EA045361B1 - ANTIBODIES AGAINST CD47 AND THEIR USE FOR THE TREATMENT OF ONCOLOGICAL DISEASES - Google Patents

Description

Cross reference to related applications

This application claims priority to US Provisional Application No. 62/593,712, filed December 1, 2017, which is incorporated herein in its entirety and for all purposes by reference.

Field of invention

This invention relates to the field of treatment of oncological diseases based on the use of antibodies. In particular, the present invention relates to novel humanized anti-CD47 antibodies and antigen-binding fragments thereof or conjugates thereof, which may optionally be associated with a removable screening agent, and their use for the treatment of cancers characterized by CD47 expression.

State of the art

Cluster of differentiation 47 (CD47), also known as integrin associated protein (IAP), is a transmembrane receptor belonging to the immunoglobulin protein superfamily. CD47 is ubiquitously expressed on cells and serves as a self-recognition marker, preventing phagocytosis, serving as a don't-eat-me signal. CD47 mediates its effects through interactions with several other proteins, including thrombospondin (TSP) and signal regulatory protein alpha (SIRPa). The interaction between SIRPa on phagocytic cells and CD47 on target cells helps ensure that target cells are not engulfed.

Some types of cancer utilize a CD47-based cell immune evasion mechanism by increasing the expression of CD47 on the cell surface of the cancer cell, thereby evading removal by the immune system. However, treatments known in the art that target CD47 expressing cells in a subject target both cancer cells and non-cancerous cells, resulting in toxicity in the subject, such as a decrease in the number of red blood cells and platelets in the peripheral blood . Therefore, there is a need for compositions and methods for selectively targeting CD47 in cancer cells without targeting non-cancerous cells.

Brief description of the invention

This disclosure is based on the discovery of new humanized antibodies against CD47 and antigen binding fragments thereof. Some aspects of the invention provide humanized anti-CD47 antibodies or antigen-binding fragments thereof that contain a removable screening agent (eg, a coiled-coil shielding agent) that prevents the anti-CD47 antibodies or antigen-binding fragments thereof from binding to the CD47 protein. In some embodiments of the present invention, the screening agent can be removed (eg, cleaved) by one or more molecules (eg, proteases) that are present in the environment of cancer cells. Removal of the shielding agent restores the ability of anti-CD47 antibodies or antigen-binding fragments thereof to bind CD47, thereby allowing anti-CD47 antibodies or antigen-binding fragments thereof to specifically target the CD47 protein in the context of cancer cells.

Some embodiments of the present invention provide a humanized antibody or antigen binding fragment thereof that specifically binds human CD47, wherein the antibody or antigen binding fragment comprises a light chain variable region and a heavy chain variable region, wherein said heavy chain variable region comprises HCDR1 selected from SEQ ID NO: 16, 19, 21 and 23; HCDR2 selected from SEQ ID NO: 17, 20, 22 and 24; and HCDR3 SEQ ID NO: 18; wherein said light chain variable region comprises LCDR1 selected from SEQ ID NOs: 31 and 34; LCDR2 selected from SEQ ID NO: 32 and 35; and LCDR3 selected from SEQ ID NOs: 33 and 36; wherein said heavy chain variable region contains an amino acid sequence with at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity to the selected amino acid sequence from SEQ ID NO: 2, 3, 4, 5, 6, 7 and 8; wherein said light chain variable region contains an amino acid sequence with an identity of at least 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% relative to the amino acid sequence selected from SEQ ID NO: 10, 11, 12, 13, 14 and 15.

Some embodiments of the present invention provide a humanized antibody or antigen binding fragment thereof that specifically binds human CD47, wherein the antibody or antigen binding fragment comprises a light chain variable region and a heavy chain variable region, wherein said heavy chain variable region comprises HCDR1 selected from SEQ ID NO: 16, 19, 21 and 23; HCDR2 selected from SEQ ID NO: 17, 20, 22 and 24; and HCDR3 SEQ ID NO: 18; wherein said light chain variable region comprises LCDR1 selected from SEQ ID NOs: 31 and 34; LCDR2 selected from SEQ ID NO: 32 and 35; and LCDR3 selected from SEQ ID NOs: 33 and 36; wherein said heavy chain variable region contains:

the human IGHV3-23/HJ4 framework sequence shown in SEQ ID NO: 88, wherein the framework sequence positions H44, H49, H82, H89, H91 and H94, according to Kabat numbering, represent donor residues; or

- 1045361 human IGHV3-48/HJ4 framework sequence shown in SEQ ID NO: 89, wherein the position of the H49 framework sequence, according to Kabat numbering, represents a donor residue; or the human IGHV3-66/HJ4 framework sequence shown in SEQ ID NO: 90, wherein the framework sequence positions H29, H49, and H82, according to Kabat numbering, represent donor residues; or the human IGHV3-74/HJ4 framework sequence shown in SEQ ID NO: 91, wherein the framework sequence position H49, according to Kabat numbering, represents a donor residue; and wherein said light chain variable region comprises:

the human IGKV6-21/KJ2 framework sequence shown in SEQ ID NO: 92, wherein the framework sequence positions L4, L21, L69 and L85, according to Kabat numbering, represent donor residues; or the human IGKV1-27/KJ2 framework sequence shown in SEQ ID NO: 93, wherein framework sequence positions L21, L49 and L69, according to Kabat numbering, represent donor residues.

In some embodiments of the present invention, the heavy chain variable region comprises HCDR1, HCDR2 and HCDR3 selected from: SEQ ID NO: 16, 17 and 18; SEQ ID NO: 19, 20 and 18; SEQ ID NO: 21, 22 and 18; SEQ ID NO: 16, 20 and 18; and SEQ ID NOs: 23, 24 and 18. In some embodiments of the present invention, the light chain variable region comprises LCDR1, LCDR2 and LCDR3 selected from: 31, 32 and 33; SEQ ID NO: 31, 32 and 36; and SEQ ID NOs: 34, 35 and 33. In some embodiments of the present invention, the antibody or antigen binding fragment thereof comprises HCDR1, HCDR2,

HCDR3, LCDR1, LCDR2 and LCDR3 selected from SEQ ID NO: 16, 17, 18, 31, 32

18, 34, 35

18, 31, 32

18, 34, 35

18, 31, 32

18, 31, 32 and 33; SEQ ID NO and 33; SEQ ID NO and 33; SEQ ID NO and 36; SEQ ID NO

19, 20, 18, 31, 32

21, 22, 18, 34, 35

23, 24, 18, 31, 32

19, 20, 18, 31, 32 and and and and 36; and SEQ ID NOs: 23, 24, 18, 31, 32 and 36.

33; SEQ ID NO

33; SEQ ID NO

33; SEQ ID NO

36; SEQ ID NO

19, 20, 18, 34, 35

16, 20, 18, 31, 32

23, 24, 18, 34, 35

21, 22, 18, 31, 32 and and and and

33; SEQ ID NO: 33; SEQ ID NO: 33; SEQ ID NO: 33; SEQ ID NO: 36; SEQ ID NO:

16, 17,

21, 22,

16, 20,

16, 17,

16, 20,

Some embodiments of the present invention provide a humanized antibody or antigen binding fragment thereof that specifically binds human CD47, wherein the antibody or antigen binding fragment comprises a light chain variable region and a heavy chain variable region, wherein said heavy chain variable region comprises HCDR1 selected from SEQ ID NO: 25, 28 and 29; HCDR2 selected from SEQ ID NO: 26 and 30; and HCDR3 SEQ ID NO: 27; and wherein said light chain variable region comprises LCDR1 selected from SEQ ID NOs: 37 and 40; and LCDR2 SEQ ID NO: 38; and LCDR3 selected from SEQ ID NO: 39 and 41; wherein said heavy chain variable region contains an amino acid sequence with at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity to the selected amino acid sequence from SEQ ID NO: 2, 3, 4, 5, 6, 7 and 8; wherein said light chain variable region contains an amino acid sequence with an identity of at least 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% relative to the amino acid sequence selected from SEQ ID NO: 10, 11, 12, 13, 14 and 15.

Some embodiments of the present invention provide a humanized antibody or antigen binding fragment thereof that specifically binds human CD47, wherein the antibody or antigen binding fragment comprises a light chain variable region and a heavy chain variable region, wherein said heavy chain variable region comprises HCDR1 selected from SEQ ID NO: 25, 28 and 29; HCDR2 selected from SEQ ID NO: 26 and 30; and HCDR3 SEQ ID NO: 27; and wherein said light chain variable region comprises LCDR1 selected from SEQ ID NOs: 37 and 40; and LCDR2 SEQ ID NO: 38; and LCDR3 selected from SEQ ID NO: 39 and 41; wherein said heavy chain variable region contains:

the human IGHV3-23/HJ4 framework sequence shown in SEQ ID NO: 88, wherein the framework sequence positions H44, H49, H82, H89, H91 and H94, according to Kabat numbering, represent donor residues; or the human IGHV3-48/HJ4 framework sequence shown in SEQ ID NO: 89, wherein the position of the H49 framework sequence, according to Kabat numbering, represents a donor residue; or the human IGHV3-66/HJ4 framework sequence shown in SEQ ID NO: 90, wherein the framework sequence positions H29, H49, and H82, according to Kabat numbering, represent donor residues; or the human IGHV3-74/HJ4 framework sequence shown in SEQ ID NO: 91, wherein the framework sequence position H49, according to Kabat numbering, represents a donor residue; and wherein said light chain variable region comprises:

human IGKV6-21/KJ2 framework sequence shown in SEQ ID NO: 92, with

- 2 045361 In this case, the positions of the framework sequence L4, L21, L69 and L85, in accordance with Kabat numbering, represent donor residues; or the human IGKV1-27/KJ2 framework sequence shown in SEQ ID NO: 93, wherein framework sequence positions L21, L49 and L69, according to Kabat numbering, represent donor residues.

In some embodiments of the present invention, the heavy chain variable region comprises HCDR1, HCDR2, and HCDR3 selected from: SEQ ID NO: 25, 26, and 27; SEQ ID NO: 28, 26 and 27; SEQ

ID NO: 29, 30 and 27; and SEQ ID NOs: 29, 26, and 27. In some embodiments of the present invention, the light chain variable region comprises LCDR1, LCDR2, and LCDR3 selected from: SEQ ID NOs: 37, 38, and 39; SEQ ID NO: 40, 38 and 39; and SEQ ID NOs: 37, 38, and 41. In some embodiments of the invention, the antibody or antigen binding fragment thereof comprises HCDR1, HCDR2, HCDR3, LCDR1, LCDR2, and LCDR3 selected from SEQ ID NOs: 25, 26, 27, 37, 38 and 39; SEQ ID NO: 25, 26, 27, 40, 38 and 39;

SEQ ID NO: 25, 26, 27, 37, 38 and 41; SEQ ID NO: 28, 26, 27, 37, 38 and 39; SEQ ID NO: 28, 26, 27, 40, 38 and 39;

SEQ ID NO: 28, 26, 27, 37, 38 and 41; SEQ ID NO: 29, 30, 27, 37, 38 and 39; SEQ ID NO: 29, 30, 27, 40, 38 and 39;

SEQ ID NO: 29, 30, 27, 37, 38 and 41; SEQ ID NO: 29, 26, 27, 37, 38 and 39; SEQ ID NO: 29, 26, 27, 40, 38 and 39;

and SEQ ID NOs: 29, 26, 27, 37, 38 and 41.

In some embodiments of the present invention, the heavy chain variable region comprises an amino acid sequence selected from SEQ ID NOs: 2, 3, 4, 5, 6, 7, and 8. In some embodiments of the present invention, the light chain variable region comprises an amino acid sequence selected from SEQ ID NOs: 10, 11, 12, 13, 14 and 15. In some embodiments of the present invention, the heavy chain variable region and the light chain variable region comprise SEQ ID NOs: 2 and 10; SEQ ID NO: 3 and 11; SEQ ID NO: 3 and 12; SEQ ID NO: 3 and 13; SEQ ID NO: 3 and 14; SEQ ID NO:4 and 11; SEQ ID NO: 4 and 12; SEQ ID NO: 4 and 13; SEQ ID NO: 4 and 14; SEQ ID NO: 5 and 11; SEQ ID NO: 5 and 12; SEQ ID NO: 5 and 13; SEQ ID NO: 5 and 14; SEQ ID NO: 6 and 11; SEQ ID NO: 6 and 12; SEQ ID NO: 6 and 13; SEQ ID NO: 6 and 14; SEQ ID NO: 7 and 11; SEQ ID NO: 7 and 12; SEQ ID NO: 7 and 13; SEQ ID NO: 7 and 14; SEQ ID NO: 8 and 11; SEQ ID NO: 8 and 12; SEQ ID NO: 8 and 13; SEQ ID NO: 8 and 14; SEQ ID NO: 3 and 15.

Some embodiments of the present invention provide a humanized antibody or antigen binding fragment thereof that specifically binds human CD47, wherein the antibody or antigen binding fragment comprises a light chain variable region and a heavy chain variable region, wherein said heavy chain variable region comprises HCDR1 SEQ ID NO: 16, HCDR2 SEQ ID NO: 17, and HCDR3 SEQ ID NO: 18; and wherein said light chain variable region comprises LCDR1 SEQ ID NO: 31, LCDR2 SEQ ID NO: 32, and LCDR3 SEQ ID NO: 33; and wherein said heavy chain variable region comprises an amino acid sequence with at least 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identity to SEQ ID NO: 3, and the light chain variable region contains an amino acid sequence of at least 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity to SEQ ID NO: 13; and the antibody has reduced hemagglutination of red blood cells compared to Ab47. In some embodiments of the present invention, the heavy chain variable region contains the amino acid sequence of SEQ ID NO: 3, and the light chain variable region contains the amino acid sequence of SEQ ID NO: 13.

In some embodiments of a humanized antibody or antigen binding fragment thereof provided herein, the heavy chain variable region comprises:

the human IGHV3-23/HJ4 framework sequence shown in SEQ ID NO: 88, wherein the framework sequence positions H44, H49, H82, H89, H91 and H94, according to Kabat numbering, represent donor residues; or the human IGHV3-48/HJ4 framework sequence shown in SEQ ID NO: 89, wherein the position of the H49 framework sequence, according to Kabat numbering, represents a donor residue; or the human IGHV3-66/HJ4 framework sequence shown in SEQ ID NO: 90, wherein the framework sequence positions H29, H49, and H82, according to Kabat numbering, represent donor residues; or the human IGHV3-74/HJ4 framework sequence shown in SEQ ID NO: 91, wherein the framework sequence position H49, according to Kabat numbering, represents a donor residue.

In some such embodiments, H29 is F, H44 is R or G, H49 is A, H82 is M or I, H89 is I or V, H91 is F or Y, and H94 is R , according to the numbering according to Kabat.

In some embodiments of a humanized antibody or antigen binding fragment thereof provided herein, the light chain variable region comprises:

human IGKV6-21/KJ2 framework sequence shown in SEQ ID NO: 92, with framework sequence positions L4, L21, L69 and L85, according to Ka numbering

- 3,045,361 baht, representing the remains of the donor; or the human IGKV1-27/KJ2 framework sequence shown in SEQ ID NO: 93, wherein framework sequence positions L21, L49 and L69, according to Kabat numbering, represent donor residues.

In some such embodiments of the present invention, L4 is M, L21 is L, L49 is K, L69 is T or S, and L85 is V or T, according to Kabat numbering.

In some embodiments of the present invention, the antibody or antigen-binding fragment thereof is an IgG1 isotype. In some embodiments of the present invention, the antibody or antigen-binding fragment thereof has increased antibody-dependent cellular cytotoxicity (ADCC) compared to its parent antibody. In some embodiments of the present invention, the antibody or antigen-binding fragment thereof exhibits increased antibody-dependent cellular phagocytosis (ADCP) compared to its parent antibody. In some embodiments of the present invention, the antibody or antigen-binding fragment thereof has increased complement-dependent cytotoxicity (CDC) compared to its parent antibody. In some embodiments of the present invention, the antibody or antigen binding fragment thereof is a Fab, Fab', F(ab') ₂ , Fv fragment, diabody, single chain antibody, scFv fragment, or scFv-Fc. In some embodiments of the present invention, the antibody or antigen-binding fragment thereof induces apoptosis of cells expressing CD47 in vitro and/or in vivo. In some embodiments of the present invention, the antibody or antigen binding fragment thereof has reduced backbone fucosylation compared to its parent antibody. In some embodiments of the present invention, the antibody or antigen-binding fragment thereof is afucosylated. In some embodiments of the present invention, the antibody or antigen-binding fragment thereof blocks the interaction between CD47 and SIRPa. In some embodiments of the present invention, the antibody or antigen-binding fragment thereof has reduced erythrocyte hemagglutination compared to Ab47.

In some embodiments, the present invention provides a nucleic acid sequence that encodes an antibody or antigen binding fragment provided herein. In some embodiments of the present invention, an expression vector is provided that contains the specified nucleic acid sequence. In some embodiments, the present invention provides a host cell that contains said nucleic acid or expression vector. In some embodiments, the present invention provides a host cell that expresses an antibody or antigen binding fragment provided herein. In some embodiments of the present invention, a method for producing an antibody or antigen-binding fragment provided herein involves culturing a host cell. In some embodiments of the present invention, the method further comprises isolating the antibody or antigen-binding fragment thereof.

Some embodiments of the present invention provide methods for treating a cancer characterized by CD47 expression in a subject, comprising administering to the subject a therapeutically effective amount of an anti-CD47 antibody or antigen-binding fragment thereof provided herein.

Some embodiments of the present invention provide methods for treating a cancer characterized by CD47 expression in a subject, comprising:

identifying the subject as having a cancer that is characterized by CD47 expression; and administering to the subject a therapeutically effective amount of an anti-CD47 antibody or antigen binding fragment thereof provided herein.

In some such embodiments of the present invention, step a) includes:

i) obtaining cancer tissue; and ii) detection of CD47 in isolated cancer tissue.

Some embodiments of the present invention provide methods for treating a cancer characterized by CD47 expression in a subject, comprising:

identifying an entity with increased levels of macrophage infiltration in cancerous tissue compared to non-cancerous tissue; and administering to the subject a therapeutically effective amount of an anti-CD47 antibody or antigen binding fragment thereof provided herein.

In some such embodiments of the present invention, step a) includes:

i) obtaining cancerous tissue and surrounding non-cancerous tissue from the subject;

ii) detection of macrophages in isolated cancerous tissue and in non-cancerous tissue; and iii) comparison of quantification of staining in cancerous versus noncancerous tissue. In some embodiments of the present invention, macrophage staining is performed using an anti-CD163 antibody.

- 4 045361

In some embodiments of the present invention, methods of treating cancer include identifying a subject as having a cancer characterized by CD47 expression and increased levels of macrophage infiltration in cancer tissue compared to non-cancerous tissue.

In some embodiments, the present invention provides a method of inducing apoptosis of a cell expressing CD47, comprising contacting the cell with an antibody or antigen-binding fragment thereof provided herein. In some embodiments of the present invention, the cells are in vitro. In some embodiments of the present invention, the cells are in vivo.

In some embodiments, the present invention provides a shielded antibody comprising an antibody or antigen binding fragment thereof that specifically binds to the human CD47 protein and at least one shielding domain, wherein the at least one shielding domain comprises an amino acid sequence selected from SEQ ID NOs: 44-55, 75-86, 94 and 95. In some embodiments, the present invention provides a shielded antibody that comprises an antibody or antigen binding fragment thereof provided herein and at least one shielding domain. In some embodiments of the present invention, the at least one shielding domain comprises an amino acid sequence selected from SEQ ID NOs: 44-55, 75-86, 94 and 95.

In some embodiments of the present invention, the at least one shielding domain reduces the binding affinity of the antibody or antigen binding fragment to the human CD47 protein compared to an antibody or antigen binding fragment thereof without at least one shielding domain. In some embodiments of the present invention, the binding affinity is reduced by at least about 100-fold compared to an antibody or antigen-binding fragment thereof without at least one shielding domain. In some embodiments of the present invention, the binding affinity is reduced from about 200-fold to about 1500-fold compared to an antibody or antigen binding fragment thereof without at least one shielding domain.

In some embodiments of the present invention, an antibody or antigen binding fragment thereof comprises a heavy chain and a light chain, wherein the heavy chain is associated with a first shielding domain; or wherein the light chain is associated with a second shielding domain; or wherein the heavy chain is associated with the first shielding domain and the light chain is associated with the second shielding domain. In some embodiments of the present invention, the first shielding domain comprises an amino acid sequence selected from SEQ ID NO: 44, 46, 48, 50, 52, 54, 75, 77, 79, 81, 83, 85, and 94; and the second shielding domain comprises an amino acid sequence selected from SEQ ID NOs: 45, 47, 49, 51, 53, 55, 76, 78, 80, 82, 84, 86, and 95. In some embodiments of the present invention, the first shielding domain and the second shielding domain is a pair of shielding domains selected from: SEQ ID NO: 44 and 45; SEQ ID NO: 46 and 47; SEQ ID NO: 48 and 49; SEQ ID NO: 50 and 51; SEQ ID NO: 52 and 53; SEQ ID NO: 54 and 55; SEQ ID NO: 75 and 76; SEQ ID NO: 77 and 78; SEQ ID NO: 79 and 80; SEQ ID NO: 81 and 82; SEQ ID NO: 83 and 84; SEQ ID NO: 85 and 86; and SEQ ID NOs: 94 and 95. In some embodiments of the present invention, the first shielding domain is associated with the N-terminus of the heavy chain and the second shielding domain is associated with the N-terminus of the light chain.

In some embodiments of the present invention, each shielding domain contains a protease-cleavable linker and is linked to a heavy chain or light chain through a protease-cleavable linker. In some embodiments of the present invention, the protease-cleavable linker comprises a matrix metalloprotease (MMP) cleavage site. In some embodiments of the present invention, the MMP cleavage site is selected from the MMP2 cleavage site, MMP7 cleavage site, MMP9 cleavage site, and MMP3 cleavage site. In some embodiments of the present invention, after cleavage by MMP, the heavy chain and/or light chain of the antibody or antigen binding fragment thereof comprises a fragment of the amino acid residues of the MMP cleavage site. In some embodiments of the present invention, the cleavage site amino acid residue fragment comprises an LRSG, SG, or VR sequence at the N-terminus of the antibody.

In some embodiments of the present invention, the antibody or antigen-binding fragment thereof is an antibody or antigen-binding fragment thereof provided herein. In some embodiments of the present invention, the shielded antibody comprises a heavy chain associated with a first shielding domain and having the amino acid sequence of SEQ ID NO: 42, and a light chain associated with a second shielding domain and having the amino acid sequence of SEQ ID NO: 43.

In some embodiments, the present invention provides a nucleic acid sequence that encodes a screened antibody provided herein. In some embodiments of the present invention, an expression vector is provided that contains the specified nucleic acid. Some embodiments of this invention provide

- 5 045361 host cell that expresses the shielded antibody provided herein. In some embodiments of the present invention, the method of producing a screened antibody provided herein includes culturing a host cell. In some embodiments of the present invention, the method further comprises isolating the screened antibody.

Some embodiments of the present invention provide methods for treating a cancer characterized by CD47 expression in a subject, comprising administering to the subject a therapeutically effective amount of a screened antibody provided herein.

Some embodiments of the present invention provide methods for treating a cancer characterized by CD47 expression in a subject, comprising:

a) identifying the subject as having elevated levels of MMPs in the tumor compared to surrounding non-cancerous tissue; And

b) administering to a subject a therapeutically effective amount of a shielded antibody provided herein, wherein each shield domain of the shielded antibody contains a protease-cleavable linker, and wherein the protease-cleavable linker contains a matrix metalloprotease (MMP) cleavage site. In some embodiments of the present invention, the MMP cleavage site is selected from the MMP2 cleavage site, MMP7 cleavage site, MMP9 cleavage site, and MMP13 cleavage site. In some embodiments of the present invention, the MMP is selected from MMP2, MMP7, MMP9, and MMP13. In some embodiments of the present invention, step a) comprises:

i) obtaining cancerous and non-cancerous tissue from the subject;

ii) detection of MMPs in isolated cancer tissue and in non-cancerous tissue; and iii) comparison of quantification of staining in cancerous versus noncancerous tissue.

Some embodiments of the present invention provide methods for treating a cancer characterized by CD47 expression in a subject, comprising:

identifying the subject as having a cancer that is characterized by CD47 expression; and administering to the subject a therapeutically effective amount of the screened antibody provided herein.

In some embodiments of the present invention, step a) comprises:

i) obtaining cancer tissue; and ii) detection of CD47 in isolated cancer tissue.

Some embodiments of the present invention provide methods for treating a cancer characterized by CD47 expression in a subject, comprising:

identifying an entity with increased levels of macrophage infiltration in cancerous tissue compared to non-cancerous tissue; and administering to the subject a therapeutically effective amount of the screened antibody provided herein.

In some embodiments of the present invention, step a) comprises:

i) obtaining cancerous tissue and surrounding non-cancerous tissue from the subject;

ii) detection of macrophages in isolated cancerous tissue and in non-cancerous tissue; and iii) comparison of quantification of staining in cancerous versus noncancerous tissue. In some embodiments of the present invention, macrophage staining is performed using an anti-CD 163 antibody.

Some embodiments of the present invention provide methods for treating a cancer characterized by CD47 expression in a subject, comprising identifying the subject as having (a) elevated levels of MMPs in the tumor relative to surrounding non-cancerous tissue and (b) a cancer characterized by the expression CD47. Some embodiments of the present invention provide methods for treating a cancer characterized by CD47 expression in a subject, comprising identifying the subject as having (a) increased levels of MMPs in the tumor relative to surrounding non-cancerous tissue and (b) increased levels of macrophage infiltration in cancerous tissue compared to non-cancerous tissue. Some embodiments of the present invention provide methods for treating a cancer characterized by CD47 expression in a subject, comprising identifying the subject as having (a) a cancer characterized by CD47 expression and (b) increased levels of macrophage infiltration in cancer tissue compared to tissue not being cancerous. Some embodiments of the present invention provide methods for treating a cancer characterized by CD47 expression in a subject, comprising identifying the subject as having (a) elevated levels of MMPs in the tumor relative to surrounding noncancerous tissue, and (b) a cancer characterized by expression of CD47, and (c) increased levels of macrophage infiltration into cancer tissue by

- 6 045361 compared to non-cancerous tissue.

In some embodiments of the present invention, the shielded antibody comprises at least one shielding domain containing a protease-cleavable linker, and wherein the protease-cleavable linker is cleaved in the tumor microenvironment. In some embodiments of the present invention, upon cleavage of the protease-cleavable linker in the tumor microenvironment, the shielding domain is released from the anti-CD47 antibody or antigen-binding fragment thereof. In some embodiments of the present invention, the protease-cleavable linker comprises the amino acid sequence IPVSLRSG (SEQ ID NO: 73) or GPLGVR (SEQ ID no: 57). In some embodiments of the present invention, the protease-cleavable linker comprises an MMP cleavage site. In some embodiments of the present invention, the MMP cleavage site is selected from the MMP2 cleavage site, MMP7 cleavage site, MMP9 cleavage site, and MMP3 cleavage site. In some embodiments of the present invention, upon release of the anti-CD47 antibody or antigen-binding fragment thereof, the anti-CD47 antibody or antigen-binding fragment thereof has a moiety of amino acid residues of a protease-cleavable linker. In some embodiments of the present invention, the cleavage site amino acid residue fragment comprises an LRSG, SG, or VR sequence at the N-terminus of the antibody.

In various embodiments of the present invention, the cancer characterized by CD47 expression is a hematologic malignancy or a solid cancer. In some embodiments of the present invention, the cancer characterized by CD47 expression is selected from non-Hodgkin's lymphoma, B-lymphoblastic lymphoma; B-cell chronic lymphocytic leukemia/small lymphocyte lymphoma, Richter's syndrome, follicular lymphoma, multiple myeloma, myelofibrosis, polycythemia vera, cutaneous T-cell lymphoma, monoclonal gammopathy of unknown etiology (MGUS), myelodysplastic syndrome (MDS), immunoblastic large cell lymphoma, B-lymphoblastic progenitor cell leukemia, acute myeloid leukemia (AML) and anaplastic large cell lymphoma. In some embodiments of the present invention, the cancer characterized by CD47 expression is selected from lung cancer, pancreatic cancer, breast cancer, liver cancer, ovarian cancer, testicular cancer, kidney cancer, bladder cancer, spine cancer, brain cancer, cancer cervical cancer, endometrial cancer, colorectal cancer, anal cancer, esophageal cancer, gallbladder cancer, gastrointestinal cancer, stomach cancer, head and neck cancer, skin cancer, melanoma, prostate cancer, pituitary cancer, uterine cancer, vaginal cancer and thyroid cancer. In some embodiments of the present invention, the cancer characterized by CD47 expression is selected from lung cancer, sarcoma, colorectal cancer, head and neck cancer, ovarian cancer, pancreatic cancer, gastric cancer, melanoma, and breast cancer.

In some embodiments of the present invention, an anti-CD47 antibody or an antigen-binding fragment thereof or a shielded antibody is administered in combination with an immune checkpoint molecule inhibitor selected from one or more of programmed cell death protein 1 (PD-1), a programmed cell death ligand (PD -Ll), PD-L2, cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), T-cell immunoglobulin and mucin domain-containing protein 3 (TIM-3), lymphocyte activation gene 3 (LAG-3), carcinomaembryonic antigen-dependent molecule cell adhesion 1 (CEACAM-1), SEACAM-5, V-domain immunoglobulin suppressor of T-cell activation (VISTA), B- and T-lymphocyte attenuator (BTLA), T-cell immunoreceptor with Ig and ITIM domains (TIGIT), leukocyte-associated immunoglobulin receptor (type 1) (LAIR1), CD160, 2B4 or TGFR. In some embodiments of the present invention, an anti-CD47 antibody or antigen-binding fragment or shielded antibody thereof is administered in combination with an agonistic anti-CD40 antibody. In some embodiments of the present invention, the CD40 agonist antibody has low levels of fucosylation or is afucosylated. In some embodiments of the present invention, an anti-CD47 antibody or antigen-binding fragment or shielded antibody thereof is administered in combination with an antibody-drug conjugate (ADC), wherein the ADC antibody specifically binds to a protein that is expressed on the extracellular surface of the cancer cell and the antibody is conjugated to the compound linker-drug containing a cytotoxic agent. In some embodiments of the present invention, the cytotoxic agent is auristatin. In some embodiments of the present invention, the ADC antibody is conjugated to a linker-drug compound selected from vcMMAE and mcMMAF.

In some embodiments of the present invention, an anti-CD47 antibody or antigen-binding fragment thereof, or an anti-CD47 antibody or a screened antibody antigen-binding fragment thereof exhibits reduced hemagglutination in vitro compared to its parent anti-CD47 antibody. In some embodiments of the present invention, the parent antibody is Ab47. In some embodiments of the present invention, administration of the anti-CD47 antibody or screened antibody does not cause hemagglutination in the subject. In some variants

- 7 045361 implementations of the present invention, an anti-CD47 antibody or a shielded antibody causes apoptosis of cells expressing CD47 in vitro and/or in vivo. In some embodiments of the present invention, the anti-CD47 antibody or shielded antibody causes apoptosis of cells expressing CD47 in vivo. In some embodiments of the present invention, the cells expressing CD47 are cancer cells.

In one aspect, a humanized antibody or antigen binding fragment thereof is provided that specifically binds human CD47, wherein the antibody or antigen binding fragment comprises a light chain variable region and a heavy chain variable region, wherein said heavy chain variable region comprises CDRs represented by SEQ ID NO : 16 (GYGMS), 17 (TITSGGTYTYYPDSVKG) and 18 (SLAGNAMDY), and the human IGHV323/HJ4 framework sequence shown in SEQ ID NO: 88 (Fig. 1A), with the framework sequence positions H44, H49, H82, H89, H91 and H94, in accordance with Kabat's numbering, represent the remains of the donor.

In another aspect, a humanized antibody or antigen binding fragment thereof is provided that specifically binds human CD47, wherein the antibody or antigen binding fragment comprises a light chain variable region and a heavy chain variable region, wherein said heavy chain variable region comprises CDRs represented by SEQ ID NO : 16 (GYGMS), 17 (TITSGGTYTYYPDSVKG) and 18 (SLAGNAMDY), and the human IGHV348/HJ4 framework sequence shown in SEQ ID NO: 89 (Fig. 1B), with the framework sequence position H49, according to Kabat numbering , represents the donor residue.

In yet another aspect, a humanized antibody or antigen binding fragment thereof is provided that specifically binds human CD47, wherein the antibody or antigen binding fragment comprises a light chain variable region and a heavy chain variable region, wherein said heavy chain variable region comprises CDRs represented by SEQ ID NO: 16 (GYGMS), 17 (TITSGGTYTYYPDSVKG) and 18 (SLAGNAMDY), and the human IGHV366/HJ4 framework sequence shown in SEQ ID NO: 90 (Fig. 1C), with framework sequence positions H29, H49 and H82, in according to Kabat's numbering, they represent the remains of the donor.

In yet another aspect, a humanized antibody or antigen binding fragment thereof is provided that specifically binds human CD47, wherein the antibody or antigen binding fragment comprises a light chain variable region and a heavy chain variable region, wherein said heavy chain variable region comprises CDRs represented by SEQ ID NO: 16 (GYGMS), 17 (TITSGGTYYPDSVKG) and 18 (SLAGNAMDY), and the human IGHV374/HJ4 framework sequence shown in SEQ ID NO: 91 (Fig. 1D), with the position of the framework sequence H49, according to numbering Kabatu, represents the remainder of the donor.

In yet another aspect, a humanized antibody or antigen binding fragment thereof is provided that specifically binds human CD47, wherein the antibody or antigen binding fragment comprises a light chain variable region and a heavy chain variable region, wherein said heavy chain variable region comprises CDRs represented by SEQ ID NO: 31 (RASQTISDYLH), 32 (FASQSIS) and 33 (QNGHGFPRT); and the human IGKV6-21/KJ2 framework sequence shown in SEQ ID NO: 92 (Fig. 1G), wherein the framework sequence positions L4, L21, L69 and L85, according to Kabat numbering, represent donor residues.

In yet another aspect, a humanized antibody or antigen binding fragment thereof is provided that specifically binds human CD47, wherein the antibody or antigen binding fragment comprises a light chain variable region and a heavy chain variable region, wherein said heavy chain variable region comprises CDRs represented by SEQ ID NO: 31 (RASQTISDYLH), 32 (FASQSIS) and 33 (QNGHGFPRT); and the human IGKV1-27/KJ2 framework sequence shown in SEQ ID NO: 93 (Fig. 1H), wherein framework sequence positions L21, L49 and L69, according to Kabat numbering, represent donor residues.

In one embodiment of the present invention, the frame sequence position L4 is occupied by M, L21 is occupied by L, L49 is occupied by K, L69 is occupied by T or S, and L85 is occupied by V or T, in accordance with Kabat numbering.

In one embodiment of the present invention, the antibody or antigen binding fragment comprises a heavy chain variable region (HCVR) having at least 90% sequence identity to any of SEQ ID NOs: 2, 3, 4, 5, 6, 7 and 8. and a light chain variable region (LCVR) having at least 90% sequence identity to any of SEQ ID NOs: 10, 11, 12, 13, 14 and 15.

In one embodiment of the present invention, the antibody or antigen binding fragment further comprises the G91A mutation in LCDR3, according to Kabat numbering.

In one embodiment of the present invention, the antibody or antigen-binding fragment thereof is an IgG1 isotype.

In one embodiment of the present invention, the antibody or antigen-binding fragment has increased antibody-dependent cellular cytotoxicity (ADCC) compared to its counterpart.

- 8 045361 running antibody.

In one embodiment of the present invention, the antibody or antigen binding fragment has increased antibody dependent cellular phagocytosis (ADCP) compared to its parent antibody.

In one embodiment of the present invention, the antibody or antigen binding fragment has reduced backbone fucosylation compared to its parent antibody.

In one embodiment of the present invention, the antibody or antigen binding fragment blocks the interaction between CD47 and SIRPa.

In one embodiment of the present invention, the antibody or antigen binding fragment has reduced red blood cell hemagglutination compared to its parent antibody.

In one aspect, a nucleic acid sequence is provided encoding a humanized antibody or antigen binding fragment thereof that specifically binds human CD47.

In one embodiment of the present invention, the antigen binding fragment comprises a Fab, Fab', F(ab')2, Fv fragment, diabody, single chain antibody, scFv fragment, or scFv-Fc.

In one aspect, there is provided a method of treating a cancer characterized by CD47 in a subject, comprising administering to the subject a therapeutically effective amount of an anti-CD47 antibody or antigen-binding fragment thereof comprising a shielding agent (also referred to as a shielding domain), wherein the shielding agent comprises one or more supercoiled peptides, which reduce the binding affinity of the antibody or antigen-binding fragment to human CD47 compared to the antibody or antigen-binding fragment without a screening agent.

In one embodiment of the present invention, a protease-cleavable linker attaches a screening agent to an antibody or antigen-binding fragment thereof.

