Therapeutic methods using constrained, conditionally activated binding proteins - Patents.com

Abstract

The present invention relates to a conditional bispecific redirecting activation construct, or COBRA, administered in an active prodrug format. Upon exposure to tumor proteases, the construct is cleaved and activated, thereby allowing it to bind both tumor target antigen (TTA) and CD3, thereby recruiting CD3-expressing T cells to the tumor and providing therapy.

Description

[Technical field]

[0001] I. CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to U.S. Provisional Patent Application No. 63 / 171,556, filed April 6, 2021, and U.S. Provisional Patent Application No. 63 / 297,662, filed January 7, 2022, the disclosures of which are incorporated by reference herein in their entireties for all purposes.

[0002] II. REFERENCE TO A "SEQUENCE LISTING", TABLE, OR COMPUTER PROGRAM LISTING EXHIBIT SUBMITTED ON A COMPACT DISC. The Sequence Listing contained in the file entitled "118459-5016-WO_ST25.txt" having a size of 984 kilobytes has been submitted electronically herewith, and the contents of the txt file are incorporated herein by reference in their entirety. [Background technology]

[0003] III. Background technology Selective destruction of individual cells or specific cell types is often desirable in various clinical settings.For example, specifically destroying tumor cells while leaving healthy cells and tissues as intact and undamaged as possible is a major goal of cancer treatment.One such method is by inducing an immune response against tumors, allowing immune effector cells such as natural killer (NK) cells or cytotoxic T lymphocytes (CTLs) to attack and destroy tumor cells.

[0004] The use of intact monoclonal antibodies (mAbs), which offer excellent binding specificity and affinity for tumor-associated antigens, has been successfully applied in the areas of cancer therapy and diagnosis. However, the large size of intact mAbs, their poor biodistribution, low potency, and poor long-term persistence in the blood pool limit their clinical use. For example, intact antibodies can show specific accumulation in the tumor region. Biodistribution studies have shown heterogeneous antibody distribution with initial accumulation in peripheral regions when tumors are probed. Tumor necrosis leads to heterogeneous antigen distribution and high interstitial pressure, making it impossible for intact antibody constructs to reach the core of the tumor. In contrast, smaller antibody fragments show rapid tumor localization, penetrate deeper into the tumor, and are cleared relatively quickly from the bloodstream.

[0005] However, many antibodies, including scFv and other constructs, exhibit "on-target / off-tumor" effects, meaning that the molecules are active against non-tumor cells and cause side effects, some of which may be toxic.

[0006] One form of immunotherapy that has been effectively used to treat hematological cancers utilizes bispecific antibodies that have specificity for both tumor cells and the CD3 antigen present on T cells. Bispecific antibodies that bind to T cells have two different binding sites, one that specifically targets an antigen on the surface of tumor cells and the other that binds to an activating receptor on the surface of T cells. The first therapeutic molecules of this type, called bispecific T cell engagers, have been shown to be highly potent in inducing cytotoxic T cell responses to specific target cells both in vitro and in vivo. See, for example, Jitschin R.et al., J Immunother.Cancer,6:116(2018);Huehls AM et al.,Immunol Cell Biol.,93:290-96(2015);Fu M.et al.,Front.Immunol.,10:1396(2019);and Dreier T.et al.,J Immunol.,(170):4397-402(2003). When the selected target antigen is expressed only on tumor cells and nonessential normal cells, this strategy has generated highly potent therapeutic molecules such as blinatumomab, which targets the CD19 antigen on the surface of certain B cell leukemias (and normal B cells). See, e.g., Wu J. et al., J. Hematol. Oncol., 8:104 (2015); Goebeler ME et al., Leuk. Lymphoma, 57:1021-32 (2016); and Bargou R. et al., Science, 321:974-77 (2008).

[0007] However, because most solid tumor antigens are also expressed on essential normal cells, it has been more difficult to utilize this strategy for the treatment of solid tumors. See, e.g., Zhang S. et al., Clin. Cancer Res. 4: 2669-76 (1998); Silver DA et al., Clin. Cancer Res. 3: 81-85 (1997); Parkhurst MR et al., Mole. Ther., 19: 620-26 (2011); Palmer DC et al., PNASUSA 105: 8061-66 (2008); and Cheever MA et al., Clin. Cancer Res., 15: 5323-37 (2009). This problem is further complicated by the need to deliver high dose levels to patients with solid tumors due to high interstitial pressure and poor tumor penetration of therapeutic agents. See, e.g., Thurber GM et al., Adv. Drug Deliv. Rev., 60:1421-34 (2008); Minchinton and Tannock, Nat. Rev. Cancer, 6:583-92 (2006); and Jain RK, Cancer Res., 50:814s-9s (1990).

[0008] Therefore, if a therapeutic agent also targets normal cells expressing solid tumor antigens with a cytotoxic T cell response, it may not be possible to deliver a dose that would be effective in patients with cancer. Strategies to address this problem have included designing bispecific T cell-binding prodrugs that can be selectively activated in the tumor microenvironment (TME). Several physiological characteristics of the TME set it apart from most normal tissues. For example, unregulated tissue growth in tumors can lead to an anoxic environment with reduced extracellular pH. See, e.g., Webb BA et al., Nat. Rev. Cancer, 11:671-77 (2011); Stubbs M. et al., Mol. Med. Today, 6:15-19 (2000); Pillai SR et al., Cancer, 38:205-22 (2019); Muz B. et al., Hypoxia, 3:83-92 (2015); Moulder and Rockwell, Cancer Metastasis Rev., 5:313-41 (1987); and Engin K. et al., Int. J. Hyperther., 11:211-16 (1995). Furthermore, rapid cell division in tumors is often associated with increased extracellular matrix remodeling and increased protein degradation. See, e.g., Rakashanda S. et al., Biotechnol. Mole. Biol. Rev., 7:90-101 (2012); and DeClerck YA et al., Am. J. Pathol., 164:1131-39 (2004). Due to dysregulated gene expression, tumors also frequently co-express cell surface antigens that are not co-expressed in normal tissues. To exploit these features of the TME, antibodies have been engineered to bind preferentially at low pH or only in the presence of aptamers that are activated under hypoxic conditions. See, for example, Zhou F. et al., J. Am. Chem. Soc., 141:18421-27 (2019); Sharp L. et al., Cancer Res., 79:2708 (2019); and Chang C. et al., Cancer Res., 79:356 (2019). Other groups have designed protease-activated bispecific T cell engagers that either block one or both antibody binding sites using masks attached via protease-cleavable linkers, or use protein domain complementarity and Boolean logic gating to ensure bispecific activation only in the presence of two different tumor antigens. See, e.g., Minogue E. et al., Blood, 134:2653 (2019); Boustany LM et al., Mole. Cancer Ther., 17:A164-A (2018); and Ellerman D. et al., Methods, 154:102-17 (2019). [Prior art documents] [Non-patent literature]

[0009] [Non-Patent Document 1] Jitschin R.et al.,J Immunother.Cancer,6:116(2018) [Non-Patent Document 2] Huehls AM et al.,Immunol Cell Biol.,93:290-96(2015) [Non-Patent Document 3] Fu M. et al.,Front.Immunol.,10:1396(2019) [Non-Patent Document 4] Dreier T.et al.,J Immunol.,(170):4397-402(2003)

Non-licensed Document 5

Non-licensed Document 6

Non-licensed Document 7

Non-licensed literature 9

Non-licensed literature 10

Non-licensed Document 11

Non-licensed Document 12

Non-licensed Document 13

Non-licensed Document 14

Non-licensed Document 15

Non-licensed Document 16

Non-licensed Document 17

Non-licensed Document 18

Non-licensed Document 19

Non-licensed Document 20

Non-licensed Document 21

Non-licensed Document 22

Non-licensed Document 23

Non-licensed Document 24

Non-licensed Document 25

Non-licensed Document 26

Non-licensed Document 27

Non-licensed Document 28

[0010] IV. Overview of the Invention Thus, there is a need in the art for improved bispecific cancer immunotherapies for the treatment of solid tumors. In particular, there is a need in the art for bispecific immunotherapies with reduced off-target toxicity. Among other aspects, the present invention relates to novel constructs that are selectively activated in the presence of tumor proteases and methods for treating solid tumors using such constructs.

[0011] The present invention provides many different protein compositions for the treatment of cancer. Thus, in one aspect, the present invention provides a constrained Fv domain comprising, from N-terminus to C-terminus, a first single domain antigen binding domain (sdABD) (sdABD-TTA) that binds to a human tumor target antigen (TTA), b) a first domain linker, c) a constrained Fv domain comprising i) a first variable heavy domain comprising vhCDR1, vhCDR2 and vhCDR3, ii) a constrained non-cleavable linker (CNCL), and iii) a first variable light domain comprising vlCDR1, vlCDR2 and vlCDR3, d) a second domain linker, e) a second sdABD-TTA, f) a cleavable linker (CL), and g) a constrained Fv domain comprising i) a first pseudo light variable domain, ii) a non-cleavable linker (NCL), and iii) a first pseudo heavy variable domain. and h) a third domain linker; and i) a third sdABD that binds human serum albumin, wherein the first variable heavy domain and the first variable light domain are capable of binding to human CD3, but the constrained Fv domain does not bind to CD3, and the first variable heavy domain and the first pseudo-variable light domain associate intramolecularly to form an inactive Fv, and the first variable light domain and the first pseudo-variable heavy domain associate intramolecularly to form an inactive Fv.

[0012] In a further aspect, the invention provides a constrained Fv domain comprising, from N-terminus to C-terminus, (a) a first sdABD-TTA; (b) a first domain linker; (c) a constrained Fv domain comprising (i) a first variable heavy domain comprising vhCDR1, vhCDR2, and vhCDR3, ii) a constrained cleavable linker (CCL), and iii) a first variable light domain comprising vlCDR1, vlCDR2, and vlCDR3; d) a second domain linker; e) a second sdABD-TTA; f) a cleavable linker (CL); and g) i) a first pseudo light variable domain, ii) a non-cleavable linker (NCL), and iii) a constrained pseudo-Fv domain comprising a first pseudo heavy variable domain; h) a third domain linker; and i) a third sdABD that binds human serum albumin, wherein the first variable heavy domain and the first variable light domain are capable of binding to human CD3, but the constrained Fv domain does not bind to CD3, and the first variable heavy domain and the first pseudo-variable light domain associate intramolecularly to form an inactive Fv, and the first variable light domain and the first pseudo-variable heavy domain associate intramolecularly to form an inactive Fv.In a further aspect, the invention provides a constrained Fv domain comprising, from N-terminus to C-terminus, a) a single domain antigen binding domain (sdABD) that binds a human tumor target antigen (TTA) (sdABD-TTA); b) a first domain linker; c) a constrained Fv domain comprising i) a first variable heavy domain comprising vhCDR1, vhCDR2 and vhCDR3, ii) a constrained non-cleavable linker (CNCL), and iii) a first variable light domain comprising vlCDR1, vlCDR2 and vlCDR3; d) a cleavable linker (CL); e) a second sdABD that binds human serum albumin; and f) a domain and g) a constrained pseudo-Fv domain comprising i) a first pseudo light variable domain, ii) a non-cleavable linker (NCL), and iii) a first pseudo heavy variable domain, wherein the first variable heavy domain and the first variable light domain are capable of binding to human CD3, but the constrained Fv domain does not bind to CD3, and the first variable heavy domain and the first pseudo variable light domain associate intramolecularly to form an inactive Fv, and the first variable light domain and the first pseudo variable heavy domain associate intramolecularly to form an inactive Fv.

[0013] In further embodiments of the above Format 1, Format 2 and Format 4 proteins, the first variable heavy domain is N-terminal to the first variable light domain and the pseudo light variable domain is N-terminal to the pseudo variable heavy domain.

[0014] In further embodiments of the above Format 1, Format 2 and Format 4 proteins, the first variable heavy domain is N-terminal to the first variable light domain and the pseudo variable heavy domain is N-terminal to the pseudo variable light domain.

[0015] In further embodiments of the above Format 1, Format 2 and Format 4 proteins, the first variable light domain is N-terminal to the first variable heavy domain and the pseudo light variable domain is N-terminal to the pseudo variable heavy domain.

[0016] In further embodiments of the above Format 1, Format 2 and Format 4 proteins, the first variable light domain is N-terminal to the first variable heavy domain and the pseudo variable heavy domain is N-terminal to the pseudo variable light domain.

[0017] In a further aspect, the invention provides Type 1 and Type 2 proteins, wherein the first and second TTAs are the same.

[0018] In another aspect, the invention provides Type 1 and Type 2 proteins, wherein said first TTA and said second TTA are different.

[0019] In a further aspect, the present invention provides Format 1, 2 and 4 proteins, wherein the first TTA and the second TTA are selected from EGFR, EpCAM, FOLR1 and B7H3, and the sequences may be selected from the group consisting of SEQ ID NO:1, SEQ ID NO:5, SEQ ID NO:9, SEQ ID NO:13, SEQ ID NO:17, SEQ ID NO:21, SEQ ID NO:25, SEQ ID NO:29, SEQ ID NO:33, SEQ ID NO:37 and SEQ ID NO:41.

[0020] In another aspect, the invention provides Format 1, 2 and 4 proteins, wherein said half-life prolonging domain has SEQ ID NO:45.

[0021] In a further aspect, the invention provides Type 1, 2 and 4 proteins, wherein the cleavable linker is cleaved by a human protease selected from the group consisting of MMP2, MMP9, meprin A, meprin B, cathepsin S, cathepsin K, cathepsin L, granzyme B, uPA, calecrein 7, matriptase and thrombin.

[0022] In another aspect, the present invention provides a protein selected from the group consisting of Pro186, Pro225, Pro226, Pro233, Pro311, Pro312, Pro313, Pro495, Pro246, Pro254, Pro255, Pro256, Pro420, Pro421, Pro432, Pro479, Pro480, Pro187, Pro221, Pro222, Pro223, Pro224, Pro393, Pro394, Pro395, Pro396, Pro429, Pro430 and Pro431.

[0023] In a further aspect, the invention provides nucleic acid encoding a Format 1, Format 2 or Format 4 protein described herein, as well as expression vectors and host cells comprising the nucleic acid encoding the proteins.

[0024] In further aspects, the invention provides methods of making the proteins of the invention and methods of treating patients in need thereof.

[0025] In a further aspect, the invention provides a method for the preparation of a quasi-Fv domain comprising: a) a first protein comprising, from N-terminus to C-terminus, i) a first sdABD-TTA; ii) a first domain linker; iii) a pseudo-Fv domain comprising, from N-terminus to C-terminus, 1) a variable heavy chain comprising vhCDR1, vhCDR2 and vhCDR3, 2) a cleavable linker and 3) a first pseudo-variable light domain comprising iVLCDR1, iVLCDR2 and iVLCDR3; iv) a second domain linker; and v) an sdABD-HSA; and a) a first protein comprising, from N-terminus to C-terminus, i) a third sdABD that binds a human tumor target antigen; ii) a third domain linker; iii) a variable light chain comprising, from N-terminus to C-terminus, 1) a variable light chain comprising VLCDR1, VLCDR2 and VLCDR3, 2) a cleavable linker and 3) iVHCDR1, iVHCDR2 and iVHCDR3. and a first second protein comprising iv) a fourth domain linker and v) a sdABD-HSA, wherein the first variable heavy domain and the first variable light domain are capable of binding to human CD3 when associated, the first variable heavy domain and the first pseudo variable light domain intermolecularly associate to form an inactive Fv, the first variable light domain and the first pseudo variable heavy domain intermolecularly associate to form an inactive Fv, and the first and third sdABDs are selected from the group consisting of SEQ ID NO:1, SEQ ID NO:5, SEQ ID NO:9, SEQ ID NO:13, SEQ ID NO:17, SEQ ID NO:21, SEQ ID NO:25, SEQ ID NO:29, SEQ ID NO:33, SEQ ID NO:37, and SEQ ID NO:41.

[0026] In a further aspect, the invention provides a method for the preparation of a polypeptide comprising: a) a first protein comprising, from N-terminus to C-terminus, i) a first sdABD-TTA, ii) a first domain linker, iii) a second sdABD-TTA, iv) a second domain linker, and iii) a pseudo-Fv domain comprising, from N-terminus to C-terminus, 1) a variable heavy chain comprising vhCDR1, vhCDR2 and vhCDR3, 2) a cleavable linker and 3) a first pseudo-variable light domain comprising iVLCDR1, iVLCDR2 and iVLCDR3, iv) a third domain linker, and v) an sdABD-HSA; and a) from N-terminus to C-terminus, i) a third sdABD-TTA, ii) a fourth domain linker, iii) a fourth sdABD-TTA, iv) a fifth domain linker, and iii) a second domain linker. to the C-terminus, 1) a variable light chain comprising VLCDR1, VLCDR2 and VLCDR3, 2) a cleavable linker and 3) a first pseudo-variable heavy domain comprising iVHCDR1, iVHCDR2 and iVHCDR3, iv) a sixth domain linker, and v) a first second protein comprising sdABD-HSA, wherein said first variable heavy domain and said first variable light domain are capable of binding to human CD3 when associated, said first variable heavy domain and said first pseudo-variable light domain associate intermolecularly to form an inactive Fv, and said first variable light domain and said first pseudo-variable heavy domain associate intermolecularly to form an inactive Fv.

[0027] In a further aspect, the Format 3A and Format 3B proteins have a sdABD-HSA having SEQ ID NO:45.

[0028] In another embodiment, the Format 3A and Format 3B proteins have a sdABD-TTA that binds to a TTA selected from EGFR, EpCAM, FOLR1 and B7H3. The sdABD-TTA may be selected from the group consisting of SEQ ID NO:1, SEQ ID NO:5, SEQ ID NO:9, SEQ ID NO:13, SEQ ID NO:17, SEQ ID NO:21, SEQ ID NO:25, SEQ ID NO:29, SEQ ID NO:33, SEQ ID NO:37 and SEQ ID NO:41.

[0029] In another aspect, the invention provides nucleic acid compositions comprising a first nucleic acid encoding a first protein member of a prodrug pair and a second nucleic acid encoding a second protein member of the pair, as well as expression vectors and host cells containing the nucleic acids. [Brief description of the drawings]

