresponding to any one of L79, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78 and D80 of SEQ ID NO: 2. (Item 100) 100. The heteromultimer of item 99, wherein the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79A, L79D, L79E, L79P, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N and D80R. (Item 101) A heteromultimer described in any one of items 93 to 100, wherein the first ActRIIB polypeptide and / or the second ActRIIB polypeptide contains one or more amino acid modifications that promote heteromultimer formation. (Item 102) A heteromultimer described in any one of items 93 to 101, wherein the first ActRIIB polypeptide and / or the second ActRIIB polypeptide contain one or more amino acid modifications that inhibit heteromultimer formation. (Item 103) 103. The heteromultimer of any one of items 93 to 102, which is a heterodimer. (Item 104) A heteromultimer comprising a first ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 42, and a second ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 5, wherein the first ActRIIB polypeptide does not contain the amino acid sequence of the second ActRIIB polypeptide. (Item 105) 105. The heteromultimer of claim 104, wherein the first ActRIIB polypeptide comprises a lysine at the amino acid position corresponding to 82 of SEQ ID NO: 2. (Item 106) 105. The heteromultimer of claim 103 or 104, wherein the second ActRIIB polypeptide does not contain a lysine at the amino acid position corresponding to 82 of SEQ ID NO: 2. (Item 107) 107. The heteromultimer of claim 106, wherein the second ActRIIB polypeptide comprises a phenylalanine at the amino acid position corresponding to 82 of SEQ ID NO: 2. (Item 108) A heteromultimer described in any one of items 104 to 107, wherein the first ActRIIB polypeptide comprises one or more amino acid substitutions at amino acid positions corresponding to any one of L79, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78 and D80 of SEQ ID NO: 2. (Item 109) 109. The heteromultimer of item 108, wherein the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79A, L79D, L79E, L79P, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N and D80R. (Item 110) A heteromultimer described in any one of items 104 to 109, wherein the second ActRIIB polypeptide comprises one or more amino acid substitutions at amino acid positions corresponding to any one of L79, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78 and D80 of SEQ ID NO: 2. (Item 111) 111. The heteromultimer of item 110, wherein the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79A, L79D, L79E, L79P, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N and D80R. (Item 112) A heteromultimer described in any one of items 104 to 111, wherein the first ActRIIB polypeptide and / or the second ActRIIB polypeptide contains one or more amino acid modifications that promote heteromultimer formation. (Item 113) A heteromultimer described in any one of items 104 to 112, wherein the first ActRIIB polypeptide and / or the second ActRIIB polypeptide contains one or more amino acid modifications that inhibit heteromultimer formation. (Item 114) 114. The heteromultimer of any one of items 104 to 113, which is a heterodimer. (Item 115) A heteromultimer comprising a first ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 45, and a second ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 48, wherein the first ActRIIB polypeptide does not contain the amino acid sequence of the second ActRIIB polypeptide. (Item 116) 116. The heteromultimer of item 115, wherein the first ActRIIB polypeptide comprises an acidic amino acid position corresponding to 79 of SEQ ID NO:2. (Item 117) 117. The heteromultimer of claim 116, wherein the acidic amino acid is aspartic acid. (Item 118) 117. The heteromultimer of item 116, wherein the acidic amino acid is glutamic acid. (Item 119) A heteromultimer described in any one of items 115 to 117, wherein the second ActRIIB polypeptide does not contain an acidic amino acid (e.g., aspartic acid or glutamic acid) at the amino acid position corresponding to 79 of SEQ ID NO: 2. (Item 120) 120. The heteromultimer of claim 119, wherein the second ActRIIB polypeptide comprises a leucine at an amino acid position corresponding to 79 of SEQ ID NO: 2. (Item 121) A heteromultimer described in any one of items 115 to 120, wherein the first ActRIIB polypeptide comprises one or more amino acid substitutions at amino acid positions corresponding to any one of F82, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78, D80 and F82 of SEQ ID NO: 2. (Item 122) 122. The heteromultimer of item 121, wherein the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79P, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N, D80R and F82A. (Item 123) A heteromultimer described in any one of items 115 to 122, wherein the second ActRIIB polypeptide comprises one or more amino acid substitutions at amino acid positions corresponding to any one of F82, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78, D80 and F82 of SEQ ID NO: 2. (Item 124) 124. The heteromultimer of item 123, wherein the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N, D80R and F82A. (Item 125) A heteromultimer described in any one of items 115 to 124, wherein the first ActRIIB polypeptide and / or the second ActRIIB polypeptide contains one or more amino acid modifications that promote heteromultimer formation. (Item 126) A heteromultimer described in any one of items 115 to 125, wherein the first ActRIIB polypeptide and / or the second ActRIIB polypeptide contains one or more amino acid modifications that inhibit heteromultimer formation. (Item 127) 127. The heteromultimer of any one of items 115 to 126, which is a heterodimer. (Item 128) A heteromultimer comprising a first ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 50, and a second ActRIIB polypeptide that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 52, wherein the first ActRIIB polypeptide does not contain the amino acid sequence of the second ActRIIB polypeptide. (Item 129) 129. The heteromultimer of claim 128, wherein the first ActRIIB polypeptide comprises an acidic amino acid position corresponding to 79 of SEQ ID NO: 2. (Item 130) 130. The heteromultimer of claim 129, wherein the acidic amino acid is aspartic acid. (Item 131) 130. The heteromultimer of item 129, wherein the acidic amino acid is glutamic acid. (Item 132) A heteromultimer described in any one of items 128 to 131, wherein the second ActRIIB polypeptide does not contain an acidic amino acid (e.g., aspartic acid or glutamic acid) at the amino acid position corresponding to 79 of SEQ ID NO: 2. (Item 133) 133. The heteromultimer of claim 132, wherein the second ActRIIB polypeptide comprises a leucine at an amino acid position corresponding to 79 of SEQ ID NO: 2. (Item 134) A heteromultimer described in any one of items 128 to 133, wherein the first ActRIIB polypeptide comprises one or more amino acid substitutions at amino acid positions corresponding to any one of F82, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78, D80 and F82 of SEQ ID NO: 2. (Item 135) 135. The heteromultimer of item 134, wherein the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, L79P, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N, D80R and F82A. (Item 136) A heteromultimer described in any one of items 128 to 135, wherein the second ActRIIB polypeptide comprises one or more amino acid substitutions at amino acid positions corresponding to any one of F82, A24, K74, R64, P129, P130, E37, R40, D54, R56, W78, D80 and F82 of SEQ ID NO: 2. (Item 137) 137. The heteromultimer of item 136, wherein the one or more amino acid substitutions are selected from the group consisting of A24N, K74A, R64K, R64N, K74A, P129S, P130A, P130R, E37A, R40A, D54A, R56A, K74F, K74I, K74Y, W78A, D80A, D80F, D80G, D80I, D80K, D80M, D80M, D80N, D80R and F82A. (Item 138) A heteromultimer described in any one of items 128 to 137, wherein the first ActRIIB polypeptide and / or the second ActRIIB polypeptide contains one or more amino acid modifications that promote heteromultimer formation. (Item 139) A heteromultimer described in any one of items 128 to 138, wherein the first ActRIIB polypeptide and / or the second ActRIIB polypeptide contains one or more amino acid modifications that inhibit heteromultimer formation. (Item 140) 139. The heteromultimer of any one of items 128 to 139, which is a heterodimer. (Item 141) An ActRIIB polypeptide described in any of the preceding items, comprising one or more amino acid modifications selected from the group consisting of glycosylated amino acids, PEGylated amino acids, farnesylated amino acids, acetylated amino acids, biotinylated amino acids, and amino acids conjugated to a lipid moiety. (Item 142) An ActRIIB polypeptide according to item 141, which is glycosylated and has a glycosylation pattern obtainable from said polypeptide in a CHO cell. (Item 143) A heteromultimer described in any of the preceding items, wherein the first ActRIIB polypeptide and / or the second ActRIIB polypeptide comprises one or more amino acid modifications selected from the group consisting of glycosylated amino acids, PEGylated amino acids, farnesylated amino acids, acetylated amino acids, biotinylated amino acids, and amino acids conjugated to a lipid moiety. (Item 144) The heteromultimer described in Item 143, wherein the first ActRIIB polypeptide and / or the second ActRIIB polypeptide is glycosylated and has a glycosylation pattern available from the polypeptide in a CHO cell. (Item 145) 2. The polypeptide or heteromultimer of any preceding item, which binds to one or more TGF-beta superfamily ligands. (Item 146) At least 1 x 10 -7 K of M D 146. The polypeptide or heteromultimer according to item 145, which binds to one or more TGF-beta superfamily ligands. (Item 147) 147. The polypeptide or heteromultimer according to item 145 or 146, wherein the one or more TGF-beta superfamily ligands are selected from the group consisting of BMP6, BMP7, BMP9, BMP10, GDF3, GDF7, GDF8, GDF11, GDF15, activin A, activin B, activin C, activin E, activin AB, activin AC, activin AE, activin BC and activin BE. (Item 148) A polypeptide or heteromultimer according to any preceding item, which inhibits one or more TGF-beta superfamily ligands. (Item 149) 149. The polypeptide or heteromultimer according to item 148, which inhibits signal transduction of one or more TGF-beta superfamily ligands. (Item 150) 150. The polypeptide or heteromultimer according to item 149, which inhibits Smad signaling of one or more TGF-beta superfamily ligands. (Item 151) 151. The polypeptide or heteromultimer of any one of items 148 to 150, which inhibits signaling of one or more TGF-beta superfamily ligands in a cell-based assay. (Item 152) 152. The polypeptide or heteromultimer of any one of items 148 to 151, which inhibits one or more TGF-beta superfamily ligands selected from the group consisting of BMP6, BMP7, BMP9, BMP10, GDF3, GDF7, GDF8, GDF11, GDF15, activin A, activin B, activin C, activin E, activin AB, activin AC, activin AE, activin BC and activin BE. (Item 153) A heteromultimeric protein comprising an ActRIIB polypeptide described in any one of items 1 to 42 and a second polypeptide or a functional fragment thereof selected from the group consisting of ALK1, ALK2, ALK3, ALK4, ALK5, ALK6, ALK7, ActRIIA, TGFBRII, BMPRII and MISRII polypeptides. (Item 154) 154. The heteromultimer of claim 153, wherein the second polypeptide is an ALK1 polypeptide or a functional fragment thereof. (Item 155) 155. The heteromultimer of item 154, wherein the ALK1 polypeptide or functional fragment thereof comprises an amino acid sequence at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to SEQ ID NO: 54 or a functional fragment thereof. (Item 156) 155. The heteromultimer of item 154, wherein the ALK1 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of any one of SEQ ID NOs: 54, 55, 56, 57, 58, 59, 60, 61, 62 and 63, or a functional fragment thereof. (Item 157) 154. The heteromultimer of item 153, wherein the second polypeptide is an ALK2 polypeptide or a functional fragment thereof. (Item 158) 158. The heteromultimer of item 157, wherein the ALK2 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 64 or 65, or a functional fragment thereof. (Item 159) 158. The heteromultimer of item 157, wherein the ALK2 polypeptide or functional fragment thereof comprises an amino acid sequence at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NOs: 64, 65, 66, 67, 68, 69, 70, 71, 72 and 73, or a functional fragment thereof. (Item 160) 154. The heteromultimer of item 153, wherein the second polypeptide is an ALK3 polypeptide or a functional fragment thereof. (Item 161) 161. The heteromultimer of item 160, wherein the ALK3 polypeptide or functional fragment thereof comprises an amino acid sequence at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 74 or a functional fragment thereof. (Item 162) 161. The heteromultimer of claim 160, wherein the ALK3 polypeptide or functional fragment thereof comprises an amino acid sequence at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 74, 75, 76, 77, 80 or 81, or a functional fragment thereof. (Item 163) 154. The heteromultimer of item 153, wherein the second polypeptide is an ALK4 polypeptide or a functional fragment thereof. (Item 164) 164. The heteromultimer of item 163, wherein the ALK4 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 84 or 85, or a functional fragment thereof. (Item 165) 164. The heteromultimer according to item 163, wherein the ALK4 polypeptide or functional fragment thereof comprises an amino acid sequence at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NOs: 84, 86, 85, 87, 88, 89, 90, 91, 92, 93, 94 and 95, or a functional fragment thereof. (Item 166) 154. The heteromultimer of item 153, wherein the second polypeptide is an ALK5 polypeptide or a functional fragment thereof. (Item 167) 167. The heteromultimer of item 166, wherein the ALK5 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 96 or 97, or a functional fragment thereof. (Item 168) 167. The heteromultimer according to item 166, wherein the ALK5 polypeptide or functional fragment thereof comprises an amino acid sequence at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NOs: 96, 98, 97, 99, 100, 101, 102, 103, 104, 105, 106 and 107, or a functional fragment thereof. (Item 169) 154. The heteromultimer of item 153, wherein the second polypeptide is an ALK6 polypeptide or a functional fragment thereof. (Item 170) 169. The heteromultimer according to Item 169, wherein the ALK6 polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 108 or 110, or a functional fragment thereof. (Item 171) 169. The heteromultimer according to item 169, wherein the ALK6 polypeptide or functional fragment thereof comprises an amino acid sequence at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NOs: 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118 and 119, or a functional fragment thereof. (Item 172) 154. The heteromultimer of item 153, wherein the second polypeptide is an ALK7 polypeptide or a functional fragment thereof. (Item 173) 173. The heteromultimer of item 172, wherein the ALK7 polypeptide or functional fragment thereof comprises an amino acid sequence at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 120, 121 or 122, or a functional fragment thereof. (Item 174) 173. The heteromultimer of item 172, wherein the ALK7 polypeptide or functional fragment thereof comprises an amino acid sequence at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NOs: 120, 123, 124, 125, 121, 126, 122, 127, 128, 129, 130, 131, 132, 133, 134, 135 and 136, or a functional fragment thereof. (Item 175) The heteromultimer described in Item 153, wherein the second polypeptide is an ActRIIA polypeptide or a functional fragment thereof. (Item 176) The heteromultimer described in Item 175, wherein the ActRIIA polypeptide or a functional fragment thereof comprises an amino acid sequence or a functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 137. (Item 177) 176. The heteromultimer of item 175, wherein the ActRIIA polypeptide or a functional fragment thereof comprises an amino acid sequence or a functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NOs: 137, 138, 139, 140, 141, 142, 143, 144, 145, 146 and 147. (Item 178) 154. The heteromultimer according to item 153, wherein the second polypeptide is a TGFBRII polypeptide or a functional fragment thereof. (Item 179) 179. The heteromultimer of item 178, wherein the TGFBRII polypeptide or functional fragment thereof comprises an amino acid sequence at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 204 or a functional fragment thereof. (Item 180) 179. The heteromultimer of claim 178, wherein the TGFBRII polypeptide or functional fragment thereof comprises an amino acid sequence at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NOs: 161, 162, 160, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178 and 179, or a functional fragment thereof. (Item 181) 154. The heteromultimer according to item 153, wherein the second polypeptide is a BMPRII polypeptide or a functional fragment thereof. (Item 182) 182. The heteromultimer of item 181, wherein the BMPRII polypeptide or functional fragment thereof comprises an amino acid sequence or functional fragment thereof that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 148 or 149, or a functional fragment thereof. (Item 183) 182. The heteromultimer according to item 181, wherein the BMPRII polypeptide or functional fragment thereof comprises an amino acid sequence at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NOs: 148, 150, 149, 151, 152, 153, 154, 155, 156, 157, 158 and 159, or a functional fragment thereof. (Item 184) 154. The heteromultimer according to item 153, wherein the second polypeptide is a MISRII polypeptide or a functional fragment thereof. (Item 185) 185. The heteromultimer according to item 184, wherein the MISRII polypeptide or functional fragment thereof comprises an amino acid sequence at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 180, 181 or 182, or a functional fragment thereof. (Item 186) 185. The heteromultimer according to item 184, wherein the MISRII polypeptide or functional fragment thereof comprises an amino acid sequence at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NOs: 180, 183, 181, 184, 182, 185, 186, 187, 188, 189, 190, 191, 192 and 193, or a functional fragment thereof. (Item 187) A pharmaceutical preparation comprising an ActRIIB polypeptide according to any of the preceding items and a pharmaceutically acceptable carrier. (Item 188) A pharmaceutical preparation comprising the heteromultimer of any of the preceding items and a pharmaceutically acceptable carrier. (Item 189) 189. The pharmaceutical preparation of item 188, comprising less than about 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2% or less than about 1% homomultimers. (Item 190) An isolated and / or recombinant nucleic acid comprising one or more coding sequences for the ActRIIB polypeptide(s) described in any of the preceding items. (Item 191) 191. The isolated and / or recombinant nucleic acid of Item 190, which is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to a nucleic acid sequence corresponding to any one of SEQ ID NOs: 3, 10, 31, 35, 38, 41, 44 or 47. (Item 192) 192. An isolated and / or recombinant polynucleotide sequence comprising a promoter sequence operably linked to the coding sequence of item 190 or 191. (Item 193) 193. A vector comprising the isolated and / or recombinant nucleic acid of any one of items 190 to 192. (Item 194) A cell comprising the isolated and / or recombinant polynucleotide sequence of any one of items 190 to 192, or the vector of item 193. (Item 195) A method for producing an ActRIIB polypeptide described in any of the preceding items, comprising culturing cells under conditions suitable for expression of the ActRIIB polypeptide, wherein the cells contain a nucleic acid or vector described in any one of items 190 to 192. (Item 196) Item 196. The method of item 195, further comprising recovering the expressed ActRIIB polypeptide. (Item 197) A method for producing a heteromultimer described in any of the preceding items, comprising a step of culturing cells under conditions suitable for expression of a first ActRIIB polypeptide and a second ActRIIB polypeptide, wherein the cells contain a nucleic acid or vector comprising a coding sequence for a first ActRIIB polypeptide described in any one of items 190 to 192, and a second nucleic acid or vector comprising a coding sequence for a second ActRIIB polypeptide described in any one of items 190 to 192. (Item 198) 198. The method of claim 197, further comprising recovering the expressed heteromultimeric polypeptide. (Item 199) A method for producing a heteromultimer according to any of the preceding items, comprising: a. culturing a first cell under conditions suitable for expression of a first ActRIIB polypeptide, wherein the cell contains a nucleic acid or vector comprising a coding sequence for the first ActRIIB polypeptide described in any one of items 190 to 192; b. recovering the first ActRIIB polypeptide; c. culturing a second cell under conditions suitable for expression of the first ActRIIB polypeptide, wherein the cell contains a nucleic acid or vector comprising a coding sequence for the first ActRIIB polypeptide described in any one of items 190 to 192; d. recovering the second ActRIIB polypeptide; e. combining the first and second ActRIIB polypeptides under conditions suitable for heteromultimer formation; A method comprising: (Item 200) 200. The method of claim 199, further comprising recovering the expressed heteromultimeric polypeptide. (Item 201) A method for increasing red blood cell levels or hemoglobin levels in a patient, comprising administering to a patient in need thereof an ActRIIB polypeptide or heteromultimer described in any of the preceding items. (Item 202) A method for treating anemia or an anemia-related disorder in a patient, comprising administering to a patient in need thereof an ActRIIB polypeptide or heteromultimer described in any of the preceding items. (Item 203) A method for increasing muscle mass and / or muscle strength in a patient, comprising administering to a patient in need thereof an ActRIIB polypeptide or heteromultimer described in any of the preceding items. (Item 204) A method for treating a muscle-related disorder in a patient, comprising administering to a patient in need thereof an ActRIIB polypeptide or heteromultimer described in any of the preceding items. (Item 205) 205. The method of claim 204, wherein the disorder is associated with undesirably low muscle growth and / or muscle wasting. [Brief explanation of the drawings]

[0065] [Figure 1-1]FIG. 1 shows a schematic example of a heteromeric protein complex comprising a first variant ActRIIB polypeptide (designated "X") and either a second variant ActRIIB polypeptide (designated "Y") or an unmodified ActRIIB polypeptide (designated "Y"). In the illustrated embodiment, the first variant ActRIIB polypeptide is part of a fusion polypeptide comprising a first member of an interacting pair ("C1"), and either the second variant ActRIIB polypeptide or an unmodified ActRIIB polypeptide is part of a fusion polypeptide comprising a second member of the interacting pair ("C2"). Suitable interacting pairs include, for example, heavy and / or light chain immunoglobulin interacting pairs, truncations, and variants thereof, such as those described herein (e.g., Spiess et al. (2015) Molecular Immunology 67(2A):95-106). In each fusion polypeptide, a linker may be disposed between the first variant ActRIIB polypeptide, the second variant ActRIIB polypeptide, or the unmodified ActRIIB polypeptide and the corresponding member of the interaction pair.The first and second members of the interaction pair may be uninduced, which means that the members of the pair may associate with each other or self-associate without substantial preference, and they may have the same or different amino acid sequences.See Figure 1A.Alternatively, the interaction pair may be an induced (asymmetric) pair, which means that the members of the pair preferentially associate with each other rather than self-associate.See Figure 1B. [Figure 1-2]FIG. 1 shows a schematic example of a heteromeric protein complex comprising a first variant ActRIIB polypeptide (designated "X") and either a second variant ActRIIB polypeptide (designated "Y") or an unmodified ActRIIB polypeptide (designated "Y"). In the illustrated embodiment, the first variant ActRIIB polypeptide is part of a fusion polypeptide comprising a first member of an interacting pair ("C1"), and either the second variant ActRIIB polypeptide or an unmodified ActRIIB polypeptide is part of a fusion polypeptide comprising a second member of the interacting pair ("C2"). Suitable interacting pairs include, for example, heavy and / or light chain immunoglobulin interacting pairs, truncations, and variants thereof, such as those described herein (e.g., Spiess et al. (2015) Molecular Immunology 67(2A):95-106). In each fusion polypeptide, a linker may be disposed between the first variant ActRIIB polypeptide, the second variant ActRIIB polypeptide, or the unmodified ActRIIB polypeptide and the corresponding member of the interaction pair.The first and second members of the interaction pair may be uninduced, which means that the members of the pair may associate with each other or self-associate without substantial preference, and they may have the same or different amino acid sequences.See Figure 1A.Alternatively, the interaction pair may be an induced (asymmetric) pair, which means that the members of the pair preferentially associate with each other rather than self-associate.See Figure 1B.

