Constrained and conditionally activated binding proteins

Abstract

The present invention relates to conditional bispecific redirecting activation constructs, or COBRAs, administered in an active prodrug format. Upon exposure to tumor proteases, the constructs are cleaved and activated, allowing them to bind to one or more tumor target antigens (TTAs) and CD3, resulting in the recruitment of CD3-expressing T cells to the tumor and treatment. In some embodiments, tumor target antigens include B7H3, CA9 (CAIX), EGFR, EpCAM, FOLR1, HER2, LyPD3, and / or Trop2.

Description

[Technical Field]

[0001] Cross-reference of related applications This application claims priority to U.S. Provisional Application No. 63 / 066,565, filed on 17 August 2020, the disclosure of which is incorporated herein by reference in its entirety.

[0002] Sequence List This application includes a sequence listing, which is filed electronically in ASCII format and is incorporated in its entirety herein by reference. The ASCII copy, created on 13 August 2021, is named 118459-5014-WO_SL.txt and has a size of 1,341,889 bytes. [Background technology]

[0003] The selective destruction of individual cells or specific cell types is often desirable in a variety of clinical settings. For example, specifically destroying tumor cells while leaving healthy cells and tissues as intact and undamaged as possible is a major objective in cancer treatment. One such method involves inducing an immune response against the tumor, causing immune effector cells, such as natural killer (NK) cells or cytotoxic T lymphocytes (CTLs), to attack and destroy the tumor cells.

[0004] The use of intact monoclonal antibodies (mAbs), which exhibit excellent binding specificity and affinity for tumor-associated antigens, has been successfully applied in the treatment and diagnosis of cancer. However, the large size of intact mAbs, their poor distribution in the body, low potency, and poor long-term persistence in the blood pool limit their clinical applications. For example, intact antibodies may show specific accumulation within tumor regions. Distribution studies have shown heterogeneous antibody distribution with initial accumulation in peripheral regions when tumors are closely examined. Due to the heterogeneous antigen distribution and increased interstitial pressure caused by tumor necrosis, it is impossible for intact antibody constructs to reach the central part of the tumor. In contrast, smaller antibody fragments show rapid localization to tumors, penetrate deeper into the tumor, and are removed from the bloodstream relatively quickly. However, many antibodies, including scFv and other constructs, exhibit an "on target / off tumor" effect, meaning the molecules are active against non-tumor cells, causing side effects, some of which may be toxic. This invention relates to a novel construct that is selectively activated in the presence of tumor proteases. [Overview of the project]

[0005] In one embodiment, from the N-terminus to the C-terminus, (a) a first sdABD (sdABD-HER2) that binds to HER2, (b) a first domain linker, (c) a restrictive Fv domain including (i) a first variable heavy domain containing vhCDR1, vhCDR2 and vhCDR3, (ii) a restrictive non-cleavable linker (CNCL), and (iii) a first variable light domain containing vlCDR1, vlCDR2 and vlCDR3, (d) a second domain linker, (e) a second sdABD-HER2, (f) a cleavable linker (CL), and (g) a first pseudo-variable light domain and (ii) a non-cleavable linker (NCL). A fusion protein is provided comprising (iii) a restrictive pseudo-Fv domain including a first pseudo-variable weight domain, (h) a third domain linker, and (i) a third sdABD (sdABD-HSA) that binds to human serum albumin, wherein the first variable weight domain and the first variable light domain of the restrictive Fv domain can bind to human CD3, but the restrictive pseudo-Fv domain does not bind to CD3, the first variable weight 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 weight domain associate intramolecularly to form an inactive Fv.

[0006] In some embodiments, the first and / or second sdABD-HER2 is (a) sdCDR1 of SEQ ID NO: 194, sdCDR2 of SEQ ID NO: 195, and sdCDR3 of SEQ ID NO: 196, (b) sdCDR1 of SEQ ID NO: 218, sdCDR2 of SEQ ID NO: 219, and sdCDR3 of SEQ ID NO: 220, (c) sdCDR1 of SEQ ID NO: 226, sdCDR2 of SEQ ID NO: 227, and sdCDR3 of SEQ ID NO: 228, (d) sdCDR1 of SEQ ID NO: 238, sdCDR2 of SEQ ID NO: 239, and sdCDR3 of SEQ ID NO: 240, (e) sdC (f) sdCDR1, sdCDR2 of sequence number 143 and sdCDR3 of sequence number 144, (g) sdCDR1 of sequence number 146, sdCDR2 of sequence number 147 and sdCDR3 of sequence number 148, (h) sdCDR1 of sequence number 150, sdCDR2 of sequence number 151 and sdCDR3 of sequence number 152, (i) sdCDR1 of sequence number 158, sdCDR2 of sequence number 159 and sdCDR3 of sequence number 160, (j) sdCDR1 of sequence number 162 R1, sdCDR2 of SEQ ID NO: 163 and sdCDR3 of SEQ ID NO: 164, (k) sdCDR1 of SEQ ID NO: 166, sdCDR2 of SEQ ID NO: 167 and sdCDR3 of SEQ ID NO: 168, (l) sdCDR1 of SEQ ID NO: 170, sdCDR2 of SEQ ID NO: 171 and sdCDR3 of SEQ ID NO: 172, (m) sdCDR1 of SEQ ID NO: 174, sdCDR2 of SEQ ID NO: 175 and sdCDR3 of SEQ ID NO: 176, (n) sdCDR1 of SEQ ID NO: 178, sdCDR2 of SEQ ID NO: 179 and sdCDR3 of SEQ ID NO: 180, (o) sdCDR of SEQ ID NO: 182 1. sdCDR2 of sequence number 183 and sdCDR3 of sequence number 184, (p) sdCDR1 of sequence number 186, sdCDR2 of sequence number 187 and sdCDR3 of sequence number 188, (q) sdCDR1 of sequence number 190, sdCDR2 of sequence number 191 and sdCDR3 of sequence number 192, (r) sdCDR1 of sequence number 194, sdCDR2 of sequence number 195 and sdCDR3 of sequence number 196, (s) sdCDR1 of sequence number 198, sdCDR2 of sequence number 199 and sdCDR3 of sequence number 200, (t) sdCDR1 of sequence number 202,(u) sdCDR2 of sequence number 203 and sdCDR3 of sequence number 204, (v) sdCDR1 of sequence number 206, sdCDR2 of sequence number 207 and sdCDR3 of sequence number 203, (w) sdCDR1 of sequence number 210, sdCDR2 of sequence number 211 and sdCDR3 of sequence number 212, (x) sdCDR1 of sequence number 214, sdCDR2 of sequence number 215 and sdCDR3 of sequence number 216, (x) sequence number 218 (y) sdCDR1, sdCDR2 of sequence number 219 and sdCDR3 of sequence number 220, (z) sdCDR1 of sequence number 226, sdCDR2 of sequence number 227 and sdCDR3 of sequence number 228, (aa) sdCDR1 of sequence number 230, sdCDR2 of sequence number 231 and sdCDR3 of sequence number 232, (ab) sdCDR1 of SEQ ID NO: 234, sdCDR2 of SEQ ID NO: 235, and sdCDR3 of SEQ ID NO: 236, (ac) sdCDR1 of SEQ ID NO: 238, sdCDR2 of SEQ ID NO: 239, and sdCDR3 of SEQ ID NO: 240, (ad) sdCDR1 of SEQ ID NO: 242, sdCDR2 of SEQ ID NO: 243, and sdCDR3 of SEQ ID NO: 244, (ae) sdCDR1 of SEQ ID NO: 500, sdCDR2 of SEQ ID NO: 501, and sdCDR3 of SEQ ID NO: 502 It has an amino acid sequence containing a set of CDRs selected from the group consisting of dCDR3, (af) sdCDR1 of SEQ ID NO: 504, sdCDR2 of SEQ ID NO: 505, and sdCDR3 of SEQ ID NO: 506, (ag) sdCDR1 of SEQ ID NO: 508, sdCDR2 of SEQ ID NO: 509, and sdCDR3 of SEQ ID NO: 510, and (ah) sdCDR1 of SEQ ID NO: 512, sdCDR2 of SEQ ID NO: 513, and sdCDR3 of SEQ ID NO: 5.