In one embodiment of the present invention, the protease-cleavable linker has an amino acid sequence comprising IPVSLRSG (SEQ ID NO: 73) or GPLGVR (SEQ ID NO: 57).

In one embodiment of the present invention, the protease-cleavable linker comprises a matrix metalloprotease (MMP) cleavage site.

In one embodiment of the present invention, the MMP cleavage site is selected from the MMP2 cleavage site, the MMP7 cleavage site, the MMP9 cleavage site, and the MMP3 cleavage site.

In one embodiment of the present invention, the screening agent is released from the anti-CD47 antibody or antigen-binding fragment thereof following cleavage of the MMP cleavage site in the tumor microenvironment by MMPs.

In one embodiment of the present invention, the cleaved anti-CD47 antibody has a fragment of the amino acid residues of the MMP cleavage site.

In one embodiment of the present invention, the cleavage site amino acid residue fragment contains the sequence LRSG, SG or VR at the N-terminus of the antibody.

In one embodiment of the present invention, one or more coiled-coil peptides comprise one or more sequences selected from SEQ ID NO: 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, and 55. In one embodiment implementations of the present invention, one or more coiled-coil peptides comprise one or more sequences selected from SEQ ID NOs: 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85 and 86. In one embodiment of the present invention one or more coiled-coil peptides contain one or more sequences selected from SEQ ID NOs: 94 and 95.

In one embodiment of the present invention, the binding of the antibody or antigen binding fragment to CD47 is reduced by at least about 100-fold compared to the antibody or antigen binding fragment thereof without a screening agent.

In one embodiment of the present invention, the binding of the antibody or antigen binding fragment to CD47 is reduced from about 200-fold to about 1500-fold compared to the antibody or antigen binding fragment thereof without a screening agent.

In one embodiment of the present invention, the cancer characterized by CD47 expression is a hematologic malignancy that causes solid cancer.

In one embodiment of the present invention, the hematologic malignancy is selected from non-Hodgkin's lymphoma, B-lymphoblastic lymphoma; B-cell chronic lymphocytic leukemia/small lymphocyte lymphoma, Richter's syndrome, follicular lymphoma, multiple myeloma, myelofibrosis, polycythemia vera, cutaneous T-cell lymphoma, monoclonal gammopathy of unknown etiology (MGUS), myelodysplastic syndrome (MDS), immunoblastic large cell lymphoma, B-lymphoblastic progenitor cell leukemia, acute myeloid leukemia (AML) and anaplastic large cell lymphoma.

In one embodiment of the present invention, the cancer characterized by CD47 expression is a solid cancer.

In one embodiment of the present invention, the solid cancer is selected from lung cancer, pancreatic cancer, breast cancer, liver cancer, ovarian cancer, testicular cancer, kidney cancer, breast cancer

- 9 045361 bladder cancer, spine cancer, brain cancer, cervical cancer, endometrial cancer, colorectal cancer, anal cancer, esophageal cancer, gall bladder cancer, gastrointestinal cancer, stomach cancer, sarcoma, head and neck cancer, melanoma, skin cancer, prostate cancer, pituitary cancer, uterine cancer, vaginal cancer and thyroid cancer.

In one embodiment of the present invention, the solid cancer is selected from lung cancer, sarcoma, ovarian cancer, pancreatic cancer, gastric cancer, melanoma, colorectal cancer, head and neck cancer, and breast cancer.

In one embodiment of the present invention, the subject is a person with solid cancer.

In one embodiment of the present invention, an anti-CD47 antibody is administered in combination with an immune checkpoint molecule inhibitor selected from one or more of programmed cell death protein 1 (PD-1), programmed cell death ligand 1 (PD-L1), PD-L2 , cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), T-cell immunoglobulin and mucin domain-containing protein 3 (TIM-3), lymphocyte activation gene 3 (LAG-3), carcinoembryonic antigen-dependent cell adhesion molecule 1 (CEACAM-1 ), SEACAM-5, V-domain immunoglobulin suppressor of T-cell activation (VISTA), B- and T-lymphocyte attenuator (BTLA), T-cell immunoreceptor with Ig and ITIM domains (TIGIT), immunoglobulin-like receptor associated with leukocytes (type 1) (LAIR1), CD160, 2B4 or TGFR.

In one aspect, an antibody or antigen-binding fragment thereof is provided that specifically binds to the human CD47 protein, which comprises a screening agent, wherein the screening agent comprises one or more coiled-coil peptides containing the sequence

SEQ ID NO: 95 (QGASTTVAQLEEKVKTLRAENYELKSEVQRLEEQVAQLGS) and/or

SEQ ID NO: 94 (QGASTSVDELQAEVDQLEDENYALKTKVAQLRKKVEKLGS), and wherein one or more coiled-coil peptides reduce the binding affinity of the antibody or antigen binding fragment thereof to human CD47 protein compared to the antibody or antigen binding fragment thereof without a screening agent.

In one embodiment of the present invention, the antibody or antigen binding fragment comprises a heavy chain variable region (HCVR) having at least 90% sequence identity to any of SEQ ID NOs: 2, 3, 4, 5, 6, 7 and 8. and a light chain variable region (LCVR) having at least 90% sequence identity to any of SEQ ID NOs: 10, 11, 12, 13, 14 and 15.

In one embodiment of the present invention, the screening agent is attached to the antibody or antigen-binding fragment thereof via a protease-cleavable linker.

In one embodiment of the present invention, the protease-cleavable linker has an amino acid sequence comprising IPVSLRSG (SEQ ID NO: 73) or GPLGVR (SEQ ID NO: 57).

In one embodiment of the present invention, the protease-cleavable linker comprises a matrix metalloprotease (MMP) cleavage site.

In one embodiment of the present invention, the MMP cleavage site is selected from the MMP2 cleavage site, the MMP7 cleavage site, the MMP9 cleavage site, and the MMP3 cleavage site.

In one embodiment of the present invention, the shielding agent is removed from the anti-CD47 antibody following cleavage of the MMP cleavage site by MMP.

In one embodiment of the present invention, the anti-CD47 antibody has a fragment of MMP cleavage site amino acid residues after cleavage of the MMP cleavage site by MMP.

In one embodiment of the present invention, the cleavage site amino acid residue fragment contains the sequence LRSG, SG or VR at the N-terminus of the antibody.

In one embodiment of the present invention, binding is reduced by at least about 100-fold compared to an antibody or antigen-binding fragment thereof without a screening agent.

In one embodiment of the present invention, binding is reduced from about 200-fold to about 1500-fold compared to an antibody or antigen binding fragment thereof without a screening agent.

In one embodiment of the present invention, the antibody or antigen binding fragment thereof comprises a heavy chain sequence of SEQ ID NO: 42 and a light chain sequence of SEQ ID NO: 43.

In one embodiment of the present invention, the antibody or antigen binding fragment comprises a variant Fc region that confers enhanced effector function selected from ADCC and/or CDC activity.

In one embodiment of the present invention, the antibody or antigen-binding fragment thereof is afucosylated.

In one aspect, a humanized antibody or antigen-binding fragment thereof is provided that specifically binds human CD47, wherein the antibody is an IgG1 isotype.

- 10 045361

In one embodiment of the present invention, the antibody contains increased ADCC, increased ADCC, and/or increased ADCC activity.

In one aspect, there is provided a method of treating a cancer characterized by CD47 expression in a subject, comprising the steps of:

a) identifying the subject as having elevated levels of MMPs in the tumor compared to surrounding non-cancerous tissue; And

b) administering to a subject a therapeutically effective amount of an anti-CD47 antibody or antigen-binding fragment thereof containing a shielding agent, wherein the shielding agent contains coiled-coil peptides that reduce the binding affinity of the antibody or antigen-binding fragment thereof to human CD47 compared to the antibody or antigen-binding fragment thereof without the shielding agent, if the subject has elevated levels of MMPs in the tumor compared to surrounding non-cancerous tissue.

In one embodiment of the present invention, the MMP is selected from the group consisting of: MMP2, MMP7, MMP9 and MMP13.

In one embodiment of the present invention, step a) includes:

i) obtaining cancerous and non-cancerous tissue from the subject;

ii) detection of MMPs in isolated cancer tissue and in non-cancerous tissue; and iii) comparison of quantification of staining in cancerous versus noncancerous tissue.

In one aspect, there is provided a method of treating a cancer characterized by CD47 expression in a subject, comprising the steps of:

a) identifying the subject as having elevated levels of CD47 in the tumor compared to surrounding non-cancerous tissue; And

b) administering to a subject a therapeutically effective amount of an anti-CD47 antibody or antigen-binding fragment thereof containing a shielding agent, wherein the shielding agent contains coiled-coil peptides that reduce the binding affinity of the antibody or antigen-binding fragment thereof to human CD47 compared to the antibody or antigen-binding fragment thereof without the shielding agent, if the subject has elevated levels of CD47 in the tumor compared to surrounding non-cancerous tissue.

In one embodiment of the present invention, step a) includes:

i) obtaining cancerous tissue and surrounding non-cancerous tissue from the subject;

ii) detection of CD47 in isolated cancer tissue and surrounding non-cancerous tissue; and iii) comparison of quantification of CD47 staining in cancerous tissue versus CD47 staining in noncancerous tissue.

In one aspect, there is provided a method of treating a cancer characterized by CD47 expression in a subject, comprising the steps of:

a) identifying an entity with increased levels of macrophage infiltration in cancerous tissue compared to non-cancerous tissue; And

b) administering to a subject a therapeutically effective amount of an anti-CD47 antibody or antigen-binding fragment thereof containing a screening agent, wherein the screening agent comprises one or more coiled-coil peptides that reduces the binding affinity of the antibody or antigen-binding fragment thereof to human CD47 compared to the antibody or antigen-binding fragment thereof without screening agent if the subject has increased levels of macrophage infiltration into the tumor compared to non-cancerous tissue.

In one embodiment of the present invention, step a) includes:

i) obtaining cancerous tissue and surrounding non-cancerous tissue from the subject;

ii) detection of macrophages in isolated cancerous tissue and in non-cancerous tissue; and iii) comparison of quantification of staining in cancerous versus noncancerous tissue.

In one embodiment of the present invention, macrophage staining is performed using an anti-CD163 antibody.

In one aspect, a humanized antibody or antigen binding fragment thereof that specifically binds human CD47 comprises a heavy chain variable region (HCVR) having at least 90% sequence identity to any of SEQ ID NO: 2, 3, 4, 5, 6, 7 and 8, and a light chain variable region (LCVR) having at least 90% sequence identity to any of SEQ ID NOs: 10, 11, 12, 13, 14 and 15, wherein the antibody further comprises the sequence LRSG , SG or VR at the N-terminus of HCVR and/or LCVR.

In one aspect, a method is provided for treating cancer by administering a combination of a shielded anti-CD47 antibody of the present invention with an agonistic anti-CD40 antibody.

In one embodiment of the present invention, the CD40 agonist antibody has low levels of fucosylation, for example, the SEA-CD40 antibody.

- 11 045361

In one aspect, a method of treating cancer is provided by administering a combination of a shielded anti-CD47 antibody according to claim 37 with an antibody-drug conjugate (ADC), wherein the ADC antibody specifically binds to a protein that is expressed on the extracellular surface of the cancer cell and the antibody is conjugated to the compound linker-drug containing a cytotoxic agent.

In one embodiment of the present invention, the cytotoxic agent is auristatin.

In one embodiment of the present invention, the ADC antibody is conjugated to a linker-drug compound selected from vcMMAE and mcMMAF.

The summary of the disclosure described above is not limiting, and other features and advantages of the disclosed antibodies, as well as methods for their manufacture and use, will be apparent from the following drawings, detailed description of the invention, examples and claims.

Brief description of graphic materials

In fig. Ία-J shows the antibody sequence alignment. In fig. 1A depicts an alignment of hB6H12 heavy chain sequence variants with the human VH donor sequence, HV3-23/HJ4. In fig. 1B shows an alignment of hB6H12 heavy chain sequence variants with the human VH donor sequence, HV3-48/HJ4. In fig. 1C depicts an alignment of hB6H12 heavy chain sequence variants with the human VH donor sequence, HV3-66/HJ4. In fig. 1D depicts an alignment of hB6H12 heavy chain sequence variants with the human VH donor sequence, HV3-74/HJ4. In fig. 1E depicts an alignment of hB6H12 heavy chain sequence variants. In fig. 1F depicts the heavy chain sequence alignment of hB6H12.3 (hvH1) compared to tB6H12 and Ab47. In fig. 1G depicts an alignment of hB6H12 light chain sequence variants with the human VH donor sequence, KV1-3/KJ2. In fig. 1H depicts an alignment of hB6H12 light chain sequence variants with the human VH donor sequence, KV1-27/KJ2. In fig. 1I depicts an alignment of hB6H12 light chain sequence variants. In fig. 1J depicts the light chain sequence alignment of hB6H12.3 (hvK3) compared to tB6H12 and Ab47.

In fig. 2A-C depict the affinity and kinetics of antibody binding. In fig. 2A depicts CD47-enriched fluorescent cell sorting (FACS) with typical anti-CD47 antibodies. In fig. 2B shows a CD47-loading ELISA with typical anti-CD47 antibodies. In fig. 2C depicts the binding kinetics of CD47 to typical anti-CD47 antibodies.

In fig. 3A, B depicts antibody-mediated phagocytosis. In fig. FOR and FIG. 3B depicts antibody-mediated phagocytosis of human CD47+ red blood cells (RBCs) with typical anti-CD47 antibodies.

In fig. 4A, B depicts antibody-mediated hemagglutination. In fig. Figure 4A shows a snapshot of the formation of dispersed non-precipitated red blood cells. In fig. 4B shows the percentage of hemagglutination of erythrocytes by antibodies against CD47 Ab47 and hB6H12.3.

In fig. 5A,B depicts antibody-mediated activation of Fcy receptors. Luciferase activity containing the NFAT reporter was measured from Jurkat cells transfected with FcyRI (Fig. 5A) or high-affinity FcyRIIIa-H (Fig. 5B) and exposed WIL2S cells coated with increasing concentrations of either mouse B6H12, Ab47, or hB6H12 .3.

In fig. 6A,B depict NK cell-mediated ADCC and FcyRIIIa activation. Chromium-labeled WIL2S cells coated with mB6H12, Ab47, or hB6H12.3 were exposed to Jurkat cells stably expressing the high-affinity V/V variant of FcyRIIIa, and receptor activation was assessed as NFAT-dependent luciferase activity. ADCC (Fig. 6A) and FcyRIIIa activation (Fig. 6B) were compared between mB6H12, Ab47 and hB6H12.3.

In fig. 7A-D depict the suppressor function of anti-CD47 antibodies. Differentiated monocytes with a tumor-associated macrophage (MAO (TAM)) phenotype showed increased levels of macrophage activation markers CD86 (Fig. 7A) and MHCII (Fig. 7B) when exposed to anti-CD47 antibodies. TCR-mediated T cell activation was assessed by detecting upregulation of MHCII (Fig. 7C) and IFNy secretion (Fig. 7D) in T cells.

In fig. 8A-D depicts a comparison between hB6H12.3 and the anti-CD47 antibody 5F9. Activation of FcyRI was detected in Jurkat cells harboring the luciferase NFAT reporter (Fig. 8A). FcyRII activation was assessed in Jurkat cells harboring the luciferase NFAT reporter (Fig. 8B). NK-mediated ADCC activity was determined (Fig. 8C). The level of INFu secretion by T cells was assessed (Fig. 8D).

In fig. 9A,B depicts mass spectrometry data for a reactivated MMP2-screened antibody. Shown is the deconvoluted light chain mass for Vel-IPV-hB6H12.3 before (Fig. 9A) and after (Fig. 9B) digestion of recombinant human MMP2. The expected m/z value for the intact light chain is 28681 (recorded: 28680.8). The expected m/z value for the MMP2 cleaved antibody (LRSG-hB6H12.3) is 23969 (reported: 23968.4).

- 12 045361

In fig. 10 depicts saturation of anti-CD47 antibodies into human bladder cancer SW780 cells. Vel-IPV-screened hB6H12 antibodies were tested together with pre-MMP2-activated comparators. The cleaved Vel-IPV antibodies contained the remainder of the LRSG sequence at the N-terminus of the antibody.

In fig. 11 depicts saturation of anti-CD47 antibodies into human bladder cancer SW780 cells. Vel-IPV-screened hB6H12 antibodies were tested together with pre-MMP2-activated comparators. Vel-IPV cleaved antibodies and IPV antibodies with a fragment of amino acid residues of the cleavage site contained the remainder of the LRSG sequence at the N-terminus of the antibody. The cleaved antibody was produced by cleavage with MMP2, whereas the IPV antibody with a cleavage site moiety was produced recombinantly.

In fig. 12 shows the saturation of human erythrocytes with anti-CD47 antibodies. Vel-IPV-screened hB6H12 antibodies were tested together with reactivated comparators (IPV-hB6H12.3 with cleavage site fragment or MMP2-cleaved Vel-IPV-hB6H12.3). Vel-IPV cleaved antibodies and IPV antibodies with a fragment of amino acid residues of the cleavage site contained the remainder of the LRSG sequence at the N-terminus of the antibody. The cleaved antibody was produced by cleavage with MMP2, whereas the IPV antibody with a cleavage site moiety was produced recombinantly.

In fig. 13 depicts anti-CD47 antibody saturation of rCD47 using ELISA detection. Vel-IPV-hB6H12.3 exhibited significantly reduced binding. Binding can be restored after cleavage of rhMMP2.

In fig. 14 depicts anti-CD47 antibody saturation of rCD47 by ELISA detection. Both hB6H12.3 and hB6H12.3 G91A show higher Bmax than Ab47.

In fig. 15 depicts saturation of anti-CD47 antibodies into human bladder cancer SW780 cells. Binding of Ab47 and hB6H12.3 was compared to variants carrying the G91A mutation in CDR-L3.

In fig. 16 shows the saturation of human erythrocytes with anti-CD47 antibodies. Binding of Ab47 and hB6H12.3 was compared to variants carrying the G91A mutation in CDR-L3.

In fig. 17A, B depicts the activity of anti-CD47 antibodies in the L428 tumor xenograft model in NSG mice. Antibodies were administered intraperitoneally (ip) every four days for four doses (q4dx4) of 1 or 10 mg/kg (Fig. 17A). Analysis of tumor tissue using the macrophage marker antiT4/80 showed the presence of murine macrophages in the L428 xenograft tumor model (Figure 17B).

In fig. 18A-0 depicts the activity of anti-CD47 antibodies in the L428 tumor xenograft model in NSG mice. Antibodies were administered IP q4dx4 (4 doses every 4 days) using a dosage of 1 or 10 mg/kg (Fig. 18A). Analysis of tumor tissue using the macrophage marker antiT4/80 revealed the presence of murine macrophages in the Detroit 562 tumor xenograft model (Fig. 18B). Anti-CD47 antibody activity was also analyzed in the SUDHL8 xenograft tumor model in NSG mice (Fig. 18C). Antibodies were administered IP q4dx4 using a dosage of 1 or 10 mg/kg. Analysis of tumor tissue using the macrophage marker antiT4/80 revealed the presence of murine macrophages in the SUDHL8 xenograft tumor model (Fig. 18D).

In fig. 19A-D depict the activity of anti-CD47 antibodies in tumor models having low levels of intratumoral macrophages. The HT1080 Fibrosarcoma cancer model (FIG. 19A and FIG. 19B) and the HepG2 Hepatocellular carcinoma model (FIG. 19C and FIG. 19D) are shown. Antibodies were administered intravenously with q4dx4 using a dosage of 10 mg/kg. Analysis of tumor tissue using the macrophage marker antiT4/80 showed the presence of murine macrophages in the HT1080 fibrosarcoma cancer model (Fig. 19B) and the HepG2 hepatocellular carcinoma model (Fig. 19D).

In fig. 20 depicts the activity of anti-CD47 antibodies in tumor models having low intratumoral macrophage content, which can be enhanced in combination with the antibody-drug conjugate (ADC) monomethyl auristatin E (MMAE) (which is known to stimulate macrophage infiltration). Anti-CD47 antibody was administered IP q4dx4 using a dosage of 5 mg/kg, while with the ADC MMAE was administered once using a dosage of 1 mg/kg.

In fig. 21A,B shows that the mouse anti-CD47 reactive antibody mIAP301 can be screened using the same VEL and IPV sequences as the human hB6H12.3 antibody (FIG. 21A). Shielding with these constructs blocked antibody binding to murine CD47-positive tumors (FIG. 21B) and prevented functional activity as measured by erythrocyte phagocytosis. In tumor models with low intratumoral macrophage content, anti-CD47 antibody was administered IP q4dx4 using a dosage of 5 mg/kg, while with the ADC MMAE was administered once using a dosage of 1 mg/kg.

In fig. 22A,B depict ³ H-labeled parent and shielded antibodies administered to BALB/c mice. Antibodies were assessed using scintillation measurements. Platelet counts in mice are shown (FIG. 22A) and pharmacokinetics using mouse antibodies (FIG. 22B). In fig. 22 The reference time point for mIAP301 antibody analysis was 24 hours after the last dose of antibody.

- 13 045361

Fig. 23A-D The anti-mouse anti-CD47 antibody mIAP301 exhibited antitumor activity in the A20 lymphoma model, but caused a concomitant decrease in the number of red blood cells (Fig. 23A). The shielded antibody Vel-IPV-mIAP301 showed similar activity, but neutralized the effect on changes in red blood cell count (Fig. 23B). Vel-IPV-mIAP301 prevented erythrocyte antigen binding but supported tumor binding (Figure 23C and Figure 23D).

In fig. 24 depicts that the anti-mouse anti-CD47 antibody mIAP301 stimulates antitumor activity in the MC38 colon cancer model, which is known to be sensitive to immuno-oncology (I/O) agents. The activity of the screened antibody against mIAP301 in this model demonstrated excellent efficacy in animals exhibiting a complete response. Reinfection of this animal resulted in complete tumor rejection, demonstrating the induction of a long-lasting memory T cell response.

Fig. 25A, B. The parent antibody and the screened anti-mouse CD47 antibody mIAP301 induced increased antitumor activity when combined with the surrogate anti-PD-1 antibody, resulting in 4/6 animals demonstrating complete responses (CR) (FIG. 25A). The parent antibody and the shielded anti-mouse anti-CD47 antibody mIAP301 induced increased antitumor activity when combined with the macrophage-activating CD40-targeting SEA-enhanced surrogate antibody 1C10 (Figure 25B).

In fig. 26 depicts the stability of various supercoil humanized B6H12 variants in BALB/c mice three days after dosing. Stability was assessed by densitometry of Western blot results after separation of shielded and unshielded heavy chains using DNS-PAGE under reducing conditions.

Fig. 27A, B. RBC levels were determined following a single IV bolus dose of 0.1, 1, 10, or 30 mg/kg humanized IgG1 hB6H12 Ab47 (FIG. 27A). Doses exceeding 1 mg/kg were not tolerated, and animals at all dose levels exhibited clinical signs associated with hemolysis and exposure to the test drug. Shown are red blood cell levels following a single IV bolus dose of 0.1, 1, or 10 mg/kg of the alternatively screened humanized IgG1 hB6H12 Ab47 Vel-IPV, which demonstrates approximately 10-fold increased tolerability at the highest dose level tested (FIG. 27B). All doses of Vel-IPV-Ab47 were tolerable, and no clinical signs were detected at any dose level.

In fig. Figure 28 shows circulating antibody levels following a single IV bolus dose of 1 mg/kg Ab47 and Vel-IPV-Ab47. While the 1 mg/kg dose of Ab47 was below the detection limit using the nonspecific TAb (total antibody) ELISA assay on day 3 of the study, 1 mg/kg VelIPV-Ab47 was detectable throughout the study course ending at 15th day of the study.

In fig. Figure 29 shows red blood cell levels after a single IV bolus dose of 1 mg/kg Ab47 and Vel-IPV-Ab47. Both Ab47 and hB6H12.3 showed a decrease in red blood cell counts.

In fig. 30 shows platelet levels after a single IV bolus dose of 1 mg/kg Ab47, hB6H12.3, or control. Ab47 demonstrated a 60% reduction in platelet levels relative to pre-dose platelet levels, whereas hB6H12.3 did not cause a platelet reduction above 20%, which was also observed in the control group. Both Ab47 and hB6H12.3 resulted in increased platelet levels from Day 7 to the end of the study, i.e. Day 15.

In fig. 31 shows red blood cell levels following a single IV bolus dose of 10 or 20 mg/kg Vel-IPV-hB6H12.3, which demonstrated approximately 20-fold increased tolerability compared to unshielded hB6H12.3 using a 1 mg/kg dosage. In addition to increased tolerability in hematological parameters, no clinical signs were observed in groups receiving the shielded antibody, whereas they were observed with the 1 mg/kg dosage of unshielded hB6H12.3.

In fig. Figure 32 shows red blood cell levels after a single IV bolus dose of 20 mg/kg VelIPV-hB6H12.3, 20 mg/kg SEA-Vel-IPV-hB6H12.3, 1 mg/kg hB6H12.3, and control for comparison. hB6H12.3 antibodies, both SEA-shielded and non-SEA-shielded, were tolerated equally despite the enhancement of effector functionality by antibody engineering.

In fig. 33 depicts detection of macrophages by immunohistochemistry using anti-CD 163 antibody in breast tumor and normal breast tissue samples.

In fig. 34 shows the sequences of superhelices M11, M15 and Vel. MMP2-cleavage sequences are also depicted.

In fig. 35A,B depict tumor MMP levels in several mouse and human tumors relative to those present in cell culture systems.

In fig. 36A, B depicts selection of MMAE that contains auristatins (LIV1A and CD30) in combination with anti-CD47 antibody in the MCSF7 breast cancer xenograft model for the Liv1A ADC and the L428 lymphoma model for the CD30 ADC.

In fig. 37A-B depict the concentrations of mIAP301 and shielded Vel-IPV-mIAP301 and Vel-M2mIAP301 in plasma (FIG. 37A and FIG. 37B), spleen (FIG. 37C and FIG. 37D), and tumor (FIG. 37E and FIG. 37F). .

- 14 045361

In fig. 38A, B shows the percentage of cleaved antibody in plasma, liver and tumor (FIG. 38A). HT1080 tumors collected from mice exposed to Ab47 or Vel-IPVAb47 for 4 or 7 days were subjected to flow cytometry to determine the extent of antibody that was able to bind to and saturate tumor cells that express CD47 (FIG. 38B).

Fig. 39A-C. Tolerance of shielded Ab47 (FIG. 39A) or shielded hB6H12.3 (FIG. 39B) was determined by measuring the amount of circulating plasma cytokine monocyte chemoattractant protein-1 (MCP-1, MCP-1). Pharmacokinetic analysis using non-specific total antibody (TAb) ELISA was performed on screened and unscreened hB6H12.3 (Fig. 39C).

In fig. 40A, B depicts relative antitumor activity determined by measuring mean tumor volume in Vel-IPV-hB6H12.3 and SEA-Vel-IPVhB6H12.3 xenograft models, and fucosylated and non-fucosylated SEA hB6H12.3 in high (Detroit562) ( Fig. 40A) and low (HT1080) (Fig. 40B) macrophage content.

In fig. 41A, B depicts measurement of circulating MCB-1 cytokine levels using Vel-IPV-hB6H12.3 and SEA-Vel-IPV-hB6H12.3 antibodies (FIG. 41A). A pharmacokinetic analysis using a non-specific total antibody ELISA (TAb) was performed between the Vel-IPVhB6H12.3 antibody with SEA and the Vel-IPV-hB6H12.3 antibody without SEA (Fig. 41B).

In fig. 42 shows the measurement of phagocytosis of CD47-positive human erythrocytes after incubation with hB6H12.3 (anti-CD47 antibody), Vel-IPV-hB6H12.3 (screened anti-CD47 antibody), MMP-cleaved Vel-IPV-hB6H12.3 (MMP-activated screened anti-CD47 antibody) or in the absence of antibody (No effect).

In fig. 43 depicts human red blood cells in a round bottom plate after incubation with Vel-IPVAb47 (Ab47 Shielded), MMP-cleaved Vel-IPV-Ab47 (MMP-cleaved Ab47 Shielded) or without antibodies (No Exposure).

In fig. 44 shows the measurement of Annexin V positive cells after incubation with hB6H12.3, 5F9 or IgG1 isotype control.

In fig. 45A-C depicts measurement of binding of Vel-IPV-hB6H12.3-FITC and hB6H12.3-FITC to a whole blood sample representing 16 of 17 patient samples (FIG. 45A) or one sample with an abnormal value (FIG. 45B); and EC50 values obtained using ELISA after incubation of plasma with recombinant CD47 and hB6H12.3 (Donor l-hB6H12.3, coated) or Vel-IPVhB6H12.3 (Patient 1-10 with Sarcoma) (Fig. 45C).

In fig. 46A, B depicts representative cytokine production induced by incubating whole blood samples from cancer patients with hB6H12.3 or Vel-IPV-hB6H12.3 for 20 hours at 37°C. In fig. 46A depicts the production of IP-10, and FIG. 46B depicts the production of IL-1RA.

In fig. 47 depicts Annexin V staining of Ht1080 tumor cells from the HT1080 tumor xenograft model in mice treated with hB6H12.3, Vel-IPV-hB6H12.3, or the hIgG1 isotype control (h00 isotype).

Detailed description of the invention

Anti-CD47 IgG3 antibodies known in the art at the time of application exhibit toxicity, such as a decrease in the number of red blood cells and platelets in the peripheral blood, which reduces their usefulness as effective therapeutic agents against diseases associated with CD47, such as cancer diseases characterized by CD47 expression. Applicants have unexpectedly discovered novel anti-CD47 IgG1 antibodies and antigen binding fragments thereof that can be activated by deshielding in the context of the tumor microenvironment to effectively target the antibodies and antigen binding fragments thereof of the present invention specifically to solid tumors expressing CD47. The humanized anti-CD47 antibodies and antigen-binding fragments thereof (either shielded or unshielded) described herein are useful for the treatment of diseases associated with CD47, such as cancers characterized by CD47 expression.

Some illustrative embodiments of the present invention provide antibodies and antigen binding fragments thereof that comprise a removable shielding moiety (eg, a coiled-coil barrier) that blocks binding of the antibody or antigen binding fragment thereof to its antigenic target. In some embodiments of the present invention, a removable shielding moiety is attached to the N-terminus of one or more heavy and/or light chains of an antibody or antigen binding fragment thereof via a matrix metalloproteinase (MMP) cleavable linker sequence.

In the tumor microenvironment, altered proteolysis leads to unregulated tumor growth, tissue remodeling, inflammation, tissue invasion and metastasis (Kessenbrock (2011) Cell 141:52). MMPs are the most prominent family of proteinases associated with tumorigenesis, and MMPs mediate many changes in the tumor microenvironment during tumor progression. Additionally. When an antibody or antigen-binding fragment thereof of this invention is exposed to MMP, the MMP linker sequence is cleaved, allowing the screen to be removed.

- 15 045361 binding fragment of the superhelix and enable the antibody or its antigen-binding fragment to bind its target antigen in a manner specific to the tumor microenvironment.

The novel anti-CD47 IgG1 antibodies and antigen-binding fragments thereof of this invention (both shielded and unshielded) advantageously demonstrate increased pharmacokinetics and reduced target effects compared to anti-CD47 IgG3 antibodies known in the art at the time of filing. The novel humanized anti-CD47 antibodies described herein advantageously exhibit one or more of: 1) enhanced antigen binding compared to a reference antibody (eg, the parent mouse antibody); 2) increased antibody-dependent cellular cytotoxicity (ADCC) compared to a reference antibody (eg, the parent mouse antibody); 3) increased phagocytosis (eg, enhanced antibody-dependent cellular phagocytosis (ADCP)) compared to a reference antibody (eg, parental mouse antibody); 4) decreased erythrocyte hemagglutination (HA) compared to a reference antibody (eg, the original mouse antibody); 5) binding to the same three-dimensional (i.e. non-linear) epitope of CD47.

To make the invention easier to understand, certain technical and scientific terms are specifically defined below. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this invention relates.

I. Definitions

As used in the context of this document, including the appended claims, singular nouns include the corresponding plural nouns unless the context clearly indicates otherwise.

An antibody-drug conjugate refers to an antibody conjugated to a cytotoxic agent or a cytostatic agent. Typically, antibody-drug conjugates bind to a target antigen (eg, CD47) on the cell surface, followed by internalization of the antibody-drug conjugate into the cell and subsequent release of the drug into the cell.

A polypeptide or polypeptide chain is a polymer of amino acid residues linked by peptide bonds, whether produced naturally or synthetically. Polypeptides with about less than 10 amino acid residues are usually called peptides.

A protein is a macromolecule containing one or more polypeptide chains. The protein may also contain non-peptide components such as carbohydrate groups. The cell in which the protein is produced may add carbohydrates and other non-peptide substituents, and these will vary depending on the cell type. Proteins are defined herein in terms of their amino acid backbone structures. Substituents such as carbohydrate groups are usually not specified but may nevertheless be present.

The terms amino-terminus and carboxy-terminus refer to positions within polypeptides. Where context permits, these terms are used with reference to a specific sequence or portion of a polypeptide to indicate proximity or relative position. For example, a particular sequence located at the carboxy terminus of a control sequence in a polypeptide is located proximal to the carboxy terminus of the control sequence, but is not necessarily located at the carboxy terminus of the complete polypeptide.

For purposes of classifying amino acid substitutions as conservative or non-conservative, the following amino acid substitutions are considered conservative substitutions: serine substituted with threonine, alanine or asparagine; threonine substituted with proline or serine; asparagine substituted with aspartic acid, histidine or serine; aspartic acid substituted with glutamic acid or asparagine; glutamic acid substituted with glutamine, lysine or aspartic acid; glutamine substituted with arginine, lysine or glutamic acid; histidine substituted with tyrosine or asparagine; arginine substituted with lysine or glutamine; methionine substituted with isoleucine, leucine or valine; isoleucine substituted with leucine, valine or methionine; leucine substituted with valine, isoleucine or methionine; phenylalanine substituted with tyrosine or tryptophan; tyrosine substituted with tryptophan, histidine or phenylalanine; proline substituted with threonine; alanine substituted with serine; lysine substituted with glutamic acid, glutamine or arginine; valine substituted with methionine, isoleucine or leucine; and tryptophan substituted with phenylalanine or tyrosine. Conservative substitutions can also mean substitutions between amino acids in the same class. The classes are as follows: Group I (hydrophobic side chains): met, ala, val, leu, ile; Group II (neutral hydrophilic side chains): cys, ser, thr; Group III (acidic side chains): asp, glu; Group IV (main side chains): asn, gin, his, lys, arg; Group V (residues affecting chain orientation): gly, pro; and group VI (aromatic side chains): trp, tyr, phe.

Two amino acid sequences have 100% amino acid sequence identity if the amino acid residues of the two amino acid sequences are the same when aligned for maximum matching. Sequence comparisons can be made

- 16 045361 using standard programs such as those included in the LASERGENE bioinformatics computing package produced by DNASTAR (Madison, Wisconsin). Other methods for comparing two nucleotide or amino acid sequences by determining optimal alignment are well known to those skilled in the art. (See, for example, Peruski and Peruski, The Internet and the New Biology: Tools for Genomic and Molecular Research (ASM Press, Inc. 1997); Wu et al. (eds.), Information Superhighway and Computer Databases of Nucleic Acids and Proteins, in Methods in Gene Biotechnology 123-151 (CrC Press, Inc. 1997); Bishop (ed.), Guide to Human Genome Computing (2nd ed., Academic Press, Inc. 1998).) Two amino acid sequences are believed to have significant sequence identity if two sequences have at least about 80%, at least about 85%, at least about 90%, or at least about 95% sequence identity relative to each other.

The percentage of sequence identity is determined using antibody sequences maximally aligned according to the Kabat numbering convention. After alignment, if a region of a test antibody (e.g., the entire variable domain of a heavy or light chain) is compared to the same region of a reference antibody, the percent sequence identity between the regions of the test antibody and the reference antibody is the number of positions occupied by the same amino acid in the region of both test and reference antibodies, divided by the total number of aligned positions of the two regions, excluding gaps, multiplied by 100 to convert to a percentage.

Compositions or methods containing one or more of the listed elements may include other elements not specifically listed. For example, a composition that contains an antibody may contain the antibody alone or in combination with other components.

The value range designation includes all integers within or defining the range.

In the antibodies or other proteins described herein, reference to amino acid residues corresponding to the residues specified in SEQ ID NO includes post-translational modifications of such residues.

The term antibody refers to immunoglobulin proteins produced by the body in response to the presence of an antigen and that bind to the antigen, as well as antigen-binding fragments and genetically engineered variants thereof. Therefore, the term antibody includes, for example, intact monoclonal antibodies (eg, antibodies produced using hybridoma technology) and antigen binding antibody fragments such as F(ab') ₂ , Fv fragment, diabody, single chain antibody, scFv fragment or scFv-Fc . Also included are genetically engineered intact antibodies and fragments, such as chimeric antibodies, humanized antibodies, single chain Fv fragments, single chain antibodies, diabodies, minibodies, linear epitope recognition antibodies, multivalent or multispecific (eg, bispecific) hybrid antibodies, and the like. Thus, the term antibody is broadly used to include any protein that contains the antigen-binding site of an antibody and is capable of specifically binding to its antigen.