[0066] [Figure 2] Figure 2 shows an alignment of the extracellular domains of human ActRIIA and human ActRIIB, with residues (boxed) predicted herein to make direct contact with the ligand based on a synthetic analysis of multiple ActRIIB and ActRIIA crystal structures.

[0067] [Figure 3]FIG. 3 shows a multiple sequence alignment of various vertebrate ActRIIB precursor proteins without their intracellular domains, the human ActRIIA precursor protein without its intracellular domains, and the consensus ActRII precursor protein.

[0068] [Figure 4] Figure 4 shows a multiple sequence alignment of Fc domains from human IgG isotypes using Clustal 2.1. The hinge region is indicated by a dotted underline. Double underlining indicates examples of engineered positions in IgG1 (SEQ ID NO: 13) Fc to promote asymmetric chain pairing, as well as corresponding positions for other isotypes IgG4 (SEQ ID NO: 17), IgG2 (SEQ ID NO: 14), and IgG3 (SEQ ID NO: 15).

[0069] [Figure 5] Figure 5 shows the amino acid sequence of the human ActRIIB extracellular domain polypeptide (SEQ ID NO: 1), with numbering based on the native human ActRIIB precursor sequence (see SEQ ID NO: 2).

[0070] [Figure 6] Figure 6 shows the amino acid sequence of the human ActRIIB precursor protein (SEQ ID NO: 2; NCBI Reference Sequence NP_001097.2). The signal peptide is underlined, the extracellular domain is in bold (also referred to as SEQ ID NO: 1), and potential N-linked glycosylation sites are boxed.

[0071] [Figure 7] FIG. 7 shows the nucleic acid sequence encoding the human ActRIIB(20-134) extracellular domain polypeptide.

[0072] [Figure 8] 8 shows the nucleic acid sequence encoding the human ActRIIB precursor protein. SEQ ID NO: 4 consists of nucleotides 25 to 1560 of the NCBI reference sequence NM_001106.

[0073] [Figure 9] 9 shows the ligand binding kinetics of homodimeric Fc fusion proteins containing ActRIIB variant domains or unmodified ActRIIB domains as determined by surface plasmon resonance at 37° C. Amino acid numbering is based on SEQ ID NO:2.

[0074] [Figure 10] 10 shows the ligand binding kinetics of homodimeric Fc fusion proteins containing ActRIIB variant domains or unmodified ActRIIB domains as determined by surface plasmon resonance at 25° C. Amino acid numbering is based on SEQ ID NO: 2.

[0075] [Figure 11] FIG. 11 shows the change in body weight from baseline for wild-type mice treated with vehicle or homodimeric Fc fusion proteins containing an ActRIIB variant domain or an unmodified ActRIIB domain.

[0076] [Figure 12] FIG. 12 shows hemoglobin concentrations in wild-type mice treated with vehicle or homodimeric Fc fusion proteins containing an ActRIIB variant domain or an unmodified ActRIIB domain.

[0077] [Figure 13] Figure 13 shows red blood cell counts in cynomolgus monkeys treated with 9 mg / kg (sc) ActRIIB-Fc, ActRIIA-Fc, or ActRIIB(F82I)-Fc on days 1 and 15. ActRIIB(F82I)-Fc treatment increased RBC counts compared to ActRIIB-Fc (negative control) by an amount similar to that of ActRIIA-Fc (positive control). DETAILED DESCRIPTION OF THE INVENTION

[0078] 1. Overview In a specific embodiment, the present invention relates to an ActRIIB polypeptide. As used herein, the term "ActRIIB" refers to a family of activin receptor type IIB (ActRIIB) proteins and ActRIIB-related proteins from any species. All members of the ActRIIB family are generally transmembrane proteins composed of a ligand-binding extracellular domain with a cysteine-rich region, a transmembrane domain, and a cytoplasmic domain with predicted serine / threonine kinase specificity. The amino acid sequence of human ActRIIB precursor protein is shown in Figure 6 (SEQ ID NO: 2).

[0079] The term "ActRIIB polypeptide" refers to a polypeptide including any naturally occurring polypeptide of an ActRIIB family member, and any variant thereof (including mutants, fragments, fusions, and peptidomimetics) that retains useful activity. For example, ActRIIB polypeptide includes a polypeptide derived from any known ActRIIB sequence that has at least about 80% identity with the sequence of an ActRIIB polypeptide, preferably at least 85%, 90%, 95%, 97%, 99% or more identity.

[0080] In a specific embodiment, the present invention relates to a soluble ActRIIB polypeptide. As described herein, the term "soluble ActRIIB polypeptide" generally refers to a polypeptide comprising the extracellular domain of an ActRIIB protein. As used herein, the term "soluble ActRIIB polypeptide" includes any naturally occurring extracellular domain of an ActRIIB protein and any variant thereof (including mutants, fragments, and peptidomimetic forms) that retains useful activity. For example, the extracellular domain of an ActRIIB protein binds to a ligand and is generally soluble. Examples of soluble ActRIIB polypeptides include the ActRIIB extracellular domain shown in FIG. 5 (SEQ ID NO: 1) and SEQ ID NO: 53. Other examples of soluble ActRIIB polypeptides include a signal sequence in addition to the extracellular domain of an ActRIIB protein (see Example 1). The signal sequence can be the native signal sequence of ActRIIB or a signal sequence from another protein, such as the tissue plasminogen activator (TPA) signal sequence or the honeybee melatin signal sequence.

[0081] TGF-β signaling is mediated by a heteromeric complex of type I and type II serine / threonine kinase receptors, which phosphorylate and activate downstream Smad proteins after ligand stimulation (Massague, 2000, Nat. Rev. Mol. Cell Biol. 1:169-178). These type I and type II receptors are all transmembrane proteins composed of a ligand-binding extracellular domain with a cysteine-rich region, a transmembrane domain, and a cytoplasmic domain with predicted serine / threonine specificity. Type I receptors are essential for signal transduction, while type II receptors are required for ligand binding. Type I and type II activin receptors form a stable complex after ligand binding, resulting in phosphorylation of the type I receptor by the type II receptor.

[0082] Two related type II receptors, ActRIIA and ActRIIB, have been identified as type II receptors for activin (Mathews and Vale, 1991, Cell 65:973-982; Attisano et al., 1992, Cell 68:97-108). In addition to activin, ActRIIA and ActRIIB can biochemically interact with several other TGF-β family proteins, including BMP7, Nodal, GDF8, and GDF11 (Yamashita et al., 1995, J. Cell Biol. 130:217-226; Lee and McPherron, 2001, Proc. Natl. Acad. Sci. 98:9306-9311; Yeo and Whitman, 2001, Mol. Cell 7:949-957; Oh et al., 2002, Genes Dev. 16:2749-54). The applicants found that soluble ActRIIA-Fc fusion protein and ActRIIB-Fc fusion protein have substantially different effects in vivo, with ActRIIA-Fc having a primary effect on bone and ActRIIB-Fc having a primary effect on skeletal muscle.

[0083] In certain embodiments, the present invention relates to the antagonism of the ligand of ActRIIB receptor (also referred to as ActRIIB ligand) by the subject ActRIIB polypeptide (e.g., soluble ActRIIB polypeptide).Thus, the compositions and methods of the present invention are useful for treating disorders related to the abnormal activity of one or more ligands of ActRIIB receptor.Exemplary ligands of ActRIIB receptor include some TGF-β family members, such as activin, nodal, GDF8, GDF11 and BMP7.

[0084] Activins are dimeric polypeptide growth factors that belong to the TGF-beta superfamily. They are homo- / heterodimers of two closely related beta subunits (β A β A , β B β B and β A β BThere are three activins (A, B, and AB), which are the TGF-beta superfamily. Within the TGF-beta superfamily, activins are unique and multifunctional factors that can stimulate hormone production in ovarian and placental cells, support neuronal survival, positively or negatively influence cell cycle progression depending on the cell type, and induce mesodermal differentiation in at least amphibian embryos (DePaolo et al., 1991, Proc SocEp Biol Med. 198:500-512; Dyson et al., 1997, Curr Biol. 7:81-84; Woodruff, 1998, Biochem Pharmacol. 55:953-963). Furthermore, erythroid differentiation factor (EDF), isolated from stimulated human monocytic leukemic cells, was found to be identical to activin A (Murata et al., 1988, PNAS 85:2434). It has been suggested that activin A acts as a natural regulator of erythropoiesis in bone marrow. In some tissues, activin signaling is antagonized by its related heterodimer, inhibin. For example, upon release of follicle-stimulating hormone (FSH) from the pituitary gland, activin promotes FSH secretion and synthesis, while inhibin prevents FSH secretion and synthesis. Other proteins that can regulate activin bioactivity and / or bind to activin include follistatin (FS), follistatin-related protein (FSRP), α2-macroglobulin, cerberus, and endoglin.

[0085] Nodal proteins function in mesoderm and endoderm induction and formation, as well as the subsequent organization of axial structures such as the heart and stomach during early embryogenesis. It has been demonstrated that dorsal tissue in the developing vertebrate embryo primarily contributes to the axial structure of the notochord and prechordal plate, while recruiting surrounding cells to form non-axial embryonic structures. Nodal appears to signal through both type I and type II receptors and intracellular effectors known as Smad proteins. Recent studies support the idea that ActRIIA and ActRIIB function as type II receptors for Nodal (Sakuma et al., Genes Cells. 2002, 7:401-12). It has been suggested that Nodal ligands interact with their cofactors (e.g., cripto) to activate activin type I and type II receptors, which phosphorylate Smad2. Nodal proteins have been implicated in many events crucial to the early vertebrate embryo, including mesoderm formation, anterior patterning, and left-right axis specification. Experimental evidence demonstrated that Nodal signaling activates pAR3-Lux, a luciferase reporter previously shown to respond specifically to activin and TGF-beta. However, Nodal fails to induce pTlx2-Lux, a reporter specifically responsive to bone morphogenetic proteins. Recent results provide direct biochemical evidence that Nodal signaling is mediated by both activin-TGF-beta pathway Smads, Smad2 and Smad3. Further evidence indicates that extracellular Crypto proteins are required for Nodal signaling, making it distinct from activin or TGF-beta signaling.

[0086] Growth differentiation factor-8 (GDF8) is also known as myostatin. GDF8 is a negative regulator of skeletal muscle mass. GDF8 is highly expressed in developing and adult skeletal muscle. GDF8 null mutations in transgenic mice are characterized by significant skeletal muscle hypertrophy and hyperplasia (McPherron et al., Nature, 1997, 387:83-90). Similar increases in skeletal muscle mass are evident in naturally occurring mutations of GDF8 in cattle (Ashmore et al., 1974, Growth, 38:501-507; Swatland and Kieffer, J. Anim. Sci., 1994, 38:752-757; McPherron and Lee, Proc. Natl. Acad. Sci. USA, 1997, 94:12457-12461; and Kambadur et al., Genome Res., 1997, 7:910-915) and, strikingly, in humans (Schuelke et al., N Engl J Med 2004; 350:2682-8). Studies have also shown that muscle wasting associated with HIV infection in humans is accompanied by increased GDF8 protein expression (Gonzalez-Cadavid et al., PNAS, 1998, 95:14938-43). In addition, GDF8 can modulate the production of muscle-specific enzymes (e.g., creatine kinase) and modulate myoblast cell proliferation (WO00 / 43781). The GDF8 propeptide can noncovalently bind to the mature GDF8 domain dimer and inactivate its biological activity (Miyazono et al. (1988) J. Biol. Chem., 263:6407-6415; Wakefield et al. (1988) J. Biol. Chem., 263:7646-7654; and Brown et al. (1990) Growth Factors, 3:35-43). Other proteins that bind to and inhibit the biological activity of GDF8 or structurally related proteins include follistatin and potentially follistatin-related proteins (Gamer et al. (1999) Dev. Biol. 208:222-232).

[0087] Growth differentiation factor-11 (GDF11), also known as BMP11, is a secreted protein (McPherron et al., 1999, Nat. Genet. 22:260-264). GDF11 is expressed in the tail bud, limb bud, maxillary and mandibular arches, and dorsal root ganglia during mouse development (Nakashima et al., 1999, Mech. Dev. 80:185-189). GDF11 plays a unique role in the patterning of both mesodermal and neural tissues (Gamer et al., 1999, Dev. Biol. 208:222-32). GDF11 has been shown to be a negative regulator of chondrogenesis and myogenesis in the developing chick limb (Gamer et al., 2001, Dev. Biol. 229:407-20). The expression of GDF11 in muscle also suggests its role in regulating muscle growth in a similar manner to GDF8.In addition, the expression of GDF11 in the brain suggests that GDF11 can also have activity related to the function of the nervous system.Interestingly, GDF11 has been found to inhibit neurogenesis in olfactory epithelium (Wu et al., 2003, Neuron. 37:197-207).Therefore, GDF11 can have in vitro and in vivo applications in the treatment of diseases, such as muscle diseases and neurodegenerative diseases (e.g., amyotrophic lateral sclerosis).

[0088] Bone morphogenetic protein 7 (BMP7), also known as osteogenic protein-1 (OP-1), is well known to induce cartilage and bone formation. In addition, BMP7 regulates a wide variety of physiological processes. For example, BMP7 can be an osteoinductive factor responsible for the phenomenon of epithelial bone formation. BMP7 has also been shown to play a role in calcium regulation and bone homeostasis. Like activin, BMP7 binds to type II receptors, ActRIIA and IIB. However, BMP7 and activin recruit distinct type I receptors into heteromeric receptor complexes. While the predominant BMP7 type I receptor observed was ALK2, activin exclusively bound to ALK4 (ActRIIB). BMP7 and activin elicit distinct biological responses and activated distinct Smad pathways (Macias-Silva et al., 1998, J Biol Chem. 273:25628-36).

[0089] In certain aspects, the present invention relates to the use of certain ActRIIB polypeptides (e.g., soluble ActRIIB polypeptides) for generally antagonizing the signal transduction of ActRIIB ligands in any process related to ActRIIB activity.Optionally, the ActRIIB polypeptides of the present invention can antagonize one or more ligands of ActRIIB receptors, such as activin, nodal, GDF8 and GDF11, and therefore can be useful in the treatment of additional disorders.

[0090] Therefore, the present invention contemplates the use of ActRIIB polypeptide in the treatment or prevention of diseases or conditions associated with the abnormal activity of ActRIIB or ActRIIB ligand.ActRIIB or ActRIIB ligand is involved in the control of many critical biological processes.Due to their important functions in such processes, ActRIIB or ActRIIB ligand can be a desirable target for therapeutic intervention.For example, ActRIIB polypeptide (e.g., soluble ActRIIB polypeptide) can be used to treat human or animal disorders or conditions. Examples of such disorders or conditions include, but are not limited to, metabolic disorders such as type 2 diabetes, impaired glucose tolerance, metabolic syndrome (e.g., syndrome X), and insulin resistance induced by trauma (e.g., burns or nitrogen imbalance); adipose tissue disorders (e.g., obesity); muscle and neuromuscular disorders such as muscular dystrophies (including Duchenne muscular dystrophy); amyotrophic lateral sclerosis (ALS); muscle atrophy; organ atrophy; frailty; carpal tunnel syndrome; congestive obstructive pulmonary disease; and sarcopenia, cachexia, and other muscle wasting syndromes. Other examples include osteoporosis, particularly in elderly and / or postmenopausal women; glucocorticoid-induced osteoporosis; osteopenia; osteoarthritis; and osteoporosis-related fractures. Further examples include low bone mass due to chronic glucocorticoid therapy, premature hypogonadism, androgen suppression, vitamin D deficiency, secondary hyperparathyroidism, nutritional deficiencies, and anorexia nervosa. These disorders and conditions are described below under "Exemplary Therapeutic Uses."

[0091] The terms used herein generally have their ordinary meaning in the art, within the context of this invention and in the specific context in which each term is used. Certain terms are described below or elsewhere in this specification to provide the practitioner with additional guidance regarding the description of the compositions and methods of the invention and how to make and use them. The scope or meaning of any use of a term will be apparent from the specific context in which it is used.

[0092] "About" and "approximately" are generally intended to mean an acceptable degree of error for the measured content, given the nature or precision of the measurement. Typically, exemplary degrees of error are within 20 percent (%), preferably within 10%, and more preferably within 5% of a given value or range of values.

[0093] Alternatively, and particularly in biological systems, the terms "about" and "approximately" can mean values ​​that are within an order of magnitude, preferably within 5-fold, and more preferably within 2-fold of a given value. Numerical values ​​given herein are approximate unless otherwise noted, meaning that the term "about" or "approximately" can be inferred when not expressly stated.

[0094] The methods of the present invention can include comparing sequences to one another, including unmodified (wild-type) sequences to one or more mutants (sequence variants). Such comparisons typically involve aligning polymer sequences, for example, using sequence alignment programs and / or algorithms well known in the art (e.g., BLAST, FASTA, and MEGALIGN, to name a few). Those skilled in the art can readily recognize that in such alignments where the mutations contain residue insertions or deletions, the sequence alignment will introduce "gaps" (typically represented by a dash or "A") in the polymer sequence that do not contain the inserted or deleted residues.

[0095] "Homologous," in all its grammatical forms and spelling variations, refers to the relationship between two proteins that share a "common evolutionary origin," including proteins from a superfamily in the same species of organism and homologous proteins from different species of organisms. Such proteins (and their encoding nucleic acids) have sequence homology, as reflected by their sequence similarity, whether in terms of percent identity or by the presence of specific residues or motifs and conserved positions.

[0096] The term "sequence similarity," in all its grammatical forms, refers to the degree of identity or correspondence between nucleic acid or amino acid sequences that may or may not share a common evolutionary origin.

[0097] However, in common usage and in this application, the term "homologous," when modified by an adverb such as "highly," can refer to sequence similarity, which may or may not relate to a common evolutionary origin.

[0098] "Turn on," in all its grammatical forms, refers to the process of activating a protein and / or gene (e.g., by activating or amplifying gene expression of that protein or by inducing an inactive protein to proceed to an active state) or increasing the activity of a protein and / or gene.