[0030] [Figure 1]The "Format 1" type of protease activation of the invention is shown, referred to herein as a "constrained cleavable construct" or "cc construct". In this embodiment, the representative construct is Pro140 in which there are two ABDs for TTA (as shown, these are both the same, but as described herein they can be different). Upon cleavage, the prodrug construct is split into three components, one containing the α-TTA domain linked to the active VH of αCD3 via a domain linker, the second component containing the α-TTA domain linked to the active VL of αCD3 via a domain linker, and the "remaining" piece containing the half-life extending domain linked to the inactive VH and VL. The two active variable domains are then free to associate to form a functional anti-CD3 binding domain. In the "Format 1" embodiment, the resulting active component is trivalent. Note that there is a monovalent that binds CD3 and a bivalent that binds TTA, resulting in a bispecific binding protein, although in some cases this trivalent can be trispecific, with a monovalent that binds CD3, a monovalent that binds the first TTA, and a monovalent that binds the second TTA. Also shown is a half-life prolonging domain, in many embodiments an anti-human serum albumin (HSA) domain, as sdABD as defined herein, although as discussed herein this is optional and / or can be replaced by other half-life prolonging domains. In addition, the half-life prolonging domain can also be at the N-terminus of the construct, or can be internal as well. Also shown are the VH and VL of the Fv, and the iVH and iVL of the pseudo-Fv in a specific order, for example, VH-linker-VL (and iVL-linker-iVH) from N-terminus to C-terminus, although as will be appreciated by those skilled in the art, these can be reversed (VL-linker-VH and iVH-linker-iVL). Alternatively, one of these Fvs could be in one orientation and the other in the other, but expression of the protein in the orientation shown here was surprisingly higher than in the other orientations. [Diagram 2]The "Format 2" type of protease activation of the present invention is shown, herein referred to as "constrained non-cleavable construct" or "CNCL construct" and sometimes referred to herein as "dimerization construct" as discussed herein. These constructs do not isomerize as discussed herein. Upon cleavage, the two prodrug constructs are split into four components: two half-life extending domains (in this case sdABD to HSA) linked to two pseudo domains (which may or may not be able to self-assemble depending on the linker length and inactivation mutation), and two active moieties that self-assemble into a dimeric active moiety containing four anti-TTA domains (which may all be the same, or two are the same and the other two are different). In the "Format 2" embodiment, the resulting active component is hexavalent. Note that there is a bivalent that binds to CD3 and a tetravalent that binds to TTA, resulting in a bispecific binding protein, but in some cases, this hexavalent can be trispecific, with a bivalent that binds to CD3, a bivalent that binds to a first TTA, and a bivalent that binds to a second TTA. Also, a half-life prolonging domain, in many embodiments an anti-human serum albumin (HSA) domain, as sdABD as defined herein, is shown, but as discussed herein, this is optional and / or can be replaced by other half-life prolonging domains. In addition, the half-life prolonging domain can also be at the N-terminus of the construct, or can be internal as well. Also, the VH and VL of the Fv and the iVH and iVL of the pseudo-Fv are in a specific order, for example, VH-linker-VL (and iVL-linker-iVH) from N-terminus to C-terminus, but as will be understood by those skilled in the art, these can be reversed (VL-linker-VH and iVH-linker-iVL). Alternatively, one of these Fvs could be in one orientation and the other in the other, but expression of the protein in the orientation shown here was surprisingly higher than in the other orientations. [Figure 3A]As discussed further herein, there are two different polypeptide chains that together constitute the MCE therapeutic, and therefore a "Format 3" type of construct is shown, sometimes referred to as a "half construct" or "half COBRA™" as outlined herein. In this embodiment, the constructs are delivered in pairs, and intramolecular self-assembly before cleavage results in an inactive anti-CD3 Fv domain. Upon cleavage, the inactive variable domain is released, and then the two active variable domains assemble intermolecularly to form the active anti-CD3 binding domain. The two sdABD-TTAs bind to corresponding receptors on the tumor cell surface, and cleavage is performed by a protease. This allows for intermolecular assembly, as the molecules are physically held in place and promote assembly of the active anti-CD3 domain. As described above for Forms 1 and 2, in this embodiment, the N-terminal to C-terminal order of the variable domains may be reversed or mixed as well. Additionally, the sdABD(HSA) may be at either the N-terminus or C-terminus of each half construct. Pro16 has a sdABD(HSA) at the C-terminus and Pro17 has a sdABD(HSA) at the N-terminus (see Pro19, SEQ ID NO:XX, has a sdABD(HSA) at the C-terminus). Format 3 constructs with a single sdABD-TTA domain per half construct are shown. [Figure 3B]As discussed further herein, there are two different polypeptide chains that together constitute the MCE therapeutic, and therefore a "Format 3" type of construct is shown, sometimes referred to as a "half construct" or "half COBRA™" as outlined herein. In this embodiment, the constructs are delivered in pairs, and intramolecular self-assembly before cleavage results in an inactive anti-CD3 Fv domain. Upon cleavage, the inactive variable domain is released, and then the two active variable domains assemble intermolecularly to form the active anti-CD3 binding domain. The two sdABD-TTAs bind to corresponding receptors on the tumor cell surface, and cleavage is performed by a protease. This allows for intermolecular assembly, as the molecules are physically held in place and promote assembly of the active anti-CD3 domain. As described above for Forms 1 and 2, in this embodiment, the N-terminal to C-terminal order of the variable domains may be reversed or mixed as well. Additionally, the sdABD(HSA) may be at either the N-terminus or C-terminus of each half construct. Pro16 has a sdABD(HSA) at the C-terminus and Pro17 has a sdABD(HSA) at the N-terminus (see Pro19, SEQ ID NO:XX, has a sdABD(HSA) at the C-terminus). Format 3 constructs are shown with two sdABD-TTAs per half construct in a "dual targeting" or "hetero-targeting" format. Note that FOLR1 and EGFR are used as the two TTAs, but other combinations as outlined herein can also be used. [Figure 4]A "Format 4" type construct is shown, similar to the "Format 2" construct, but with only a single sdABD-TTA. The figure shows EGFR from sdABD-TTA, but other TTAs can be used, as will be appreciated by those skilled in the art. Upon cleavage, the prodrug construct splits into two components, a half-life extending domain (in this case sdABD to HSA) linked to a pseudo-Fv, and an active portion that self-assembles into a dimeric active portion containing two anti-TTA domains in the presence of a second active portion from a different cleavage molecule. Note that in the "Format 4" embodiment, the resulting active portion is tetravalent, with a bivalent that binds CD3 and a bivalent that binds TTA, resulting in a bispecific binding protein. Also shown is a half-life extending domain, in many embodiments an anti-human serum albumin (HSA) domain, as sdABD(1 / 2) as defined herein, but as discussed herein this is optional and / or can be replaced by other half-life extending domains. In addition, the half-life extending domain may also be at the N-terminus of the construct, or may be internal as well. Also, the VH and VL of the Fv and the iVH and iVL of the pseudo-Fv are in a specific order, e.g., VH-linker-VL (and iVL-linker-iVH) from N-terminus to C-terminus, although as will be appreciated by those skilled in the art, they may be reversed (VL-linker-VH and iVH-linker-iVL). Alternatively, one of these Fvs may be in one orientation and the other in another orientation, although expression of the protein in the orientation shown here was surprisingly higher than in other orientations. [Figure 5A] A number of sequences of the invention are shown. For antigen binding domains, the CDRs are underlined. As more fully outlined herein, these domains can be assembled in a wide variety of configurations in the present invention, including "Format 1", "Format 2", "Format 3" and "Format 4" orientations. [Figure 5B] A number of sequences of the invention are shown. For antigen binding domains, the CDRs are underlined. As more fully outlined herein, these domains can be assembled in a wide variety of configurations in the present invention, including "Format 1", "Format 2", "Format 3" and "Format 4" orientations. [Figure 5C] A number of sequences of the invention are shown. For antigen binding domains, the CDRs are underlined. As more fully outlined herein, these domains can be assembled in a wide variety of configurations in the present invention, including "Format 1", "Format 2", "Format 3" and "Format 4" orientations. [Figure 5D] A number of sequences of the invention are shown. For antigen binding domains, the CDRs are underlined. As more fully outlined herein, these domains can be assembled in a wide variety of configurations in the present invention, including "Format 1", "Format 2", "Format 3" and "Format 4" orientations. [Figure 5E] A number of sequences of the invention are shown. For antigen binding domains, the CDRs are underlined. As more fully outlined herein, these domains can be assembled in a wide variety of configurations in the present invention, including "Format 1", "Format 2", "Format 3" and "Format 4" orientations. [Figure 5F] A number of sequences of the present invention are shown. For antigen binding domains, CDRs are underlined. As more fully outlined herein, these domains can be assembled in a wide variety of configurations in the present invention, including "Format 1", "Format 2", "Format 3" and "Format 4" orientations. Of note, SEQ ID NO:90 is cleaved by MMP9 slightly faster than SEQ ID NOs:75 and 76, and SEQ ID NO:91 is cleaved slower than SEQ ID NOs:75 and 76. [Figure 5G] A number of sequences of the invention are shown. For antigen binding domains, the CDRs are underlined. As more fully outlined herein, these domains can be assembled in a wide variety of configurations in the present invention, including "Format 1", "Format 2", "Format 3" and "Format 4" orientations. [Figure 6A] A number of suitable protease cleavage sites are shown. As will be appreciated by those of skill in the art, these cleavage sites can be used as cleavable linkers. In some embodiments, for example, when a more flexible cleavable linker is required, there may be additional amino acids (typically glycine and serine) at either or both the N-terminus and C-terminus of these cleavage sites. [Figure 6B] A number of suitable protease cleavage sites are shown. As will be appreciated by those of skill in the art, these cleavage sites can be used as cleavable linkers. In some embodiments, for example, when a more flexible cleavable linker is required, there may be additional amino acids (typically glycine and serine) at either or both the N-terminus and C-terminus of these cleavage sites. [Figure 7A] We present some data relating to the "Format 3" or "half-COBRA™" construct, which shows that the Format 3 construct binds cooperatively to CD3 and creates a CD3 binding site after cleavage by a protease (in this case the EK protease, although any of the protease cleavage sites outlined herein and shown in Figures 5 and 6 may be used) as shown by sandwich FACS analysis. [Figure 7B] We present some data relating to the "Format 3" or "half-COBRA™" construct, which shows that the Format 3 construct binds cooperatively to CD3 and creates a CD3 binding site after cleavage by a protease (in this case the EK protease, although any of the protease cleavage sites outlined herein and shown in Figures 5 and 6 may be used) as shown by sandwich FACS analysis. [Figure 7C] We present some data relating to the "Format 3" or "half-COBRA™" construct, which shows that the Format 3 construct binds cooperatively to CD3 and creates a CD3 binding site after cleavage by a protease (in this case the EK protease, although any of the protease cleavage sites outlined herein and shown in Figures 5 and 6 may be used) as shown by sandwich FACS analysis. [Figure 7D]We present some data relating to the "Format 3" or "half-COBRA™" construct, which shows that the Format 3 construct binds cooperatively to CD3 and creates a CD3 binding site after cleavage by a protease (in this case the EK protease, although any of the protease cleavage sites outlined herein and shown in Figures 5 and 6 may be used) as shown by sandwich FACS analysis. [Figure 8A] We show that protease cleavage synergizes with complementary half-COBRA™ pairs to activate T cell killing of EGFR+ target cells, and that constructs cleaved with proteases alone, but at different concentrations, do not affect target cell viability. [Figure 8B] We show that protease cleavage synergizes with complementary half-COBRA™ pairs to activate T cell killing of EGFR+ target cells, and that constructs cleaved with proteases alone, but at different concentrations, do not affect target cell viability. [Figure 8C] We show that protease cleavage synergizes with the complementary half-COBRA™ pair to activate T cell killing of EGFR+ target cells, however, the combination shows a marked decrease in target cell viability in the presence of the protease. [Figure 8D] 1 shows that protease cleavage activates T cell killing of EGFR+ target cells in cooperation with complementary half-COBRA™ pairs. A general mechanism is shown. [Figure 9] Several non-targeting controls are shown for use in assays testing the effectiveness of Format 1 constructs. [Figure 10A] It is shown that the formation of an active CD3 binding domain depends on target binding of both "arms", one of which is in each of the two constructs, e.g., the sdABD-TTA domain. TDCC assays were performed as described in the Examples. [Figure 10B]It is shown that the formation of an active CD3 binding domain depends on target binding of both "arms", one of which is in each of the two constructs, e.g., the sdABD-TTA domain. TDCC assays were performed as described in the Examples. [Figure 10C] It is shown that the formation of an active CD3 binding domain depends on target binding of both "arms", one of which is in each of the two constructs, e.g., the sdABD-TTA domain. TDCC assays were performed as described in the Examples. [Figure 10D] It is shown that the formation of an active CD3 binding domain depends on target binding of both "arms", one of which is in each of the two constructs, e.g., the sdABD-TTA domain. TDCC assays were performed as described in the Examples. [Figure 10E] It is shown that the formation of an active CD3 binding domain depends on target binding of both "arms", one of which is in each of the two constructs, e.g., the sdABD-TTA domain. TDCC assays were performed as described in the Examples. [Figure 10F] It is shown that the formation of an active CD3 binding domain depends on target binding of both "arms", one of which is in each of the two constructs, e.g., the sdABD-TTA domain. TDCC assays were performed as described in the Examples. [Figure 11] A schematic diagram of a suitable half-COBRA™ pair is shown, where "Mep" stands for meprin protease cleavage site, "His-6" is a tag as described more fully herein, ST14 is a matriptase protease cleavage site, and "Thb" is a thrombin protease cleavage site. [Figure 12A] TDCC data associated with the constructs of FIG. 11 are shown, demonstrating that the addition of half-COBRA pairs prior to cleavage provides an effect in cancer cells, OvCAR8 cells. [Figure 12B] TDCC data associated with the constructs of FIG. 11 are shown, demonstrating that addition of the half COBRA pair prior to cleavage provides an effect in cancer cells, HCT116 cells. [Figure 12C] TDCC data associated with the constructs of FIG. 11 are shown, demonstrating that the addition of a half COBRA pair prior to cleavage provides an effect in cancer cells, LoVo cells. [Figure 13] We demonstrate that the MMP9 linker is stable in vivo. A single intravenous bolus dose of either Pro40 (MMP9 cleavable) or Pro74 (non-cleavable) was administered to NSG mice via the tail vein at a dose level of 0.5 mg / kg. Dosing solutions for each compound were prepared in a vehicle of 25 mM citric acid, 75 mM L-arginine, 75 mM NaCl, and 4% sucrose, pH 7.0. Two blood samples were collected from each animal at preselected times; one was collected by orbital or submandibular bleed toward the start of the study, and the other was collected by cardiac puncture at the end. Blood collection time points were 0.083, 1, 6, 24, 72, and 168 hours. Plasma was prepared from each individual blood sample using K2EDTA tubes. Concentrations were determined using the MSD assay with MAb specific for anti-HSA sdABD and detected in the EGFR extracellular domain. [Figure 14] A schematic of the format 3A half COBRA™ constructs used in the experiments shown in Figure 15 is shown. Pro51 is a positive control because it forms an active anti-CD3 Fv and is therefore "always on". Pro98 is a negative control because its sdABD is directed against hen egg white lysozyme, which is not expressed by tumors. Pro77 and Pro53 are prodrug format 3A pairs using the sdABD against EGFR and an MMP9 cleavage site. Pro74 and Pro72 are negative control format 3A pairs because they lack a cleavage site. [Figure 15] Using two different tumor cell lines implanted in mice using the protocol of the examples, we show that the Format 1 construct works to regress tumors in vivo. Antitumor activity with the half-COBRA constructs (Pro77 and Pro53) was dependent on the inclusion of both the anti-EGFR sdABD and the MMP9 cleavable linker along with the active anti-CD3 Fv. [Figure 16]1 shows a schematic diagram of the next generation format, a full-length construct with two pseudo-Fv domains with a cleavable site between them, as generally described in US2018 / 0134789, which is incorporated herein by reference. However, as shown in the diagram below, this first generation full-length construct does not show very good conditioning, as it can isomerize to form both active and inactive constructs. [Figure 17] 1 shows that the Format 3A construct pair indeed exhibits better conditioning than the Pro100 first generation full length construct. [Figure 18] FIG. 19 shows additional first generation full length constructs that were tested. [Figure 19] It is shown that the first generation constructs exhibit high activity even in the non-cleavable form, eg, under poor conditions. [Figure 20] FIG. 1 shows that the first generation full-length construct shows two monomer peaks in analytical SEC. [Figure 21] A schematic diagram of the reason for the non-cleavable activity is shown, because the full-length first generation construct isomerizes to form two conformations, one of which is inactive due to the absence of active anti-CD3 Fv ("bivalent scFv"), and the other is a "single-chain diabody" type configuration that is active in the absence of proteases. See PEDS 23(8):667-677 (2010). [Figure 22] TDCC assay results for 2 days at 37° C. are shown using first generation single chain constructs. The results show that the non-cleavable constructs exhibit robust killing. These results led to the generation of Format 1 constructs. [Figure 23A] 1 shows the format 1 constructs used in the present invention. As will be understood by those of skill in the art and described herein, these are shown with sdABD-EGFR targeting moieties, although sdABDs for other TTAs may be used. [Figure 23B] 1 shows the format 1 constructs used in the present invention. As will be understood by those of skill in the art and described herein, these are shown with sdABD-EGFR targeting moieties, although sdABDs for other TTAs may be used. [Figure 23C]1 shows the format 1 constructs used in the present invention. As will be understood by those of skill in the art and described herein, these are shown with sdABD-EGFR targeting moieties, although sdABDs for other TTAs may be used. [Figure 23D] 1 shows the format 1 constructs used in the present invention. As will be understood by those of skill in the art and described herein, these are shown with sdABD-EGFR targeting moieties, although sdABDs for other TTAs may be used. [Figure 23E] 1 shows the format 1 constructs used in the present invention. As will be understood by those of skill in the art and described herein, these are shown with sdABD-EGFR targeting moieties, although sdABDs for other TTAs may be used. [Figure 23F] 1 shows the format 1 constructs used in the present invention. As will be understood by those of skill in the art and described herein, these are shown with sdABD-EGFR targeting moieties, although sdABDs for other TTAs may be used. [Figure 23G] 1 shows the format 1 constructs used in the present invention. As will be understood by those of skill in the art and described herein, these are shown with sdABD-EGFR targeting moieties, although sdABDs for other TTAs may be used. [Figure 24] It is shown that the format 1 construct (in this case Pro140) forms one isomer that is stable at 37°C. [Diagram 25] Figure 1 shows that the Format 1 construct has very low binding to human CD3 in the uncleaved format as measured by the Octet assay. The top line is Pro120, the middle line is Pro51 (positive control) and the bottom line is Pro140 held at either 4°C or 37°C for 3 days. [Figure 26] Similarly, it is shown that the Format 1 construct has very low TDCC activity in the uncleaved form. [Figure 27] The specific Format 1 construct, Pro140, using sdABD-EGFR and an MMP9 cleavage site as the targeting moiety was shown for in vivo studies. [Figure 28]AB show tumor regression using Format 1 constructs. [Figure 29] We show that the cleavage site within the constrained Fv can generate several different fragments, both partially and completely cleaved. Surprisingly, the partially cleaved form is more active than the completely cleaved form, leading to the generation of form 2. [Diagram 30] Schematics of several Format 2 are shown, all of which use the sdABD-EGFR targeting domain, although sdABDs for other TTAs may be used as outlined herein and listed in the sequences. Pro51 and Pro201 are positive controls (active "half" constructs) and Pro214 is a full-length negative control as it has no cleavage site. [Diagram 31] Figure 2 shows the TDCC activity of the format 2 construct Pro187, which uses a meprin cleavage site. Pro187 was 1200-fold more active in the TDCC assay when added prior to cleavage than when added uncleaved. Pre-cleaved Pro187 showed activity between the positive controls Pro51 and Pro201. Uncleaved Pro187 showed activity similar to Pro214, which does not contain a protease-cleavable linker. [Diagram 32] Figure 2 shows the TDCC activity of the format 2 construct Pro186 using the MMP9 cleavage site. Pro186 was 18-fold more active in the TDCC assay when added prior to cleavage than when added uncleaved. Uncleaved Pro186 showed activity between the positive controls Pro51 and Pro201. Uncleaved Pro186 showed greater activity than Pro214, which does not contain a protease-cleavable linker. [Diagram 33] We show that the Pro186 construct binds to cells with different levels of EGFR receptors, and CHO cells do not express EGFR on the cell surface. Pro186 saturates cells expressing different levels of EGFR at similar COBRA concentrations. [Diagram 34] FIG. 35 shows a schematic of the Format 2 constructs used for in vivo testing, all of which use the sdABD-EGFR targeting domain. [Diagram 35]The format 2 construct Pro186 is shown to be highly effective at both concentrations, and better than the format 1 construct Pro140 at lower concentrations. [Diagram 36] Shown are several format 2 constructs based on Pro186 but with different protease cleavage sites. All of these constructs utilize sdABD-EGFR for both targeting domains, but other sdABDs for different TTAs may be used, which may be the same or different. That is, both homo-targeting (both sdABDs targeting the same TTA) or hetero-targeting (one sdABD to the first TTA and the other to a different TTA) can be performed. [Figure 37] Schematic diagrams of various format 2 constructs with varying linker length between Fv domains are shown. These are shown using MMP9 cleavage sites, but others may be used as outlined herein. Similarly, these constructs all utilize sdABD-EGFR for both targeting domains, but other sdABDs for different TTAs may be used, and they may be the same or different. [Figure 38] We show that the linker length of the pseudo-Fv can be varied, for example, format 2 constructs with a short linker between the active Fvs ("short active") and a long linker between the pseudo-Fvs ("long inactive") show similar activity to "short active" and "short inactive". Thus, the conditioning of the COBRA constructs does not depend on both the active and inactive scFv linkers being constrained, and as long as one of them is constrained, single-chain diabody folding appears to be favored over bivalent scFv folding. [Figure 39] It is shown that the linker length of the active Fv may be altered, e.g., a format 2 construct with a "long active" and a "short inactive" behaves similarly to a "short active" and a "short inactive" construct. Thus, the conditioning of the COBRA construct does not depend on both the active and inactive scFv linkers being constrained, and as long as one of them is constrained, single-chain diabody folding appears to be favored over bivalent scFv folding. [Figure 40A] Schematics of several different constructs are shown. Pro188 is a format 1 construct that is similar to Pro140 except that it has a long linker (16mer) in the pseudo-Fv. Pro189 and Pro190 (format 2 constructs) are similar to Pro186 and Pro187 except that they have a long linker (16mer) in the pseudo-Fv domain. [Figure 40B] Schematics of several different constructs are shown. Pro191 and Pro192 (also format 2 constructs) are similar to Pro189 and Pro190, except that they have an additional cleavage site upstream of the sdABD(1 / 2). Pro193 (format 4) has one EGFR targeting domain, rearranged iVH and iVL in reverse order, and an additional cleavage site upstream of the sdABD(1 / 2). [Figure 40C] Schematic diagrams of several different constructs are shown. Pro195 is a format 2 construct similar to Pro186 with the same TTA, EGFR targeting domain but binding to a different epitope. Pro196, Pro197 and Pro198 are format 2 constructs with rearranged variable domains. [Diagram 41] The fact that different sdABD clones against human FOLR1 show differential killing shows that the Pro22-type construct (Pro51 with FLAG sequence instead of NCL) that binds human FOLR1 was compared with the Pro22-EGFR construct against several cell line families. [Diagram 42] Schematics of four sdABD-FOLR1 constructs are shown, including a positive control, Pro201, using sdABD-EGFR2 (two molecules intermolecularly associate to form two active Fvs against CD3), and two format 2 test articles, Pro311 using h77.2 sdABD and Pro312 using h59.3 sdABD, and two negative controls, Pro299 using h77.2 sdABD and Pro303 using h59.3 sdABD. [Diagram 43] Schematic diagram of Format 2 construct for FOLR / MMP9 in vivo design. [Diagram 44] The effect of the Pro312 construct in vivo is shown, demonstrating that the MMP9 cleavable linker is required for antitumor activity. [Diagram 45] Schematics of several formats using a sdABD against human B7H3 (sdABD-B7H3) are shown, including Pro244, a positive control (sdABD-B7H3(hF7) was used (two molecules intermolecularly associate to form two active Fvs against CD3), and two Format 2 test articles, Pro225, a Format 2 construct, and Pro295, a negative control with no cleavage site. [Diagram 46] 4 shows that Pro225 has greater conditioning compared to the control, Pro295. [Figure 47] 1 shows that a format 2 construct using a meprin linker, Pro373, shows greater conditionality compared to Pro295. [Figure 48] Shown are several sdABD-B7H3 (using the hF12 sequence) constructs, showing a Pro51 positive control using sdABD-EGFR, a Pro244 positive control using sdABD-hF12 B7H3, a test construct, Pro226, and a negative control with no cleavage site, Pro296. [Figure 49] 1 shows good conditioning of the Pro226 construct in the TDCC assay. [Figure 50] 1 shows humanization of the sdABD for human EpCAM. [Figure 51] Schematics of several formats are shown. Pro22hVIB13 and Pro205 are positive controls, Pro199 is a format 2 construct, and Pro175 is a negative control. [Figure 52] FIG. 1 shows TDCC activity of the sdABD-EpCAM construct, indicating good conditioning. [Diagram 53] AB show the TDCC activity of the sdABD-EpCAM Pro199 construct, which shows good conditioning in HT29 and LoVo cell models. [Figure 54]AB show the TDCC activity of the sdABD-EpCAM Pro200 construct, which shows good conditioning in HT29 and LoVo cell models. [Figure 55] Shown is a schematic of Pro255 using two different sdABD-TTAs, one against EGFR (sdABD-EGFR) and the other against EpCAM (sdABD-EpCAM), compared to Pro199, which has a double EpCAM sdABD. These are sometimes referred to herein as "hetero-targeting" constructs, in this case Format 2 constructs. [Figure 56] 1 shows that a Pro255 dual targeting molecule with an MMP9 cleavage site exhibits good conditionality. [Figure 57A] The results of experiments on three different cell types are shown. First, Raji transfectants with similar expression levels of EpCAM, EGFR and EpCAM+EGFR were generated (data not shown). Next, Pro255, which targets both EpCAM and EGFR, was tested in the TDCC assay using each cell type. The parental Raji line, which does not express either receptor, is shown. [Figure 57B] The results of the experiment on three different cell types are shown. First, Raji transfectants with similar expression levels of EpCAM, EGFR and EpCAM+EGFR were created (data not shown). Next, Pro255, which targets both EpCAM and EGFR, was tested in the TDCC assay using each cell type. Conditioning on the EpCAM line is shown. [Figure 57C] The results of the experiment on three different cell types are shown. First, Raji transfectants with similar expression levels of EpCAM, EGFR and EpCAM+EGFR were created (data not shown). Next, Pro255, which targets both EpCAM and EGFR, was tested in the TDCC assay using each cell type. Conditioning on the EGRF line is shown. [Fig. 57D]The results of the experiment on three different cell types are shown. First, Raji transfectants with similar expression levels of EpCAM, EGFR and EpCAM+EGFR were created (data not shown). Next, Pro255, which targets both EpCAM and EGFR, was tested in the TDCC assay using each cell type. Conditioning on the EpCAM / EGFR line is shown. [Figure 58] 1 shows a schematic diagram of the Format 4 construct, Pro258. [Figure 59] A-B show that Pro258 is conditional in both FBS and human serum. The conditioning of the MMP9 linker is underestimated by MMP9 activity in culture. Interestingly, the TDCC activity of Pro51 is inhibited by HSA binding, whereas the TDCC activity of Pro258 is similar to that of Pro51 in the presence of HSA. Finally, the conditioning of Pro258 is somewhat enhanced in the presence of HSA, up to 6-fold. [Figure 60] AC show cleavage of MMP9 substrates by other MMPs. [Figure 61A] Some exemplary constructs and their formats are shown. [Figure 61B] Some exemplary constructs and their formats are shown. [Figure 62A] Many sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and bold, linkers are double underlined (cleavable linkers are italicized and double underlined), and domain separation is indicated by " / ". All His6 tags are optional, as they can be used to reduce immunogenicity in humans and can also be purification tags. [Figure 62B] Many sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and bold, linkers are double underlined (cleavable linkers are italicized and double underlined), and domain separation is indicated by " / ". All His6 tags are optional, as they can be used to reduce immunogenicity in humans and can also be purification tags. [Figure 62C] Many sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and bold, linkers are double underlined (cleavable linkers are italicized and double underlined), and domain separation is indicated by " / ". All His6 tags are optional, as they can be used to reduce immunogenicity in humans and can also be purification tags. [Fig.62D] Many sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and bold, linkers are double underlined (cleavable linkers are italicized and double underlined), and domain separation is indicated by " / ". All His6 tags are optional, as they can be used to reduce immunogenicity in humans and can also be purification tags. [Figure 62E] Many sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and bold, linkers are double underlined (cleavable linkers are italicized and double underlined), and domain separation is indicated by " / ". All His6 tags are optional, as they can be used to reduce immunogenicity in humans and can also be purification tags. [Fig.62F] Many sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and bold, linkers are double underlined (cleavable linkers are italicized and double underlined), and domain separation is indicated by " / ". All His6 tags are optional, as they can be used to reduce immunogenicity in humans and can also be purification tags. [Fig.62G] Many sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and bold, linkers are double underlined (cleavable linkers are italicized and double underlined), and domain separation is indicated by " / ". All His6 tags are optional, as they can be used to reduce immunogenicity in humans and can also be purification tags. [Fig. 62H] Many sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and bold, linkers are double underlined (cleavable linkers are italicized and double underlined), and domain separation is indicated by " / ". All His6 tags are optional, as they can be used to reduce immunogenicity in humans and can also be purification tags. [Fig.62I] Many sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and bold, linkers are double underlined (cleavable linkers are italicized and double underlined), and domain separation is indicated by " / ". All His6 tags are optional, as they can be used to reduce immunogenicity in humans and can also be purification tags. [Fig.62J] Many sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and bold, linkers are double underlined (cleavable linkers are italicized and double underlined), and domain separation is indicated by " / ". All His6 tags are optional, as they can be used to reduce immunogenicity in humans and can also be purification tags. [Figure 62K] Many sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and bold, linkers are double underlined (cleavable linkers are italicized and double underlined), and domain separation is indicated by " / ". All His6 tags are optional, as they can be used to reduce immunogenicity in humans and can also be purification tags. [Figure 62L] Many sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and bold, linkers are double underlined (cleavable linkers are italicized and double underlined), and domain separation is indicated by " / ". All His6 tags are optional, as they can be used to reduce immunogenicity in humans and can also be purification tags. [Figure 62M] Many sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and bold, linkers are double underlined (cleavable linkers are italicized and double underlined), and domain separation is indicated by " / ". All His6 tags are optional, as they can be used to reduce immunogenicity in humans and can also be purification tags. [Figure 62N] Many sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and bold, linkers are double underlined (cleavable linkers are italicized and double underlined), and domain separation is indicated by " / ". All His6 tags are optional, as they can be used to reduce immunogenicity in humans and can also be purification tags. [Fig.62O] Many sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and bold, linkers are double underlined (cleavable linkers are italicized and double underlined), and domain separation is indicated by " / ". All His6 tags are optional, as they can be used to reduce immunogenicity in humans and can also be purification tags. [Figure 62P] Many sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and bold, linkers are double underlined (cleavable linkers are italicized and double underlined), and domain separation is indicated by " / ". All His6 tags are optional, as they can be used to reduce immunogenicity in humans and can also be purification tags. [Figure 62Q]Many sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and bold, linkers are double underlined (cleavable linkers are italicized and double underlined), and domain separation is indicated by " / ". All His6 tags are optional, as they can be used to reduce immunogenicity in humans and can also be purification tags. [Fig.62R] Many sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and bold, linkers are double underlined (cleavable linkers are italicized and double underlined), and domain separation is indicated by " / ". All His6 tags are optional, as they can be used to reduce immunogenicity in humans and can also be purification tags. [Fig. 62S] Many sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and bold, linkers are double underlined (cleavable linkers are italicized and double underlined), and domain separation is indicated by " / ". All His6 tags are optional, as they can be used to reduce immunogenicity in humans and can also be purification tags. [Fig.62T] Many sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and bold, linkers are double underlined (cleavable linkers are italicized and double underlined), and domain separation is indicated by " / ". All His6 tags are optional, as they can be used to reduce immunogenicity in humans and can also be purification tags. [Fig.62U] Many sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and bold, linkers are double underlined (cleavable linkers are italicized and double underlined), and domain separation is indicated by " / ". All His6 tags are optional, as they can be used to reduce immunogenicity in humans and can also be purification tags. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0031] A. Overview The present invention relates to a method for reducing the toxicity and side effects of bispecific antibodies (including antibody-like functional proteins) that bind to important physiological targets, such as CD3 and tumor antigens. Many antigen-binding proteins, such as antibodies, can exhibit significant off-target side effects, and therefore it is necessary to activate the binding ability of therapeutic molecules only in the vicinity of diseased tissues to avoid off-target interactions. Therefore, the present invention relates to multivalent conditionally effective ("MCE") proteins that have several functional protein domains. Generally, one of these domains is an antigen-binding domain (ABD) that binds to a target tumor antigen (TTA), and another is an ABD that binds to a T-cell antigen, such as CD3, under certain conditions. In addition, MCE proteins also contain one or more protease cleavage sites. This means that the therapeutic molecule is generated in a "pro-drug"-like form, in which the CD3-binding domain is inactive until it is exposed to the tumor environment. Since the tumor environment contains proteases, when exposed to the proteases, the pro-drug is cleaved and activated.

[0032] This is generally accomplished herein by constraining the CD3 Fv of the MCE to an inactive conformation as discussed herein and using a protein that includes a "pseudo" variable heavy domain and a "pseudo" variable light domain that targets a T cell antigen such as CD3. The TTA targets the MCE to the vicinity of the tumor, thereby exposing the MCE to proteases. Upon cleavage, the active variable heavy and light domains can now pair to form one or more active ABDs against CD3, thereby recruiting T cells to the tumor and providing therapy.

[0033] Generally, CD3 binding domains ("Fv") are traditionally in a constrained format in which the linker between the active variable heavy domain and the active variable light domain that form the Fv is too short to allow the two active variable domains to bind to each other. This is called a "constrained linker". These can be constrained and cleavable (CCL, used in format 1) or constrained and non-cleavable (CNCL, used in format 2). Rather, in the prodrug (e.g., uncleaved) format, the prodrug polypeptide also contains a "pseudo-Fv domain". The pseudo-Fv domain contains a variable heavy domain and a variable light domain, with standard framework regions, but the CDRs are "inert" or "inactive". The pseudo-Fv domain also has a constrained linker between the inactive variable heavy domain and the inactive variable light domain. Since neither the Fv domain nor the pseudo-Fv domain can self-assemble due to steric constraints, there is an intramolecular assembly that pairs aVL with iVH and aVH with iVL due to the affinity of the respective framework regions. However, due to the "inactive" CDRs of the pseudo domain, the resulting ABD does not bind to CD3, thus preventing off-target toxicity. However, in the presence of proteases in or near the tumor, the prodrug constructs are cleaved such that the pseudo Fv domains are released from the surface and the "authentic" variable heavy and variable light domains can associate intermolecularly (e.g., two cleaved constructs come together), thereby inducing effective CD3 binding and providing an anti-tumor effect. These constructs are generally referred to herein as conditional bispecific redirection activation constructs or "COBRA™". The stability of the intramolecular assembly is shown by conditional experiments herein, where in the absence of proteases, the uncleaved constructs have no activity (e.g., no active CD3 binding domain is formed).

[0034] Interestingly, while for ease of explanation, all these constructs are referred to herein as "constrained", further studies show that intramolecular assembly is promoted even when one of the Fv domains is not constrained, for example, one of the domains can have a longer, more flexible linker. That is, as shown in Figures 37, 38 and 39, if only one of the Fv domains, either those with active VL and VH or pseudo-Fv domains, is constrained, intramolecular assembly still occurs (e.g., in the absence of protease cleavage, the uncleaved construct is inactive). However, in this system, the protein has better expression when both linkers are constrained. However, as will be understood by those skilled in the art, any of the Format 1, Format 2 or Format 4 constructs herein can have one of these Fv domains with a "non-constrained" or "flexible" linker. For ease of reference, the constructs are shown in a format in which both Fv domains are constrained.

[0035] The constructs and formats of the present invention are variations of the invention described in WO2017 / 156178, which is expressly incorporated herein by reference in its entirety. As shown in Figure 21, the previous construct has the ability to isomerize to form both bivalent scFvs and single-chain diabodies due to the presence of two sets of VH and VL domains in a single polypeptide. Even after purification of each isoform, the bivalent construct can still reach equilibrium with the diabody isoforms. The single-chain diabody has the ability to bind to CD3 in the absence of protease cleavage, reducing the usefulness of the construct.

[0036] To solve this problem, the present invention provides four separate types of constructs to achieve this conditional activation. Prodrug activation can occur in one of four general ways, as generally shown in the figures. In Figure 1, a "Form 1" mechanism is shown. In this embodiment, the prodrug construct has two cleavage sites. One is between the VH and vl domains of the constrained Fv, thus allowing the two variable domains to freely associate, and the second cleavage site is at the site that releases the pseudo-Fv domain from the prodrug construct, leaving two molecules that associate by the inherent self-assembly of the variable heavy and variable light domains, each of which also has an antigen-binding domain against a tumor antigen, thus allowing the recruitment of T cells to the tumor site.