[0007] In some embodiments, the first and / or second sdABD-HER2 includes an amino acid sequence selected from the group consisting of SEQ ID NOs: 193, 217, 225, 237, 141, 145, 149, 153, 157, 161, 165, 169, 173, 177, 181, 185, 189, 197, 201, 205, 209, 213, 221, 229, 233, 241, 499, 503, 507, and 511.

[0008] In some embodiments, the first sdABD-HER2 and the second sdABD-HER2 are the same.

[0009] In some embodiments, the first sdABD-HER2 and the second sdABD-HER2 are different.

[0010] In some embodiments, the first variable heavy domain is located at the N-terminus of the first variable light domain, and the pseudo-variable light domain is located at the N-terminus of the pseudo-variable heavy domain.

[0011] In some embodiments, the first variable heavy domain is located at the N-terminus of the first variable light domain, and the pseudo-variable heavy domain is located at the N-terminus of the pseudo-variable light domain.

[0012] In some embodiments, the first variable light domain is located at the N-terminus of the first variable heavy domain, and the pseudo-variable light domain is located at the N-terminus of the pseudo-variable heavy domain.

[0013] In some embodiments, the first variable light domain is located at the N-terminus of the first variable heavy domain, and the pseudo-variable heavy domain is located at the N-terminus of the pseudo-variable light domain.

[0014] In some embodiments, the third sdABD that binds to HSA (sdABD-HSA) has an amino acid sequence comprising: (a) a set of CDRs selected from the group consisting of (i) sdCDR1 of SEQ ID NO: 246, sdCDR2 of SEQ ID NO: 247, and sdCDR3 of SEQ ID NO: 248, and (ii) sdCDR1 of SEQ ID NO: 250, sdCDR2 of SEQ ID NO: 251, and sdCDR3 of SEQ ID NO: 252; or (b) an amino acid sequence selected from the group consisting of SEQ ID NO: 245 and SEQ ID NO: 249.

[0015] In some embodiments, the cleavable linker comprises a cleavage domain sequence selected from the group consisting of SEQ ID NOs: 339-408 and SEQ ID NOs: 532-535.

[0016] In some embodiments, 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, kallikrein 7, matriptase, and thrombin.

[0017] In some embodiments, the fusion protein has an amino acid sequence selected from the group consisting of SEQ ID NOs: 459-484 and SEQ ID NOs: 491-494.

[0018] Nucleic acids encoding any of the described fusion proteins are provided herein.

[0019] Expression vectors containing any of the described nucleic acids are provided herein.

[0020] Host cells containing any of the described expression vectors are provided herein.

[0021] In some aspects, provided is a method for producing a fusion protein of the present disclosure, comprising: (i) culturing the described host cells under conditions where the fusion protein is expressed; and (ii) recovering the fusion protein.

[0022] In some embodiments, a method is provided for treating a target cancer, comprising administering one of the described fusion proteins to the target.

[0023] In some embodiments, (i) SEQ ID NOs: 141, 145, 149, 153, 157, 161, 165, 169, 173, 177, 181, 185, 189, 193, 197, 201, 205, 209, 213, 217, 221, 225, 229, 233, 237, 241, 499, 503, 507 An amino acid sequence selected from the group consisting of (a) sdCDR1 of SEQ ID NO: 194, sdCDR2 of SEQ ID NO: 195 and sdCDR3 of SEQ ID NO: 196, (b) sdCDR1 of SEQ ID NO: 218, sdCDR2 of SEQ ID NO: 219 and sdCDR3 of SEQ ID NO: 220, (c) sdCDR1 of SEQ ID NO: 226, sdCDR2 of SEQ ID NO: 227 and sdCDR3 of SEQ ID NO: 228, (d) sdCDR1 of SEQ ID NO: 238, sdCDR2 of SEQ ID NO: 239 and sdCDR3 of SEQ ID NO: 240, (e) sdCDR1 of SEQ ID NO: 142, (f) sdCDR2 of SEQ ID NO: 143 and sdCDR3 of SEQ ID NO: 144, (g) sdCDR1 of SEQ ID NO: 146, sdCDR2 of SEQ ID NO: 147 and sdCDR3 of SEQ ID NO: 148, (h) sdCDR1 of SEQ ID NO: 150, sdCDR2 of SEQ ID NO: 151 and sdCDR3 of SEQ ID NO: 152, (h) sdCDR1 of SEQ ID NO: 154, sdCDR2 of SEQ ID NO: 155 and sdCDR3 of SEQ ID NO: 156, (i) sdCDR1 of SEQ ID NO: 158, sdCDR2 of SEQ ID NO: 159 and sdCDR3 of SEQ ID NO: 160, (j) sdCDR1 of SEQ ID NO: 162 (k) sdCDR2 of sequence number 163 and sdCDR3 of sequence number 164, (l) sdCDR1 of sequence number 166, sdCDR2 of sequence number 167 and sdCDR3 of sequence number 168, (m) sdCDR1 of sequence number 170, sdCDR2 of sequence number 171 and sdCDR3 of sequence number 172, (n) sdCDR1 of sequence number 178, sdCDR2 of sequence number 179 and sdCDR3 of sequence number 180, (o) sdCDR1 of sequence number 182,(p) sdCDR2 of SEQ ID NO: 183 and sdCDR3 of SEQ ID NO: 184, (p) sdCDR1 of SEQ ID NO: 186, sdCDR2 of SEQ ID NO: 187 and sdCDR3 of SEQ ID NO: 188, (q) sdCDR1 of SEQ ID NO: 190, sdCDR2 of SEQ ID NO: 191 and sdCDR3 of SEQ ID NO: 192, (r) sdCDR1 of SEQ ID NO: 194, sdCDR2 of SEQ ID NO: 195 and sdCDR3 of SEQ ID NO: 196, (s) sdCDR1 of SEQ ID NO: 198, sdCDR2 of SEQ ID NO: 199 and sdCDR3 of SEQ ID NO: 2 (u) sdCDR1 of 02, sdCDR2 of sequence number 203 and sdCDR3 of sequence number 204, (v) sdCDR1 of sequence number 210, sdCDR2 of sequence number 211 and sdCDR3 of sequence number 212, (w) sdCDR1 of sequence number 214, sdCDR2 of sequence number 215 and sdCDR3 of sequence number 216, (x) sdCDR1 of sequence number 218, sdCDR2 of sequence number 219 and sdCDR3 of sequence number 220 (y) sdCDR1 of sequence number 222, sdCDR2 of sequence number 223 and sdCDR3 of sequence number 224, (z) sdCDR1 of sequence number 226, sdCDR2 of sequence number 227 and sdCDR3 of sequence number 228, (aa) sdCDR1 of sequence number 230, sdCDR2 of sequence number 231 and sdCDR3 of sequence number 232, (ab) sdCDR1 of sequence number 234, sdCDR2 of sequence number 235 and sdCDR3 of sequence number 236, (ac) sdCDR1 of sequence number 238, sdCDR2 of sequence number 239 and sequence (ad) sdCDR3 of sequence number 240, sdCDR1 of sequence number 242, sdCDR2 of sequence number 243 and sdCDR3 of sequence number 244, (ae) sdCDR1 of sequence number 500, sdCDR2 of sequence number 501 and sdCDR3 of sequence number 502, (af) sdCDR1 of sequence number 504, sdCDR2 of sequence number 505 and sdCDR3 of sequence number 506, (ag) sdCDR1 of sequence number 508, sdCDR2 of sequence number 509 and sdCDR3 of sequence number 510, and (ah) sdCDR1 of sequence number 512,A single-domain antigen-binding domain (sdABD) (sdABD-HER2) that binds to human HER2 is provided, comprising an amino acid sequence containing a set of CDRs selected from the group consisting of sdCDR2 (SEQ ID NO: 513) and sdCDR3 (SEQ ID NO: 514).