The term antibody or antigen binding fragment thereof includes a naked antibody or antigen binding fragment thereof that is not bound (ie not covalently or non-covalently linked) to the shielding compound of the present invention. The term antibody also includes a shielded antibody or antigen-binding fragment thereof that is covalently or non-covalently linked to one or more shielding compounds, such as, for example, coiled-coil peptides, as described later herein. The term antibody or antigen-binding fragment thereof includes a conjugated antibody or antigen-binding fragment thereof or an antibody-drug conjugate (ADC) in which the antibody or antigen-binding fragment thereof is covalently or non-covalently linked to a pharmaceutical agent, such as a cytostatic or cytotoxic drug. In some embodiments of the present invention, the antibody or antigen binding fragment thereof is a naked antibody or antigen binding fragment that is optionally conjugated to a pharmaceutical agent, such as a cytostatic or cytotoxic drug. In other embodiments of the present invention, the antibody or antigen binding fragment thereof is a shielded antibody or antigen binding fragment that is optionally conjugated to a pharmaceutical agent, such as a cytostatic or cytotoxic drug.

The term genetically engineered antibodies refers to an antibody in which the amino acid sequence differs from the sequence of the native or parent antibody. The possible variations are many and range from substituting only one or more amino acids to completely changing the design, such as the variable or constant region. Changes in the constant region are typically made to improve or change characteristics, such as complement fixation reactions and other effector functions. Typically, changes in the variable region are made to improve antigen-binding characteristics, improve the stability of the variable

- 17 045361 white area and/or reducing the risk of immunogenicity.

The term chimeric antibody refers to an antibody in which part of the heavy and/or light chain is identical or homologous to corresponding sequences in an antibody derived from a particular species (eg, human) or belonging to a particular class or subclass of antibodies, while the remainder of the chain(s) is identical or homologous to corresponding sequences in an antibody derived from another species (eg, mouse) or belonging to another class or subclass of antibodies, as well as fragments of such antibodies, provided that they exhibit the desired biological activity.

The antigen-binding site of an antibody is that portion of the antibody that is sufficient to bind it to an antigen. The minimal such region is usually a variable domain or a genetically engineered variant thereof. Single domain binding sites can be derived from camel antibodies (see Muyldermans and Lauwereys, Mol. Recog. 12: 131-140, 1999; Nguyen et al., EMBO J. 19:921-930, 2000) or from VH domains other species to produce single domain antibodies (dAbs, see Ward et al., Nature 341:544-546, 1989; US Patent No. 6,248,516 to Winter et al). Typically, the antigen binding site of an antibody contains both a heavy chain variable domain (VH) and a light chain variable domain (VL) that bind to a common epitope. In the context of the present invention, an antibody may include one or more components in addition to an antigen binding site, such as, for example, a second antigen binding site of the antibody (which may bind the same or a different epitope or the same or a different antigen), a peptide linker, an immunoglobulin constant region , immunoglobulin hinge, amphipathic helix (see Pack and Pluckthun, Biochem. 31: 1579-1584, 1992), non-peptide linker, oligonucleotide (see Chaudri et al, FEBS Letters 450:23-26, 1999), cytostatic or cytotoxic drug agent and the like, and may be a monomeric or multimeric protein. Examples of molecules containing an antibody antigen binding site are known in the art and include, for example, Fv, single chain Fv (scFv), Fab, Fab', F(ab') ₂ , F(ab)c, diabodies, minibodies, nanoantibodies, Fab-scFv, bispecific (scFv)4-IgG, and bispecific (scFv)2-Fab fusions. (See, for example, Hu et al., Cancer Res. 56:30553061, 1996; Atwell et al., Molecular Immunology 33: 1301-1312, 1996; Carter and Merchant, Curr. Op. Biotechnol. 8:449-454, 1997; Zuo et al., Protein Engineering 13:361-367, 2000; and Lu et al., J. Immunol. Methods 267:213-226, 2002.)

The term immunoglobulin refers to a protein consisting of one or more polypeptides, primarily encoded by the immunoglobulin gene(s). One form of immunoglobulin represents the basic structural unit of native (that is, natural or parent) antibodies in vertebrates. This form is a tetramer and consists of two identical pairs of immunoglobulin chains, each with one light chain and one heavy chain. In each pair, the light and heavy chain variable regions (VL and VH) together are primarily responsible for antigen binding, and the constant regions are primarily responsible for the effector functions of the antibody. Five classes of immunoglobulin proteins have been identified in higher vertebrates (IgG, IgA, IgM, IgD and IgE). IgG is a major class and is typically found in plasma as the second most abundant protein. In humans, IgG consists of four subclasses, designated IgG1, IgG2, IgG3 and IgG4. Each immunoglobulin heavy chain possesses a constant region that consists of constant region protein domains (CH1, hinge, CH2, and CH3; IgG3 also contains a CH4 domain) that are essentially invariant for a given subclass within a species.

DNA sequences encoding human and non-human immunoglobulin chains are known in the art. (See, for example, Ellison et al., DNA 1:11-18, 1981; Ellison et al., Nucleic Acids Res. 10:4071-4079, 1982; Kenten et al., Proc. Natl. Acad. Set USA 79 :6661-6665, 1982; Seno et al., Nucl. Acids Res. 11:719-726, 1983; Riechmann et al., Nature 332:323-327, 1988; Amster et al., Nucl. Acids Res. 8 :20552065, 1980; Rusconi and Kohler, Nature 314:330-334, 1985; Boss et al., Nucl. Acids Res. 12:3791-3806, 1984; Bothwell et al., Nature 298:380-382, 1982; van der Loo et al., Immunogenetics 42:333-341, 1995; Karlin et al., J. Mol. Evol. 22: 195-208, 1985; Kindsvogel et al., DNA 1:335-343, 1982; Breiner et al., Gene 18: 165-174, 1982; Kondo et al., Eur J Immunol 23:245-249, 1993; and GenBank Accession No. J00228.) For the structure and function of immunoglobulins, see Putnam , The Plasma Proteins, Vol V, Academic Press, Inc., 49-140, 1987; and Padlan, Mol. Immunol. 31: 169-217, 1994. The term immunoglobulin is used herein for its general meaning to mean the intact antibody, components of its chains or fragments of chains, depending on the context.

Full-length immunoglobulin light chains (about 25 kDa or 214 amino acids) are encoded by a variable region gene at the amino terminus (encoding about 110 amino acids) and a kappa or lambda constant region gene at the carboxyl terminus. Full-length immunoglobulin heavy chains (about 50 kDa or 446 amino acids) are encoded by a variable region gene (encoding about 116 amino acids) and a gamma, mu, alpha, delta or epsilon constant region gene (encoding about 330 amino acids), with the latter defining the antibody isotype as IgG , IgM, IgA, IgD or IgE, respectively. Within the light and heavy chains, the variable and constant regions are connected by a J region of about 12 or more amino acids, with the heavy chain also including a D region of about another

- 18045361 amino acids. (See generally Fundamental Immunology (Paul, ed., Raven Press, N.Y., 2nd ed. 1989), Ch. 7).

An immunoglobulin light or heavy chain variable region (also referred to herein as a light chain variable domain (VL domain) or a heavy chain variable domain (VH domain), respectively) consists of a framework region interrupted by three complementarity determining regions or CDRs. Framework regions serve to align CDRs for specific binding to an antigen epitope. Thus, the term CDR refers to the amino acid residues of an antibody that are primarily responsible for antigen binding. From the amino terminus to the carboxyl terminus, both VL and VH domains contain the following framework (FR) and CDR regions: FR1, CDR1, FR2, CDR2, FR3, CDR3, FR4.

The definition of amino acid positions in each variable region domain follows the definitions according to Kabat, Sequences of Proteins of Immunological Interest (National Institutes of Health, Bethesda, MD, 1987 and 1991). The Kabat method also provides for a commonly used numbering convention (Kabat numbering) in which corresponding residues between different heavy chain variable regions or between different light chain variable regions are assigned the same number. CDRs 1, 2 and 3 of the VL domain are also referred to herein as CDRL1, CDR-L2 and CDR-L3, respectively. CDRs 1, 2 and 3 of the VH domain are also referred to herein as CDR-H1, CDR-H2 and CDR-H3, respectively. If so noted, the determination of CDR may be according to IMGT® (Lefranc et al., Developmental & Comparative Immunology 27:55-77; 2003) instead of the Kabat method.

Numbering of the heavy chain constant region is done using the EC index, as indicated in Kabat (Sequences of Proteins of Immunological Interest, National Institutes of Health, Bethesda, MD, 1987 and 1991).

Unless the context otherwise indicates, the term monoclonal antibody is not limited to antibodies produced using hybridoma technology. The term monoclonal antibody may include an antibody derived from a single clone, including any eukaryotic, prokaryotic or phage clone. In specific embodiments of the present invention, the antibodies described herein are monoclonal antibodies.

The term humanized VH domain or humanized VL domain refers to a VH or VL domain of an immunoglobulin comprising some or all of the CDRs in whole or in large part from a non-human donor immunoglobulin (eg, mouse or rat), and the entire variable domain framework sequences or substantially from human immunoglobulin sequences. The non-human immunoglobulin providing the CDR is called a donor, and the human immunoglobulin providing the framework sequence is called an acceptor. In some cases, humanized antibodies retain some non-human residues in the framework regions of the human variable domain to improve proper binding characteristics (for example, mutations in the framework sequences may be required to maintain binding affinity when the antibody is humanized).

A humanized antibody is an antibody containing one or both of a humanized VH domain and a humanized VL domain. Immunoglobulin constant region(s) do not necessarily have to be present, but if present, they are wholly or substantially human immunoglobulin constant regions.

A humanized antibody is a genetically engineered antibody in which CDRs from a non-human donor antibody are grafted onto human acceptor antibody sequences (see, for example, Queen, US 5,530,101 and 5,585,089; Winter, US 5,225,539; Carter, US 6,407,213; Adair , US 5,859,205; and Foote, US 6,881,557). The acceptor antibody sequences may be, for example, a mature human antibody sequence, a composition of such sequences, a consensus sequence of human antibody sequences, or a germline region sequence. Human acceptor sequences can be selected for high sequence identity in variable region frameworks with donor sequences to match the canonical shapes between the acceptor and donor CDRs, among other criteria. Thus, a humanized antibody is an antibody having CDRs wholly or substantially from donor antibodies and variable region framework sequences and constant regions, if present, wholly or substantially from human antibody sequences. Likewise, a humanized heavy chain typically has all three CDRs derived entirely or substantially from the donor antibody heavy chain, and the heavy chain variable region framework sequence and the heavy chain constant region, if present, substantially from the human heavy chain variable region framework sequence. and human antibody heavy chain constant region sequences. Likewise, a humanized light chain typically has all three CDRs derived entirely or substantially from the donor antibody light chain and the light chain variable region framework sequence and the light chain constant region, if present, substantially from the human light chain variable region framework sequence and sequence human light chain constant region. CDR in a humanized antibody in

- 19 045361 are substantially derived from the corresponding CDR in a non-human antibody, with at least about 80%, about 81%, about 82%, about 83%, about 84%, about 85%, about 86 %, about 87%, about 88%, about 89%, about 90%, about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98% or about 99% of the corresponding residues (as determined by Kabat numbering), or about 100% of the corresponding residues (as determined by Kabat numbering) are identical between the corresponding CDRs. The antibody variable chain or antibody constant region framework sequences are substantially derived from the human variable region or human antibody constant region framework sequence, respectively, with at least about 80%, about 81%, about 82%, about 83% , about 84%, about 85%, about 86%, about 87%, about 88%, about 89%, about 90%, about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98%, or about 99% of the corresponding residues (as determined by Kabat numbering for the variable region and EU numbering for the constant region), or about 100% of the corresponding residues (as determined by Kabat numbering for the variable region and numbering EU for constant region) are identical.

Although humanized antibodies often include all six CDRs (preferably as defined by Kabat or IMGT®) from a murine antibody, they can also be prepared using fewer than all six CDRs (e.g., at least 3, 4, or 5) CDRs from mouse antibody (eg, Pascalis et al., J. Immunol. 169:3076, 2002; Vajdos et al., Journal of Molecular Biology, 320: 415-428, 2002; Iwahashi et al., Mol. Immunol. 36:1079 -1091, 1999; Tamura et al, Journal of Immunology, 164: 14321441, 2000).

The CDR in the humanized antibody is derived substantially from the corresponding CDR in the non-human antibody, with at least 60%, at least 85%, at least 90%, at least 95%, or 100% of the corresponding residues (as determined by Kabat (or IMGT)) are identical between the corresponding CDRs. In particular variations of a humanized VH or VL domain in which the CDRs are substantially derived from non-human immunoglobulin, the humanized VH or VL domain CDRs have no more than six (e.g., no more than five, no more than four, no more than three, no more than two or no more than one) amino acid substitutions (preferably conservative substitutions) in all three CDRs relative to the corresponding non-human VH or VL CDRs. The antibody VH or VL variable domain framework sequences, or, if present, the immunoglobulin constant region sequences, are substantially derived from the human VH or VL framework sequence or the human antibody constant region, respectively, with at least about 80% , about 81%, about 82%, about 83%, about 84%, about 85%, about 86%, about 87%, about 88%, about 89%, about 90%, about 91%, about 92%, about 93%, about 94%, about 95%, about 96%, about 97%, about 98%, or about 99% of the corresponding residues (as defined by Kabat numbering for the variable region and EU numbering for the constant region), or about 100% of the corresponding residues (as defined by Kabat numbering for the variable region and EU numbering for the constant region) are identical. Therefore, all portions of a humanized antibody, with the exception of the CDR, are generally derived entirely or substantially from corresponding portions of naturally occurring human immunoglobulin sequences.

Antibodies are usually offered in isolated form. This means that the antibody typically has a purity of at least about 50% w/w. from interfering proteins and other contaminants resulting from its manufacture or purification, but does not exclude the possibility that the antibody is combined with an excess of pharmaceutically acceptable carrier(s) or other excipient intended to facilitate use of the antibody. Sometimes the antibodies are at least about 60%, about 70%, about 80%, about 90%, about 95%, or about 99% w/w purity. from interfering proteins and contaminants resulting from production or purification. Antibodies, including isolated antibodies, can be conjugated to cytotoxic agents and presented as drug-antibody conjugates and/or shielded, for example, using appropriate supercoils.

The specific binding of an antibody to its target antigen typically has an affinity of at least about 10 ⁶ , about 10 ⁷ , about 10 ⁸ , about 10 ⁹ or about 10 ¹⁰ M ^-1 . Specific binding is markedly higher in magnitude and differs from nonspecific binding, which occurs to at least one nonspecific target. Specific binding may result from the formation of bonds between specific functional groups or a specific spatial correspondence (eg, type of lock and key), whereas nonspecific binding is usually the result of van der Waals forces.

The term epitope refers to the site of an antigen to which an antibody binds. An epitope can be formed from contiguous amino acids or non-contiguous amino acids located nearby by folding into the tertiary structure of one or more proteins. Epitopes formed from contiguous amino acids are usually retained when exposed to denaturing agents such as solvents, whereas epitopes formed by protein folding into a tertiary structure are usually lost when exposed to

- 20 045361 action of denaturing agents, for example solvents. An epitope typically comprises at least about 3 or more, typically at least about 5, at least about 6, at least about 7, or about 8-10 amino acids in a unique spatial conformation. Methods for determining the spatial conformation of epitopes include, for example, x-ray crystallography and two-dimensional nuclear magnetic resonance. See, for example, Epitope Mapping Protocols, in Methods in Molecular Biology, Vol. 66, Glenn E. Morris, Ed. (1996).

Antibodies that recognize the same or overlapping epitopes can be identified by a simple immunoassay showing the ability of one antibody to compete with the binding of another antibody to the target antigen. An antibody epitope can also be determined by X-ray crystallography of the antibody bound to its antigen to identify contact residues.

Alternatively, two antibodies have the same epitope if all amino acid mutations in the antigen that reduce or eliminate the binding of one antibody reduce or eliminate the binding of the other (provided that such mutations do not cause a global change in the structure of the antigen). Two antibodies have overlapping epitopes if certain amino acid mutations that reduce or eliminate the binding of one antibody reduce or eliminate the binding of the other antibody.

Competition between antibodies can be determined using an assay in which a test antibody inhibits the specific binding of a reference antibody to a common antigen (see, for example, Junghans et al., Cancer Res. 50: 1495, 1990). A test antibody competes with a reference antibody if an excess of test antibody inhibits binding of the reference antibody.

Antibodies identified by competition analysis (competitive antibodies) include antibodies that bind to the same epitope as the reference antibody and antibodies that bind to an adjacent epitope sufficiently proximal to the epitope bound by the reference antibody to cause a steric mismatch . Antibodies identified by a competition assay also include antibodies that indirectly compete with a reference antibody by causing a conformational change in the target protein, thereby preventing the reference antibody from binding to an epitope different from that associated with the test antibody.

The effector function of an antibody refers to the function provided by the Fc region of the Ig. Such functions may be, for example, antibody-dependent cellular cytotoxicity (ADCC), antibody-dependent cellular phagocytosis (ADCP) or complement-dependent cytotoxicity (CDC). Such a function may be accomplished, for example, by binding the Fc region to an Fc receptor on an immune cell with phagocytic or lytic activity, or by binding the Fc region to components of the complement system. Typically, the effect(s) mediated by Fc-binding cells or complement components results in inhibition and/or reduction in the number of CD47-targeted cells. Fc regions of antibodies can recruit cells expressing the Fc receptor (FcR) and match them with antibody-coated target cells. Cells expressing surface FcR for IgG, including FcyRIII (CD16), FcyRII (CD32), and FcyRIII (CD64), can act as effector cells to destroy IgG-coated cells. Such effector cells include monocytes, macrophages, natural killer (NK) cells, neutrophils, and eosinophils. Binding of FcyR IgG activates ADCC or ADCP. ADCC is mediated by CD16+ effector cells through the secretion of membrane pore-forming proteins and proteases, while phagocytosis is mediated by CD32+ and CD64+ effector cells (see Fundamental Immunology, ^4th ed., Paul ed., Lippincott-Raven, NY, 1997, Chapters 3, 17 and 30; Uchida et al., J. Exp. Med. 199:1659-69, 2004; Akewanlop et al., Cancer Res. 61:4061-65, 2001; Watanabe et al., Breast Cancer Res. Treat. 53: 199-207, 1999).

In addition to ADCC and ADCP, the Fc regions of cell-bound antibodies can also activate the classical complement pathway to cause ADCC. Clq of the complement system binds to the F regions of antibodies when they form a complex with antigens. Binding of Clq to cellular antibodies can initiate a cascade of events including proteolytic activation of C4 and C2 to generate C3 convertase. Cleavage of C3 into C3b by C3 convertase allows activation of terminal complement components, including C5b, C6, C7, C8 and C9. Together, these proteins form a membrane attack complex that forms pores on antibody-coated cells. These pores disrupt the integrity of the cell membrane, killing the target cell (see Immunobiology, ^6th ed., Janeway et al, Garland Science, NY, 2005, Chapter 2).

The term antibody-dependent cellular cytotoxicity or ADCC refers to a mechanism of cell death induction that depends on the interaction of antibody-coated target cells with lytic immune cells (also called effector cells). Such effector cells include natural killer cells, monocytes/macrophages and neutrophils. Effector cells attach to the Fc region of Ig, which is associated with target cells through their antigen-binding sites. Death of the antibody-coated target cell occurs as a result of the activity of effector cells. In some illustrative embodiments of the present invention

- 21 045361 the anti-CD47 IgG1 antibody of this invention mediates the same or increased ADCC relative to the parent antibody and/or relative to the anti-CD47 IgG3 antibody.

The term antibody-dependent cellular phagocytosis or ADCP refers to the process by which antibody-coated cells are engulfed, in whole or in part, by phagocytic immune cells (eg, macrophages, neutrophils and/or dendritic cells) that bind to the Fc region of the Ig. In some illustrative embodiments of the present invention, the anti-CD47 IgG1 antibody of the present invention mediates equal or increased ADCP relative to the parent antibody and/or relative to the anti-CD47 IgG3 antibody.

The term complement-dependent cytotoxicity or CDC refers to a cell death induction mechanism in which the Fc region of a target-bound antibody activates a series of enzymatic reactions that culminate in the formation of holes in the target cell membrane.

Typically, antigen-antibody complexes, such as those on target cells coated with antibodies, bind and activate the complement component C1q, which in turn activates the complement cascade leading to the death of the target cells. Complement activation can also lead to the deposition of complement components on the surface of the target cell, which promote ADCC by binding complement receptors (eg, CR3) on leukocytes.

The cytotoxic effect refers to the reduction in the number, elimination and/or destruction of the target cell. A cytotoxic agent refers to a compound that has a cytotoxic effect on a cell, thereby mediating reduction in the number, elimination and/or destruction of the target cell. In some embodiments of the present invention, the cytotoxic agent is conjugated to an antibody or administered in combination with an antibody. Suitable cytotoxic agents are described later herein.

The cytostatic effect refers to the inhibition of cell proliferation. A cytostatic agent refers to a compound that has a cytostatic effect on a cell, thereby mediating the inhibition of the growth and/or proliferation of a particular cell type and/or subpopulation of cells. Suitable cytostatic agents are described later in this document.

The terms expression unit and expression cassette are used interchangeably herein to refer to a nucleic acid segment encoding a polypeptide of interest and capable of causing expression of the nucleic acid segment in a host cell. The expression unit typically contains a transcription promoter, an open reading frame encoding the polypeptide of interest, and a transcription terminator operably linked. In addition to the promoter and transcription terminator, the expression unit may further include other nucleic acid segments, such as, for example, an enhancer or a polyadenylation signal.

The term expression vector refers to a nucleic acid molecule, linear or cyclic, containing one or more expression units. In addition to one or more expression units, the expression vector may also include additional nucleic acid segments, such as, for example, one or more origins of replication or one or more selectable markers.

Expression vectors are typically derived from plasmid or viral DNA or may contain elements of both.

The term patient or subject includes humans and other mammals, such as non-human primates, rabbits, rats, mice and the like, and transgenic species thereof, which are subject to either prophylactic or therapeutic treatment. In some illustrative embodiments of the present invention, the subject is a human patient suffering from or at risk of developing a cancer, such as a solid tumor, that optionally secretes one or more proteases capable of cleaving the shielding domain (e.g., the supercoiled shielding domain) of the anti-CD47 antibody described in this document.

The term effective amount, in the context of treating a disease characterized by CD47 expression by administering an anti-CD47 antibody as described herein, refers to an amount of such antibody that is sufficient to inhibit the occurrence or amelioration of one or more symptoms of a CD47-related disease (eg, cancer characterized by the expression of CD47). An effective amount of antibody is administered in an effective schedule. The term effective regimen refers to a combination of an amount of antibody administered and a dosing frequency adequate to provide prophylactic or therapeutic treatment for a disease (eg, prophylactic or therapeutic treatment for a cancer characterized by CD47 expression).

The term pharmaceutically acceptable preferably refers to a compound that is licensed for medical use or approved by a regulatory agency of the Federal or State government or included in the United States Pharmacopoeia or other generally accepted pharmacopoeia. The term pharmaceutically compatible component refers to a pharmaceutically acceptable diluent, adjuvant, excipient or excipient with which an anti-CD47 antibody is formulated.

The phrase pharmaceutically acceptable salt refers to pharmaceutically acceptable organic or inorganic salts. Typical salts include sulfate, citrate, acetate, oxalate

- 22 045361 chloride, bromide, iodide, nitrate, bisulfate, phosphate, hydrophosphate, isonicotinate, lactate, salicylate, citrate, tartrate, oleate, tannate, pantothenate, bitartrate, ascorbate, succinate, maleate, gentisinate, fumarate, gluconate, glucuronate , saccharate, formate, benzoate, glutamate, methanesulfonate, ethanesulfonate, benzenesulfonate, p-toluenesulfonate and pamoate (i.e. 1,1'-methylene-bis-(2-hydroxy-3naphthoate)) salts. The pharmaceutically acceptable salt may further contain an additional molecule, such as, for example, an acetate ion, succinate ion, or other counterion. The counterion can be any organic or inorganic moiety that stabilizes the charge of the parent compound. In addition, a pharmaceutically acceptable salt may have more than one charged atom in its structure. In cases where multiple charged atoms are part of a pharmaceutically acceptable salt, they may have multiple counterions. Therefore, a pharmaceutically acceptable salt may have one or more charged atoms and/or one or more counterions.

Unless otherwise obvious from the context, when a value is expressed as being about X or approximately X, the stated value of X will be taken to be accurate to within ±10%.

Solvates in the context of the invention are those forms of the compounds of this invention that form a complex in the solid or liquid state in coordination with solvent molecules. Hydrates are one special form of solvate in which coordination occurs with water. In some illustrative embodiments of this invention, solvates in the context of this invention are hydrates.

The term hemagglutination refers to the process by which red blood cells clump together. Hemagglutination is a known unwanted side effect of anti-CD47 antibodies in the art. The humanized anti-CD47 antibodies of the present invention optionally mediate reduced hemagglutination compared to the mouse parental anti-CD47 antibody and/or compared to one or more anti-CD47 antibodies known in the art.

II. Antibodies against CD47 and antigen binding fragments

This invention provides isolated, recombinant and/or synthetic antibodies against human, primate, rodent, mammalian CD47, chimeric, humanized and/or CDR-grafted antibodies and antigen-binding fragments thereof, as well as nucleic acid compositions and molecules containing at least one polynucleotide , encoding at least a portion of one anti-CD47 antibody molecule. The invention further includes, but is not limited to, methods for producing and using such nucleic acids and antibodies, including diagnostic and therapeutic compositions, methods and articles. Some illustrative embodiments of the present invention provide humanized anti-CD47 IgG1 antibodies. Other exemplary embodiments of the present invention provide anti-CD47 IgG1 shielded antibodies. Other illustrative embodiments of the present invention provide humanized anti-CD47 IgG1 antibodies comprising a fragment of cleavage site amino acid residues.

Specific embodiments of the present invention provide humanized anti-CD47 antibodies having one or more of the following activities: 1) enhancing antigen binding compared to a reference antibody (eg, a parent mouse antibody); 2) increased antibody-dependent cellular cytotoxicity (ADCC) compared to a reference antibody (eg, the parent mouse antibody); 3) increased phagocytosis (eg, antibody-dependent cellular phagocytosis (ADCP)) compared to a reference antibody (eg, parent mouse antibody); 4) decreased erythrocyte hemagglutination (HA) compared to a reference antibody (eg, the original mouse antibody); 5) binding to the three-dimensional (i.e. non-linear) epitope of CD47.

Exemplary anti-CD47 antibodies and antigen-binding fragments thereof of this invention include the following CD47 antibody heavy chain/light chain pairs: hB6H12.1-hvH1/hvK1; hB6H12.2-hvH1/hvK2; hB6H12.3-hvH1/hvK3; hB6H12.4-hvH1/hvK4; hB6H12.5-hvH2/hvK1; hB6H12.6hvH2/hvK2; hB6H12.7-hvH2/hvK3; hB6H12.8-hvH2/hvK4; hB6H12.9-hvH3/hvK1; hB6H12.10-hvH3/hvK2; hB6H12.11-hvH3/hvK3; hB6H12.12-hvH3/hvK4; hB6H12.13-hvH4/hvK1; hB6H12.14-hvH4/hvK2;

hB6H12.15-hvH4/hvK3; hB6H12.16-hvH4/hvK4; hB6H12.17-hvH5/hvK1; hB6H12.18-hvH5/hvK2;

hB6H12.19-hvH5/hvK3; hB6H12.20-hvH5/hvK4; hB6H12.21-hvH6/hvK1; hB6H12.22-hvH6/hvK2;

hB6H12.23-hvH6/hvK3; hB6H12.24-hvH6/hvK4; hB6H12.3 (deamidated mutant)-hvH1/hvK3 G91A; Ab47-HV3-7/HJ4/KV3D-11/KJ1; and mB6H12-vH1/vL. Exemplary sequences of the anti-CD47 antibody heavy chain variable region, light chain variable regions, heavy chain CDRs, and light chain CDRs can be found in Tables 1-6. The amino acid sequences for the heavy chain and light chain of a typical humanized anti-CD47 antibody can be found in Table 1. 7.

- 23 045361

Table 1. Heavy chain variable sequences obtained from mouse antibody B6H12. Kabat CDRs are underlined and IMGT CDRs are in bold.

Table 1

Option Subsequence mB6hl2 vH EVQLVESGGDLVKPGGSLKLSCAASGFTFSGYGMSWVRQTPDKRLE WVATITSGGTYTYYPDSVKGRFTISRDNAKNTLYLQIDSLKSEDTAIY FCARSLAGNAMDYWGQGTSVTVSS (SEQ ID NO: 1) Ab47vH (HV3-7/HJ4) EVOLVESGGGLVQPGGSLRLSCAASGFTFSGYGMSWVROAPGKGLE WVATITSGGTYYPDSVKGRFTISRDNAKNSLYLQMNSLRAEDTAV YYCARSLAGNAMDYWGOGTLVTVSS (SEQ ID NO: 2) hvHl EVOLLESGGGLVQPGGSLRLSCAASGFTFSGYGMSWVROAPGKRLE WVATITSGGTYYPDSVKGRFTISRDNSKNTLYLQMNSLRAEDTAIY FCARSLAGNAMDYWGOGTLVTVSS (SEQ ID NO: 3) hvH2 EVOLLESGGGLVQPGGSLRLSCAASGFFSSYAMSWVROAPGKGLEW VATITSGGTYTYYADSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVY FCARSLAGNAMDYWGOGTLVTVSS (SEQ ID NO: 4) hvH3 EVOLLESGGGLVQPGGSLRLSCAASGFFSSYAMSWVROAPGKGLEW VATITSGGTYTYYADSVKGRFTISRDNSKNTLYLOINSLRAEDTAVYF CARSLAGNAMDYWGOGTLVTVSS (SEQ ID NO: 5) hvH4 EVOLVESGGGLVQPGGSLRLSCAASGFFSSYGMNWVROAPGKGLE WVATITSGGTYIYYADSVKGRFTISRDNAKNSLYLQMNSLRAEDTAV YYCARSLAGNAMDYWGOGTLVTVSS (SEQ ID NO: 6) hvH5 EVOLVESGGGLVQPGGSLRLSCAASGFTFSGYGMSWVROAPGKGLE WVATITSGGTYYADSVKGRFTISRDNSKNTLYLOINSLRAEDTAVY YCARSLAGNAMDYWGOGTLVTVSS (SEQ ID NO: 7) hvH6 EVOLVESGGGLVOPGGSLRLSCAASGFTFSSYGMHWVROAPGKGLV WVATITSGGTYTSYADSVKGRFTISRDNAKNTLYLOMNSLRAEDTAV YYCARSLAGNAMDYWGOGTLVTVSS (SEQ ID NO: 8)

Table 2. Light chain variable sequences derived from mouse antibody B6H12. Kabat CDRs are underlined and IMGT CDRs are in bold.

table 2

Option Subsequence mB6hl2 vL DIVMTOSPATLSVTPGDRVSLSCRASOTISDYLHWYOQKSHESPRLLIK FASQSISGIPSRFSGSGSGSDFTLSINSVEPEDVGVYYCQNGHGFPRTF GGGTKLEIKR (SEQ ID NO: 9) Ab47vL (KV3D- 11/KJ1) EIVLTQSPATLSLSPGERATLSCRASOTISDYLHWYOQKPGOAPRLLIK FASOSISGIP ARF SGSGSGTDFTLTIS SLEPEDF AVYYCQNGHGFPRTFG QGTKVEIKR (SEQ ID NO: 10) hvKl EIVMTQSPDFOSVTPKEKVTLTCRASOTISDYLHWYOQKPDOSPKLLI KFASQSISGVPSRFSGSGSGTDFTLTINSLEAEDAAVYYCQNGHGFPR TFGQGTKLEIK(R) (SEQ ID NO: 11) hvK2 EIVMTOSPDFOSVTPKEKVTLTCRASQTISDYLHWYOQKPDOSPKLLI KFASQSISGVPSRFSGSGSGTDFTLTINSLEAEDAATYYCQNGHGFPRT FGQGTKLEIK(R) (SEQ ID NO: 12) hvK3 EIVMTQSPDFOSVTPKEKVTLTCRASQTISDYLHWYQQKPDQSPKLLI KFASQSISGVPSRFSGSGSGSDFTLTINSLEAEDAATYYCQNGHGFPRT FGQGTKLEIK(R) (SEQ ID NO: 13)

- 24 045361

hvK4 DIOMTOSPSSLSASVGDRVTLTCRASQTISNYLAWYOOKPGKVPKLLI KFASTLQSGVPSRFSGSGSGSDFTLTISSLQPEDVATYYCQNGHGFPRT FGQGTKLEIK(R) (SEQ ID NO: 14) hvK3 (G91A) EIVMTOSPDFOSVTPKEKVTLTCRASQTISDYLHWYQQKPDOSPKLLI KFASQSISGVPSRFSGSGSGSDFTLTINSLEAEDAATYYCQNAHGFPRT FGQGTKLEIKR (SEQ ID NO: 15)

Table 3

Heavy chain CDR sequences of antibody variants (according to Kabat)

CDR Subsequence hvHl & hvH5 HCDR1 (according to Kabat) GYGMS (SEQ ID NO: 16) hvHl HCDR2 (according to Kabat) TITSGGTYTYYPDSVKG (SEQ ID NO: 17) hvHl - hvH6 HCDR3 (according to Kabat) SLAGNAMDY (SEQ ID NO: 18) hvH2 & hvH3 HCDR1 (according to Kabat) SYAMS (SEQ ID NO: 19) hvH2, hvH3, & hvH5 HCDR2 (no Kabatu) TITSGGTYTYYADSVKG (SEQ ID NO: 20) hvH4 HCDR1 (according to Kabat) SYGMN (SEQ ID NO: 21) hvH4 HCDR2 (according to Kabat) TITSGGTYIYYADSVKG (SEQ ID NO: 22) hvH6 HCDR1 (according to Kabat) SYGMH (SEQ ID NO: 23) hvH6 HCDR2 (according to Kabat) TITSGGTYTSYADSVKG (SEQ ID NO: 24)

Table 4

Heavy chain CDR sequences of antibody variants (IMGT)

CDR Subsequence hvHl & hvH5 HCDR1 (IMGT) GFTFSGYG (SEQ ID NO: 25) hvHl-hvH3, hvH5-hvH6 HCDR2 (IMGT) ITSGGTYT (SEQ ID NO: 26) hvHl - hvH6 HCDR3 (IMGT) ARSLAGNAMDY (SEQ ID NO: 27) hvH2 & hvH3 HCDR1 (IMGT) GFTFSSYA (SEQ ID NO: 28) hvH4 & hvH6 HCDR1 (IMGT) GFTFSSYG (SEQ ID NO: 29) hvH4 HCDR2 (IMGT) ITSGGTYI (SEQ ID NO: 30)

Table 5

Light chain CDR sequences of antibody variants (according to Kabat)

CDR Subsequence hvKl-hvK3 LCDR1 (according to Kabat) RASQTISDYLH (SEQ ID NO: 31) hvKl-hvK3 LCDR2 (according to Kabat) FASQSIS (SEQ ID NO: 32) hvKl- hvK4 LCDR3 (according to Kabat) QNGHGFPRT (SEQ ID NO: 33) hvK4 LCDR1 (according to Kabat) RASQTISNYLA (SEQ ID NO: 34) hvK4 LCDR2 (according to Kabat) FASTLQS (SEQ ID NO: 35) hvK3 (G91 A) LCDR3 (according to Kabat) QNAHGFPRT (SEQ ID NO: 36)

- 25 045361

Table 6

Light chain CDR sequences of antibody variants (IMGT)

CDR Subsequence hvKl-hvK3 LCDR1 (IMGT) QTISDY (SEQ Ш NO: 37) hvKl-hvK4 LCDR2 (IMGT) FAS (SEQ ID NO: 38) hvKl- hvK4 LCDR3 (IMGT) QNGHGFPRT (SEQ ID NO: 39) hvK4 LCDR1 (IMGT) QTISNY (SEQ ID NO: 40) hvK3 (G91A) LCDR3 (IMGT) QNAHGFPRT (SEQ ID NO: 41)

Table 7. Complete sequences of the heavy and light chains of the screened anti-CD47 antibody in accordance with a preferred embodiment of the present invention. The heavy chain and light chain sequences are shown in plain text (SEQ ID NOs: 42 and 43, respectively), the shielding sequences are in bold, and the protease cleavage sequences are underlined.