[0037] In an alternative embodiment, the prodrug construct is shown in the "Format 2" mechanism in Figure 2. In this embodiment, the domain linker between the active variable heavy and active light chains is a constrained but non-cleavable linker ("CNCL"). In the prodrug format, the inactive VH and VL of the constrained pseudo-Fv domain associate with the VH and VL of the constrained Fv domain such that there is no CD3 binding. However, once cleavage in the tumor environment occurs, two different activating proteins, each containing the active variable heavy and light domains, associate to form two anti-CD3 binding domains.

[0038] In addition to the "single-chain protein" COBRA format discussed above, in which all components are contained on a single amino acid sequence, there are also constructs that rely on two proteins acting in pairs, "half-COBRAs," as shown in Figure 3. In this embodiment, each protein has one active and one inactive variable domain separated by a protease cleavage site. Each molecule contains a TTA binding domain, such that when the molecule is bound to TTA and exposed to tumor proteases, the inactive domain is cleaved and the two active variable domains self-assemble to form the anti-CD3 binding domain.

[0039] Additionally, the present invention also provides "Format 4" constructs, as shown in Figure 4. These are similar to the "Format 2" design, except that a single ABD is used for the TTA, which upon cleavage forms a tetravalent bispecific construct, where two of the prodrug molecules now contain two active anti-CD3 domains, as further described below.

[0040] Thus, the formats and constructs of the present invention are used in the treatment of diseases.

[0041] B. Definition In order that this application may be more fully understood, certain definitions are set forth below. Such definitions are meant to encompass grammatical equivalents.

[0042] "Amino acid" and "amino acid identity" as used herein refer to one of the 20 naturally occurring amino acids or any non-natural analogs that may be present at a particular given position. In many embodiments, "amino acid" refers to one of the 20 naturally occurring amino acids. "Protein" as used herein refers to at least two covalently attached amino acids, and includes proteins, polypeptides, oligopeptides, and peptides.

[0043] As used herein, "amino acid modification" refers to amino acid substitutions, insertions and / or deletions in a polypeptide sequence, or modifications to moieties chemically bound to a protein. For example, modifications may be altered carbohydrates or PEG structures bound to a protein. For clarity, unless otherwise indicated, amino acid modifications are always to amino acids encoded by DNA, such as the 20 amino acids that have codons in DNA and RNA. The preferred amino acid modification herein is a substitution.

[0044] As used herein, "amino acid substitution" or "substitution" refers to replacing an amino acid at a particular position in a parent polypeptide sequence with a different amino acid. Notably, in some embodiments, the substitution is for an amino acid that does not naturally occur at the particular position, either in an organism or not naturally occurring in any organism. To be clear, a protein that has been engineered to change the nucleic acid coding sequence but not the starting amino acid (e.g., replacing CGG (which codes for arginine) with CGA (which also codes for arginine) to increase expression levels in a host organism) is not an "amino acid substitution." That is, if a new gene is created that codes for the same protein, but the protein has the same amino acid at the starting particular position, it is not an amino acid substitution.

[0045] An "amino acid insertion" or "insertion" as used herein refers to the addition of an amino acid sequence at a particular position in a parent polypeptide sequence.

[0046] An "amino acid deletion" or "deletion" as used herein means removing an amino acid sequence at a particular position in a parent polypeptide sequence.

[0047] The polypeptides of the invention specifically bind to CD3 and target tumor antigens (TTAs), such as target cell receptors as outlined herein. "Specific binding," or "specifically binds to" or "specific for" a particular antigen or epitope, refers to binding that is measurably different from non-specific interactions. Specific binding can be measured, for example, by determining the binding of a molecule compared to the binding of a control molecule, which is typically a molecule of similar structure that has no binding activity. For example, specific binding can be measured by competition with a control molecule that is similar to the target.

[0048] Specific binding to a particular antigen or epitope is, for example, at least about 10 -4 M, at least about 10 -5M, at least about 10 -6 M, at least about 10 -7 M, at least about 10 -8 M, at least about 10 -9 M, or at least about 10 -10 M, at least about 10 -11 M, at least about 10 -12 A specific binding rate of an antibody to an antigen may be exhibited by an antibody having a KD (KD refers to the off-rate of a particular antibody-antigen interaction) of M, M, or greater. Generally, an antibody that specifically binds to an antigen has a KD for the antigen or epitope that is 20-fold, 50-fold, 100-fold, 500-fold, 1000-fold, 5,000-fold, 10,000-fold or greater than that of a control molecule.

[0049] Similarly, specific binding to a particular antigen or epitope can be exhibited by an antibody having a K or K (K or K refers to the association rate of a particular antibody-antigen interaction) for the antigen or epitope that is at least 20-fold, 50-fold, 100-fold, 500-fold, 1000-fold, 5,000-fold, 10,000-fold or more times greater than a control. Binding affinity is typically measured using a Biacore assay or Octet, as is well known in the art.

[0050] "Parent polypeptide" or "precursor polypeptide" (including Fc parent or Fc precursor), as used herein, refers to a polypeptide that is subsequently modified to generate a variant. The parent polypeptide may be a naturally occurring polypeptide, or a variant or modified version of a naturally occurring polypeptide. A parent polypeptide may refer to the polypeptide itself, a composition that includes the parent polypeptide, or the amino acid sequence that encodes the parent polypeptide. Thus, "parent Fc polypeptide", as used herein, refers to an unmodified Fc polypeptide that is modified to generate a variant, and "parent antibody", as used herein, refers to an unmodified antibody that is modified to generate a variant antibody.

[0051] "Position" as used herein means a location within a sequence of a protein. Positions may be numbered sequentially or according to an established system, such as the EU index for antibody numbering.

[0052] "Target antigen," as used herein, refers to a molecule to which a variable region of a given antibody specifically binds. A target antigen may be a protein, carbohydrate, lipid, or other compound. A variety of suitable exemplary target antigens are described herein.

[0053] "Target cell" as used herein means a cell expressing a target antigen. Generally, for the purposes of the present invention, a target cell is either a tumor cell expressing a TTA or a T cell expressing a CD3 antigen.

[0054] "Fv" or "Fv domain" or "Fv region" as used herein generally refers to a polypeptide comprising a VL domain and a VH domain of an antibody binding domain derived from an antibody. Fv domains usually form an "antigen binding domain" or "ABD" as discussed herein when they contain active VH and VL domains (although in some cases, Fvs containing constrained linkers are used, whereby an active ABD is not formed prior to cleavage). As discussed below, Fv domains may be organized in the present invention using several methods and may be "active" or "inactive", such as in scFv format, constrained Fv format, pseudo-Fv format, etc. It should be understood that in some cases, in the present invention, an Fv domain is composed of a VH and VL domain on a single polypeptide chain, as shown in Figures 1 and 2, but with a constrained linker such that an intramolecular ABD cannot be formed. In these embodiments, it is after cleavage that two active ABDs are formed. In some cases, an Fv domain is composed of a VH domain and a VL domain, one of which is inactive, so that an intermolecular ABD is formed only after cleavage. As discussed below, Fv domains can be organized in many ways according to the invention and can be "active" or "inactive", in scFv format, constrained Fv format, pseudo-Fv format, etc. In addition, as discussed herein, Fv domains containing a VH and a VL can be / form an ABD, and other ABDs that do not contain a VH and a VL domain can be formed using sdABDs.

[0055] As used herein, "variable domain" refers to a region of an immunoglobulin that includes one or more Ig domains substantially encoded by any of the Vκ, Vλ and / or VH genes that constitute the kappa, lambda and heavy chain immunoglobulin loci, respectively. In some cases, a single variable domain, such as an sdFv (also referred to herein as sdABD), may be used.

[0056] In embodiments utilizing both variable heavy (VH) and variable light (VL) domains, each VH and VL is composed of three hypervariable regions ("complementarity determining regions", "CDRs") and four "framework regions" or "FRs" arranged from amino-terminus to carboxy-terminus in the following order: FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4. Thus, the VH domain has the structure vhFR1-vhCDR1-vhFR2-vhCDR2-vhFR3-vhCDR3-vhFR4 and the VL domain has the structure vlFR1-vlCDR1-vlFR2-vlCDR2-vlFR3-vlCDR3-vlFR4. As described in more detail herein, the vhFR and vlFR regions self-assemble to form the Fv domain. Generally, in the prodrug format of the invention, there are "constrained Fv domains" in which the VH and VL domains are unable to self-associate, and "pseudo-Fv domains" in which the CDRs do not form an antigen-binding domain when self-associated.

[0057] The hypervariable region confers antigen-binding specificity and generally includes amino acid residues from about amino acid residues 24-34 (LCDR1, "L" represents light chain), 50-56 (LCDR2), and 89-97 (LCDR3) in the light chain variable region, and amino acid residues from about amino acid residues 31-35B (HCDR1, "H" represents heavy chain), 50-65 (HCDR2), and 95-102 (HCDR3) in the heavy chain variable region (Kabat et al., SEQUENCES OF PROTEINS OF IMMUNOLOGICAL INTEREST, 5th Ed. Public Health Service, National Institutes of Health, Bethesda, Md. (1991)), and / or amino acid residues forming hypervariable loops (e.g., residues 26-32 (LCDR1), 50-52 (LCDR2), and 91-96 (LCDR3) in the light chain variable region, and residues 26-32 (HCDR1), 53-55 (HCDR2), and 96-101 (HCDR3) in the heavy chain variable region) (Chothia and Lesk (1987) J. Mol. Biol. 196:901-917). Specific CDRs of the present invention are described below.

[0058] As will be appreciated by those skilled in the art, the exact numbering and arrangement of the CDRs may vary between individual numbering systems. However, it should be understood that the present disclosure of a variable heavy chain sequence and / or a variable light chain sequence includes the present disclosure of the associated (unique) CDRs. Thus, the present disclosure of each variable heavy region is a disclosure of the vhCDRs (e.g., vhCDR1, vhCDR2 and vhCDR3) and the present disclosure of each variable light region is a disclosure of the vlCDRs (e.g., vlCDR1, vlCDR2 and vlCDR3).

[0059] For a useful comparison of CDR numbering, see Lafranc et al., Dev. Comp. Immunol. 27(1):55-77 (2003). [Table 1]

[0060] Throughout this specification, when referring to variable domain residues (approximately residues 1-107 for light chain variable regions and residues 1-113 for heavy chain variable regions), the Kabat numbering system is generally used, with the EU numbering system being used for the Fc region (e.g., Kabat et al., supra (1991)).

[0061] The present invention provides a number of different CDR sets. In this case, a "complete CDR set" in the context of an anti-CD3 construct comprises three variable light chain CDRs and three variable heavy chain CDRs, e.g., vlCDR1, vlCDR2, vlCDR3, vhCDR1, vhCDR2 and vhCDR3. As the skilled artisan will appreciate, each set of CDRs, VH CDRs and VL CDRs, can bind antigen both individually and as a set. For example, in a constrained Fv domain, the vhCDRs can bind, for example, to CD3, and the vlCDRs can bind to CD3, but in a constrained format, the vhCDRs and the vlCDRs cannot bind to CD3.

[0062] In the context of a single domain ABD ("sdABD") as generally used herein to bind a target tumor antigen (TTA), a CDR set is only three CDRs. These are sometimes also referred to in the art as "VHH" domains.

[0063] Each of these CDRs may be part of a larger variable light or variable heavy domain. In addition, as more fully outlined herein, the variable heavy and variable light domains may be on separate polypeptide chains, or, in the case of scFv sequences, on a single polypeptide chain, depending on the format and arrangement of the moieties herein.

[0064] CDRs contribute to the formation of antigen-binding sites, or more specifically epitope-binding sites. "Epitope" refers to a determinant that interacts with a specific antigen-binding site of a variable region, known as a paratope. An epitope is a group of molecules, such as amino acids or sugar side chains, that usually have specific structural characteristics as well as specific charge characteristics. A single antigen may have more than one epitope.

[0065] An epitope may include amino acid residues that are directly involved in binding (also referred to as the major antigenic component of the epitope) and other amino acid residues that are not directly involved in binding, e.g., amino acid residues that are effectively blocked by the specific antigen-binding peptide (i.e., the amino acid residues are within the footprint of the specific antigen-binding peptide).

[0066] Epitopes may be either conformational or linear. Conformational epitopes are generated by spatially ordered amino acids from different segments of a linear polypeptide chain. Linear epitopes are epitopes generated by adjacent amino acid residues in a polypeptide chain. Conformational and non-conformational epitopes can be distinguished in that the binding to the former but not the latter is lost in the presence of denaturing solvents.

[0067] An epitope usually comprises at least 3, more commonly at least 5, or 8-10 amino acids in a unique spatial conformation. Antibodies that recognize the same epitope can be verified by a simple immunoassay, e.g., "binning," showing the ability of one antibody to block the binding of another antibody to its target antigen. As outlined below, the present invention includes not only the antigen-binding domains and antibodies listed herein, but also those that compete for binding to the epitope bound by the listed antigen-binding domains.

[0068] The variable heavy and variable light domains of the present invention may be "active" or "inactive".

[0069] As used herein, "inactive VH" ("iVH") and "inactive VL" ("iVL") refer to components of pseudo-Fv domains that, when paired with their cognate VL or cognate VH partners, respectively, form a resulting VH / VL pair that does not specifically bind to the antigen bound by the "active" VH or "active" VL, but is not "inactive" when bound to a similar VL or similar VH. Exemplary "inactive VH" and "inactive VL" domains are formed by mutation of the wild-type VH or VL sequence, as more fully outlined below. Exemplary mutations are within CDR1, CDR2 or CDR3 of the VH or VL. Exemplary mutations include placing a domain linker within CDR2 to form an "inactive VH" or "inactive VL" domain. In contrast, an "active VH" or "active VL" is a VH or VL that can specifically bind to its target antigen when paired with its "active" cognate partner, i.e., a VL or VH, respectively. It should therefore be understood that a pseudo-Fv can be a VH / iVL pair, an iVH / VL pair, or an iVH / iVL pair.

[0070] In contrast, as used herein, the term "active" refers to a CD3 binding domain capable of specifically binding to CD3. This term is used in two contexts: (a) when referring to one member of an Fv binding pair (i.e., VH or VL), the sequence of which can be paired with its cognate partner to specifically bind to CD3, and (b) a cognate pair of sequences (i.e., VH and VL) that can specifically bind to CD3. Exemplary "active" VH, VL, or VH / VL pairs are wild-type or parental sequences.

[0071] "CD-x" refers to a Cluster of Differentiation (CD) protein. In an exemplary embodiment, CD-x is selected from those CD proteins that have a role in recruiting or activating T cells in a subject administered the polypeptide construct of the invention. In an exemplary embodiment, CD-x is CD3, the sequence of which is shown in FIG. 5.

[0072] The term "binding domain" in the context of the present invention is characterized as a domain that (specifically) binds to / interacts with / recognizes a given target epitope or a given target site on a target molecule (antigen), e.g., EGFR and CD3, respectively. The structure and function of the target antigen binding domain (that recognizes EGFR) and preferably also the structure and / or function of the CD3 binding domain (that recognizes CD3) are based on the structure and / or function of an antibody, e.g., a full-length immunoglobulin molecule or the entire immunoglobulin molecule, comprising the sdABD. According to the present invention, the target antigen binding domain is usually characterized by the presence of three CDRs that bind to the target tumor antigen (commonly referred to in the art as variable heavy domains, although there are no corresponding light chain CDRs). Alternatively, the ABD for TTA may comprise three light chain CDRs (i.e., CDR1, CDR2 and CDR3 of the VL region) and / or three heavy chain CDRs (i.e., CDR1, CDR2 and CDR3 of the VH region). Preferably, the CD3 binding domain also comprises at least the minimal structural requirements of an antibody to enable target binding. More preferably, the CD3 binding domain comprises at least three light chain CDRs (i.e., CDR1, CDR2 and CDR3 of the VL region) and / or three heavy chain CDRs (i.e., CDR1, CDR2 and CDR3 of the VH region). In exemplary embodiments, it is envisaged that the target antigen and / or CD3 binding domain can be generated or obtained by phage display or library screening methods.

[0073] "Domain" as used herein means a protein sequence having a function as outlined herein. Domains of the present invention include tumor target antigen binding domains (TTA domains), variable heavy domains, variable light domains, linker domains, and half-life extending domains.

[0074] By "domain linker" herein is meant an amino acid sequence that connects two domains as outlined herein. A domain linker may be a cleavable linker, a constrained cleavable linker, a non-cleavable linker, a constrained non-cleavable linker, an scFv linker, etc.

[0075] By "cleavable linker" ("CL") herein is meant an amino acid sequence that can be cleaved by a protease, preferably a human protease, in disease tissue as outlined herein. Cleavable linkers are usually at least 3 amino acids in length, and depending on the flexibility required, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or more amino acids are useful in the present invention. A number of cleavable linker sequences can be found in Figures 6 and 5.

[0076] As used herein, a "non-cleavable linker" ("NCL") refers to an amino acid sequence that cannot be cleaved by human proteases under normal physiological conditions.

[0077] As used herein, a "constrained cleavable linker" ("CCL") refers to a short polypeptide containing a protease cleavage site (as defined herein) that links two domains as outlined herein in such a way that the two domains cannot significantly interact with each other until after the two domains are present on different polypeptide chains, e.g., after cleavage. When a CCL links a VH domain and a VL domain as defined herein, the VH and VL cannot self-assemble to form a functional Fv prior to cleavage due to intramolecular steric hindrance (although they can intermolecularly assemble into a pseudo-Fv domain). Upon cleavage by an appropriate protease, the VH and VL can intermolecularly associate to form an active antigen-binding domain. Typically, CCLs are less than 10 amino acids long, with 9, 8, 7, 6, 5 and 4 amino acids being useful in the present invention. In general, protease cleavage sites are usually at least 4+ amino acids long to confer sufficient specificity, as shown in FIG. 6.

[0078] As used herein, a "constrained non-cleavable linker" ("CNCL") refers to a short polypeptide that links two domains as outlined herein in such a way that the two domains cannot significantly interact with each other and are not significantly cleaved by human proteases under physiological conditions.

[0079] By "constrained Fv domain" herein is meant an Fv domain comprising an active variable heavy domain and an active variable light domain covalently linked to a constraining linker as outlined herein in such a manner that the active heavy and light variable domains cannot interact intramolecularly to form an active Fv that binds an antigen, such as CD3. Constrained Fv domains are therefore domains which are similar to scFvs but which are unable to bind antigen due to the presence of the constraining linker (although they can assemble intermolecularly with inactive variable domains to form pseudo-Fv domains).

[0080] As used herein, "pseudo-Fv domain" refers to a domain that includes a pseudo or inactive variable heavy domain or a pseudo or inactive variable light domain, or both, linked with a domain linker (which may be cleavable, constrained, non-cleavable, non-constrained, etc.). The iVH and iVL domains of the pseudo-Fv domain do not bind human antigens, either when associated with each other (iVH / iVL) or when associated with an active VH or active VL. Thus, the iVH / iVL Fv domain, the iVH / VL Fv domain, and the iVL / VH Fv domain do not measurably bind human proteins, and as a result, these domains are inactive in the human body.

[0081] As used herein, "single chain Fv" or "scFv" refers to a variable heavy (VH) domain covalently linked to a variable light (VL) domain, generally using a domain linker as discussed herein, to form an scFv or scFv domain. The scFv domain can be in either N-terminal to C-terminal orientation (VH-linker-VL or VL-linker-VH).

[0082] As used herein, "single domain Fv", "sdFv" or "sdABD" refers to an antigen-binding domain having only three CDRs, generally based on camelid antibody technology. See Protein Engineering 9(7):1129-35(1994); Rev Mol Biotech 74:277-302(2001); Ann Rev Biochem 82:775-97(2013). As outlined herein, two general types of sdABDs are used herein: sdABDs that bind TTA and are annotated as such (commonly referred to as sdABD-TTA, or sdABD-EGFR for those that bind EGFR, sdABD-FOLR1 for those that bind FOLR1, etc.), and sdABDs that bind HSA ("sdABD-HSA" or "sdABD(1 / 2)").

[0083] "Protease cleavage site" refers to an amino acid sequence that a protease recognizes and cleaves. Suitable protease cleavage sites are outlined below and shown in Figures 5 and 6.

[0084] As used herein, a "protease cleavage domain" refers to a "protease cleavage site" and any linkers between individual protease cleavage sites and between the protease cleavage site(s), as well as other functional components of the constructs of the invention (e.g., V H , V L By protease cleavage domain is meant a peptide sequence incorporating a protease cleavage domain (e.g., iVH, iVL, target antigen binding domain(s), half-life extending domain, etc.). As outlined herein, the protease cleavage domain may also include additional amino acids as needed, for example to provide flexibility.

[0085] The terms "COBRA™" and "conditional bispecific redirected activation" refer to a bispecific conditionally enabled protein having several functional protein domains. In some embodiments, one of the functional domains is an antigen binding domain (ABD) that binds to a target tumor antigen (TTA). In certain embodiments, another domain is an ABD that binds to a T cell antigen under certain conditions. T cell antigens include, but are not limited to, CD3. The term "half COBRA™" refers to a conditionally enabled protein that can bind to a T cell antigen when its unique self-assembly upon aggregation on the surface of a target-expressing cell allows the variable heavy chain of the half COBRA to associate with the variable light chain of another half COBRA™ (complementary half COBRA™).

[0086] VII. Fusion Proteins of the Invention The fusion proteins of the present invention have several different components, generally referred to herein as domains, linked together in various ways. Some of the domains are binding domains, each binding to a target antigen (e.g., TTA or CD3). When the binding domains bind to more than one antigen, they are referred to herein as "multispecific", for example, the prodrug construct of the present invention can bind to TTA and CD3, and is therefore "bispecific". Proteins can also have higher specificity, for example, if a first αTTA binds to EGFR, a second αTTA binds to EpCAM, and there is an anti-CD3 binding domain, it is a "trispecific" molecule. Similarly, the addition of an anti-HSA binding domain to this construct is "tetraspecific", as shown in Figure 3B.

[0087] As will be appreciated by those skilled in the art, the proteins of the invention can have different valencies and can be multispecific, i.e., they can bind to a target by more than one binding site, e.g., Pro140 is bivalent for EGFR.

[0088] The proteins of the invention may comprise CD3 antigen binding domains, tumor targeting antigen binding domains, half-life extending domains, linkers, etc. arranged in various manners as outlined herein.

[0089] A. CD3 antigen-binding domain The specificity of the T cell response involves the recognition of antigens (presented in the context of the major histocompatibility complex, MHC) by the T cell receptor complex. CD3, as part of the T cell receptor complex, is a protein complex present on the cell surface that includes the CD3γ (gamma) chain, the CD3δ (delta) chain, two CD3e (epsilon) chains, and two CD3ζ (zeta) chains. The CD3 molecule associates with the α (alpha) and β (beta) chains of the T cell receptor (TCR) to comprise the TCR complex. Clustering of CD3 on the T cell, such as by Fv domains that bind to CD3, results in T cell activation, similar to T cell receptor binding, but independent of the specificity typical of that clone.

[0090] However, as is well known in the art, activation of CD3 leads to many toxic side effects, and therefore the present invention relates to providing effective CD3 binding of the polypeptides of the present invention only in the presence of tumor cells where the presence of a specific protease cleaves the prodrug polypeptide of the present invention to provide the active CD3 binding domain. Thus, in the present invention, binding of the anti-CD3 Fv domain to CD3 is controlled by a protease cleavage domain that restricts binding of the CD3 Fv domain to only those microenvironments of diseased cells or tissues with high protease levels, such as the tumor microenvironment described herein.

[0091] Thus, the present invention provides two sets of VH and VL domains, an active set (VH and VL) and an inactive set (iVH and iVL), all four of which are present in the prodrug construct. The construct is formatted such that the VH and VL sets cannot self-associate, but rather associate with an inactive partner, e.g., iVH and VL, and iVL and VH, as shown herein.

[0092] 1. Active anti-CD3 variable heavy and variable light domains There are many suitable active CDR sets and / or VH and VL domains known in the art that may be found to be useful in the present invention. For example, the CDRs and / or VH and VL domains may be derived from known anti-CD3 antibodies, such as muromonab-CD3 (OKT3), otelixizumab (TRX4), teplizumab (MGA031), visilizumab (Nuvion), SP34 or I2C, TR-66 or X35-3, VIT3, BMA030 (BW264 / 56), CLB-T3 / 3, C These include RIS7, YTH12.5, F111-409, CLB-T3.4.2, TR-66, WT32, SPv-T3b, 11D8, XIII-141, XIII-46, XIII-87, 12F6, T3 / RW2-8C8, T3 / RW2-4B6, OKT3D, M-T301, SMC2, F101.01, UCHT-1, and WT-31.

[0093] In one embodiment, the VH and VL sequences that form an active Fv domain that binds to human CD3 are shown in Figure 5. As shown herein, these active VH ("aVH") and active VL ("aVL") domains can be used in various configurations and formats 1, 2, 3, and 4.

[0094] 2. Inactive anti-CD3 variable heavy and variable light domains The inactive iVH domain and the inactive iVL domain contain "normal" framework regions (FRs) that enable them to associate, such that the inactive variable domain associates with the active variable domain to render the pair inactive, e.g., unable to bind CD3.

[0095] As will be appreciated by those skilled in the art, there are several "inactive" variable domains that may be found to be useful in the present invention. Essentially, any variable domain may be used that has human framework regions that allow self-assembly with another variable domain, no matter what amino acids are in the CDR positions of the variable domain. For the sake of clarity, the inactive variable domain is said to contain the CDRs, even though strictly speaking the inactive variable domain does not confer binding capacity.

[0096] As understood in the art, generating an inactive VH or VL domain is generally straightforward and can be done in a variety of ways. In some embodiments, the generation of an inactive variable domain is generally done by modifying one or more of the CDRs of an active Fv, including changing one or more of the three CDRs of the active variable domain. This can be done by making one or more amino acid substitutions at functionally important residues of one or more CDRs, replacing some or all CDR residues with random sequences, replacing one or more CDRs with tag or flag sequences, and / or exchanging the CDRs and / or variable regions with those from an unrelated antibody (e.g., from a protein of a different organism).

[0097] In some cases, only one of the CDRs in the variable region may be altered, rendering it inactive, although other embodiments include alterations in 1, 2, 3, 4, 5 or 6 CDRs.

[0098] In some cases, inactive domains can be engineered to promote selective binding in a prodrug format to favor intramolecular iVH-VL and VH-iVL domain formation (e.g., rather than intermolecular pairing) prior to cleavage. See, e.g., Igawa et al., Protein Eng. Des. Selection 23(8):667-677 (2010), expressly incorporated herein by reference in its entirety, particularly for interface residue amino acid substitutions.

[0099] In certain embodiments, the CD3 binding domain of the polypeptide construct described herein not only exhibits strong CD3 binding affinity to human CD3, but also exhibits good cross-reactivity with the corresponding cynomolgus monkey CD3 protein. In some cases, the CD3 binding domain of the polypeptide construct cross-reacts with CD3 from cynomolgus monkey. In certain examples, the human:cynomolgus monkey KD ratio for CD3 is between 5 and 0.2.

[0100] In some embodiments, the CD3 binding domain of the antigen binding protein may be any domain that binds to CD3, including but not limited to domains derived from monoclonal, polyclonal, recombinant, human, or humanized antibodies.In some cases, it is beneficial for the CD3 binding domain to be derived from the same species in which the antigen binding protein will ultimately be used.For example, when used in humans, it may be advantageous to include human or humanized residues derived from the antigen binding domain of an antibody or antibody fragment in the CD3 binding domain of the antigen binding protein.