[0024] In some embodiments, from the N-terminus to the C-terminus, (a) a first sdABD (sdABD-TTA) that binds to a tumor target antigen, (b) a first domain linker, (c) a restrictive Fv domain comprising (i) a first variable heavy domain containing vhCDR1, vhCDR2 and vhCDR3, (ii) a restrictive non-cleavageable linker (CNCL), and (iii) a first variable light domain containing vlCDR1, vlCDR2 and vlCDR3, (d) a second domain linker, (e) a second sdABD-TTA, (f) A cleavable linker (CL), (g) A restrictive pseudo-Fv domain comprising (i) a first pseudo-variable light domain and (ii) a non-cleavable linker (NCL) and (iii) a first pseudo-variable heavy domain is provided, and a fusion protein comprising (h) a third domain linker and (i) a third sdABD (sdABD-HSA) that binds to human serum albumin, wherein the first variable heavy domain and the first variable light domain of the restrictive Fv domain can bind to human CD3, but the restrictive pseudo-Fv domain It 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, (1) the first sdABD-TTA is sdABD-HER2 or sdABD-LyPD3, and the second sdABD-TTA is sdABD-B7H3, sdABD-CA9, sdABD-EGFR, sdABD-EpCAM, sdABD-FOLR1, sdAB (1) The first sdABD-TTA is selected from the group consisting of D-HER2, sdABD-LyPD3, and sdABD-Trop2, or (2) the first sdABD-TTA is selected from the group consisting of sdABD-B7H3, sdABD-CA9, sdABD-EGFR, sdABD-EpCAM, sdABD-FOLR1, sdABD-HER2, sdABD-LyPD3, and sdABD-Trop2, and the second sdABD-TTA is either sdABD-HER2 or sdABD-LyPD3.

[0025] In some embodiments, the first and second sdABD-TTAs are each sdABD-LyPD3. In some embodiments, the first and second sdABD-LPYD3s are the same. In some embodiments, the first and second sdABD-LPYD3s are different.

[0026] In some embodiments of the fusion protein, (a) the first sdABD-TTA is sdABD-HER2 and the second sdABD-TTA is selected from the group consisting of sdABD-B7H3, sdABD-CA9, sdABD-EGFR, sdABD-EpCAM, sdABD-FOLR1, sdABD-LyPD3 and sdABD-Trop2, and (b) the first sdABD-TTA is sdABD-LyPD3 and the second sdABD-TTA is selected from the group consisting of sdABD-B7H3, sdABD-CA9, sdABD-EGFR, sdABD-EpCAM, sdABD-FOLR1, sdABD-HER2 and sdABD-Trop2. (c) The first sdABD-TTA is selected from the group consisting of p2, and the second TTA is sdABD-HER2, or (d) The first sdABD-TTA is selected from the group consisting of sdABD-B7H3, sdABD-CA9, sdABD-EGFR, sdABD-EpCAM, sdABD-FOLR1, sdABD-LyPD3, and sdABD-Trop2, and the second TTA is sdABD-HER2.

[0027] In some embodiments, sdABD-HER2 includes (a) a set of CDRs including sdCDR1 of SEQ ID NO: 194, sdCDR2 of SEQ ID NO: 195, and sdCDR3 of SEQ ID NO: 196; (b) a set of CDRs including sdCDR1 of SEQ ID NO: 218, sdCDR2 of SEQ ID NO: 219, and sdCDR3 of SEQ ID NO: 220; (c) a set of CDRs including sdCDR1 of SEQ ID NO: 226, sdCDR2 of SEQ ID NO: 227, and sdCDR3 of SEQ ID NO: 228; and (d) sdCDR1 of SEQ ID NO: 238, sdCDR2 of SEQ ID NO: 239, and SEQ ID NO: 240. (e) A set of CDRs including sdCDR1 of sequence number 142, sdCDR2 of sequence number 143, and sdCDR3 of sequence number 144, (f) A set of CDRs including sdCDR1 of sequence number 146, sdCDR2 of sequence number 147, and sdCDR3 of sequence number 148, (g) A set of CDRs including sdCDR1 of sequence number 150, sdCDR2 of sequence number 151, and sdCDR3 of sequence number 152, (h) A set of CDRs including sdCDR1 of sequence number 154, sdCDR2 of sequence number 155, and sdCDR3 of sequence number 156, (i) A set of CDRs containing sdCDR1 of SEQ ID NO: 158, sdCDR2 of SEQ ID NO: 159, and sdCDR3 of SEQ ID NO: 160; (j) A set of CDRs containing sdCDR1 of SEQ ID NO: 162, sdCDR2 of SEQ ID NO: 163, and sdCDR3 of SEQ ID NO: 164; (k) A set of CDRs containing sdCDR1 of SEQ ID NO: 166, sdCDR2 of SEQ ID NO: 167, and sdCDR3 of SEQ ID NO: 168; (l) A set of CDRs containing sdCDR1 of SEQ ID NO: 170, sdCDR2 of SEQ ID NO: 171, and sdCDR3 of SEQ ID NO: 172; (m) SEQ ID NO: 17 (n) A set of CDRs including sdCDR1 of sequence number 4, sdCDR2 of sequence number 175 and sdCDR3 of sequence number 176, (o) A set of CDRs including sdCDR1 of sequence number 178, sdCDR2 of sequence number 179 and sdCDR3 of sequence number 180, (o) A set of CDRs including sdCDR1 of sequence number 182, sdCDR2 of sequence number 183 and sdCDR3 of sequence number 184, (p) A set of CDRs including sdCDR1 of sequence number 186, sdCDR2 of sequence number 187 and sdCDR3 of sequence number 188, (q) sdCDR1 of sequence number 190,(r) A set of CDRs including sdCDR2 of sequence number 191 and sdCDR3 of sequence number 192, (r) A set of CDRs including sdCDR1 of sequence number 194, sdCDR2 of sequence number 195 and sdCDR3 of sequence number 196, (s) A set of CDRs including sdCDR1 of sequence number 198, sdCDR2 of sequence number 199 and sdCDR3 of sequence number 200, (t) A set of CDRs including sdCDR1 of sequence number 202, sdCDR2 of sequence number 203 and sdCDR3 of sequence number 204, (u) sdCDR1 of sequence number 206 and sdCDR of sequence number 207 (v) A set of CDRs including sdCDR3 of sequence number 203, (v) A set of CDRs including sdCDR1 of sequence number 210, sdCDR2 of sequence number 211 and sdCDR3 of sequence number 212, (w) A set of CDRs including sdCDR1 of sequence number 214, sdCDR2 of sequence number 215 and sdCDR3 of sequence number 216, (x) A set of CDRs including sdCDR1 of sequence number 218, sdCDR2 of sequence number 219 and sdCDR3 of sequence number 220, (y) sdCDR1 of sequence number 222, sdCDR2 of sequence number 223 and sdCDR3 of sequence number 224 (z) A set of CDRs including CDR3, (aa) A set of CDRs including sdCDR1 of sequence number 226, sdCDR2 of sequence number 227, and sdCDR3 of sequence number 228, (ab) A set of CDRs including sdCDR1 of sequence number 230, sdCDR2 of sequence number 231, and sdCDR3 of sequence number 232, (ab) A set of CDRs including sdCDR1 of sequence number 234, sdCDR2 of sequence number 235, and sdCDR3 of sequence number 236, (ac) A set of CDRs including sdCDR1 of sequence number 238, sdCDR2 of sequence number 239, and sdCDR3 of sequence number 240 A set of (ad) sdCDR1 of sequence number 242, sdCDR2 of sequence number 243 and sdCDR3 of sequence number 244, (ae) sequence number 141, (af) sequence number 145, (ag) sequence number 149, (ag) sequence number 153, (ai) sequence number 157, (aj) sequence number 161, (ak) sequence number 165, (al) sequence number 169, (am) sequence number 173, (an) sequence number 177, (ao) sequence number 181, (ap) sequence number 185, (aq) sequence number 189, (ar) sequence number 193, (as) sequence number 197,It contains an amino acid sequence selected from the group consisting of (at)sequence number 201, (au)sequence number 205, (av)sequence number 209, (aw)sequence number 213, (ax)sequence number 217, (ay)sequence number 221, (az)sequence number 225, (ba)sequence number 229, (bb)sequence number 233, (bc)sequence number 237, and (bd)sequence number 241.

[0028] sdABD-LyPD3 is a set of CDs containing (a) sdCDR1 of SEQ ID NO: 118, sdCDR2 of SEQ ID NO: 119, and sdCDR3 of SEQ ID NO: 120; (b) a set of CDs containing sdCDR1 of SEQ ID NO: 122, sdCDR2 of SEQ ID NO: 123, and sdCDR3 of SEQ ID NO: 124; (c) a set of CDs containing sdCDR1 of SEQ ID NO: 126, sdCDR2 of SEQ ID NO: 127, and sdCDR3 of SEQ ID NO: 128; and (d) a set of CDs containing sdCDR1 of SEQ ID NO: 130, sdCDR2 of SEQ ID NO: 131, and sdCDR3 of SEQ ID NO: 132. A fusion protein according to any one of claims 20 to 25, comprising a set of DRs, a set of CDRs including (e) sdCDR1 of SEQ ID NO: 134, sdCDR2 of SEQ ID NO: 135, and sdCDR3 of SEQ ID NO: 136, a set of CDRs including sdCDR1 of SEQ ID NO: 138, sdCDR2 of SEQ ID NO: 139, and sdCDR3 of SEQ ID NO: 140, an amino acid sequence selected from the group consisting of (g) SEQ ID NO: 117, (h) SEQ ID NO: 121, (i) SEQ ID NO: 125, (j) SEQ ID NO: 129, (k) SEQ ID NO: 133, and (l) SEQ ID NO: 137.