[0099] "Antagonize," in all its grammatical forms, refers to the process of inhibiting a protein and / or gene (e.g., by inhibiting or reducing gene expression of that protein or by inducing an active protein to go to an inactive state) or reducing the activity of a protein and / or gene.

[0100] The terms "about" and "approximately," when used throughout this specification and claims in connection with numerical values, denote intervals of accuracy familiar and accepted by those of ordinary skill in the art. Generally, such intervals of accuracy are ±10%; alternatively, and particularly in biological systems, the terms "about" and "approximately" can mean values ​​within an order of magnitude of a given value, preferably ≦5-fold, and more preferably ≦2-fold.

[0101] The numerical ranges disclosed herein are inclusive of the numbers defining the range.

[0102] The terms "a" and "an" include plural referents unless the context in which such terms are used clearly dictates otherwise. The terms "a" (or "an") and "one or more" and "at least one" can be used interchangeably herein. Furthermore, "and / or," when used herein, should be construed as a specific disclosure of each of the two or more specified features or components, with or without the others. Thus, the term "and / or," when used herein in phrases such as "A and / or B," is intended to include "A and B," "A or B," "A" (alone) and "B" (alone). Similarly, the term "and / or," when used in phrases such as "A, B and / or C," is intended to encompass each of the following embodiments: A, B and C; A, B or C; A or C; A or B; B or C; A and C; A and B; B and C; A (alone); B (alone); and C (alone). 2. ActRIIB polypeptide

[0103] In certain aspects, the present invention relates to ActRIIB variant polypeptides (e.g., soluble ActRIIB polypeptides). Optionally, the fragments, functional variants, and modified forms have similar or identical biological activity to their corresponding wild-type ActRIIB polypeptides. For example, the ActRIIB variants of the present invention can bind to and inhibit the function of ActRIIB ligands (e.g., activin A, activin AB, activin B, nodal, GDF8, GDF11, or BMP7). Optionally, the ActRIIB polypeptides modulate the growth of tissues such as bone, cartilage, muscle, or fat. Examples of ActRIIB polypeptides include human ActRIIB precursor polypeptide (SEQ ID NO: 2) and soluble human ActRIIB polypeptides (e.g., SEQ ID NOs: 1, 5, 6, and 12).

[0104] The present disclosure identifies functionally active portions and variants of ActRIIB. Applicants have confirmed that an Fc fusion protein having the sequence disclosed by Hilden et al. (Blood., April 15, 1994; Vol. 83(8):2163-70), which has an alanine (A64) at position corresponding to amino acid 64 of SEQ ID NO:2, has relatively low affinity for activin and GDF11. In contrast, the same Fc fusion protein having an arginine (R64) at position 64 has affinity for activin and GDF-11 in the low nanomolar to high picomolar range. Therefore, the sequence having R64 is used as the wild-type reference sequence of human ActRIIB in the present disclosure.

[0105] Attisano et al. (Cell., January 10, 1992; Vol. 68(1):97-108) showed that deletion of the proline knot at the C-terminus of the extracellular domain of ActRIIB reduced the receptor's affinity for activin. Data disclosed in WO2008097541 show that an ActRIIB-Fc fusion protein, "ActRIIB(20-119)-Fc," containing amino acids 20-119 of SEQ ID NO:2, exhibits reduced binding to GDF11 and activin compared with ActRIIB(20-134)-Fc, which contains the proline knot region and the complete juxtamembrane domain. However, the ActRIIB(20-129)-Fc protein retains similar, though slightly reduced, activity compared to the wild-type protein, even though the proline knot region has been disrupted. Thus, ActRIIB extracellular domains ending at amino acids 134, 133, 132, 131, 130, and 129 are all expected to be active, but constructs ending at 134 or 133 may be the most active. Similarly, mutations at any of residues 129-134 are not expected to significantly alter ligand binding affinity. Supporting this, mutations at P129 and P130 do not substantially reduce ligand binding. Thus, ActRIIB-Fc fusion proteins can end as early as amino acid 109 (the final cysteine), but forms ending at or between 109 and 119 are expected to have reduced ligand binding. Amino acid 119 is poorly conserved and therefore easily modified or truncated. Forms ending at or after 128 retain ligand binding activity. Forms ending at or between 119 and 127 will have intermediate binding capacity. Either of these forms may be desirable for use, depending on the clinical or experimental setting.

[0106] Proteins beginning at or before amino acid 29 at the N-terminus of ActRIIB are predicted to retain ligand-binding activity. Amino acid 29 represents the first cysteine. Mutation of alanine to asparagine at position 24 introduces an N-linked glycosylation sequence without substantially affecting ligand binding. This confirms that mutations in the region between the signal cleavage peptide and the cysteine ​​bridge region corresponding to amino acids 20-29 are well tolerated. In particular, constructs beginning at positions 20, 21, 22, 23, and 24 are predicted to retain activity, as are constructs beginning at positions 25, 26, 27, 28, and 29. Data presented in WO2008097541 demonstrate, surprisingly, that constructs beginning at positions 22, 23, 24, or 25 are most likely to be active.

[0107] In summary, the active portion of ActRIIB comprises amino acids 29-109 of SEQ ID NO:2, and the construct can, for example, begin at a residue corresponding to amino acids 20-29 and end at a position corresponding to amino acids 109-134. Other examples include constructs beginning at positions 20-29 or 21-29 and ending at positions 119-134, 119-133, or 129-134, or 129-133. Other examples include constructs beginning at positions 20-24 (or 21-24 or 22-25) and ending at positions 109-134 (or 109-133), 119-134 (or 119-133), or 129-134 (or 129-133). Variants within these ranges are also contemplated, particularly those having at least 80%, 85%, 90%, 95%, or 99% identity to the corresponding portions of SEQ ID NO:4.

[0108] The present disclosure includes the results of an analysis of the composite ActRIIB structure shown in Figure 2, which demonstrates that the ligand-binding pocket is defined by residues Y31, N33, N35, L38 to T41, E47, E50, Q53 to K55, L57, H58, Y60, S62, K74, W78 to N83, Y85, R87, A92, and E94 to F101. Conservative mutations will be tolerated at these positions, but the K74A mutation is predicted to be well tolerated, as are R40A, K55A, F82A, and mutations at position L79. R40 is K in Xenopus, indicating that basic amino acids will be tolerated at this position. Q53 is R in bovine ActRIIB and K in Xenopus ActRIIB, therefore, amino acids containing R, K, Q, N, and H will be tolerated at this position. Thus, a general formula for an active ActRIIB variant protein includes amino acids 29-109, but optionally begins at positions spanning 20-24 or 22-25 and ends at positions spanning 129-134, and includes no more than 1, 2, 5, 10, or 15 conservative amino acid changes in the ligand-binding pocket, and zero, one, or more non-conservative changes at positions 40, 53, 55, 74, 79, and / or 82 in the ligand-binding pocket. Such proteins can retain greater than 80%, 90%, 95%, or 99% sequence identity with the sequence of amino acids 29-109 of SEQ ID NO: 2. Sites outside the binding pocket where variability may be particularly tolerated include the amino- and carboxy-termini of the extracellular domain (as noted above), as well as positions 42-46 and 65-73. An asparagine to alanine change at position 65 (N65A) actually improves ligand binding in the A64 background and is therefore not expected to have a detrimental effect on ligand binding in the R64 background. This change likely eliminates glycosylation at N65 in the A64 background, demonstrating that significant changes in this region are likely to be tolerated. The R64A change is poorly tolerated, whereas R64K is well tolerated, suggesting that another basic residue, such as H, may be tolerated at position 64.

[0109] ActRIIB is well conserved across nearly all vertebrates, with the large, extended extracellular domain being completely conserved (see Figure 3). Many of the ligands that bind to ActRIIB are also highly conserved. Therefore, comparison of ActRIIB sequences from various vertebrate organisms provides insight into residues that can be altered. Thus, an active human ActRIIB variant can contain one or more amino acids at corresponding positions from another vertebrate ActRIIB sequence, or can contain similar residues as in human or other vertebrate sequences. The following example illustrates this approach to defining active ActRIIB variants. Because L46 is a valine in Xenopus ActRIIB, this position can be altered, optionally to another hydrophobic residue such as V, I, or F, or a nonpolar residue such as A. E52 is K in Xenopus, indicating that this site can tolerate a wide variety of changes, including polar residues such as E, D, K, R, H, S, T, P, G, Y, and possibly A. T93 is K in Xenopus, indicating that a wide range of structural variation is tolerated at this position, with polar residues such as S, K, R, E, D, H, G, P, G, and Y being preferred. F108 is Y in Xenopus, and therefore Y or other hydrophobic groups such as I, V, or L should be tolerated. E111 is K in Xenopus, indicating that charged residues including D, R, K, and H, as well as Q and N, will be tolerated at this position. R112 is K in Xenopus, indicating that basic residues including R and H are tolerated at this position. A at position 119 is relatively poorly conserved, occurring as P in rodents and V in Xenopus, so essentially any amino acid should be tolerated at this position.

[0110] The data disclosed in WO2008097541 demonstrate that the addition of an additional N-linked glycosylation site (NXS / T) does not affect the activity of the ActRIIB-Fc fusion protein compared to the ActRIIB(R64)-Fc form. Other N-X(T / S) sequences are found at positions 42-44 (NQS) and 65-67 (NSS), although the latter may not be efficiently glycosylated due to the R at position 64. The N-X-S / T sequence can generally be introduced at a position outside the ligand-binding pocket defined in Figure 2. Particularly suitable sites for the introduction of a non-endogenous N-X-S / T sequence include amino acids 20-29, 20-24, 22-25, 109-134, 120-134, or 129-134. The N-X-S / T sequence can also be introduced into the linker between the ActRIIB sequence and the Fc or other fusion component. Such sites can be introduced with minimal effort by introducing an N at the correct position relative to an existing S or T, or by introducing an S or T at the position corresponding to an existing N. Thus, desirable changes that would create N-linked glycosylation sites are: A24N, R64N, S67N (possibly in combination with an N65A change), E106N, R112N, G120N, E123N, P129N, A132N, R112S, and R112T. Any S predicted to be glycosylated can be changed to a T without creating an immunogenic site because of the protection provided by glycosylation. Similarly, any T predicted to be glycosylated can be changed to an S. Thus, the changes S67T and S44T are contemplated. Similarly, in the A24N variant, an S26T change can be used. Thus, an ActRIIB variant can contain one or more additional non-endogenous N-linked glycosylation consensus sequences.

[0111] Position L79 can be altered to confer altered activin-myostatin (GDF-11) binding properties. L79A or L79P reduces GDF-11 binding to a greater extent than activin binding. L79E or L79D retains GDF-11 binding. Notably, the L79E and L79D variants have significantly reduced activin binding. In vivo experiments show that these non-activin receptors retain significant muscle mass-increasing abilities but exhibit reduced efficacy in other tissues. These data demonstrate the desirability and feasibility of obtaining polypeptides with reduced efficacy in activin.

[0112] The variations described can be combined in various ways. Moreover, the results of the mutagenesis program described herein indicate that there are amino acid positions in ActRIIB that are often beneficial to conserve. These include position 64 (basic amino acid), position 80 (acidic or hydrophobic amino acid), position 78 (hydrophobic, particularly tryptophan), position 37 (acidic, particularly aspartic acid or glutamic acid), position 56 (basic amino acid), and position 60 (hydrophobic amino acid, particularly phenylalanine or tyrosine). Thus, in each of the variants disclosed herein, the disclosure provides a framework of amino acids that may be conserved. Other positions where conservation may be desirable include: position 52 (acidic amino acid), position 55 (basic amino acid), position 81 (acidic), and position 98 (polar or charged, particularly E, D, R, or K).

[0113] In certain embodiments, isolated fragments of ActRIIB polypeptides can be obtained by screening polypeptides recombinantly produced from corresponding fragments of nucleic acids encoding ActRIIB polypeptides (e.g., SEQ ID NOS: 3 and 4). In addition, fragments can be chemically synthesized using techniques known in the art, such as conventional Merrifield solid-phase f-Moc or t-Boc chemistry. Fragments can be produced (recombinantly or by chemical synthesis) and tested to identify peptidyl fragments that can function, for example, as antagonists (inhibitors) or agonists (activators) of ActRIIB protein or ActRIIB ligand.

[0114] In certain embodiments, a functional variant of an ActRIIB polypeptide has an amino acid sequence that is at least 75% identical to an amino acid sequence selected from SEQ ID NOs: 1, 2, and 53. In certain cases, a functional variant has an amino acid sequence that is at least 80%, 85%, 90%, 95%, 97%, 98%, 99%, or 100% identical to an amino acid sequence selected from SEQ ID NOs: 1, 2, and 53.

[0115] In certain embodiments, the present invention contemplates the creation of functional variants by modifying the structure of an ActRIIB polypeptide, for purposes such as enhancing therapeutic efficacy or stability (e.g., ex vivo shelf life and resistance to proteolytic degradation in vivo). Modified ActRIIB polypeptides can also be produced, for example, by amino acid substitution, deletion, or addition. For example, it is reasonable to expect that isolated replacement of leucine with isoleucine or valine, aspartic acid with glutamic acid, or threonine with serine, or similar replacement of an amino acid with a structurally related amino acid (e.g., conservative mutation), will have no major effect on the biological activity of the resulting molecule. Conservative replacements are those that occur within a family of amino acids whose side chains are related. Whether a change in the amino acid sequence of an ActRIIB polypeptide results in a functional homolog can be readily determined by evaluating the ability of the variant ActRIIB polypeptide to produce a response in cells in a manner similar to that of the wild-type ActRIIB polypeptide or to bind to one or more ligands, such as activin, GDF11, or GDF8, in a manner similar to that of the wild-type.

[0116] In certain specific embodiments, the present invention contemplates making mutations in the extracellular domain (also referred to as the ligand-binding domain) of an ActRIIB polypeptide such that the variant (or mutant) ActRIIB polypeptide has altered ligand-binding activity (e.g., binding affinity or binding selectivity). In certain cases, such variant ActRIIB polypeptides have altered (increased or decreased) binding affinity for a specific ligand. In other cases, the variant ActRIIB polypeptides have altered binding selectivity for that ligand.

[0117] In certain embodiments, the present invention contemplates specific mutations of ActRIIB polypeptides that can alter the glycosylation of the polypeptide. Exemplary glycosylation sites in ActRIIB polypeptides are illustrated in Figure 6. Such mutations can be selected to introduce or eliminate one or more glycosylation sites, such as O-linked or N-linked glycosylation sites. Asparagine-linked glycosylation recognition sites generally contain the tripeptide sequence asparagine-X-threonine (where "X" is any amino acid), which is specifically recognized by appropriate cellular glycosylation enzymes. Alterations can also be made by adding or substituting one or more serine or threonine residues to the sequence of the wild-type ActRIIB polypeptide (for O-linked glycosylation sites). Various amino acid substitutions or deletions at one or both of the first or third amino acid positions (and / or amino acid deletions at the second position) of the glycosylation recognition site result in a lack of glycosylation in the modified tripeptide sequence. Another way to increase the number of carbohydrate moieties in ActRIIB polypeptides is by chemical or enzymatic coupling of glycosides to ActRIIB polypeptides. Depending on the coupling mode used, sugar(s) can be attached to (a) arginine and histidine; (b) free carboxyl groups; (c) free sulfhydryl groups, such as in cysteine; (d) free hydroxyl groups, such as in serine, threonine, or hydroxyproline; (e) aromatic residues, such as in phenylalanine, tyrosine, or tryptophan; or (f) the amide group of glutamine. These methods are described in WO87 / 05330 published September 11, 1987, and Aplin and Wriston (1981) CRC Crit., which are incorporated herein by reference. Rev. Biochem., pp. 259-306. Removal of one or more carbohydrate moieties present on an ActRIIB polypeptide can be accomplished chemically and / or enzymatically. Chemical deglycosylation can involve, for example, exposure of the ActRIIB polypeptide to the compound trifluoromethanesulfonic acid or an equivalent compound. This treatment results in the cleavage of most or all sugars except the linking sugar (N-acetylglucosamine or N-acetylgalactosamine), while leaving the amino acid sequence intact. Chemical deglycosylation is further described by Hakimuddin et al. (1987) Arch. Biochem. Biophys., 259:52 and Edge et al. (1981) Anal. Biochem., 118:131. Enzymatic cleavage of carbohydrate moieties in ActRIIB polypeptides can be achieved by the use of various endo- and exoglycosidases, as described by Thotakura et al. (1987) Meth. Enzymol. 138:350. Because mammalian, yeast, insect, and plant cells can all introduce different glycosylation patterns that can be influenced by the amino acid sequence of the peptide, the sequence of the ActRIIB polypeptide can be adjusted accordingly depending on the type of expression system used. Generally, ActRIIB proteins for use in humans will be expressed in mammalian cell lines that provide proper glycosylation, such as HEK293 or CHO cell lines, although other mammalian expression cell lines are expected to be useful as well.

[0118] The present disclosure further contemplates a method for generating a set of ActRIIB polypeptide variants, particularly combinatorial variants, optionally including truncated variants; a pool of combinatorial mutants is particularly useful for identifying functional variant sequences. The purpose of screening such combinatorial libraries can be to generate ActRIIB polypeptide variants with altered properties, such as, for example, altered pharmacokinetics or altered ligand binding. Various screening assays are presented below, and such assays can be used to evaluate variants. For example, ActRIIB polypeptide variants can be screened for their ability to bind to ActRIIB polypeptides and prevent the binding of ActRIIB ligands to ActRIIB polypeptides.

[0119] The activity of an ActRIIB polypeptide or variant thereof can be determined by cell-based or in vivo assays. It can also be tested in vivo assays. For example, the effect of an ActRIIB polypeptide variant on the expression of genes involved in bone production in osteoblasts or precursors can be evaluated. This can be done, if necessary, in the presence of one or more recombinant ActRIIB ligand proteins (e.g., BMP7), and the cells can be transfected to produce an ActRIIB polypeptide and / or its variant, and optionally an ActRIIB ligand. Similarly, an ActRIIB polypeptide can be administered to mice or other animals, and one or more bone properties, such as density or volume, can be evaluated. The healing rate of fractures can also be evaluated. Similarly, the activity of an ActRIIB polypeptide or its variant can be tested in muscle cells, adipocytes, and nerve cells for any effect on the growth of these cells, for example, by assays as described below. Such assays are well known and routine in the art. SMAD-responsive reporter genes can be used in such cell systems to monitor the effect on downstream signal transduction.

[0120] Combinatorially derived variants can be generated that have selective potency compared to naturally occurring ActRIIB polypeptides. When expressed from a recombinant DNA construct, such variant proteins can be used in gene therapy protocols. Similarly, mutagenesis can result in variants with dramatically different intracellular half-lives than the corresponding wild-type ActRIIB polypeptide. For example, the altered protein can be more or less stable to proteolytic degradation or other processes that destroy or otherwise inactivate the native ActRIIB polypeptide. Such variants and the genes encoding them can be used to modulate the half-life of the ActRIIB polypeptide, thereby altering ActRIIB polypeptide levels. For example, a shorter half-life can result in a more transient biological effect and, in the case of an inducible expression system, can allow for tighter regulation of recombinant ActRIIB polypeptide levels in cells.

[0121] In certain embodiments, the ActRIIB polypeptides of the present invention can further include post-translational modifications in addition to those naturally occurring in the ActRIIB polypeptide. Such modifications include, but are not limited to, acetylation, carboxylation, glycosylation, phosphorylation, lipidation, and acylation. As a result, modified ActRIIB polypeptides can contain non-amino acid elements, such as polyethylene glycol, lipids, polysaccharides or monosaccharides, and phosphates. The effect of such non-amino acid elements on the functionality of ActRIIB polypeptides can be examined as described herein for other ActRIIB polypeptide variants. When ActRIIB polypeptides are produced in cells by cleavage of nascent ActRIIB polypeptides, post-translational processing can also be important for the correct folding and / or function of the protein. Different cells (such as CHO, HeLa, MDCK, 293, WI38, NIH-3T3, or HEK293) have specific cellular mechanisms and characteristic mechanisms for such post-translational activities and can be selected to ensure the correct modification and processing of ActRIIB polypeptides.