[0101] Thus, in one aspect, the antigen binding domain comprises a humanized or human binding domain. In one embodiment, the humanized or human anti-CD3 binding domain comprises one or more (e.g., all three) of the light chain complementarity determining region 1 (LC CDR1), light chain complementarity determining region 2 (LC CDR2) and light chain complementarity determining region 3 (LC CDR3) of a humanized or human anti-CD3 binding domain described herein, and / or one or more (e.g., all three) of the heavy chain complementarity determining region 1 (HC CDR1), heavy chain complementarity determining region 2 (HC CDR2) and heavy chain complementarity determining region 3 (HC CDR3) of a humanized or human anti-CD3 binding domain described herein, e.g., the humanized or human anti-CD3 binding domain comprises one or more, e.g., all three LC CDRs, and one or more, e.g., all three HC CDRs.

[0102] In some embodiments, the humanized or human anti-CD3 binding domain comprises a humanized or human light chain variable region specific for CD3, which comprises a human or non-human light chain CDR within a human light chain framework region. In a particular example, the light chain framework region is a λ (lambda) light chain framework. In another example, the light chain framework region is a κ (kappa) light chain framework.

[0103] In some embodiments, one or more CD3 binding domains are humanized or fully human. In some embodiments, one or more activating CD3 binding domains have a KD binding of 1000 nM or less to CD3 on CD3-expressing cells. In some embodiments, one or more activating CD3 binding domains have a KD binding of 100 nM or less to CD3 on CD3-expressing cells. In some embodiments, one or more activating CD3 binding domains have a KD binding of 10 nM or less to CD3 on CD3-expressing cells. In some embodiments, one or more CD3 binding domains have cross-reactivity with cynomolgus CD3. In some embodiments, one or more CD3 binding domains comprise an amino acid sequence provided herein.

[0104] In some embodiments, the humanized or human anti-CD3 binding domain comprises a humanized or human heavy chain variable region specific for CD3, which heavy chain variable region specific for CD3 comprises human or non-human heavy chain CDRs within human heavy chain framework regions.

[0105] In one embodiment, the anti-CD3 binding domain is an Fv comprising a light chain and a heavy chain of an amino acid sequence provided herein. In one embodiment, the anti-CD3 binding domain comprises a light chain variable region comprising an amino acid sequence having at least one, two or three modifications (e.g., substitutions) but not more than 30, 20 or 10 modifications (e.g., substitutions) of the amino acid sequence of the light chain variable region provided herein, or a sequence having 95-99% identity to the amino acid sequence provided herein, and / or a heavy chain variable region comprising an amino acid sequence having at least one, two or three modifications (e.g., substitutions) but not more than 30, 20 or 10 modifications (e.g., substitutions) of the amino acid sequence of the heavy chain variable region provided herein, or a sequence having 95-99% identity to the amino acid sequence provided herein. In one embodiment, the humanized anti-CD3 binding domain or human anti-CD3 binding domain is an scFv, and a light chain variable region comprising an amino acid sequence as provided herein is linked to a heavy chain variable region comprising an amino acid sequence as provided herein via an scFv linker. The light and heavy chain variable regions of an scFv may be, for example, in one of the following orientations: light chain variable region-scFv linker-heavy chain variable region or heavy chain variable region-scFv linker-light chain variable region.

[0106] In some embodiments, the CD3 binding domain of the antigen binding protein has an affinity for CD3 on CD3 expressing cells with a KD of 1000 nM or less, 100 nM or less, 50 nM or less, 20 nM or less, 10 nM or less, 5 nM or less, 1 nM or less, or 0.5 nM or less. In some embodiments, the CD3 binding domain of the antigen binding protein has an affinity for CD3ε with a KD of 1000 nM or less, 100 nM or less, 50 nM or less, 20 nM or less, 10 nM or less, 5 nM or less, 1 nM or less, or 0.5 nM or less. In further embodiments, the CD3 binding domain of the antigen binding protein has a low affinity for CD3, i.e., an affinity of about 100 nM or more.

[0107] For example, the binding affinity to CD3 may be measured by the ability of the antigen binding protein itself or its CD3 binding domain to CD3, such as coated on an assay plate, displayed on a bacterial cell surface, in solution, etc., as is well known in the art (typically using a Biacore assay or an Octet assay). The antigen binding protein of the present disclosure, or its CD3 binding domain, may be assayed for binding activity to CD3 by immobilizing the ligand (e.g., CD3) or the antigen binding protein itself or its CD3 binding domain to beads, substrates, cells, etc. The agent may be added to the appropriate buffer and binding partner and incubated for a period of time at a given temperature. After washing to remove unbound material, the bound protein may be stripped, for example, using SDS, high pH buffer, etc., and analyzed, for example, by surface plasmon resonance (SPR).

[0108] In many embodiments, suitable active and inactive binding domains are those shown in FIG.

[0109] B. Antigen-binding domains for tumor target antigens In addition to the CD3 and half-life extending domains described herein, the polypeptide constructs described herein also include targeting domains that bind to one or more target antigens, or one or more regions on a single target antigen. For example, it is envisioned that the polypeptide constructs of the present invention will be cleaved by the protease cleavage domain to activate the CD3 binding domains that bind to T cells by binding the respective target antigen binding domains to the target antigen on the target cell in the disease-specific microenvironment or blood of the subject. Typically, the TTA binding domains can bind to their targets before protease cleavage, so that they can "wait" on the target cell to be activated as T cell engagers. At least one target antigen is involved and / or associated with a disease, disorder, or condition. Exemplary target antigens include target antigens associated with a proliferative disease, a neoplastic disease, an inflammatory disease, an immunological disorder, an autoimmune disease, an infectious disease, a viral disease, an allergic response, a parasitic response, a graft versus host disease, or a host versus graft disease. In some embodiments, the target antigen is a tumor antigen expressed on a tumor cell. Alternatively, in some embodiments, the target antigen is associated with a pathogen, such as a virus or bacteria. At least one target antigen may also be associated with healthy tissue.

[0110] In some embodiments, the target antigen is a cell surface molecule, such as a protein, lipid, or polysaccharide, etc. In some embodiments, the target antigen is on a tumor cell, a virus-infected cell, a bacteria-infected cell, a damaged red blood cell, an arterial plaque cell, or a fibrous tissue cell.

[0111] Preferred embodiments of the invention utilize sdABDs as targeting domains. These are preferred over scFv ABDs because the addition of other VH and VL domains to the constructs of the invention can complicate the formation of pseudo-Fv domains.

[0112] In some embodiments, the prodrug constructs of the invention utilize a single TTA binding domain, generally as shown in Figure 3A as the sdABD-TTA pair and in Figure 4 as the "Format 4" configuration. Although Figure 4 shows the use of a single anti-EGFR ABD, other TTA binding domains may be used.

[0113] In some embodiments, particularly in the format 1 and format 2 constructs, the prodrug construct of the present invention utilizes two TTA ABDs, again preferably in sdABD-TTA format. When dual targeting domains are used, they can bind to the same epitope of the same TTA. For example, as discussed herein, many of the constructs herein utilize two identical targeting domains. In some embodiments, two targeting domains that bind to different epitopes of the same TTA can be used, for example, as shown in Figure 5, two EGFR sdABDs bind to different epitopes on human EGFR. In some embodiments, the two targeting domains bind to different TTAs. See, for example, Figure 55.

[0114] The polypeptide constructs contemplated herein comprise at least one antigen binding domain, which binds to at least one target antigen. In some embodiments, the target antigen binding domain specifically binds to a cell surface molecule. In some embodiments, the target antigen binding domain specifically binds to a tumor antigen. In some embodiments, the target antigen binding domain specifically and independently binds to a tumor target antigen ("TTA") selected from at least one of EpCAM, EGFR, HER-2, HER-3, cMet, LyPD3, B7H3, CEA, and FOLR1.

[0115] Of particular use in the present invention is the sdABD for human EGFR shown in FIG.

[0116] A further embodiment for use in the present invention is the sdABD for human FOLR1 shown in FIG.

[0117] Another embodiment for use in the present invention is the sdABD against human B7H3 shown in FIG.

[0118] A further embodiment for use in the present invention is the sdABD against human EpCAM shown in FIG.

[0119] In some embodiments, the protein before cleavage of the protease cleavage domain is less than about 100 kDa. In some embodiments, the protein after cleavage of the protease cleavage domain is about 25 to about 75 kDa. In some embodiments, the protein before protease cleavage has a size that exceeds the renal threshold for first-pass clearance. In some embodiments, the protein before protease cleavage has an elimination half-life of at least about 50 hours. In some embodiments, the protein before protease cleavage has an elimination half-life of at least about 100 hours. In some embodiments, the protein has increased tissue penetration compared to IgG against the same target antigen. In some embodiments, the protein has increased tissue distribution compared to IgG against the same target antigen.

[0120] C. Half-life extending domain The MCE proteins of the present invention (again, also referred to herein as "COBRA™" proteins or constructs) optionally comprise a half-life prolonging domain. Such domains are contemplated to include, but are not limited to, HSA binding domains, Fc domains, small molecules, and other half-life prolonging domains known in the art.

[0121] Human serum albumin (HSA) (molecular weight approximately 67 kDa) is the most abundant protein in plasma, present at approximately 50 mg / ml (600 uM) and with a half-life of approximately 20 days in humans. HSA is responsible for maintaining plasma pH, contributes to oncotic blood pressure, functions as a carrier for many metabolites and fatty acids, and serves as the major drug transport protein in plasma.

[0122] Non-covalent association with albumin extends the elimination half-life of short-lived proteins. For example, recombinant fusion of albumin-binding domains with Fab fragments reduced in vivo clearance by 25-fold and 58-fold, and extended half-life by 26-fold and 37-fold when administered intravenously to mice and rabbits, respectively, compared with administration of the Fab fragment alone. In another example, when insulin is acylated with fatty acids to promote association with albumin, a sustained effect was observed when injected subcutaneously into rabbits or pigs. Taken together, these studies show a link between albumin binding and sustained action.

[0123] In one aspect, the antigen binding protein described herein comprises a half-life extending domain, e.g., a domain that specifically binds to HSA. In other embodiments, the HSA binding domain is a peptide. In further embodiments, the HSA binding domain is a small molecule. It is contemplated that the HSA binding domain of the antigen binding protein is quite small in some embodiments, 25 kD or less, 20 kD or less, 15 kD or less, or 10 kD or less. In certain examples, the HSA binding domain is 5 kD or less when it is a peptide or a small molecule.

[0124] In many embodiments, the half-life extending domain is a single domain antigen binding domain from a single domain antibody that binds to HSA. This domain is generally referred to herein as "sdABD" for human HSA (sdABD-HSA), or alternatively as "sdABD(1 / 2)" to distinguish these binding domains from sdABD for TTA. A particularly useful sdABD(1 / 2) is shown in FIG. 5.

[0125] The half-life prolonging domain of an antigen-binding protein results in changes in the pharmacodynamics and pharmacokinetics of the antigen-binding protein itself. As described above, the half-life prolonging domain extends the elimination half-life. The half-life prolonging domain also modifies the pharmacological properties of the antigen-binding protein, including altering tissue distribution, penetration and diffusion. In some embodiments, the half-life prolonging domain provides improved tissue (including tumor) targeting, tissue penetration, tissue distribution, diffusion within tissues and enhanced efficacy compared to the protein without the half-life prolonging binding domain. In one embodiment, the therapeutic method effectively and efficiently utilizes a reduced amount of the antigen-binding protein, resulting in reduced side effects, such as reduced cytotoxicity of non-tumor cells.

[0126] Further characteristics of the half-life extending domain, e.g., HSA binding domain, include the binding affinity of the HSA binding domain to HSA. The affinity of the HSA binding domain may be selected to target a particular elimination half-life of a particular polypeptide construct. Thus, in some embodiments, the HSA binding domain has a high binding affinity. In other embodiments, the HSA binding domain has a moderate binding affinity. In still other embodiments, the HSA binding domain has a low or no binding affinity. Exemplary binding affinities include KD concentrations of 10 nM or less (high), 10 nM to 100 nM (medium), and more than 100 nM (low). As described above, the binding affinity to HSA is determined by known methods, such as surface plasmon resonance (SPR).

[0127] D. Protease Cleavage Sites The protein compositions of the present invention, particularly the prodrug constructs, contain one or more protease cleavage sites, typically located within a cleavable linker, as outlined herein.

[0128] As described herein, the prodrug constructs of the present invention contain at least one protease cleavage site that comprises an amino acid sequence that is cleaved by at least one protease. In some cases, the MCE protein described herein comprises 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 or more protease cleavage sites that are cleaved by at least one protease. As discussed more fully herein, when two or more protease cleavage sites are used in a prodrug construct, they may be the same (e.g., multiple sites cleaved by a single protease) or different (two or more cleavage sites cleaved by at least two different proteases). As will be understood by those skilled in the art, constructs containing three or more protease cleavage sites may utilize one, two, three, etc., e.g., some constructs may utilize three sites for two different proteases, etc.

[0129] The amino acid sequence of the protease cleavage site depends on the protease being targeted. As is well known in the art, there are many human proteases that can be found in the body and potentially implicated in disease states.

[0130] Proteases are known to be secreted by some diseased cells and diseased tissues, such as tumor or cancer cells, creating a protease-rich microenvironment. In some cases, the blood of a subject is rich in proteases. In some cases, cells surrounding a tumor secrete proteases into the tumor microenvironment. Cells surrounding a tumor that secrete proteases include, but are not limited to, neoplastic stromal cells, myofibroblasts, blood cells, mast cells, B cells, NK cells, regulatory T cells, macrophages, cytotoxic T lymphocytes, dendritic cells, mesenchymal stem cells, polymorphonuclear cells, and other cells. In some cases, proteases, such as proteases that target amino acid sequences found in bacterial peptides, are present in the blood of a subject. This property allows for additional specificity for targeted therapeutics, such as antigen binding proteins, since antigen binding proteins do not bind to T cells except in the protease-rich microenvironment of the target cells or tissues.

[0131] Proteases are proteins that cleave proteins, optionally in a sequence-specific manner. Proteases include serine proteases, cysteine ​​proteases, aspartic acid proteases, threonine proteases, glutamic acid proteases, metalloproteases, asparagine peptide lyases, serum proteases, cathepsins (e.g., cathepsin B, cathepsin C, cathepsin D, cathepsin E, cathepsin K, cathepsin L, cathepsin S), kallikrein, hK1, hK10, hK15, KLK7, granzyme B, plasmin, collagenase, type IV collagenase, stromelysin, factor XA, chymotrypsin-like proteases, trypsin-like proteases, elastase-like proteases, subtilisin-like proteases, actinidain, bromelain, calpain, and caspases (e.g., caspase 3). ), Mir1-CP, papain, HIV-1 protease, HSV protease, CMV protease, chymosin, renin, pepsin, matriptase, legumain, plasmepsin, nepenthesin, metalloexopeptidase, metalloendopeptidase, matrix metalloprotease (MMP), MMP1, MMP2, MMP3, MMP8, MMP9, MMP13, MMP11, MMP14, meprin, urokinase plasminogen activator (uPA), enterokinase, prostate specific antigen (PSA, hK3), interleukin-1β converting enzyme, thrombin, FAP (FAP-α), dipeptidyl peptidase, and dipeptidyl peptidase IV (DPPIV / CD26).

[0132] Some suitable proteases and protease cleavage sequences are shown in FIG.

[0133] E. Linker As discussed herein, the different domains of the invention are generally linked together using amino acid linkers, which can confer functionality including flexibility or inflexibility (e.g., steric constraints), as well as the ability to be cleaved using a protease at that site. These linkers can be classified in a number of ways.

[0134] The present invention provides "domain linkers" that are used to link two or more domains (e.g., VH and VL, VH or VL and a target tumor antigen binding domain (TTABD, sometimes referred to herein as "αTTA" (instead of "anti-TTA")), another component and a half-life extending domain, etc.). Domain linkers can be, for example, non-cleavable (NCL), cleavable ("CL"), constrained and cleavable (CCL), and constrained and non-cleavable (CNCL).

[0135] 1. Non-cleavable linker In some embodiments, the domain linker is non-cleavable. In general, they can be one of two types: non-cleavable and flexible, which allows the components of the construct "upstream" and "downstream" of the linker to self-assemble intramolecularly in a specific way, or non-cleavable and restrictive, which means that the two components separated by the linker cannot self-assemble intramolecularly. However, it should be noted that in the latter case, the two component domains separated by the non-cleavable restrictive linker do not self-assemble intramolecularly, while other intramolecular components self-assemble to form pseudo-Fv domains.

[0136] (i) A non-cleavable but flexible linker In this embodiment, a linker is used to link the domains to maintain their functionality by a generally longer flexible domain that is not cleaved by the protease at that site in the patient. Examples of internal non-cleavable linkers suitable for linking domains in the polypeptides of the invention include, but are not limited to, (GS)n, (GGS)n, (GGGS)n, (GGSG)n, (GGSGG)n, or (GGGGS)n, where n is 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10. In some embodiments, the length of the linker may be about 15 amino acids.

[0137] (ii) Non-cleavable and constrained linkers In some cases, the linker is a "constrained non-cleavable linker" or "CNCL" that does not contain a cleavage site and is too short to allow the protein domains separated by the linker to self-assemble intramolecularly. For example, in Pro186, the active VH and active VL are separated by eight amino acids ("8mer") that do not allow the VH and VL to self-assemble into an active antigen-binding domain. In some embodiments, the linker is still flexible, e.g., (GGGS)n (n=2). In other embodiments, more rigid linkers such as those containing proline or bulky amino acids may be used, although they are generally less preferred.

[0138] 2. Cleavable Linker All of the prodrug constructs herein include at least one cleavable linker. Thus, in one embodiment, the domain linker is cleavable (CL), sometimes referred to herein as a "protease cleavage domain" ("PCD"). In this embodiment, the CL includes a protease cleavage site, as outlined herein and shown in Figures 5 and 6. In some cases, the CL includes only a protease cleavage site. Optionally, depending on the length of the cleavage recognition site, there may be several more linking amino acids at either or both the N-terminus or C-terminus of the CL. For example, there may be 1, 2, 3, 4, or 5 amino acids at either or both the N-terminus and C-terminus of the cleavage site. Thus, the cleavable linker may also be constrained (e.g., 8-mer) or flexible.

[0139] Of particular interest in the present invention are MMP9 cleavable linkers and meprin cleavable linkers, in particular MMP9 restricted cleavable linkers and meprin restricted cleavable linkers.

[0140] VIII. Domains of the Invention The present invention provides many different formats of the prodrug polypeptides of the present invention. The present invention provides constrained Fv domains and constrained pseudo-Fv domains. Additionally, the present invention provides multivalent conditionally enabled ("MCE") proteins that contain two Fv domains but are non-isomerizing constructs. As outlined herein, these can be non-isomerizing cleavable or non-isomerizing non-cleavable formats, but all constructs contain at least one protease cleavage domain.

[0141] Importantly, both of these domains (the Fv domain and the pseudo-Fv domain) are referred to herein as "constrained", meaning that only one of them needs to be constrained, as discussed above and shown in Figures 37, 38 and 39, but generally the protein has better expression when both linkers are constrained.

[0142] Those skilled in the art will understand that for formats 1, 2 and 4, there are four possibilities for the N- to C-terminal order of the constraints and pseudo-Fv domains of the present invention (linkers not shown): aVH-aVL and iVL-iVH, aVH-aVL and iVH-iVL, aVL-aVH and iVL-iVH, aVL-aVH and iVH-iVL. All four have been tested, and all four have activity, but the first order, aVH-aVL and iVL-iVH, shows better expression than the other three. Thus, although the description herein is generally presented in this aVH-aVL and iVL-iVH format, all disclosures herein include other orders of these domains.

[0143] It should be noted that, in general, the N-terminal to C-terminal order of the full-length constructs of the present invention is based on the aVH-aVL and iVL-iVH orientation.

[0144] In addition, it is known in the art that there may be immunogenicity in humans derived from the C-terminal sequence of a particular ABD. Therefore, in general, a histidine tag (either His6 or His10) can be used, especially when the C-terminus of the construct ends with a sdABD (e.g., the sdABD-HSA domain of many of the constructs). Many or most of the sequences herein were generated using a His6 C-terminal tag for purification reasons, but these sequences can also be used to reduce immunogenicity in humans, as shown by Holland et al., DOI 10.1007 / s10875-013-9915-0 and WO2013 / 024059.

[0145] A. Constrained Fv Domains The present invention provides constrained Fv domains, which comprise active VH and active VL domains covalently linked using a constrained linker (which can be cleavable (formats 1 and 3) or non-cleavable (formats 2 and 4) as outlined herein). The constrained linker prevents intramolecular association between aVH and aVL in the absence of cleavage. Thus, a constrained Fv domain generally comprises a set of six CDRs contained within the variable domains, where vhCDR1, vhCDR2 and vhCDR3 of VH bind human CD3, and vlCDR1, vCDR2 and vlCDR3 of VL bind human CD3, but in a prodrug format (e.g., non-cleavable), VH and VL cannot sterically associate to form an active binding domain, but rather prefer to pair intramolecularly with a pseudo-Fv.

[0146] The restricted Fv domains can comprise an active VH and an active VL (aVH and aVL) or an inactive VH and an inactive VL (iVH and iVL, in which case they are restricted pseudo-Fv domains), or a combination thereof, as described herein.

[0147] As will be appreciated by those skilled in the art, the order of VH and VL in a constrained Fv domain can be either VH-linker-VL or VL-linker-VH (from N-terminus to C-terminus).

[0148] As outlined herein, for format 1 constructs, the constrained Fv domain can comprise a VH and a VL linked using a cleavable linker (such as those shown in Figures 5 and 6). In this embodiment, the constrained Fv domain has the structure (N-terminus to C-terminus): vhFR1-vhCDR1-vhFR2-vhCDR2-vhFR3-vhCDR3-vhFR4-CCL-vlFR1-vlCDR1-vlFR2-vlCDR2-vlFR3-vlCDR3-vlFR4. Generally, a constrained Fv domain comprises active VH and VL domains (e.g., capable of binding to CD3 when associated) and thus has the structure (from N-terminus to C-terminus): vhFR1-avhCDR1-vhFR2-avhCDR2-vhFR3-avhCDR3-vhFR4-CCL-vlFR1-avlCDR1-vlFR2-avlCDR2-vlFR3-avlCDR3-vlFR4.

[0149] As outlined herein, for a Format 2 construct, the constrained Fv domain can comprise a VH and a VL linked using a non-cleavable linker. In this embodiment, the constrained Fv domain has the structure (N-terminus to C-terminus): vhFR1-vhCDR1-vhFR2-vhCDR2-vhFR3-vhCDR3-vhFR4-CNCL-vlFR1-vlCDR1-vlFR2-vlCDR2-vlFR3-vlCDR3-vlFR4. Generally, a constrained Fv domain comprises active VH and VL domains (e.g., capable of binding to CD3 when associated) and thus has the structure (from N-terminus to C-terminus): vhFR1-avhCDR1-vhFR2-avhCDR2-vhFR3-avhCDR3-vhFR4-CNCL-vlFR1-avlCDR1-vlFR2-avlCDR2-vlFR3-avlCDR3-vlFR4.

[0150] Of particular use in the present invention is a constrained non-cleavable Fv domain having aVH with SEQ ID NO:61, aVL with SEQ ID NO:49 and a domain linker with SEQ ID NO:74.

[0151] B. Constrained pseudo-Fv domains The present invention provides constrained pseudo-Fv domains comprising inactive or pseudo iVH and iVL domains covalently linked using a constrained linker (which may be cleavable or non-cleavable as outlined herein), which prevents intramolecular association between iVH and iVL in the absence of cleavage. Thus, a constrained pseudo-Fv domain generally comprises an iVH and an iVL with framework regions that allow for the association of the iVH and iVL (when in unconstrained form), but the resulting pseudo-Fv domain does not bind to human proteins: the iVH domain can assemble with the aVL domain, and the iVL domain can assemble with the aVH domain, but the resulting structure does not bind CD3.

[0152] The constrained pseudo-Fv domains comprise an inactive VH and VL (iVH and iVL).

[0153] As will be appreciated by those skilled in the art, the order of VH and VL in the constrained pseudo-Fv domain can be either VH-linker-VL or VL-linker-VH (from N-terminus to C-terminus).

[0154] As outlined herein, the constrained pseudo-Fv domains can comprise an iVH and an iVL linked using a non-cleavable linker, as shown in formats 1, 2 and 4, or with a cleavable linker, as shown in format 3.

[0155] Generally, a constrained Fv domain contains inactive VH and VL domains (e.g., capable of binding to CD3 when assembled) and thus has the structure (from N-terminus to C-terminus): vhFR1-ivlCDR1-vhFR2-ivlCDR2-vhFR3-ivlCDR3-vhFR4-CNCL-vlFR1-ivhCDR1-vlFR2-ivhCDR2-vlFR3-ivhCDR3-vlFR4.

[0156] Of particular use in the present invention is a constrained non-cleavable pseudo-Fv domain having an iVH having SEQ ID NO:65 or SEQ ID NO:69, an iVL having SEQ ID NO:53 or SEQ ID NO:57 and a domain linker having SEQ ID NO:74.

[0157] IX. Form of the Invention As discussed herein, the prodrug constructs of the present invention can take many different formats, including cleavable formats with dual TTA binding domains, non-cleavable formats with dual TTA binding domains (either of which can have the same TTA binding domain or different binding domains), and non-cleavable formats with a single targeting domain.

[0158] A. Cleavable format with dual targeting The present invention provides a non-isomerizing cleavable format of the type "Format 1" of Figure 1. In this embodiment, the constrained Fv domain comprises a VH and a VL domain linked using a constrained cleavable linker, and the constrained pseudo-Fv domain uses a constrained non-cleavable linker. Although both of these are referred to herein as "constrained", as discussed above and shown in Figures 37, 38 and 39, only one of these needs to be constrained, but generally the protein has better expression when both linkers are constrained.

[0159] All constructs of Format 1 (as well as other formats) also have a cleavable linker (CL) that is cleaved by human tumor proteases.

[0160] The present invention provides a prodrug protein comprising, from N-terminus to C-terminus, (sdABD-TTA1)-domain linker-constrained Fv domain-domain linker-(sdABD-TTA2)-CL-constrained pseudo-Fv domain-domain linker-sdABD-HSA.

[0161] As will be appreciated by those skilled in the art, the order of VH and VL of either the constrained Fv domain or the constrained pseudo-Fv domain can be (from N-terminus to C-terminus) either VH-linker-VL or VL-linker-VH.

[0162] Thus, in one embodiment, the prodrug protein comprises, from N-terminus to C-terminus, (sdABD-TTA1)-domain linker-aVH-CCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-sdABD-HSA.

[0163] Thus, in one embodiment, the prodrug protein comprises, from N-terminus to C-terminus, (sdABD-TTA1)-domain linker-aVH-CCL-aVL-domain linker-(sdABD-TTA2)-CL-iVH-CCL-iVL-domain linker-sdABD-HSA.

[0164] Thus, in one embodiment, the prodrug protein comprises, from N-terminus to C-terminus, (sdABD-TTA1)-domain linker-aVL-CCL-aVH-domain linker-(sdABD-TTA2)-CL-iVL-CCL-iVH-domain linker-sdABD-HSA.

[0165] Thus, in one embodiment, the prodrug protein comprises, from N-terminus to C-terminus, (sdABD-TTA1)-domain linker-aVL-CCL-aVH-domain linker-(sdABD-TTA2)-CL-iVH-CCL-iVL-domain linker-sdABD-HSA.

[0166] In some embodiments, the prodrug construct comprises sdABD(TTA1)-domain linker-aVH-CCL-aVL-domain linker-sdABD(TTA2)-CL-iVL-CNCL-iVH-NCL-sdABD(1 / 2), where aVH, aVL, iVH and iVL have the sequences shown in FIG.

[0167] In some embodiments, the prodrug construct comprises sdABD(TTA1)-domain linker-aVH-CCL-aVL-domain linker-sdABD(TTA2)-CL-iVL-CNCL-iVH-domain linker-sdABD(1 / 2). In this embodiment, the aVH, aVL, iVH, iVL have the sequence shown in Figure 5. In this embodiment, the two targeting domains bind to the same TTA, which may be EGFR, EpCAM, FOLR1 or B7H3, the sequence of which is shown in Figure 5.