[0029] In some embodiments, sdABD-B7H3 is (i) a set of CDRs including sdCDR1 of SEQ ID NO: 34, sdCDR2 of SEQ ID NO: 35, and sdCDR3 of SEQ ID NO: 36, (ii) a set of CDRs including sdCDR1 of SEQ ID NO: 38, sdCDR2 of SEQ ID NO: 39, and sdCDR3 of SEQ ID NO: 40, (iii) a set of CDRs including sdCDR1 of SEQ ID NO: 42, sdCDR2 of SEQ ID NO: 43, and sdCDR3 of SEQ ID NO: 44, (iv) a set of CDRs including sdCDR1 of SEQ ID NO: 46, sdCDR2 of SEQ ID NO: 47, and sdCDR3 of SEQ ID NO: 48, (v) sdCDR1 of SEQ ID NO: 50 (vi) A set of CDRs including CDR1, sdCDR2 of SEQ ID NO: 51, and sdCDR3 of SEQ ID NO: 52; (vi) A set of CDRs including sdCDR1 of SEQ ID NO: 54, sdCDR2 of SEQ ID NO: 55, and sdCDR3 of SEQ ID NO: 56; (vii) A set of CDRs including sdCDR1 of SEQ ID NO: 58, sdCDR2 of SEQ ID NO: 59, and sdCDR3 of SEQ ID NO: 60; (ix) An amino acid sequence selected from the group consisting of SEQ ID NO: 33, (x) SEQ ID NO: 37, (xi) SEQ ID NO: 41, (xii) SEQ ID NO: 45, (xiii) SEQ ID NO: 49, (xiv) SEQ ID NO: 53, and (xv) SEQ ID NO: 57.

[0030] In some embodiments, sdABD-CA9 includes an amino acid sequence selected from the group consisting of (i) a set of CDRs including sdCDR1 of SEQ ID NO: 102, sdCDR2 of SEQ ID NO: 103, and sdCDR3 of SEQ ID NO: 104; (ii) a set of CDRs including sdCDR1 of SEQ ID NO: 106, sdCDR2 of SEQ ID NO: 107, and sdCDR3 of SEQ ID NO: 108; (iii) a set of CDRs including sdCDR1 of SEQ ID NO: 110, sdCDR2 of SEQ ID NO: 111, and sdCDR3 of SEQ ID NO: 112; (iv) a set of CDRs including sdCDR1 of SEQ ID NO: 114, sdCDR2 of SEQ ID NO: 115, and sdCDR3 of SEQ ID NO: 116; (v) SEQ ID NO: 101, (vi) SEQ ID NO: 105, (vii) SEQ ID NO: 109, and (viii) SEQ ID NO: 113.

[0031] In some embodiments, sdABD-EGFR includes an amino acid sequence selected from the group consisting of (i) a set of CDRs including sdCDR1 of SEQ ID NO: 2, sdCDR2 of SEQ ID NO: 3, and sdCDR3 of SEQ ID NO: 4; (ii) a set of CDRs including sdCDR1 of SEQ ID NO: 6, sdCDR2 of SEQ ID NO: 7, and sdCDR3 of SEQ ID NO: 8; (iii) a set of CDRs including sdCDR1 of SEQ ID NO: 10, sdCDR2 of SEQ ID NO: 11, and sdCDR3 of SEQ ID NO: 12; (iv) a set of CDRs including sdCDR1 of SEQ ID NO: 14, sdCDR2 of SEQ ID NO: 15, and sdCDR3 of SEQ ID NO: 16; (v) a set of CDRs including sdCDR1 of SEQ ID NO: 18, sdCDR2 of SEQ ID NO: 19, and sdCDR3 of SEQ ID NO: 20; (vi) SEQ ID NO: 1, (vii) SEQ ID NO: 5, (viii) SEQ ID NO: 9, (ix) SEQ ID NO: 13, and (x) SEQ ID NO: 17.

[0032] In some embodiments, sdABD-EpCAM includes an amino acid sequence selected from the group consisting of (i) a set of CDRs including sdCDR1 of SEQ ID NO: 62, sdCDR2 of SEQ ID NO: 63, and sdCDR3 of SEQ ID NO: 64; (ii) a set of CDRs including sdCDR1 of SEQ ID NO: 66, sdCDR2 of SEQ ID NO: 67, and sdCDR3 of SEQ ID NO: 68; (iii) a set of CDRs including sdCDR1 of SEQ ID NO: 70, sdCDR2 of SEQ ID NO: 71, and sdCDR3 of SEQ ID NO: 72; (iv) a set of CDRs including sdCDR1 of SEQ ID NO: 74, sdCDR2 of SEQ ID NO: 75, and sdCDR3 of SEQ ID NO: 76; (v) a set of CDRs including sdCDR1 of SEQ ID NO: 496, sdCDR2 of SEQ ID NO: 497, and sdCDR3 of SEQ ID NO: 498; (vi) SEQ ID NO: 61; (vii) SEQ ID NO: 65; (viii) SEQ ID NO: 69; (ix) SEQ ID NO: 73; and (x) SEQ ID NO: 495.

[0033] In some embodiments, sdABD-FOLR1 includes an amino acid sequence selected from the group consisting of (i) a set of CDRs including sdCDR1 of SEQ ID NO: 22, sdCDR2 of SEQ ID NO: 23, and sdCDR3 of SEQ ID NO: 24; (ii) a set of CDRs including sdCDR1 of SEQ ID NO: 26, sdCDR2 of SEQ ID NO: 27, and sdCDR3 of SEQ ID NO: 28; (iii) a set of CDRs including sdCDR1 of SEQ ID NO: 30, sdCDR2 of SEQ ID NO: 31, and sdCDR3 of SEQ ID NO: 32; (iv) SEQ ID NO: 21; (v) SEQ ID NO: 25; and (vi) SEQ ID NO: 29.

[0034] In some embodiments, sdABD-Trop2 is (i) a set of CDRs including sdCDR1 of SEQ ID NO: 78, sdCDR2 of SEQ ID NO: 79, and sdCDR3 of SEQ ID NO: 80, (ii) a set of CDRs including sdCDR1 of SEQ ID NO: 82, sdCDR2 of SEQ ID NO: 83, and sdCDR3 of SEQ ID NO: 84, (iii) a set of CDRs including sdCDR1 of SEQ ID NO: 86, sdCDR2 of SEQ ID NO: 87, and sdCDR3 of SEQ ID NO: 88, (iv) sdCDR1 of SEQ ID NO: 90, sdCDR2 of SEQ ID NO: 91, and (v) a set of CDRs including sdCDR3 of SEQ ID NO: 92, (v) a set of CDRs including sdCDR1 of SEQ ID NO: 94, sdCDR2 of SEQ ID NO: 95, and sdCDR3 of SEQ ID NO: 96, (vi) a set of CDRs including sdCDR1 of SEQ ID NO: 98, sdCDR2 of SEQ ID NO: 99, and sdCDR3 of SEQ ID NO: 100, (vii) an amino acid sequence selected from the group consisting of SEQ ID NO: 77, (viii) SEQ ID NO: 81, (ix) SEQ ID NO: 85, (x) SEQ ID NO: 89, (xi) SEQ ID NO: 93, and (xii) SEQ ID NO: 97.

[0035] In some embodiments, the first variable heavy domain is located at the N-terminus of the first variable light domain, and the pseudo-variable light domain is located at the N-terminus of the pseudo-variable heavy domain.

[0036] In some embodiments, the first variable heavy domain is located at the N-terminus of the first variable light domain, and the pseudo-variable heavy domain is located at the N-terminus of the pseudo-variable light domain.