[0122] In certain embodiments, functional variants or modified forms of ActRIIB polypeptides include fusion proteins having at least a portion of an ActRIIB polypeptide and one or more fusion domains. Well-known examples of such fusion domains include, but are not limited to, polyhistidine, Glu-Glu, glutathione S-transferase (GST), thioredoxin, protein A, protein G, immunoglobulin heavy chain constant regions (e.g., Fc), maltose-binding protein (MBP), or human serum albumin. The fusion domain can be selected to confer desired properties. For example, some fusion domains are particularly useful for isolating fusion proteins by affinity chromatography. For affinity purification, relevant matrices for affinity chromatography, such as glutathione, amylase, and nickel- or cobalt-conjugated resins, are used. Many such matrices are available in "kit" form, such as the Pharmacia GST purification system and the QIAexpress™ system (Qiagen), which are useful with (HIS6) fusion partners. As another example, the fusion domain can be selected to facilitate detection of the ActRIIB polypeptide. Examples of such detection domains include various fluorescent proteins (e.g., GFP), as well as "epitope tags," which are usually short peptide sequences for which specific antibodies are available. Well-known epitope tags for which specific monoclonal antibodies are readily available include FLAG, influenza virus hemagglutinin (HA), and c-myc tags. In some cases, the fusion domain has a protease cleavage site, such as that for factor Xa or thrombin, which allows the relevant protease to partially digest the fusion protein, thereby liberating the recombinant protein therefrom. The liberated protein can then be isolated from the fusion domain by subsequent chromatographic separation. In certain preferred embodiments, the ActRIIB polypeptide is fused to a domain that stabilizes the ActRIIB polypeptide in vivo (a "stabilizer" domain)."Stabilization" means anything that increases serum half-life, whether it be through reduced destruction, reduced renal clearance, or other pharmacokinetic effects. Fusion with the Fc portion of an immunoglobulin is known to confer desirable pharmacokinetic properties to a wide range of proteins. Similarly, fusion to human serum albumin can confer desirable properties. Other types of fusion domains that can be selected include multimerization (e.g., dimerization, tetramerization) domains and functional domains (which confer additional biological functions, such as further stimulation of muscle growth).

[0123] In certain embodiments, the polypeptides disclosed herein can form homomeric variant ActRIIB polypeptides, meaning that each fusion polypeptide chain in a protein complex contains the same ActRIIB variant as any other such chain in the complex. In certain embodiments, the polypeptides disclosed herein can form heteromultimers comprising at least one variant ActRIIB polypeptide covalently or non-covalently associated with at least one unmodified ActRIIB polypeptide or at least one variant ActRIIB polypeptide different from the first ActRIIB variant. In certain embodiments, the polypeptides disclosed herein can form heteromultimers comprising at least one variant ActRIIB polypeptide covalently or non-covalently associated with at least one TGF-beta superfamily type I serine / threonine kinase receptor polypeptide (e.g., ALK1, ALK2, ALK3, ALK4, ALK5, ALK6, and ALK7 polypeptides), including fragments and variants thereof. In certain aspects, the polypeptides disclosed herein can form heteromultimers comprising at least one variant ActRIIB polypeptide covalently or non-covalently associated with at least one TGF-beta superfamily type II serine / threonine kinase receptor polypeptide (e.g., ActRIIA, TGFBRII, BMPRII, and MISRII), including fragments and variants thereof. Preferably, the heteromeric polypeptides disclosed herein form heterodimers, but also include higher order heteromultimers, including, but not limited to, heterotrimers, heterotetramers, and further oligomeric structures. In some embodiments, the variant ActRIIB polypeptides of the present disclosure comprise at least one multimerization domain. As disclosed herein, the term "multimerization domain" refers to an amino acid or sequence of amino acids that promotes covalent or non-covalent interactions between at least a first polypeptide and at least a second polypeptide.The variant ActRIIB polypeptides disclosed herein can be covalently or non-covalently linked to a multimerization domain. Preferably, the multimerization domain promotes the interaction between a first polypeptide (e.g., a variant ActRIIB polypeptide) and a second polypeptide (e.g., an unmodified ActRIIB polypeptide or a variant ActRIIB polypeptide different from that present in the first polypeptide) to promote heteromultimer formation (e.g., heterodimer formation), and optionally prevents or does not support homomultimer formation (e.g., homodimer formation), thereby increasing the yield of the desired heteromultimer (see, e.g., Figure 1).

[0124] In certain embodiments, the present disclosure relates to a heteromultimer comprising at least one ALK1-Fc fusion protein and at least one ActRIIB-Fc fusion protein. In some embodiments, the ALK1-Fc:ActRIIB-Fc heteromultimer binds to one or more TGF-beta superfamily ligands, such as those described herein. In some embodiments, the ALK1-Fc:ActRIIB-Fc heteromultimer inhibits signal transduction of one or more TGF-beta superfamily ligands, such as those described herein. In some embodiments, the ALK1-Fc:ActRIIB-Fc heteromultimer is a heterodimer.

[0125] In certain embodiments, the present disclosure relates to a heteromultimer comprising at least one ALK2-Fc fusion protein and at least one ActRIIB-Fc fusion protein. In some embodiments, the ALK2-Fc:ActRIIB-Fc heteromultimer binds to one or more TGF-beta superfamily ligands, such as those described herein. In some embodiments, the ALK2-Fc:ActRIIB-Fc heteromultimer inhibits signal transduction of one or more TGF-beta superfamily ligands, such as those described herein. In some embodiments, the ALK2-Fc:ActRIIB-Fc heteromultimer is a heterodimer.

[0126] In certain embodiments, the present disclosure relates to a heteromultimer comprising at least one ALK3-Fc fusion protein and at least one ActRIIB-Fc fusion protein. In some embodiments, the ALK3-Fc:ActRIIB-Fc heteromultimer binds to one or more TGF-beta superfamily ligands, such as those described herein. In some embodiments, the ALK3-Fc:ActRIIB-Fc heteromultimer inhibits signal transduction of one or more TGF-beta superfamily ligands, such as those described herein. In some embodiments, the ALK3-Fc:ActRIIB-Fc heteromultimer is a heterodimer.

[0127] In certain embodiments, the present disclosure relates to a heteromultimer comprising at least one ALK4-Fc fusion protein and at least one ActRIIB-Fc fusion protein. In some embodiments, the ALK4-Fc:ActRIIB-Fc heteromultimer binds to one or more TGF-beta superfamily ligands, such as those described herein. In some embodiments, the ALK4-Fc:ActRIIB-Fc heteromultimer inhibits signal transduction of one or more TGF-beta superfamily ligands, such as those described herein. In some embodiments, the ALK4-Fc:ActRIIB-Fc heteromultimer is a heterodimer.

[0128] In certain embodiments, the present disclosure relates to a heteromultimer comprising at least one ALK5-Fc fusion protein and at least one ActRIIB-Fc fusion protein. In some embodiments, the ALK5-Fc:ActRIIB-Fc heteromultimer binds to one or more TGF-beta superfamily ligands, such as those described herein. In some embodiments, the ALK5-Fc:ActRIIB-Fc heteromultimer inhibits signal transduction of one or more TGF-beta superfamily ligands, such as those described herein. In some embodiments, the ALK5-Fc:ActRIIB-Fc heteromultimer is a heterodimer.

[0129] In certain embodiments, the present disclosure relates to a heteromultimer comprising at least one ALK6-Fc fusion protein and at least one ActRIIB-Fc fusion protein. In some embodiments, the ALK6-Fc:ActRIIB-Fc heteromultimer binds to one or more TGF-beta superfamily ligands, such as those described herein. In some embodiments, the ALK6-Fc:ActRIIB-Fc heteromultimer inhibits signal transduction of one or more TGF-beta superfamily ligands, such as those described herein. In some embodiments, the ALK6-Fc:ActRIIB-Fc heteromultimer is a heterodimer.

[0130] In certain embodiments, the present disclosure relates to a heteromultimer comprising at least one ALK7-Fc fusion protein and at least one ActRIIB-Fc fusion protein. In some embodiments, the ALK7-Fc:ActRIIB-Fc heteromultimer binds to one or more TGF-beta superfamily ligands, such as those described herein. In some embodiments, the ALK7-Fc:ActRIIB-Fc heteromultimer inhibits signal transduction of one or more TGF-beta superfamily ligands, such as those described herein. In some embodiments, the ALK7-Fc:ActRIIB-Fc heteromultimer is a heterodimer.

[0131] In certain embodiments, the present disclosure relates to a heteromultimer comprising at least one ActRIIB-Fc fusion protein and at least one ActRIIA-Fc fusion protein. In some embodiments, the ActRIIB-Fc:ActRIIA-Fc heteromultimer binds to one or more TGF-beta superfamily ligands, such as those described herein. In some embodiments, the ActRIIB-Fc:ActRIIA-Fc heteromultimer inhibits signal transduction of one or more TGF-beta superfamily ligands, such as those described herein. In some embodiments, the ActRIIB-Fc:ActRIIA-Fc heteromultimer is a heterodimer.

[0132] In certain embodiments, the present disclosure relates to a heteromultimer comprising at least one ActRIIB-Fc fusion protein and at least one BMPRII-Fc fusion protein. In some embodiments, the ActRIIB-Fc:BMPRII-Fc heteromultimer binds to one or more TGF-beta superfamily ligands, such as those described herein. In some embodiments, the ActRIIB-Fc:BMPRII-Fc heteromultimer inhibits signal transduction of one or more TGF-beta superfamily ligands, such as those described herein. In some embodiments, the ActRIIB-Fc:BMPRII-Fc heteromultimer is a heterodimer.

[0133] In certain embodiments, the present disclosure relates to a heteromultimer comprising at least one ActRIIB-Fc fusion protein and at least one TGFBRII-Fc fusion protein. In some embodiments, the ActRIIB-Fc:TGFBRII-Fc heteromultimer binds to one or more TGF-beta superfamily ligands, such as those described herein. In some embodiments, the ActRIIB-Fc:TGFBRII-Fc heteromultimer inhibits signal transduction of one or more TGF-beta superfamily ligands, such as those described herein. In some embodiments, the ActRIIB-Fc:TGFBRII-Fc heteromultimer is a heterodimer.

[0134] In certain embodiments, the present disclosure relates to a heteromultimer comprising at least one ActRIIB-Fc fusion protein and at least one MISRII-Fc fusion protein. In some embodiments, the ActRIIB-Fc:MISRII-Fc heteromultimer binds to one or more TGF-beta superfamily ligands, such as those described herein. In some embodiments, the ActRIIB-Fc:MISRII-Fc heteromultimer inhibits signal transduction of one or more TGF-beta superfamily ligands, such as those described herein. In some embodiments, the ActRIIB-Fc:MISRII-Fc heteromultimer is a heterodimer.

[0135] In certain aspects, the present disclosure relates to heteromultimers comprising an ALK1-Fc fusion protein. In some embodiments, the ALK1-Fc fusion protein begins with any one of amino acids 22-34 (e.g., amino acid residues 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, and 34) of SEQ ID NO: 54 and continues with any one of amino acids 95-118 (e.g., amino acid residues 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 30, 31, 32, 33, and 34) of SEQ ID NO: 54. 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, and 118). In some embodiments, the ALK1-Fc fusion protein comprises an ALK1 domain comprising an amino acid sequence at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to amino acids 22-118 of SEQ ID NO: 54. In some embodiments, the ALK1-Fc fusion protein comprises an ALK1 domain comprising an amino acid sequence at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to amino acids 34-95 of SEQ ID NO:54. In some embodiments, the ALK1-Fc fusion protein comprises an ALK1 domain comprising an amino acid sequence that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of any one of SEQ ID NOs: 54, 55, 56, 57, 58, 59, 60, 61, 62, and 63. A representative ALK1-Fc fusion polypeptide (SEQ ID NO: 60) is shown below: [ka] [ka]

[0136] The leader and linker sequences are underlined.

[0137] The mature ALK1-Fc fusion protein sequence (SEQ ID NO: 61) is shown below and can optionally be provided with a lysine removed from the C-terminus. [ka] In some embodiments, the ALK1-Fc fusion polypeptide (SEQ ID NO: 56) is shown below: [ka]

[0138] The leader and linker sequences are underlined. To induce heterodimer formation by certain Fc fusion polypeptides disclosed herein, two amino acid substitutions (replacing lysine with aspartic acid) can be introduced into the Fc domain of the ALK1-Fc fusion polypeptide, as shown double underlined above. The amino acid sequence of SEQ ID NO: 56 can optionally be provided with an additional lysine at the C-terminus. This ALK1-Fc fusion protein is encoded by the following nucleic acid (SEQ ID NO:258): [ka] [ka]

[0139] The mature ALK1-Fc fusion protein sequence (SEQ ID NO: 57) is shown below and can optionally be provided with an additional C-terminal lysine. [ka]

[0140] In certain embodiments, the present disclosure relates to a protein complex comprising ALK2 polypeptide.As used herein, the term "ALK2" refers to the activin receptor-like kinase-2 protein from any species, and the family of variants derived from such ALK2 protein by mutagenesis or other modifications.It is understood that the reference to ALK2 herein refers to any one of the currently identified forms.Members of the ALK2 family are generally transmembrane proteins composed of a ligand-binding extracellular domain with a cysteine-rich region, a transmembrane domain, and a cytoplasmic domain with predicted serine / threonine kinase activity.

[0141] The term "ALK2 polypeptide" includes polypeptides including any naturally occurring polypeptide of an ALK2 family member, as well as any variants thereof (including mutants, fragments, fusions, and peptidomimetic forms) that retain useful activity. The human ALK2 precursor protein sequence (NCBI reference sequence NP_001096.1) is as follows: [ka]

[0142] The signal peptide is indicated by a single underline and the extracellular domain is indicated in bold font. The processed extracellular ALK2 polypeptide sequence is shown below: [ka]

[0143] The nucleic acid sequence encoding the human ALK2 precursor protein is set forth in SEQ ID NO: 217, which corresponds to nucleotides 431 to 1957 of Genbank reference sequence NM_001105.4. The nucleic acid sequence encoding the extracellular ALK2 polypeptide is as set forth in SEQ ID NO: 218.

[0144] In certain embodiments, the present disclosure relates to heteromultimers comprising at least one ALK2 polypeptide, including fragments, functional variants, and modified forms thereof. Preferably, the ALK2 polypeptides for use in accordance with the disclosed invention (e.g., heteromultimers comprising ALK2 polypeptides and their uses) are soluble (e.g., the extracellular domain of ALK2). In other preferred embodiments, the ALK2 polypeptides for use in accordance with the disclosed invention bind to one or more TGF-beta superfamily ligands and / or inhibit (antagonize) their activity (e.g., induction of Smad2 / 3 and / or Smad1 / 5 / 8 signaling). In some embodiments, the heteromultimers of the present disclosure comprise at least one ALK2 polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of SEQ ID NO: 64 or 65. In some embodiments, the heteromultimeric complexes of the present disclosure consist of or consist essentially of at least one ALK2 polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 64 or 65.

[0145] In certain aspects, the present disclosure relates to heteromultimers comprising an ALK2-Fc fusion protein. In some embodiments, the ALK2-Fc fusion protein begins with any one of amino acids 21 to 35 of SEQ ID NO:64 (e.g., amino acid residues 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, and 35) and includes any one of amino acids 99 to 123 of SEQ ID NO:64 (e.g., amino acid residues 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, and 135). 10, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, and 123). In some embodiments, the ALK2-Fc fusion protein comprises an ALK2 domain comprising an amino acid sequence at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 35-99 of SEQ ID NO: 64. In some embodiments, the ALK2-Fc fusion protein comprises an ALK2 domain comprising an amino acid sequence at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 21-123 of SEQ ID NO: 64. In some embodiments, the ALK2-Fc fusion protein comprises an ALK2 domain comprising an amino acid sequence that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NOs: 64, 65, 66, 67, 68, 69, 70, 71, 72 and 73. In some embodiments, the ALK2-Fc fusion protein uses the TPA leader, which is shown below (SEQ ID NO: 66): [ka] [ka]

[0146] The signal sequence and linker sequence are underlined. To promote heterodimer formation with certain other Fc fusions disclosed herein, two amino acid substitutions (replacing lysine with aspartic acid) can be introduced into the Fc domain of the fusion protein, as shown double underlined above. The amino acid sequence of SEQ ID NO: 66 can optionally be provided with an additional lysine at the C-terminus. This ALK2-Fc fusion protein is encoded by the following nucleic acid (SEQ ID NO:244): [ka]

[0147] The mature ALK2-Fc fusion protein sequence (SEQ ID NO: 67) is shown below and can optionally be provided with an additional C-terminal lysine. [ka] [ka]

[0148] In another approach to promoting the formation of heteromultimeric complexes using asymmetric Fc fusion proteins, the Fc domain is altered to introduce complementary hydrophobic interactions and additional intermolecular disulfide bonds. In some embodiments, the ALK2-Fc fusion polypeptide (SEQ ID NO: 70) is shown below: [ka]

[0149] The leader and linker sequences are underlined. To induce heterodimer formation by certain Fc fusion polypeptides disclosed herein, four amino acid substitutions can be introduced into the Fc domain of the ALK2 fusion polypeptide, as shown by double underlines above. Furthermore, the C-terminal lysine residue of the Fc domain can be deleted. The amino acid sequence of SEQ ID NO: 70 can be provided, optionally, by removing the lysine from the C-terminus.

[0150] The mature ALK2-Fc fusion protein sequence (SEQ ID NO: 71) is shown below and can optionally be provided with a lysine removed from the C-terminus. [ka] [ka]

[0151] In certain embodiments, the present disclosure relates to a protein complex comprising an ALK2 polypeptide. As used herein, the term "ALK2" refers to a family of activin receptor-like kinase-2 proteins from any species, and variants derived from such ALK2 proteins by mutagenesis or other modifications. Reference to ALK2 herein is understood to refer to any one of the currently identified forms. Members of the ALK2 family are generally transmembrane proteins composed of a ligand-binding extracellular domain with a cysteine-rich region, a transmembrane domain, and a cytoplasmic domain with predicted serine / threonine kinase activity.

[0152] The term "ALK2 polypeptide" includes polypeptides including any naturally occurring polypeptide of an ALK2 family member, and any variant thereof (including mutants, fragments, fusions and peptidomimetic forms) that retain useful activity. A representative human ALK2 precursor protein sequence (NCBI Ref Seq NP_001096.1) is shown below: [ka]

[0153] The signal peptide is shown with a single underline and the extracellular domain is shown in bold font. The processed extracellular ALK2 polypeptide sequence is shown below: [ka]

[0154] The nucleic acid sequence encoding the human ALK2 precursor protein is set forth in SEQ ID NO: 217, which corresponds to nucleotides 431 to 1957 of Genbank reference sequence NM_001105.4. The nucleic acid sequence encoding the extracellular ALK2 polypeptide is as set forth in SEQ ID NO: 218.

[0155] In certain embodiments, the present disclosure relates to heteromultimers comprising at least one ALK2 polypeptide, including fragments, functional variants, and modified forms thereof. Preferably, the ALK2 polypeptide for use in accordance with the disclosed invention (e.g., heteromultimers comprising ALK2 polypeptides and their uses) is soluble (e.g., the extracellular domain of ALK2). In other preferred embodiments, the ALK2 polypeptide for use in accordance with the disclosed invention binds to one or more TGF-beta superfamily ligands and / or inhibits (antagonizes) the activity of such ligands (e.g., induction of Smad2 / 3 and / or Smad1 / 5 / 8 signaling). In some embodiments, the heteromultimers of the present disclosure comprise at least one ALK2 polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of SEQ ID NO: 64 or 65. In some embodiments, the heteromultimeric complexes of the present disclosure consist of or consist essentially of at least one ALK2 polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 64 or 65.

[0156] In certain embodiments, the present disclosure relates to a protein complex comprising an ALK3 polypeptide. As used herein, the term "ALK3" refers to a family of activin receptor-like kinase-3 proteins from any species, and variants derived from such ALK3 proteins by mutagenesis or other modifications. Reference to ALK3 herein is understood to refer to any one of the currently identified forms. Members of the ALK3 family are generally transmembrane proteins composed of a ligand-binding extracellular domain with a cysteine-rich region, a transmembrane domain, and a cytoplasmic domain with predicted serine / threonine kinase activity.