[0168] In some embodiments, the prodrug construct comprises sdABD(TTA1)-domain linker-aVH-CCL-aVL-domain linker-sdABD(TTA2)-CL-iVL-CNCL-iVH-domain linker-sdABD(1 / 2). In this embodiment, aVH, aVL, iVH, iVL have the sequences shown in Figure 5. In this embodiment, the two targeting domains bind to different TTAs.

[0169] In some embodiments, the prodrug construct comprises sdABD(TTA1)-domain linker-aVH-CCL-aVL-domain linker-sdABD(TTA2)-CL-iVL-CNCL-iVH-domain linker-sdABD(1 / 2). In this embodiment, aVH, aVL, iVH, iVL have the sequences shown in Figure 5. In this embodiment, the two targeting domains bind to EGFR and EpCAM, and sdABD-TTA has the sequence of Figure 5.

[0170] In some embodiments, the prodrug construct comprises sdABD(TTA1)-domain linker-aVH-CCL-aVL-domain linker-sdABD(TTA2)-CL-iVL-CNCL-iVH-domain linker-sdABD(1 / 2). In this embodiment, aVH, aVL, iVH, iVL have the sequences shown in Figure 5. In this embodiment, the two targeting domains bind to EGFR and FOLR1, and sdABD-TTA has the sequence of Figure 5.

[0171] In some embodiments, the prodrug construct comprises sdABD(TTA1)-domain linker-aVH-CCL-aVL-domain linker-sdABD(TTA2)-CL-iVL-CNCL-iVH-domain linker-sdABD(1 / 2). In this embodiment, aVH, aVL, iVH, iVL have the sequences shown in Figure 5. In this embodiment, the two targeting domains bind EGFR and B7H3, and sdABD-TTA has the sequence of Figure 5.

[0172] In some embodiments, the prodrug construct comprises sdABD(TTA1)-domain linker-aVH-CCL-aVL-domain linker-sdABD(TTA2)-CL-iVL-CNCL-iVH-domain linker-sdABD(1 / 2). In this embodiment, aVH, aVL, iVH, iVL have the sequences shown in Figure 5. In this embodiment, the two targeting domains bind to EpCAM and FOLR1, and sdABD-TTA has the sequence of Figure 5.

[0173] In some embodiments, the prodrug construct comprises sdABD(TTA1)-domain linker-aVH-CCL-aVL-domain linker-sdABD(TTA2)-CL-iVL-CNCL-iVH-domain linker-sdABD(1 / 2). In this embodiment, aVH, aVL, iVH, iVL have the sequences shown in Figure 5. In this embodiment, the two targeting domains bind EpCAM and B7H3, and sdABD-TTA has the sequence of Figure 5.

[0174] In some embodiments, the prodrug construct comprises sdABD(TTA1)-domain linker-aVH-CCL-aVL-domain linker-sdABD(TTA2)-CL-iVL-CNCL-iVH-domain linker-sdABD(1 / 2). In this embodiment, aVH, aVL, iVH, iVL have the sequences shown in Figure 5. In this embodiment, the two targeting domains bind to B7H3 and FOLR1, and sdABD-TTA has the sequence of Figure 5.

[0175] In some embodiments, the prodrug construct comprises sdABD(TTA1)-domain linker-aVH-CCL-aVL-domain linker-sdABD(TTA2)-CL-iVL-CNCL-iVH-domain linker-sdABD(1 / 2). In this embodiment, aVH, aVL, iVH, iVL have the sequence shown in Figure 5. In this embodiment, the two targeting domains bind to the same TTA, which may be EGFR, FOLR1, B7H3 or EpCAM, the sequence of which is shown in Figure 5, CCL and CL are selected from linkers that are cleaved by MMP9 or meprin, and sdABD(1 / 2) has SEQ ID NO:45.

[0176] In format 1, a preferred domain linker is SEQ ID NO: 74 (which also serves as a preferred constrained non-cleavable linker).

[0177] In format 1, the preferred constructs are Pro140 and Pro140b.

[0178] B. Non-cleavable form As shown in Figure 2, the present invention provides a non-isomerizing non-cleavable format. In this embodiment, it is understood that "non-cleavable" applies only to the linkage of the constrained Fv domain, since an activation cleavage site is present in the prodrug construct. In this embodiment, the constrained Fv domain comprises a VH and a VL domain linked using a constrained non-cleavable linker, and the constrained pseudo-Fv domain uses a constrained non-cleavable linker.

[0179] As will be appreciated by those skilled in the art, the order of VH and VL of either the constrained Fv domain or the constrained pseudo-Fv domain can be (from N-terminus to C-terminus) either VH-linker-VL or VL-linker-VH.

[0180] The present invention provides a prodrug protein comprising, from N-terminus to C-terminus, sdABD(TTA1)-domain linker-constrained Fv domain-domain linker-sdABD(TTA2)-cleavable linker-constrained pseudo-Fv domain-domain linker-sdABD-HSA.

[0181] As will be appreciated by those skilled in the art, the order of VH and VL of either the constrained Fv domain or the constrained pseudo-Fv domain can be (from N-terminus to C-terminus) either VH-linker-VL or VL-linker-VH.

[0182] Thus, in one embodiment, the prodrug protein comprises, from N-terminus to C-terminus, (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-sdABD-HSA.

[0183] Thus, in one embodiment, the prodrug protein comprises, from N-terminus to C-terminus, (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVH-CNCL-iVL-domain linker-sdABD-HSA.

[0184] Thus, in one embodiment, the prodrug protein comprises, from N-terminus to C-terminus, (sdABD-TTA1)-domain linker-aVL-CNCL-aVH-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-sdABD-HSA.

[0185] Thus, in one embodiment, the prodrug protein comprises, from N-terminus to C-terminus, (sdABD-TTA1)-domain linker-aVL-CNCL-aVH-domain linker-(sdABD-TTA2)-CL-iVH-CNCL-iVL-domain linker-sdABD-HSA.

[0186] In some embodiments, the prodrug protein comprises, from N-terminus to C-terminus, (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-sdABD-HSA. In this embodiment, the aVH, aVL, iVH, iVL have the sequence shown in Figure 5. In this embodiment, the two targeting domains bind to the same TTA, which may be EGFR, EpCAM, FOLR1 or B7H3, the sequence of which is shown in Figure 5.

[0187] In some embodiments, the prodrug protein comprises, from N-terminus to C-terminus, (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-sdABD-HSA. In this embodiment, aVH, aVL, iVH, iVL have the sequences shown in Figure 5. In this embodiment, the two targeting domains bind to different TTAs.

[0188] In some embodiments, the prodrug protein comprises, from N-terminus to C-terminus, (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-sdABD-HSA. In this embodiment, aVH, aVL, iVH, iVL have the sequences shown in Figure 5. In this embodiment, the two targeting domains bind to EGFR and EpCAM, and sdABD-TTA has the sequence of Figure 5.

[0189] In some embodiments, the prodrug protein comprises, from N-terminus to C-terminus, (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-sdABD-HSA. In this embodiment, aVH, aVL, iVH, iVL have the sequences shown in Figure 5. In this embodiment, the two targeting domains bind to EGFR and FOLR1, and sdABD-TTA has the sequence of Figure 5.

[0190] In some embodiments, the prodrug protein comprises, from N-terminus to C-terminus, (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-sdABD-HSA. In this embodiment, aVH, aVL, iVH, iVL have the sequence shown in Figure 5. In this embodiment, the two targeting domains bind EGFR and B7H3, and sdABD-TTA has the sequence of Figure 5.

[0191] In some embodiments, the prodrug protein comprises, from N-terminus to C-terminus, (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-sdABD-HSA. In this embodiment, aVH, aVL, iVH, iVL have the sequences shown in Figure 5. In this embodiment, the two targeting domains bind to EpCAM and FOLR1, and sdABD-TTA has the sequence of Figure 5.

[0192] In some embodiments, the prodrug protein comprises, from N-terminus to C-terminus, (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-sdABD-HSA. In this embodiment, aVH, aVL, iVH, iVL have the sequences shown in Figure 5. In this embodiment, the two targeting domains bind EpCAM and B7H3, and sdABD-TTA has the sequence of Figure 5.

[0193] In some embodiments, the prodrug protein comprises, from N-terminus to C-terminus, (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-sdABD-HSA. In this embodiment, aVH, aVL, iVH, iVL have the sequences shown in Figure 5. In this embodiment, the two targeting domains bind to FOLR1 and B7H3, and sdABD-TTA has the sequence of Figure 5.

[0194] In some embodiments, the prodrug protein comprises, from N-terminus to C-terminus, (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-sdABD-HSA. In this embodiment, aVH, aVL, iVH, iVL have the sequence shown in Figure 5. In this embodiment, the two targeting domains bind to the same TTA, which may be EGFR, FOLR1, B7H3 or EpCAM, the sequence of which is shown in Figure 5, CCL and CL are selected from linkers that are cleaved by MMP9 or meprin, and sdABD(1 / 2) has SEQ ID NO:45.

[0195] In format 2, a preferred domain linker is SEQ ID NO: 74 (which also serves as a preferred constrained non-cleavable linker).

[0196] In Format 2, specific use embodiments include, but are not limited to, Pro186, Pro225, Pro226, Pro233, Pro311, Pro312, Pro313, Pro495, Pro246, Pro254, Pro255, Pro256, Pro420, Pro421, Pro432, Pro479, Pro480, Pro187, Pro221, Pro222, Pro223, Pro224, Pro393, Pro394, Pro395, Pro396, Pro429, Pro430 and Pro431.

[0197] C. Single TTA Constructs As shown in Figure 4, "Format 4" constructs similar to the Format 2 constructs but lacking the second TTA ABD are also included in the compositions of the invention. In this embodiment, it is understood that "non-cleavable" applies only to the linkage of the constrained Fv domains, due to the presence of an activation cleavage site in the prodrug construct. In this embodiment, the constrained Fv domains comprise VH and VL domains linked using a constrained non-cleavable linker, and the constrained pseudo-Fv domains use a constrained non-cleavable linker.

[0198] As will be appreciated by those skilled in the art, the order of VH and VL of either the constrained Fv domain or the constrained pseudo-Fv domain can be (from N-terminus to C-terminus) either VH-linker-VL or VL-linker-VH.

[0199] The present invention provides a prodrug protein comprising, from N-terminus to C-terminus, sdABD(TTA)-domain linker-constrained Fv domain-cleavable linker-sdABD-HSA-constrained pseudo-Fv domain. (Note that for all constructs of this type, sdABD-HSA generally does not have a His6 tag, although one may be included.)

[0200] As will be appreciated by those skilled in the art, the order of VH and VL of either the constrained Fv domain or the constrained pseudo-Fv domain can be (from N-terminus to C-terminus) either VH-linker-VL or VL-linker-VH.

[0201] Thus, in one embodiment, the prodrug protein comprises, from N-terminus to C-terminus, (sdABD-TTA)-domain linker-aVH-CNCL-aVL-CL-(sdABD-HSA)-domain linker-iVL-CNCL-iVH.

[0202] Thus, in one embodiment, the prodrug protein comprises, from N-terminus to C-terminus, (sdABD-TTA)-domain linker-aVH-CNCL-aVL-CL-(sdABD-HSA)-domain linker-iVH-CNCL-iVL.

[0203] Thus, in one embodiment, the prodrug protein comprises, from N-terminus to C-terminus, (sdABD-TTA)-domain linker-aVL-CNCL-aVH-CL-(sdABD-HSA)-domain linker-iVH-CNCL-iVL.

[0204] Thus, in one embodiment, the prodrug protein comprises, from N-terminus to C-terminus, (sdABD-TTA)-domain linker-aVL-CNCL-aVH-CL-(sdABD-HSA)-domain linker-iVL-CNCL-iVH.

[0205] Thus, in one embodiment, the prodrug protein comprises, from N-terminus to C-terminus: (sdABD-TTA)-domain linker-aVH-CNCL-aVL-CL-(sdABD-HSA)-domain linker-iVL-CNCL-iVH. In this embodiment, aVH, aVL, iVH and iVL have the sequence shown in Figure 5. In this embodiment, the targeting domain binds to a TTA, which may be EGFR, EpCAM, FOLR1 or B7H3, the sequence of which is shown in Figure 5.

[0206] In format 4, a preferred domain linker is SEQ ID NO: 74 (which also serves as a preferred constrained non-cleavable linker).

[0207] In Format 4, a preferred sdABD-HSA is that of SEQ ID NO:45.

[0208] D. Two Protein Compositions In some embodiments, the compositions of the invention comprise two distinct molecules, sometimes referred to as "half-COBRA™" or "half-constructs," that, in the absence of cleavage, associate intramolecularly to form a pseudo-Fv. In the presence of a protease, the cleavage site is cleaved, releasing the inactive variable domain, and the protein pair then forms an active antigen-binding domain against CD3, as generally shown in FIG.

[0209] What is important in the design of the half construct is that the active variable domain and the sdABD-TTA remain together after cleavage, so that the two cleaved portions can be held together by the tumor antigen receptor on the tumor surface, thereby forming the active anti-CD3 binding domain.

[0210] There are two different general format 3 constructs: one with a single sdABD-TTA in each member of the pair (Figure 3A), and one with two different sdABD-TTAs, each directed against a different TTA (Figure 3B).

[0211] 1. Half-COBRA™ Construct with a Single TTA Binding Domain (Format 3A) In some embodiments, the first COBRA™ half has, from N-terminus to C-terminus, sdABD(TTA1)-domain linker-aVH-CL-iVL-domain linker-sdABD(1 / 2) and the second COBRA™ half has sdABD(1 / 2)-domain linker-iVH-CL-aVL-domain linker-sdABD(TTA2). In this embodiment, the aVH, aVL, iVH, iVL and sdABD(1 / 2) have the sequence shown in Figure 5, and sdABD-TTAa binds human EGFR, EpCAM, FOLR1 and / or B7H3 and has the sequence shown in Figure 5.

[0212] 2. Half-COBRA™ Construct with Dual TTA ABD In some embodiments, the paired prodrug construct can have two sdABD-TTA binding domains per construct, as shown in Figure 3B. In this embodiment, the first member of the pair comprises, from N-terminus to C-terminus, sdABD-TTA1-domain linker-sdABD-TTA2-domain linker-aVH-CL-iVL-domain linker-sdABD(HAS) and the second member comprises, from N-terminus to C-terminus, sdABD-TTA1-domain linker-sdABD-TTA2-aVL-CL-iVH-domain linker-sdABD-HSA.

[0213] The two sdABD-TTAs on each member of the pair are different, but generally both members (half COBRA™) have the same two sdABD-TTAs, for example both have EGFR and FOLR1 or EGFR and B7H3.

[0214] The two sdABD-TTAs are, in some embodiments, selected from those shown in FIG.

[0215] X. Methods of Making the Compositions of the Invention The prodrug compositions of the present invention are made as generally recognized by those of skill in the art and as outlined below.

[0216] The present invention provides a nucleic acid composition that encodes the prodrug composition of the present invention. As will be understood by those skilled in the art, the nucleic acid composition depends on the format of the prodrug polypeptide(s). Thus, for example, if the format requires two amino acid sequences, such as a "format 3" construct, the two nucleic acid sequences can be incorporated into one or more expression vectors for expression. Similarly, the prodrug constructs that are a single polypeptide (formats 1, 2 and 4) require a single nucleic acid in a single expression vector for production.

[0217] As known in the art, the nucleic acid encoding the components of the present invention can be incorporated into an expression vector, as known in the art and depending on the host cell used to produce the prodrug composition of the present invention. In general, the nucleic acid is operably linked to any number of regulatory elements (promoter, origin of replication, selectable marker, ribosome binding site, inducer, etc.). The expression vector can be an extrachromosomal or integrating vector.

[0218] The nucleic acids and / or expression vectors of the invention are then transformed into any number of different types of host cells as are known in the art, including mammalian, bacterial, yeast, insect and / or fungal cells, with mammalian cells (e.g., CHO cells, 293 cells) being used in many embodiments.

[0219] The prodrug compositions of the present invention are produced by culturing host cells containing the expression vector(s), as is well known in the art. Once produced, conventional antibody purification steps are performed, including Protein A affinity and / or ion exchange chromatography steps.

[0220] XI. Formulations of the Prodrug Compositions of the Invention and Methods for Treating Cancer. Formulations of the prodrug compositions used in accordance with the present invention are prepared for storage by mixing the prodrug (single protein in the case of formats 1, 2 and 4, and two proteins in the case of format 3) having the desired purity (generally as outlined in Remington's Pharmaceutical Sciences 16th edition, Osol, A. Ed.

[1980] ) with any pharma- ceutically acceptable carriers, excipients or stabilizers in the form of a lyophilized formulation or aqueous solution.

[0221] The prodrug compositions of the present invention are administered to a subject according to known methods, such as intravenously, as a bolus or by continuous infusion over a period of time.

[0222] The prodrug compositions of the present invention are useful for the treatment of cancer. In an exemplary embodiment, the cancer is a solid cancer. In various embodiments, the cancer is selected from resectable cancer and non-resectable cancer. In some embodiments, the cancer is selected from advanced localized resectable cancer and advanced localized non-resectable cancer. In various embodiments, the cancer is selected from resectable metastatic cancer and non-resectable metastatic cancer.

[0223] Thus, in one aspect, the disclosure provides a method for treating a solid cancer in a patient in need of such treatment by administering to the patient a therapeutically effective amount of a polypeptide prodrug described herein. In some embodiments, the polypeptide comprises, from N-terminus to C-terminus, a) a first single domain antigen binding domain (sdABD) that binds to a first human tumor target antigen (TTA), b) a first domain linker, and c) a constrained Fv domain comprising i) a first variable heavy domain comprising vhCDR1, vhCDR2 and vhCDR3, ii) a constrained non-cleavable linker (CNCL), and a first variable light domain comprising vlCDR1, vlCDR2 and vlCDR3, wherein the CNCL is between the first variable heavy domain and the first variable light domain. and prevents the first variable heavy domain and the first variable light domain from interacting to form an active Fv capable of binding to CD3; d) a second domain linker; e) a second sdABD that binds to a second human TTA; f) a cleavable linker (CL); g) a pseudo-Fv domain comprising i) a first pseudo-variable light domain, ii) a non-cleavable linker (NCL), and iii) a first pseudo-variable heavy domain; h) a third domain linker; and i) a third sdABD that binds to human serum albumin. The first variable heavy domain and the first variable light domain are capable of binding to human CD3, but the constrained Fv domain does not bind to CD3. The polypeptide does not bind to CD3 when the CL is intact.

[0224] In various embodiments, the cancer is head and neck cancer, e.g., squamous cell carcinoma. In an exemplary embodiment, the cancer is lung cancer, e.g., non-small cell lung cancer. In an exemplary embodiment, the cancer is prostate cancer. In some embodiments, the cancer is a colorectal neoplasm.

[0225] In an exemplary embodiment, the subject to whom the compound of the invention is administered has non-small cell lung cancer that has progressed during or after treatment with platinum-based chemotherapy and anti-PDx therapy (e.g., anti-PD1 therapy). In some embodiments, the subject to whom the compound of the invention is administered has non-small cell lung cancer with an EGFR mutation or ALK rearrangement. In some embodiments, the cancer has progressed after treatment with platinum-based chemotherapy plus an available EGFR or ALK targeted therapy.

[0226] In an exemplary embodiment, where the cancer is head and neck squamous cell carcinoma, the cancer in the subject to whom the compound of the invention is administered has progressed during or after treatment with anti-PDx (e.g., anti-PD1 therapy) (unless ineligible, e.g., chemotherapy not working, patients with PD-L1 CPS<1) and platinum-based chemotherapy (unless ineligible / intolerant to platinum-based chemotherapy) for metastatic or recurrent disease.

[0227] In exemplary embodiments, the subject to whom the compound of the present invention is administered has advanced or stopped prostate cancer. In some embodiments, the progression or stopping of prostate cancer is evaluated by measuring the level of prostate specific antigen (PSA) and / or prostate specific membrane antigen (PSMA) by blood test and / or positron emission tomography (PET) scan. In some embodiments, the success of the treatment is evaluated based on the level of PSA being reduced by 50%, about 50%, or more than 50% compared to the original level of PSA when the treatment is administered to the patient for at least 2 weeks, at least 3 weeks, at least 4 weeks, at least 5 weeks, at least 6 weeks, at least 7 weeks, at least 8 weeks, at least 9 weeks, at least 10 weeks, at least 11 weeks, at least 12 weeks, at least 13 weeks, at least 14 weeks, at least 16 weeks, at least 17 weeks, at least 18 weeks, at least 19 weeks, at least 20 weeks, at least 25 weeks, at least 30 weeks, at least 35 weeks, at least 40 weeks, at least 45 weeks, at least 50 weeks, at least 55 weeks, or at least 60 weeks. In some embodiments, the success of the treatment is assessed based on a 50%, about 50%, or more than 50% decrease in the level of PSA compared to the original level of PSA when the treatment has been administered to the patient for at least 1 month, at least 2 months, at least 3 months, at least 4 months, at least 5 months, at least 6 months, at least 7 months, at least 8 months, at least 9 months, at least 10 months, at least 11 months, at least 12 months, at least 13 months, at least 14 months, or at least 15 months. In some embodiments, the success of the treatment is assessed based on a 50%, about 50%, or more than 50% decrease in the level of PSA compared to the original level of PSA when the treatment has been administered to the patient for about 6 months.In some embodiments, the reduction in PSA levels is about 50%, about 51%, about 52%, about 53%, about 54%, about 55%, about 56%, about 57%, about 58%, about 59%, about 60%, about 61%, about 62%, about 63%, about 64%, about 65%, about 66%, about 67%, about 68%, about 69%, about 70%, about 71%, about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 78%, about 79%, 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%, about 99%, about 100%, about 101%, about 102%, about 103%, about 104%, about 105%, about 106%, about 107%, about 108%, about 109%, about 110%, about 111%, about 112%, about 113%, about 114%, about 115%, about 116%, about 117%, about 118%, about 119%, about 120%, about 121%, about 122%, about 123%, about 124%, about 125%, about 126%, about 127%, about 128%, about 129%, about 130%, about 131%, about 132%, about 133%, about 134%, about 135%, about 136%, about 137%, about 138%, about 139%, about 140%, about 14 The PSA level is about 71%, about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 78%, about 79%, 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%, about 99%, or about 100%. In some embodiments, the success of the treatment is evaluated based on a 50%, about 50%, or more than 50% decrease in the level of PSA compared to the original level of PSA, when the 50%, about 50%, or more than 50% decrease in the level of PSA is sustained for up to about 6 months or more than 6 months after treatment. In some embodiments, the success of the treatment is evaluated based on the absence or reduction of bone metastases. In some embodiments, the reduction in detectable metastases, the stability of metastasis size, and / or the increase in metastasis size is determined by PET scan. In some embodiments, the success of treatment is determined based on a reduction in tumor size of at least 20%, at least 21%, at least 22%, at least 23%, at least 24%, at least 25%, at least 26%, at least 27%, at least 28%, at least 29%, at least 30%, at least 31%, at least 32%, at least 33%, at least 34%, at least 35%, at least 36%, at least 37%, at least 38%, at least 39%, at least 40%, at least 41%, at least 42%, at least 43%, at least 44%, at least 45%, at least 46%, at least 47%, at least 48%, at least 49%, at least 50% compared to the original size before or at the start of treatment (see exemplary details in Supplement A).

[0228] In some embodiments, the subject to whom the compound of the invention is administered has prostate cancer that has progressed during or after treatment with platinum-based chemotherapy and anti-PDx therapy (e.g., anti-PD1 therapy).

[0229] In some embodiments, compounds of the invention are administered simultaneously to patients in combination with other cancer treatments, such as, but not limited to, radiation.

[0230] In various embodiments, the compounds of the invention are administered to a subject whose cancer is colorectal cancer, such as: (a) K-Ras WT: subjects who have progressed on or after, or are ineligible for, both irinotecan- and oxaliplatin-based chemotherapy and subjects who have relapsed or are refractory to at least one prior systemic therapy, including an anti-EGFR antibody; (b) K-Ras mutation: subjects who have progressed on or after, or are ineligible for, both irinotecan- and oxaliplatin-based chemotherapy.

[0231] In some embodiments, the subject has an ECOG performance status of ≦1.

[0232] In various embodiments, subjects have disease evaluable according to RECIST v1.1 criteria and documented by CT and / or MRI.

[0233] In an exemplary embodiment, the subject has undergone immune checkpoint therapy prior to being administered a compound of the invention, and has resolved immune-related toxicity of the checkpoint inhibitor to either grade ≦1 or baseline.

[0234] In an exemplary embodiment, the subject has symptomatic central nervous system metastases that have been treated, have been asymptomatic for 14 days or more, and are free of concomitant treatment or leptomeningeal disease / spinal cord compression.

[0235] As will be understood by those skilled in the art, the choice of dosage is within the ability of a physician experienced in cancer treatment.In an exemplary embodiment, the subject is administered the compound of the present invention at a dose of about 0.3 μg / kg to about 30 μg / kg.In an exemplary embodiment, the compound of the present invention is administered to the subject about once a week.

[0236] As will be appreciated by those skilled in the art, the duration of treatment is within the capabilities of a physician experienced in cancer treatment.

[0237] As will be appreciated by those skilled in the art, the vehicle in which the compounds of the present invention are formulated may be any useful, safe, non-toxic, pharma- ceutically acceptable formulation. In an exemplary embodiment, the compound is in an isotonic, sterile aqueous solution of 0.9 mg / ml sodium chloride.

[0238] A. Treating Cancer with EGFR COBRA In some embodiments, the disclosure provides a method for treating a solid tumor in a patient in need of such treatment by administering to the patient a therapeutically effective amount of a polypeptide prodrug comprising at least one anti-EGFR single domain antigen binding domain described herein. In some embodiments, the polypeptide comprises, from N-terminus to C-terminus, a) a first single domain antigen binding domain (sdABD) that binds to a first human tumor target antigen (TTA), which is EGFR, b) a first domain linker, and c) a constrained Fv domain comprising i) a first variable heavy domain comprising vhCDR1, vhCDR2, and vhCDR3, ii) a constrained non-cleavable linker (CNCL), and a first variable light domain comprising vlCDR1, vlCDR2, and vlCDR3, wherein CNCL is a linker between the first variable heavy domain and the first variable light domain. a constrained Fv domain located between the first variable heavy domain and the first variable light domain, preventing the first variable heavy domain and the first variable light domain from interacting to form an active Fv capable of binding to CD3; d) a second domain linker; e) a second sdABD that binds to a second human TTA; f) a cleavable linker (CL); g) a pseudo-Fv domain comprising i) a first pseudo-variable light domain, ii) a non-cleavable linker (NCL), and iii) a first pseudo-variable heavy domain; h) a third domain linker; and i) a third sdABD that binds to human serum albumin. The first variable heavy domain and the first variable light domain are capable of binding to human CD3, but the constrained Fv domain does not bind to CD3. The polypeptide does not bind to CD3 when the CL is intact.