[0037] In some embodiments, the first variable light domain is located at the N-terminus of the first variable heavy domain, and the pseudo-variable light domain is located at the N-terminus of the pseudo-variable heavy domain.

[0038] In some embodiments, the first variable light domain is located at the N-terminus of the first variable heavy domain, and the pseudo-variable heavy domain is located at the N-terminus of the pseudo-variable light domain.

[0039] In some embodiments, the third sdABD that binds to the HSA has an amino acid sequence comprising (a)(i) a set of CDRs selected from the group consisting of sdCDR1 of SEQ ID NO: 246, sdCDR2 of SEQ ID NO: 247, and sdCDR3 of SEQ ID NO: 248, and (ii) sdCDR1 of SEQ ID NO: 250, sdCDR2 of SEQ ID NO: 251, and sdCDR3 of SEQ ID NO: 252, or (b) an amino acid sequence selected from the group consisting of SEQ ID NO: 245 and SEQ ID NO: 249.

[0040] In some embodiments, the cleavage linker includes a cleavage domain sequence selected from the group consisting of SEQ ID NOs: 339-408 and SEQ ID NOs: 532-535.

[0041] In some embodiments, the cleavable linker is cleaved by a human protease selected from the group consisting of MMP2, MMP9, meprin A, meprin B, cathepsin S, captepsin K, captepsin L, granzyme B, uPA, kalecreiin 7, matryptase, and thrombin.

[0042] In some embodiments, the fusion protein includes an amino acid sequence selected from the group consisting of SEQ ID NOs: 453, 454, 455, 456, 457, and 458.

[0043] Nucleic acids encoding any of the described fusion proteins are provided herein. Expression vectors containing any of the described nucleic acids are provided herein. Host cells containing any of the described expression vectors are provided herein.

[0044] In some embodiments, a method for producing the fusion protein of the present disclosure is provided, comprising (i) culturing the described host cells under conditions in which the fusion protein is expressed, and (ii) recovering the fusion protein.

[0045] In some embodiments, (i) an amino acid sequence selected from the group consisting of SEQ ID NOs: 117, 121, 125, 129, 133, and 137, or (ii) (a) sdCDR1 of SEQ ID NOs: 118, sdCDR2 of SEQ ID NOs: 119, and sdCDR3 of SEQ ID NOs: 120, (b) sdCDR1 of SEQ ID NOs: 122, sdCDR2 of SEQ ID NOs: 123, and sdCDR3 of SEQ ID NOs: 124, (c) sdCDR1 of SEQ ID NOs: 126, sdCDR2 of SEQ ID NOs: 127, and sdC A single-domain antigen-binding domain (sdABD-LyPD3) that binds to human LyPD3 is provided, comprising an amino acid sequence containing a set of CDRs selected from the group consisting of DR3, (d) sdCDR1 of SEQ ID NO: 130, sdCDR2 of SEQ ID NO: 131, and sdCDR3 of SEQ ID NO: 132, (e) sdCDR1 of SEQ ID NO: 134, sdCDR2 of SEQ ID NO: 135, and sdCDR3 of SEQ ID NO: 136, and (f) sdCDR1 of SEQ ID NO: 138, sdCDR2 of SEQ ID NO: 139, and sdCDR3 of SEQ ID NO: 140.

[0046] Nucleic acids encoding any of the described single-domain antigen-binding domains (sdABDs) are also provided. Expression vectors containing any of the nucleic acids are also provided. Host cells containing any of the described expression vectors are provided herein.

[0047] In some embodiments, a method is provided for producing a single-domain antigen-binding domain (sdABD), comprising (a) culturing one of the host cells described herein under conditions in which sdABD is expressed, and (b) recovering the sdABD.

[0048] A pharmaceutical composition comprising any of the fusion proteins described or any of the single-domain antigen-binding domains (sdABDs) described is also provided.

[0049] In some embodiments, the pharmaceutical composition further comprises a pharmaceutically acceptable carrier or excipient.

[0050] In some embodiments, a method is provided for treating a target cancer, comprising administering one of the described fusion proteins, one of the described single-domain antigen-binding domains (sdABD), or one of the pharmaceutical compositions of the present disclosure. [Brief explanation of the drawing]