[0157] The term "ALK3 polypeptide" includes polypeptides including any naturally occurring polypeptide of an ALK3 family member, as well as any variants thereof (including mutants, fragments, fusions, and peptidomimetic forms) that retain useful activity. The human ALK3 precursor protein sequence (NCBI reference sequence NP_004320.2) is as follows: [ka] [ka]

[0158] The signal peptide is indicated by a single underline and the extracellular domain is indicated in bold font. The processed extracellular ALK3 polypeptide sequence is as follows: [ka]

[0159] The nucleic acid sequence encoding the human ALK3 precursor protein is set forth in SEQ ID NO: 219, which corresponds to nucleotides 549-2144 of Genbank reference sequence NM_004329.2. The signal sequence is underlined, and the extracellular domain is shown in bold font. The nucleic acid sequence encoding the extracellular human ALK3 polypeptide is set forth in SEQ ID NO: 220.

[0160] A general formula for an active (e.g., ligand-binding) ALK3 polypeptide is one that includes a polypeptide beginning at any amino acid position 25-31 (i.e., positions 25, 26, 27, 28, 29, 30, or 31) of SEQ ID NO: 74 and ending at any amino acid position 140-152 (i.e., 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, or 152) of SEQ ID NO: 74. See U.S. Patent No. 8,338,377, the teachings of which are incorporated herein by reference in their entirety.

[0161] In certain embodiments, the present disclosure relates to heteromultimers comprising at least one ALK3 polypeptide, including fragments, functional variants, and modified forms thereof.Preferably, the ALK3 polypeptide for use according to the present disclosure's invention (e.g., heteromultimers comprising ALK3 polypeptides and their uses) is soluble (e.g., the extracellular domain of ALK3).In other preferred embodiments, the ALK3 polypeptide for use according to the present disclosure's invention binds to one or more TGF-beta superfamily ligands and / or inhibits (antagonizes) the activity of such ligands (e.g., inducing Smad2 / 3 and / or Smad1 / 5 / 8 signal transduction). In some embodiments, heteromultimers of the disclosure comprise at least one ALK3 polypeptide comprising amino acids beginning at any of amino acid positions 25-31 (i.e., positions 25, 26, 27, 28, 29, 30, or 31) of SEQ ID NO: 74 and ending at any of amino acid positions 140-153 (i.e., 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, or 152) of SEQ ID NO: 74. In some embodiments, heteromultimeric complexes of the disclosure comprise at least one ALK3 polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 74, 75, 76, 77, 80, or 81. In some embodiments, the heteromultimeric complexes of the present disclosure consist of or consist essentially of at least one ALK3 polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98% or 99% identical to the amino acid sequence of SEQ ID NO: 74, 75, 76, 77, 80 or 81.

[0162] In certain aspects, the present disclosure relates to heteromultimers comprising an ALK3-Fc fusion protein. In some embodiments, the ALK3-Fc fusion protein begins with any one of amino acids 24 to 61 of SEQ ID NO: 74 (e.g., amino acid residues 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, and 61) and continues with any one of amino acids 130 to 152 of SEQ ID NO: 74 (e.g., amino acid residues 130, 152, 153, 154, 155, 156, 157, 158, 159, 60, and 61). 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, and 152). In some embodiments, the ALK3-Fc fusion protein comprises an ALK3 domain comprising an amino acid sequence at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 61-130 of SEQ ID NO: 74. In some embodiments, the ALK3-Fc fusion protein comprises an ALK3 domain comprising an amino acid sequence at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 24-152 of SEQ ID NO: 74. In some embodiments, the ALK3-Fc fusion protein comprises an ALK3 domain comprising an amino acid sequence that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of any one of SEQ ID NOs: 74, 75, 76, 77, 78, 79, 80, 81, 82 and 83. In some embodiments, the ALK3-Fc fusion protein uses the TPA leader, as follows: [ka] [ka]

[0163] The leader and linker sequences are underlined. To promote the formation of ActRIIB-Fc:ALK3-Fc heterodimers rather than either of the possible homodimeric complexes, two amino acid substitutions (replacing lysine with aspartic acid) can be introduced into the Fc domain of the fusion protein, shown double underlined above. The amino acid sequence of SEQ ID NO: 76 can optionally be provided with an additional lysine at the C-terminus.

[0164] This ALK3-Fc fusion protein is encoded by the following nucleic acid (SEQ ID NO:245). [ka]

[0165] The mature ALK3-Fc fusion protein sequence is shown below (SEQ ID NO: 77) and can optionally be provided with an additional C-terminal lysine. [ka] The complementary form of the ALK3-Fc fusion polypeptide (SEQ ID NO: 80) is shown below: [ka]

[0166] The leader sequence and linker are underlined. To induce heterodimer formation by certain Fc fusion polypeptides disclosed herein, four amino acid substitutions can be introduced into the Fc domain of the ALK3 fusion polypeptide, as shown by double underlining above. The amino acid sequence of SEQ ID NO: 80 can optionally be provided with the lysine removed from the C-terminus.

[0167] The mature ALK3-Fc fusion protein sequence (SEQ ID NO: 81) is shown below and can optionally be provided with a lysine (K) removed from the C-terminus. [ka] [ka]

[0168] In certain embodiments, the present disclosure relates to a protein complex comprising an ALK4 polypeptide. As used herein, the term "ALK4" refers to the activin receptor-like kinase-4 protein family from any species, and variants derived from such ALK4 proteins by mutagenesis or other modifications. Reference to ALK4 herein is understood to refer to any one of the currently identified forms. Members of the ALK4 family are generally transmembrane proteins composed of a ligand-binding extracellular domain with a cysteine-rich region, a transmembrane domain, and a cytoplasmic domain with predicted serine / threonine kinase activity.

[0169] The term "ALK4 polypeptide" includes polypeptides including any naturally occurring polypeptide of an ALK4 family member, as well as any variants thereof (including mutants, fragments, fusions, and peptidomimetic forms) that retain useful activity. The human ALK4 precursor protein sequence (NCBI reference sequence NP_004293) is as follows: [ka]

[0170] The signal peptide is indicated by a single underline and the extracellular domain is indicated in bold font. The processed extracellular human ALK4 polypeptide sequence is as follows: [ka]

[0171] The nucleic acid sequence encoding the ALK4 precursor protein is set forth in SEQ ID NO: 221, which corresponds to nucleotides 78 to 1592 of Genbank reference sequence NM_004302.4. The nucleic acid sequence encoding the extracellular ALK4 polypeptide is set forth in SEQ ID NO: 222. An alternative isoform of the human ALK4 precursor protein sequence, isoform C (NCBI Ref Seq NP_064733.3), is shown below: [ka]

[0172] The signal peptide is indicated by a single underline and the extracellular domain is indicated in bold font. The processed extracellular ALK4 polypeptide sequence (isoform C) is as follows: [ka]

[0173] The nucleic acid sequence encoding the ALK4 precursor protein (isoform C) is set forth in SEQ ID NO: 223, which corresponds to nucleotides 78 to 1715 of Genbank reference sequence NM_020328.3. The nucleic acid sequence encoding the extracellular ALK4 polypeptide (isoform C) is set forth in SEQ ID NO: 224.

[0174] In certain embodiments, the present disclosure relates to heteromultimers comprising at least one ALK4 polypeptide, including fragments, functional variants, and modified forms thereof. Preferably, the ALK4 polypeptide for use according to the present disclosure's invention (e.g., heteromultimers comprising ALK4 polypeptides and their uses) is soluble (e.g., the extracellular domain of ALK4). In other preferred embodiments, the ALK4 polypeptide for use according to the present disclosure's invention binds to one or more TGF-beta superfamily ligands and / or inhibits (antagonizes) the activity of such ligands (e.g., induction of Smad2 / 3 and / or Smad1 / 5 / 8 signaling). In some embodiments, the heteromultimer of the present disclosure comprises at least one ALK4 polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 84, 86, 85, 87, 88, 89, 92, or 93. In some embodiments, the heteromultimers of the present disclosure consist of or essentially consist of at least one ALK4 polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 84, 86, 85, 87, 88, 89, 92 or 93.

[0175] In certain aspects, the present disclosure relates to heteromultimers comprising an ALK4-Fc fusion protein. In some embodiments, the ALK4-Fc fusion protein begins with any one of amino acids 23-34 (e.g., amino acid residues 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34) of SEQ ID NO: 84 or 85 and continues with any one of amino acids 101-126 (e.g., amino acid residues 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134) of SEQ ID NO: 84 or 85. 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, and 126). In some embodiments, the ALK4-Fc fusion protein comprises an ALK4 domain comprising an amino acid sequence at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 34-101 of SEQ ID NO: 84 or 85. In some embodiments, the ALK4-Fc fusion protein comprises an ALK4 domain comprising an amino acid sequence at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 23-126 of SEQ ID NO: 84 or 85. In some embodiments, the ALK4-Fc fusion protein comprises an ALK4 domain comprising an amino acid sequence that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of any one of SEQ ID NOs: 84, 86, 85, 87, 88, 89, 90, 91, 92, 93, 94 and 95. In certain embodiments, the polypeptide comprises the following ALK4-Fc fusion polypeptide (SEQ ID NO: 88): [ka]

[0176] The leader and linker sequences are underlined. To induce heterodimer formation by certain Fc fusion polypeptides of the present disclosure, two amino acid substitutions (replacing lysine with aspartic acid) can be introduced into the Fc domain of the ALK4-Fc fusion polypeptide, as shown double underlined above. The amino acid sequence of SEQ ID NO: 88 can optionally be provided with an additional lysine at the C-terminus. This ALK4-Fc fusion protein is encoded by the following nucleic acid (SEQ ID NO:243): [ka]

[0177] The mature ALK4-Fc fusion protein sequence (SEQ ID NO: 89) is shown below and can optionally be provided with an additional C-terminal lysine. [ka]

[0178] In some embodiments, the ALK4-Fc fusion polypeptide (or any Fc fusion polypeptide disclosed herein) uses a tissue plasminogen activator (TPA) leader: [ka]

[0179] In some embodiments, the ALK4-Fc fusion polypeptide (SEQ ID NO: 92) is shown below and can optionally be provided with the lysine removed from the C-terminus. [ka]

[0180] The leader sequence and linker are underlined. To induce heterodimer formation by certain Fc fusion polypeptides disclosed herein, four amino acid substitutions can be introduced into the Fc domain of the ALK4 fusion polypeptide, as shown by double underlining above. The amino acid sequence of SEQ ID NO: 92 can optionally be provided with the lysine removed from the C-terminus.

[0181] The mature ALK4-Fc fusion protein sequence is as follows, optionally provided with the lysine removed from the C-terminus: [ka]

[0182] Purification of various ActRIIB-Fc:ALK4-Fc complexes can be achieved, for example, by a series of column chromatography steps, including three or more of the following, in any order: Protein A chromatography, Q Sepharose chromatography, phenyl Sepharose chromatography, size exclusion chromatography, and cation exchange chromatography. Purification can be completed by viral filtration and buffer exchange.

[0183] In some embodiments, the ALK4-Fc fusion polypeptide (SEQ ID NO: 247) is shown below and can optionally be provided with the lysine removed from the C-terminus. [ka]

[0184] The leader sequence and linker are underlined. To induce heterodimer formation by certain Fc fusion polypeptides disclosed herein, four amino acid substitutions (replacing tyrosine with cysteine, threonine with serine, leucine with alanine, and tyrosine with valine) can be introduced into the Fc domain of the ALK4 fusion polypeptide, as shown double underlined above. To facilitate purification of the ALK4-Fc:ActRIIB-Fc heterodimer, two amino acid substitutions (replacing asparagine with arginine and aspartic acid with arginine) can also be introduced into the Fc domain of the ALK4-Fc fusion polypeptide, as shown double underlined above. The amino acid sequence of SEQ ID NO:247 can optionally be provided with the lysine removed from the C-terminus. This ALK4-Fc fusion polypeptide is encoded by the following nucleic acid (SEQ ID NO:248): [ka] [ka]

[0185] The mature ALK4-Fc fusion polypeptide sequence is shown below (SEQ ID NO: 249), and can optionally be provided with a lysine removed from the C-terminus. [ka] This ALK4-Fc fusion polypeptide is encoded by the following nucleic acid (SEQ ID NO:250): [ka] [ka]

[0186] In certain embodiments, the ALK4-Fc fusion polypeptide is SEQ ID NO: 92 (shown above) containing four amino acid substitutions, optionally provided with a lysine removed from the C-terminus, to induce heterodimerization of certain Fc fusion polypeptides disclosed herein. This ALK4-Fc fusion polypeptide is encoded by the following nucleic acid (SEQ ID NO:251): [ka] [ka]

[0187] The mature ALK4-Fc fusion polypeptide sequence is SEQ ID NO:93 (shown above), which can optionally be provided with a lysine removed from the C-terminus. This ALK4-Fc fusion polypeptide is encoded by the following nucleic acid (SEQ ID NO:252): [ka]

[0188] Purification of various ALK4-Fc:ActRIIB-Fc complexes can be achieved, for example, by a series of column chromatography steps, including, in any order, three or more of the following: Protein A chromatography, Q Sepharose chromatography, phenyl Sepharose chromatography, size exclusion chromatography, and epitope-based affinity chromatography (e.g., using antibodies or functionally equivalent ligands against epitopes on ALK4 or ActRIIB), and multimodal chromatography (using resins containing both electrostatic and hydrophobic ligands). Purification can be completed using viral filtration and buffer exchange.

[0189] In certain embodiments, the present disclosure relates to a protein complex comprising an ALK5 polypeptide. As used herein, the term "ALK5" refers to a family of activin receptor-like kinase-5 proteins from any species, and variants derived from such ALK4 proteins by mutagenesis or other modifications. Reference to ALK5 herein is understood to refer to any one of the currently identified forms. Members of the ALK5 family are generally transmembrane proteins composed of a ligand-binding extracellular domain with a cysteine-rich region, a transmembrane domain, and a cytoplasmic domain with predicted serine / threonine kinase activity.

[0190] The term "ALK5 polypeptide" includes polypeptides including any naturally occurring polypeptide of an ALK5 family member, as well as any variants thereof (including mutants, fragments, fusions, and peptidomimetic forms) that retain useful activity. The human ALK5 precursor protein sequence (NCBI reference sequence NP_004603.1) is as follows: [ka]

[0191] The signal peptide is shown with a single underline and the extracellular domain is shown in bold font. The processed extracellular ALK5 polypeptide sequence is shown below: [ka]

[0192] The nucleic acid sequence encoding the ALK5 precursor protein is set forth in SEQ ID NO: 225, which corresponds to nucleotides 77 to 1585 of Genbank reference sequence NM_004612.2. The nucleic acid sequence encoding the extracellular human ALK5 polypeptide is set forth in SEQ ID NO: 226. An alternative isoform of the human ALK5 precursor protein sequence, isoform 2 (NCBI reference sequence XP_005252207.1), is as follows: [ka]

[0193] The signal peptide is indicated by a single underline and the extracellular domain is indicated in bold font. The processed extracellular ALK5 polypeptide sequence (isoform 2) is shown below: [ka]

[0194] The nucleic acid sequence encoding the human ALK5 precursor protein (isoform 2) is set forth in SEQ ID NO: 227, which corresponds to nucleotides 77 to 1597 of Genbank reference sequence XM_005252150.1. The nucleic acid sequence encoding the processed extracellular ALK5 polypeptide is set forth in SEQ ID NO: 228.

[0195] In certain embodiments, the present disclosure relates to heteromultimers comprising at least one ALK5 polypeptide, including fragments, functional variants, and modified forms thereof. Preferably, the ALK5 polypeptide for use in accordance with the disclosed invention (e.g., heteromultimers comprising ALK5 polypeptides and their uses) is soluble (e.g., the extracellular domain of ALK5). In other preferred embodiments, the ALK5 polypeptide for use in accordance with the disclosed invention binds to one or more TGF-beta superfamily ligands and / or inhibits (antagonizes) the activity of such ligands (e.g., induction of Smad2 / 3 and / or Smad1 / 5 / 8 signaling). In some embodiments, the heteromultimers of the present disclosure comprise at least one ALK5 polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of SEQ ID NO: 96, 98, 97, or 99. In some embodiments, the heteromultimeric complexes of the present disclosure consist of or consist essentially of at least one ALK5 polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 96, 98, 97 or 99.

[0196] In certain aspects, the present disclosure relates to heteromultimers comprising an ALK5-Fc fusion protein. In some embodiments, the ALK5-Fc fusion protein begins with any one of amino acids 25 to 36 (e.g., amino acid residues 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, and 36) of SEQ ID NO: 96 or 97, and continues with any one of amino acids 106 to 126 (e.g., amino acid residues 106, 107, 108, 109, 110, 111, 112, 113, 114, and 116) of SEQ ID NO: 96 or 97. , 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125 and 126). In some embodiments, the ALK5-Fc fusion protein comprises an ALK5 domain comprising an amino acid sequence at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to amino acids 36-106 of SEQ ID NO: 96 or 97. In some embodiments, the ALK5-Fc fusion protein comprises an ALK5 domain comprising an amino acid sequence at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to amino acids 25-126 of SEQ ID NO: 96 or 97. In some embodiments, the ALK5-Fc fusion protein comprises an ALK5 domain comprising an amino acid sequence that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of any one of SEQ ID NOs: 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, and 106. The complementary ALK5-Fc fusion protein uses the TPA leader, which is shown below (SEQ ID NO: 100): [ka]

[0197] The signal sequence and linker sequence are underlined. To promote the formation of ActRIIB-Fc:ALK5-Fc heterodimers rather than either of the possible homodimeric complexes, two amino acid substitutions (replacing lysine with aspartic acid) can be introduced into the Fc domain of the fusion protein, shown double underlined above. The amino acid sequence of SEQ ID NO: 100 can optionally be provided with an additional lysine at the C-terminus. This ALK5-Fc fusion protein is encoded by the following nucleic acid (SEQ ID NO:253): [ka]

[0198] The mature ALK5-Fc fusion protein sequence (SEQ ID NO: 101) is shown below and can optionally be provided with an additional C-terminal lysine. [ka] [ka]

[0199] In another approach to promoting the formation of heteromultimeric complexes using asymmetric Fc fusion proteins, the Fc domain is altered to introduce complementary hydrophobic interactions and additional intermolecular disulfide bonds. In some embodiments, the ALK5-Fc fusion polypeptide (SEQ ID NO: 104) is shown below: [ka]

[0200] The leader and linker sequences are underlined. To induce heterodimer formation by certain Fc fusion polypeptides disclosed herein, four amino acid substitutions can be introduced into the Fc domain of the ALK5 fusion polypeptide, as shown by double underlines above. Furthermore, the C-terminal lysine residue of the Fc domain can be deleted. The amino acid sequence of SEQ ID NO: 104 can optionally be provided by removing the lysine from the C-terminus.

[0201] The mature ALK5-Fc fusion protein sequence (SEQ ID NO: 105) is shown below and can optionally be provided with a lysine removed from the C-terminus. [ka] [ka]

[0202] In certain embodiments, the present disclosure relates to a protein complex comprising an ALK6 polypeptide. As used herein, the term "ALK6" refers to the activin receptor-like kinase-6 protein family from any species, and variants derived from such ALK6 proteins by mutagenesis or other modifications. Reference to ALK6 herein is understood to refer to any one of the currently identified forms. Members of the ALK6 family are generally transmembrane proteins composed of a ligand-binding extracellular domain with a cysteine-rich region, a transmembrane domain, and a cytoplasmic domain with predicted serine / threonine kinase activity.