[0239] In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is selected from the group consisting of SEQ ID NO:143 (Pro140), SEQ ID NO:144 (Pro140b), SEQ ID NO:185 (Pro141), SEQ ID NO:199 (Pro176), SEQ ID NO:208 (Pro188), SEQ ID NO:200 (Pro178), SEQ ID NO:201 (Pro179), SEQ ID NO:202 (Pro180), SEQ ID NO:203 (Pro181), SEQ ID NO:204 (Pro182), SEQ ID NO:205 (Pro183), SEQ ID NO:206 (Pro184), SEQ ID NO:207 (Pro188), SEQ ID NO:208 (Pro189), SEQ ID NO:209 (Pro190), SEQ ID NO:300 (Pro191), SEQ ID NO:301 (Pro192), SEQ ID NO:302 (Pro193), SEQ ID NO:303 (Pro194), SEQ ID NO:304 (Pro195), SEQ ID NO:305 (Pro196), SEQ ID NO:306 (Pro197), SEQ ID NO:307 (Pro198), SEQ ID NO:308 (Pro199), SEQ ID NO:309 (Pro199), SEQ ID NO:409 (Pro198), SEQ ID NO:410 (Pro199), SEQ ID NO:411 (Pro199), SEQ ID NO:412 (Pro199), SEQ ID NO:413 (Pro199), SEQ ID NO:414 (Pro199), SEQ ID NO:415 (Pro199), SEQ ID NO:416 (Pro199), SEQ ID NO:417 (Pro199), SEQ ID NO:41 ro185), SEQ ID NO: 145 (Pro186), SEQ ID NO: 146 (Pro187), SEQ ID NO: 189 (Pro189), SEQ ID NO: 2 (Pro190), SEQ ID NO: 191 (Pro191), SEQ ID NO: 212 (Pro192), SEQ ID NO: 214 (Pro195), SEQ ID NO: 215 (Pro196), SEQ ID NO: 216 (Pro197), SEQ ID NO: 217 (Pro198), SEQ ID NO: 165 (SEQ ID NO: 221), SEQ ID NO: 166 (Pro222), SEQ ID NO: 167 (Pro223), SEQ ID NO: 168 (Pro224), SEQ ID NO: 149 (Pro233), SEQ ID NO: 226 (Pro247), SEQ ID NO: 227 (Pro248), SEQ ID NO: 228 (Pro249), SEQ ID NO: 299 (Pro250), SEQ ID NO: 230 (Pro251), SEQ ID NO: 231 (Pro252), SEQ ID NO: 232 (Pro253), SEQ ID NO: 153 (Pro246), SEQ ID NO: 169 (Pro254), SEQ ID NO: 170 (Pro255), SEQ ID NO: 154 (Pro256), SEQ ID NO: 233 (Pro261), SEQ ID NO: 171 (Pro262), SEQ ID NO: 234 (Pro294), SEQ ID NO: 250 (Pro345), SEQ ID NO: 259 (Pro375) ), SEQ ID NO:260 (Pro376), SEQ ID NO:157 (Pro393), SEQ ID NO:158 (Pro394), SEQ ID NO:159 (Pro395), SEQ ID NO:160 (Pro396), SEQ ID NO:263 (Pro412), SEQ ID NO:264 (Pro413), SEQ ID NO:265 (Pro414), SEQ ID NO:265 (Pro415), SEQ ID NO:266 (Pro416), SEQ ID NO:267 (Pro417), SEQ ID NO:269 (Pro418), SEQ ID NO:272 (Pro429), SEQ ID NO:162 (Pro430), and SEQ ID NO:163 (Pro431).

[0240] In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is selected from SEQ ID NO:145 (Pro186), SEQ ID NO:149 (Pro233), SEQ ID NO:153 (Pro246), SEQ ID NO:169 (Pro254), SEQ ID NO:232 (Pro253), and SEQ ID NO:234 ​​(Pro294).

[0241] In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 95% identical to SEQ ID NO: 145 (Pro186). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 96% identical to SEQ ID NO: 145 (Pro186). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 97% identical to SEQ ID NO: 145 (Pro186). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 98% identical to SEQ ID NO: 145 (Pro186). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 99% identical to SEQ ID NO: 145 (Pro186). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 99.5% identical to SEQ ID NO: 145 (Pro186). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 25 amino acids compared to SEQ ID NO: 145 (Pro186). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 20 amino acids compared to SEQ ID NO: 145 (Pro186). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 15 amino acids compared to SEQ ID NO: 145 (Pro186). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 10 amino acids compared to SEQ ID NO: 145 (Pro186). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 5 amino acids compared to SEQ ID NO: 145 (Pro186).In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than four amino acids compared to SEQ ID NO: 145 (Pro186). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than three amino acids compared to SEQ ID NO: 145 (Pro186). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than two amino acids compared to SEQ ID NO: 145 (Pro186). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than one amino acid compared to SEQ ID NO: 145 (Pro186).

[0242] In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is SEQ ID NO: 145 (Pro186).

[0243] In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 95% identical to SEQ ID NO: 149 (Pro233). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 96% identical to SEQ ID NO: 149 (Pro233). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 97% identical to SEQ ID NO: 149 (Pro233). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 98% identical to SEQ ID NO: 149 (Pro233). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 99% identical to SEQ ID NO: 149 (Pro233). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 99.5% identical to SEQ ID NO: 149 (Pro233). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 25 amino acids compared to SEQ ID NO: 149 (Pro233). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 20 amino acids compared to SEQ ID NO: 149 (Pro233). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 15 amino acids compared to SEQ ID NO: 149 (Pro233). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 10 amino acids compared to SEQ ID NO: 149 (Pro233). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 5 amino acids compared to SEQ ID NO: 149 (Pro233).In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that has no more than four amino acid differences compared to SEQ ID NO: 149 (Pro233). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that has no more than three amino acid differences compared to SEQ ID NO: 149 (Pro233). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that has no more than two amino acid differences compared to SEQ ID NO: 149 (Pro233). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that has no more than one amino acid difference compared to SEQ ID NO: 149 (Pro233).

[0244] In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is SEQ ID NO: 149 (Pro233).

[0245] In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 95% identical to SEQ ID NO: 153 (Pro246). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 96% identical to SEQ ID NO: 153 (Pro246). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 97% identical to SEQ ID NO: 153 (Pro246). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 98% identical to SEQ ID NO: 153 (Pro246). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 99% identical to SEQ ID NO: 153 (Pro246). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 99.5% identical to SEQ ID NO: 153 (Pro246). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 25 amino acids compared to SEQ ID NO: 153 (Pro246). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 20 amino acids compared to SEQ ID NO: 153 (Pro246). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 15 amino acids compared to SEQ ID NO: 153 (Pro246). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 10 amino acids compared to SEQ ID NO: 153 (Pro246). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 5 amino acids compared to SEQ ID NO: 153 (Pro246).In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that has no more than four amino acid differences compared to SEQ ID NO: 153 (Pro246). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that has no more than three amino acid differences compared to SEQ ID NO: 153 (Pro246). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that has no more than two amino acid differences compared to SEQ ID NO: 153 (Pro246). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that has no more than one amino acid difference compared to SEQ ID NO: 153 (Pro246).

[0246] In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is SEQ ID NO: 153 (Pro246).

[0247] In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 95% identical to SEQ ID NO: 169 (Pro254). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 96% identical to SEQ ID NO: 169 (Pro254). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 97% identical to SEQ ID NO: 169 (Pro254). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 98% identical to SEQ ID NO: 169 (Pro254). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 99% identical to SEQ ID NO: 169 (Pro254). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 99.5% identical to SEQ ID NO: 169 (Pro254). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 25 amino acids compared to SEQ ID NO: 169 (Pro254). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 20 amino acids compared to SEQ ID NO: 169 (Pro254). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 15 amino acids compared to SEQ ID NO: 169 (Pro254). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 10 amino acids compared to SEQ ID NO: 169 (Pro254). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 5 amino acids compared to SEQ ID NO: 169 (Pro254).In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that has no more than four amino acid differences compared to SEQ ID NO: 169 (Pro254). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that has no more than three amino acid differences compared to SEQ ID NO: 169 (Pro254). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that has no more than two amino acid differences compared to SEQ ID NO: 169 (Pro254). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that has no more than one amino acid difference compared to SEQ ID NO: 169 (Pro254).

[0248] In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is SEQ ID NO: 169 (Pro254).

[0249] In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 95% identical to SEQ ID NO: 232 (Pro253). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 96% identical to SEQ ID NO: 232 (Pro253). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 97% identical to SEQ ID NO: 232 (Pro253). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 98% identical to SEQ ID NO: 232 (Pro253). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 99% identical to SEQ ID NO: 232 (Pro253). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 99.5% identical to SEQ ID NO: 232 (Pro253). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 25 amino acids compared to SEQ ID NO: 232 (Pro253). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 20 amino acids compared to SEQ ID NO: 232 (Pro253). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 15 amino acids compared to SEQ ID NO: 232 (Pro253). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 10 amino acids compared to SEQ ID NO: 232 (Pro253). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 5 amino acids compared to SEQ ID NO: 232 (Pro253).In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than four amino acids compared to SEQ ID NO:232 (Pro253). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than three amino acids compared to SEQ ID NO:232 (Pro253). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than two amino acids compared to SEQ ID NO:232 (Pro253). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than one amino acid compared to SEQ ID NO:232 (Pro253).

[0250] In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is SEQ ID NO: 232 (Pro253).

[0251] In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 95% identical to SEQ ID NO: 234 (Pro294). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 96% identical to SEQ ID NO: 234 (Pro294). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 97% identical to SEQ ID NO: 234 (Pro294). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 98% identical to SEQ ID NO: 234 (Pro294). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 99% identical to SEQ ID NO: 234 (Pro294). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence at least 99.5% identical to SEQ ID NO: 234 (Pro294). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 25 amino acids compared to SEQ ID NO: 234 (Pro294). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 20 amino acids compared to SEQ ID NO: 234 (Pro294). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 15 amino acids compared to SEQ ID NO: 234 (Pro294). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 10 amino acids compared to SEQ ID NO: 234 (Pro294). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than 5 amino acids compared to SEQ ID NO: 234 (Pro294).In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than four amino acids compared to SEQ ID NO:234 ​​(Pro294). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than three amino acids compared to SEQ ID NO:234 ​​(Pro294). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than two amino acids compared to SEQ ID NO:234 ​​(Pro294). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain comprises an amino acid sequence that differs by no more than one amino acid compared to SEQ ID NO:234 ​​(Pro294).

[0252] In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is SEQ ID NO: 234 (Pro294).

[0253] In various embodiments, the cancer is head and neck cancer, e.g., squamous cell carcinoma. In an exemplary embodiment, the cancer is lung cancer, e.g., non-small cell lung cancer. In an exemplary embodiment, the cancer is prostate cancer. In some embodiments, the cancer is a colorectal neoplasm.

[0254] Thus, in some embodiments, the method comprises treating head and neck cancer, e.g., squamous cell carcinoma, by administering to the patient a therapeutically effective amount of a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain described herein. In some embodiments, the method comprises treating lung cancer, e.g., non-small cell lung cancer, by administering to the patient a therapeutically effective amount of a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain described herein. In some embodiments, the method comprises treating prostate cancer, by administering to the patient a therapeutically effective amount of a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain described herein. In some embodiments, the method comprises treating colorectal cancer, by administering to the patient a therapeutically effective amount of a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain described herein.

[0255] In some embodiments, the method comprises treating head and neck cancer, e.g., squamous cell carcinoma, with a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain described herein. In some embodiments, the method comprises treating head and neck cancer, e.g., squamous cell carcinoma, with a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain selected from SEQ ID NO: 145 (Pro186), SEQ ID NO: 149 (Pro233), SEQ ID NO: 153 (Pro246), SEQ ID NO: 169 (Pro254), SEQ ID NO: 232 (Pro253), and SEQ ID NO: 234 (Pro294). In some embodiments, the method comprises treating head and neck cancer, e.g., squamous cell carcinoma, with Pro186 (SEQ ID NO: 145).

[0256] In some embodiments, the methods include treating lung cancer, e.g., non-small cell lung cancer, with a polypeptide prodrug that includes an anti-EGFR single domain antigen binding domain described herein. In some embodiments, the methods include treating lung cancer, e.g., non-small cell lung cancer, with a polypeptide prodrug that includes an anti-EGFR single domain antigen binding domain selected from SEQ ID NO: 145 (Pro186), SEQ ID NO: 149 (Pro233), SEQ ID NO: 153 (Pro246), SEQ ID NO: 169 (Pro254), SEQ ID NO: 232 (Pro253), and SEQ ID NO: 234 (Pro294). In some embodiments, the methods include treating lung cancer, eg, non-small cell lung cancer, with Pro186 (SEQ ID NO: 145).

[0257] In some embodiments, the methods include treating prostate cancer with a polypeptide prodrug that includes an anti-EGFR single domain antigen binding domain described herein. In some embodiments, the methods include treating prostate cancer with a polypeptide prodrug that includes an anti-EGFR single domain antigen binding domain selected from SEQ ID NO: 145 (Pro186), SEQ ID NO: 149 (Pro233), SEQ ID NO: 153 (Pro246), SEQ ID NO: 169 (Pro254), SEQ ID NO: 232 (Pro253), and SEQ ID NO: 234 (Pro294). In some embodiments, the methods include treating prostate cancer with Pro186 (SEQ ID NO: 145).

[0258] In some embodiments, the methods include treating colorectal cancer with a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain described herein. In some embodiments, the methods include treating colorectal cancer with a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain selected from SEQ ID NO: 145 (Pro186), SEQ ID NO: 149 (Pro233), SEQ ID NO: 153 (Pro246), SEQ ID NO: 169 (Pro254), SEQ ID NO: 232 (Pro253), and SEQ ID NO: 234 (Pro294). In some embodiments, the methods include treating colorectal cancer with Pro186 (SEQ ID NO: 145).

[0259] In exemplary embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is administered to a subject with non-small cell lung cancer that has progressed during or after treatment with platinum-based chemotherapy and anti-PDx therapy (e.g., anti-PD1 therapy). In some embodiments, the subject has non-small cell lung cancer with an EGFR mutation or ALK rearrangement. In some embodiments, the cancer has progressed after treatment with platinum-based chemotherapy plus available EGFR or ALK targeted therapy. In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is selected from SEQ ID NO: 145 (Pro186), SEQ ID NO: 149 (Pro233), SEQ ID NO: 153 (Pro246), SEQ ID NO: 169 (Pro254), SEQ ID NO: 232 (Pro253), and SEQ ID NO: 234 (Pro294). In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is SEQ ID NO: 145 (Pro186).

[0260] In exemplary embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is administered to a subject with head and neck squamous cell carcinoma that has progressed during or after treatment with anti-PDx (e.g., anti-PD1 therapy) and platinum-based chemotherapy for metastatic or recurrent disease. In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is selected from SEQ ID NO: 145 (Pro186), SEQ ID NO: 149 (Pro233), SEQ ID NO: 153 (Pro246), SEQ ID NO: 169 (Pro254), SEQ ID NO: 232 (Pro253), and SEQ ID NO: 234 (Pro294). In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is SEQ ID NO: 145 (Pro186).

[0261] In exemplary embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is administered to a subject with advanced or halted prostate cancer. In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is selected from SEQ ID NO: 145 (Pro186), SEQ ID NO: 149 (Pro233), SEQ ID NO: 153 (Pro246), SEQ ID NO: 169 (Pro254), SEQ ID NO: 232 (Pro253), and SEQ ID NO: 234 (Pro294). In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is SEQ ID NO: 145 (Pro186).

[0262] In some embodiments, progression or cessation of prostate cancer is assessed by measuring the levels of prostate specific antigen (PSA) and / or prostate specific membrane antigen (PSMA) by blood tests and / or positron emission tomography (PET) scans. In some embodiments, the success of the treatment is assessed based on a 50%, about 50%, or more than 50% reduction in the level of PSA compared to the original level of PSA when the treatment has been administered to the patient for at least 2 weeks, at least 3 weeks, at least 4 weeks, at least 5 weeks, at least 6 weeks, at least 7 weeks, at least 8 weeks, at least 9 weeks, at least 10 weeks, at least 11 weeks, at least 12 weeks, at least 13 weeks, at least 14 weeks, at least 16 weeks, at least 17 weeks, at least 18 weeks, at least 19 weeks, at least 20 weeks, at least 25 weeks, at least 30 weeks, at least 35 weeks, at least 40 weeks, at least 45 weeks, at least 50 weeks, at least 55 weeks, or at least 60 weeks. In some embodiments, the success of the treatment is assessed based on a 50%, about 50%, or more than 50% decrease in the level of PSA compared to the original level of PSA when the treatment has been administered to the patient for at least 1 month, at least 2 months, at least 3 months, at least 4 months, at least 5 months, at least 6 months, at least 7 months, at least 8 months, at least 9 months, at least 10 months, at least 11 months, at least 12 months, at least 13 months, at least 14 months, or at least 15 months. In some embodiments, the success of the treatment is assessed based on a 50%, about 50%, or more than 50% decrease in the level of PSA compared to the original level of PSA when the treatment has been administered to the patient for about 6 months.In some embodiments, the reduction in PSA levels is about 50%, about 51%, about 52%, about 53%, about 54%, about 55%, about 56%, about 57%, about 58%, about 59%, about 60%, about 61%, about 62%, about 63%, about 64%, about 65%, about 66%, about 67%, about 68%, about 69%, about 70%, about 71%, about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 78%, about 79%, 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%, about 99%, about 100%, about 101%, about 102%, about 103%, about 104%, about 105%, about 106%, about 107%, about 108%, about 109%, about 110%, about 111%, about 112%, about 113%, about 114%, about 115%, about 116%, about 117%, about 118%, about 119%, about 120%, about 121%, about 122%, about 123%, about 124%, about 125%, about 126%, about 127%, about 128%, about 129%, about 130%, about 131%, about 132%, about 133%, about 134%, about 135%, about 136%, about 137%, about 138%, about 139%, about 140%, about 14 The PSA level is about 71%, about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 78%, about 79%, 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%, about 99%, or about 100%. In some embodiments, the success of the treatment is evaluated based on a 50%, about 50%, or more than 50% decrease in the level of PSA compared to the original level of PSA, when the 50%, about 50%, or more than 50% decrease in the level of PSA is sustained for up to about 6 months or more than 6 months after treatment. In some embodiments, the success of the treatment is measured based on the absence or reduction of bone metastases. In some embodiments, the reduction in detectable metastases, the stability of the size of metastases, and / or the increase in the size of metastases is determined by PET scan. In some embodiments, the success of the treatment is determined based on a reduction in tumor size of at least 20%, at least 21%, at least 22%, at least 23%, at least 24%, at least 25%, at least 26%, at least 27%, at least 28%, at least 29%, at least 30%, at least 31%, at least 32%, at least 33%, at least 34%, at least 35%, at least 36%, at least 37%, at least 38%, at least 39%, at least 40%, at least 41%, at least 42%, at least 43%, at least 44%, at least 45%, at least 46%, at least 47%, at least 48%, at least 49%, at least 50% compared to the original size before or at the start of the treatment.

[0263] In exemplary embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is administered to a subject with prostate cancer that has progressed during or after treatment with platinum-based chemotherapy and anti-PDx therapy (e.g., anti-PD1 therapy). In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is selected from SEQ ID NO: 145 (Pro186), SEQ ID NO: 149 (Pro233), SEQ ID NO: 153 (Pro246), SEQ ID NO: 169 (Pro254), SEQ ID NO: 232 (Pro253), and SEQ ID NO: 234 (Pro294). In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is SEQ ID NO: 145 (Pro186).

[0264] In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is administered simultaneously to a patient in combination with other cancer treatments, such as, but not limited to, radiation. In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is selected from SEQ ID NO: 145 (Pro186), SEQ ID NO: 149 (Pro233), SEQ ID NO: 153 (Pro246), SEQ ID NO: 169 (Pro254), SEQ ID NO: 232 (Pro253), and SEQ ID NO: 234 (Pro294). In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is SEQ ID NO: 145 (Pro186).

[0265] In various embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is administered to a subject with colorectal cancer having a wild-type K-Ras genotype. In some embodiments, the subject with a wild-type K-Ras genotype has progressed during or after chemotherapy, for example, irinotecan and / or oxaliplatin-based chemotherapy. In some embodiments, the subject with a wild-type K-Ras genotype is ineligible for irinotecan and / or oxaliplatin-based chemotherapy. In some embodiments, the subject with a wild-type K-Ras genotype has relapsed or is refractory to at least one prior systemic therapy comprising an anti-EGFR antibody. In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is selected from SEQ ID NO: 145 (Pro186), SEQ ID NO: 149 (Pro233), SEQ ID NO: 153 (Pro246), SEQ ID NO: 169 (Pro254), SEQ ID NO: 232 (Pro253), and SEQ ID NO: 234 (Pro294). In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is SEQ ID NO: 145 (Pro186).

[0266] In various embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is administered to a subject with colorectal cancer having a mutant K-Ras genotype. In some embodiments, the subject with a mutant K-Ras genotype has progressed during or after chemotherapy, for example, irinotecan and / or oxaliplatin-based chemotherapy. In some embodiments, the subject with a mutant K-Ras genotype is ineligible for irinotecan and / or oxaliplatin-based chemotherapy. In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is selected from SEQ ID NO: 145 (Pro186), SEQ ID NO: 149 (Pro233), SEQ ID NO: 153 (Pro246), SEQ ID NO: 169 (Pro254), SEQ ID NO: 232 (Pro253), and SEQ ID NO: 234 (Pro294). In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is SEQ ID NO: 145 (Pro186).

[0267] In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is administered to a subject with an ECOG performance status of 1 or less. In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is administered to a subject with an ECOG performance status of 0.5, 0.6, 0.7, 0.8, 0.9, 1.0, 1.1, 1.2, 1.3, 1.4, or 1.5 or less. In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is selected from SEQ ID NO: 145 (Pro186), SEQ ID NO: 149 (Pro233), SEQ ID NO: 153 (Pro246), SEQ ID NO: 169 (Pro254), SEQ ID NO: 232 (Pro253), and SEQ ID NO: 234 (Pro294). In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is SEQ ID NO: 145 (Pro186).

[0268] In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is administered to a subject with disease evaluable according to RECIST v1.1 criteria and documented by CT and / or MRI. In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is selected from SEQ ID NO: 145 (Pro186), SEQ ID NO: 149 (Pro233), SEQ ID NO: 153 (Pro246), SEQ ID NO: 169 (Pro254), SEQ ID NO: 232 (Pro253), and SEQ ID NO: 234 (Pro294). In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is SEQ ID NO: 145 (Pro186).

[0269] In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is administered to a subject who has undergone immune checkpoint therapy prior to administration. In some embodiments, the subject who has undergone immune checkpoint therapy prior to administration has recovered from checkpoint inhibitor immune-related toxicity to either grade ≦1 or baseline. In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is selected from SEQ ID NO: 145 (Pro186), SEQ ID NO: 149 (Pro233), SEQ ID NO: 153 (Pro246), SEQ ID NO: 169 (Pro254), SEQ ID NO: 232 (Pro253), and SEQ ID NO: 234 (Pro294). In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is SEQ ID NO: 145 (Pro186).

[0270] In some embodiments, a polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is administered to a subject with symptomatic central nervous system metastases who has been treated and has been asymptomatic for at least 14 days, without concomitant concomitant treatment, or leptomeningeal disease / spinal cord compression. In some embodiments, the subject has been asymptomatic for at least 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 128, 19, 20, 21, 28, 35, 42, 49 days or more. In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is selected from SEQ ID NO: 145 (Pro186), SEQ ID NO: 149 (Pro233), SEQ ID NO: 153 (Pro246), SEQ ID NO: 169 (Pro254), SEQ ID NO: 232 (Pro253), and SEQ ID NO: 234 (Pro294). In some embodiments, the polypeptide prodrug comprising an anti-EGFR single domain antigen binding domain is SEQ ID NO: 145 (Pro186).

[0271] As will be understood by those skilled in the art, the choice of dosage is within the ability of a physician experienced in cancer treatment.In an exemplary embodiment, the subject is administered the compound of the present invention at a dose of about 0.3 μg / kg to about 30 μg / kg.In an exemplary embodiment, the compound of the present invention is administered to the subject about once a week.

[0272] As will be appreciated by those skilled in the art, the duration of treatment is within the capabilities of a physician experienced in cancer treatment.

[0273] As will be appreciated by those skilled in the art, the vehicle in which the compounds of the present invention are formulated may be any useful, safe, non-toxic, pharma- ceutically acceptable formulation. In an exemplary embodiment, the compound is in an isotonic, sterile aqueous solution of 0.9 mg / ml sodium chloride.

[0274] B.B7H3 COBRA CURE FOR CANCER In some embodiments, the disclosure provides a method for treating a solid cancer in a patient in need of such treatment by administering to the patient a therapeutically effective amount of a polypeptide prodrug comprising at least one anti-B7H3 single domain antigen binding domain described herein. In some embodiments, the polypeptide comprises, from N-terminus to C-terminus, a) a first single domain antigen binding domain (sdABD) that binds a first human tumor target antigen (TTA), which is B7H3, b) a first domain linker, and c) a constrained Fv domain comprising i) a first variable heavy domain comprising vhCDR1, vhCDR2, and vhCDR3, ii) a constrained non-cleavable linker (CNCL), and a first variable light domain comprising vlCDR1, vlCDR2, and vlCDR3, wherein CNCL is a linker between the first variable heavy domain and the first variable light domain. a constrained Fv domain located between the first variable heavy domain and the first variable light domain, preventing the first variable heavy domain and the first variable light domain from interacting to form an active Fv capable of binding to CD3; d) a second domain linker; e) a second sdABD that binds to a second human TTA; f) a cleavable linker (CL); g) a pseudo-Fv domain comprising i) a first pseudo-variable light domain, ii) a non-cleavable linker (NCL), and iii) a first pseudo-variable heavy domain; h) a third domain linker; and i) a third sdABD that binds to human serum albumin. The first variable heavy domain and the first variable light domain are capable of binding to human CD3, but the constrained Fv domain does not bind to CD3. The polypeptide does not bind to CD3 when the CL is intact.

[0275] In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is selected from SEQ ID NO:147 (Pro225), SEQ ID NO:148 (Pro226), SEQ ID NO:173 (Pro359), SEQ ID NO:257 (Pro373), SEQ ID NO:258 (Pro374), SEQ ID NO:181 (Pro479), SEQ ID NO:182 (Pro480), and SEQ ID NO:183 (Pro495).

[0276] In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 95% identical to SEQ ID NO: 147 (Pro225). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 96% identical to SEQ ID NO: 147 (Pro225). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 97% identical to SEQ ID NO: 147 (Pro225). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 98% identical to SEQ ID NO: 147 (Pro225). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 99% identical to SEQ ID NO: 147 (Pro225). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 99.5% identical to SEQ ID NO: 147 (Pro225). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 25 amino acids compared to SEQ ID NO: 147 (Pro225). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 20 amino acids compared to SEQ ID NO: 147 (Pro225). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 15 amino acids compared to SEQ ID NO: 147 (Pro225). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 10 amino acids compared to SEQ ID NO: 147 (Pro225). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 5 amino acids compared to SEQ ID NO: 147 (Pro225).In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than four amino acid differences compared to SEQ ID NO: 147 (Pro225). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than three amino acid differences compared to SEQ ID NO: 147 (Pro225). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than two amino acid differences compared to SEQ ID NO: 147 (Pro225). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than one amino acid difference compared to SEQ ID NO: 147 (Pro225).

[0277] In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is SEQ ID NO: 147 (Pro225).

[0278] In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 95% identical to SEQ ID NO: 148 (Pro226). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 96% identical to SEQ ID NO: 148 (Pro226). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 97% identical to SEQ ID NO: 148 (Pro226). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 98% identical to SEQ ID NO: 148 (Pro226). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 99% identical to SEQ ID NO: 148 (Pro226). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 99.5% identical to SEQ ID NO: 148 (Pro226). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 25 amino acids compared to SEQ ID NO: 148 (Pro226). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 20 amino acids compared to SEQ ID NO: 148 (Pro226). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 15 amino acids compared to SEQ ID NO: 148 (Pro226). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 10 amino acids compared to SEQ ID NO: 148 (Pro226). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 5 amino acids compared to SEQ ID NO: 148 (Pro226).In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than four amino acid differences compared to SEQ ID NO: 148 (Pro226). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than three amino acid differences compared to SEQ ID NO: 148 (Pro226). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than two amino acid differences compared to SEQ ID NO: 148 (Pro226). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than one amino acid difference compared to SEQ ID NO: 148 (Pro226).