[0051] [Figure 1]This describes the "Form 1" type protease activation of the present invention, referred to herein as the "constrained cleavage construct" or "cc construct." In this embodiment, a typical construct has two TTAs and ABDs (as shown in Figure 1, these are both the same, but they may be different as described herein). Upon cleavage, the prodrug construct splits into three components: one containing an α-TTA domain linked to the active VH of αCD3 via a domain linker; a second containing an α-TTA domain linked to the active VL of αCD3 via a domain linker; and the "remaining" piece containing half-life extension domains linked to the inactive VH and VL. The two activity-variable domains then freely associate to form a functional anti-CD3 binding domain. In the "Form 1" embodiment, the resulting active component is trivalent. There are monovalent and divalent forms that bind to CD3 and TTA, resulting in a bispecific binding protein, but in some cases this trivalent form may be triply specific, so note that the monovalent binds to CD3, the monovalent binds to a first TTA, and the monovalent binds to a second TTA. Figure 1 also shows an anti-human serum albumin (HSA) domain, which in many embodiments is as sdABD as defined herein, but as considered herein, this is optional and / or can be replaced by other half-life extension domains. In addition, the half-life extension domain may also be at the N-terminus of the construct or as similarly internal. Figure 1 also has VH and VL of Fv, as well as iVH and iVL of pseudo-Fv in a specific order, for example, from N-terminus to C-terminus as VH-linker-VL (and iVL-linker-iVH), but as will be understood by those skilled in the art, these may also be in reverse (VL-linker-VH and iVH-linker-iVL). Alternatively, one of these Fvs may be in one orientation and another in another, but the expression of the protein in the orientations shown here was surprisingly higher than in the others. [Figure 2]The “Form 2” of the protease activation of the present invention is described herein as a “constrained non-cleavable construct” or “CNCL construct,” and may also be described herein as a “dimerized construct,” as considered herein. These constructs do not isomerize as considered herein. Upon cleavage, the two prodrug constructs split into four components: two half-life extension domains (in this case, sdABD for HSA) linked to pseudodomains (which may or may not be capable of self-assembly depending on the linker length and inactivating mutations); 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 may be the same and the other two different). In the “Form 2” embodiment, the resulting active components are hexavalent. There are divalent and tetravalent forms that bind to CD3 and TTA, resulting in a bispecific binding protein, but in some cases, this hexavalent form may be tripspecific, with the divalent binding to CD3, the divalent binding to a first TTA, and the divalent binding to a second TTA. Figure 2 also shows an anti-human serum albumin (HSA) domain, which in many embodiments is as sdABD as defined herein, but as considered herein, this is optional and / or can be replaced by other half-life extension domains. In addition, the half-life extension domain may also be at the N-terminus of the construct or as similarly internal. Figure 2 also has VH and VL of Fv, as well as iVH and iVL of pseudo-Fv in a specific order, for example, from N-terminus to C-terminus as VH-linker-VL (and iVL-linker-iVH), but as will be understood by those skilled in the art, these may also be in reverse (VL-linker-VH and iVH-linker-iVL). Alternatively, one of these Fvs may be in one orientation and another in another, but the expression of the protein in the orientations shown here was surprisingly higher than in the others. [Figure 3A]As will be further discussed herein, we present a construct of type "Form 3," which is sometimes referred to as a "semi-construct" or "semi-COBRA(trademark)" as outlined herein, as these are two different polypeptide chains that together constitute an MCE therapeutic agent. 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, an inactive variable domain is released, and then the two active variable domains intermolecularly assemble to form an active anti-CD3 binding domain. The two sdABD-TTAs bind to their corresponding receptors on the tumor cell surface, and cleavage is carried out by a protease. This allows for intermolecular assembly, as the molecules are physically held in place and prefer the assembly of the active anti-CD3 domains. As with Forms 1 and 2 described above, in this embodiment, the order of the variable domains from N-terminus to C-terminus may be reversed or mixed as well. Furthermore, sdABD(HSA) may be located at either the N-terminus or C-terminus of each semi-construct. Pro16 has an sdABD(HSA) at its C-terminus, and Pro17 has an sdABD(HSA) at its N-terminus. Pro19 has an sdABD(HSA) at its C-terminus. A shows a form 3 construct having a single sdABD-TTA domain per semi-construct, and B shows a form 3 construct having two sdABD-TTAs per semi-construct in a "dual-targeted" or "hetero-targeted" form. B uses FOLR1 and EGFR as the two TTAs, but it should be noted that other combinations, as outlined herein, may also be used. [Figure 3B]As will be further discussed herein, we present a construct of type "Form 3," which is sometimes referred to as a "semi-construct" or "semi-COBRA(trademark)" as outlined herein, as these are two different polypeptide chains that together constitute an MCE therapeutic agent. 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, an inactive variable domain is released, and then the two active variable domains intermolecularly assemble to form an active anti-CD3 binding domain. The two sdABD-TTAs bind to their corresponding receptors on the tumor cell surface, and cleavage is carried out by a protease. This allows for intermolecular assembly, as the molecules are physically held in place and prefer the assembly of the active anti-CD3 domains. As with Forms 1 and 2 described above, in this embodiment, the order of the variable domains from N-terminus to C-terminus may be reversed or mixed as well. Furthermore, sdABD(HSA) may be located at either the N-terminus or C-terminus of each semi-construct. Pro16 has an sdABD(HSA) at its C-terminus, and Pro17 has an sdABD(HSA) at its N-terminus. Pro19 has an sdABD(HSA) at its C-terminus. A shows a form 3 construct having a single sdABD-TTA domain per semi-construct, and B shows a form 3 construct having two sdABD-TTAs per semi-construct in a "dual-targeted" or "hetero-targeted" form. B uses FOLR1 and EGFR as the two TTAs, but it should be noted that other combinations, as outlined herein, may also be used. [Figure 4]The diagram shows a "Form 4" construct, similar to the "Form 2" construct, but containing only a single sdABD-TTA. The diagram shows EGFR from sdABD-TTA, but other TTAs may also be used, as will be understood by those skilled in the art. Upon cleavage, the prodrug construct splits into an active moiety that self-assembles into a dimeric active moiety containing two anti-TTA domains in the presence of two components: a half-life extension domain linked to a pseudo-Fv (in this case, sdABD to HSA) and a second active moiety from a different cleavage molecule. Note that in the "Form 4" embodiment, the resulting active component is tetravalent, with a divalent that binds to CD3 and a divalent that binds to TTA, resulting in a bispecific binding protein. Figure 4 also shows an anti-human serum albumin (HSA) domain, which in many embodiments is defined as sdABD(1 / 2) as herein, but as discussed herein, this is optional and / or can be replaced by other half-life extension domains. In addition, the half-life extension domain may also be at the N-terminus of the construct or similarly internal. Figure 4 also has VH and VL of Fv, as well as iVH and iVL of pseudo-Fv 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 may also be in reverse (VL-linker-VH and iVH-linker-iVL). Alternatively, one of these Fv may be in one orientation and another in the other, but the expression of the protein in the orientation shown herein was remarkably higher than in the other orientations. [Figure 5A] The present invention shows numerous single-domain tumor target antigen-binding domain (sdTTA-ABD) sequences, with the CDR underlined. As will be further outlined herein, these domains can assemble in a broad configuration of the invention, including "Form 1," "Form 2," "Form 3," and "Form 4" orientations. [Figure 5B]The present invention shows numerous single-domain tumor target antigen-binding domain (sdTTA-ABD) sequences, with the CDR underlined. As will be further outlined herein, these domains can assemble in a broad configuration of the invention, including "Form 1," "Form 2," "Form 3," and "Form 4" orientations. [Figure 5C] The present invention shows numerous single-domain tumor target antigen-binding domain (sdTTA-ABD) sequences, with the CDR underlined. As will be further outlined herein, these domains can assemble in a broad configuration of the invention, including "Form 1," "Form 2," "Form 3," and "Form 4" orientations. [Figure 5D] The present invention shows numerous single-domain tumor target antigen-binding domain (sdTTA-ABD) sequences, with the CDR underlined. As will be further outlined herein, these domains can assemble in a broad configuration of the invention, including "Form 1," "Form 2," "Form 3," and "Form 4" orientations. [Figure 5E] The present invention shows numerous single-domain tumor target antigen-binding domain (sdTTA-ABD) sequences, with the CDR underlined. As will be further outlined herein, these domains can assemble in a broad configuration of the invention, including "Form 1," "Form 2," "Form 3," and "Form 4" orientations. [Figure 5F] The present invention shows numerous single-domain tumor target antigen-binding domain (sdTTA-ABD) sequences, with the CDR underlined. As will be further outlined herein, these domains can assemble in a broad configuration of the invention, including "Form 1," "Form 2," "Form 3," and "Form 4" orientations. [Figure 5G] The present invention shows numerous single-domain tumor target antigen-binding domain (sdTTA-ABD) sequences, with the CDR underlined. As will be further outlined herein, these domains can assemble in a broad configuration of the invention, including "Form 1," "Form 2," "Form 3," and "Form 4" orientations. [Figure 5H]The present invention shows numerous single-domain tumor target antigen-binding domain (sdTTA-ABD) sequences, with the CDR underlined. As will be further outlined herein, these domains can assemble in a broad configuration of the invention, including "Form 1," "Form 2," "Form 3," and "Form 4" orientations. [Figure 5I] The present invention shows numerous single-domain tumor target antigen-binding domain (sdTTA-ABD) sequences, with the CDR underlined. As will be further outlined herein, these domains can assemble in a broad configuration of the invention, including "Form 1," "Form 2," "Form 3," and "Form 4" orientations. [Figure 5J] The present invention shows numerous single-domain tumor target antigen-binding domain (sdTTA-ABD) sequences, with the CDR underlined. As will be further outlined herein, these domains can assemble in a broad configuration of the invention, including "Form 