[0203] The term "ALK6 polypeptide" includes polypeptides including any naturally occurring polypeptide of an ALK6 family member, as well as any variants thereof (including mutants, fragments, fusions, and peptidomimetic forms) that retain useful activity. The human ALK6 precursor protein sequence (NCBI reference sequence NP_001194.1) is as follows: [ka]

[0204] The signal peptide is indicated by a single underline and the extracellular domain is indicated in bold font. The processed extracellular ALK6 polypeptide sequence is shown below: [ka]

[0205] The nucleic acid sequence encoding the ALK6 precursor protein is set forth in SEQ ID NO: 229, which corresponds to nucleotides 275 to 1780 of Genbank reference sequence NM_001203.2. The nucleic acid sequence encoding the processed extracellular ALK6 polypeptide is set forth in SEQ ID NO: 230. An alternative isoform of the human ALK6 precursor protein sequence, isoform 2 (NCBI reference sequence NP_001243722.1), is as follows: [ka]

[0206] The signal peptide is indicated by a single underline and the extracellular domain is indicated in bold font. The processed extracellular ALK6 polypeptide sequence (isoform 2) is shown below: [ka]

[0207] The nucleic acid sequence encoding the human ALK6 precursor protein (isoform 2) is set forth in SEQ ID NO: 231, which corresponds to nucleotides 22 to 1617 of Genbank reference sequence NM_001256793.1. The nucleic acid sequence encoding the processed extracellular ALK6 polypeptide is set forth in SEQ ID NO: 232.

[0208] In certain embodiments, the present disclosure relates to heteromultimers comprising at least one ALK6 polypeptide, including fragments, functional variants, and modified forms thereof. Preferably, the ALK6 polypeptide for use in accordance with the presently disclosed invention (e.g., heteromultimers comprising ALK6 polypeptides and their uses) is soluble (e.g., the extracellular domain of ALK6). In other preferred embodiments, the ALK6 polypeptide for use in accordance with the presently disclosed invention binds to one or more TGF-beta superfamily ligands and / or inhibits (antagonizes) the activity of such ligands (e.g., induction of Smad2 / 3 and / or Smad1 / 5 / 8 signaling). In some embodiments, the heteromultimers of the present disclosure comprise at least one ALK6 polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 108, 109, 110, or 111. In some embodiments, the heteromultimers of the present disclosure consist of or essentially consist of at least one ALK6 polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 108, 109, 110 or 111.

[0209] In certain aspects, the present disclosure relates to heteromultimers comprising an ALK6-Fc fusion protein. In some embodiments, the ALK6-Fc fusion protein begins with any one of amino acids 14 to 32 (e.g., amino acid residues 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, and 32) of SEQ ID NO: 108 and includes any one of amino acids 102 to 126 (e.g., amino acid residues 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, and 132) of SEQ ID NO: 108. 11, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, and 126). In some embodiments, the ALK6-Fc fusion protein comprises an ALK6 domain comprising an amino acid sequence at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to amino acids 32-102 of SEQ ID NO: 108. In some embodiments, the ALK6-Fc fusion protein comprises an ALK6 domain comprising an amino acid sequence at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to amino acids 14-126 of SEQ ID NO: 108. In some embodiments, the ALK6-Fc fusion protein comprises an ALK6 domain comprising an amino acid sequence that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of any one of SEQ ID NOs: 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, and 119.

[0210] In some embodiments, the ALK6-Fc fusion protein begins with any one of amino acids 26-62 of SEQ ID NO: 110 (e.g., amino acid residues 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, and 62) and continues with any one of amino acids 132-156 of SEQ ID NO: 110 (e.g., amino acid residues 132, 133, 134, 135, 136, 137, 138, 139, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, and 62). 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, and 156). In some embodiments, the ALK6-Fc fusion protein comprises an ALK6 domain comprising an amino acid sequence at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to amino acids 62-132 of SEQ ID NO: 110. In some embodiments, the ALK6-Fc fusion protein comprises an ALK6 domain comprising an amino acid sequence at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to amino acids 26-156 of SEQ ID NO: 110. The complementary ALK6-Fc fusion protein uses the TPA leader and is shown below (SEQ ID NO: 112): [ka]

[0211] The signal sequence and linker sequence are underlined. To promote the formation of the ActRIIB-Fc:ALK6-Fc heterodimer rather than one of the possible homodimeric complexes, two amino acid substitutions (replacing lysine with aspartic acid) can be introduced into the Fc domain of the fusion protein, as shown double underlined above. The amino acid sequence of SEQ ID NO: 112 can optionally be provided with an additional lysine at the C-terminus. This ALK6-Fc fusion protein is encoded by the following nucleic acid (SEQ ID NO:254): [ka] [ka]

[0212] The mature ALK6-Fc fusion protein sequence (SEQ ID NO: 113) is shown below and can optionally be provided with an additional C-terminal lysine. [ka]

[0213] In another approach to promoting heteromultimeric complex formation using asymmetric Fc fusion proteins, the Fc domain can be altered to introduce complementary hydrophobic interactions and additional intermolecular disulfide bonds. The complementary form of the ALK6-Fc fusion polypeptide (SEQ ID NO: 116) is shown below: [ka] [ka]

[0214] The leader and linker sequences are underlined. To induce heterodimer formation by certain Fc fusion polypeptides disclosed herein, four amino acid substitutions can be introduced into the Fc domain of the ALK6 fusion polypeptide, as shown by double underlines above. Furthermore, the C-terminal lysine residue of the Fc domain can be deleted. The amino acid sequence of SEQ ID NO: 116 can optionally be provided by removing the lysine from the C-terminus.

[0215] The mature ALK6-Fc fusion protein sequence (SEQ ID NO: 117) can be as follows, optionally provided with the lysine removed from the C-terminus: [ka]

[0216] In certain embodiments, the present disclosure relates to heteromultimers comprising at least one ALK6 polypeptide, including fragments, functional variants, and modified forms thereof. Preferably, the ALK6 polypeptide for use in accordance with the presently disclosed invention (e.g., heteromultimers comprising ALK6 polypeptides and their uses) is soluble (e.g., the extracellular domain of ALK6). In other preferred embodiments, the ALK6 polypeptide for use in accordance with the presently disclosed invention binds to one or more TGF-beta superfamily ligands and / or inhibits (antagonizes) the activity of such ligands (e.g., induction of Smad2 / 3 and / or Smad1 / 5 / 8 signaling). In some embodiments, the heteromultimers of the present disclosure comprise at least one ALK6 polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of SEQ ID NO: 34, 35, 91, or 92. In some embodiments, the heteromultimers of the present disclosure consist of, or consist essentially of, at least one ALK6 polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of SEQ ID NO: 34, 35, 91, or 92.

[0217] In certain embodiments, the present disclosure relates to a protein complex comprising an ALK7 polypeptide. As used herein, the term "ALK7" refers to the activin receptor-like kinase-7 protein family from any species, and variants derived from such ALK7 proteins by mutagenesis or other modifications. Reference to ALK7 herein is understood to refer to any one of the currently identified forms. Members of the ALK7 family are generally transmembrane proteins composed of a ligand-binding extracellular domain with a cysteine-rich region, a transmembrane domain, and a cytoplasmic domain with predicted serine / threonine kinase activity.

[0218] The term "ALK7 polypeptide" includes polypeptides including any naturally occurring polypeptide of an ALK7 family member, as well as any variants thereof (including mutants, fragments, fusions, and peptidomimetic forms) that retain useful activity. Four naturally occurring isoforms of human ALK7 have been described. The sequence of the human ALK7 isoform 1 precursor protein (NCBI reference sequence NP_660302.2) is as follows: [ka] The signal peptide is shown with a single underline and the extracellular domain is shown in bold font. The processed extracellular ALK7 isoform 1 polypeptide sequence is shown below: [ka]

[0219] The nucleic acid sequence encoding the human ALK7 isoform 1 precursor protein is shown below in SEQ ID NO: 233, which corresponds to nucleotides 244 to 1722 of Genbank reference sequence NM_145259.2. The nucleic acid sequence encoding the processed extracellular ALK7 polypeptide (isoform 1) is shown in SEQ ID NO: 234.

[0220] The amino acid sequence of an alternative isoform of human ALK7, isoform 2 (NCBI Ref Seq NP_001104501.1) is shown in its processed form as follows (SEQ ID NO: 124), in which the extracellular domain is shown in bold font: [ka] The amino acid sequence of the extracellular ALK7 polypeptide (isoform 2) is shown below:

[0221] [ka]

[0222] The nucleic acid sequence encoding the processed ALK7 polypeptide (isoform 2) is shown below in SEQ ID NO: 235, which corresponds to nucleotides 279 to 1607 of the NCBI reference sequence NM_001111031.1.

[0223] The nucleic acid sequence encoding the extracellular ALK7 polypeptide (isoform 2) is set forth in SEQ ID NO:236.

[0224] The amino acid sequence of the alternative human ALK7 precursor protein, isoform 3 (NCBI Ref Seq NP_001104502.1) is shown below (SEQ ID NO: 121), in which the signal peptide is indicated by a single underline. [ka]

[0225] The amino acid sequence of the processed ALK7 polypeptide (isoform 3) is shown below (SEQ ID NO: 126). This isoform lacks the transmembrane domain and is therefore proposed to be soluble in its entirety (Roberts et al., 2003, Biol Reprod 68:1719-1726). N-terminal variants of SEQ ID NO: 126 are predicted as described below. [ka]

[0226] The nucleic acid sequence encoding the unprocessed ALK7 polypeptide precursor protein (isoform 3) is set forth in SEQ ID NO: 237, which corresponds to nucleotides 244 to 1482 of the NCBI reference sequence NM_001111032.1. The nucleic acid sequence encoding the processed ALK7 polypeptide (isoform 3) is set forth in SEQ ID NO: 238.

[0227] The amino acid sequence of the alternative human ALK7 precursor protein, isoform 4 (NCBI Ref Seq NP_001104503.1) is shown below (SEQ ID NO: 122), where the signal peptide is indicated by a single underline. [ka]

[0228] The amino acid sequence of the processed ALK7 polypeptide (isoform 4) is shown below (SEQ ID NO: 127). Like ALK7 isoform 3, isoform 4 lacks the transmembrane domain and is therefore proposed to be soluble in its entirety (Roberts et al., 2003, Biol Reprod 68:1719-1726). N-terminal variants of SEQ ID NO: 127 are predicted as described below. [ka]

[0229] The nucleic acid sequence encoding the unprocessed ALK7 polypeptide precursor protein (isoform 4) is set forth in SEQ ID NO: 239, which corresponds to nucleotides 244 to 1244 of the NCBI reference sequence NM_001111033.1. The nucleic acid sequence encoding the processed ALK7 polypeptide (isoform 4) is set forth in SEQ ID NO: 240.

[0230] Based on the signal sequence of full-length ALK7 (isoform 1) in rat (see NCBI reference sequence NP_620790.1) and the high degree of sequence identity between human and rat ALK7, the processed form of human ALK7 isoform 1 is predicted to be as follows (sequence number 128): [ka]

[0231] SEQ ID NO: 123 predicts an active variant of processed ALK7 isoform 1 truncated by 1, 2, 3, 4, 5, 6, or 7 amino acids at the N-terminus, and SEQ ID NO: 128 predicts an active variant of processed ALK7 isoform 1 truncated by 1 or 2 amino acids at the N-terminus. Consistent with SEQ ID NO: 128, it is further predicted that leucine is the N-terminal amino acid in the processed forms of human ALK7 isoform 3 (SEQ ID NO: 126) and human ALK7 isoform 4 (SEQ ID NO: 127).

[0232] In certain embodiments, the present disclosure relates to heteromultimers comprising at least one ALK7 polypeptide, including fragments, functional variants, and modified forms thereof. Preferably, the ALK7 polypeptide for use in accordance with the disclosed invention (e.g., heteromultimers comprising ALK7 polypeptides and their uses) is soluble (e.g., the extracellular domain of ALK7). In other preferred embodiments, the ALK7 polypeptides for use in accordance with the disclosed invention bind to one or more TGF-beta superfamily ligands and / or inhibit (antagonize) the activity of such ligands (e.g., induction of Smad2 / 3 and / or Smad1 / 5 / 8 signaling). In some embodiments, the heteromultimers of the present disclosure comprise at least one ALK7 polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 120, 123, 129, 130, 124, 125, 121, 126, 122, 127, 128, 133 or 134. In some embodiments, the heteromultimers of the present disclosure consist of or consist essentially of at least one ALK7 polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 120, 123, 129, 130, 124, 125, 121, 126, 122, 127, 128, 133 or 134.

[0233] In certain aspects, the present disclosure relates to heteromultimers comprising an ALK7-Fc fusion protein. In some embodiments, the ALK7-Fc fusion protein begins with any one of amino acids 21 to 28 (e.g., amino acid residues 21, 22, 23, 24, 25, 26, 27, and 28) of SEQ ID NO: 120, 121, or 122, and continues with any one of amino acids 92 to 113 (e.g., amino acid residues 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 16 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, and 113). In some embodiments, the ALK7-Fc fusion protein comprises an ALK7 domain comprising an amino acid sequence that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 28-92 of SEQ ID NO: 120, 121 or 122. In some embodiments, the ALK7-Fc fusion protein comprises an ALK7 domain comprising an amino acid sequence that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 21-113 of SEQ ID NO: 120, 121 or 122. In some embodiments, the ALK7-Fc fusion protein comprises an ALK7 domain comprising an amino acid sequence that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of any one of SEQ ID NOs: 120, 123, 124, 125, 121, 126, 122, 127, 128, 129, 130, 131, 132, 133, 134, 135 and 136. In some embodiments, the ALK7-Fc fusion protein uses the TPA leader, shown below (SEQ ID NO: 129): [ka]

[0234] The signal sequence and linker sequence are underlined. To promote the formation of the ActRIIB-Fc:ALK7-Fc heterodimer rather than one of the possible homodimeric complexes, two amino acid substitutions (replacing lysine with aspartic acid) can be introduced into the Fc domain of the fusion protein, as shown double underlined above. The amino acid sequence of SEQ ID NO: 129 can optionally be provided with an additional lysine at the C-terminus. This ALK7-Fc fusion protein is encoded by the following nucleic acid (SEQ ID NO:255): [ka] [ka]

[0235] The mature ALK7-Fc fusion protein sequence (SEQ ID NO: 130) is expected to be as follows, optionally provided with an additional C-terminal lysine: [ka] The complementary form of the ALK7-Fc fusion polypeptide (SEQ ID NO: 133) is shown below: [ka]

[0236] The leader and linker sequences are underlined. To induce heterodimer formation by certain Fc fusion polypeptides disclosed herein, four amino acid substitutions can be introduced into the Fc domain of the ALK7 fusion polypeptide, as shown by double underlines above. Furthermore, the C-terminal lysine residue of the Fc domain can be deleted. The amino acid sequence of SEQ ID NO: 133 can optionally be provided by removing the lysine from the C-terminus.

[0237] The mature ALK7-Fc fusion protein sequence (SEQ ID NO: 134) is expected to be as follows, optionally provided with a lysine removed from the C-terminus: [ka]

[0238] In certain embodiments, the present disclosure relates to a protein complex comprising an ActRIIA polypeptide. As used herein, the term "ActRIIA" refers to the family of activin receptor type IIA (ActRIIA) proteins from any species, and variants derived from such ActRIIA proteins by mutagenesis or other modifications. Reference to ActRIIA herein is understood to refer to any one of the currently identified forms. Members of the ActRIIA family are generally transmembrane proteins composed of a ligand-binding extracellular domain containing a cysteine-rich region, a transmembrane domain, and a cytoplasmic domain with predicted serine / threonine kinase activity.

[0239] The term "ActRIIA polypeptide" includes polypeptides including any naturally occurring polypeptide of an ActRIIA family member, as well as any variants thereof (including mutants, fragments, fusions, and peptidomimetic forms) that retain useful activity. Examples of such variant ActRIIA polypeptides are provided throughout this disclosure and in International Patent Application Publication No. WO2006 / 012627, which is incorporated herein by reference in its entirety. The human ActRIIA precursor protein sequence is as follows: [ka] [ka]

[0240] The signal peptide is indicated by a single underline; the extracellular domain is indicated in bold font; potential endogenous N-linked glycosylation sites are indicated by a double underline. The processed extracellular human ActRIIA polypeptide sequence is shown below: [ka] The C-terminal "tail" of the extracellular domain is indicated by a single underline. The sequence with the "tail" deleted (Δ15 sequence) is shown below: [ka]

[0241] The nucleic acid sequence encoding the human ActRIIA precursor protein is set forth in SEQ ID NO: 241, which corresponds to nucleotides 159 to 1700 of Genbank reference sequence NM_001616.4. The nucleic acid sequence encoding the processed extracellular ActRIIA polypeptide is set forth in SEQ ID NO: 242.

[0242] A general formula for an active (e.g., ligand-binding) ActRIIA polypeptide is one that includes a polypeptide beginning at amino acid 30 and ending at amino acid 110 of SEQ ID NO: 137. Accordingly, ActRIIA polypeptides of the present disclosure can include polypeptides that are at least 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99%, or 100% identical to amino acids 30-110 of SEQ ID NO: 137. Optionally, ActRIIA polypeptides of the present disclosure optionally begin at positions spanning 1-5 (e.g., 1, 2, 3, 4, or 5) or 3-5 (e.g., 3, 4, or 5), respectively, and end at positions spanning 110-116 (e.g., 110, 111, 112, 113, 114, 115, or 116) or 110-115 (e.g., 110, 111, 112, 113, 114, or 115), respectively, with respect to SEQ ID NO: 137. and polypeptides that are at least 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 12-82 of SEQ ID NO: 137, including no more than 1, 2, 5, 10 or 15 conservative amino acid changes in the ligand binding pocket and zero, one or more non-conservative changes at positions 40, 53, 55, 74, 79 and / or 82 in the ligand binding pocket.

[0243] In certain embodiments, the present disclosure relates to heteromultimers comprising at least one ActRIIA polypeptide, including fragments, functional variants, and modified forms thereof. Preferably, the ActRIIA polypeptide for use in accordance with the disclosed invention (e.g., heteromultimers comprising ActRIIA polypeptides and their uses) is soluble (e.g., the extracellular domain of ActRIIA). In other preferred embodiments, the ActRIIA polypeptide for use in accordance with the disclosed invention binds to one or more TGF-beta superfamily ligands and / or inhibits (antagonizes) the activity of such ligands (e.g., induction of Smad2 / 3 and / or Smad1 / 5 / 8 signaling). In some embodiments, a heteromultimer of the present disclosure comprises at least one ActRIIA polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of any one of SEQ ID NOs: 137, 138, 139, 140, 141, 144 or 145. In some embodiments, a heteromultimer of the present disclosure comprises at least one ActRIIA polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of any one of SEQ ID NOs: 137, 138, 139, 140, 141, 144 or 145.

[0244] In certain aspects, the present disclosure relates to heteromultimers comprising an ActRIIA-Fc fusion protein. In some embodiments, the ActRIIA-Fc fusion protein begins with any one of amino acids 21 to 30 (e.g., amino acid residues 21, 22, 23, 24, 25, 26, 27, 28, 29, or 30) of SEQ ID NO: 137 and continues with any one of amino acids 110 to 135 (e.g., 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 178, 179, 180, 181, 182, 183 , 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134 or 135). In some embodiments, an ActRIIA-Fc fusion protein comprises an ActRIIA domain comprising an amino acid sequence at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to amino acids 30-110 of SEQ ID NO: 137. In some embodiments, an ActRIIA-Fc fusion protein comprises an ActRIIA domain comprising an amino acid sequence at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to amino acids 21-135 of SEQ ID NO: 137. In some embodiments, the ActRIIA-Fc fusion protein comprises an ActRIIA domain comprising an amino acid sequence that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of any one of SEQ ID NOs: 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, and 147. The ActRIIA-Fc polypeptide sequence (SEQ ID NO: 140) is shown below: [ka]

[0245] The leader and linker sequences are underlined. To promote the formation of the ActRIIA-Fc:ActRIIB-Fc heterodimer rather than one of the possible homodimeric complexes, two amino acid substitutions (replacing acidic amino acids with lysines) can be introduced into the Fc domain of the ActRIIA fusion protein, as shown double underlined above. The amino acid sequence of SEQ ID NO: 140 can optionally be provided with the lysine removed from the C-terminus. This ActRIIA-Fc fusion protein is encoded by the following nucleic acid sequence (SEQ ID NO:256): [ka] [ka]

[0246] The mature ActRIIA-Fc fusion polypeptide (SEQ ID NO: 141) is shown below and can optionally be provided with the lysine removed from the C-terminus. [ka] The ActRIIA-Fc polypeptide sequence (SEQ ID NO: 144) is shown below: [ka]

[0247] The leader and linker sequences are underlined. To promote the formation of the ActRIIA-Fc:ALK4-Fc heterodimer rather than one of the possible homodimeric complexes, two amino acid substitutions (replacing serine with cysteine ​​and threonine with trytophan) can be introduced into the Fc domain of the fusion protein, as shown double underlined above. The amino acid sequence of SEQ ID NO: 144 can optionally be provided with the lysine removed from the C-terminus.