[0279] In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is SEQ ID NO: 148 (Pro226).

[0280] In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 95% identical to SEQ ID NO: 173 (Pro359). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 96% identical to SEQ ID NO: 173 (Pro359). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 97% identical to SEQ ID NO: 173 (Pro359). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 98% identical to SEQ ID NO: 173 (Pro359). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 99% identical to SEQ ID NO: 173 (Pro359). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 99.5% identical to SEQ ID NO: 173 (Pro359). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 25 amino acids compared to SEQ ID NO: 173 (Pro359). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 20 amino acids compared to SEQ ID NO: 173 (Pro359). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 15 amino acids compared to SEQ ID NO: 173 (Pro359). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 10 amino acids compared to SEQ ID NO: 173 (Pro359). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 5 amino acids compared to SEQ ID NO: 173 (Pro359).In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than four amino acid differences compared to SEQ ID NO: 173 (Pro359). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than three amino acid differences compared to SEQ ID NO: 173 (Pro359). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than two amino acid differences compared to SEQ ID NO: 173 (Pro359). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than one amino acid difference compared to SEQ ID NO: 173 (Pro359).

[0281] In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is SEQ ID NO: 173 (Pro359).

[0282] In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 95% identical to SEQ ID NO:257 (Pro373). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 96% identical to SEQ ID NO:257 (Pro373). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 97% identical to SEQ ID NO:257 (Pro373). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 98% identical to SEQ ID NO:257 (Pro373). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 99% identical to SEQ ID NO:257 (Pro373). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 99.5% identical to SEQ ID NO:257 (Pro373). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 25 amino acids compared to SEQ ID NO:257 (Pro373). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 20 amino acids compared to SEQ ID NO:257 (Pro373). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 15 amino acids compared to SEQ ID NO:257 (Pro373). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 10 amino acids compared to SEQ ID NO:257 (Pro373). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 5 amino acids compared to SEQ ID NO:257 (Pro373).In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than four amino acid differences compared to SEQ ID NO:257 (Pro373). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than three amino acid differences compared to SEQ ID NO:257 (Pro373). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than two amino acid differences compared to SEQ ID NO:257 (Pro373). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than one amino acid difference compared to SEQ ID NO:257 (Pro373).

[0283] In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is SEQ ID NO: 257 (Pro373).

[0284] In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 95% identical to SEQ ID NO:258 (Pro374). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 96% identical to SEQ ID NO:258 (Pro374). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 97% identical to SEQ ID NO:258 (Pro374). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 98% identical to SEQ ID NO:258 (Pro374). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 99% identical to SEQ ID NO:258 (Pro374). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 99.5% identical to SEQ ID NO:258 (Pro374). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 25 amino acids compared to SEQ ID NO:258 (Pro374). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 20 amino acids compared to SEQ ID NO:258 (Pro374). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 15 amino acids compared to SEQ ID NO:258 (Pro374). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 10 amino acids compared to SEQ ID NO:258 (Pro374). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 5 amino acids compared to SEQ ID NO:258 (Pro374).In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than four amino acid differences compared to SEQ ID NO:258 (Pro374). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than three amino acid differences compared to SEQ ID NO:258 (Pro374). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than two amino acid differences compared to SEQ ID NO:258 (Pro374). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than one amino acid difference compared to SEQ ID NO:258 (Pro374).

[0285] In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is SEQ ID NO: 258 (Pro374).

[0286] In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 95% identical to SEQ ID NO: 181 (Pro479). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 96% identical to SEQ ID NO: 181 (Pro479). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 97% identical to SEQ ID NO: 181 (Pro479). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 98% identical to SEQ ID NO: 181 (Pro479). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 99% identical to SEQ ID NO: 181 (Pro479). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 99.5% identical to SEQ ID NO: 181 (Pro479). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 25 amino acids compared to SEQ ID NO: 181 (Pro479). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 20 amino acids compared to SEQ ID NO: 181 (Pro479). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 15 amino acids compared to SEQ ID NO: 181 (Pro479). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 10 amino acids compared to SEQ ID NO: 181 (Pro479). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 5 amino acids compared to SEQ ID NO: 181 (Pro479).In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than four amino acid differences compared to SEQ ID NO: 181 (Pro479). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than three amino acid differences compared to SEQ ID NO: 181 (Pro479). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than two amino acid differences compared to SEQ ID NO: 181 (Pro479). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than one amino acid difference compared to SEQ ID NO: 181 (Pro479).

[0287] In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is SEQ ID NO: 181 (Pro479).

[0288] In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 95% identical to SEQ ID NO: 182 (Pro480). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 96% identical to SEQ ID NO: 182 (Pro480). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 97% identical to SEQ ID NO: 182 (Pro480). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 98% identical to SEQ ID NO: 182 (Pro480). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 99% identical to SEQ ID NO: 182 (Pro480). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 99.5% identical to SEQ ID NO: 182 (Pro480). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 25 amino acids compared to SEQ ID NO: 182 (Pro480). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 20 amino acids compared to SEQ ID NO: 182 (Pro480). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 15 amino acids compared to SEQ ID NO: 182 (Pro480). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 10 amino acids compared to SEQ ID NO: 182 (Pro480). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 5 amino acids compared to SEQ ID NO: 182 (Pro480).In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than four amino acid differences compared to SEQ ID NO: 182 (Pro480). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than three amino acid differences compared to SEQ ID NO: 182 (Pro480). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than two amino acid differences compared to SEQ ID NO: 182 (Pro480). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than one amino acid difference compared to SEQ ID NO: 182 (Pro480).

[0289] In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is SEQ ID NO: 182 (Pro480).

[0290] In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 95% identical to SEQ ID NO: 183 (Pro495). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 96% identical to SEQ ID NO: 183 (Pro495). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 97% identical to SEQ ID NO: 183 (Pro495). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 98% identical to SEQ ID NO: 183 (Pro495). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 99% identical to SEQ ID NO: 183 (Pro495). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence at least 99.5% identical to SEQ ID NO: 183 (Pro495). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 25 amino acids compared to SEQ ID NO: 183 (Pro495). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 20 amino acids compared to SEQ ID NO: 183 (Pro495). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 15 amino acids compared to SEQ ID NO: 183 (Pro495). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 10 amino acids compared to SEQ ID NO: 183 (Pro495). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that differs by no more than 5 amino acids compared to SEQ ID NO: 183 (Pro495).In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than four amino acid differences compared to SEQ ID NO: 183 (Pro495). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than three amino acid differences compared to SEQ ID NO: 183 (Pro495). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than two amino acid differences compared to SEQ ID NO: 183 (Pro495). In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain comprises an amino acid sequence that has no more than one amino acid difference compared to SEQ ID NO: 183 (Pro495).

[0291] In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is SEQ ID NO: 183 (Pro495).

[0292] In various embodiments, the cancer is head and neck cancer, e.g., squamous cell carcinoma, by administering to a patient a therapeutically effective amount of a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain described herein. In an exemplary embodiment, the cancer is lung cancer, e.g., non-small cell lung cancer, by administering to a patient a therapeutically effective amount of a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain described herein. In an exemplary embodiment, the cancer is prostate cancer, by administering to a patient a therapeutically effective amount of a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain described herein. In some embodiments, the cancer is colorectal neoplasm, by administering to a patient a therapeutically effective amount of a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain described herein.

[0293] In some embodiments, the method comprises treating head and neck cancer, e.g., squamous cell carcinoma, with a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain described herein. In some embodiments, the method comprises treating head and neck cancer, e.g., squamous cell carcinoma, with a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain selected from SEQ ID NO: 147 (Pro225), SEQ ID NO: 148 (Pro226), SEQ ID NO: 173 (Pro359), SEQ ID NO: 257 (Pro373), SEQ ID NO: 258 (Pro374), SEQ ID NO: 181 (Pro479), SEQ ID NO: 182 (Pro480), and SEQ ID NO: 183 (Pro495). In some embodiments, the method comprises treating head and neck cancer, e.g., squamous cell carcinoma, with Pro225 (SEQ ID NO: 147).

[0294] In some embodiments, the method comprises treating lung cancer, e.g., non-small cell lung cancer, with a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain described herein. In some embodiments, the method comprises treating lung cancer, e.g., non-small cell lung cancer, with a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain selected from SEQ ID NO: 147 (Pro225), SEQ ID NO: 148 (Pro226), SEQ ID NO: 173 (Pro359), SEQ ID NO: 257 (Pro373), SEQ ID NO: 258 (Pro374), SEQ ID NO: 181 (Pro479), SEQ ID NO: 182 (Pro480), and SEQ ID NO: 183 (Pro495). In some embodiments, the method comprises treating lung cancer, e.g., non-small cell lung cancer, with Pro225 (SEQ ID NO: 147).

[0295] In some embodiments, the method comprises treating prostate cancer with a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain described herein. In some embodiments, the method comprises treating prostate cancer with a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain selected from SEQ ID NO: 147 (Pro225), SEQ ID NO: 148 (Pro226), SEQ ID NO: 173 (Pro359), SEQ ID NO: 257 (Pro373), SEQ ID NO: 258 (Pro374), SEQ ID NO: 181 (Pro479), SEQ ID NO: 182 (Pro480), and SEQ ID NO: 183 (Pro495). In some embodiments, the method comprises treating prostate cancer with Pro225 (SEQ ID NO: 147).

[0296] In some embodiments, the method comprises treating colorectal cancer with a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain described herein. In some embodiments, the method comprises treating colorectal cancer with a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain selected from SEQ ID NO: 147 (Pro225), SEQ ID NO: 148 (Pro226), SEQ ID NO: 173 (Pro359), SEQ ID NO: 257 (Pro373), SEQ ID NO: 258 (Pro374), SEQ ID NO: 181 (Pro479), SEQ ID NO: 182 (Pro480), and SEQ ID NO: 183 (Pro495). In some embodiments, the method comprises treating colorectal cancer with Pro225 (SEQ ID NO: 147).

[0297] In exemplary embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is administered to a subject with non-small cell lung cancer that has progressed during or after treatment with platinum-based chemotherapy and anti-PDx therapy (e.g., anti-PD1 therapy). In some embodiments, the subject has non-small cell lung cancer with a B7H3 mutation or ALK rearrangement. In some embodiments, the cancer has progressed after treatment with platinum-based chemotherapy plus available B7H3 or ALK targeted therapy. In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is selected from SEQ ID NO: 147 (Pro225), SEQ ID NO: 148 (Pro226), SEQ ID NO: 173 (Pro359), SEQ ID NO: 257 (Pro373), SEQ ID NO: 258 (Pro374), SEQ ID NO: 181 (Pro479), SEQ ID NO: 182 (Pro480), and SEQ ID NO: 183 (Pro495). In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is SEQ ID NO: 147 (Pro225).

[0298] In exemplary embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is administered to a subject with head and neck squamous cell carcinoma that has progressed during or after treatment with anti-PDx (e.g., anti-PD1 therapy) and platinum-based chemotherapy for metastatic or recurrent disease. In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is selected from SEQ ID NO: 147 (Pro225), SEQ ID NO: 148 (Pro226), SEQ ID NO: 173 (Pro359), SEQ ID NO: 257 (Pro373), SEQ ID NO: 258 (Pro374), SEQ ID NO: 181 (Pro479), SEQ ID NO: 182 (Pro480), and SEQ ID NO: 183 (Pro495). In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is SEQ ID NO: 147 (Pro225).

[0299] In exemplary embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is administered to a subject with advanced or halted prostate cancer. In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is selected from SEQ ID NO: 147 (Pro225), SEQ ID NO: 148 (Pro226), SEQ ID NO: 173 (Pro359), SEQ ID NO: 257 (Pro373), SEQ ID NO: 258 (Pro374), SEQ ID NO: 181 (Pro479), SEQ ID NO: 182 (Pro480), and SEQ ID NO: 183 (Pro495). In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is SEQ ID NO: 147 (Pro225).

[0300] In some embodiments, progression or cessation of prostate cancer is assessed by measuring the levels of prostate specific antigen (PSA) and / or prostate specific membrane antigen (PSMA) by blood tests and / or positron emission tomography (PET) scans. In some embodiments, the success of the treatment is assessed based on a 50%, about 50%, or more than 50% reduction in the level of PSA compared to the original level of PSA when the treatment has been administered to the patient for at least 2 weeks, at least 3 weeks, at least 4 weeks, at least 5 weeks, at least 6 weeks, at least 7 weeks, at least 8 weeks, at least 9 weeks, at least 10 weeks, at least 11 weeks, at least 12 weeks, at least 13 weeks, at least 14 weeks, at least 16 weeks, at least 17 weeks, at least 18 weeks, at least 19 weeks, at least 20 weeks, at least 25 weeks, at least 30 weeks, at least 35 weeks, at least 40 weeks, at least 45 weeks, at least 50 weeks, at least 55 weeks, or at least 60 weeks. In some embodiments, the success of the treatment is assessed based on a 50%, about 50%, or more than 50% decrease in the level of PSA compared to the original level of PSA when the treatment has been administered to the patient for at least 1 month, at least 2 months, at least 3 months, at least 4 months, at least 5 months, at least 6 months, at least 7 months, at least 8 months, at least 9 months, at least 10 months, at least 11 months, at least 12 months, at least 13 months, at least 14 months, or at least 15 months. In some embodiments, the success of the treatment is assessed based on a 50%, about 50%, or more than 50% decrease in the level of PSA compared to the original level of PSA when the treatment has been administered to the patient for about 6 months.In some embodiments, the reduction in PSA levels is about 50%, about 51%, about 52%, about 53%, about 54%, about 55%, about 56%, about 57%, about 58%, about 59%, about 60%, about 61%, about 62%, about 63%, about 64%, about 65%, about 66%, about 67%, about 68%, about 69%, about 70%, about 71%, about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 78%, about 79%, 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 The PSA level is about 71%, about 72%, about 73%, about 74%, about 75%, about 76%, about 77%, about 78%, about 79%, 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%, about 99%, or about 100%. In some embodiments, the success of the treatment is evaluated based on a 50%, about 50%, or more than 50% decrease in the level of PSA compared to the original level of PSA, when the 50%, about 50%, or more than 50% decrease in the level of PSA is sustained for up to about 6 months or more than 6 months after treatment. In some embodiments, the success of the treatment is evaluated based on the absence or reduction of bone metastases. In some embodiments, the reduction in detectable metastases, the stability of the size of metastases, and / or the increase in the size of metastases is determined by PET scan. In some embodiments, the success of the treatment is determined based on a reduction in tumor size of at least 20%, at least 21%, at least 22%, at least 23%, at least 24%, at least 25%, at least 26%, at least 27%, at least 28%, at least 29%, at least 30%, at least 31%, at least 32%, at least 33%, at least 34%, at least 35%, at least 36%, at least 37%, at least 38%, at least 39%, at least 40%, at least 41%, at least 42%, at least 43%, at least 44%, at least 45%, at least 46%, at least 47%, at least 48%, at least 49%, at least 50% compared to the original size before or at the start of the treatment.

[0301] In exemplary embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is administered to a subject with prostate cancer that has progressed during or after treatment with platinum-based chemotherapy and anti-PDx therapy (e.g., anti-PD1 therapy). In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is selected from SEQ ID NO: 147 (Pro225), SEQ ID NO: 148 (Pro226), SEQ ID NO: 173 (Pro359), SEQ ID NO: 257 (Pro373), SEQ ID NO: 258 (Pro374), SEQ ID NO: 181 (Pro479), SEQ ID NO: 182 (Pro480), and SEQ ID NO: 183 (Pro495). In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is SEQ ID NO: 147 (Pro225).

[0302] In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is administered simultaneously to a patient in combination with other cancer treatments, such as, but not limited to, radiation. In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is selected from SEQ ID NO: 147 (Pro225), SEQ ID NO: 148 (Pro226), SEQ ID NO: 173 (Pro359), SEQ ID NO: 257 (Pro373), SEQ ID NO: 258 (Pro374), SEQ ID NO: 181 (Pro479), SEQ ID NO: 182 (Pro480), and SEQ ID NO: 183 (Pro495). In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is SEQ ID NO: 147 (Pro225).

[0303] In various embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is administered to a subject with colorectal cancer having a wild-type K-Ras genotype. In some embodiments, the subject with a wild-type K-Ras genotype has progressed during or after chemotherapy, for example, irinotecan and / or oxaliplatin-based chemotherapy. In some embodiments, the subject with a wild-type K-Ras genotype is ineligible for irinotecan and / or oxaliplatin-based chemotherapy. In some embodiments, the subject with a wild-type K-Ras genotype has relapsed or is refractory to at least one prior systemic therapy comprising an anti-B7H3 antibody. In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is selected from SEQ ID NO: 147 (Pro225), SEQ ID NO: 148 (Pro226), SEQ ID NO: 173 (Pro359), SEQ ID NO: 257 (Pro373), SEQ ID NO: 258 (Pro374), SEQ ID NO: 181 (Pro479), SEQ ID NO: 182 (Pro480), and SEQ ID NO: 183 (Pro495). In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is SEQ ID NO: 147 (Pro225).

[0304] In various embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is administered to a subject with colorectal cancer having a mutant K-Ras genotype. In some embodiments, the subject with a mutant K-Ras genotype has progressed during or after chemotherapy, for example, irinotecan and / or oxaliplatin-based chemotherapy. In some embodiments, the subject with a mutant K-Ras genotype is ineligible for irinotecan and / or oxaliplatin-based chemotherapy. In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is selected from SEQ ID NO: 147 (Pro225), SEQ ID NO: 148 (Pro226), SEQ ID NO: 173 (Pro359), SEQ ID NO: 257 (Pro373), SEQ ID NO: 258 (Pro374), SEQ ID NO: 181 (Pro479), SEQ ID NO: 182 (Pro480), and SEQ ID NO: 183 (Pro495). In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is SEQ ID NO: 147 (Pro225).

[0305] In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is administered to a subject with an ECOG performance status of less than or equal to 1. In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is administered to a subject with an ECOG performance status of less than or equal to 0.5, 0.6, 0.7, 0.8, 0.9, 1.0, 1.1, 1.2, 1.3, 1.4, or 1.5. In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is selected from SEQ ID NO:147 (Pro225), SEQ ID NO:148 (Pro226), SEQ ID NO:173 (Pro359), SEQ ID NO:257 (Pro373), SEQ ID NO:258 (Pro374), SEQ ID NO:181 (Pro479), SEQ ID NO:182 (Pro480), and SEQ ID NO:183 (Pro495). In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is SEQ ID NO: 147 (Pro225).

[0306] In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is administered to a subject with disease evaluable according to RECIST v1.1 criteria and documented by CT and / or MRI. In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is selected from SEQ ID NO: 147 (Pro225), SEQ ID NO: 148 (Pro226), SEQ ID NO: 173 (Pro359), SEQ ID NO: 257 (Pro373), SEQ ID NO: 258 (Pro374), SEQ ID NO: 181 (Pro479), SEQ ID NO: 182 (Pro480), and SEQ ID NO: 183 (Pro495). In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is SEQ ID NO: 147 (Pro225).

[0307] In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is administered to a subject who has undergone immune checkpoint therapy prior to administration. In some embodiments, the subject who has undergone immune checkpoint therapy prior to administration has resolved immune-related toxicity of the checkpoint inhibitor to either grade ≦1 or baseline. In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is selected from SEQ ID NO: 147 (Pro225), SEQ ID NO: 148 (Pro226), SEQ ID NO: 173 (Pro359), SEQ ID NO: 257 (Pro373), SEQ ID NO: 258 (Pro374), SEQ ID NO: 181 (Pro479), SEQ ID NO: 182 (Pro480), and SEQ ID NO: 183 (Pro495). In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is SEQ ID NO: 147 (Pro225).

[0308] In some embodiments, a polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is administered to a subject with symptomatic central nervous system metastases who has been treated and has been asymptomatic for at least 14 days, with no concomitant concomitant leptomeningeal disease / spinal cord compression, hi some embodiments, the subject has been asymptomatic for at least 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 128, 19, 20, 21, 28, 35, 42, 49 or more days. In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is selected from SEQ ID NO: 147 (Pro225), SEQ ID NO: 148 (Pro226), SEQ ID NO: 173 (Pro359), SEQ ID NO: 257 (Pro373), SEQ ID NO: 258 (Pro374), SEQ ID NO: 181 (Pro479), SEQ ID NO: 182 (Pro480), and SEQ ID NO: 183 (Pro495). In some embodiments, the polypeptide prodrug comprising an anti-B7H3 single domain antigen binding domain is SEQ ID NO: 147 (Pro225).

[0309] As will be understood by those skilled in the art, the choice of dosage is within the ability of a physician experienced in cancer treatment.In an exemplary embodiment, the subject is administered the compound of the present invention at a dose of about 0.3 μg / kg to about 30 μg / kg.In an exemplary embodiment, the compound of the present invention is administered to the subject about once a week.

[0310] As will be appreciated by those skilled in the art, the duration of treatment is within the capabilities of a physician experienced in cancer treatment.

[0311] As will be appreciated by those skilled in the art, the vehicle in which the compounds of the present invention are formulated may be any useful, safe, non-toxic, pharma- ceutically acceptable formulation. In an exemplary embodiment, the compound is in an isotonic, sterile aqueous solution of 0.9 mg / ml sodium chloride. EXAMPLES

[0312] A. Example 1: Construction and purification of Pro constructs Transfection Each protein (e.g. single protein for formats 1, 2 and 4) or pair of constructs (format 3) was expressed from a separate expression vector (pcdna3.4 derivative). Equal amounts of plasmid DNA encoding half-cobra or single-chain construct pairs were mixed and transfected into Expi293 cells according to the manufacturer's transfection protocol. Conditioned media was harvested 5 days after transfection by centrifugation (6000 rpm x 25') and filtration (0.2uM filter). Protein expression was confirmed by SDS-PAGE. Constructs were purified and the final buffer composition was 25mM citrate, 75mM arginine, 75mM NaCl, 4% sucrose, pH 7. Final preparations were stored at -80°C.

[0313] Activation of MMP9 Recombinant human (rh) MMP9 was activated according to the following protocol: Recombinant human MMP-9 (R&D #911-MP-010) was at 0.44 mg / ml (4.7 uM). p-Aminophenylmercuric acetate (APMA) (Sigma) is prepared at a stock concentration of 100 mM in DMSO. Assay buffer is 50 mM Tris, pH 7.5, 10 mM CaCl2, 150 mM NaCl, 0.05% Brij®-35.

[0314] - Dilute rhMMP9 to approximately 100ug / ml in assay buffer (25ul hMMP9 + 75uL assay buffer)

[0315] - Add p-aminophenylmercuric acetate (APMA) from a 100 mM stock in DMSO to a final concentration of 1 mM (1 uL to 100 uL)

[0316] - Incubate at 37 °C for 24 hours

[0317] - Dilute MMP9 to 10ng / ul (add 900ul of assay buffer to 100ul of activation solution)

[0318] The concentration of activated rhMMP9 is approximately 100 nM.

[0319] Cleavage of constructs for TDCC assay To cleave the constructs, 100ul of protein sample at a concentration of 1mg / ml (10.5uM) in formulation buffer (25mM citric acid, 75mM L-arginine, 75mM NaCl, 4% sucrose) was supplemented with CaCl2 up to 10mM. Activated rhMMP9 was added to a concentration of 20-35nM. Samples were incubated overnight (16-20h) at room temperature. Completeness of cleavage was verified using SDS PAGE (10-20% TG, TG running buffer, 200v, 1h). Samples were typically 98% cleaved.

[0320] B. Example 2: T-cell dependent cytotoxicity (TDCC) assay Firefly luciferase-transduced HT-29 cells were grown to approximately 80% confluency and detached with Versene (0.48 mM EDTA in PBS-Ca-Mg). Cells were centrifuged and resuspended in TDCC medium (5% heat-inactivated FBS in RPMI1640 containing HEPES, GlutaMax, sodium pyruvate, non-essential amino acids and β-mercaptoethanol). Purified human Pan-T cells were thawed, centrifuged and resuspended in TDCC medium.

[0321] Co-cultures of HT-29_Luc and T cells were added to 384-well cell culture plates. Serially diluted COBRA was then added to the co-cultures and incubated at 37°C for 48 hours. Finally, an equal volume of SteadyGlo luciferase assay reagent was added to the plates and incubated for 20 minutes. The plates were read on a Perkin Elmer Envision with an exposure time of 0.1 seconds per well. Total luminescence was recorded and the data was analyzed with GraphPad Prism7.

[0322] C. Example 3: General Protocol Design of In Vivo Adoptive T Cell Transfer Efficacy Model These protocols were used for many of the experiments shown. Tumor cells were implanted subcutaneously (SC) into the right flank of NSG (NOD.Cg-Prkdcscid Il2rgtm1Wjl / SzJ) mice (The Jackson Laboratory, Cat. No. 005557) and grew to approximately 200 mm 3 The mice were allowed to grow until established tumors with an average volume of 1000 μg / ml were achieved. In parallel, human T cells were cultured in T cell medium (X-VIVO 15 [Lonza, Cat. No. 04-418Q], 5% human serum, 1% penicillin / streptomycin, 0.01 mM 2-mercaptoethanol) in G-Rex100M gas-permeable flasks (Wilson Wolf, Cat. No. 81100S) with MACSiBeads from a T cell activation / expansion kit (Miltenyi, Cat. No. 130-091-441) for approximately 10 days, supplemented with recombinant human IL-2 protein. Tumor growth and human T cell activation / expansion in mice were regulated, whereby on day 0 of the experiment, mice were randomly divided into groups (N=6) based on tumor size, and then each was treated with 2.5×10 6 Mice were injected intravenously (IV) with 1000 mg of cultured human T cells and given an initial dose of COBRA or a control molecule. Mice were administered 7 doses every 3 days (days 0, 3, 6, 9, 12, 15, and 18), and then the tumors were grown until they reached 2000 mm 3 Treatment was continued for an additional 2-3 weeks until a tumor volume exceeding 100 μg / kg was reached or the study was terminated. Tumor volumes were measured every 3 days.

[0323] D. Example 4: In vivo activity with EGFR / MMP9 half-COBRA versus Pro77 and Pro53 5×10 6 or 5 x 10 LoVo cells 6HT29 cells were implanted subcutaneously into the right flank of NSG (NOD.Cg-Prkdcscid Il2rgtm1Wjl / SzJ) mice (The Jackson Laboratory, Catalog No. 005557) and allowed to grow until tumors were established. In parallel, human T cells were cultured in T cell medium (X-VIVO 15 [Lonza, Catalog No. 04-418Q], 5% human serum, 1% penicillin / streptomycin, 0.01 mM 2-mercaptoethanol) in G-Rex100M gas-permeable flasks (Wilson Wolf, Catalog No. 81100S) with MACSiBeads from a T cell activation / expansion kit (Miltenyi, Catalog No. 130-091-441) for 10 days, supplemented with recombinant human IL-2 protein. To control for mouse tumor growth and human T cell activation / expansion, mice were randomly divided into groups (N=6) based on tumor size on day 0 of the experiment, and then each group was assigned 2.5×10 6 Mice were injected intravenously (IV) with 1000 mg of cultured human T cells and given an initial dose of COBRA or a control molecule. Mice were administered 7 doses every 3 days (days 0, 3, 6, 9, 12, 15, and 18), and then the tumors were grown until they reached 2000 mm 3 Tumors were administered until a tumor volume exceeding 100 μg / kg (mpk) was reached or the study was terminated. Groups received 0.2 mg / kg (mpk) of anti-EGFR x CD3 positive control, Pro51 bispecific antibody (bsAb), 0.5 mpk of negative control, anti-hen egg white lysozyme (HEL) x CD3 bsAb Pro98, 0.5 mpk each of anti-EGFR half-COBRA with MMP9 cleavable linker vs. Pro77 and Pro53, or 0.5 mpk each of anti-EGFR half-COBRA with non-cleavable (NCL) linker vs. Pro74 and Pro72. Tumor volumes were measured every 3 days.