Table 7

Chain Antibodies Subsequence Heavy Chain QGASTSVDELQAEVDQLEDENYALKTKVAQLRKKVEKLGSIPV SLRSGEVQLLESGGGLVQPGGSLRLSCAASGFTFSGYGMSWVRQA PGKRLEWVATITSGGTYTYYPDSVKGRFTISRDNSKNTLYLQMNSL RAEDTAIYFCARSLAGNAMDYWGQGTLVTVSSASTKGPSVFPLAPS SKSTSGGTAALGCLVKDYFPEPV TVSWNSGALTSGVHTFPAVLQSS GL YSL S S VVT VP S S SLGTQT YICNVNHKP SNTK VDKK VEPK SCDKT HTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHED PEVI<FNWYVDGVEVHNAI<TI<PREEQYNSTYRVVSVLTVLHQDWL NGKEYKCKVSNKA LPAPIEKTISKAKGQPREPQVYTLPPSRDELTKN Q VSLTCL VKGF YP SDIAVEWESNGQPENNYKTTPP VLD SDGSFFL Y SKLTVDKSRWQQGNVFSCSVMHEALHNHYTQK (SEQ ID NO: 42) Heavy Circuit Shield Sequence QGASTSVDELQAEVDQLEDENYALKTKVAQLRKKVEKLGS (SEQ ID NO: 94) Light Chain QGASTTVAQLEEKVKTLRAENYELKSEVQRLEEQVAQLGSIPVS LRSGEIVMTQSPDFQSVTPKEKVTLTCRASQTISDYLHWYQQKPDQ SPKLLIKFASQSISGVPSRFSGSGSGSDFTLTINSLEAEDAATYYCQN GHGFPRTFGQGTKLEIKRTVAAPSVFIFPPSDEQLKSGTASVVCLLN NFYPREA KVQWKVDNALQSGNSQESVTEQDSKDSTYSLSSTLTLSK ADYEKHKVYACEVTHQGLSSPVTKSFNRGEC (SEQ ID NO: 43) Light Circuit Shielding Sequence QGASTTVAQLEEKVKTLRAENYELKSEVQRLEEQVAQLGS (SEQ ID NO: 95)

hB6H12.1

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 3 and/or the CDRs from LCVR represented by SEQ ID NO: 11. In other embodiments of the present invention, the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 16, 17 and 18 and/or light chain CDRs SEQ ID NOs: 31, 32 and 33. In some embodiments,

- 26 045361 tions of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 25, 26 and 27 and/or light chain CDRs SEQ ID NOs: 37, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody. CD47 or an antigen binding fragment thereof comprises the pair HCVR/LCVR SEQ ID NO: 3/SEQ ID NO: 11. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99%) to SEQ ID NO: 3 and/or contains LCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 11.

hB6H12.2

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 3 and/or the CDRs from LCVR represented by SEQ ID NO: 12. In other embodiments of the present invention, the anti-CD47 antibody or an antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 16, 17 and 18 and/or light chain CDRs SEQ ID NOs: 31, 32 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NOs: 25, 26 and 27 and/or light chain CDRs SEQ ID NOs: 37, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 3/SEQ GO NO: 12. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%, 94%, 95%, 96%, 97%, 98%, or 99%) relative to SEQ ID NO: 3 and/or contains an LCVR that has at least about 80% homology or identity (e.g. , 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 12.

hB6H12.3

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 3 and/or the CDRs from LCVR represented by SEQ ID NO: 13. In other embodiments of the present invention, the anti-CD47 antibody or an antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 16, 17 and 18 and/or light chain CDRs SEQ ID NOs: 31, 32 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NOs: 25, 26 and 27 and/or light chain CDRs SEQ ID NOs: 37, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 3/SEQ ID NO: 13. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%, 94%, 95%, 96%, 97%, 98%, or 99%) relative to SEQ ID NO: 3 and/or contains an LCVR that has at least about 80% homology or identity (e.g. , 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 13.

hB6H12.4

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 3 and/or the CDRs from LCVR represented by SEQ ID NO: 14. In other embodiments of the present invention the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 16, 17 and 18 and/or light chain CDRs SEQ ID NOs: 34, 35 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NOs: 25, 26 and 27 and/or light chain CDRs SEQ ID NOs: 40, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 3/SEQ ID NO: 14. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%, 94%, 95%, 96%, 97%, 98%, or 99%) relative to SEQ ID NO: 3 and/or contains an LCVR that has at least about 80% homology or identity (e.g. , 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 14.

hB6H12.5

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 4 and/or the CDRs from LCVR represented by SEQ ID NO: 11. In other embodiments of the present invention, the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 19, 20 and 18 and/or light chain CDRs SEQ ID NOs: 31, 32 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NO: 28, 26 and 27 and/or light chain CDR SEQ ID NO: 37, 38 and 39. Otherwise

- 27 045361 each embodiment of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the pair HCVR/LCVR SEQ ID NO: 4 /SEQ ID NO: 11. In other embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises an HCVR, which has, at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99%) relative to SEQ ID NO: 4 and/or contains an LCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95% , 96%, 97%, 98% or 99%) relative to SEQ ID nO: 11.

hB6H12.6

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 4 and/or the CDRs from LCVR represented by SEQ ID NO: 12. In other embodiments of the present invention the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 19, 20 and 18 and/or light chain CDRs SEQ ID NOs: 31, 32 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SeQ ID NOs: 28, 26 and 27 and/or light chain CDRs SEQ ID NOs: 37, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 4 /SEQ ID NO: 12. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID NO: 4 and/or contains an LCVR that has at least about 80% homology or identity (e.g. , 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 12.

hB6H12.7

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 4 and/or the CDRs from LCVR represented by SEQ ID NO: 13. In other embodiments of the present invention, the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 19, 20 and 18 and/or light chain CDRs SEQ ID NOs: 31, 32 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NOs: 28, 26 and 27 and/or light chain CDRs SEQ ID NOs: 37, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 4 /SEQ ID NO: 13. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID NO: 4 and/or contains an LCVR that has at least about 80% homology or identity (e.g. , 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 13.

hB6H12.8

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 4 and/or the CDRs from LCVR represented by SEQ ID NO: 14. In other embodiments of the present invention the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 19, 20 and 18 and/or light chain CDRs SEQ ID NOs: 34, 35 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NOs: 28, 26 and 27 and/or light chain CDRs SEQ ID NOs: 40, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 4 /SEQ ID NO: 14. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID NO: 4 and/or contains an LCVR that has at least about 80% homology or identity (e.g. , 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 14.

hB6H12.9

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 5 and/or the CDRs from LCVR represented by SEQ ID NO: 11. In other embodiments of the present invention, the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 19, 20 and 18 and/or light chain CDRs SEQ ID NOs: 31, 32 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NOs: 28, 26 and 27 and/or light chain CDRs SEQ ID NOs: 37, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 5 /SEQ ID NO: 11. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID NO: 5 and/or contains an LCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99%) with respect to SEQ ID nO: 11.

hB6H12.10

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 5 and/or the CDRs from LCVR represented by SEQ ID NO: 12. In other embodiments of the present invention, the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 19, 20 and 18 and/or light chain CDRs SEQ ID NOs: 31, 32 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NOs: 28, 26 and 27 and/or light chain CDRs SEQ ID NOs: 37, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 5 /SEQ ID NO: 12. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID NO: 5 and/or contains an LCVR that has at least about 80% homology or identity (e.g. , 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 12.

hB6H12.11

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 5 and/or the CDRs from LCVR represented by SEQ ID NO: 13. In other embodiments of the present invention, the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 19, 20 and 18 and/or light chain CDRs SEQ ID NOs: 31, 32 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NOs: 28, 26 and 27 and/or light chain CDRs SEQ ID NOs: 37, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 5 /SEQ ID NO: 13. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%, 94%, 95%, 96%, 97%, 98%, or 99%) relative to SEQ ID NO: 5 and/or contains an LCVR that has at least about 80% homology or identity (e.g. , 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 13.

hB6H12.12

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 5 and/or the CDRs from LCVR represented by SEQ ID NO: 14. In other embodiments of the present invention, the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 19, 20 and 18 and/or light chain CDRs SEQ ID NOs: 34, 35 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NOs: 28, 26 and 27 and/or light chain CDRs SEQ ID NOs: 40, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 5 /SEQ ID NO: 14. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%, 94%, 95%, 96%, 97%, 98%, or 99%) relative to SEQ ID NO: 5 and/or contains an LCVR that has at least about 80% homology or identity (e.g. , 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 14.

hB6H12.13

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 6 and/or the CDRs from LCVR represented by SEQ ID NO: 11. In other embodiments of the present invention, the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 21, 22 and 18 and/or light chain CDRs SEQ ID NOs: 31, 32 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NOs: 29, 30 and 27 and/or light chain CDRs SEQ ID NOs: 37, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 6/SEQ ID NO: 11. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%,

- 29 045361

94%, 95%, 96%, 97%, 98%, or 99%) relative to SEQ ID NO: 6 and/or contains an LCVR that has at least about 80% homology or identity (e.g., 80%, 85 %, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 11.

hB6H12.14

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 6 and/or the CDRs from LCVR represented by SEQ ID NO: 12. In other embodiments of the present invention the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 21, 22 and 18 and/or light chain CDRs SEQ ID NOs: 31, 32 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NOs: 29, 30 and 27 and/or light chain CDRs SEQ ID NOs: 37, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 6/SEQ ID NO: 12. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%, 94%, 95%, 96%, 97%, 98%, or 99%) relative to SEQ ID NO: 6 and/or contains an LCVR that has at least about 80% homology or identity (e.g. , 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 12.

hB6H12.15

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 6 and/or the CDRs from LCVR represented by SEQ ID NO: 13. In other embodiments of the present invention, the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 21, 22 and 18 and/or light chain CDRs SEQ ID NOs: 31, 32 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SeQ ID NOs: 29, 30 and 27 and/or light chain CDRs SEQ ID NOs: 37, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 6/SEQ ID NO: 13. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%, 94%, 95%, 96%, 97%, 98%, or 99%) relative to SEQ ID NO: 6 and/or contains an LCVR that has at least about 80% homology or identity (e.g. , 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 13.

hB6H12.16

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 6 and/or the CDRs from LCVR represented by SEQ ID NO: 14. In other embodiments of the present invention the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 21, 22 and 18 and/or light chain CDRs SEQ ID NOs: 34, 35 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NOs: 29, 30 and 27 and/or light chain CDRs SEQ ID NOs: 40, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 6/SEQ ID NO: 14. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%, 94%, 95%, 96%, 97%, 98%, or 99%) relative to SEQ ID NO: 6 and/or contains an LCVR that has at least about 80% homology or identity (e.g. , 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 14.

hB6H12.17

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 7 and/or the CDRs from LCVR represented by SEQ ID NO: 11. In other embodiments of the present invention the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 16, 20 and 18 and/or light chain CDRs SEQ ID NOs: 31, 32 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NOs: 25, 26 and 27 and/or light chain CDRs SEQ ID NOs: 37, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 7 /SEQ ID NO: 11. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%, 94%, 95%, 96%, 97%, 98%, or 99%) relative to SEQ ID NO: 7 and/or contains an LCVR that has at least about 80% homology or identity (e.g. , 80%, 85%, 90%, 91%, 92%, 93%,

- 30 045361

94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID NO: 11.

hB6H12.18

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 7 and/or the CDRs from LCVR represented by SEQ ID NO: 12. In other embodiments of the present invention, the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 16, 20 and 18 and/or light chain CDRs SEQ ID NOs: 31, 32 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NOs: 25, 26 and 27 and/or light chain CDRs SEQ ID NOs: 37, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 7 /SEQ ID NO: 12. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%, 94%, 95%, 96%, 97%, 98%, or 99%) relative to SEQ ID NO: 7 and/or contains an LCVR that has at least about 80% homology or identity (e.g. , 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 12.

hB6H12.19

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 7 and/or the CDRs from LCVR represented by SEQ ID NO: 13. In other embodiments of the present invention the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 16, 20 and 18 and/or light chain CDRs SEQ ID NOs: 31, 32 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NOs: 25, 26 and 27 and/or light chain CDRs SEQ ID NOs: 37, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 7 /SEQ ID NO: 13. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%, 94%, 95%, 96%, 97%, 98%, or 99%) relative to SEQ ID NO: 7 and/or contains an LCVR that has at least about 80% homology or identity (e.g. , 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 13.

hB6H12.20

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 7 and/or the CDRs from LCVR represented by SEQ ID NO: 14. In other embodiments of the present invention the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 16, 20 and 18 and/or light chain CDRs SEQ ID NOs: 34, 35 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NOs: 25, 26 and 27 and/or light chain CDRs SEQ ID NOs: 40, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 7 /SEQ ID NO: 14. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%, 94%, 95%, 96%, 97%, 98%, or 99%) relative to SEQ ID NO: 7 and/or contains an LCVR that has at least about 80% homology or identity (e.g. , 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 14.

hB6H12.21

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 8 and/or the CDRs from LCVR represented by SEQ ID NO: 11. In other embodiments of the present invention the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 23, 24 and 18 and/or light chain CDRs SEQ ID NOs: 31, 32 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NOs: 29, 26 and 27 and/or light chain CDRs SEQ ID NOs: 37, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 8/SEQ ID NO: 11. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%, 94%, 95%, 96%, 97%, 98%, or 99%) relative to SEQ ID NO: 8 and/or contains an LCVR that has at least about 80% homology or identity (e.g. , 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 11.

- 31 045361 hB6H12.22

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 8 and/or the CDRs from LCVR represented by SEQ ID NO: 12. In other embodiments of the present invention, the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 23, 24 and 18 and/or light chain CDRs SEQ ID NOs: 31, 32 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NOs: 29, 26 and 27 and/or light chain CDRs SEQ ID NOs: 37, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 8/SEQ ID NO: 12. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%, 94%, 95%, 96%, 97%, 98%, or 99%) relative to SEQ ID NO: 8 and/or contains an LCVR that has at least about 80% homology or identity (e.g. , 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 12.

hB6H12.23

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 8 and/or the CDRs from LCVR represented by SEQ ID NO: 13. In other embodiments of the present invention, the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 23, 24 and 18 and/or light chain CDRs SEQ ID NOs: 31, 32 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NOs: 29, 26 and 27 and/or light chain CDRs SEQ ID NOs: 37, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 8/SEQ ID NO: 13. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%, 94%, 95%, 96%, 97%, 98%, or 99%) relative to SEQ ID NO: 8 and/or contains an LCVR that has at least about 80% homology or identity (e.g. , 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 13.

hB6H12.24

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 8 and/or the CDRs from LCVR represented by SEQ ID NO: 14. In other embodiments of the present invention the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 23, 24 and 18 and/or light chain CDRs SEQ ID NOs: 34, 35 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SeQ ID NOs: 29, 26 and 27 and/or light chain CDRs SEQ ID NOs: 40, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 8/SEQ ID NO: 14. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%, 94%, 95%, 96%, 97%, 98%, or 99%) relative to SEQ ID NO: 8 and/or contains an LCVR that has at least about 80% homology or identity (e.g. , 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 14.

НВ6Н12.3 G91A

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 3 and/or the CDRs from LCVR represented by SEQ ID NO: 15. In other embodiments of the present invention the anti-CD47 antibody or the antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 16, 17 and 18 and/or light chain CDRs SEQ ID NOs: 34, 35 and 36. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NOs: 25, 26 and 27 and/or light chain CDRs SEQ ID NOs: 40, 38 and 41. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 3/SEQ ID NO: 15. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%, 94%, 95%, 96%, 97%, 98%, or 99%) relative to SEQ ID NO: 3 and/or contains an LCVR that has at least about 80% homology or identity (e.g. , 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 15.

Ab47

In some illustrative embodiments of the present invention, the anti-CD47 antibody

- 32 045361 or an antigen binding fragment thereof comprises a CDR from HCVR represented by SEQ ID NO: 2 and/or a CDR from LCVR represented by SEQ ID NO: 10. In other embodiments of the present invention, an anti-CD47 antibody or an antigen binding fragment thereof comprises a CDR severe chains SEQ ID NOs: 16, 17 and 18 and/or light chain CDRs SEQ ID NOs: 31, 32 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NOs: 25, 26 and 27 and/or light chain CDRs SEQ ID NO: 37, 38 and 39. In other embodiments, the anti-CD47 antibody or antigen binding fragment thereof comprises the pair HCVR/LCVR SEQ ID NO: 2/SEQ ID NO: 10. In others In embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92%, 93%, 94% , 95%, 96%, 97%, 98%, or 99%) relative to SEQ ID NO: 2 and/or contains an LCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID nO: 10.

mB6H12

In some illustrative embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof comprises the CDRs from HCVR represented by SEQ ID NO: 1 and/or the CDRs from LCVR represented by SEQ ID NO: 9. In other embodiments of the present invention the anti-CD47 antibody or an antigen binding fragment thereof comprises heavy chain CDRs SEQ ID NOs: 16, 17 and 18 and/or light chain CDRs SEQ ID NOs: 31, 32 and 33. In some embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises a heavy chain CDR SEQ ID NOs: 25, 26 and 27 and/or light chain CDRs SEQ ID NOs: 37, 38 and 39. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR/LCVR pair SEQ ID NO: 1/SEQ ID NO: 9. In other embodiments of the present invention, the anti-CD47 antibody or antigen binding fragment thereof comprises an HCVR that has at least about 80% homology or identity (e.g., 80%, 85%, 90%, 91%, 92 %, 93%, 94%, 95%, 96%, 97%, 98%, or 99%) relative to SEQ ID NO: 1 and/or contains an LCVR that has at least about 80% homology or identity (e.g. , 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99%) relative to SEQ ID NO: 9.

The anti-CD47 antibodies and antigen-binding fragments thereof described herein may be expressed in a modified form. For example, a region of additional amino acids, especially charged amino acids, can be added to the N-terminus of an anti-CD47 antibody or antigen binding fragment thereof to improve stability and persistence in a host cell during purification or subsequent handling and storage. Also, peptide fragments can be added to the anti-CD47 antibody or antigen binding fragment thereof of the present invention to facilitate purification. Such regions may be removed prior to final production of the antibody molecule or at least one fragment thereof. Such methods are described in many standard laboratory manuals, such as Sambrook, cited above; Ausubel, et al., ed., Current Protocols In Molecular Biology, John Wiley & Sons, Inc., NY, N.Y. (1987-2001).

The anti-CD47 antibodies or antigen-binding fragments thereof described herein typically bind CD47 with an equilibrium binding constant of =1 μM, e.g., =100 nM, preferably -10 nM, and more preferably -1 nM, as measured using standard binding assays, e.g. ,binding assay based on Biacore technology.

Antibody molecules of the present invention can be characterized relative to a reference anti-CD47 antibody, for example, B6H12, 2D3, MABL, CC2C6 or BRIC126. Antibody B6H12 is described, for example, in US Patent Nos. 5,057,604 and 9,017,675, commercially available from Abcam, PLC, Santa Cruz Biotechnology, Inc. and eBioscience, Inc.

Glycosylation variants.

Antibodies against CD47 and their antigen binding fragments can be glycosylated at conserved positions in their constant regions (Jefferis and Lund, (1997) Chem. Immunol. 65:111128; Wright and Morrison, (1997) TibTECH 15:26-32). The oligosaccharide side chains of immunoglobulins influence protein function (Boyd et al., (1996) Mol. Immunol. 32:1311-1318; Wittwe and Howard, (1990) Biochem. 29:4175-4180) and intramolecular interactions between the parts of the glycoprotein that can influence the conformation and three-dimensional surface presentation of the glycoprotein (Jefferis and Lund, cited above; Wyss and Wagner, (1996) Current Op. Biotech. 7:409-416). Oligosaccharides can also serve to target a given glycoprotein to specific molecules based on specific recognition structures. For example, it has been reported that in agalactosylated IgG, the oligosaccharide moiety "slips" out of the space between CH2 and the terminal N-acetylglucosamine residues become available for binding by mannose binding protein (Malhotra et al., (1995) Nature Med. 1:237-243) . Removal of oligosaccharides from CAMPATHI (a recombinant humanized mouse monoclonal antibody IgG1 that recognizes human lymphocyte CDw52 antigen) produced in Chinese hamster ovary (CHO) cells by glycopeptidase resulted in a complete reduction in

- 33 045361 complement-mediated lysis (CMCL) (Boyd et al., (1996) Mol. Immunol. 32:1311-1318), while selective removal of sialic acid residues using neuraminidase did not result in loss of CMCL. Antibody glycosylation has also been reported to influence antibody-dependent cellular cytotoxicity (ADCC). In particular, CHO cells with tetracycline-regulated expression of a(l,4)-N-acetylglucosaminyltransferase III, a glycosyltransferase that catalyzes the formation of bisecting GlcNAc, have been reported to have increased ADCC activity. (1999) Mature Biotech. 17:176-180).

Glycosylation of antibodies is usually either N-linked or O-linked. N-linking refers to the attachment of a carbohydrate moiety to the side chain of an asparagine residue. The tripeptide sequences asparagine-X-serine and asparagine-X-threonine, where X is any amino acid other than proline, are recognition sequences for enzymatically attaching a carbohydrate moiety to the asparagine side chain. Thus, the presence of any of these tripeptide sequences in a polypeptide creates a potential glycosylation site. O-linked glycosylation refers to the addition of one of the sugars N-acetylgalactosamine, galactose or xylose, to a hydroxyamino acid, most commonly serine or threonine, although 5-hydroxyproline or 5-hydroxylysine can also be used.

Antibody glycosylation variants are variants in which the glycosylation pattern of an antibody is changed. By alteration is meant the removal of one or more carbohydrate groups found in the antibody, the addition of one or more carbohydrate groups to the antibody, a change in the glycosylation composition (glycosylation pattern), the degree of glycosylation, etc.

The addition of glycosylation sites to an anti-CD47 antibody or antigen binding fragment thereof can be achieved by altering the amino acid sequence to contain one or more of the tripeptide sequences described above (for N-linked glycosylation sites). The change can also be made by adding or replacing one or more serine or threonine residues to the original antibody sequence (for O-linked glycosylation sites). Likewise, removal of glycosylation sites can be achieved by changing the amino acids at the native glycosylation sites of an antibody.

The amino acid sequence is usually changed by changing the corresponding nucleic acid sequence. These methods include the isolation from a natural source (in the case of naturally occurring amino acid sequence variants) or the production by oligonucleotide-mediated (or site-directed) mutagenesis, PCR mutagenesis, and cassette mutagenesis of a previously obtained antibody variant or non-variant antibody version.

Glycosylation (including glycosylation pattern) of antibodies can also be changed without changing the amino acid sequence or the corresponding nucleotide sequence. Glycosylation is highly dependent on the host cell used to express the antibody. Since the cell type used to express recombinant glycoproteins, such as antibodies, as potential therapeutics is rarely the native cell, significant changes in the glycosylation pattern of antibodies can be expected. See, for example, Hse et al., (1997) J. Biol. Chem. 272:9062–9070. In addition to the choice of host cells, factors that influence glycosylation during recombinant antibody production include growth mode, medium composition, culture density, oxygenation, pH, purification schemes, and the like. Various methods have been proposed to alter the glycosylation pattern achieved in a particular host, including the introduction or overexpression of certain enzymes involved in the production of oligosaccharides (US patent No. 5047335; 5510261; 5278299). Glycosylation or certain types of glycosylation can be enzymatically removed from a glycoprotein, for example, using endoglycosidase H (Endo H). In addition, the recombinant host cell may be genetically engineered, for example, damaged in processing certain types of polysaccharides. Such and similar methods are well known in the art.

The glycosylation structure of antibodies can be easily analyzed using conventional carbohydrate analysis techniques, including lectin chromatography, NMR, mass spectrometry, HPLC, GPC, monosaccharide composition analysis, sequential enzymatic digestion and HPAEC-PAD. , which uses high pH anion exchange chromatography to separate oligosaccharides based on charge. Methods for releasing oligosaccharides for analytical purposes are also known and include, but are not limited to, enzymatic treatment (typically performed using peptide-N-glycosidase F/endo-P-galactosidase), elimination using a harsh alkaline environment to release primarily O-linked structures, and chemical methods using anhydrous hydrazine to release both N- and O-linked oligosaccharides.

The preferred form of antibody glycosylation modification is reduced backbone fucosylation. Backbone fucosylation refers to the addition of fucose (fucosylation) to N-acetylglucosamine (GlcNAc) at the reducing end of the N-linked glycan.

The complex N-glycoside-linked sugar chain is usually linked to asparagine 297 (according to number

- 34 045361 radios on Kabat). As used herein, an N-glycoside-linked sugar chain has a biantenar sugar chain generally having the following structure:

+/-Fuca1 +/-Galp1 —► 4GlcNAcp1 —►ZManal v |

6 +/- GIcNAcpI ► 4Manp1 ► 4GlcNAcp1—► 4GlcNAc +/-Galp1—► 4GlcNAcp1—► 2Mana1 where +/- indicates that the sugar molecule may or may not be present, and the numbers indicate the position of the bonds between the sugar molecules. In the above structure, the end of the sugar chain that binds to asparagine is called the reducing end (right), and the opposite side is called the non-reducing end. Fucose is typically linked to the N-acetylglucosamine (GlcNAc) of the reducing end, typically via an α1,6 linkage (position 6 of GlcNAc linked to position 1 of fucose). Gal refers to galactose and Man refers to mannose.

A complex N-glycoside-linked sugar chain includes 1) a complex type in which the non-reducing end of the backbone structure has one or more galactose-Nacetylglucosamine (also called gal-GlcNAc) branches and the non-reducing end of GalGlcNAc optionally contains sialic acid, bisected GlcNAcN-acetylglucosamine or the like; or 2) a hybrid type in which the non-reducing end side of the backbone structure has both branches of a high-mannose sugar chain linked to an N-glycoside and a complex sugar chain linked to an N-glycoside.

In some embodiments of the present invention, the complex N-glycoside-linked sugar chain includes 1) a complex type in which the non-reducing end of the backbone structure has no or one or more galactose-N-acetylglucosamine (also called galGlcNAc) branches and a non-reducing end of Gal -GlcNAc optionally further contains sialic acid, bisected GlcNAcN-acetylglucosamine or the like.

In accordance with the present methods, typically only a trace amount of fucose is included in the complex N-glycoside-linked sugar chain(s) of humanized or chimeric antibodies. For example, in various embodiments of the present invention, less than about 60%, less than about 50%, less than about 40%, less than about 30%, less than about 20%, less than about 15%, less than about 10% , less than about 5% or less than about 3% of the antibody molecules have backbone fucosylation with fucose. In some embodiments of the present invention, about 2% of the antibody molecules have backbone fucosylation with fucose.

In some embodiments of the present invention, only a trace amount of the fucose analog (or metabolite or product of the fucose analog) is included in the complex N-glycoside-linked sugar chain(s). For example, in various embodiments of the present invention, less than about 60%, less than about 50%, less than about 40%, less than about 30%, less than about 20%, less than about 15%, less than about 10% , less than about 5% or less than about 3% of humanized or chimeric antibodies have a fucosylation of the backbone with a fucose analogue or a metabolite or product of a fucose analogue. In some embodiments of the present invention, about 2% of the humanized or chimeric antibodies have a fucosylation backbone with a fucose analog or a metabolite or product of a fucose analog.

Methods for producing non-fucosylated antibodies by incubating antibody-producing cells with a fucose analogue are described, for example, in WO 2009/135181. Briefly, cells that have been engineered to express humanized or chimeric antibodies are incubated in the presence of a fucose analog or an intracellular metabolite or product of a fucose analog. The intracellular metabolite may be, for example, a GDP-modified analogue or a fully or partially de-esterified analogue. The product may, for example, be a fully or partially de-esterified analogue. In some embodiments of the present invention, the fucose analogue may inhibit the enzyme(s) in the fucose utilization pathway. For example, a fucose analog (or an intracellular metabolite or product of a fucose analog) can inhibit the activity of fucokinase or GDP-fucosa pyrophosphorylase. In some embodiments of the present invention, the fucose analog (or intracellular metabolite or product of the fucose analog) inhibits fucosyltransferase (preferably 1,6-fucosyltransferase, such as FUT8 protein). In some embodiments of the present invention, a fucose analog (or an intracellular metabolite or product of a fucose analog) can inhibit the activity of an enzyme in the de novo synthetic pathway for fucose. For example, a fucose analog (or intracellular metabolite or product of a fucose analog) can inhibit the activity of GDP-mannose 4,6-dehydratase and/or GDP-fucose synthetase. In some embodiments of the present invention, the fucose analog (or intracellular metabolite or product of the fucose analog) can inhibit a fucose transporter (eg, the GDP-fucose transporter).

- 35 045361

In one embodiment of this invention, the fucose analogue is 2flurofucose. Methods for using fucose analogs in culture media and other fucose analogs are described, for example, in WO/2009/135181, which is incorporated herein by reference.

Other methods for constructing cell lines to reduce backbone fucosylation have included knock-out genes, knock-in genes, and RNA interference (RNAi). In gene knockout, the gene encoding FUT8 (alpha-1,6-fucosyltransferase enzyme) is inactivated. FUT8 catalyzes the transfer of a fucosyl moiety from GDP-fucose to position 6 of the Asn-linked (N-linked) GlcNac N-glycan. FUT8 is reported to be the sole enzyme responsible for the addition of fucose to the N-linked bianthenary carbohydrate at Asn297. In gene knockin, genes encoding enzymes such as GNTIII or Golgi alpha-mannosidase II are added. Increasing the level of such enzymes in cells diverts monoclonal antibodies from the fucosylation pathway (leading to a decrease in backbone fucosylation) and increases the amount of bisected N-acetylglucosamines. RNAi typically also targets FUT8 gene expression, resulting in decreased levels of mRNA transcripts or completely disrupting gene expression. Any of these methods can be used to create a cell line that is capable of producing a non-fucosylated antibody, such as a humanized or chimeric antibody.

There are many ways to determine the degree of fucosylation on an antibody. Methods include, for example, LC-MS using PLRP-S chromatography and electrospray ionization quadrupole MS (TOF MS).

Supercoiled screening agents.

In some embodiments of the present invention, the anti-CD47 antibody or antigen-binding fragment thereof is coupled to a coiled-coil shielding agent (also referred to as a coiled-coil shielding domain or shielding domain) that prevents the anti-CD47 antibody or antigen-binding fragment thereof from binding to CD47. In various embodiments of the present invention, an anti-CD47 antibody or an antigen-binding fragment thereof associated with a shielding domain is referred to as a shielded antibody.

A supercoil is a structural motif in proteins and peptides in which two or more alpha helices wrap around each other to form a supercoil. There can be two, three, or four helices in a supercoil bundle, and the helices can be oriented in either the same (parallel) or opposite (antiparallel) directions.

Supercoils typically contain three- and four-residue sequence elements whose hydrophobicity pattern and residue composition are compatible with the structure of amphipathic alphahelices. The alternating sequence elements of three and four residues are heptad repeats, in which the amino acids are designated a, b, c, d, e, f' and g. Residues at positions a and d are typically hydrophobic and form a zigzag pattern of protrusions and holes that interlock with a similar pattern on the other strand to form a tight hydrophobic core. Of the remaining residues, b, c and f' are generally charged. Therefore, the formation of a heptad repeat depends on the physical properties of hydrophobicity and charge, which are required at a specific position rather than at a specific amino acid. In some illustrative embodiments of the present invention, the supercoils of the present invention are formed from two peptides forming a supercoil.

Examples of consensus formulas for heptad repeats in coiled-coil peptides are given in WO 2011034605, which is incorporated herein by reference in its entirety for all purposes.

Exemplary consensus claims in accordance with some embodiments of the present invention are presented below:

Formula 1: (X1, X2, X3, X4, X5, X6, X7)n, where:

X1 is a hydrophobic amino acid or asparagine;

X2, X3 and X6 are any amino acid;

X4 is a hydrophobic amino acid;

X5 and X7 each represent a charged amino acid residue; and n is a positive integer.

Formula 2: (X1', X2', X3', X4', X5, X6, X7)p, where:

X1' is a hydrophobic amino acid or asparagine;

X2', X'3 and X'6 each represent any amino acid residue;

X4' is a hydrophobic amino acid;

X5' and X7' each represent a charged amino acid residue;

where n in formulas 1 and 2 is greater than or equal to 2; and n is a positive integer.

In some embodiments in which the peptides in Formula 1 and Formula 2 form a supercoil, X5 in Formula 1 is opposite in charge to X'7 in Formula 2, and X7 in Formula 1 is opposite in charge to X'5 in Formula 2. Heptad repeats within peptide that forms super

- 36 045361 spiral, may be the same or different from each other in accordance with Formula 1 and/or 2.

Supercoils can be homodimeric or heterodimeric. Examples of peptides that can form a supercoil in accordance with some exemplary embodiments of this invention are presented in table. 8. Peptide sequences can be used as is, or their components can be used in other combinations. For example, the coiled-coil Vel peptide can be used with other linker sequences. The sequences presented for light chains can also be used with heavy chains and vice versa.

Some illustrative embodiments of the present invention provide a divalent antibody comprising two pairs of light and heavy chains, wherein the N-termini of one or more light chains and/or heavy chains are linked through linkers containing a protease cleavage site in the coiled-coil peptides that bind , forming a supercoil, reducing the binding affinity of the light-heavy chain pair to the target. Optionally, the peptides bind without forming a disulfide bridge.

Not necessarily, the two pairs of light and heavy chains are the same. Optionally, the two pairs of light and heavy chains are different. Optionally, light chains include a light chain variable region and a light chain constant region, and heavy chains include a heavy chain variable region and a heavy chain constant region. Optionally, the heavy chain region includes the CH1, hinge, CH2 and CH3 regions. Optionally, two light chains are associated with the first heterologous peptide and two heavy chains are associated with the second heterologous peptide.

Optionally, the protease cleavage site is a MMP1 or MMP2 cleavage site.

Optionally, the target is CD47.

Optionally, antigen binding is reduced by at least 100-fold due to the presence of a screening agent (eg, a supercoiled screening agent). Optionally, antigen binding is reduced 200-1500-fold due to the presence of a screening agent (eg, a supercoiled screening agent). Optionally, the cytotoxicity of the conjugate is reduced by at least 100-fold due to the presence of a screening agent (eg, a supercoiled screening agent). Optionally, the cytotoxicity of the conjugate is reduced by at least 200-1500 times due to the presence of a shielding agent (eg, a supercoiled shielding agent).

Optionally, the coiled-coil peptides are linked to the N-terminus of the heavy and light chains in the same orientation. Optionally, the coiled-coil peptides are linked to the N-terminus of the heavy and light chains in opposite orientations. Optionally, multiple copies of the coiled-coil peptide are linked in tandem to the N-terminus of the heavy and light chains.

Exemplary coiled-coil peptide linkers and protease sites in accordance with some embodiments of the present invention are depicted in FIG. 34.

In accordance with some exemplary embodiments of the present invention, a peptide containing or consisting of SEQ ID NO: 51 (Vel LC in Table 8) is used to form a linker containing a protease cleavage site and a coiled-coil peptide linked to the N-terminus of the lung chain, and a peptide of the sequence SEQ ID NO: 50 (Vel HC in Table 8) is used to form a linker containing a protease cleavage site and a coiled-coil peptide linked to the N-terminus of the heavy chain, or vice versa. Peptides containing these sequences can be linked to any of the antibodies described herein.

In some illustrative embodiments of the present invention, the amino acid substitutions in the coiled-coil variant of the peptide are conservative substitutions. In other illustrative embodiments of the present invention, the repeating heptad pattern is retained in the peptide variant, whereby the coiled-coil peptide can be subdivided into contiguous heptad segments according to a formula classifying the amino acids occupying positions in the formula by amino acid type, as in Formula 1 and/or in Formula 2. In some embodiments of the present invention, there are no more than 1 or 2 substitutions per amino acid heptad, and any such substitutions are conservative. In other embodiments of this invention, the variant may be at least about 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91% , 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity to the coiled-coil forming peptide described herein and is capable of coiled-coil formation.

Exemplary peptides that form supercoils are listed in Table 1. 8A and table. 8B.

- 37 045361

Table 8A

Sequences of shielding domains in accordance with some exemplary embodiments of the present invention.