[0248] The mature ActRIIA-Fc fusion polypeptide (SEQ ID NO: 145) is shown below and can optionally be provided with the lysine removed from the C-terminus. [ka]

[0249] In certain aspects, the present disclosure relates to a protein complex comprising a BMPRII polypeptide. As used herein, the term "BMPRII" refers to a family of bone morphogenetic protein receptor type II (BMPRII) proteins from any species, and variants derived from such BMPRII proteins by mutagenesis or other modifications. Reference to BMPRII herein is understood to refer to any one of the currently identified forms. Members of the BMPRII family are generally transmembrane proteins composed of a ligand-binding extracellular domain with a cysteine-rich region, a transmembrane domain, and a cytoplasmic domain with predicted serine / threonine kinase activity.

[0250] The term "BMPRII polypeptide" includes polypeptides including any naturally occurring polypeptide of a BMPRII family member, as well as any variants thereof (including mutants, fragments, fusions, and peptidomimetic forms) that retain useful activity. The human BMPRII precursor protein sequence (NCBI Reference Sequence NP_001195.2) is as follows: [ka] [ka]

[0251] The signal peptide is shown with a single underline and the extracellular domain is shown in bold font. The processed extracellular BMPRII polypeptide sequence is shown below: [ka]

[0252] The nucleic acid sequence encoding the BMPRII precursor protein is set forth in SEQ ID NO: 205, according to nucleotides 1149 to 4262 of Genbank reference sequence NM_001204.6. The nucleic acid sequence encoding the extracellular BMPRII polypeptide is set forth in SEQ ID NO: 206. An alternative isoform of BMPRII, isoform 2 (GenBank: AAA86519.1) is shown below: [ka] [ka]

[0253] The signal peptide is indicated by a single underline and the extracellular domain is indicated in bold font. The processed extracellular BMPRII polypeptide sequence (isoform 2) is shown below: [ka]

[0254] The nucleic acid sequence encoding the human BMPRII precursor protein (isoform 2) is set forth in SEQ ID NO: 207, which corresponds to nucleotides 163-1752 of Genbank reference sequence U25110.1. The signal sequence is underlined. The nucleic acid sequence encoding the extracellular BMPRII polypeptide (isoform 2) is set forth in SEQ ID NO: 208.

[0255] In certain embodiments, the present disclosure relates to heteromultimers comprising at least one BMPRII polypeptide, including fragments, functional variants, and modified forms thereof. Preferably, the BMPRII polypeptide for use in accordance with the disclosed invention (e.g., heteromultimers comprising BMPRII polypeptides and their uses) is soluble (e.g., the extracellular domain of BMPRII). In other preferred embodiments, the BMPRII polypeptide for use in accordance with the disclosed invention binds to one or more TGF-beta superfamily ligands and / or inhibits (antagonizes) their activity (e.g., induction of Smad2 / 3 and / or Smad1 / 5 / 8 signaling). In some embodiments, heteromultimers of the disclosure comprise at least one BMPRII polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 148, 150, 149, 151, 152, 153, 156 or 157. In some embodiments, heteromultimeric complexes of the disclosure consist of or consist essentially of at least one BMPRII polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99% or 100% identical to the amino acid sequence of SEQ ID NO: 148, 150, 149, 151, 152, 153, 156 or 157.

[0256] In certain aspects, the present disclosure relates to heteromultimers comprising a BMPRII-Fc fusion protein. In some embodiments, the BMPRII-Fc fusion protein begins with any one of amino acids 27 to 34 (e.g., amino acid residues 27, 28, 29, 30, 31, 32, 33, and 34) of SEQ ID NO: 148 or 149 and continues with any one of amino acids 123 to 150 (e.g., amino acid residues 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, and 150). In some embodiments, the BMPRII-Fc fusion protein comprises a BMPRII domain comprising an amino acid sequence that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to amino acids 34-123 of SEQ ID NO: 148 or 149. In some embodiments, the BMPRII-Fc fusion protein comprises a BMPRII domain comprising an amino acid sequence that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to amino acids 27-150 of SEQ ID NO: 148 or 149. In some embodiments, the BMPRII-Fc fusion protein comprises a BMPRII domain comprising an amino acid sequence that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of any one of SEQ ID NOs: 148, 150, 149, 151, 152, 153, 154, 155, 156, 157, 158, and 159. The BMPRII-Fc polypeptide sequence (SEQ ID NO: 152) is shown below: [ka]

[0257] The leader and linker sequences are underlined. To promote the formation of the BMPRII-Fc:ActRIIB-Fc heterodimer rather than one of the possible homodimeric complexes, two amino acid substitutions (replacing acidic amino acids with lysines) can be introduced into the Fc domain of the BMPRII-Fc fusion protein, as shown double underlined above. The amino acid sequence of SEQ ID NO: 152 can optionally be provided with the lysine removed from the C-terminus. This BMPRII-Fc fusion protein is encoded by the following nucleic acid sequence (SEQ ID NO:257): [ka]

[0258] The mature BMPRII-Fc fusion polypeptide (SEQ ID NO: 153) is shown below and can optionally be provided with the lysine removed from the C-terminus. [ka] [ka] The BMPRII-Fc polypeptide sequence (SEQ ID NO: 156) is shown below: [ka]

[0259] The leader and linker sequences are underlined. To promote the formation of the BMPRII-Fc:ActRIIB-Fc heterodimer rather than either of the possible homodimeric complexes, two amino acid substitutions (replacing serine with cysteine ​​and threonine with tryptophan) can be introduced into the Fc domain of the fusion protein, as shown double underlined above. The amino acid sequence of SEQ ID NO: 156 can optionally be provided with the lysine removed from the C-terminus.

[0260] The mature BMPRII-Fc fusion polypeptide (SEQ ID NO: 157) is shown below and can optionally be provided with a lysine (K) removed from the C-terminus. [ka]

[0261] In certain aspects, the present disclosure relates to a protein complex comprising a TGFBRII polypeptide. As used herein, the term "TGFBRII" refers to the family of transforming growth factor-beta receptor II (TGFBRII) proteins from any species, as well as variants derived from such proteins by mutagenesis or other modifications. Reference to TGFBRII herein is understood to refer to any one of the currently identified forms. Members of the TGFBRII family are generally transmembrane proteins composed of a ligand-binding extracellular domain with a cysteine-rich region, a transmembrane domain, and a cytoplasmic domain with predicted serine / threonine kinase activity.

[0262] The term "TGFBRII polypeptide" includes polypeptides including any naturally occurring polypeptide of a TGFBRII family member, as well as any variants thereof (including mutants, fragments, fusions, and peptidomimetic forms) that retain useful activity. The human TGFBRII precursor protein sequence (NCBI Reference Sequence NP_003233.4) is as follows: [ka]

[0263] The signal peptide is shown with a single underline and the extracellular domain is shown in bold font. The processed extracellular TGFBRII polypeptide sequence is as follows: [ka]

[0264] The nucleic acid sequence encoding the human TGFBRII precursor protein is set forth in SEQ ID NO: 196, which corresponds to nucleotides 383 to 2083 of Genbank reference sequence NM_003242.5. The nucleic acid sequence encoding the processed extracellular TGFBRII polypeptide is set forth in SEQ ID NO: 197. An alternative isoform of TGFBRII, isoform A (NP_001020018.1), is as follows: [ka]

[0265] The signal peptide is indicated by a single underline and the extracellular domain is indicated in bold font. The processed extracellular TGFBRII polypeptide sequence (isoform A) is as follows: [ka]

[0266] The nucleic acid sequence encoding the TGFBRII precursor protein (isoform A) is set forth in SEQ ID NO: 202, which corresponds to nucleotides 383 to 2158 of Genbank reference sequence NM_001024847.2. The nucleic acid sequence encoding the processed extracellular TGFBRII polypeptide (isoform A) is set forth in SEQ ID NO: 203.

[0267] Any of the above-mentioned TGFβRII isoforms (SEQ ID NOs: 194, 195, 198 and 199) could incorporate a 36 amino acid (SEQ ID NO: 204) insertion between a pair of glutamic acid residues located near the C-terminus of the TGFβRII ECD (positions 151 and 152 of SEQ ID NO: 194; positions 129 and 130 of SEQ ID NO: 195; positions 176 and 177 of SEQ ID NO: 198; or positions 154 and 155 of SEQ ID NO: 199), as occurs naturally in TGFβRII isoform C (Konrad et al., BMC Genomics 8:318, 2007). [ka]

[0268] In certain embodiments, the present disclosure relates to heteromultimers comprising at least one TGFBRII polypeptide, including fragments, functional variants, and modified forms thereof. Preferably, the TGFBRII polypeptides for use in accordance with the disclosed invention (e.g., heteromultimers comprising TGFBRII polypeptides and their uses) are soluble (e.g., the extracellular domain of TGFBRII). In other preferred embodiments, the TGFBRII polypeptides for use in accordance with the disclosed invention bind to one or more TGF-beta superfamily ligands and / or inhibit (antagonize) the activity of such ligands (e.g., induce Smad2 / 3 and / or Smad1 / 5 / 8 signaling). In some embodiments, heteromultimers of the present disclosure comprise at least one TGFBRII polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of SEQ ID NO: 194, 195, 198, or 199, with or without the insertion of SEQ ID NO: 204, as described above. In some embodiments, heteromultimeric complexes of the present disclosure consist of, or consist essentially of, at least one TGFBRII polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of SEQ ID NO: 194, 195, 198, or 199, with or without the insertion of SEQ ID NO: 204.

[0269] In certain aspects, the present disclosure relates to heteromultimers comprising TGFBII-Fc fusion proteins. In some embodiments, the TGFBII-Fc fusion protein begins with any one of amino acids 23-44 (e.g., 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, or 44) of SEQ ID NO: 160 and includes any one of amino acids 168-191 (e.g., 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 310, 311, 312, 313, 314, 315, 316, 317, 318, 319, 320, 321, 322, 323, 324, 325, 326, 327, 328, 329, 330, 331, 332, 333, 334, 335, 336, 337, 338, 339, 440, 441, 442, 443, or 444) of SEQ ID NO: 160. 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190 or 191). In some embodiments, the TGFBRII-Fc fusion protein comprises a TGFBRII domain comprising an amino acid sequence that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 44-168 of SEQ ID NO: 160. In some embodiments, the TGFBRII-Fc fusion protein comprises a TGFBRII domain comprising an amino acid sequence that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 23-191 of SEQ ID NO: 160. In some embodiments, the TGFBRII-Fc fusion protein comprises a TGFBRII domain comprising an amino acid sequence that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of any one of SEQ ID NOs: 161, 162, 160, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178 and 179.In some embodiments, the TGFBII-Fc fusion protein begins with any one of amino acids 23-51 (e.g., 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, and 51) of SEQ ID NO: 161 and continues with any one of amino acids 143-166 (e.g., 143, 144, 145, 146, 147, 148, 149, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 330, 331, 332, 333, 340, 341, 342, 343, 344, 350, 351, 352, 353, 354, 355, 356, 357, 358, 359, 400, 410, 411, 412, 413, 414, 415, 416, 417, 41 9, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, and 166). In some embodiments, the TGFBRII-Fc fusion protein comprises a TGFBRII domain comprising an amino acid sequence that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 51-143 of SEQ ID NO: 161. In some embodiments, the TGFBRII-Fc fusion protein comprises a TGFBRII domain comprising an amino acid sequence that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 23-166 of SEQ ID NO: 161. The human TGFBRII precursor protein sequence (NCBI Ref Seq NP_003233.4) is shown below: [ka]

[0270] The signal peptide is shown with a single underline and the extracellular domain is shown in bold font. The processed extracellular TGFBRII polypeptide sequence is shown below: [ka] An alternative isoform of TGFBRII, isoform A (NP_001020018.1), is shown below: [ka]

[0271] The signal peptide is shown with a single underline and the extracellular domain is shown in bold font. The processed extracellular TGFBRII polypeptide sequence (isoform A) is shown below: [ka] TGFβRII SHORT -Fc polypeptide sequence (SEQ ID NO: 164) is shown below: [ka]

[0272] The leader and linker sequences are underlined. SHORT - To promote the formation of ActRIIB-Fc heterodimers, two amino acid substitutions (replacing acidic amino acids with lysines) were made in TGFβRII as shown double underlined above. SHORT -Fc fusion protein. The amino acid sequence of SEQ ID NO: 164 can be provided with the lysine optionally removed from the C-terminus. This TGFβRII SHORT -Fc fusion protein is encoded by the following nucleic acid sequence (SEQ ID NO: 259): [ka]

[0273] mature TGFβRII SHORTThe -Fc fusion polypeptide (SEQ ID NO: 165) is shown below and can optionally be provided with the lysine removed from the C-terminus. [ka] TGFβRII LONG -Fc polypeptide sequence (SEQ ID NO: 166) is shown below: [ka]

[0274] The leader and linker sequences are underlined. LONG - To promote the formation of ActRIIB-Fc heterodimers, two amino acid substitutions (replacing acidic amino acids with lysines) were made in TGFβRII as shown double underlined above. LONG -Fc fusion protein. The amino acid sequence of SEQ ID NO: 166 can be provided with the lysine optionally removed from the C-terminus. This TGFβRII LONG -Fc fusion protein is encoded by the following nucleic acid sequence (SEQ ID NO: 260): [ka] [ka]

[0275] mature TGFβRII LONG The -Fc fusion polypeptide (SEQ ID NO: 167) is shown below and can optionally be provided with the lysine removed from the C-terminus. [ka]

[0276] In a second approach to promoting heteromultimeric complex formation using asymmetric Fc fusion proteins, the Fc domain is altered to introduce complementary hydrophobic interactions and additional intermolecular disulfide bonds. TGFβRII SHORT -Fc polypeptide sequence (SEQ ID NO: 172) is shown below: [ka]

[0277] The leader and linker sequences are underlined. SHORT -To promote the formation of the ActRIIB-Fc heterodimer, two amino acid substitutions (replacing serine with cysteine ​​and threonine with tryptophan) can be introduced into the Fc domain of the fusion protein, as shown double underlined above. The amino acid sequence of SEQ ID NO: 172 can optionally be provided with the lysine removed from the C-terminus.

[0278] mature TGFβRII SHORT The -Fc fusion polypeptide (SEQ ID NO: 173) is shown below and can optionally be provided with the lysine removed from the C-terminus. [ka]

[0279] To induce heterodimer formation with certain Fc fusion polypeptides disclosed herein, four amino acid substitutions can be introduced into the Fc domain of the ALK1 fusion polypeptide. In some embodiments, TGFβRII LONG -Fc polypeptide sequence (SEQ ID NO: 174) is as shown below: [ka]

[0280] The leader and linker sequences are underlined. LONG -To promote the formation of the ActRIIB-Fc heterodimer, two amino acid substitutions (replacing serine with cysteine ​​and threonine with tryptophan) can be introduced into the Fc domain of the fusion protein, as shown double underlined above. The amino acid sequence of SEQ ID NO: 174 can optionally be provided with the lysine removed from the C-terminus.

[0281] mature TGFβRII LONG The -Fc fusion polypeptide (SEQ ID NO: 175) is shown below and can optionally be provided with the lysine removed from the C-terminus. [ka]

[0282] In certain aspects, the present disclosure relates to a protein complex comprising a MISRII polypeptide. As used herein, the term "MISRII" refers to the family of Müllerian inhibitory factor receptor type II (MISRII) proteins from any species, as well as variants derived from such MISRII proteins by mutagenesis or other modifications. Reference to MISRII herein is understood to refer to any one of the currently identified forms. Members of the MISRII family are generally transmembrane proteins composed of a ligand-binding extracellular domain with a cysteine-rich region, a transmembrane domain, and a cytoplasmic domain with predicted serine / threonine kinase activity.

[0283] The term "MISRII polypeptide" includes polypeptides including any naturally occurring polypeptide of a MISRII family member, as well as any variants thereof (including mutants, fragments, fusions, and peptidomimetic forms) that retain useful activity. The human MISRII precursor protein sequence (NCBI reference sequence NP_065434.1) is as follows: [ka]

[0284] The signal peptide is indicated by a single underline and the extracellular domain is indicated in bold font. The processed extracellular MISRII polypeptide sequence is as follows: [ka]

[0285] The nucleic acid sequence encoding the MISRII precursor protein is set forth in SEQ ID NO: 209, which corresponds to nucleotides 81 to 1799 of Genbank reference sequence NM_020547.2. The nucleic acid sequence encoding the extracellular human MISRII polypeptide is set forth in SEQ ID NO: 210. An alternative isoform of the human MISRII precursor protein sequence, isoform 2 (NCBI reference sequence NP_001158162.1), is as follows: [ka]

[0286] The signal peptide is shown with a single underline and the extracellular domain is shown in bold font. The processed extracellular MISRII polypeptide sequence (isoform 2) is as follows: [ka]

[0287] The nucleic acid sequence encoding the MISRII precursor protein (isoform 2) is set forth in SEQ ID NO:211, which corresponds to nucleotides 81 to 1514 of Genbank reference sequence NM_001164690.1. The nucleic acid sequence encoding the processed soluble (extracellular) human MISRII polypeptide (isoform 2) is set forth in SEQ ID NO:212. An alternative isoform of the human MISRII precursor protein sequence, isoform 3 (NCBI reference sequence NP_001158163.1), is as follows: [ka]

[0288] The signal peptide is shown with a single underline and the extracellular domain is shown in bold font. The processed extracellular MISRII polypeptide sequence (isoform 3) is shown below: [ka]

[0289] The nucleic acid sequence encoding the human MISRII precursor protein (isoform 3) is set forth in SEQ ID NO:213, which corresponds to nucleotides 81 to 1514 of Genbank reference sequence NM_001164691.1. The nucleic acid sequence encoding the processed soluble (extracellular) human MISRII polypeptide (isoform 3) is set forth in SEQ ID NO:214.

[0290] In certain embodiments, the present disclosure relates to heteromultimers comprising at least one MISRII polypeptide, including fragments, functional variants, and modified forms thereof. Preferably, the MISRII polypeptide for use in accordance with the disclosed invention (e.g., heteromultimers comprising MISRII polypeptides and their uses) is soluble (e.g., the extracellular domain of MISRII). In other preferred embodiments, the MISRII polypeptide for use in accordance with the disclosed invention binds to one or more TGF-beta superfamily ligands and / or inhibits (antagonizes) the activity of such ligands (e.g., induction of Smad2 / 3 and / or Smad1 / 5 / 8 signaling). In some embodiments, the heteromultimers of the present disclosure comprise at least one MISRII polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of SEQ ID NO: 180, 183, 184, 182, or 185. In some embodiments, a heteromultimer of the present disclosure consists of, or consists essentially of, at least one MISRII polypeptide that is at least 70%, 75%, 80%, 85%, 90%, 95%, 97%, 98%, 99%, or 100% identical to the amino acid sequence of SEQ ID NO: 180, 183, 184, 182, or 185.

[0291] In certain aspects, the present disclosure relates to heteromultimers comprising MISRII-Fc fusion proteins. In some embodiments, the MISRII-Fc fusion protein begins with any one of amino acids 17-24 (e.g., amino acid residues 17, 18, 19, 20, 21, 22, 23, and 24) of SEQ ID NO: 180, 181, or 182, and continues with any one of amino acids 116-149 (e.g., amino acid residues 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189 140, 141, 142, 143, 144, 145, 146, 147, 148, and 149). In some embodiments, a MISRII-Fc fusion protein comprises a MISRII domain comprising an amino acid sequence that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 24-116 of SEQ ID NO: 180, 181 or 182. In some embodiments, a MISRII-Fc fusion protein comprises a MISRII domain comprising an amino acid sequence that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to amino acids 17-149 of SEQ ID NO: 180, 181 or 182.In some embodiments, the MISRII-Fc fusion protein comprises a MISRII domain comprising an amino acid sequence that is at least 70%, 75%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identical to the amino acid sequence of any one of SEQ ID NOs: 180, 183, 181, 184, 182, 185, 186, 187, 188, 189, 190, 191, 192 and 193.