[0324] E. Example 5: EGFR / MMP9 In vivo activity with COBRA Pro140. 5×10 6 or 5 x 10 LoVo cells 6HT29 cells were implanted subcutaneously into the right flank of NSG (NOD.Cg-Prkdcscid Il2rgtm1Wjl / SzJ) mice (The Jackson Laboratory, Catalog No. 005557) and allowed to grow until tumors were established. In parallel, human T cells were cultured in T cell medium (X-VIVO 15 [Lonza, Catalog No. 04-418Q], 5% human serum, 1% penicillin / streptomycin, 0.01 mM 2-mercaptoethanol) in G-Rex100M gas-permeable flasks (Wilson Wolf, Catalog No. 81100S) with MACSiBeads from a T cell activation / expansion kit (Miltenyi, Catalog No. 130-091-441) for 10 days, supplemented with recombinant human IL-2 protein. To control for mouse tumor growth and human T cell activation / expansion, mice were randomly divided into groups (N=6) based on tumor size on day 0 of the experiment, and then each group was assigned 2.5×10 6 Mice were injected intravenously (IV) with 1000 mg of cultured human T cells and given an initial dose of COBRA or a control molecule. Mice were administered 7 doses every 3 days (days 0, 3, 6, 9, 12, 15, and 18), and then the tumors were grown until they reached 2000 mm 3 Tumor volumes were measured every 3 days. Tumor volumes were measured every 3 days.

[0325] F. Example 6: EGFR / MMP9 In vivo activity with COBRA Pro186. 5×10 6HT29 cells were implanted subcutaneously into the right flank of NSG (NOD.Cg-Prkdcscid Il2rgtm1Wjl / SzJ) mice (The Jackson Laboratory, Catalog No. 005557) and allowed to grow until tumors were established. In parallel, human T cells were cultured in T cell medium (X-VIVO 15 [Lonza, Catalog No. 04-418Q], 5% human serum, 1% penicillin / streptomycin, 0.01 mM 2-mercaptoethanol) in G-Rex100M gas-permeable flasks (Wilson Wolf, Catalog No. 81100S) with MACSiBeads from a T cell activation / expansion kit (Miltenyi, Catalog No. 130-091-441) for 10 days, supplemented with recombinant human IL-2 protein. To control for mouse tumor growth and human T cell activation / expansion, mice were randomly divided into groups (N=6) based on tumor size on day 0 of the experiment, and then each group was assigned 2.5×10 6 Mice were injected intravenously (IV) with 1000 mg of cultured human T cells and given an initial dose of COBRA or a control molecule. Mice were administered 7 doses every 3 days (days 0, 3, 6, 9, 12, 15, and 18), and then the tumors were grown until they reached 2000 mm 3 Tumors were dosed until a tumor volume exceeding 0.1 mg / kg (mpk) was reached or the study was terminated. Groups were dosed with 0.1 mg / kg (mpk) of anti-EGFRxCD3 positive control, Pro51 bispecific antibody (bsAb), 0.3 mpk of anti-EGFR COBRA Pro214 with a non-cleavable (NCL) control linker, 0.1 or 0.3 mpk of anti-EGFR COBRA Pro140 with an MMP9 cleavable linker, or 0.1 or 0.3 mpk of anti-EGFR COBRA Pro186 with an MMP9 cleavable linker. Tumor volumes were measured every 3 days.

[0326] G. Example 7: Successful humanization of anti-EGFR sequences The results are shown below. [Table 2]

[0327] These results indicate both successful humanization of the EGFR binding domain and strong avidity for the target EGFR when two binding sites are present on the molecule.

[0328] H. Example 8: Successful humanization of EpCAM sdABD The results are shown below. [Table 3]

[0329] Together these results indicate that the EpCAM-binding domain was successfully humanized.

[0330] I. Example 9: Dose escalation and expansion study of EGFR COBRA Pro186 in patients with advanced cancer. The objective of this study is to evaluate the safety, tolerability, pharmacokinetics (PK), pharmacodynamics, and preliminary antitumor activity of EGFR COBRA Pro186 (SEQ ID NO: 145) in patients with unresectable locally advanced or metastatic cancer. The study will characterize the safety, dose-limiting toxicities (DLTs), and maximum tolerated dose / recommended phase 2 dose (MTD / RP2D) of Pro186. Dose escalation will be performed in a 1+3 design, followed by a 3+3 design in patients with advanced solid tumors. Once the MTD / RP2D is determined, a cohort expansion phase will be enrolled to further characterize safety and initial antitumor activity in patients with squamous cell carcinoma of the head and neck, non-small cell lung cancer, or colorectal neoplasms.

[0331] Approximately 68 participants will be enrolled in the study. The study will follow a 3+3 protocol followed by an initial 1+3 protocol to identify the recommended phase II dose (RP2D) / maximum tolerated dose (MDT). Briefly, in the 3+3 protocol, the study will progress in cohorts of three patients receiving escalating dose levels. Escalation will be terminated if at least two or three patients experience dose-limiting toxicity (DTL). The MDT will be determined as the highest possible dose at which 1 / 6 or less patients experience DTL. Thus, escalating doses of Pro186 will be administered to patients with histologically proven unresectable locally advanced or metastatic solid tumors believed to express EGFR.

[0332] Once RP2D has been determined, we will select patients with: (i) non-small cell lung cancer (NSCLC) who have progressed during or after treatment with platinum-based chemotherapy and anti-PDx therapy, and patients with NSCLC harboring EGFR mutations or ALK rearrangements who have progressed after treatment with platinum-based chemotherapy plus available EGFR- or ALK-targeted therapy; (ii) patients with anti-PDx for metastatic or recurrent disease (unless otherwise eligible, e.g., chemotherapy is not working, PD-L1 Cohort expansion will be performed to further characterize the safety and initial antitumor activity of Pro186 in patients with squamous cell carcinoma of the head and neck who have progressed during or after treatment with or are ineligible for platinum-based chemotherapy (patients with CPS<1) and platinum-based chemotherapy (unless ineligible / intolerant for platinum-based chemotherapy), and (iii) in patients with colorectal cancer who have (a) K-Ras wild-type patients who have progressed during or after or are ineligible for both irinotecan- and oxaliplatin-based chemotherapy and K-Ras wild-type patients who have relapsed or are resistant to at least one prior systemic therapy including an anti-EGFR antibody, or (b) K-Ras mutations and have progressed during or after or are ineligible for both irinotecan- and oxaliplatin-based chemotherapy.

[0333] Inclusion criteria for the study require that patients have an Eastern Cooperative Oncology Group (ECOG) performance status of ≤1 and disease evaluable according to RECIST v1.1 criteria and documented by CT and / or MRI. Patients must allow for the acquisition of current archived tumor samples, either mass or unstained slides. Patients must consent to paired tumor biopsies, one at the screening period and one during the first cycle of treatment. Patients must also have acceptable laboratory parameters and adequate organ reserve. Inclusion criteria also include that patients who received prior immune checkpoint therapy prior to enrollment must have resolved immune-related toxicity of checkpoint inhibitors to either grade ≤1 or baseline. Patients with symptomatic CNS metastases must have been treated and asymptomatic for ≥14 days and must not have concomitant leptomeningeal disease / spinal cord compression.

[0334] Potential participants will be excluded using the following criteria: patients with a known history of autoimmune disease, with certain exceptions; patients with clinically significant cardiovascular / vascular disease; patients with clinically significant inflammatory gastrointestinal disorders; patients with clinically significant lung injury; major surgery or trauma within 8 weeks from the date of first study dose; unhealed wounds from surgery or trauma; treatment with radiation therapy less than 2 weeks from the date of first study dose; and inflammatory processes that have not resolved for more than 4 weeks from the date of first study dose.

[0335] The primary endpoint of the study includes the incidence of treatment-emergent adverse events from the time of Pro186 administration through the end of treatment or 28 days after the last dose of Pro186, based on signs, symptoms, physical examination findings, and / or laboratory results. Secondary endpoints for the study also include the pharmacokinetics of Pro186 administration, as assessed by plasma concentrations. Secondary endpoints also include the immunogenicity of Pro186, as assessed by testing for plasma anti-Pro186 antibodies, for up to 54 weeks after initiation of the study. Secondary endpoints for the study also include radiographic antitumor activity, as assessed using both the traditional Response Evaluation Criteria in Solid Tumors version 1.1 (RECIST v1.1) and modified RECIST v1.1, for up to 54 weeks after initiation of the study.

[0336] The study is evaluating Pro186 in (i) patients with non-small cell lung cancer (NSCLC) who have progressed during or after treatment with platinum-based chemotherapy and anti-PDx therapy, and in patients with NSCLC harboring EGFR mutations or ALK rearrangements who have progressed after treatment with platinum-based chemotherapy plus available EGFR- or ALK-targeted therapy, and (ii) patients with anti-PDx for metastatic or recurrent disease (unless eligible, e.g., chemotherapy is not working, PD-L1 The study is expected to show that Pro186 has antitumor activity in patients with squamous cell carcinoma of the head and neck who have progressed during or after treatment with (patients with a CPS<1) and platinum-based chemotherapy (unless ineligible / intolerant for platinum-based chemotherapy), and (iii) in patients with (a) K-Ras wild-type patients who have progressed during or after or are ineligible for both irinotecan and oxaliplatin-based chemotherapy and K-Ras wild-type patients who have relapsed or are resistant to at least one prior systemic therapy including an anti-EGFR antibody, or (b) colorectal cancer patients who have a K-Ras mutation and have progressed during or after or are ineligible for both irinotecan and oxaliplatin-based chemotherapy. The study is also expected to show that Pro186 has lower toxicity than previously tested anti-CD3 bispecific immunotherapies.

[0337] J. Example 10: Dose escalation and expansion study of B7H3 COBRA Pro225 in patients with unresectable locally advanced or metastatic cancer. The objective of this study is to evaluate the safety and tolerability of B7H3 COBRA Pro225 in participants with unresectable locally advanced or metastatic cancer who have failed or cannot tolerate standard treatments. The study consists of a dose escalation phase to determine the recommended dose of Pro225 for the cohort expansion phase, where the safety and initial efficacy of Pro225 will be further defined.

[0338] Approximately 186 participants at least 18 years of age will be enrolled in the study. Study inclusion criteria will require patients to have an Eastern Cooperative Oncology Group (ECOG) performance status of ≦1 and disease evaluable according to RECIST v1.1 criteria and documented by CT and / or MRI, excluding participants with PC with bone metastases only. Dose escalation will be performed in participants with histologically or pathologically confirmed unresectable locally advanced or metastatic cancer, using a Bayesian Optimal Interval (BOIN) design. A dose escalation phase will be used to determine a recommended phase II dose (RP2D) that does not exceed the maximum tolerated dose (MDT) as confirmed according to the BOIN design.

[0339] Participants will be eligible for the cohort expansion phase of the study if they have histologically proven unresectable locally advanced or metastatic malignant neoplasms. Briefly, patients will be treated with Pro225 for up to 14 treatment cycles with a maximum total of 56 doses of Pro225. Each treatment cycle will last 28 days. Participants will be treated with Pro225 until disease progression, unacceptable toxicity, or withdrawal from the study. Participants' cancer will be treated by physicians according to the physician's usual clinical practice. After the last dose of study drug, participants will be followed up every 12 weeks for survival for at least 52 weeks.

[0340] Prospective participants will be excluded using the following criteria: history of known autoimmune disease, major surgery or trauma within 8 weeks prior to the first dose of Pro225, unhealed wounds from surgery or trauma, persistent or active infection of at least grade 2, oxygen saturation >92% in room air at screening or during pre-C1D1 evaluation, inflammatory process that has not resolved for >4 weeks prior to the first dose of study drug. Participants with low-grade chronic inflammatory processes, such as radiation-induced interstitial pneumonia, will be excluded regardless of vaccination with any live virus vaccine within 4 weeks prior to starting study drug or any other vaccine within 2 weeks during that period. Annual inactivated influenza vaccination will be permitted, as well as known hypersensitivity to Pro225 or any of its excipients.

[0341] The primary endpoint of the study includes the incidence of treatment-emergent adverse events from the time of Pro225 administration through approximately 37 months. The primary endpoint of the study also includes the incidence of dose-limiting toxicities (DLTs) from the time of first Pro225 administration through day 28 of cycle 1. The primary endpoint of the study also includes determination of the maximum tolerated dose (MTD) of Pro225, selected as the dose at which an isotonic estimate of toxicity rate from the time of Pro225 administration through approximately 37 months is closest to a target toxicity rate of 30%.

[0342] Secondary evaluation criteria for the study included: - Proportion of participants achieving confirmed overall response rate (ORR) based on modified Response Evaluation Criteria in Solid Tumors version 1.1 (mRECIST 1.1), from time of Pro225 administration up to approximately 37 months. Duration of response (DOR) per mRECIST 1.1 was approximately 37 months from the time of Pro225 administration. ● Progression-free survival (PFS), from the start of first dose to disease progression or death, whichever occurs first, up to approximately 37 months. Overall survival (OS), from start of first dose of study drug to death, up to approximately 37 months. • Proportion of participants with prostate cancer (PC) who experienced a ≥50% decrease in prostate-specific antigen (PSA) from baseline up to 6 months. Maximum observed plasma concentrations (Cmax) of Pro225, Cycle 1, Day 1: 0 (pre-dose) to 48 hours, Days 8, 15: 0 to 6 hours, Day 22: 0 to 1 hour; Cycle 2, Days 1, 8, 15, 22: 0 to 1 hour; Cycle 3, Day 1: 0 to 4 hours; Days 8, 15, 22: 0 to 1 hour; Cycle 5, Day 1: 0 to 1 hour (post-dose) (each cycle = 28 days). Cmax of Pro225 is reported. Area under the plasma concentration-time curve (AUC) of Pro225, Cycle 1, Day 1: 0 (predose) to 48 hours, Days 8, 15: 0 to 6 hours, Day 22: 0 to 1 hour; Cycle 2, Days 1, 8, 15, 22: 0 to 1 hour; Cycle 3, Day 1: 0 to 4 hours; Days 8, 15, 22: 0 to 1 hour; Cycle 5, Day 1: 0 to 1 hour (postdose) (each cycle = 28 days). The AUC of Pro225 is reported. Observed times to maximum plasma concentration (tmax) of Pro225: Cycle 1, Day 1: 0 (pre-dose) to 48 hours, Days 8, 15: 0 to 6 hours, Day 22: 0 to 1 hour; Cycle 2, Days 1, 8, 15, 22: 0 to 1 hour; Cycle 3, Day 1: 0 to 4 hours; Days 8, 15, 22: 0 to 1 hour; Cycle 5, Day 1: 0 to 1 hour (post-dose) (each cycle = 28 days). tmax of Pro225 is reported. Terminal elimination half-life (t1 / 2) of Pro225, Cycle 1, Day 1: 0 (predose) to 48 hours, Days 8, 15: 0 to 6 hours, Day 22: 0 to 1 hour; Cycle 2, Days 1, 8, 15, 22: 0 to 1 hour; Cycle 3, Day 1: 0 to 4 hours; Days 8, 15, 22: 0 to 1 hour; Cycle 5, Day 1: 0 to 1 hour (postdose) (each cycle = 28 days). The t1 / 2 of Pro225 is reported. Systemic clearance (CL) of Pro225, Cycle 1, Day 1: 0 (pre-dose) to 48 hours, Days 8, 15: 0 to 6 hours, Day 22: 0 to 1 hour; Cycle 2, Days 1, 8, 15, 22: 0 to 1 hour; Cycle 3, Day 1: 0 to 4 hours; Days 8, 15, 22: 0 to 1 hour; Cycle 5, Day 1: 0 to 1 hour (post-dose) (each cycle = 28 days). CL of Pro225 will be reported. Volume of distribution (Vss) at steady state following intravenous administration of Pro225: Day 1 of Cycle 1: 0 (pre-dose) to 48 hours, Days 8 and 15: 0 to 6 hours, Day 22: 0 to 1 hour; Days 1, 8, 15 and 22 of Cycle 2: 0 to 1 hour; Day 1 of Cycle 3: 0 to 4 hours; Days 8, 15, 22: 0 to 1 hour; Day 1 of Cycle 5: 0 to 1 hour (post-dose) (each cycle = 28 days). VSS for Pro225 will be reported. Percentage of participants producing cytokines in response to Pro225, Cycles 1 to 5: Pre-dose (each cycle = 28 days). The percentage of participants producing cytokines in response to Pro225 will be reported. Percentage of participants positive for induction of anti-drug antibodies (ADA) to Pro225, Cycles 1 to 5: Pre-dose (each cycle = 28 days). The percentage of participants positive for induction of ADA to Pro225 will be reported.

[0343] This study is expected to show that Pro225 has antitumor activity, at least in patients with prostate cancer. This study is also expected to show that Pro225 has lower toxicity than previously tested anti-CD3 bispecific immunotherapies.

Claims

1. 1. A composition for use in a method of treating cancer in a patient in need thereof, the composition comprising a polypeptide, the polypeptide comprising, from N-terminus to C-terminus: a) a first single domain antigen binding domain (sdABD) that binds to a first human tumor target antigen (TTA); b) a first domain linker; and c) i) a first variable heavy domain comprising a vhCDR1, a vhCDR2 and a vhCDR3; ii) a constrained non-cleavable linker (CNCL), and iii) a first variable light domain comprising vlCDR1, vlCDR2 and vlCDR3; wherein the CNCL is disposed between the first variable heavy domain and the first variable light domain, and prevents the first variable heavy domain and the first variable light domain from interacting to form an active Fv capable of binding to CD3; d) a second domain linker; and e) a second sdABD that binds to a second human TTA; and f) a cleavable linker (CL); and g) i) a first pseudo-variable light domain; ii) a non-cleavable linker (NCL), and iii) a first pseudo-variable heavy domain and a pseudo-Fv domain comprising: h) a third domain linker; and i) a third sdABD that binds human serum albumin; Including, the first variable heavy domain and the first variable light domain are capable of binding to human CD3, but the constrained Fv domain does not bind to CD3; the polypeptide does not bind to CD3 when the CL is intact; The method comprises administering to the patient a therapeutically effective amount of the polypeptide, a composition.

2. 1. A composition for use in a method of treating cancer in a patient in need thereof, the composition comprising a polypeptide, the polypeptide comprising, from N-terminus to C-terminus: a) a first single domain antigen binding domain (sdABD) that binds to human serum albumin; b) a first domain linker; and c) i) a first pseudo-variable light domain; ii) a non-cleavable linker (NCL), and iii) a first pseudo-variable heavy domain and a pseudo-Fv domain comprising: d) a cleavable linker (CL); and e) a second sdABD that binds to a first human tumor target antigen (TTA); and f) a second domain linker; and g) i) a second variable heavy domain comprising a vhCDR1, a vhCDR2 and a vhCDR3; ii) a constrained non-cleavable linker (CNCL), and iii) a first variable light domain comprising vlCDR1, vlCDR2 and vlCDR3; wherein the CNCL is disposed between the first variable heavy domain and the first variable light domain, and prevents the first variable heavy domain and the first variable light domain from interacting to form an active Fv capable of binding to CD3; h) a third domain linker; and i) a third second sdABD that binds to a second human TTA; and Including, the first variable heavy domain and the first variable light domain are capable of binding to human CD3, but the constrained Fv domain does not bind to CD3; the polypeptide does not bind to CD3 when the CL is intact; The method comprises administering to the patient a therapeutically effective amount of the polypeptide, a composition.

3. The cancer is (i) a solid tumor, (ii) Unresectable cancer; or (iii) metastatic cancer The composition according to claim 1 or claim 2,

4. 3. The composition of claim 1 or claim 2, wherein the cancer is head and neck cancer, optionally the cancer is squamous cell carcinoma, optionally the squamous cell carcinoma has progressed during or after combination therapy with a platinum-based chemotherapy agent and a checkpoint inhibitor, and further optionally the checkpoint inhibitor is an anti-PD1 antibody or antigen-binding fragment thereof.

5. The composition of claim 1 or claim 2, wherein the cancer is lung cancer.

6. 6. The composition of claim 5, wherein the cancer is non-small cell lung cancer, optionally wherein the non-small cell lung cancer has progressed during or after combination therapy with a platinum-based chemotherapy agent and a checkpoint inhibitor, and further optionally wherein the checkpoint inhibitor is an anti-PD1 antibody or antigen-binding fragment thereof. (i) the non-small cell lung cancer has an EGFR mutation, and optionally the cancer has progressed during or after treatment with an EGFR targeted therapy; and / or (ii) the non-small cell lung cancer harbors an ALK rearrangement, and optionally the cancer has progressed during or after treatment with an ALK-targeted therapy. The composition of claim 6.

8. 3. The composition of claim 1 or claim 2, wherein the cancer is colorectal cancer, and optionally the colorectal cancer has progressed during or after combination therapy with a platinum-based chemotherapy agent and a checkpoint inhibitor, and further optionally the checkpoint inhibitor is an anti-PD1 antibody or antigen-binding fragment thereof.

9. The colorectal cancer does not have a K-Ras mutation, and optionally (i) the colorectal cancer progresses during or after a combination of irinotecan or oxaliplatin-based chemotherapy and the patient has relapsed or is refractory to at least one prior systemic therapy, including an EGFR-targeted therapy; or (ii) the patient is ineligible for both irinotecan and oxaliplatin-based chemotherapy, and the patient has relapsed or is refractory to at least one prior systemic therapy, including an EGFR-targeted therapy; The composition of claim 8.

10. The composition of claim 8, wherein the colorectal cancer has a K-Ras mutation, and optionally (i) the colorectal cancer progresses during or after irinotecan or oxaliplatin-based chemotherapy, or (ii) the patient is ineligible for both irinotecan and oxaliplatin-based chemotherapy.

11. The composition of claim 1 or claim 2, wherein the cancer is prostate cancer. (i) the prostate cancer has progressed during or after platinum-based chemotherapy; (ii) the prostate cancer has progressed during or after anti-PDx therapy; and / or (iii) the efficacy of said treatment is assessed by measuring prostate-specific antigen (PSA) and / or prostate-specific membrane antigen (PSMA) levels by blood tests and / or positron emission tomography (PET) scans, and a decrease in said PSA level or said PSMA level in at least a threshold amount in said patient following a treatment regimen including said administration of said polypeptide, optionally for at least a minimum threshold time, indicates successful treatment. The composition of claim 11.

13. (i) the patient has an ECOG performance status of at least Grade 1. (ii) the patient has disease evaluable according to RECIST v1.1 criteria and documented by CT and / or MRI; (iii) the patient has previously treated symptomatic central nervous system metastases; and / or (iv) the patient does not have concurrent leptomeningeal disease; The composition according to claim 1 or claim 2. (i) (a) the first TTA and the second TTA are the same; or (b) the first TTA and the second TTA are different. (ii) (a) the first human TTA and the second human TTA are selected from EGFR and B7H3; or (b) the first sdABD-TTA and the second sdABD-TTA are selected from the group consisting of SEQ ID NO:1, SEQ ID NO:5, SEQ ID NO:9, SEQ ID NO:13, SEQ ID NO:17, SEQ ID NO:21, SEQ ID NO:25, SEQ ID NO:29, SEQ ID NO:33, SEQ ID NO:37, and SEQ ID NO:

41. (iii) the half-life prolonging domain has SEQ ID NO: 45; and / or (iv) the cleavable linker is cleaved by a human protease selected from the group consisting of MMP2, MMP9, meprin A, meprin B, cathepsin S, cathepsin K, cathepsin L, granzyme B, uPA, calecrein 7, matriptase, and thrombin; The composition according to claim 1 or claim 2. (i) the polypeptide is selected from the group consisting of SEQ ID NO:143 (Pro140), SEQ ID NO:144 (Pro140b), SEQ ID NO:185 (Pro141), SEQ ID NO:199 (Pro176), SEQ ID NO:208 (Pro188), SEQ ID NO:200 (Pro178), SEQ ID NO:201 (Pro179), SEQ ID NO:202 (Pro180), SEQ ID NO:203 (Pro181), SEQ ID NO:204 (Pro182), SEQ ID NO:205 (Pro183), SEQ ID NO:206 (Pro184), SEQ ID NO:207 (Pro185), SEQ ID NO:145 (Pro186), SEQ ID NO:146 (Pro187), SEQ ID NO:189 (Pro189), SEQ ID NO:2 (Pro190), SEQ ID NO:191 (Pro191), SEQ ID NO:212 (Pro192), SEQ ID NO:214 (Pro195), SEQ ID NO:215 (Pro196), SEQ ID NO:216 (Pro197), SEQ ID NO:217 (Pro198), SEQ ID NO:165 (SEQ ID NO:221), SEQ ID NO:166 (Pro222), SEQ ID NO:167 (Pro223), SEQ ID NO:168 (Pro224), SEQ ID NO:149 (Pro233), SEQ ID NO:226 (Pro247), SEQ ID NO:227 (Pro24 8), SEQ ID NO:228 (Pro249), SEQ ID NO:299 (Pro250), SEQ ID NO:230 (Pro251), SEQ ID NO:231 (Pro252), SEQ ID NO:232 (Pro253), SEQ ID NO:153 (Pro246), SEQ ID NO:169 (Pro254), SEQ ID NO:170 (Pro255), SEQ ID NO:154 (Pro256), SEQ ID NO:233 (Pro261), SEQ ID NO:171 (Pro262), SEQ ID NO:234 ​​(Pro294), SEQ ID NO:250 (Pro345), SEQ ID NO:259 (Pro375), SEQ ID NO:260 (Pro376), SEQ ID NO: SEQ ID NO:157 (Pro393), SEQ ID NO:158 (Pro394), SEQ ID NO:159 (Pro395), SEQ ID NO:160 (Pro396), SEQ ID NO:263 (Pro412), SEQ ID NO:264 (Pro413), SEQ ID NO:265 (Pro414), SEQ ID NO:265 (Pro415), SEQ ID NO:266 (Pro416), SEQ ID NO:267 (Pro417), SEQ ID NO:269 (Pro418), SEQ ID NO:272 (Pro429), SEQ ID NO:162 (Pro430), and SEQ ID NO:163 (Pro431), (ii) the polypeptide comprises the amino acid sequence of SEQ ID NO: 145 (Pro186); (iii) the polypeptide comprises the amino acid sequence of SEQ ID NO: 149 (Pro233); (iv) the polypeptide comprises the amino acid sequence of SEQ ID NO: 153 (Pro246); (v) the polypeptide comprises the amino acid sequence of SEQ ID NO: 169 (Pro254); (vi) the polypeptide comprises the amino acid sequence of SEQ ID NO: 232 (Pro253); (vii) the polypeptide comprises the amino acid sequence of SEQ ID NO: 234 (Pro294); (viii) the polypeptide comprises an amino acid sequence selected from the group consisting of SEQ ID NO: 147 (Pro225), SEQ ID NO: 148 (Pro226), SEQ ID NO: 173 (Pro359), SEQ ID NO: 257 (Pro373), SEQ ID NO: 258 (Pro374), SEQ ID NO: 181 (Pro479), SEQ ID NO: 182 (Pro480), and SEQ ID NO: 183 (Pro495); (ix) the polypeptide comprises the amino acid sequence of SEQ ID NO: 147 (Pro225); (x) the polypeptide comprises the amino acid sequence of SEQ ID NO: 148 (Pro226); (xi) the polypeptide comprises the amino acid sequence of SEQ ID NO: 173 (Pro359); (xii) the polypeptide comprises the amino acid sequence of SEQ ID NO: 257 (Pro373); (xiii) the polypeptide comprises the amino acid sequence of SEQ ID NO: 258 (Pro374); (xiv) the polypeptide comprises the amino acid sequence of SEQ ID NO: 181 (Pro479); (xv) the polypeptide comprises the amino acid sequence of SEQ ID NO: 182 (Pro480); or (xvi) the polypeptide comprises the amino acid sequence of SEQ ID NO: 183 (Pro495); The composition according to claim 1 or claim 2.

Images