Cleavage sequences are underlined

Shielding Peptide Subsequence A2B1 NS GASTSVDELQAEVDQLQDENYALKTKVAQLRKKVEKLSEGGGGG GPLGVRGGGGS (SEQ ID NO: 44) A2B1 LC GASTTVAQLRERVKTLRAQNYELESEVQRLREQVAQLASGGGG GPLGVRGGGGS (SEQ ID NO: 45) MI NS LEIEAAFLERENTALETRVAELRQRVQRARNRVSQYRTRYGGGG GPLGVRGGGGS (SEQ ID NO: 46) MI LC LEIRAAFLRQRNTALRTEVAELEQEVQRLENEVSQYETRYGGGG GPLGVRGGGGS (SEQ ID NO: 47) Ml 5 NS LEIRAAFLRRRNTALRTRVAELRQRVQRLRNIVSQYETRYGGGG GGPLGVRGGGGS (SEQ ID NO: 48) M15LC LEIEAAFLEQENTALETEVAELEQEVQRLENIVSQYETRYGGGGG GPLGVRGGGGS (SEQ ID NO: 49) Vel NS GASTSVDELQAEVDQLEDENYALKTKVAQLRKKVEKLGSIPVSL RSG(SEQIDNO: 50) Vel LC GASTTVAQLEEKVKTLRAENYELKSEVQRLEEQVAQLGSIPVSL RSG(SEQIDNO: 51) Fos-Jun NS GALTDTLQAETDQLEDKKSALQTEIANLLKEKEKLEFILAAHGG GGGPLGVRGGGGS (SEQ ID NO: 52) Fos-Jun L.C. GARIARLEEKVKTLKAQNSELASTANMLREQVAQLKQKVMNY GGGGGPLGVRGGGGS (SEQ ID NO: 53) A4B4 NS GKIAALKQKIAALKYKNAALKKKIAALKQGGGGGPLGVRGGGG S (SEQ ID NO: 54) A4B4 LC GEIAALEQEIAALEKENALEWEIAALEQGGGGGPLGVRGGGGS (SEQ ID NO: 55)

- 38 045361

Table 8B

Sequences of shielding domains in accordance with some exemplary embodiments of the present invention. Cleavage sequences are underlined. Remaining EAS included

Shielding Peptide Subsequence CA2B1 NS EACGASTSVDELQAEVDQLQDENYALKTKVAQLRKKVEKLSEG GGGGPLGVRGGGGS (SEQ ID NO: 75) CA2B1LC EACGASTTVAQLRERVKTLRAQNYELESEVQRLREQVAQLASG GGGGPLGVRGGGGS (SEQ ID NO: 76) CM11 NS EACLEIEAAFLERENTALETRVAELRQRVQRARNRVSQYRTRYG GGGGPLGVRGGGGS (SEQ ID NO: 77) Media LC EACLEIRAAFLRQRNTALRTEVAELEQEVQRLENEVSQYETRYG GGGGPLGVRGGGGS (SEQ ID NO: 78) CM15HC EACLEIRAAFLRRRNTALRTRVAELRQRVQRLRNIVSQYETRYG GGGGGPLGVRGGGGS (SEQ ID NO: 79) CM15LC EACLEIEAAFLEQENTALETEVAELEQEVQRLENIVSQYETRYGG GGGGPLGVRGGGGS (SEQ ID NO: 80) CVel NS EACGASTSVDELQAEVDQLEDENYALKTKVAQLRKKVEKLGSIP VSLRSG (SEQ ID NO: 81) CVel LC EACGASTTVAQLEEKVKTLRAENYELKSEVQRLEEQVAQLGSIP VSLRSG (SEQ ID NO: 82) CFos-Jun HC EACGALTTDTLQAETDQLEDKKSALQTEIANLLKEKEKLEFILAAH GGGGGPLGVRGGGGS (SEQ ID NO: 83) CFos-Jun LC EACGARIARLEEKVKTLKAQNSELASTANMLREQVAQLKQKVM NYGGGGPLGVRGGGGS (SEQ ID NO: 84) CA4B4HC EACGKIAALKQKIAALKYKNAALKKKIAALKQGGGGGPLGVRG GGGS (SEQ ID NO: 85) CA4B4LC EACGEIAALEQEIAALEKENAALEWEIAALEQGGGGGPLGVRGG GGS (SEQ ID NO: 86)

Linkers and cleavage sites.

In some embodiments of the present invention, a linker is used to link the coiled-coil shielding agent to an anti-CD47 antibody or antigen-binding fragment thereof. Linkers can be any amino acid segments commonly used as a linker to attach peptide domains. Suitable linkers can vary in length, for example from 1 to 20, from 2 to 15, from 3 to 12, from 4 to 10, 5, 6, 7, 8, 9 or 10. Some such linkers include a polyglycine segment. Some such linkers include one or more serine residues, often in positions flanking glycine residues. Other linkers include one or more alanine residues. Glycine and glycine-serine polymers are relatively unstructured and therefore can serve as a neutral bond between components. Glycine gains access to significantly more phi-psi space than even alanine, and is much less restricted than residues with longer side chains (see Scheraga, Rev. Computational Chem. 11173-142 (1992)). Some exemplary linkers are of the form S(G)nS, where n is 5-20. Other exemplary linkers are (G)n, glycine-serine polymers (including, for example, (GS)n, (GSGGS)n [(GSGGS) SEQ ID NO: 59) and (GGGS)n, [(GGGS) - SEQ ID NO: 60), where n is an integer of at least one), glycine-alanine polymers, alanine-serine polymers and other flexible linkers known in the art. Some examples of linkers are Ser-(Gly)10-Ser (SEQ ID NO: 61), Gly-Gly-Ala-Ala (SEQ ID NO: 62), Gly-Gly-Gly-Gly-Ser (SEQ ID NO: 63 ), Leu-Ala-Ala- 39 045361

Ala-Ala (SEQ ID NO: 64), Gly-Gly-Ser-Gly (SEQ ID NO: 65), Gly-Gly-Ser-Gly-Gly (SEQ ID NO: 66), GlySer-Gly-Ser-Gly (SEQ ID NO: 67), Gly-Ser-Gly-Gly-Gly (SEQ ID NO: 68), Gly-Gly-Gly-Ser-Gly (SEQ ID NO: 69), Gly-Ser-Ser-Ser- Gly (SEQ ID NO: 70) and the like.

The protease site is preferentially recognized and cleaved by a protease expressed extracellularly, so it binds to the shielded antibody, releasing the shielded antibody and allowing it to bind to its target, such as an extracellular domain receptor or soluble ligand. Several matrix metalloproteinase sites (MMP1-28) are suitable. MMPs play a role in tissue remodeling and are involved in tumor processes such as morphogenesis, angiogenesis and metastasis. Some typical protease sites are PLG-XXX (SEQ ID NO: 71), a well-known endogenous sequence for MMPs, PLG-VR (SEQ ID NO: 72) (W02014193973) and IPVSLRSG (SEQ ID NO: 73) (Turk et al. , Nat. Biotechnol., 2001, 19, 661-667), LSGRSDNY (SEQ ID NO: 74) (Cytomyx) and GPLGVR (SEQ ID NO: 57) (Chang et al., Clin. Cancer Res. 2012 Jan 1; 18(1):238-47). Additional examples of MMPs are provided in US 2013/0309230, WO 2009/025846, WO 2010/081173, WO 2014/107599, WO 2015/048329, US 20160160263 and Ratnikov et al., Proc. Natl. Acad. Sci. USA, 111: E4148-E4155 (2014).

Table 9

Protease cleavage sequences. The MMP cleavage site is indicated by * and the uPA/matriptase/legumain cleavage sites are indicated by **

Cleavage Site Name Subsequence M2 GPLG*VR** (SEQ Ш NO: 57) IPV IPVS*LR**SG (SEQ ID NO: 58)

Binding of supercoiled shielding agents to anti-CD47 antibodies.

The coiled-coil peptides are linked to the N-terminus of the antibody variable regions through a linker including a protease cleavage site. A typical antibody includes a heavy and light chain variable region, in which case a coiled-coil peptide is linked to the N-terminus of each chain. A divalent antibody has two binding sites, which may or may not be the same. In a normal monospecific antibody, the binding sites are identical and the antibody has two identical light and heavy chain pairs. In this case, each heavy chain is associated with the same coiled-coil peptide, and each light chain is associated with the same coiled-coil peptide (which may or may not be the same as the peptide associated with the heavy chain). In a bispecific antibody, the binding sites are different and are formed from two different pairs of heavy and light chains. In such a case, the heavy and light chain variable regions of one binding site are respectively associated with the coiled-coil peptides, as are the heavy and light chain variable regions of the other binding site. Typically, both heavy chain variable regions are associated with the same type of coiled-coil peptide as both light chain variable regions.

The coiled-coil peptide may be linked to the variable region of the antibody through a linker including a protease cleavage site. Typically, the same linker with the same protease cleavage site is used to link each antibody heavy or light chain variable region to the coiled-coil peptide. The protease cleavage site must be one that can be cleaved by a protease present extracellularly in the intended target tissue or pathology, such as a tumor, such that cleavage of the linker releases the antibody from the coiled coil shielding its activity, allowing the antibody to bind to its intended target, such as an antigen cell surface or soluble ligand.

Along with the variable regions, the shielded antibody typically includes all or part of a constant region, which may include any or all of the light chain, CH1, hinge, CH2, and CH3 constant regions. As with other antibodies, one or more C-terminal residues may be proteolytically processed or derivatized.

Supercoils can be formed from the same peptide, forming a homodimer, or two different peptides, forming a heterodimer. To form a homodimer, the light and heavy chains of antibodies bind to the same peptide, forming a supercoil. To form a heterodimer, the light and heavy chains of antibodies bind to different peptides that form a superhelix. For some peptide pairs forming a supercoil, it is preferable for one of the pair to be associated with the heavy chain and the other with the light chain of the antibody, although the reverse orientation is also possible.

Each antibody chain may be associated with a single coiled-coil peptide or multiple such peptides in tandem (eg, two, three, four, or five copies of the peptide). If the latter, then the peptides in tandem bonding are usually the same. Also, if tandem linkage is used, the light and heavy chains are usually linked to the same number of peptides.

Binding of antibody chains to coiled-coil peptides can reduce the binding affinity of the antibody by at least about 10-fold, about 50-fold, about 100-fold, about 200-fold,

-40045361 about 500 times, about 1000 times or about 1500 times relative to the same antibody without such a bond or after cleavage of such a bond. In some such antibodies, the binding affinity is reduced in the range of about 50-1500-fold, about 100-1500-fold, about 200-1500-fold, about 500-1500-fold, about 50-1000-fold, about 100-1000-fold , about 200-1000 times, about 500-1000 times, about 50-500 times or about 100-500 times. Antibody effector functions such as ADCC, phagocytosis and CDC, or cytotoxicity resulting from drug binding in an antibody-drug conjugate can be reduced by the same factors or within the same ranges. After proteolytic cleavage, which serves to strip the shielding moiety from the antibody or otherwise remove the shielding moiety from the antibody, the recovered antibody typically has an affinity or effector function that is within a factor of 2, 1.5, or preferably constant within experimental error. compared to another identical control antibody that was never screened.

Antibody-drug conjugates.

In some embodiments of the present invention, the anti-CD47 antibodies of the present invention can be combined with antibody-drug conjugates (ADCs). Specific ADCs may contain cytotoxic agents (eg, chemotherapeutic agents), prodrug-converting enzymes, radioactive isotopes or compounds, or toxins (these groups are collectively referred to as therapeutic agents). For example, the ADC may be conjugated to a cytotoxic agent, such as a chemotherapeutic agent, or a toxin (eg, a cytostatic or cytocidal agent such as, for example, abrin, ricin A, pseudomonas exotoxin, or diphtheria toxin). Examples of useful classes of cytotoxic agents include, for example, DNA minor groove binders, DNA alkylating agents, and tubulin inhibitors. Illustrative cytotoxic agents include, for example, auristatin, camptotecins, kalichamicins, duokocracy, ethosides, mayanzinoids (for example, DM1, DM2, DM3, DM4), dactia, benzodiazepines (for example, pyrroolo [1,4] benzodiazepines, indolidiazodiazepines and oxasolyidinonosepines Azepins including dimmers pyrrolo [1,4] benzodiazepine, indolinebenzodiazepine dimers and oxazolidinone benzodiazepine dimers) and vinca alkaloids.

The ADC can be conjugated to a prodrug converting enzyme. The prodrug-converting enzyme can be recombinantly fused to or chemically conjugated to the antibody using known methods. Typical prodrug-converting enzymes are carboxypeptidase G2, beta-glucuronidase, penicillin V-amidase, penicillin O-amidase, β-lactamase, β-glucosidase, nitroreductase, and carboxypeptidase A.

Methods for conjugating therapeutic agents to proteins and, in particular, antibodies are well known. (See, for example, Alley et al., Current Opinion in Chemical Biology 2010 14: 1-9; Senter, Cancer J., 2008, 14(3): 154-169.) The therapeutic agent can be conjugated in a manner that reduces its activity unless it is cleaved from the antibody (eg by hydrolysis, proteolytic degradation or a cleavage agent). In some aspects, the therapeutic agent is attached to the antibody via a cleavable linker that is sensitive to cleavage in the intracellular environment of the CD47-expressing cancer cell, but is substantially insensitive to the extracellular environment, such that the conjugate is cleaved from the antibody when it is taken up by the cancer cell. expressing CD47, (eg, in the endosomal or, for example, depending on pH sensitivity or protease, in the lysosomal environment or in the caveolar environment). In some embodiments of the present invention, the therapeutic agent may also be attached to the antibody via a non-cleavable linker.

In some illustrative embodiments of the present invention, the ADC may include a linker region between the cytotoxic or cytostatic agent and the antibody. As noted above, generally, the linker can be cleaved under intracellular conditions such that cleavage of the linker releases the therapeutic agent from the antibody in the intracellular environment (eg, a lysosome, endosome, or caveola). The linker may be, for example, a peptidyl linker that is cleaved by an intracellular peptidase or protease enzyme, including a lysosomal or endosomal protease. Degrading agents may include cathepsins B and D and plasmin (see, for example, Dubowchik and Walker, Pharm. Therapeutics 83:67-123, 1999). The most typical are peptidyl linkers, which are cleaved by enzymes present in cells expressing CD47. For example, a peptidyl linker may be used that is cleaved by the thiol-dependent protease cathepsin-B, which is highly expressed in cancer tissue (eg, a linker containing a Phe-Leu or Val-Cit peptide).

The cleavable linker may be pH sensitive, that is, sensitive to hydrolysis at certain pH values. Typically, the pH-sensitive linker is hydrolyzed under acidic conditions. For example, an acid-labile linker that is hydrolyzed in the lysosome (eg, hydrazone, semicarbazone, thiosemicarbazone, cis-aconitic acid amide, orthoester, acetal, ketal, or the like) may be used. (See, for example, US Patents No. 5,122,368; 5,824,805; 5,622,929; Dubowchik

- 41 045361 and Walker, Pharm. Therapeutics 83:67-123, 1999; Neville et al, Biol. Chem. 264: 14653-14661, 1989.) Such linkers are relatively stable under neutral pH conditions, such as in blood, but are unstable below pH 5.5 or 5.0, the approximate pH of the lysosome.

Other linkers are cleaved under reducing conditions (eg, disulfide linker). Disulfide linkers include those that can be formed using SATA (Nsuccinimidyl-S-acetylthioacetate), SPDP (N-succinimidyl-3-(2-pyridyldithio)propionate), SPDB (Nsuccinimidyl-3-(2-pyridyldithio)butyrate), and SMPT (N-succinimidyl-oxycarbonyl-alpha-methylalpha-(2-pyridyl-dithio)toluene), SPDB and SMPT. (See, for example, Thorpe et al., Cancer Res. 47:59245931, 1987; Wawrzynczak et al., In Immunoconjugates: Antibody Conjugates in Radioimagery and Therapy of Cancer (C. W. Vogel ed., Oxford U. Press, 1987. See (see also US Patent No. 4,880,935.)

The linker may also be a malonate linker (Johnson et al., Anticancer Res. 15: 138793, 1995), a maleimidobenzoyl linker (Lau et al., Bioorg-Med-Chem. 3: 1299-1304, 1995) or a 3'-Namide analogue (Lau et al., Bioorg-Med-Chem. 3: 1305-12, 1995).

The linker may also be a non-cleavable linker, such as a maleimido-alkylene or maleimide-aryl linker, that is directly attached to the therapeutic agent and released upon proteolytic degradation of the antibody.

Typically, the linker is not significantly sensitive to the extracellular environment, meaning that no more than about 20%, typically no more than about 15%, more typically no more than about 10%, and even more typically no more than about 5%, no more than about 3% or no more than about 1% of the linkers in an ADC sample are cleaved when the ADC is present in the extracellular environment (eg, plasma). Whether a linker is substantially insensitive to the extracellular environment can be determined, for example, by independently incubating with plasma both (a) the ADC (ADC sample) and (b) an equal molar amount of unconjugated antibody or therapeutic agent (control sample) for a predetermined period of time (eg, 2, 4, 8, 16 or 24 hours) and then compare the amount of unconjugated antibody or therapeutic agent present in the ADC sample with the amount present in the control sample, measured, for example, using high performance liquid chromatography .

The linker can also stimulate cellular internalization. The linker can promote cellular internalization when it is conjugated to a therapeutic agent (ie, in the environment of a linker-therapeutic agent component of an ADC or an ADC derivative as described herein). Alternatively, the linker can promote cellular internalization when it is conjugated to both a therapeutic agent and an antibody (ie, in an ADC medium as described herein).

The antibody may be conjugated to a linker via an antibody heteroatom. These heteroatoms may be present on the antibody in its natural state or may be introduced into the antibody. In some aspects, the antibody will be conjugated to a linker via the nitrogen atom of a lysine residue. In other aspects, the antibody will be conjugated to the linker via the sulfur atom of the cysteine residue. Methods for conjugating a linker and a drug linker to antibodies are known in the art.

Exemplary antibody-drug conjugates include auristatin-based antibody-drug conjugates, which means that the drug component is the drug auristatin. Auristatins have been shown to bind tubulin, affect microtubule dynamics, nuclear and cell division, and have antitumor activity. Typically, an auristatin antibody-drug conjugate contains a linker between the auristatin drug and an anti-CD47 antibody. The linker may be, for example, a cleavable linker (eg, a peptidyl linker) or a non-cleavable linker (eg, a linker released upon antibody degradation). Auristatins include MMAF and MMAE. The synthesis and structure of exemplary auristatins are described in US Publication Nos. 7,659,241, 7,498,298, 2009-0111756, 2009-0018086, and 7,968,687, each of which is incorporated herein in its entirety and for all purposes by reference.

Other exemplary antibody-drug conjugates include maytansinoid antibody-drug conjugates, meaning that the drug component is a maytansinoid drug, and benzodiazepine-based antibody-drug conjugates, meaning that the drug component is a benzodiazepine (e.g. pyrrolo[1,4]benzodiazepine dimers, indolinebenzodiazepine dimers and oxazolidinone benzodiazepine dimers).

In some embodiments of the present invention, an anti-CD47 antibody of the present invention can be combined with an ADC with binding specificity for another target. Exemplary ADCs that can be combined with an anti-CD47 antibody include brentuximabvedotin (anti-CD30-ADC), enfortumab-vedotin (anti-Nectin-4-ADC), ladiratuzumab-vedotin (anti-LIV-1 ADC), denintuzumab-mafodotin (anti- CD19 ADC), glembatumumab-vedotin (anti-GPNMB ADC), anti-TIM-1 ADC, polatuzumab-vedotin (anti-CD79b ADC), anti-MUC16 ADC, depatuxizumab mafodotin, telizotuzumab-vedotin, anti-PSMA ADC, anti- C4.4a ADC, anti-BCMA ADC, anti-AXL ADC, tizotumab vedotin (anti-tissue factor ADC).

- 42 045361

Expression of an antibody molecule.

The nucleic acids of the present invention can be expressed in a host cell that contains endogenous DNA encoding an antibody or screened antibody of the present invention. Such methods are well known in the art, for example, as described in US Patent Nos. 5,580,734, 5,641,670, 5,733,746, and 5,733,761. Also see, for example, Sambrook, et al., cited above, and Ausubel, et al., cited above. Those skilled in the art are aware of numerous expression systems available for expressing the nucleic acid encoding the protein of the present invention. Illustrative cell cultures useful for the production of antibodies, shielded antibodies, these portions, or variants thereof are mammalian cells. Mammalian cell systems can often be in the form of cell monolayers, although mammalian cell suspensions or bioreactors can also be used. A number of suitable host cell lines capable of expressing intact glycosylated proteins have been developed in the art and include COS-1 (eg ATCC CRL 1650), C0S-7 (eg ATCC CRL-1651), HEK293, BHK21 (eg cell lines ATCC CRL-10), CHO (eg ATCC CRL 1610) and BSC1 (eg ATCC CRL-26), hep G2 cells, P3X63Ag8.653 cells, SP2/0-Ag14, HeLa cells and the like, which are readily available from , for example, the American Type Culture Collection (Manassas, Virginia). Yeast and bacterial host cells, which are well known to those skilled in the art, may also be used. Other cells useful for producing the nucleic acids or proteins of this invention are known and/or available, for example from the American Type Culture Collection Catalog of Cell Lines and Hybridomas or from other known or commercial sources.

Expression vectors may include one or more of the following expression control sequences, such as, but not limited to, an origin of replication; promoter (e.g., SV40 late or early promoters, CMV promoter (US Pat. No. 5,168,062; 5,385,839), HSV-tk promoter, pgk (phosphoglycerate kinase) promoter, EF-1 alpha promoter (US Pat. No. 5,266,491), at least , one human immunoglobulin promoter; enhancer and/or processing sites such as ribosome binding sites, RNA splicing sites, polyadenylation sites (e.g., poly-A addition site to the SV40 large T antigen sequence) and transcription terminator sequences). Also see, for example, Ausubel et al., cited above; Sambrook, et al., cited above.

Expression vectors optionally include at least one selectable marker. Such markers include, for example, but are not limited to, methotrexate (MTX), dihydrofolate reductase (DHFR, US Pat. No. 4,399,216; 4,634,665; 4,656,134; 4,956,288; 5,149,636; 5,179,017), ampicillin, neomycin (G418), mi caffenolic acid or glutamine synthetase (GS , US Pat. No. 5,122,464; 5,770,359 and 5,827,739), eukaryotic cell culture resistance and tetracycline or ampicillin resistance genes for cultivation in E. coli and other bacteria or prokaryotes. Suitable culture media and conditions for the above-described host cells are known in the art. Suitable vectors will be apparent to one skilled in the art. Introduction of the vector construct into a host cell can be accomplished by calcium phosphate transfection, DEAE-dextran-mediated transfection, cationic lipid-mediated transfection, electroporation, transduction, infection, or other known methods. Such methods are described in the art, for example Sambrook, cited above; Ausubel, quoted above.

The nucleic acid insert must be operably linked to an appropriate promoter. Expression constructs will additionally contain transcription initiation and termination sites and, in the transcribed region, a ribosome binding site for translation. The coding portion of mature transcripts expressed by these constructs will preferably include a translation initiation at the beginning and a stop codon (e.g., UAA, UGA, or UAG) appropriately located at the end of the mRNA for translation, with UAA and UAG being preferred for expression in mammalian cells or eukaryotes.

The nucleic acid insert is not necessarily in frame with a coiled-coil sequence and/or an MMP cleavage sequence, for example, at the N-terminus of one or more heavy chain and/or light chain sequences. Alternatively, a coiled-coil sequence and/or an MMP cleavage sequence may be post-translationally added to the antibody or antigen-binding fragment thereof, for example, through a disulfide bond or the like.

When eukaryotic host cells are used, polyadenylation or transcription terminator sequences are typically included in the vector. An example of a terminator sequence is a polyadenylation sequence from the bovine growth hormone gene. Sequences for precise splicing of the transcript may also be included. An example of a splice sequence is the VP1 intron of SV40 (Sprague, et al. (1983) J. Virol. 45:773-781). In addition, gene sequences for controlling replication in a host cell can be included in the vector, as is known in the art.

Isolation and purification of antibodies.

Anti-CD47 antibodies or screened antibodies described herein may be

- 43 045361 isolated and purified from recombinant cell cultures by well known methods including, but not limited to, protein A purification, ammonium sulfate or ethanol precipitation, acid extraction, anion or cation exchange chromatography, phosphocellulose chromatography, hydrophobic interaction chromatography, affinity chromatography, hydroxyapatite chromatography and lectin chromatography. High performance liquid chromatography (HPLC) can also be used for purification. See, for example, Colligan, Current Protocols in Immunology, or Current Protocols in Protein Science, John Wiley & Sons, New York, N.Y., (1997-2001).

Antibodies and screened antibodies described herein may include purified products, products of chemical synthetic processes and products obtained by recombinant methods from a eukaryotic host, including, for example, cells of yeast, higher plants, insects and mammals. Depending on the host used in the recombinant production procedure, the antibody or screened antibody of this invention may be glycosylated or non-glycosylated, with glycosylation being preferred. Such methods are described in many standard laboratory manuals, such as Sambrook, cited above; Ausubel, cited above, Colligan, Protein Science, cited above.

Nucleic acid molecules.

The nucleic acid molecules of the present invention may be in the form of RNA, such as mRNA, hnRNA, tRNA or any other form, or in the form of DNA, including, but not limited to, cDNA and genomic DNA, whether cloned or synthetically produced, or any combinations of them. DNA can be triple-stranded, double-stranded, single-stranded, or any combination thereof. Any portion of at least one strand of DNA or RNA may be a coding strand, also known as a sense strand, or it may be a non-coding strand, also known as an antisense strand.

Isolated nucleic acid molecules of the present invention may include nucleic acid molecules containing an open reading frame (ORF), optionally with one or more nitrones, for example, but not limited to, at least one specified portion, at least one CDR, such as CDR1, CDR2 and/or CDR3 of at least one heavy chain or light chain; nucleic acid molecules containing the coding sequence of an anti-CD47 antibody or a shielded antibody, or a variable region of an anti-CD47 antibody or a shielded antibody; and nucleic acid molecules containing a nucleotide sequence substantially different from that described above, but which, due to the degeneracy of the genetic code, still encode at least one anti-CD47 antibody or shielded antibody as described herein and/or as known in this field of technology. Given that the genetic code is well known in the art, it is common for one skilled in the art to generate such degenerate nucleic acid variants that encode specific anti-CD47 antibodies or screened antibodies of the present invention.

As indicated herein, the nucleic acid molecules of the present invention, which contain a nucleic acid encoding an anti-CD47 antibody molecule, may include, but are not limited to, sequences directly encoding the amino acid sequence of the antibody fragment; the coding sequence of all or part of the antibody; the coding sequence of the antibody, fragment or portion, as well as additional sequences, such as sequences of one or both of the shielding agents (eg, coiled-coil shielding agent) and/or MMP cleavage sequences, or such as the coding sequence of at least one leader signal peptide or fusion peptide with or without the above-mentioned additional coding sequences, such as at least one intron, together with additional non-coding sequences, including, but not limited to, 5 and 3' non-coding sequences, such as transcribed untranslated sequences, which play a role in transcription, mRNA processing, including splicing and polyadenylation signals (eg, ribosome binding and mRNA stability); an additional coding sequence that encodes additional amino acids, such as those that provide additional functionality. In some embodiments of the present invention, the antibody coding sequence may be fused to a marker sequence, such as a peptide coding sequence, that facilitates purification of the fusion antibody containing a fragment or portion of the antibody.

Construction of nucleic acids.

The isolated nucleic acids of this invention can be prepared using (a) recombinant methods, (b) synthetic methods, (c) purification methods, or combinations thereof, well known in the art. The nucleic acids may easily contain sequences in addition to the polynucleotide of the present invention. For example, a multiple cloning site containing one or more restriction endonuclease sites may be inserted into a nucleic acid to aid in the isolation of a polynucleotide. In addition, translation sequences may be inserted to assist in isolating the translatable polynucleotide of the present invention.

- 44 045361

For example, the hexa-histidine tag sequence provides a convenient method for purifying the proteins of the present invention. The nucleic acid of the present invention, excluding the coding sequence, is optionally a vector, adapter, or linker for cloning and/or expression of a polynucleotide of the present invention. Additional sequences may be added to such cloning and/or expression sequences to optimize their function during cloning and/or expression, to assist in the isolation of the polynucleotide, or to improve the introduction of the polynucleotide into the cell. The use of cloning vectors, expression vectors, adapters and linkers is well known in the art. (See, for example, Ausubel, cited above; or Sambrook, cited above.)

The isolated nucleic acid compositions of this invention, such as RNA, cDNA, genomic DNA, or any combination thereof, can be obtained from biological sources using any number of cloning technologies known to those skilled in the art. In some embodiments of the present invention, oligonucleotide probes that selectively hybridize under stringent conditions with polynucleotides of the present invention are used to identify a desired sequence in a cDNA library or genomic DNA. Isolation of RNA and construction of cDNA and genomic libraries are well known to those skilled in the art. (See, for example, Ausubel, cited above; or Sambrook, cited above.)

III. Therapeutic Applications

The invention provides methods for treating disorders associated with cells that express CD47, such as cancer. Cells may or may not express elevated levels of CD47 relative to cells that are not associated with the disorder of interest. As a result, the invention provides a method of treating a subject, for example a subject with cancer, using anti-CD47 antibodies or shielded antibodies described herein. The method includes administering an effective amount of an anti-CD47 antibody or a shielded antibody, or a composition comprising an anti-CD47 antibody or a shielded antibody, to a subject in need thereof.

As used herein, the terms subject and patient refer to organisms to be treated by the methods of this invention. Such organisms preferably include, but are not limited to, mammals (eg, mice, monkeys, horses, cattle, pigs, dogs, cats, and the like) and more preferably include humans. As used herein, the terms treatment, therapy, and effect include any effect, such as reduction, restoration, modulation, normalization, or elimination, that results in the improvement of a condition, disease, disorder, or the like, or the alleviation of a symptom thereof, such as, for example, reducing the number of cancer cells, reducing tumor size, reducing the rate of cancer cell infiltration into peripheral organs, or reducing the rate of tumor metastasis or tumor growth.

Beneficial therapeutic effects in cancer can be measured in several ways (see WA Weber, J. Null. Med. 50:1S-10S (2009); Eisenhauer et al., cited above). In some preferred embodiments of the present invention, the response to the anti-CD47 antibody or screened antibody is assessed using the RECIST 1.1 criterion. In some embodiments of the present invention, the treatment achieved by a therapeutically effective amount is any of partial response (PR), complete response (CR), progression-free survival (PFS), disease-free survival (DFS), objective response (OR, OR) or overall survival (OB, OS).

The dosage regimen described herein for therapy that is effective for treating primary or secondary liver cancer may vary depending on factors such as disease status, age and weight of the patient, and the ability of the therapy to elicit an anticancer response in the subject. Although an embodiment of the treatment, medicaments and uses of this invention may not be effective in achieving a beneficial therapeutic effect in every subject, it should do so in a statistically significant number of subjects, as determined by any statistical test known in the art, such as t -Student's t test, chi-square test, Mann-Whitney U test, Kruskal-Wallis test (N test), Jonckheere-Terpstra test and Wilcoxon test.

As used in this document, RECIST 1.1 Tumor Response Evaluation Criteria means the definitions set forth in Eisenhauer et al., E. A. et al., Eur. J Cancer 45:228–247 (2009) for target or non-target lesions, as appropriate, depending on the context in which the response is assessed.

Tumor, as applied to a subject diagnosed or suspected of having primary or secondary liver cancer, refers to a malignant or potentially malignant neoplasm or mass of tissue of any size. A solid tumor is an abnormal growth or mass of tissue that usually does not contain cysts or fluid areas. The different types of solid tumors are named for the type of cells that form them. Examples of solid tumors are sarcomas, carcinomas and

- 45 045361 lymphoma. Leukemias (blood cancers) do not usually form solid tumors (National Cancer Institute, Dictionary of Oncology Terms). Non-limiting illustrative sarcomas include soft tissue sarcoma and osteosarcoma.

Tumor burden, also called tumor load, refers to the total amount of tumor material distributed throughout the body. Tumor burden refers to the total number of cancer cells or the total size of the tumor throughout the body, including the lymph nodes and bone marrow. Tumor burden can be determined by various methods known in the art, such as, for example, measuring the size of the tumor(s) when removed from a subject, for example, using a caliper, or while in the body using imaging techniques, such as ultrasound, bone scintigraphy, computed tomography (CT) or magnetic resonance imaging (MRI).

The term tumor size refers to the overall size of the tumor, which may include measurements of both the length and width of the tumor. Tumor size can be determined by various methods known in the art, such as, for example, measuring the size of the tumor(s) when removed from a subject, for example, using a caliper, or while in the body using imaging techniques, for example, bone scintigraphy , Ultrasound, CT or MRI.

As used herein, the term effective amount refers to an amount of a compound (eg, anti-CD47 antibody or screened antibody) sufficient to achieve beneficial or desired results. An effective amount may be administered in one or more administrations, applications, or dosages and is not intended to be limited to a particular formulation or route of administration. Typically, a therapeutically effective amount of the active ingredient is in the range of 0.1 mg/kg to 100 mg/kg, for example, 1 mg/kg to 100 mg/kg, 1 mg/kg to 10 mg/kg. The dose administered may vary depending on known factors such as the pharmacodynamic characteristics of the particular agent and its mode and route of administration; age, health and weight of the recipient; type and extent of disease or indication for treatment, nature and extent of symptoms, type of concomitant treatment, frequency of therapy and desired effect. The initial dose may be increased beyond the upper level to quickly achieve the desired blood level or tissue level. Alternatively, the initial dose may be less than optimal and the daily dose may be gradually increased over the course of therapy. Human dosing may be optimized, for example, in a conventional Phase I dose escalation study targeting doses ranging from 0.5 to 20 mg/kg. Dosing frequency may vary depending on factors such as route of administration, dosage, serum half-life of the antibody, and the disease being treated. Exemplary dosing frequencies are once daily, once weekly, and once every two weeks. The composition of monoclonal antibody drugs is known to those skilled in the art. In some embodiments of the present invention, the monoclonal antibody is lyophilized and then reconstituted in a buffered saline solution at the time of administration.

In some illustrative embodiments of the present invention, the present invention relates to a method of treating cancer in a cell, tissue, organ, animal, or patient. In specific embodiments of the present invention, the present invention relates to a method for treating solid cancer in a human. Typical cancers are those that are characterized by the expression of CD47 in a cancer cell (ie, a cancer characterized by CD47 expression). Examples of cancer include, but are not limited to, solid tumors, soft tissue tumors, hematopoietic tumors that cause solid tumors, and metastatic lesions. Examples of hematopoietic tumors that can lead to the formation of solid tumors include, but are not limited to, diffuse large B-cell lymphoma (DLBCL), follicular lymphoma, myelodysplastic syndrome (MDS), lymphoma, Hodgkin's disease, malignant lymphoma, non-Hodgkin's lymphoma, Burkitt's lymphoma, multiple myeloma, Richter syndrome (Richter transformation) and the like. Examples of solid tumors include, but are not limited to, malignancies such as sarcomas (including soft tissue sarcoma and osteosarcoma), adenocarcinomas, and carcinomas of various organ systems such as head and neck (including pharynx), thyroid, lung (small cell or non-small cell lung cancer (NSCLC)), breast, lymphoid tissues, gastrointestinal tract (eg, oral cavity, esophagus, stomach, liver, pancreas, small intestine, colon and rectum, anal canal), genitals and genitourinary tract (eg , kidney, urothelium, bladder, ovary, uterus, cervical, endometrium, prostate, testis), central nervous system (eg, affecting nerve or glial cells, such as neuroblastoma or glioma), skin (eg, melanoma), and the like . In some embodiments of the present invention, the solid tumor is an NMDA receptor-positive teratoma. In other embodiments of the present invention, the cancer is selected from breast cancer, colon cancer, pancreatic cancer (eg, pancreatic neuroendocrine tumor (PNET) or pancreatic ductal adenocarcinoma (PDAC)), stomach cancer, uterine cancer, and ovarian cancer.

- 46 045361

In some embodiments of the present invention, the cancer is selected from, but is not limited to, leukemias, such as acute lymphoblastic leukemia (ALL), chronic lymphocytic leukemia (CLL), acute myelogenous leukemia (AML), chronic myelogenous leukemia (CML), villous leukemia cells (HCL), T-cell prolymphocytic leukemia (T-PLL), large granular lymphocyte leukemia, adult T-cell leukemia and acute monocytic leukemia (AMoL).

In one embodiment of the present invention, the cancer is a solid tumor that is associated with ascites. Ascites is a symptom of many types of cancer and can also be caused by a number of conditions, such as advanced liver disease. Types of cancers that can cause ascites include, but are not limited to, breast, lung, large intestine (colon), stomach, pancreatic, ovarian, uterine (endometrial), abdominal, and the like. In some embodiments of the present invention, the solid tumor associated with ascites is selected from breast cancer, colon cancer, pancreatic cancer, gastric cancer, uterine cancer, and ovarian cancer. In some embodiments of the present invention, the cancer is associated with pleural effusions, such as lung cancer.