[0292] In some embodiments, the present disclosure contemplates creating functional variants by modifying the structure of TGF-beta superfamily type I receptor polypeptides (e.g., ALK1, ALK2, ALK3, ALK4, ALK5, ALK6, and ALK7) and / or TGF-beta superfamily type II receptor polypeptides (e.g., ActRIIA, ActRIIB, TGFBRII, BMPRII, and MISRII), such as to enhance therapeutic efficacy or stability (e.g., shelf life and resistance to in vivo proteolysis). Variants can be produced by amino acid substitution, deletion, addition, or a combination thereof. For example, it is reasonable to expect that isolated replacement of leucine with isoleucine or valine, aspartate with glutamate, threonine with serine, or similar replacement of amino acids with structurally related amino acids (e.g., conservative mutations) will not significantly affect the biological activity of the resulting molecule. Conservative substitutions are those that occur within a family of amino acids that are related in their side chains. Whether a change in the amino acid sequence of a polypeptide of the present disclosure results in a functional homologue can be determined by the ability of the variant polypeptide to produce a response in a cell in a manner similar to that of the wild-type polypeptide, or by the ability of the variant polypeptide to produce a response in a cell in a manner similar to that of the wild-type polypeptide, for example, BMP2, BMP2 / 7, BMP3, BMP4, BMP4 / 7, BMP5, BMP6, BMP7, BMP8a, BMP8b, BMP9, BMP10, GDF3, GDF5, GDF6 / BMP13, GDF7, GDF8, GDF9b / BMP15, GDF11 / BMP11, GDF15 / MIC1, This can be readily determined by assessing the ability of the variant polypeptide to bind to one or more TGF-beta superfamily ligands, including TGF-β1, TGF-β2, TGF-β3, activin A, activin B, activin C, activin E, activin AB, activin AC, activin AE, activin BC, activin BE, Nodal, glial cell line-derived neurotrophic factor (GDNF), neurturin, artemin, persephin, MIS, and lefty.

[0293] In some embodiments, the present disclosure contemplates modifying the structure of TGF-beta superfamily type I receptor polypeptides and / or TGF-beta superfamily type II receptor polypeptides to create functional variants, such as to enhance therapeutic efficacy or stability (e.g., prolonged shelf life and / or increased resistance to proteolysis).

[0294] In certain embodiments, the present disclosure contemplates specific mutations of the disclosed TGF-beta superfamily type I receptor polypeptides (e.g., ALK1, ALK2, ALK3, ALK4, ALK5, ALK6, and ALK7) and / or TGF-beta superfamily type II receptor polypeptides (e.g., ActRIIA, ActRIIB, TGFBRII, BMPRII, and MISRII) receptors to alter the glycosylation of the polypeptides. Such mutations can be selected to introduce or eliminate one or more glycosylation sites, such as O-linked or N-linked glycosylation sites. Asparagine-linked glycosylation recognition sites generally comprise asparagine-X-threonine or asparagine-X-serine (where "X" is any amino acid), tripeptide sequences that are specifically recognized by appropriate cellular glycosylation enzymes. Alterations can also be made to the polypeptide sequence by adding one or more serine or threonine residues (for O-linked glycosylation sites) or by substitution with such residues. Substitution or deletion of various amino acids at one or both of the first or third amino acid positions of the glycosylation recognition site (and / or deletion of an amino acid at the second position) results in a non-glycosylated tripeptide sequence. Another means of increasing the number of carbohydrate moieties on a polypeptide is by chemical or enzymatic coupling of glycosides to the polypeptide. Depending on the coupling method used, sugars can be attached to (a) arginine and histidine; (b) free carboxyl groups; (c) free sulfhydryl groups, such as those of cysteine; (d) free hydroxyl groups, such as those of serine, threonine, or hydroxyproline; (e) aromatic residues, such as those of phenylalanine, tyrosine, or tryptophan; or (f) the amide group of glutamine. Removal of one or more carbohydrate moieties present on the polypeptide can be accomplished chemically and / or enzymatically. Chemical deglycosylation can involve, for example, exposure of the polypeptide to the compound trifluoromethanesulfonic acid, or an equivalent compound.This treatment results in the cleavage of most or all sugars except the linking sugar (N-acetylglucosamine or N-acetylgalactosamine), while leaving the amino acid sequence intact. Enzymatic cleavage of carbohydrate moieties on polypeptides can be achieved using a variety of endo- and exoglycosidases, such as those described by Thotakura et al. [Meth. Enzymol. (1987) 138:350]. Because mammalian, yeast, insect, and plant cells can all introduce different glycosylation patterns that can be influenced by the amino acid sequence of the peptide, the sequence of the polypeptide can be tailored as needed depending on the type of expression system used. Generally, the TGF-beta superfamily type I and type II receptor complexes of the present disclosure for use in humans can be expressed in mammalian cell lines, such as HEK293 or CHO cell lines, that provide appropriate glycosylation, although other mammalian expression cell lines are expected to be useful.

[0295] The present disclosure further contemplates methods for generating mutants, particularly combinatorial mutant sets of the TGF-beta superfamily type I receptor polypeptides (e.g., ALK1, ALK2, ALK3, ALK4, ALK5, ALK6, and ALK7) and / or TGF-beta superfamily type II receptor polypeptides (e.g., ActRIIA, ActRIIB, TGFBRII, BMPRII, and MISRII) disclosed herein, as well as truncation mutants. Pools of combinatorial mutants are particularly useful for identifying functionally active (e.g., ligand-binding) TGF-beta superfamily type I and / or TGF-beta superfamily type II receptor sequences. The purpose of screening such combinatorial libraries can be to generate polypeptide variants with altered properties, such as, for example, altered pharmacokinetics or altered ligand binding. Various screening assays are provided below, and such assays can be used to evaluate variants. For example, TGF-beta superfamily type I and II receptor complex variants can be associated with TGF-beta superfamily ligands (e.g., BMP2, BMP2 / 7, BMP3, BMP4, BMP4 / 7, BMP5, BMP6, BMP7, BMP8a, BMP8b, BMP9, BMP10, GDF3, GDF5, GDF6 / BMP13, GDF7, GDF8, GDF9b / BMP15, GDF11 / BMP11, GDF15 / MIC1, TGF-β1, TGF-β2, TGF-β3, activin A, activin B, activin C, and the like). , activin E, activin AB, activin AC, activin AE, activin BC, activin BE, Nodal, glial cell line-derived neurotrophic factor (GDNF), neurturin, artemin, persephin, MIS, and lefty), ability to prevent binding of TGF-beta superfamily ligands to TGF-beta superfamily receptors, and / or ability to interfere with signal transduction caused by TGF-beta superfamily ligands.

[0296] The activity of the TGF-beta superfamily heteromultimers of the present disclosure can also be tested, for example, in cell-based in vivo assays. For example, the effect of the heteromultimeric complex on the expression of genes or the activity of proteins involved in muscle production in muscle cells can be assessed. This can be done, if desired, by injecting one or more recombinant TGF-beta superfamily ligand proteins (e.g., BMP2, BMP2 / 7, BMP3, BMP4, BMP4 / 7, BMP5, BMP6, BMP7, BMP8a, BMP8b, BMP9, BMP10, GDF3, GDF5, GDF6 / BMP13, GDF7, GDF8, GDF9b / BMP15, GDF11 / BMP11, GDF15 / MIC1, TGF-β1, TGF-β2, TGF-β3, activin A, activin B, activin C, activin D, activin E, activin F, activin G, activin H, activin I ... The heteromultimer complexes of the present disclosure can be administered to mice or other animals, and one or more measures, such as muscle formation and strength, can be assessed using art-recognized methods. Similarly, the activity of the heteromultimers or variants thereof can be tested in osteoblasts, adipocytes, and / or neurons for any effect on the proliferation of these cells, for example, by assays described herein and known in the art. SMAD-responsive reporter genes can be used in such cell lines to monitor effects on downstream signal transduction.

[0297] Combinatorially derived variants can be generated that have increased selectivity or generally increased potency compared to a reference TGF-beta superfamily heteromultimer. Such variants, when expressed from recombinant DNA constructs, can be used in gene therapy protocols. Similarly, mutagenesis can result in variants with intracellular half-lives that are dramatically different from the corresponding unmodified TGF-beta superfamily heteromultimer. For example, the altered protein can be made more stable or less stable to proteolysis or other cellular processes that would cause the unmodified polypeptide to break down or otherwise inactivate. Such variants, and the genes encoding them, can be used to alter polypeptide complex levels by modulating the polypeptide's half-life. For example, a shorter half-life can result in a shorter-term biological effect and, when part of an inducible expression system, can allow for tighter control of recombinant polypeptide complex levels within cells. In Fc fusion proteins, mutations can be made in the linker (if present) and / or the Fc portion to alter one or more activities of the TGF-beta superfamily heteromultimer complex, including, for example, immunogenicity, half-life, and solubility.

[0298] Many methods known in the art can be used to generate the heteromultimer of the present disclosure.For example, in the first polypeptide (for example, variant ActRIIB polypeptide), the naturally occurring amino acid can be replaced with a free thiol-containing residue such as cysteine, so that the free thiol interacts with another free thiol-containing residue in the second polypeptide (for example, unmodified ActRIIB polypeptide or variant ActRIIB polypeptide that is different from that present in the first polypeptide), so that disulfide bond is formed between the first and second polypeptides. Further examples of interactions for promoting heteromultimer formation include, but are not limited to, ionic interactions such as those described in Kjaergaard et al., WO2007147901; electrostatic steering effects such as those described in Kannan et al., US8,592,562; coiled-coil interactions such as those described in Christensen et al., US20120302737; leucine zippers such as those described in Pack and Plueckthun, (1992) Biochemistry 31:1579-1584; and helix-turn-helix motifs such as those described in Pack et al. (1993) Bio / Technology 11:1271-1277. Linkage of the various segments can be achieved by covalent bonds, such as by chemical crosslinking, peptide linkers, disulfide bridges, etc., or by affinity interactions such as by avidin-biotin or leucine zipper technology.

[0299] In certain aspects, the multimerization domain can comprise one component of an interacting pair. In some embodiments, the polypeptides disclosed herein can form a protein complex comprising a first polypeptide covalently or noncovalently associated with a second polypeptide, wherein the first polypeptide comprises the amino acid sequence of a variant ActRIIB polypeptide and the amino acid sequence of a first member of the interacting pair; and the second polypeptide comprises an amino acid sequence of a variant ActRIIB polypeptide that differs from that present in the unmodified ActRIIB polypeptide or the first polypeptide, and the amino acid sequence of a second member of the interacting pair. An interacting pair can be any two polypeptide sequences that interact to form a complex, particularly a heterodimeric complex, although useful embodiments can also employ an interacting pair that can form a homodimeric complex. The interacting pair can be selected to confer improved properties / activities, such as increased serum half-life, or to act as an adapter to which another moiety can be attached to provide improved properties / activities. For example, a polyethylene glycol moiety can be attached to one or both components of the interacting pair to provide improved properties / activities, such as increased serum half-life.

[0300] The first and second members of the interaction pair may be an asymmetric pair, meaning that the members of the pair preferentially associate with each other rather than self-associating. Thus, the first and second members of the asymmetric interaction pair may associate to form a heterodimeric complex (see, for example, Figure 1B). Alternatively, the interaction pair may be uninduced, meaning that the members of the pair may associate with each other or self-associate without substantial preference, and thus may have the same or different amino acid sequences (see, for example, Figure 1A). Thus, the first and second members of the uninduced interaction pair may associate to form a homodimeric complex or a heterodimeric complex. Optionally, the first member of the interaction pair (e.g., an asymmetric pair or an uninduced interaction pair) is covalently associated with the second member of the interaction pair. Optionally, the first member of the interaction pair (e.g., an asymmetric pair or an uninduced interaction pair) is non-covalently associated with the second member of the interaction pair.

[0301] As a specific example, the present disclosure provides a fusion protein comprising a variant ActRIIB polypeptide or an unmodified ActRIIB polypeptide fused to a polypeptide comprising an immunoglobulin constant domain, such as the CH1, CH2, or CH3 domain of an immunoglobulin or Fc domain. Fc domains derived from human IgG1, IgG2, IgG3, and IgG4 are provided herein. Other mutations that reduce either CDC or ADCC activity are known, and collectively, any of these variants are included in the present disclosure and can be used as advantageous components of the heteromultimers of the present disclosure. Optionally, the IgG1 Fc domain of SEQ ID NO: 13 has one or more mutations at residues such as Asp-265, Lys-322, and Asn-434 (numbered according to the corresponding full-length IgG1). In certain cases, variant Fc domains with one or more of these mutations (e.g., an Asp-265 mutation) have reduced ability to bind to Fcγ receptors compared to wild-type Fc domains. In other cases, variant Fc domains with one or more of these mutations (e.g., an Asn-434 mutation) have increased ability to bind to MHC class I-associated Fc receptors (FcRNs) compared to wild-type Fc domains.

[0302] An example of a naturally occurring amino acid sequence that can be used for the Fc portion of human IgG1 (G1Fc) is shown below (SEQ ID NO: 13). The dotted underline indicates the hinge region, and the solid underline indicates positions with naturally occurring variants. In part, the present disclosure provides polypeptides comprising, consisting of, or consisting essentially of an amino acid sequence having 70%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% identity to SEQ ID NO: 13. Naturally occurring variants in G1Fc would include E134D and M136L according to the numbering system used in SEQ ID NO: 13 (see Uniprot P01857). [ka]

[0303] An example of a naturally occurring amino acid sequence that can be used for the Fc portion of human IgG2 (G2Fc) is shown below (SEQ ID NO: 14). The dotted underline indicates the hinge region, and the double underline indicates positions where the sequence has a database conflict (according to UniProt P01859). In part, the disclosure provides polypeptides comprising, consisting of, or consisting essentially of an amino acid sequence having 70%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identity to SEQ ID NO: 14. [ka]

[0304] Two examples of amino acid sequences that can be used for the Fc portion of human IgG3 (G3Fc) are shown below. The hinge region in G3Fc can be up to four times longer than in other Fc chains and contains three identical 15-residue segments preceded by a similar 17-residue segment. The first G3Fc sequence shown below (SEQ ID NO: 15) contains a short hinge region consisting of a single 15-residue segment, while the second G3Fc sequence (SEQ ID NO: 16) contains a full-length hinge region. In each case, dotted underlining indicates the hinge region and solid underlining indicates positions with naturally occurring variants, according to UniProt P01859. In part, the disclosure provides polypeptides comprising, consisting of, or consisting essentially of an amino acid sequence having 70%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identity to SEQ ID NOs: 15 and 16. [ka]

[0305] Naturally occurring variants in G3Fc (see, e.g., Uniprot P01860), when converted to the numbering system used in SEQ ID NO: 15, include E68Q, P76L, E79Q, Y81F, D97N, N100D, T124A, S169N, S169del, and F221Y, and the present disclosure provides fusion proteins comprising a G3Fc domain containing one or more of these mutations. Additionally, the human immunoglobulin IgG3 gene (IGHG3) exhibits structural polymorphisms characterized by different hinge lengths [see Uniprot P01859]. Specifically, variant WIS lacks most of the V region and all of the CH1 region. Variant WIS has an additional interchain disulfide bond at position 7 in addition to the 11 normally present in the hinge region. Variant ZUC lacks most of the V region, all of the CH1 region, and part of the hinge. The variant OMMs may represent allelic forms or may represent different gamma chain subclasses. The present disclosure provides additional fusion proteins comprising a G3Fc domain containing one or more of these variants.

[0306] An example of a naturally occurring amino acid sequence that can be used for the Fc portion of human IgG4 (G4Fc) is shown below (SEQ ID NO: 17). The dotted underline indicates the hinge region. In part, the disclosure provides polypeptides comprising, consisting of, or consisting essentially of an amino acid sequence having 70%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identity to SEQ ID NO: 17. [ka]

[0307] Various engineered mutations in the Fc domain are presented herein with respect to the G1Fc sequence (SEQ ID NO: 13), and similar mutations in G2Fc, G3Fc, and G4Fc can be derived from their alignment with G1Fc in Figure 4. Due to unequal hinge lengths, similar Fc positions based on the isotype alignment (Figure 4) have different amino acid numbers in SEQ ID NOs: 13, 14, 15, and 17. Hinge, C H 2 and C H A given amino acid position in a three-region immunoglobulin sequence (e.g., SEQ ID NO: 13, 14, 15, or 17) is numbered as in the Uniprot database (C H 1. Hinge, C H 2 and C H It will be understood that the same position will be identified by a different ...

Claims

1. 1. A fusion protein comprising a variant ActRIIB polypeptide comprising an amino acid sequence that is at least 94% identical to an amino acid sequence beginning with any one of amino acids 20-29 of SEQ ID NO:2 and ending with any one of amino acids 109-134 of SEQ ID NO:2, wherein the polypeptide comprises a lysine (Lys) at the amino acid position corresponding to 82 of SEQ ID NO:2, When administered to a patient, (i) inhibiting activin A and GDF11; and (ii) a pharmaceutical preparation that exhibits weaker inhibition of BMP9 compared to a pharmaceutical preparation having a wild-type ActRIIB extracellular domain polypeptide.

2. The pharmaceutical preparation of claim 1 further for treating anemia.

3. 10. The pharmaceutical preparation of claim 1, further for increasing red blood cell levels or hemoglobin levels.

4. The pharmaceutical preparation according to claim 1, further for increasing muscle mass and / or strength.

5. The pharmaceutical preparation of claim 1 further for treating a muscle-related disorder.

6. The pharmaceutical preparation according to claim 5 , wherein the disorder is associated with undesirable low muscle growth and / or muscle wasting.

7. 6. The pharmaceutical preparation of claim 5, wherein the disorder is muscle atrophy, muscular dystrophy, amyotrophic lateral sclerosis (ALS) or a muscle wasting disorder.

8. 6. The pharmaceutical preparation of claim 5, wherein the disorder is Duchenne muscular dystrophy (DMD), Becker muscular dystrophy (BMD), Emily-Dreifuss muscular dystrophy (EDMD), limb-girdle muscular dystrophy (LGMD), facioscapulohumeral muscular dystrophy (FSH or FSHD) (also known as Landauzy-Dejerine type), myotonic dystrophy (MMD; also known as Steinert disease), oculopharyngeal muscular dystrophy (OPMD), distal muscular dystrophy (DD), or congenital muscular dystrophy (CMD).

9. 6. The pharmaceutical preparation of claim 5, wherein the disorder is cachexia, anorexia, AIDS wasting syndrome, a neuromuscular disease, a motor neuron disease, a disease of the neuromuscular junction, or an inflammatory myopathy.

10. 10. The pharmaceutical preparation of claim 1, further comprising a pharma- ceutically acceptable carrier.

11. 10. The pharmaceutical preparation of claim 1, wherein the fusion protein further comprises an Fc polypeptide domain.

12. 12. The pharmaceutical preparation of claim 11, wherein the fusion protein further comprises a linker positioned between said variant ActRIIB polypeptide and said Fc polypeptide domain.

13. The pharmaceutical preparation according to claim 12, wherein the linker is composed of an amino acid sequence of any one of SEQ ID NOs: 261 to 267.

14. A recombinant nucleic acid molecule encoding the fusion protein of claim 1.

15. 15. The recombinant nucleic acid molecule of claim 14, comprising the nucleic acid sequence of SEQ ID NO:

41.

16. A vector comprising a recombinant nucleic acid molecule according to claim 14 or 15.

17. A host cell comprising a recombinant nucleic acid molecule according to claim 14 or 15.

18. 18. The host cell of claim 17, wherein the host cell is a COS cell or a CHO cell.

19. 17. A pharmaceutical preparation comprising the vector of claim 16 and a pharma- ceutically acceptable carrier.

20. 20. The pharmaceutical preparation according to claim 19 for the treatment of cardiovascular diseases.

Images