1," "Form 2," "Form 3," and "Form 4" orientations. [Figure 5K] The present invention shows numerous single-domain tumor target antigen-binding domain (sdTTA-ABD) sequences, with the CDR underlined. As will be further outlined herein, these domains can assemble in a broad configuration of the invention, including "Form 1," "Form 2," "Form 3," and "Form 4" orientations. [Figure 5L] The present invention shows numerous single-domain tumor target antigen-binding domain (sdTTA-ABD) sequences, with the CDR underlined. As will be further outlined herein, these domains can assemble in a broad configuration of the invention, including "Form 1," "Form 2," "Form 3," and "Form 4" orientations. [Figure 5M] The CDR shows a number of single-domain tumor target antigen-binding domain (sdTTA-ABD) sequences of the present invention, with the CDR underlined. As will be further outlined herein, these domains can assemble in a broad configuration of the present invention, including “Form 1,” “Form 2,” “Form 3,” and “Form 4” orientations. [Figure 6] It exhibits many half-life extension domains. [Figure 7A] This shows many αCD3 variable weight and variable light domains, including active domains (e.g., "VL" or "VH", sometimes also referred to as "aVL" or "aVH") and inactive domains (e.g., "VLi" or "VHi", sometimes also referred to as "iVL" or "iVH"). CDR is underlined. [Figure 7B] This shows many αCD3 variable weight and variable light domains, including active domains (e.g., "VL" or "VH", sometimes also referred to as "aVL" or "aVH") and inactive domains (e.g., "VLi" or "VHi", sometimes also referred to as "iVL" or "iVH"). CDR is underlined. [Figure 8A] This exhibits many suitable protease cleavage sites. As those skilled in the art will understand, these cleavage sites can be used as cleavage linkers. In some embodiments, for example, when a more flexible cleavage linker is required, additional amino acids (generally glycine and serine) may be present at either or both of the N-terminus and C-terminus of these cleavage sites. [Figure 8B] Many suitable protease cleavage sites are shown. As those skilled in the art will understand, these cleavage sites can be used as cleavage linkers. In some embodiments, for example, when a more flexible cleavage linker is required, additional amino acids (generally glycine and serine) may be present at either or both of the N-terminus and C-terminus of these cleavage sites. [Figure 8C] Many suitable protease cleavage sites are shown. As those skilled in the art will understand, these cleavage sites can be used as cleavage linkers. In some embodiments, for example, when a more flexible cleavage linker is required, additional amino acids (generally glycine and serine) may be present at either or both of the N-terminus and C-terminus of these cleavage sites. [Figure 8D]Many suitable protease cleavage sites are shown. As those skilled in the art will understand, these cleavage sites can be used as cleavage linkers. In some embodiments, for example, when a more flexible cleavage linker is required, additional amino acids (generally glycine and serine) may be present at either or both of the N-terminus and C-terminus of these cleavage sites. [Figure 9A] Many of the sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and in bold, linkers are double underlined (cleavage linkers are italicized and double underlined), and domain isolation is indicated by " / ". All His6 tags are optional, as they may be used to reduce immunogenicity in humans and may be purification tags. [Figure 9B] Many of the sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and in bold, linkers are double underlined (cleavage linkers are italicized and double underlined), and domain isolation is indicated by " / ". All His6 tags are optional, as they may be used to reduce immunogenicity in humans and may be purification tags. [Figure 9C] Many of the sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and in bold, linkers are double underlined (cleavage linkers are italicized and double underlined), and domain isolation is indicated by " / ". All His6 tags are optional, as they may be used to reduce immunogenicity in humans and may be purification tags. [Figure 9D] Many of the sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and in bold, linkers are double underlined (cleavage linkers are italicized and double underlined), and domain isolation is indicated by " / ". All His6 tags are optional, as they may be used to reduce immunogenicity in humans and may be purification tags. [Figure 9E]Many of the sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and in bold, linkers are double underlined (cleavage linkers are italicized and double underlined), and domain isolation is indicated by " / ". All His6 tags are optional, as they may be used to reduce immunogenicity in humans and may be purification tags. [Figure 9F] Many of the sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and in bold, linkers are double underlined (cleavage linkers are italicized and double underlined), and domain isolation is indicated by " / ". All His6 tags are optional, as they may be used to reduce immunogenicity in humans and may be purification tags. [Figure 9G] Many of the sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and in bold, linkers are double underlined (cleavage linkers are italicized and double underlined), and domain isolation is indicated by " / ". All His6 tags are optional, as they may be used to reduce immunogenicity in humans and may be purification tags. [Figure 9H] Many of the sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and in bold, linkers are double underlined (cleavage linkers are italicized and double underlined), and domain isolation is indicated by " / ". All His6 tags are optional, as they may be used to reduce immunogenicity in humans and may be purification tags. [Figure 9I] Many of the sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and in bold, linkers are double underlined (cleavage linkers are italicized and double underlined), and domain isolation is indicated by " / ". All His6 tags are optional, as they may be used to reduce immunogenicity in humans and may be purification tags. [Figure 9J] Many of the sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and in bold, linkers are double underlined (cleavage linkers are italicized and double underlined), and domain isolation is indicated by " / ". All His6 tags are optional, as they may be used to reduce immunogenicity in humans and may be purification tags. [Figure 9K] Many of the sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and in bold, linkers are double underlined (cleavage linkers are italicized and double underlined), and domain isolation is indicated by " / ". All His6 tags are optional, as they may be used to reduce immunogenicity in humans and may be purification tags. [Figure 9L] Many of the sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and in bold, linkers are double underlined (cleavage linkers are italicized and double underlined), and domain isolation is indicated by " / ". All His6 tags are optional, as they may be used to reduce immunogenicity in humans and may be purification tags. [Figure 9M] Many of the sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and in bold, linkers are double underlined (cleavage linkers are italicized and double underlined), and domain isolation is indicated by " / ". All His6 tags are optional, as they may be used to reduce immunogenicity in humans and may be purification tags. [Figure 9N]Many of the sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and in bold, linkers are double underlined (cleavage linkers are italicized and double underlined), and domain isolation is indicated by " / ". All His6 tags are optional, as they may be used to reduce immunogenicity in humans and may be purification tags. [Figure 9O] Many of the sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and in bold, linkers are double underlined (cleavage linkers are italicized and double underlined), and domain isolation is indicated by " / ". All His6 tags are optional, as they may be used to reduce immunogenicity in humans and may be purification tags. [Figure 9P] Many of the sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and in bold, linkers are double underlined (cleavage linkers are italicized and double underlined), and domain isolation is indicated by " / ". All His6 tags are optional, as they may be used to reduce immunogenicity in humans and may be purification tags. [Figure 9Q] Many of the sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and in bold, linkers are double underlined (cleavage linkers are italicized and double underlined), and domain isolation is indicated by " / ". All His6 tags are optional, as they may be used to reduce immunogenicity in humans and may be purification tags. [Figure 9R] Many of the sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and in bold, linkers are double underlined (cleavage linkers are italicized and double underlined), and domain isolation is indicated by " / ". All His6 tags are optional, as they may be used to reduce immunogenicity in humans and may be purification tags. [Figure 9S] Many of the sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and in bold, linkers are double underlined (cleavage linkers are italicized and double underlined), and domain isolation is indicated by " / ". All His6 tags are optional, as they may be used to reduce immunogenicity in humans and may be purification tags. [Figure 9T] Many of the sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and in bold, linkers are double underlined (cleavage linkers are italicized and double underlined), and domain isolation is indicated by " / ". All His6 tags are optional, as they may be used to reduce immunogenicity in humans and may be purification tags. [Figure 9U] Many of the sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and in bold, linkers are double underlined (cleavage linkers are italicized and double underlined), and domain isolation is indicated by " / ". All His6 tags are optional, as they may be used to reduce immunogenicity in humans and may be purification tags. [Figure 9V] Many of the sequences of the present invention are shown, but many additional sequences can also be found in the sequence listing. CDRs are underlined and in bold, linkers are double underlined (cleavage linkers are italicized and double underlined), and domain isolation is indicated by " / ". All His6 tags are optional, as they may be used to reduce immunogenicity in humans and may be purification tags. [Figure 