Additional hematologic malignancies that cause solid tumors include, but are not limited to, non-Hodgkin lymphoma (eg, diffuse large B cell lymphoma, mantle cell lymphoma, B lymphoblastic lymphoma, peripheral T cell lymphoma, and Burkitt lymphoma); B-cell chronic lymphocytic leukemia/small cell lymphocytic lymphoma; lymphoplasmacytic lymphoma; B-cell marginal zone lymphoma of the spleen (± villous lymphocytes); plasma cell myeloma/plasmocytoma; extranodal B-cell lymphoma from marginal zone cells of the MALT type; nodular marginal zone B cell lymphoma (±monocytoid B cells); follicular lymphoma; diffuse large B-cell lymphoma; Burkitt's lymphoma; T-lymphoblastic lymphoma from progenitor cells; adult T-cell lymphoma (HTLV 1-positive); extranodal NK/T cell lymphoma of the nasal type; Enteropathic type T-cell lymphoma; hepatosplenic γ-δ T-cell lymphoma; subcutaneous panniculitis-like T-cell lymphoma; fungoid mycosis/Sézary syndrome; primary cutaneous anaplastic large cell lymphoma with T/0 cells; primary systemic anaplastic large cell lymphoma with T/0 cells; peripheral T-cell lymphoma, unspecified; angioimmunoblastic T-cell lymphoma, multiple myeloma, polycythemia vera or myelofibrosis, cutaneous T-cell lymphoma, small cell lymphocytic lymphoma (SLL), marginal zone lymphoma, CNS lymphoma, immunoblastic large cell lymphoma, B-lymphoblastic progenitor cell leukemia and the like.

In specific embodiments of the present invention, the cancer is sarcoma, colorectal cancer, head and neck cancer, lung cancer, ovarian cancer, pancreatic cancer, gastric cancer, melanoma, and/or breast cancer.

Antibodies against CD47 and shielded antibodies, as described herein, can also be used to treat disorders associated with cancer, for example, encephalopathy caused by cancer.

The methods and compositions of this invention can be used in combination with other therapeutic agents and/or methods. The term combination, as used herein, means that two (or more) different therapies are provided to a subject during the course of the subject's illness with the disease such that the effects of the patient's therapies overlap at a particular point in time. In some embodiments of the present invention, delivery of one therapy is still occurring when delivery of the second begins, so that there is an overlap in terms of drug administration. This is sometimes referred to herein as simultaneous or co-delivery of the drug. In other embodiments of the present invention, provision of one therapy ends before provision of another therapy begins. In some embodiments, therapy is more effective due to combined administration. For example, the second therapy is more effective, for example, an equivalent effect is observed with less of the second therapy, or the second therapy reduces symptoms to a greater extent than would have been the case if the second therapy had been given in the absence of the first therapy, or a similar situation occurs with the first therapy . In some embodiments of the present invention, the provision of therapy is such that the reduction in a symptom or other parameter associated with the disorder is greater than that which would be observed with one therapy provided in the absence of the other. The effect of the two therapies may be partially additive, fully additive, or greater than additive (ie, synergistic response). The provision of therapy may be such that the effect of the first therapy provided can still be detected when the second therapy is provided.

In one embodiment of the present invention, the methods of the present invention comprise administering to a subject an anti-CD47 antibody or a screened antibody as described herein

- 47 045361 those, for example, compositions or preparations, in combination with one or more additional therapies, for example surgery, radiation therapy or the administration of another therapeutic drug. In one embodiment of the present invention, the additional therapy may include chemotherapy, such as a cytotoxic agent. In one embodiment of the present invention, the adjunctive therapy may include a targeted therapy, such as a tyrosine kinase inhibitor, a proteasome inhibitor, or a protease inhibitor. In one embodiment of the present invention, the adjunctive therapy may include an anti-inflammatory, anti-angiogenic, anti-fibrotic or anti-proliferative compound, such as a steroid, an immunomodulator of biological action such as an immune checkpoint molecule inhibitor, a monoclonal antibody, an antibody fragment, an aptamer, a siRNA, an insensitive molecule, a fusion protein, cytokine, cytokine receptor, bronchodilator, statin, anti-inflammatory agent (eg, methotrexate) or NSAID. In another embodiment of the present invention, adjunctive therapy may include combining drugs from different classes. The anti-CD47 antibody or shielded antibody drug and the additional therapy may be administered simultaneously or sequentially.

An immune checkpoint molecule, as used in the context of this document, refers to a molecule in the immune system that either increases a signal (stimulatory molecule) or decreases a signal (inhibitory molecule). Many cancers evade the immune system by suppressing T-cell signaling.

Exemplary immune checkpoint molecules include, but are not limited to, programmed cell death protein 1 (PD-1), programmed cell death ligand 1 (PDL1), PD-L2, cytotoxic T lymphocyte associated protein 4 (CTLA-4), T-cell immunoglobulin and mucin domain-containing protein 3 (TIM-3), lymphocyte activation gene 3 (LAG-3), carcinoembryonic antigen-dependent cell adhesion molecule 1 (CEACAM-1), CEACAM-5, V domain immunoglobulin suppressor of T cell activation (VISTA), B- and T-lymphocyte attenuator (BTLA), T-cell immunoreceptor with Ig and ITIM domains (TIGIT), leukocyte-associated immunoglobulin receptor (type 1) (LAIR1), CD160, TGFR, A2 adenosine receptor (A2AR), B7-H3 (also known as CD276), B7-H4 (also called VTCN1), indoleamine 2,3-dioxygenase (IDO), 2B4, natural killer immunoglobulin-like receptor (KIR) and the like.

As used herein, the term immune checkpoint inhibitor refers to a molecule (eg, small molecule, monoclonal antibody, antibody fragment, etc.) that inhibits and/or blocks one or more inhibitory checkpoint molecules.

Exemplary immune checkpoint inhibitors include, but are not limited to, the following monoclonal antibodies: PD-1 inhibitors such as pembrolizumab (Keytruda, Merck) and nivolumab (Opdivo, Bristol-Myers Squibb); PD-L1 inhibitors such as atezolizumab (Tecentriq, Genentech), avelumab (Bavencio, Pfizer), durvalumab (Imfinzi, AstraZeneca); and CTLA-1 inhibitors such as ipilimumab (Yervoy, Bristol-Myers Squibb).

Exemplary cytotoxic agents include antimicrotubule agents, topoisomerase inhibitors, antimetabolites, protein synthesis and degradation inhibitors, mitotic inhibitors, alkylating agents, platinating agents, nucleic acid synthesis inhibitors, histone deacetylase inhibitors (HDAC inhibitors, e.g., vorinostat (SAHA, MK0683), entinostat (MS -275), panobinostat (LBH589), trichostatin A (TSA), mocetinostat (MGCD0103), belinostat (PXD101), romidepsin (FK228, depsipeptide)), DNA methyltransferase inhibitors, nitrogen mustards, nitrosoureas, ethylene imines, alkyl sulfonates, triazenes, analogues folate, nucleoside analogs, ribonucleotide reductase inhibitors, vinca alkaloids, taxanes, epothilones, intercalating agents, agents capable of interfering with the signal transduction pathway, agents that promote apoptosis and irradiation, or conjugates of antibody molecules that bind surface proteins to deliver a toxic agent. In one embodiment, the cytotoxic agent that can be administered with a drug described herein is a platinum-based agent (such as cisplatin), cyclophosphamide, dacarbazine, methotrexate, fluorouracil, gemcitabine, capecitabine, hydroxyurea, topotecan, irinotecan, azacitidorine, azacitidine, ixabepilone, bortezomib, taxanes (for example, paclitaxel or docetaxel), cytochalasin B, gramicidin D, ethidium bromide, emetine, mitomycin, etoposide, tenoposide, vincristine, vinblastine, vinorelbine, colchicine, anthracyclines (for example, doxorubicin or epirubicin) daunorubicin, dihydroxy anthracine dione, mitoxantrone, mithramycin, actinomycin D, adriamycin, 1-dehydrotestosterone, glucocorticoids, procaine, tetracaine, lidocaine, propranolol, puromycin, ricin or maytansinoids.

The methods and compositions of this invention can be used in the treatment of subjects with CD47-positive cancer. In one embodiment of the present invention, a CD47-positive cancer is characterized by the expression of one or more matrix metalloproteinases (MMPs). Exemplary MMPs include, but are not limited to, MMP1-MMP28. In particular, exemplary MMPs include MMP2 and MMP9. In one embodiment of the invention, the CD47-positive cancer is a tumor in which, when

- 48 045361 there are infiltrating macrophages.

The methods and compositions of this invention can be used in the treatment of subjects with CD47-positive cancer that is characterized by the expression of one or more MMPs and contains infiltrating macrophages.

Methods for determining the presence of CD47-positive cancers, MMP expression, and the presence of tumor-infiltrating macrophages are known in the art.

The assessment of CD47-positive cancers in a subject can be determined by conventional methods, which include immunohistochemistry (IHC), Western blotting, flow cytometry, or RNA sequencing methods. Any anti-CD47 antibodies known in the art, as well as the anti-CD47 antibodies described herein, can be analyzed by IHC, Western blotting and flow cytometry.

Assessing macrophage infiltration into tissues can be accomplished by monitoring macrophage surface markers, including F4/80 for mouse macrophages or CD163, CD68, or CD11b, using conventional methods that include immunohistochemistry (IHC), Western blot TING, flow cytometry, or RNA sequencing methods.

The assessment of proteases in tissues can be monitored using a variety of methods, including both those that monitor protease activity and those that can detect proteolytic activity. Common methods that can detect the presence of proteases in tissue, which may include both inactive and active forms of protease, include IHC, RNA sequencing, Western blotting or ELISA-based methods. Additional methods can be used to detect protease activity in tissues, which include zymography, in situ zymography using fluorescence microscopy, or the use of fluorescent proteolytic substrates. In addition, the use of fluorescent proteolytic substrates can be combined with immunocapture of specific proteases. In addition, antibodies directed against the active site of the protease can be used in a variety of ways, including IHC, fluorescence microscopy, Western blotting, ELISA, or flow cytometry (see Sela-Passwell et al. Nature Medicine. 18:143-147. 2012; LeBeau et al. Cancer Research. 75:1225-1235. 2015; Sun et al. Biochemistry. 42:892-900. 2003; Shiryaev et al. 2:e80. 2013.)

Throughout the specification, where compositions and kits are described as having, including or containing specific components, or where processes and methods are described as having, including or containing specific steps, it is intended that there are, in addition, compositions and kits of this invention that essentially consist of or consist of the listed components, and that there are processes and methods of this invention that consist essentially of or consist of the listed process steps and methods.

IV. Pharmaceutical compositions and preparations

For therapeutic use, the anti-CD47 antibody or shielded antibody is preferably combined with a pharmaceutically acceptable carrier. As used in this document, the term pharmaceutically acceptable carrier means buffers, carriers and excipients suitable for use in contact with tissues of humans and animals without undue toxicity, irritation, allergic response or other problems or complications, commensurate with a reasonable balance of benefit and risk. The carrier(s) must be acceptable in the sense that it is compatible with the other components of the preparations and is not harmful to the recipient. Pharmaceutically acceptable carriers include buffers, diluents, dispersion media, coatings, isotonic and absorption delaying agents, and the like, which are compatible with pharmaceutical administration. The use of such media and agents for pharmaceutically active substances is known in the art.

Accordingly, the anti-CD47 antibody or shielded antibody compositions of this invention may contain at least one of any suitable excipients such as, but not limited to, a diluent, a binder, a stabilizer, buffers, salts, lipophilic solvents, a preservative, an adjuvant or the like. Pharmaceutically acceptable excipients are preferred. Non-limiting examples and methods for preparing such sterile solutions are well known in the art, such as, but not limited to, those described in Gennaro, Ed., Remington's Pharmaceutical Sciences, 18th Edition, Mack Publishing Co. (Easton, Pa.) 1990. Pharmaceutically acceptable carriers can generally be selected from those that are suitable for the mode of administration, solubility and/or stability of the antibody molecule, fragment or composition variant that are well known in the art or described herein .

Pharmaceutical excipients and additives used in this composition include, but are not limited to, proteins, peptides, amino acids, lipids and carbohydrates (e.g., sugars, including monosaccharides, di-, tri-, tetra- and oligosaccharides; derivatized sugars such as alditols, aldonic acids, esterified sugars and the like and polysaccharides or sugar polymers), which may be present alone or in combination, containing one or a combination of 1- 49 045361

99.99% by weight or volume. Exemplary protein excipients include serum albumin such as human serum albumin (HSA), recombinant human albumin (rHA), gelatin, casein, and the like. Exemplary amino acid/antibody molecule components that may also function in the buffer capacity include alanine, glycine, arginine, betaine, histidine, glutamic acid, aspartic acid, cysteine, lysine, leucine, isoleucine, valine, methionine, phenylalanine, aspartame, and the like. .

Carbohydrate excipients suitable for use in the invention include, for example, monosaccharides such as fructose, maltose, galactose, glucose, D-mannose, sorbose and the like; disaccharides such as lactose, sucrose, trehalose, cellobiose and the like; polysaccharides such as raffinose, melezitose, maltodextrins, dextrans, starches and the like; and alditols such as mannitol, xylitol, maltitol, lactitol, xylitol, sorbitol (glucite), myoinositol and the like. Preferred carbohydrate excipients for use in this invention are mannitol, trehalose and raffinose.

The antibody molecule compositions may also include a buffer or pH adjusting agent; Typically the buffer is a salt derived from an organic acid or base. Exemplary buffers include salts of organic acids, such as salts of citric acid, acetic acid, ascorbic acid, gluconic acid, carbonic acid, grape acid, succinic acid, or phthalic acid; Tris, tromethamine hydrochloride or phosphate buffers.

In addition, the antibody molecule compositions of the present invention may include polymeric excipients/additives such as polyvinylpyrrolidones, ficolls (polymer sugars), decrates (e.g., cyclodextrins such as 2-hydroxypropyl-β-cyclodextrin), polyethylene glycols, flavoring agents, antimicrobial agents, sweeteners, antioxidants, antistatic agents, surfactants (eg, polysorbates such as Tween 20 and Tween 80), lipids (eg, phospholipids, fatty acids), steroids (eg, cholesterol) and chelating agents (eg, EDTA).

These and additional known pharmaceutical excipients and/or additives suitable for use in the antibody molecule compositions of this invention are known in the art, for example, as described in Remington: The Science & Practice of Pharmacy, 19th ed., Williams & Williams, (1995), and in Physician's Desk Reference, 52nd ed., Medical Economics, Montvale, N.J. (1998). Preferred carriers or excipients are carbohydrates (eg, saccharides and alditols) and buffers (eg, citrate) or polymeric agents.

This invention relates to stable compositions containing at least one anti-CD47 antibody molecule in a pharmaceutically acceptable composition. The canned preparations contain at least one known preservative or it is optionally selected from at least one phenol, m-cresol, p-cresol, o-cresol, chlorocresol, benzyl alcohol, phenylmercuric nitrite, phenoxyethanol, formaldehyde, chlorobutanol, magnesium chloride (eg hexahydrate) of alkylparaben (methyl, ethyl, propyl, butyl and the like), benzalkonium chloride, benzethonium chloride, sodium dehydroacetate and thimerosal or mixtures thereof in an aqueous diluent. Any suitable concentration or mixture may be used as is known in the art, for example, 0.001-5%, or any range or value therein, for example, but not limited to, 0.001, 0.003, 0.005, 0.009, 0.01, 0 .02, 0.03, 0.05, 0.09, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9 , 1.0, 1.1, 1.2, 1.3, 1.4, 1.5, 1.6, 1.7, 1.8, 1.9, 2.0, 2.1, 2 ,2, 2.3, 2.4, 2.5, 2.6, 2.7, 2.8, 2.9, 3.0, 3.1, 3.2, 3.3, 3.4 , 3.5, 3.6, 3.7, 3.8, 3.9, 4.0, 4.3, 4.5, 4.6, 4.7, 4.8, 4.9 or any range or value within it. Non-limiting examples include, without preservative, 0.12% m-cresol (for example, 0.2, 0.3, 0.4, 0.5, 0.9 or 1.0%), 0.1-3% benzyl alcohol (eg 0.5, 0.9, 1.1., 1.5, 1.9, 2.0 or 2.5%), 0.001-0.5% thimerosal (eg 0.005 or 0.01 %), 0.001-2.0% phenol (e.g. 0.05, 0.25, 0.28, 0.5, 0.9 or 1.0%), 0.0005-1.0% alkylparaben(s) ) (for example, 0.00075, 0.0009, 0.001, 0.002, 0.005, 0.0075, 0.009, 0.01, 0.02, 0.05, 0.075, 0.09, 0.1, 0.2, 0.3, 0.5, 0.75, 0.9 or 1.0%), and the like.

Pharmaceutical compositions containing an anti-CD47 antibody or a shielded antibody as described herein may be presented in unit dosage form and may be prepared by any suitable method. The pharmaceutical composition must be formulated so that it is compatible with the intended route of administration. Examples of routes of administration include intravenous (IV), intradermal, inhalation, transdermal, topical, transmucosal and rectal administration. The preferred route of administration of monoclonal antibodies is intravenous infusion. Useful preparations can be prepared by methods known in the pharmaceutical field. For example, see Remington's Pharmaceutical Sciences (1990), cited above. Components of the composition suitable for parenteral administration include a sterile diluent such as water for injection, saline, fatty oils, polyethylene glycols, glycerin, propylene glycol or other synthetic solvents; antibacterial agents such as benzyl alcohol or methylparabens; antioxidants such as ascorbic acid or sodium bisulfite; chelating agents such as EDTA; buffers such as acetate, citrate or phosphate; and tone-regulating agents such as sodium chloride or dextrose.

For intravenous administration, suitable vehicles include physiological saline,

- 50 045361 bacteriostatic water, Cremophor EL™ (BASF, Parsippany, N.J.) or phosphate buffered saline (PBS). The carrier must be stable under the conditions of manufacture and storage and must be protected from microorganisms. The carrier may be a solvent or dispersion medium containing, for example, water, ethanol, polyol (eg glycerol, propylene glycol and liquid polyethylene glycol) and suitable mixtures thereof.

Pharmaceutical preparations are preferably sterile. Sterilization can be accomplished by any suitable method, for example by filtration through sterile filter membranes. If the composition is lyophilized, filter sterilization may be performed before or after lyophilization and reconstitution.

The compositions of this invention may take various forms. These include, for example, liquid, semi-solid and solid dosage forms such as liquid solutions (eg injections and infusions), dispersions or suspensions and liposomes. The preferred dosage form depends on the intended route of administration and therapeutic use. Typical preferred compositions are in the form of injections or infusions. The preferred route of administration is parenteral (eg, intravenous, subcutaneous, intraocular, intraperitoneal, intramuscular). In a preferred embodiment of the present invention, the drug is administered by intravenous infusion or injection. In a preferred embodiment of the present invention, the drug is administered by intravenous infusion or injection.

The phrases parenteral administration and parenteral administration, as used herein, mean modes of administration other than enteral and local administration, usually by injection, and include, but are not limited to, intravenous, intramuscular, subcutaneous, intra-arterial, intrathecal, intracapsular, intraorbital, intravitreal, intracardial, intradermal, intraperitoneal, transtracheal, inhalation, subcutaneous, subcuticular, intraarticular, subcapsular, subarachnoid, intraspinal, epidural and intrasternal injection and infusion.

The present invention provides a kit comprising packaging material and at least one vial containing a solution of at least one anti-CD47 antibody or screened antibody with provided buffers and/or preservatives, optionally in an aqueous diluent. The aqueous diluent optionally further contains a pharmaceutically acceptable preservative. Preservatives include those selected from phenol, m-cresol, p-cresol, o-cresol, chlorocresol, benzyl alcohol, alkylparaben (methyl, ethyl, propyl, butyl and the like), benzalkonium chloride, benzethonium chloride, sodium dehydroacetate and thimerosal or mixtures thereof. The concentration of the preservative used in the preparation is a concentration sufficient to obtain an antimicrobial effect. Such concentrations depend on the preservative chosen and are readily determined by one skilled in the art.

Other excipients, such as isotonic agents, buffers, antioxidants, preservative enhancers, may optionally and preferably be added to the diluent. An isotonic agent such as glycerol is usually used in known concentrations. A physiologically tolerated buffer is preferably added to improve pH control. The formulations may cover a wide range of pH, for example, from about 4.0 to about 10.0, from about 5.0 to about 9.0, or from about 6.0 to about 8.0.

Other additives such as pharmaceutically acceptable solubilizers such as Tween 20 (polyoxyethylene (20) sorbitan monolaurate), Tween 40 (polyoxyethylene (20) sorbitan monopalmitate), Tween 80 (polyoxyethylene (20) sorbitan monooleate), Pluronic F68 (block copolymers polyoxyethylene and polyoxypropylene) and PEG (polyethylene glycol) or nonionic surfactants such as polysorbate 20 or 80 or poloxamer 184 or 188, Pluronic® polyls, other block copolymers and chelators such as EDTA and EGTA may optionally be added to drugs or compositions to reduce aggregation. These supplements are especially useful if a pump or plastic container is used to administer the drug. The presence of a pharmaceutically acceptable surfactant reduces the tendency of the protein to aggregate.

Various delivery systems can be used to administer anti-CD47 antibodies or shielded antibodies to a subject. In some illustrative embodiments of the present invention, administration of the anti-CD47 antibody or shielded antibody is accomplished by intravenous infusion. In some embodiments of the present invention, the administration is a two-hour intravenous infusion.

Any of the preparations described above may be stored in liquid or frozen form and may optionally be subjected to a preservation process. In some embodiments of the present invention, the preparations described above are lyophilized, that is, they are lyophilized. In some embodiments of the present invention, the preparations described above are subjected to a preservation process, such as lyophilization, and then reconstituted with a suitable liquid, such as water. By lyophilization is meant that the composition has been lyophilized in

- 51 045361 vacuum. Lyophilization is usually achieved by freezing the specific drug so that the solutes are separated from the solvent(s). The solvent is then removed by sublimation (ie primary drying) and then desorption (ie secondary drying).

The formulations of this invention may be used with the methods described herein or with other methods of treating a disease. Anti-CD47 or shielded antibody preparations may be further diluted before administration to a subject. In some embodiments of the present invention, the drugs are diluted with saline and stored in bags or syringes for intravenous infusion before administration to the subject. Accordingly, in some embodiments of the present invention, methods of treating a cancer characterized by CD47 expression in a subject will comprise administering to the subject in need thereof a weekly dose of a pharmaceutical composition comprising an anti-CD47 antibody or a shielded antibody.

It will be apparent to those skilled in the art that other suitable modifications and adaptations of the methods described herein can be made using suitable equivalents without departing from the scope of the embodiments of the present invention described herein. Having now described in detail certain embodiments of the present invention, they will be better understood by reference to the following examples, which are included for purposes of illustration only and are not intended to be limiting. All patents, patent applications and references described herein are hereby incorporated by reference in their entirety and for all purposes.

Examples

Example 1. Preparation of an antibody.

Humanized versions of the mouse anti-CD47 antibody B6H12 were obtained. Antibody variable domain sequences for DNA and vector generation were synthesized using non-template PCR. Briefly, the virtual gene sequence is converted into oligonucleotide sequences using the proprietary ATUM software package. Oligonucleotides are synthesized, combined and amplified using PCR. The full-length amplicon from the PCR reaction is cloned into a vector using the Sapl site, transformed into E. coli, and unique colonies are isolated. Colonies are grown overnight in liquid medium and plasmid DNA is isolated, purified using the Machery-Nagel Midi Prep Kit, and sequence verified using Sanger sequencing. The light chain variable domains are cloned into the Carr-Hs vector and the heavy chain domains are cloned into the IgG1.01-Fc-Hs vectors. For antibody expression, a 1:1 ratio of antibody heavy and light chain vectors is diluted in LifeTech Optimem medium with PolyPlus FectoPro transfection reagent. The DNA/transfection reagent is then added to Atum-specific HEK293 cells in LifeTech ExpiExpression medium and cultured for 5 days. The culture is collected by centrifugation and filtration through 0.2 μm pores. For IgG purification, GE's MabSelect is a must for antibody purification. Before elution, the resin is washed with 10CV 1 M NaCl in PBS and 10CV PBS. IgG is eluted using 20 mM Na-citrate buffer pH 3.2. The sample is buffer exchanged using Zeba columns from Pierce in PBS. The sample is then filter sterilized before the specimen is collected for testing. The study involved concentration of A280 and electrophoresis in reduced and unreduced states using an Agilent P200 Tape Station 2200 on Tapescreen p200 chips.

Specific mutations of the parent antibody B6H12 are described in Tables 10-13, as follows.

Table 10

Humanizing Mutations in the hB6H12 Heavy Chain Variants

vH Option Subsequence IGHV Acceptor Exon Remains of Karkasnaya Donor Mouse Sequences Leftovers Acceptor CDR Sequences Human hvHl IGHV3-23/HJ4 H44, H49, H89, H91, H94 none hvH2 IGHV3-23/HJ4 H49, H91, H94 Н31,НЗЗ, Н60 hvH3 IGHV3-23/HJ4 H49, H82, H91, H94 Н31,НЗЗ, Н60 hvH4 IGHV3-48/HJ4 H49 H31,H60 hvH5 IGHV3-66/HJ4 H29, H49, H82 H60 hvH6 IGHV3-74/HJ4 H49 H31,H58,H60

- 52 045361

Table 11

Humanizing mutations in hB6H12 kappa light chain variants

Option vH Subsequence IGKV Acceptor Exon Remains of Karkasnaya Donor Mouse Sequences Leftovers Acceptor CDR Sequences Human hvKl IGKV6-21/KJ2 L4, L21,L85 none hvK2 IGKV6-21/KJ2 L4, L21 none hvK3 IGKV6-21/KJ2 L4, L21,L69 none hvK4 IGKV 1-27 /KJ2 L21,L49, L69 L31, L34, L53, L54, L55

Table 12

Mouse framework-specific mutations in hB6H12 heavy chain variants

Option 29 44 49 82 89 91 94 % Balance Human Sequence hvHl F R* A* M I* p* R* 87.8 hvH2 F G A* M V p* R* 92.9 hvH3 F G A* I* V p* R* 91.8 hvH4 F G A* M V Y R 92.9 hvH5 p* G A* I* V Y R 89.8 hvH6 F G A* M V Y R 92.9

*Remnants of mouse sequence

Table 13

Specific mutations of mouse framework sequences in

Antibodies were expressed by transient transfection of Expi HEK or Expi CHO cells or stable transfection of CHO-DG44 and purified using MabSelect SuRe columns (GE Healthcare). Additional preparatory purification by size exclusion chromatography using Superdex columns (GE Healthcare) was performed on screened antibodies that were less than 90% monomeric.

Saturation binding study using flow cytometry and ELISA.

Humanized anti-CD47 antibodies B6H12 with different heavy and light chain sequences were assessed for their binding affinity to human CD47 by either saturating ELISA or cell-based FACS analysis, and the EC50 or Kd was calculated. Only antibodies

- 53 045361 consisting of a heavy chain 1 (hvH1) sequence (group A; antibodies 1-4) or a heavy chain 5 (hvH5) sequence (group B, antibodies 17-20) were able to bind to CD47. Antibodies from these groups bind with similar affinity to Kd (Fig. 2A) and _EC50 (Fig. 2B) as the murine mB6H12 antibody and the alternative humanized antibody Ab47.

For cellular FACS analysis, L450cy lymphoma cancer cells were exposed to increasing concentrations of humanized B6H12 antibodies, which were found to retain binding to CD47 with defined Kd values. The alternative humanized framework regions retained similar binding to Kd as the mouse parent and the alternative humanized antibody Ab47 (Fig. 2A).

In addition to cellular binding, the affinity of humanized B6H12 antibodies for CD47 was assessed using ELISA. Plates coated with human CD47 were treated with increasing concentrations of humanized B6H12 antibodies, which were found to retain binding to CD47 with defined EC ₅₀ values. The alternative humanized framework regions retained similar binding to Kd as the mouse parent and the alternative humanized antibody Ab47 (Fig. 2B).

In addition to assessing EC50 binding, the ELISA assay also assessed binding kinetics. Surprisingly, the antibodies in Group A (hB6H12.3 and hB6H12.4) showed a significantly higher maximum binding value (BMax) than the parent antibody, mB6H12, the antibodies in Group B (hB6h12.19 and hB6H12.20) or the alternative humanized antibody Ab47 ( Fig. 2C).

Saturation binding study using flow cytometry 2x105 of the indicated cells (SW780 or human erythrocytes) were combined with a serial dilution of the indicated antibody in staining buffer (PBS, 5% FBS, 0.2% NaN ₃ ). Samples were incubated for 1 h on ice and washed twice with chilled staining buffer. Cells were resuspended with anti-human IgG antibody AF647 (JacksonlmmunoResearch, 1:200 dilution in staining buffer) for 1 h on ice. Cells were washed twice with chilled staining buffer and resuspended in staining buffer. Labeled cells were examined by flow cytometry on an Invitrogen Attune NxT flow cytometer designed to exclude nonviable cells and analyzed using FlowJo 10 software. Kd was calculated using GraphPad Prism 6 using nonlinear regression.

Saturation binding study using ELISA.

Soluble recombinant human CD47-Fc (R&D Systems) was diluted to the appropriate concentration in 50 mM carbonate buffer, pH 9.6. 100 μl of soluble antigen was added to each well of a 96-well Maxisorb ELISA plate. The plate was sealed and stored at 4°C overnight. Then the plate was washed 3-5 times with PBS-T buffer (PBS, pH 7.4+0.05% Tween-20). Wells were blocked using 300 μl/well of PBS-T buffer containing BSA for 1 h at room temperature, then washed 3-5 times with PBS-T. Antibody dilutions were prepared in blocking buffer and added to each well in a volume of 100 μl. Antibodies were incubated for 1 h at room temperature and then washed 3–5 times with PBS-T. HRP-conjugated secondary antibodies (either anti-human Fc or anti-human kappa light chain) were then added and incubated for 1 h at room temperature. The plate was washed 3-5 times with PBS-T. ELISA was performed by adding 100 μl of TMB solution and incubating for 3–15 minutes at room temperature. To stop the reaction, 100 μl of 1 N was added to each well. sulfuric acid. Absorbance at 450 nm was determined using a Spectramax 190 plate reader (Molecular Biosciences) and data were plotted using GraphPad Prism 6.

B6H12-mediated phagocytosis.

Functional characterization of humanized anti-B6H12 antibodies was carried out, including phagocytosis in human erythrocytes and hemagglutination. Human erythrocytes, labeled with the fluorescent red dye PKH, were opsonized for 30 min with increasing concentrations of humanized antibodies against B6H12 from Groups A and B. The erythrocytes were washed and fed in a 10:1 ratio to monocytic macrophages for 2 h. The samples were washed 3 times with hypotonic buffer ACK lysis, which allows for the lysis and removal of unabsorbed red blood cells. Samples were subjected to flow cytometry and phagocytosis was assessed. Antibodies from Groups A and B demonstrated similar abilities to mediate phagocytosis of human erythrocytes. Surprisingly, Group A antibody hB6H12.3 mediated phagocytosis better than the mouse parent antibody and on par with the alternative humanized antibody Ab47.

Humanized antibodies B6H12 from Antibody Groups A and B mediate phagocytosis of CD47-positive human erythrocytes in a similar manner. hB6H12.3 stimulated greater phagocytosis-promoting activity at 1 μg/ml compared to mouse mB6H12 and similar phagocytosis-promoting activity compared to Ab47 (Fig. 3A and Fig. 3B).

B6H12-mediated hemagglutination.

A distinctive feature of B6H12 is the ability to stimulate hemagglutination of red blood cells

- 54,045,361 people. Hemagglutination is an example of a homotypic interaction that allows two cells expressing CD47 to aggregate or stick together when exposed to an anti-CD47 antibody. The agglutinating scaffold maintains the red blood cell mass in a suspended arrangement, which can be quantified visually by image analysis or the level of aggregation monitored by flow cytometry. Human red blood cells suspended in PBS and plated in a 96-well round bottom plate were exposed to increased concentrations of anti-CD47 antibodies. After 30 min at 37°C, hemagglutination was monitored optically as a change in erythrocyte density and using flow cytometry as an increase in cell aggregation. Antibodies in Group A, particularly hB6H12.3, showed reduced hemagglutination compared to the parent antibody mB6H12 and the alternative humanized antibody Ab47 (Fig. 4A).

To assess hemagglutination, standard image capture was performed and the formation of dispersed non-precipitated red blood cells was assessed. In addition, the plates were scanned using a GE In Cell analyzer and the diameter of the visible spot inside the well was assessed. In order to assess hemagglutination, flow cytometry was used in an assay that is amenable to clinical monitoring, and hemagglutination was assessed as the overall increase in the observed parameter SSC/FSC of red blood cells, which is a way of monitoring aggregation. This method was used to evaluate hemagglutination caused by humanized anti-CD47 antibodies hB6H12.3 and Ab47 (Fig. 4A and Fig. 4B).

B6H12-mediated activation of Fcy receptors.

In vivo, monocytes, macrophages, neutrophils and dendritic cells can mediate ADCP via FcyRIIa, FcyRI and FcyRIIIa. Although all three receptors may be involved in ADCF, FcyRIIa is thought to be the predominant Fcy receptor involved in this process. Jurkat cells stably transfected with human FcyRI or high-affinity FcyRIIa-H coupled to an NFAT-luciferase reporter construct were exposed to WIL2S cells coated with increasing concentrations of either mouse B6H12, Ab47, or hB6H12.3. FcyR activation was monitored by luciferase production. The results are presented in Fig. 5A and 5B. Cells coated with both Ab47 and hB6H12.3 dose-dependently activated the FcyRI receptor, but only hB6H12.3 was able to drive activation of FcyRIIa-H, the receptor most closely associated with directly mediating ADCP activity.

Antibody-mediated activation of Fcy receptors involved in antibody-dependent cellular phagocytosis was assessed using engineered Jurkat cells stably expressing FcyRI or FcyRIIa-H (high-affinity variant H131) and an NFAT response element driving firefly luciferase expression as effector cells. The biological activity of antibodies was quantified through luciferase produced by activation of the NFAT pathway; Luciferase activity in effector cells was quantified by luminescence readings (Figures 5A and 5B).

NK cell-mediated ASCC activity.

NK cell-mediated antibody-dependent cellular cytotoxicity (ADCC) and FcyRIIIa activation of mB6H12, Ab47, and hB6H12.3 were examined. Chromium-labeled Wils-2 cells coated with increasing concentrations of antibodies were exposed to primary natural killer (NK) cells for 4 h and specific lysis was assessed by the release of radioactive chromium into the tissue culture medium. Alternatively, Wils-2 cells coated with increasing concentrations of mB6H12, Ab47, or hB6H12.3 were exposed to Jurkat cells stably expressing the high affinity V/V variant of FcyRIIIa, and receptor activation was assessed as NFAT-driven luciferase activity. hB6H12.3 demonstrated higher activation of ADCC and FcyRIIIa compared to both Ab47 and the mouse parent antibody B6H12. The results are presented in Fig. 6A and 6B. The hB6H12.3 IgG1 scaffold mediates ADCC activity and drives FcyRIII signaling, functionality lacking in current clinical IgG4 antibodies.

For the ADCC assay, purified human NK cells, negatively selected from Astarte Biologies, were thawed and resuspended in RPMI/1% FBS at a concentration of 7.2 x 105 CD16+ cells/ml (so that 70 µl contained approximately 5 x 104 effector cells). 5x106 target cells (WIL2S cells) were collected, centrifuged and resuspended in 100 µl FBS. 100 μL (approximately 100 μCi) Cr-51 was added to the cells and mixed gently. Cells were placed in a CO ₂ incubator with a humidification system at 37°C for labeling for 1 hour. The cells were shaken occasionally to keep them suspended. Cells were washed three times with RPMI/1% FBS, taking care to remove the radioactive supernatant into an appropriate waste container. Cells were then resuspended in 10 ml RPMI/1% FBS and counted. 7.2 x 105 cells were removed and suspended in a total volume of 10 ml assay medium, so that 70 µl was equivalent to approximately 5 x 103 target cells.

- 55 045361

For antibody dilution and plate assembly, antibodies were diluted in assay media (Prep @ 3X). Antibodies were added to the plate immediately before the addition of Cr-labeled target cells. To control wells, instead of antibodies, 70 μl and 140 μl of assay medium were added. These wells represent total and experimental release controls, respectively. Target cells were mixed and 70 μl was added to each test and control well of a 96-well plate. Targets with antibodies were incubated in an incubator at 37°C for 30 min. 70 µl (5 x 10 ⁴ ) effector cells were added to each test well. 70 μL of 3% Triton X-100 was added to total release wells and pipetted up and down three times to mix. The plates were returned to 37°C in a CO2 incubator with a humidification system for 4 hours.

Taken together, the data demonstrate that the new humanized anti-CD47 antibodies of this invention have superior and unexpected properties compared to the mouse parent antibody. Humanized anti-CD47 antibodies exhibit superior phagocytic/ADCP capacity, as demonstrated by phagocytosis of erythrocytes (Figure 3 B) and FcyRII signaling (Figure 5 B). Humanized anti-CD47 antibodies exhibit excellent cell lysis/ADCC, as demonstrated by NK cell-mediated cell lysis (Figure 6A) and FcyRIII signaling (Figure 6B). Humanized anti-CD47 antibodies exhibit superior FcyRI signaling (Fig. 5A). Finally, humanized anti-CD47 antibodies also exhibit an excellent toxicity profile as demonstrated by reduced hemagglutination as demonstrated in hemagglutination assays (Figure 4A and Figure 4B). A humanized anti-CD47 antibody, hB6H12.3, was found to be particularly effective.

Suppressor function of antibodies against CD47.

In addition to blocking phagocytic activity, the interaction of CD47 with SIRPa also reduces the production of inflammatory cytokines through activation of SHP1 and inactivation of downstream signals such as Vav. Human monocytes were exposed to LPS to stimulate cytokine production in the presence of parent and various humanized anti-CD47 antibodies. Exposure to all CD47-targeting antibodies was able to significantly enhance LPS-driven cytokines. One consequence of better cytokine production and innate cellular activation is the ability to stimulate secondary T cell responses. Tumor-associated macrophages are known to have a suppressive phenotype that is unable to maintain and can actively suppress secondary T cell activation. Human mononuclear cells were differentiated and polarized with IL10 and MSCF to direct them toward an immunosuppressive TAM-like phenotype. These differentiated monocytes are able to suppress CD3/CD28-driven proliferation of autologous T cells. Adding anti-CD47 antibody exposure to this paradigm is able to activate TAMs and move them toward the M0/M1 phenotype (increased CD86/MHCII). See, for example, FIG. 7A, 7B and 7C. This change in the TAM-like phenotype correlated with the ability to support CD3/CD28-mediated T cell proliferation and activation. See, for example, FIG. 7D.

For the suppression assay, human monocytes differentiated into tumor-like macrophages with IL-10 were used, co-cultured with autologous T cells and assessed for the ability to support T-cell receptor-mediated activation. CD47 was added to cultures and was found to induce upregulation of macrophage activation markers, as measured by upregulation of CD86 and MHCII. In addition, support for TCR-mediated T cell activation was assessed by up-regulation of MHCII on T cells and INFu secretion (Figure 7A-Figure 7D).

Comparison with other anti-CD47 antibodies.

hB6H12.3 was compared with other known antibodies against CD47, 5F9 and CC-9002 (see WO 2011/143624). In comparison, FcyRI and FcyRII activities were measured in the same Jurkat cell NFAT-luciferase assay as described above. hB6H12.3 demonstrated superior ability to activate FcyRI and FcyRII compared to IgG4 antibodies 5F9 and CC-90002 (Fig. 8A and Fig. 8B).

In addition, NK-mediated secretion of ADCC and INFu was measured as described above. Again, hB6H12.3 demonstrated superior ability to mediate ADCC and stimulate INFu secretion compared with 5F9 and CC-90002 (Fig. 8C and Fig. 8D).

Example 3. Humanized anti-CD47 antibodies with a screening fragment.

Digestion of shielded antibodies using recombinant human MMPs for binding and function studies.

All recombinant MMPs were activated by incubation with 1.25 mM 4-aminophenylmercuric acetate (APMA) for 1–2 h at 37°C. Typically, 1-2 μg of activated rhMMP2 was added to 0.250.5 mg of screened antibody and incubated for 4-16 hours at 37°C. The extent of antibody cleavage was assessed using reversed-phase LC-MS of reconstituted antibodies using a Waters Acquity/Xevo UPLC equipped with a 3 μm PLRP-MS column (Agilent). Data were analyzed using UNIFI software (Waters). Reactivation with MMP results in site-specific cleavage of the shielding moiety at the putative cleavage site.

- 56 045361

After complete reactivation of the shielded antibodies, digested products were purified using MabSelect SuRe Protein A resin before use in binding assays.

Mass spectrometry data were obtained for the re-MMP2-screened antibody. The light chain mass after deconvolution was determined for Vel-IPV-hB6H12.3 before (Fig. 9A) and after (Fig. 9B) cleavage of recombinant human MMP2. The expected m/z value for the intact light chain is 28681 (recorded: 28680.8). The expected m/z value for the MMP2 cleaved antibody (LRSG-hB6H12.3) is 23969 (reported: 23968.4).

Antibodies against CD47, carrying a fragment sequence of amino acid residues of the cleavage site at the N-terminus, which would remain after proteolysis, were also generated by transient expression in Expi-HEK or Expi-CHO cells. For example, for a screened antibody using the MMP cleavage sequence IPVS-LRSG, the LRSG sequence remains after MMP activation.

Assessment of protease specificity of MMP-based cleavage sequences.

MMP was activated by incubation with 1.25 mM 4-aminophenylmercuric acetate (APMA) at 37°C for 1 hour and up to 24 hours. Legumain was activated after 2 hours of incubation at 37°C in 50 mM sodium acetate, 100 mM NaCl, pH 4.0.

To evaluate the cleavage profile of the cleavage sequences, antibodies (10 μg) were incubated overnight at 37°C with normalized proteases at 400 pmol/min as specified in the manufacturer's reports. The extent of antibody cleavage was assessed by reverse phase LC-MS of recovered antibodies using a Waters Acquity/Xevo UPLC equipped with a 3 μm PLRPMS column (Agilent). Data were analyzed using UNIFI software (Waters).

The protease cleavage profile of two protease cleavage sites (IPV and M2) was tested against a panel of tumor-associated proteases such as human and mouse/rat MMPs, ADAM family, uPA, matriptase and legumain. In addition, sequence cleavage by extracellular proteases tPA and factor Xa was also tested. Protease cleavage profiles at these three protease cleavage sites are presented in Table 1. 14.

Peptides (the N-terminus fused to the hBU12 antibody backbone) were incubated at 37°C overnight with 400 pmol/min normalized protease and assessed for degradation. The M2 site was cleaved by most MMPs, as well as uPA and matriptase. The IPV site was cleaved by almost all MMPs except MMP13 and was not affected by other classes of proteases.

- 57 045361

Table 14

Protease cleavage profiles at two different protease cleavage sites

Enzyme Cleavage Sequence: M2 IPV GPLG*VR** IPVS*LR**SG Human MMP2 Complete cleavage Complete cleavage Human MMP7 Complete cleavage Complete cleavage Human MMP9 Complete cleavage Complete cleavage Human MMP13 Complete cleavage Minimal splitting Mouse MMP2 Complete cleavage Partial splitting Mouse MMP7 Partial splitting Complete cleavage MMP9 rats Partial splitting Complete cleavage uPA Partial splitting Absent Matriptase Complete cleavage Minimal splitting Legumain Absent Minimal splitting tPA — Absent Factor Ha Minimal splitting Absent ADAM (human, mouse) Absent Absent

The MMP cleavage site is indicated by * and the uPA/matriptase/legumain cleavage sites are indicated by **. Cleavage at any site is sufficient to restore antibody binding.

Saturation with shielded anti-CD47 antibodies.

Anti-CD47 antibody saturation of human bladder cancer SW780 cells was performed. Vel-IPV-screened hB6H12 antibodies were tested together with pre-MMP2-activated comparators. The cleaved Vel-IPV antibodies had a fragment of amino acid residues of the cleavage site - LRSG at the N-termini of the antibody (Fig. 10 and Table 15).

Table 15

Determination of binding affinity (Kd) with SW780 cells obtained _________________by saturating them with antibodies (Fig. 10)._________________

Antibody Kd (nM) Split Vel-IPV-hB6H12.3 11.4 Split Vel-IPV-hB6H12.4 9.2 Split Vel-IPV-hB6H12.19 26.6 Split Vel-IPV-hB6H12.20 30.2 Vel-IPV-hB6H12.3 >2000 Vel-IPV-hB6H12.4 >2000 Vel-IPV-hB6H12.19 >2000 Vel-IPV-hB6H12.20 >2000

Vel-IPV-screened hB6H12 antibodies were tested together with pre-MMP2-activated comparators. Vel-IPV cleaved antibodies and IPV antibodies with a fragment of amino acid residues of the cleavage site contained the remainder of the LRSG sequence at the N-terminus of the antibody. The cleaved antibody was produced by cleavage with MMP2, whereas the IPV antibody with the cleavage site fragment was produced recombinantly (Fig. 11 and Table 16). As used herein, the term cleavage site amino acid fragment, cleavage site amino acid fragment antibody, or cleavage site amino acid fragment antigen binding fragment refers to an antibody or antigen binding fragment after cleavage of MMP, or an antibody or antigen binding fragment that has been produced recombinantly. method, i.e., obtained without a supercoiled domain and, therefore, no cleavage was performed. Cleavage of MMP results in a short fragment of the amino acid sequence of the cleavage site. For an IPV cleavage site, the cleavage site amino acid residue fragment must be LRSG or SG. For the M2 cleavage site, the fragment from the amino acid residues of the cleavage site must be VR.

Table 16

Determination of binding affinity (Kd) to SW780 cells obtained by saturating them with antibodies (Fig. 11).

Antibody Kd (nM) hB6H12.3 19.2 AB47 8.0 Vel-IPV-hB6H12.3 >2000 fragment1RU-LB6N12.3 18.1 Split Vel-IPV-hB6H12.3 14.2

Saturation of human erythrocytes with anti-CD47 antibodies was performed. Vel-IPV-screened hB6H12 antibodies were tested together with reactivated comparators (IPVhB6H12.3 with a cleavage site fragment or MMP2-cleaved VelIPV-hB6H12.3). Vel-IPV cleaved antibodies and IPV antibodies with a fragment of amino acid residues of the cleavage site contained the remainder of the LRSG sequence at the N-terminus of the antibody. The cleaved antibody was produced by cleavage with MMP2, whereas the IPV antibody with the cleavage site fragment was produced recombinantly (Fig. 12 and Table 17).

Table 17

Determination of binding affinity (Kd) to erythrocytes, obtained by saturating them with antibodies (Fig. 12).

Antibody Kd (nM) hB6H12.3 58.2 AB47 26.5 Vel-IPV-hB6H12.3 >2000 fragment1P V-hB 6H12.3 46.8 Split Vel-IPV-hB6H12.3 41.7

Saturation of anti-CD47 antibodies with rCD47 was performed using ELISA. Vel-IPV-hB6H12.3 exhibited significantly reduced binding. Binding can be restored after cleavage of rhMMP2 (Fig. 13 and Table 18).

Table 18

Determination of binding affinity (Kd) to rCD47 using ELISA (Fig. 13)

Antibody Kd (nM) hB6H12.3 1.7 Vel-IPV-hB6H12.3 1.7 Split Vel-IPV-hB6H12.3 >250

Anti-CD47 antibody saturation with rCD47 ELISA was performed with hB6H12.3 G91A, which has the G91A point mutation in LCDR3. Both hB6H12.3 and hB6H12.3 G91A showed a higher Bmax than Ab47 (Fig. 14 and Table 19).

Table 19

Determination of binding affinity (Kd) to rCD47 using ELISA (Fig. 14)

Antibody Kd (nM) AB47 1.0 hB6H12.3 1.4 НВ6Н12.3 G91A 2.7

Anti-CD47 antibody saturation of human bladder cancer SW780 cells was performed. Binding of Ab47 and hB6H12.3 was compared with variants carrying the G91A mutation in LCDR3 (Fig. 15 and Table 20)

Table 20

Determination of binding affinity (Kd) to SW780 human bladder cancer cells obtained by saturation (Fig. 15)

Antibody Kd (nM) AB47 6.8 Ab47 G91A 19.8 hB6H12.3 20.5 НВ6Н12.3 G91A 62.3

Saturation of human erythrocytes with anti-CD47 antibodies was performed. Binding of Ab47 and hB6H12.3 was compared with variants carrying the G91A mutation in LCDR3 (Fig. 16 and Table 21)

Table 21 Determination of binding affinity (Kd) with erythrocytes obtained by saturating them with antibodies (Fig. 16)

Antibody Kd (nM) AB47 12.1 Ab47 G91A 100.3 hB6H12.3 68.9 НВ6Н12.3 G91A > 150

Example 4. In vivo experiments.

Assessing macrophage infiltration into tissues can be accomplished by monitoring macrophage surface markers, including F4/80 for mouse macrophages or CD163, CD68, or CD11b, using conventional methods that include immunohistochemistry (IHC), Western blot TING, flow cytometry, or RNA sequencing methods.

The assessment of proteases in tissues can be monitored using a variety of methods, including both those that monitor protease activity and those that can detect proteolytic activity. Common methods that can detect the presence of proteases in tissue, which may include both inactive and active forms of protease, include IHC, RNA sequencing, Western blotting or ELISA-based methods. Additional methods can be used to detect protease activity in tissues, which include zymography, in situ zymography using fluorescence microscopy, or the use of fluorescent proteolytic substrates. In addition, the use of fluorescent proteolytic substrates can be combined with immunocapture of specific proteases. In addition, antibodies directed against the active site of the protease can be used in a variety of ways, including IHC, fluorescence microscopy, Western blotting, ELISA, or flow cytometry (see Sela-Passwell et al. Nature Medicine. 18:143-147. 2012; LeBeau et al. Cancer Research. 75:1225-1235. 2015; Sun et al. Biochemistry. 42:892-900. 2003; Shiryaev et al. 2:e80. 2013.)

The activity of anti-CD47 antibodies was determined in a xenograft model of L428 lymphoma in NSG mice. This xenograft model has high macrophage infiltration, as evidenced by strong staining using the mouse macrophage marker F4/80 on IHC. See, for example, FIG. 17V. Antibodies were administered IP q4dx4 using a dosage of 1 or 10 mg/kg. Shielded antibody Vel-IPV-hB6H12.3, as well as unshielded antibodies Ab47 and

- 60 045361 hB6H12.3 effectively reduced tumor volume during the study period using a dosage of 10 mg/kg. At a dose of 1 mg/kg, both Vel-IPV-hB6H12.3 and unshielded hB6H12.3 delayed tumor growth. The shielded Vel-IPV-hB6H12.3 antibody was slightly less active than the unshielded hB6H12.3 at this dose. (Fig. 17A).

The activity of anti-CD47 antibodies was determined in the Detroit 562 xenograft model of head and neck cancer in NSG mice. This xenograft model has high macrophage infiltration, as evidenced by strong staining using the mouse macrophage marker F4/80 on IHC. See fig. 18V. q4dx4 using a dosage of 5 mg/kg. Antibodies Ab47, hB6H12.3 and Vel-IPV-hB6H12.3 effectively reduced tumor volume during the study period (Fig. 18 A).

Anti-CD47 antibody activity was determined in the HT1080 fibrosarcoma xenograft model (FIGS. 19A and 19B) and the HEPG2 hepatocellular carcinoma xenograft model (FIGS. 19C and 19D) in NSG mice. These xenograft models have low macrophage infiltration, as evidenced by limited staining with the mouse macrophage marker F4/80 on IHC. q4dx4 using a dosage of 10 mg/kg. Antibodies Ab47, hB6H12.3 and Vel-IPV-hB6H12.3 effectively reduced tumor volume and/or slowed tumor growth during the study period.

In addition to assessing antitumor activity and its correlation with macrophage infiltration, MMP levels in these tumors were assessed, as well as quantification upon removal of the antibody shield moiety. Tumors from xenograft and syngeneic tumor models were selected and subjected to protein and RNA-seq assessments to monitor MMP levels in these tumors as well as to understand the correlation between RNA and protein levels in the TME. Protein analysis using the Luminex multiplex platform showed that all tumors used to study the antitumor activity of Vel-IPV-hB6H12.3 contained robust levels of both MMP2 and 9 (Table 22). Additionally, when we compared MMP levels in the tumor with those present in cell culture systems, we found a marked increase in MMP levels at the tumor site, which significantly exceeded the levels observed under in vitro tissue culture conditions alone (Figures 35A and 35B).

Table 22

MMP2 and MMP9 levels in selected tumors (pg/ml)

Tumors Xenograft Models MMP NT1080 HEPG2 L428 MMP2 7506 787 204 MMP9 2020 771 47 Syngeneic Models MMP NT 1080 HEPG2 L428 MMP2 8453 27765 47899 MMP9 28137 20845 22661

Anti-CD47 antibody activity in a tumor model with low internal macrophage content may be enhanced in combination with the ADC-austatin containing MMAE, which is known to drive macrophage infiltration. This was demonstrated in a xenograft HepG2 tumor model in NSG mice. Anti-CD47 antibodies were administered IP q4dx4 using a dosage of 5 mg/kg, while the ADC MMAE was administered once using a dosage of 1 mg/kg. The combination of antibodies was more effective in reducing tumor volume than either antibody alone (Figure 20). Additional experiments with other MMAE-containing auristatins (LIVIA and CD30) demonstrated similar compatibility with the anti-CD47 antibody in the MCSF7 breast cancer xenograft model for the Liv1A ADC and the L428 lymphoma model for the CD30 ADC (FIGS. 36A and 36B).

The mouse anti-CD47 reactive antibody mIAP301 (Oldenborg et al., J. Exp. Med. 193:855-861, 2001) can be screened using the same VEL and IPV sequences used for the human antibody hB6H12.3. Shielding using these constructs blocked antibody binding to murine CD47-positive tumors (Figure 21A) and prevented functionality as measured by erythrocyte phagocytosis (Figure 21B).

The anti-mouse anti-CD47 antibody mIAP301 causes a decrease in platelet counts in BALB/c mice when administered as a single 10 mg/kg IV dose. In contrast, such a decrease was not observed when mice were administered the shielded antibodies Vel-IPV-mIAP301 and Vel-M2-mIAP301 at a dose of 10 mg/kg intravenously (Fig. 22A).

The shielded antibody Vel-IPV-mIAP301 significantly improved pharmacokinetics in plasma of BALB/c mice compared with unshielded mIAP301, demonstrating that the shielded antibody is able to evade target-mediated drug targeting.

- 61 045361 typical antibodies against CD47 are encountered. Vel-IPV-mIAP301 and mIAP301 antibodies were labeled using ^3H -proprionate via lysine conjugation and administered intravenously to BALB/c mice at a dose of 1 mg/kg. Antibody concentrations were determined using scintillation measurements of plasma collected at different time points (Fig. 22B). Plasma concentrations of mIAP301 were below detectable amounts within 15 minutes, whereas Vel-IPV-mIAP301 concentrations were measurable up to 7 days post-dose.

The biodistribution of shielded Vel-IPV-mIAP301 and Vel-M2-mIAP301 and unshielded mIAP301 was tested in BALB/c mice carrying A20 using ³ H-labeled antibodies at doses of 1 and 10 mg/kg. Antibodies were administered as soon as the tumors reached 250 mm ³ . At designated time points, mice were sacrificed, and antibody concentrations in plasma, blood, tumor, spleen, and liver were determined using scintillation measurements. As shown in FIG. 37A and 37B, the plasma concentration of mIAP301 is negligible within one hour after administration. Meanwhile, significantly less shielded antibody was present in the spleen compared to unshielded mIAP301 antibody (FIGS. 37C and 37D). Similar results were obtained when studying antibodies in the liver. (Data not shown.) In contrast, by avoiding the target-mediated localization of CD47 in normal tissues, the shielded antibodies Vel-IPV-mIAP301 and Vel-M2-mIAP301 showed increased levels in the tumor compared to mIAP301 (Figures 37E and 37F).

The anti-mouse anti-CD47 antibody mIAP301 exhibited antitumor activity in the A20 lymphoma model, but caused a concomitant decrease in the number of red blood cells (Fig. 23A and Fig. 23B). The shielded antibody Vel-IPV-mIAP301 showed similar activity, but neutralized the effect on changes in red blood cell count. Vel-IPV-mIAP301 antibody avoided erythrocyte antigen uptake but also retained tumor binding (Figure 23C and Figure 23D). The A20 lymphoma model is described in more detail in Donnou et al. (Advances in Hematology, Article ID 701704, 2012) and Liu et al. (Nature Medicine, 21:1209-1215, 2015).

The anti-mouse anti-CD47 antibody mIAP301 exhibited antitumor activity in the MC38 colon cancer model, which is known to be sensitive to immuno-oncology agents. The activity of the mIAP301 screened antibody in this model demonstrated excellent efficacy in animals exhibiting a complete response (FIG. 24). Reinfection of this animal resulted in complete tumor rejection, demonstrating the induction of a long-lasting memory T cell response (data not shown). The MC38 colon cancer model is described in more detail in Liu et al. (quoted above).

In addition to testing the combinatorial activity of human anti-CD47 with ADCs versus Liv1a or CD30, we also assessed combinatorial activity with other immunomodulatory agents. To achieve this, we moved to the full mouse immune system and used the anti-mouse surrogate antibody mlAP against CD47. Using the A20 model, we demonstrated that the shielded anti-CD47 antibody acts synergistically with the surrogate anti-PD-1 antibody as well as the anti-SEA-CD40 antibody, SEA-1C10. These data suggest that the interaction of both the innate and adaptive arms of the immune system, together with a shielded CD47-targeting agent, can enhance the promotion of durable antitumor responses.

The parent and screened anti-mouse anti-CD47 antibody mIAP301 induced increased antitumor activity when combined with a surrogate anti-PD-1 antibody, resulting in 4/6 animals demonstrating CR (See, Dahan et al. Cancer Cell. 28: 285-295. 2015 , for reference to the PD-1 antibody, clone RMP1-14) (Fig. 25A). The parent anti-mouse anti-CD47 antibody mIAP301 induced enhanced antitumor activity when combined with the macrophage-activating CD40-targeting surrogate SEA-enhanced antibody 1C10 (see WO 2016/069919 for reference to antibody 1C10; see Lindberg et al. J. Biol., Chem. 269: 1567-1570. 1994, for reference to antibody mIAP301) (Fig. 25B).

Data demonstrating the effect of anti-CD47 antibody in combination with anti-PD1 antibody or anti-CD40 antibody indicate that anti-CD47 antibody therapy may enhance the activity of agents that enhance T cell activity, as in the context of anti-checkpoint inhibitor antibodies ( antibodies against PD-1) as well as agents that enhance the activity of innate cells (antibodies against CD40 or other CD40 inhibitors). This supports the notion that anti-CD47 antibody can be paired with multiple immunomodulatory agents in the clinic that support both adaptive and innate arms of the antitumor response.

Example 5. Supercoiled domain of shielded antibodies.

The stability of shielded humanized B6H12 antibodies bearing different coiled-coil domains was assessed using intravenous administration in BALB/c mice. Antibodies were administered using a dosage of 5 mg/kg. At a given time point (3 days), plasma was collected from infected mice. Human antibody was purified from plasma using IgSelect resin. Captured antibody was reconstituted and separated by SDS-PAGE and then probed by Western blotting using HRP-conjugated anti-human Fc antibody. The percentage of cleaved antibody was assessed by densitometry of bands corresponding to shielded and unshielded

- 62 045361 nym heavy chains, which differ in size by approximately 5 kDa (Fig. 26).

In table 23 shows the effects of shielding by various pairs of coiled-coil peptides included in the humanized antibody B6H12 containing a heavy chain (SEQ ID NO: 2) and a light chain (SEQ ID NO: 10) and tested in different cell lines. In the examples, the antibody is also called Ab47. The Kd (nM) is shown for each antibody that was generated from saturation in each respective cell line. A high concentration of 2000 nM was used for each binding experiment. As presented in table. 23, various coiled-coil domains were able to inhibit the binding of Ab47 to cell surface-expressed CD47, even when tested at concentrations greater than 2 micromolar, whereas the unshielded Ab47 antibody showed an IC ₅₀ of 3.3 -21 nM.

The stability and activation of screened humanized B6H12 antibodies carrying the Vel-IPV supercoil and cleavage sequence were assessed in nude mice bearing a human fibrosarcoma xenograft HT1080. Antibodies were administered intravenously using a dosage of 5 mg/kg. At specific time points (days 1, 3, 4), mice were sacrificed and tissue and plasma were collected. Tissues were homogenized and human antibody was purified from biological samples using IgSelect resin. Captured antibody was reconstituted and separated by DSNPAG and then probed by Western blotting using HRP-conjugated anti-human Fc antibody. The percentage of antibody cleaved was assessed by densitometry of bands corresponding to shielded and unshielded heavy chains, which differ in size by approximately kDa (FIGS. 38A and 38B). Very few unscreened antibodies (<5%) were detected in plasma or liver at any control time point. Meanwhile, more than 20-30% cleavage was detected in the tumors, with the maximum amount of cleavage occurring 3-4 days after dosing. In addition, HT1080 tumors collected from mice exposed to Ab47 or Vel-IPV-Ab47 for 4 or 7 days were subjected to flow cytometry to determine the extent of antibody that was able to bind to and saturate tumor cells that express CD47 (Fig. 38B).

Table 23

Affinity (nM), measured by compound saturation by flow cytometry, of humanized antibody B6H12 (Ab47) containing various antibody shielding domains

Superspiral NT1080 SW780 NSTI6 Raji AB47 21 9.6 7.8 3.3 A2B1 >2000 >2000 1775 CA2B1 >2000 >2000 >2000 Vel >2000 >2000 >2000 CVel >2000 >2000 >2000 >2000 MI >2000 >2000 >2000 mass media >2000 >2000 >2000 M15 >2000 >2000 >2000 mass media >2000 >2000 >2000 Fos-Jun >2000 >2000 >2000 CFos-Jun >2000 >2000 >2000 A4B4 1181 1684 1116 466 Hinge >2000 476 118 104

Activity and pharmacokinetics of humanized anti-CD47 antibodies and demonstration of improved tolerability by shielding in cynomolgus macaques.

To test the ability of the shielding moiety to improve the pharmacokinetics and tolerability of anti-CD47 IgG1 antibody variants, a series of studies were conducted using a single IV injection in cynomolgus macaques. The anti-CD47 IgG1 antibodies tested were cross-reactive with human and cynomolgus CD47, which are highly conserved in expression and sequence among these species. Evaluation of protease activity by in situ gel zymography on a panel of cynomolgus macaques and human tissues showed that levels of protease activity were also high.

- 63 045361 soko conservative for these species. In addition, cynomolgus macaques have very similar FcyR interactions with IgG1 antibodies and are considered toxicologically predictive of the effector function-related effects of human IgG1 antibodies, making them a suitable model for assessing the effects of IgG1 anti-CD47 antibodies (Warncke et al. J. Immunol. 188 :4405-4411. 2012). Taken together, cynomolgus macaque represents an appropriate species for assessing the differential activity of anti-CD47 antibodies and, furthermore, for assessing how this activity changes with shielding and modification of effector function.

Alternatively humanized B6H12, Ab47.

To determine the tolerability and PK of B6H12 on an alternatively humanized IgG1 construct and to further demonstrate evidence of the ability of the VelIPV shield fragment and cleavage sequence to alter tolerability and pharmacokinetics, cynomolgus monkeys were administered intravenously 0.1, 1, 10, and 30 mg/kg of the alternatively humanized anti-CD47 IgG1 antibody B6H12 ,Ab47. The humanized IgG1-based antibody Ab47 showed increased activity, demonstrated by loss of red blood cell mass in hematology assays, compared to published antibody data for mB6H12 on the humanized IgG4 platform (Liu, 2015). While IgG4 B6H12, which lacks the effector function to mediate high ADCC, ADCP, and CDC activity in addition to blocking interactions with CD47, has been shown to be tolerable at doses up to 30 mg/kg in cynomolgus macaques (Liu, 2015), Ab47 was not tolerated at doses greater than 1 mg/kg (Fig. 27A).

Ab47 versus Vel-IPV-Ab47.

To demonstrate the ability of the Vel-IPV shielding moiety to reduce Ab47 activity, cynomolgus monkeys were administered doses of 0.1, 1, and 10 mg/kg. Notably, doses approximately 10 times higher were better tolerated than uncoated Ab47, as evidenced by similar levels of red cell mass loss in blood samples collected for hematology analysis from animals exposed to 1.0 mg/kg Ab47 and 10 mg /kg Vel-IPV-Ab47 (Fig. 27B). Animals exposed to Ab47 showed persistent clinical signs associated with hemolysis and poor tolerance (red genitourinary discharge, hemolyzed specimens, vomiting and hypoactivity), while animals exposed to the screened antibody showed a complete absence of adverse clinical signs . This increase in tolerance with shielded Ab47 was also demonstrated in the ability of the shielding moiety to reduce the increase in plasma circulating cytokines such as monocyte chemoattractant protein-1, MCB-1. The highest dose of Vel-IPV-Ab47 tested (10 mg/kg) showed levels similar to the control and the lowest dose of Ab47 tested, 0.1 mg/kg (FIG. 39A). In addition, shielding significantly improved the PK of the molecule. When the dose of 1 mg/kg Ab47 was below the detection limit using the nonspecific total antibody ELISA on Study Day 3, 1 mg/kg Vel-IPV-Ab47 could be detected throughout the study course, Study Day 15 (Fig. 28). Nonspecific all antibody ELISA uses 96-well microtiter plates coated with anti-human kappa light chain mAb that binds to human kappa light chain Ab47 and Vel-IPV-Ab47. It does not cross-react with cynomolgus monkey kappa light chain. Test samples were diluted in the dynamic range of the assay for Ab47 (10 (LLOQ) to 1280 ng/ml (ULOQ)) or Vel-IPV-Ab47 (20 (LLOQ) to 2560 ng/ml (ULOQ)) with pooled K ₂ EDTA Monkey plasma Cynomolgus. Diluted samples along with QC and calibration markers were subjected to a minimum required dilution (MRD) of 1:20 with assay buffer before adding to blocked and washed plates. After incubation for 1 h at room temperature, the plates were washed and bound analyte (Ab47 or Vel-IPV-Ab47) was detected using a biotinylated anti-human kappa light chain mAb (identical to the capture reagent clone) followed by the addition of polymeric horseradish peroxidase , conjugated to streptavidin (poly-HRP-SA). After incubation and washing, the HRP substrate 3,3′,5,5′-tetramethylbenzidine (TMB) was added to the plates and stained for 10 minutes. The reaction was stopped with 1 N. HCl and plates were read on a Spectromax M5 plate reader at 450-630 nm. Net absorbance values were imported into Watson LIMS v. 7.4.2, and a nonlinear 5-PL regression was performed to convert the absorbance to ng/mL of all antibodies present in the samples.

Ab47 versus hB6H12.3.

Humanized B6H12.3, which is characterized by differential Kd, increased Bmax and reduced hemagglutination, was administered as a single intravenous bolus at a dose of 1 mg/kg to cynomolgus macaques to test for differential activity in vivo. While hB6H12.3 and Ab47 at 1 mg/kg resulted in similar levels of red blood cell reduction (FIG. 29), Ab47 also showed a reduction to approximately 40% of the pre-1 mg/kg platelet level, whereas hB6H12.3 caused only a 20% decrease in abundance, similar to control levels, likely due to sampling error (Fig. 30). An increase in platelet count starting on day 7 of the study was observed after treatment with both antibodies, likely

- 64 045361 but, due to general bone marrow stimulation.

hB6H12.3 versus Vel-IPV-hB6H12.

To demonstrate the ability of the Vel-IPV shielding fragment to reduce hB6H12.3 activity, cynomolgus macaques were administered doses of 10 and 20 mg/kg. 20 mg/kg, the highest dose tested, its hematological testing demonstrated a similar reduction in red blood cell mass loss as observed with 1 mg/kg hB6H12.3, however with greater tolerability than unshielded hB6H12.3 at a 20-fold lower dose (Fig. 31). Similarly, hB6H12.3 shielding also demonstrated a reduction in circulating plasma cytokines, including MCB-1, requiring a 20-fold dose of shielded hB6H12.3 to produce responses similar to those achieved with unshielded hB6H12.3 (Figure 39B) . Animals administered Vel-IPV-hB6H12.3 had no clinical symptoms associated with antibody exposure, whereas clinical symptoms associated with hemolysis were observed with unshielded hB6H12.3 at a dose of 1 mg/kg. Pharmacokinetic analysis using non-specific ELISA of all antibodies demonstrated an improved pharmacokinetic profile compared to unscreened hB6H12.3 (Fig. 39C).

Vel-IPV-hB6H12.3 versus SEA-Vel-IPV-hB6H12.3.

To further enhance the activity of the IgG1 shielded anti-CD47 antibody, Vel-IPVhB6H12.3, the antibody was generated using Seattle Genetics' patented SEA (sugar engineered antibodies) technology, a non-fucosylated antibody with enhanced effector function (USA). 8163551). The relative antitumor activity in xenograft models of Vel-IPV-hB6H12.3 and SEA-Vel-IPV-hB6H12.3, as well as fucosylated and non-fucosylated SEA-hB6H12.3, was assessed in models with high (Detroit562) and low (HT1080) macrophage content. Removing the backbone fucosylation and resulting in SEA-hB6H12.3 and SEA-Vel-IPV-hB6H12.3 did not provide evidence of superior antitumor activity in either model (FIGS. 40A and 40B). Because these models operate in nude mice, they may not take full advantage of the enhanced effector function conferred by the SEA backbone, the murine surrogate antibody mIAP301 is also nonfucosylated, and antitumor activity with fucosylated and nonfucosylated vel-IPV-mIAP301 was tested in a syngeneic model of A20 lymphoma with high activity, as well as in the syngeneic model of CT26 activity with lower activity. Removing the fucosylation of the backbone to produce the SEA antibody does not appear to provide additional antitumor effect.

While the SEA technology with other antibodies resulted in increased activity when the antibody was shielded, the SEA and non-SEA versions were well tolerated with similar red blood cell mass loss in the hematology assay at an identical maximum dose level tested, 20 mg/kg. (Fig. 32). No adverse clinical signs were observed during exposure to either antibody, however, circulating levels of the cytokine MCB-1 were increased for the SEA antibody (FIG. 41A). Pharmacokinetic analysis using a non-specific ELISA of all antibodies demonstrated generally similar pharmacokinetic profiles between the SEA and non-SEA Vel-IPVhB6H12.3 antibodies, with both antibodies improving over unscreened hB6H12.3 (Figure 41B).

Example 6 Treatment of Tumors in Subjects Using Selective Biomarkers.

The disclosure further provides for the treatment of subjects with tumors using the anti-CD47 antibodies of the present invention based on selective biomarkers in said subjects. The selection of subjects for treatment using anti-CD47 antibodies of this invention will be based on 1) higher levels and activity of MMPs in cancerous tissue compared to surrounding non-cancerous tissue; 2) higher levels of CD47 expression in cancerous tissue compared to surrounding noncancerous tissue; and 3) higher levels of macrophage infiltration in cancerous tissue compared to surrounding noncancerous tissue.

Methods for determining the levels and activity of individual biomarkers include immunohistochemical and enzymological analyses. For example, levels of macrophage infiltration can be determined by immunohistochemistry using anti-CD163 antibody as a marker.

To demonstrate the ability to detect tumor-infiltrating macrophages in a cancer tissue sample compared to a non-cancerous tissue sample, breast cancer tumor core samples and normal breast tissue samples were used. Immunohistochemistry was performed using anti-CD163 antibody. The results demonstrate that tumor-infiltrating macrophages can be easily detected in tumor samples compared to non-cancerous tissue samples (FIG. 33).

Example 7: Phagocytosis and hemagglutination of CD47-positive cells by Vel-IPVhB6H12.3.

Phagocytosis of CD47-positive erythrocytes was monitored to assess the effect of shielding on antibody functionality. Human erythrocytes were labeled with the fluorescent red dye PKH and opsonized for 30 min with 1 μg/ml unshielded hB6H12.3, shielded Vel-IPV-hB6H12.3, or MMP-activated Vel-IPV-hB6H12.3. The red blood cells were washed and then incubated with monocytic macrophages at a ratio of 10:1 for two hours. Then samples three

- 65 045361 washed every day with hypotonic lysis buffer ACK. The extent of phagocytosis was assessed by monitoring the uptake of fluorescently labeled human erythrocytes by human macrophages via flow cytometry.

As shown in FIG. 42, shielded Vel-IPV-hB6H12.3 did not exhibit an increase in phagocytosis above background levels of unexposed RBCs, whereas hB6H12.3 and MMP-activated Vel-IPV-hB6H12.3 showed similar levels of RBC phagocytosis.

Activation of hemagglutination of human erythrocytes by shielded Ab47 antibodies was tested as described in Example 2. Human erythrocytes were exposed to increased concentrations of shielded Vel-IPV-Ab47 or MMP-cleaved Vel-IPV-Ab47 for thirty minutes at 37°C. Hemagglutination was monitored by optically assessing the diameter of the visible spot in each well.

As shown in FIG. 43, hemagglutination was inhibited when Ab47 was shielded, and hemagglutination was restored when the shielding moiety was removed by MMP.

Example 8: hB6H12.3 directly induces apoptosis.

Ligation of CD47 on the cell surface can induce apoptosis. It has been reported that the anti-CD47 mAb clone B6H12 can only induce apoptosis when immobilized on a surface; however, hB6H12.3 demonstrated apoptotic activity without being surface bound. Eight different cell lines (A431, HEK293, Hela, HepG2, HPAF11, I540Cy, MCF7 and THP1) were seeded at 50,000 cells per well for 24 hours. Cells were exposed to hB6H12.3, 5F9 or IgG1 isotype control at concentrations ranging from 5 to 0. 05 μg/ml for 18 hours. Cells were then harvested, washed twice, stained with the apoptosis marker Annexin V, and apoptosis was quantified by flow cytometry.

As shown in FIG. 44, in all eight cell types, cells that were exposed to hB6H12.3 exhibited higher levels of apoptosis, as measured by associated levels of annexin staining, than cells of the same type treated with 5F9 or the IgG1 isotype control.

Example 9 Stability of hB6H12.3 in whole blood and plasma.

Freshly collected whole blood samples (4% sodium citrate) from 17 patients with various types of cancer (10 sarcoma, 3 NSCLC, 3 colon cancer, and 1 melanoma) were obtained from BioIVT. hB6H12.3 and Vel-IPV-hB6H12.3 screened antibodies were directly labeled with fluorescein isothiocyanate (FITC). Fresh whole blood samples were incubated with increasing concentrations (maximum concentration, 20 μg/ml) of FITC-labeled hB6H12.3 or FITC-labeled Vel-IPVhB6H12.3 for 20 h at 37°C. Antibody binding to blood cells was characterized using flow cytometry.

A portion of each of the whole blood samples from sarcoma patients was used for plasma extraction, and 20 μg/ml recombinant CD47 was added to the plasma samples. Plasma samples containing recombinant CD47 were then incubated with 20 μg/ml hB6H12.3 or screened Vel-IPVhB6H12.3 for four days at 37°C. The degree of cleavage of the Vel-IPV shielding fragment was assessed using a CD47 binding ELISA as described in Example 2. Vel-IPVhB6H12.3 did not bind to recombinant CD47 at the plasma concentration (20 μg/ml); therefore, any antibody-CD47 binding detected in plasma incubated with Vel-IPVhB6H12.3 is due to the binding of cleaved Vel-IPV-hB6H12.3 to CD47.

Of the 17 blood samples tested, shielded Vel-IPV-hB6H12.3 showed greater than 10% binding to only one sample with an abnormal value (from a sarcoma patient, see FIG. 45B) at the highest concentration. None of the 16 other samples showed binding of shielded Vel-IPV-hB6H12.3 (representative data shown in Fig. 45A). As shown in FIG. 45C, no more than 2% of screened Vel-IPV-hB6H12.3 was cleaved in any of ten plasma samples from sarcoma patients.

Example 10. Cytokine production in response to IV6H12.3.

Fresh whole blood samples from cancer patients (10 sarcoma, 3 NSCLC, 3 colon cancer, and 1 melanoma) were incubated with increasing concentrations (maximum concentration, 20 μg/ml) of FITC-labeled hB6H12.3, or FITC -labeled Vel-IPVhB6H12.3 or with 0.1 μg/ml LPS for 20 h at 37°C. Cytokine levels were assessed using a 38-plex magnetic bead-based cytokine and chemokine panel.

In the majority of patient samples tested, moderate cytokine production was induced by hB6H12.3, and minimal cytokine production was induced by Vel-IPV-hB6H12.3. The cytokines IP10, IL-Ra, MIP-1a, and MIP-1a were most frequently induced by hB6H12.3. Levels of IL1-Ra (Fig. 46B), MIP-1a and MIP-1e were below 200 pg/ml at the highest concentration of hB6H12.3 tested, while IP-10 levels reached 4000-5000 ng/ml (Fig. 46A). Levels of cytokines produced by Vel-IPV-hB6H12.3 were lower than those produced by hB6H12.3 in all cases and were typically 100- to 1000-fold lower.

Example 11 hB6H12.3 induces apoptosis in vivo.

Athymic mice bearing human fibrosarcoma xenograft tumors HT1080, i.v.

- 66 045361 administered an i.p. dose of 5 mg/kg hB6H12.3, Vel-IPV-hB6H12.3, or isotype control hlgG1 when tumors reached 200 mm ³ . At specific time points (after 24 and 96 hours), mice were sacrificed and tumors were collected. Tumors were homogenized and human fibrosarcoma xenograft HT1080 tumor cells were resuspended at 1 million cells/ml in 1X Annexin V staining buffer (10X staining buffer containing 50 mM HEPES, 700 mM NaCl, 12.5 mM CaCl ₂ , pH 7 ,4, with a dilution of 1:10 in water). Cells were transferred to a round-bottom 96-well plate (100 μl/well) and 5 μl of Annexin V-FITC staining reagent and 1 μl of 100 μg/ml Live/Dead UV staining solution were added to each well. Cells were stained for 30 min at room temperature. Samples were centrifuged at 1550 g for 5 min, the supernatant was discarded, and the cells were washed 3 times with chilled 1X Annexin V staining buffer. Cells were resuspended in 100 μl of 1X Annexin V staining buffer. Apoptosis was assessed by flow cytometry on an LSRII cytometer as the percentage of cells positive for binding of annexin V to surface phosphatidylserine. Cells that stained positively with Live/Dead dye were excluded from the analysis.

As shown in FIG. 47, tumors exposed to both hB6H12.3 and Vel-IPVhB6H12.3 showed increased annexation of V+ apoptotic cells at 96 hours post-exposure compared to tumor samples that were unexposed or exposed to the isotype control.

Some non-limiting embodiments of this invention

Embodiment 1. A humanized antibody or antigen binding fragment thereof that specifically binds human CD47, an antibody or antigen binding fragment comprising a light chain variable region and a heavy chain variable region, wherein said heavy chain variable region comprises:

CDRs represented by SEQ ID NOs: 16 (GYGMS), 17 (TITSGGTYTYYPDSVKG), and 18 (SLAGNAMDY); and the human IGHV3-23/HJ4 framework sequence shown in SEQ ID NO: 88, wherein the framework sequence positions H44, H49, H82, H89, H91 and H94, according to Kabat numbering, represent donor residues.

Embodiment 2. A humanized antibody or antigen binding fragment thereof that specifically binds human CD47, an antibody or antigen binding fragment comprising a light chain variable region and a heavy chain variable region, wherein said heavy chain variable region comprises:

CDRs represented by SEQ ID NOs: 16 (GYGMS), 17 (TITSGGTYTYYPDSVKG), and 18 (SLAGNAMDY); and the human IGHV3-48/HJ4 framework sequence shown in SEQ ID NO: 89, wherein the H49 framework sequence position, according to Kabat numbering, represents a donor residue.

Embodiment 3. A humanized antibody or antigen binding fragment thereof that specifically binds human CD47, an antibody or antigen binding fragment comprising a light chain variable region and a heavy chain variable region, wherein said heavy chain variable region comprises:

CDRs represented by SEQ ID NOs: 16 (GYGMS), 17 (TITSGGTYTYYPDSVKG), and 18 (SLAGNAMDY); and the human IGHV3-66/HJ4 framework sequence shown in SEQ ID NO: 90, wherein the framework sequence positions H29, H49, and H82, according to Kabat numbering, represent donor residues; or

Embodiment 4 A humanized antibody or antigen binding fragment thereof that specifically binds human CD47, an antibody or antigen binding fragment comprising a light chain variable region and a heavy chain variable region, wherein said heavy chain variable region comprises:

CDRs represented by SEQ ID NOs: 16 (GYGMS), 17 (TITSGGTYTYYPDSVKG), and 18 (SLAGNAMDY); and the human IGHV3-74/HJ4 framework sequence shown in SEQ ID NO: 91, wherein the framework sequence position H49, according to Kabat numbering, represents the donor residue.

Embodiment 5. A humanized antibody or antigen-binding fragment thereof according to any one of embodiments 1-4 of the present invention, wherein said light chain variable region comprises:

CDRs represented by SEQ ID NOs: 31 (RASQTISDYLH), 32 (FASQSIS), and 33 (QNGHGFPRT); and the human IGKV6-21/KJ2 framework sequence shown in SEQ ID NO: 92, wherein the framework sequence positions L4, L21 L69 and L85, according to Kabat numbering, represent donor residues.

Embodiment 6. Humanized antibody or antigen-binding fragment thereof

- 67 045361 to one of embodiments 1-4 of this invention, wherein said light chain variable region contains:

CDRs represented by SEQ ID NOs: 31 (RASQTISDYLH), 32 (FASQSIS), and 33 (QNGHGFPRT); and the human IGKV1-27/KJ2 framework sequence shown in SEQ ID NO: 93, wherein framework sequence positions L21, L49 and L69, according to Kabat numbering, represent donor residues.

Embodiment 7. An antibody or antigen-binding fragment thereof according to any of embodiments 1-4 of this invention, wherein H29 is occupied by F, H44 is occupied by R or G, H49 is occupied by A, H82 is occupied by M or I, H89 is occupied by I or V, H91 is occupied F or Y and H94 is occupied by R, in accordance with the numbering according to Kabat.

Embodiment 8. The antibody or antigen binding fragment of Embodiment 5, wherein L4 is occupied by M, L21 is occupied by L, L49 is occupied by K, L69 is occupied by T or S, and L85 is occupied by V or T, according to Kabat numbering.

Embodiment 9. An antibody or antigen binding fragment according to any one of embodiments 1-4 of the present invention, comprising a heavy chain variable region (HCVR) having at least 90% sequence identity to any of SEQ ID NO: 2, 3, 4 , 5, 6, 7 and 8, and a light chain variable region (LCVR) having at least 90% sequence identity to any of SEQ ID NOs: 10, 11, 12, 13, 14 and 15.

Embodiment 10 The antibody or antigen binding fragment of Embodiment 5, further comprising the G91A mutation in LCDR3 according to Kabat numbering.

Embodiment 11. The antibody or antigen binding fragment of any one of embodiments 1-4 of the present invention, wherein the antibody or antigen binding fragment thereof is an IgG1 isotype.

Embodiment 12. An antibody or antigen-binding fragment according to any one of embodiments 1-4 of the present invention, wherein the antibody or antigen-binding fragment thereof has increased antibody-dependent cellular cytotoxicity (ADCC) compared to its parent antibody.

Embodiment 13. The antibody or antigen-binding fragment of any one of embodiments 1-4 of the present invention, wherein the antibody or antigen-binding fragment thereof exhibits increased antibody-dependent cellular phagocytosis (ADCP) compared to its parent antibody.

Embodiment 14. The antibody or antigen binding fragment of any one of Embodiments 1-4 of the present invention having reduced backbone fucosylation compared to its parent antibody.

Embodiment 15. An antibody or antigen-binding fragment according to any one of embodiments 1-4 of the present invention, wherein the antibody or antigen-binding fragment thereof blocks the interaction between CD47 and SIRPa.

Embodiment 16. An antibody or antigen-binding fragment according to any one of embodiments 1-4 of the present invention, wherein the antibody or antigen-binding fragment has reduced hemagglutination of red blood cells compared to the parent antibody.

Embodiment 17. A nucleic acid sequence encoding an antibody or antigen binding fragment according to any of embodiments 1-4 of the present invention.

Embodiment 18. The antigen binding fragment of Embodiment 1 of the present invention, comprising a Fab, Fab', F(ab') ₂ , Fv fragment, diabody, single chain antibody, scFv fragment, or scFv-Fc.

Embodiment 19. A method of treating CD47 cancer in a subject, comprising administering to the subject a therapeutically effective amount of an anti-CD47 antibody or antigen-binding fragment thereof containing a screening agent, wherein the screening agent comprises one or more supercoiled peptides that reduce the binding affinity of the antibody or antigen binding fragment with human CD47 compared to an antibody or antigen binding fragment thereof without a screening agent.

Embodiment 20. The method of Embodiment 19, wherein a protease-cleavable linker attaches the screening agent to the antibody or antigen-binding fragment thereof.

Embodiment 21. The method of Embodiment 20, wherein the protease-cleavable linker has an amino acid sequence comprising IPVSLRSG (SEQ ID NO: 73) or GPLGVR (SEQ ID NO: 57).

Embodiment 22. The method of Embodiment 20, wherein the protease-cleavable linker comprises a matrix metalloprotease (MMP) cleavage site.

Embodiment 23. The method of Embodiment 22, wherein the MMP cleavage site is selected from the group consisting of an MMP2 cleavage site, a cleavage site

- 68 045361

MMP7, MMP9 cleavage site and MMP13 cleavage site.

Embodiment 24. The method of Embodiment 22, wherein the shielding agent is released from the anti-CD47 antibody or antigen binding fragment thereof upon cleavage of the MMP cleavage site in the tumor microenvironment by the MMP.

Embodiment 25. The method of Embodiment 24, wherein the cleaved anti-CD47 antibody has a moiety of amino acid residues of an MMP cleavage site.

Embodiment 26. The method of Embodiment 25, wherein the cleavage site amino acid residue fragment contains an LRSG, SG, or VR sequence at the N-terminus of the antibody.

Embodiment 27. The method of Embodiment 19, wherein the one or more coiled-coil peptides comprise one or more sequences selected from the group consisting of SEQ ID NOs: 44, 45, 46, 47, 48, 49, 50, 51 , 52, 53, 54, 55 and 56.

Embodiment 28. The method of Embodiment 19, wherein binding of the antibody or antigen binding fragment to CD47 is reduced by at least about 100-fold compared to the antibody or antigen binding fragment thereof without a screening agent.

Embodiment 29. The method of Embodiment 19, wherein binding of the antibody or antigen binding fragment to CD47 is reduced from about 200-fold to about 1500-fold compared to the antibody or antigen binding fragment thereof without a screening agent.

Embodiment 30. The method of Embodiment 19, wherein the cancer characterized by the expression of CD47 is a hematologic malignancy that causes solid cancer.

Embodiment 31. The method of Embodiment 30 of the present invention, wherein the hematologic malignancy is selected from the group consisting of non-Hodgkin's lymphoma, B-lymphoblastic lymphoma; Cellular chronic lymphocytic leukemia/small lymphocyte lymphoma, Richter's syndrome, follicular lymphoma, multiple myeloma, myelofibrosis, polycythemia vera, cutaneous T-cell lymphoma, monoclonal gammopathy of unknown etiology (MGUS), myelodysplastic syndrome (MDS), immunoblastic large cell lymphoma, B-lymphoblast leukemia from progenitor cells, acute myeloid leukemia (AML) and anaplastic large cell lymphoma.

Embodiment 32. The method of Embodiment 19, wherein the cancer characterized by CD47 expression is a solid cancer.

Embodiment 33. The method of Embodiment 32, wherein the solid cancer is selected from the group consisting of lung cancer, pancreatic cancer, breast cancer, liver cancer, ovarian cancer, testicular cancer, kidney cancer, bladder cancer, cancer spine, brain cancer, cervical cancer, endometrial cancer, colorectal cancer, anal cancer, esophageal cancer, gallbladder cancer, gastrointestinal cancer, stomach cancer, sarcoma, head and neck cancer, melanoma, skin cancer, prostate cancer , pituitary cancer, uterine cancer, vaginal cancer and thyroid cancer.

Embodiment 34. The method of Embodiment 32, wherein the solid cancer is selected from the group consisting of lung cancer, soft tissue sarcoma, colorectal cancer, head and neck cancer, and breast cancer.

Embodiment 35. The method of Embodiment 19, wherein the subject is a human with solid cancer.

Embodiment 36. The method of Embodiment 19 of the present invention, wherein the anti-CD47 antibody is administered in combination with an immune checkpoint molecule inhibitor selected from one or more of programmed cell death protein 1 (PD-1), programmed cell death ligand 1 ( PD-L1), PD-L2, cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), T-cell immunoglobulin and mucin domain-containing protein 3 (TIM-3), lymphocyte activation gene 3 (LAG-3), carcinoembryonic antigen-dependent cell adhesion molecule 1 (CEACAM-1), CEACAM-5, V-domain immunoglobulin suppressor of T-cell activation (VISTA), B- and T-lymphocyte attenuator (BTLA), T-cell immunoreceptor with Ig and ITIM domains ( TIGIT), leukocyte-associated immunoglobulin-like receptor (type 1) (LAIR1), CD160, 2B4 or TGFR.

Embodiment 37. An antibody or antigen-binding fragment thereof that specifically binds to the human CD47 protein, comprising a screening agent, wherein the screening agent comprises one or more coiled-coil peptides containing the sequence

SEQ ID NO: 95 (QGASTTVAQLEEKVKTLRAENYELKSEVQRLEEQVAQLGS) and/or

SEQ ID NO: 94 (QGASTSVDELQAEVDQLEDENYALKTKVAQLRKKVEKLGS), and wherein one or more coiled-coil peptides reduce the binding affinity of the antibody or antigen binding fragment thereof to human CD47 protein compared to the antibody or antigen binding fragment thereof without a screening agent.

Embodiment 38: The antibody or antigen-binding fragment of embodiment is given

- 69 045361 of invention 37, comprising a heavy chain variable region (HCVR) having at least 90% sequence identity to any of SEQ ID NOs: 2, 3, 4, 5, 6, 7 and 8, and a variable region light chain (LCVR) having at least 90% sequence identity to any of SEQ ID NOs: 10, 11, 12, 13, 14 and 15.

Embodiment 39. The antibody or antigen binding fragment of Embodiment 37, wherein the screening agent is attached to the antibody or antigen binding fragment via a protease cleavable linker.

Embodiment 40. The antibody or antigen binding fragment of Embodiment 39, wherein the protease-cleavable linker has an amino acid sequence comprising IPVSLRSG (SEQ ID NO: 73) or GPLGVR (SEQ ID NO: 57).

Embodiment 41. The antibody or antigen binding fragment of Embodiment 39, wherein the protease-cleavable linker comprises a matrix metalloprotease (MMP) cleavage site.

Embodiment 42. The antibody or antigen binding fragment of Embodiment 41, wherein the MMP cleavage site is selected from the group consisting of an MMP2 cleavage site, an MMP7 cleavage site, an MMP9 cleavage site, and an MMP3 cleavage site.

Embodiment 43. The antibody of Embodiment 37, wherein the shielding agent is removed from the anti-CD47 antibody upon cleavage of the MMP cleavage site by MMP.

Embodiment 44. The antibody of Embodiment 43, wherein the anti-CD47 antibody has a fragment of MMP cleavage site amino acid residues after cleavage of the MMP cleavage site by MMP.

Embodiment 45. The antibody of Embodiment 44, wherein the cleavage site amino acid residue fragment contains an LRSG, SG, or VR sequence at the N-terminus of the antibody.

Embodiment 46. The antibody or antigen-binding fragment of Embodiment 37, wherein binding is reduced by at least about 100-fold compared to an antibody or antigen-binding fragment thereof without a screening agent.

Embodiment 47. The antibody or antigen-binding fragment of Embodiment 37, wherein binding is reduced from about 200-fold to about 1500-fold compared to an antibody or antigen-binding fragment thereof without a screening agent.

Embodiment 48. The antibody or antigen binding fragment of Embodiment 37, comprising the heavy chain sequence of SEQ ID NO: 42 and the light chain sequence of SEQ ID NO: 43.

Embodiment 49. The antibody or antigen binding fragment of any one of embodiments 37-48 of the present invention, comprising a variant Fc region that confers enhanced effector function selected from ADCC and/or CDC activity.

Embodiment 50. An antibody or antigen binding fragment of Embodiment 49 that is afucosylated.

Embodiment 51. A humanized antibody or antigen binding fragment thereof that specifically binds human CD47, wherein the antibody is an IgG1 isotype.

Embodiment 52. The antibody of Embodiment 51 of the present invention, comprising increased ADCC, increased ADCC, and/or increased ADCC activity.

Embodiment 53. A method of treating cancer characterized by CD47 expression in a subject, comprising the steps of:

a) identifying the subject as having elevated levels of MMPs in the tumor compared to surrounding non-cancerous tissue; And

b) administering to a subject a therapeutically effective amount of an anti-CD47 antibody or antigen-binding fragment thereof containing a shielding agent, wherein the shielding agent contains coiled-coil peptides that reduce the binding affinity of the antibody or antigen-binding fragment thereof to human CD47 compared to the antibody or antigen-binding fragment thereof without the shielding agent, if the subject has elevated levels of MMPs in the tumor compared to surrounding non-cancerous tissue.

Embodiment 54. The method of embodiment 53 of the present invention, wherein the MMP is selected from the group consisting of: MMP2, MMP7, MMP9 and MMP13.

Embodiment 55. The method according to embodiment 53 of the present invention, wherein step a) includes:

i) obtaining cancerous and non-cancerous tissue from the subject;

ii) detection of MMPs in isolated cancer tissue and in non-cancerous tissue; and iii) comparison of quantification of staining in cancerous versus noncancerous tissue.

Implementation option 56. Method for treating an oncological disease characterized by ex

- 70 045361 CD47 compression, in a subject, including the stages:

a) identifying the subject as having elevated levels of CD47 in the tumor compared to surrounding non-cancerous tissue; And

b) administering to a subject a therapeutically effective amount of an anti-CD47 antibody or antigen-binding fragment thereof containing a shielding agent, wherein the shielding agent contains coiled-coil peptides that reduce the binding affinity of the antibody or antigen-binding fragment thereof to human CD47 compared to the antibody or antigen-binding fragment thereof without the shielding agent, if the subject has elevated levels of CD47 in the tumor compared to surrounding non-cancerous tissue.

Embodiment 57. The method according to embodiment 56 of the present invention, wherein step a) includes:

i) obtaining cancerous tissue and surrounding non-cancerous tissue from the subject;

ii) detection of CD47 in isolated cancer tissue and surrounding non-cancerous tissue; and iii) comparison of quantification of CD47 staining in cancerous tissue versus CD47 staining in noncancerous tissue.

Embodiment 58. A method of treating cancer characterized by CD47 expression in a subject, comprising the steps of:

a) identifying an entity with increased levels of macrophage infiltration in cancerous tissue compared to non-cancerous tissue; And

b) administering to a subject a therapeutically effective amount of an anti-CD47 antibody or antigen-binding fragment thereof containing a screening agent, wherein the screening agent comprises one or more coiled-coil peptides that reduces the binding affinity of the antibody or antigen-binding fragment thereof to human CD47 compared to the antibody or antigen-binding fragment thereof without screening agent if the subject has increased levels of macrophage infiltration into the tumor compared to non-cancerous tissue.

Embodiment 59. The method according to embodiment 58 of the present invention, wherein step a) includes:

i) obtaining cancerous tissue and surrounding non-cancerous tissue from the subject;

ii) detection of macrophages in isolated cancerous tissue and in non-cancerous tissue; and iii) comparison of quantification of staining in cancerous versus noncancerous tissue.

Embodiment 60. The method of Embodiment 58, wherein staining of macrophages is performed using an anti-CD163 antibody.

Embodiment 61. A humanized antibody or antigen binding fragment thereof that specifically binds human CD47 comprising a heavy chain variable region (HCVR) having at least 90% sequence identity to any of SEQ ID NO: 2, 3, 4, 5 , 6, 7, and 8, and a light chain variable region (LCVR) having at least 90% sequence identity to any of SEQ ID NOs: 10, 11, 12, 13, 14, and 15, wherein the antibody further comprises the sequence LRSG, SG or VR at the N-terminus of HCVR and/or LCVR.

Embodiment 62. A method of treating cancer by administering a combination of a shielded anti-CD47 antibody according to embodiment 37 of the present invention with an agonistic anti-CD40 antibody.

Embodiment 63. The method of Embodiment 62, wherein the CD40 agonist antibody has low levels of fucosylation, for example, the SEA-CD40 antibody.

Embodiment 64. A method of treating cancer by administering a combination of a shielded anti-CD47 antibody of Embodiment 37 with an antibody-drug conjugate (ADC), wherein the ADC antibody specifically binds to a protein that is expressed on the extracellular surface of the cancer cell and the antibody is conjugated with a linker-drug compound containing a cytotoxic agent.

Embodiment 65. The method of Embodiment 64, wherein the cytotoxic agent is auristatin.

The method of embodiment 64, wherein the ADC antibody is conjugated to a linker-drug compound selected from the group consisting of vcMMAE and mcMMAF.

Claims (68)

1. A humanized antibody or antigen-binding fragment thereof that specifically binds to human CD47, wherein the antibody or antigen-binding fragment thereof comprises a light chain variable region and a heavy chain variable region, wherein the heavy chain variable region comprises the amino acid sequence of SEQ ID NO: 3 and a light chain variable region chain contains the amino acid sequence SEQ ID NO: 13. - 71 045361 2. The antibody or antigen-binding fragment thereof according to claim 1, wherein the antibody or antigen-binding fragment thereof is an IgG1 isotype. 3. The antibody or antigen-binding fragment thereof as claimed in claim 1 or 2, wherein the antibody or antigen-binding fragment thereof has (i) increased antibody-dependent cellular cytotoxicity (ADCC) compared to the parent antibody, and/or (ii) increased antibody-dependent cellular phagocytosis ( ADCP) compared to the parent antibody, and/or (iii) increased complement-dependent cytotoxicity (CDC) compared to their parent antibody. 4. An antibody or antigen-binding fragment thereof according to any one of claims 1 to 3, wherein the antibody or antigen-binding fragment thereof is a Fab, Fab', F(ab') ₂ , Fv fragment, diabody, single chain antibody, scFv fragment or scFv-Fc . 5. An antibody or an antigen-binding fragment thereof according to any one of claims 1 to 4, wherein the antibody or an antigen-binding fragment thereof induces apoptosis of cells expressing CD47 in vitro and/or in vivo. 6. An antibody or antigen-binding fragment thereof according to any one of claims 1 to 5, having reduced backbone fucosylation compared to the parent antibody or which is afucosylated. 7. An antibody or an antigen-binding fragment thereof according to any one of claims 1 to 6, wherein the antibody or an antigen-binding fragment thereof blocks the interaction between CD47 and SIRPa. 8. The antibody or antigen-binding fragment thereof according to any one of claims 1 to 7, wherein the antibody or antigen-binding fragment thereof has reduced hemagglutination of erythrocytes compared to Ab47 containing the heavy chain variable region of SEQ ID NO: 2 and the light chain variable region of SEQ ID NO: 10. 9. A nucleic acid sequence encoding an antibody or an antigen-binding fragment thereof according to any one of claims 1 to 8. 10. An expression vector containing the nucleic acid according to claim 9. 11. A host cell containing the nucleic acid of claim 9 or the expression vector of claim 10. 12. A host cell that expresses the antibody or antigen-binding fragment thereof according to any one of claims 1 to 8. 13. A method for producing an antibody or an antigen-binding fragment thereof according to any one of claims 1 to 8, including culturing a host cell according to claim 11 or 12. 14. The method according to claim 13, further comprising isolating the antibody or an antigen-binding fragment thereof. 15. Composition for inducing apoptosis of a cell expressing CD47, containing an antibody or an antigen-binding fragment according to any one of claims 1 to 8. 16. The composition according to claim 15, in which the cells are in vitro. 17. The composition according to claim 15, in which the cells are in vivo. 18. Shielded antibody containing an antibody or an antigen-binding fragment according to any one of paragraphs. 1-8 and at least one shielding domain. 19. The shielded antibody of claim 18, wherein the at least one shielding domain comprises an amino acid sequence selected from SEQ ID NOs: 44-55, 75-86, 94 and 95. 20. The shielded antibody of claim 18 or 19, wherein the at least one shielding domain reduces the binding affinity of the antibody or antigen binding fragment to human CD47 protein compared to an antibody or antigen binding fragment thereof without at least one shielding domain. 21. The shielded antibody of claim 20, wherein the binding affinity is reduced by at least about 100-fold or from about 200-fold to about 1500-fold compared to an antibody or antigen binding fragment thereof without at least one shielding domain. 22. The shielded antibody according to any one of claims 18 to 21, wherein the antibody or antigen-binding fragment thereof comprises a heavy chain and a light chain, wherein the heavy chain is linked to a first shielding domain; or where the light chain is associated with a second shielding domain; or wherein the heavy chain is linked to a first shielding domain and the light chain is linked to a second shielding domain. 23. The shielded antibody of claim 22, wherein the first shielding domain comprises an amino acid sequence selected from SEQ ID NO: 44, 46, 48, 50, 52, 54, 75, 77, 79, 81, 83, 85 and 94; and the second shielding domain comprises an amino acid sequence selected from SEQ ID NOs: 45, 47, 49, 51, 53, 55, 76, 78, 80, 82, 84, 86 and 95. 24. The shielded antibody of claim 23, wherein the first shielding domain and the second shielding domain are a pair of shielding domains selected from SEQ ID NOs: 44 and 45; SEQ ID NO: 46 and 47; SEQ ID NO: 48 and 49; SEQ ID NO: 50 and 51; SEQ ID NO: 52 and 53; SEQ ID NO: 54 and 55; SEQ ID NO: - 72 045361 75 and 76; SEQ ID NO: 77 and 78; SEQ ID NO: 79 and 80; SEQ ID NO: 81 and 82; SEQ ID NO: 83 and 84; SEQ ID NO: 85 and 86; and SEQ ID NOs: 94 and 95. 25. The shielded antibody of any one of claims 22 to 24, wherein the first shielding domain is linked to the N-terminus of the heavy chain and the second shielding domain is linked to the N-terminus of the light chain. 26. The shielded antibody according to any one of claims 22 to 25, wherein each shield domain contains a protease-cleavable linker and is linked to the heavy chain or light chain through the protease-cleavable linker. 27. The shielded antibody of claim 26, wherein the protease-cleavable linker comprises a matrix metalloprotease (MMP) cleavage site. 28. The shielded antibody of claim 27, wherein the MMP cleavage site is selected from the MMP2 cleavage site, the MMP7 cleavage site, the MMP9 cleavage site, and the MMP13 cleavage site. 29. The shielded antibody according to claim 27 or 28, where after cleavage by MMP, the heavy chain and/or light chain of the antibody or antigen-binding fragment thereof comprises a fragment of amino acid residues of the MMP cleavage site. 30. The shielded antibody of claim 29, wherein the cleavage site amino acid residue fragment contains the sequence LRSG, SG or VR at the N-terminus of the antibody. 31. A shielded antibody according to any one of claims 18 to 30, comprising a heavy chain associated with a first shielding domain and having the amino acid sequence of SEQ ID NO: 42, and a light chain associated with a second shielding domain and having the amino acid sequence of SEQ ID NO: 43 . 32. Nucleic acid encoding the shielded antibody according to any one of claims 18-31. 33. An expression vector containing the nucleic acid according to claim 32. 34. A host cell containing the nucleic acid of claim 32 or the expression vector of claim 33. 35. A host cell that expresses the shielded antibody according to any one of claims. 18-31. 36. A method for producing a screened antibody according to any one of claims 18 to 31, including culturing a host cell according to claim 34 or 35. 37. The method of claim 36, further comprising isolating the screened antibody. 38. A composition for the treatment of CD47-expressing cancer in a patient, containing an anti-CD47 antibody or an antigen-binding fragment according to any one of claims 1 to 8 or a shielded antibody according to any of claims 18 to 31. 39. A composition for treating CD40-expressing cancer in a patient, comprising the screened antibody of any one of claims 18 to 31, wherein the patient is identified as having elevated levels of MMPs in the tumor compared to surrounding non-cancerous tissue; and wherein each shielding domain of the shielded antibody comprises a protease-cleavable linker, wherein the protease-cleavable linker comprises a matrix metalloprotease (MMP) cleavage site. 40. The composition according to claim 39, where the MMP is selected from MMP2, MMP7, MMP9 and MMP13. 41. The composition according to claim 39 or 40, wherein the MMP cleavage site is selected from the MMP2 cleavage site, the MMP7 cleavage site, the MMP9 cleavage site and the MMP13 cleavage site. 42. The composition according to any one of claims 39-41, where the identification includes i) taking cancerous and non-cancerous tissue from the patient; ii) detection of MMPs in isolated cancer tissue and in non-cancerous tissue; and iii) comparison of quantification of staining in cancerous versus noncancerous tissue. 43. The composition of claim 38, wherein the patient is identified as having a CD47-expressing cancer that is characterized by CD47 expression. 44. The composition according to claim 43, wherein the identification includes i) taking cancer tissue; and ii) detection of CD47 in isolated cancer tissue. 45. The composition of claim 38, wherein the patient is identified as having increased levels of macrophage infiltration in cancerous tissue compared to non-cancerous tissue. 46. The composition according to claim 45, where the identification includes i) taking cancerous tissue and surrounding non-cancerous tissue from the patient; ii) detection of macrophages in isolated cancerous tissue and in non-cancerous tissue; and iii) comparison of quantification of staining in cancerous versus noncancerous tissue. 47. The composition according to claim 46, wherein staining of macrophages is carried out using an antibody against CD163. 48. The composition according to any one of claims 38 to 47, wherein the cancer characterized by the expression of CD47 is hematological malignancy or solid cancer. 49. The composition according to any one of claims 38-48, where the CD47-expressing cancer is selected from non-Hodgkin's lymphoma, B-lymphoblastic lymphoma; B-cell chronic lymphoma - 73 045361 cytic leukemia/lymphoma of small lymphocytes, Richter syndrome, follicular lymphoma, multiple myeloma, myelofibrosis, polycythemia vera, T-cell lymphoma of the skin, monoclonal gammopathy of unknown etiology (MGUS), myelodysplastic syndrome (MDS), immunoblastic large cell lymphoma, B - lymphoblastic progenitor cell leukemia, acute myeloid leukemia (AML) and anaplastic large cell lymphoma. 50. The composition according to any one of claims 38 to 48, wherein the CD47-expressing cancer is selected from lung cancer, pancreatic cancer, breast cancer, liver cancer, ovarian cancer, testicular cancer, kidney cancer, bladder cancer, spine cancer, brain cancer, cervical cancer, endometrial cancer, colorectal cancer, anal cancer, esophageal cancer, gallbladder cancer, gastrointestinal cancer, stomach cancer, sarcoma, head and neck cancer, melanoma, skin cancer, prostate cancer, cancer pituitary gland, uterine cancer, vaginal cancer and thyroid cancer. 51. The composition of claim 50, wherein the CD47-expressing cancer is selected from lung cancer, sarcoma, colorectal cancer, head and neck cancer, ovarian cancer, pancreatic cancer, gastric cancer, melanoma and breast cancer. 52. The composition according to any one of claims 38 to 51, wherein the composition is administered in combination with an inhibitor of an immune checkpoint molecule selected from one or more of: programmed cell death protein 1 (PD-1), programmed cell death ligand 1 (PD- L1), PD-L2, cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), T-cell immunoglobulin and mucin domain-containing protein 3 (TIM-3), lymphocyte activation gene 3 (LAG-3), carcinomaembryonic antigen-dependent cell adhesion molecule 1 (CEACAM-1), CEACAM-5, V-domain immunoglobulin suppressor of T-cell activation (VISTA), B- and T-lymphocyte attenuator (BTLA), T-cell immunoreceptor with Ig and ITIM domains (TIGIT), immunoglobulin -like leukocyte associated receptor (type 1) (LAIR1), CD160, 2B4 or TGFR. 53. The composition according to any one of claims 38 to 52, wherein the composition is administered in combination with an agonistic anti-CD40 antibody. 54. The composition of claim 53, wherein the anti-CD40 agonist antibody has low levels of fucosylation or is afucosylated. 55. The composition according to any one of claims 38 to 54, wherein the composition is administered in combination with an antibody-drug conjugate (ADC), wherein the ADC antibody specifically binds to a protein that is expressed on the extracellular surface of the cancer cell and the antibody is conjugated to a linker-drug compound a product containing a cytotoxic agent. 56. The composition according to claim 55, wherein the cytotoxic agent is auristatin. 57. The composition of claim 56, wherein the ADC antibody is conjugated to a linker-drug compound selected from the group consisting of vcMMAE and mcMMAF. 58. The composition according to any one of claims 38 to 57, wherein the shielded antibody comprises at least one shielding domain containing a protease-cleavable linker, characterized in that the protease-cleavable linker is cleaved in the tumor microenvironment. 59. The composition of claim 58, wherein upon cleavage of the protease-cleavable linker in the tumor microenvironment, the shielding domain is released from the anti-CD47 antibody or antigen-binding fragment thereof. 60. The composition of claim 58 or 59, wherein the protease-cleavable linker contains the amino acid sequence IPVSLRSG (SEQ ID NO: 73) or GPLGVR (SEQ ID NO: 57). 61. The composition according to any one of claims 58-60, wherein the protease-cleavable linker contains an MMP cleavage site. 62. The composition of claim 61, wherein the MMP cleavage site is selected from the group consisting of an MMP2 cleavage site, an MMP7 cleavage site, an MMP9 cleavage site, and an MMP13 cleavage site. 63. The composition according to any one of claims 58-62, where after cleavage in the tumor microenvironment, the released anti-CD47 antibody or antigen-binding fragment thereof binds to CD47 with an affinity that is at least about 100 times higher, from 200 times to 1500 times higher affinity screened antibody to CD47. 64. The composition according to any one of claims 38 to 63, wherein the anti-CD47 antibody or antigen-binding fragment thereof or the anti-CD47 antibody or antigen-binding fragment thereof of the screened antibody exhibits reduced hemagglutination in vitro compared to the parent anti-CD47 antibody. 65. The composition of claim 64, wherein the parent antibody is Ab47 comprising a heavy chain variable region of SEQ ID NO: 2 and a light chain variable region of SEQ ID NO: 10. 66. The composition according to any one of claims 38 to 65, wherein administration of an anti-CD47 antibody or an antigen-binding fragment thereof or a screened antibody does not cause hemagglutination in the patient. 67. The composition according to any one of claims 38 to 66, wherein the anti-CD47 antibody, or an antigen-binding fragment thereof, or a shielded antibody causes apoptosis of CD47-expressing cells in vitro and/or in vivo. 68. The composition according to claim 67, wherein the cells expressing CD47 are cancer cells.