10A] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10B] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10C] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10D] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10E] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10F] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10G] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10H] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10I] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10J] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10K] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10L] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10M] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10N] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10O] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10P] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10Q] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10R] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10S] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10T] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10U] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10V] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10W] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10X] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10Y] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10Z] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10AA] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10BB] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10CC] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10DD] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 10EE] The amino acid sequences of exemplary Form 2 constructs containing numerous sdABD-B7H3 and pseudo-Fv domains (e.g., Vli2 / Vhi2 domains) are shown. [Figure 11] The design and expected folding mechanism of COBRA are shown, with the expected structure of the non-cleaved molecule shown at the top. This then binds to a tumor antigen (EGFR in the case of MVC-101), losing its CD3 binding and binding to human serum albumin. The middle shows the expected cleavage product, and the left shows the active dimer. [Figure 12A] Further sequences of some COBRAs of the present invention are shown. [Figure 12B] Further sequences of some COBRAs of the present invention are shown. [Figure 12C] Further sequences of some COBRAs of the present invention are shown. [Figure 12D] Further sequences of some COBRAs of the present invention are shown. [Figure 12E] Further sequences of some COBRAs of the present invention are shown. [Figure 12F] Further sequences of some COBRAs of the present invention are shown. [Figure 12G] Further sequences of some COBRAs of the present invention are shown. [Figure 12H] Further sequences of some COBRAs of the present invention are shown. [Figure 12I] Further sequences of some COBRAs of the present invention are shown. [Figure 12J] Further sequences of some COBRAs of the present invention are shown. [Figure 12K] Further sequences of some COBRAs of the present invention are shown. [Figure 12L] Further sequences of some COBRAs of the present invention are shown. [Figure 12M] Further sequences of some COBRAs of the present invention are shown. [Figure 12N] Further sequences of some COBRAs of the present invention are shown. [Figure 12O] Further sequences of some COBRAs of the present invention are shown. [Figure 12P] Further sequences of some COBRAs of the present invention are shown. [Figure 12Q] Further sequences of some COBRAs of the present invention are shown. [Figure 13] This invention demonstrates that the Type 2 construct of the present invention, once cleaved and dimerized, is rapidly removed from injected mice. [Figure 14] This shows the binding dynamics of Pro225. [Figure 15A] A Type 2 construct, in this case Pro225, is shown to degenerate established solid tumors in mice. [Figure 15B] A Type 2 construct, in this case Pro225, is shown to degenerate established solid tumors in mice. [Figure 16A] The present invention demonstrates that a Type 2 construct, in this case Pro225, exhibits increased tolerance compared to an intrinsically active T cell engager. [Figure 16B] The present invention demonstrates that a Type 2 construct, in this case Pro225, exhibits increased tolerance compared to an intrinsically active T cell engager. [Figure 16C] We demonstrate that treatment with Pro225 results in lower cytokine release in mice compared to an intrinsically active bispecific agent. Compared to an intrinsically active T cell engager, Pro225 does not induce IL2, TNFa, and IL10 in NHP, nor does it induce mouse IL6 in mice. [Figure 16D] We demonstrate that treatment with Pro225 results in lower cytokine release in mice compared to an intrinsically active bispecific agent. Compared to an intrinsically active T cell engager, Pro225 does not induce IL2, TNFa, and IL10 in NHP, nor does it induce mouse IL6 in mice. [Figure 17] As outlined in Example 2, the efficacy of numerous Form 2 constructs of the present invention in T cell-dependent cell-mediated cytotoxicity (TDCC) assays is demonstrated. Pro233 is an aEGFR construct with an MMP9 cleavage site. Pro565 is an aEpCAM(h664) construct with an MMP9 cleavage site. Pro566 is an aEpCAM(h665) construct with an MMP9 cleavage site. Pro623 is a hetero-COBRA and MMP9 site of aEGFR and aEpCAM(h664). Pro624 is a hetero-COBRA and MMP9 site of aEGFR and aEpCAM(h665). [Figure 18] The efficacy of several Type 2 constructs of the present invention in the TDCC assay outlined in Example 2 is demonstrated. Pro233 is an aEGFR construct having an MMP9 cleavage site. Pro311 is an aFOLR1 construct having an MMP9 cleavage site. Pro421 is a heterocobra and MMP9 site of aEGFR and aFOLR1. [Figure 19]The efficacy of several Type 2 constructs of the present invention in the TDCC assay outlined in Example 2 is demonstrated. Pro225 is an aB7H3 construct having an MMP9 cleavage site. Pro566 is an aEpCAM construct having an MMP9 cleavage site. Pro656 is a hetero-COBRA and MMP9 site of aB7H3 and aEpCAM. Pro658 is a hetero-COBRA and MMP9 site of aEpCAM and aB7H3. [Figure 20] The efficacy of numerous Form 2 constructs of the present invention in the TDCC assay outlined in Example 2 for two different cell lines is demonstrated. Pro225 is an aB7H3 construct with an MMP9 cleavage site. Pro566 is an aEpCAM construct with an MMP9 cleavage site. Pro656 is a hetero-COBRA and MMP9 site of aB7H3 and aEpCAM. HT29 is an epithelial cell line that, unlike the Raji cell line, performs well in xenotransplantation to mice. HT29 expresses both target genes (B7H3 and EpCAM), in which case B7H3 expression was knocked out using CRISPR. Therefore, hetero-COBRA and EpCAM single-target COBRA killed both, but B7H3 single-target COBRA did not. [Figure 21]The efficacy of numerous Form 2 constructs of the present invention in the TDCC assay outlined in Example 2 against HT29 cell lines with high EpCAM expression and low Trop2 expression is demonstrated. Pro824 is aEpCAM X aTrop2 with an MMP9 linker heterocobra. Pro825 is aEpCAM X aTrop2 with an NCL heterocobra (e.g., a non-cleavable control construct). Pro826 is aTrop2 X aEpCAM heterocobra with an MMP9 linker. Pro827 is aTrop2 X aEpCAM with an NCL heterocobra (e.g., a non-cleavable control construct). Pro677 is aTrop2 / MMP9 cobra, and Pro566 is aEpCAM / MMP9 cobra. As the levels of the two antigens change, heterocobras maintain good killing power, while killing power by single-specific (e.g., targeting a single tumor antigen) cobras changes. Monospecific COBRA also fails to kill when the expression level of a specific antigen (in this case, Trop2) is reduced. The same applies to Figures 22 and 23. [Figure 22] The efficacy of numerous Form 2 constructs of the present invention in the TDCC assay outlined in Example 2 against an HT116 cell line having high EpCAM expression and very low Trop2 expression is demonstrated. Pro824 is aEpCAM X aTrop2 with an MMP9 linker heterocobra. Pro825 is aEpCAM X aTrop2 with an NCL heterocobra (non-cleavage control). Pro826 is aTrop2 X aEpCAM with an MMP9 linker heterocobra. Pro827 is aTrop2 X aEpCAM with an NCL heterocobra (non-cleavage control). Pro677 is an aTrop2 / MMP9 monospecific cobra, and Pro566 is an aEpCAM / MMP9 monospecific cobra. [Figure 23]The efficacy of numerous Form 2 constructs of the present invention in the TDCC assay outlined in Example 2 against a BXPC3 cell line having moderate EpCAM expression and high Trop2 expression is demonstrated. Pro824 is an aEpCAM X aTrop2 with an MMP9 linker heterocobra. Pro825 is an aEpCAM X aTrop2 with an NCL heterocobra (non-cleavage control). Pro826 is an aTrop2 X aEpCAM heterocobra with an MMP9 linker. Pro827 is an aTrop2 X aEpCAM with an NCL heterocobra (non-cleavage control). Pro677 is an aTrop2 / MMP9 monospecific cobra, and Pro566 is an aEpCAM / MMP9 monospecific cobra. [Figure 24] The in vivo efficacy of aEpCAM COBRA with an MMP9 cleavage site is demonstrated using Protocol 2 of Example 3. Pro566 showed efficacy against LoVo tumors, as well as xenografts of HT29, BxPC3, and SW403 tumors. [Figure 25] The in vivo efficacy of aTrop2 COBRA with an MMP9 cleavage site, using Protocol 2 of Example 3, is demonstrated. Pro677 showed efficacy against BxPC3 tumors and HCC827 tumor xenografts. [Figure 26] The in vivo efficacy of aB7H3 COBRA with an MMP9 cleavage site is demonstrated using Protocol 3 of Example 3. Pro225 showed efficacy against A549 tumors. [Figure 27A]This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-HER2 (aHer2 hVIB1139) conditionally killed human or cyno-HER2 expressing tumor cell lines in a T cell-dependent cell-mediated cytotoxicity (TDCC) assay. A: Human HER2-Raji cells tested with various fusion proteins. B: Cyno-Her2-Raji cells tested with various fusion proteins. C: Raji cells tested with various fusion proteins. D: SKOV3 cells with high HER2 expression tested with various fusion proteins. E: HT29 cells with low HER2 expression tested with various fusion proteins. The fusion proteins tested were Pro1123 NCL (non-cleavable control), Pro1117 MMP9 (non-cleavable MMP9-containing COBRA), or Pro1117 MMP9cl (cleavable MMP9-containing COBRA), or Pro1060 Pro51 format (a positive control × CD3-positive control Pro51 similar to anti-EGFR, as described in U.S. Patent No. 2020 / 0347132 and International Publication No. WO2020 / 181140), and Pro1069 AD (active domain only). The amino acid sequence of Pro1117 is shown in Figure 74 and Sequence ID No. 493. [Figure 27B] This is a series of graphs demonstrating that monospecific COBRAs containing two sdABD-HER2 (aHer2 hVIB1139) conditionally killed human or cyno-HER2 expressing tumor cell lines in a T cell-dependent cytotoxicity (TDCC) assay. A: Human HER2-Raji cells were tested with various fusion proteins. The fusion proteins tested were Pro1123 NCL (non-cleavable control), Pro1117 MMP9 (non-cleavable MMP9-containing COBRA), or Pro1117 MMP9cl (cleavable MMP9-containing COBRA), or Pro1060 Pro51 format (a positive control similar to anti-EGFR × CD3-positive control Pro51, as described in U.S. Patent No. 2020 / 0347132 and International Publication No. WO2020 / 181140), and Pro1069 AD (active domain only). The amino acid sequence of Pro1117 is shown in Figure 74 and SEQ ID NO. 493. [Figure 27C]This is a series of graphs demonstrating that monospecific COBRAs containing two sdABD-HER2 (aHer2 hVIB1139) conditionally killed human or cyno-HER2 expressing tumor cell lines in a T cell-dependent cytotoxicity (TDCC) assay. C:Raji cells were tested with various fusion proteins. The tested fusion proteins were Pro1123 NCL (non-cleavable control), Pro1117 MMP9 (non-cleavable MMP9-containing COBRA), or Pro1117 MMP9cl (cleavable MMP9-containing COBRA), or Pro1060 Pro51 format (a positive control similar to anti-EGFR × CD3-positive control Pro51, as described in U.S. Patent No. 2020 / 0347132 and International Publication No. WO2020 / 181140), and Pro1069 AD (active domain only). The amino acid sequence of Pro1117 is shown in Figure 74 and SEQ ID NO. 493. [Figure 27D] This is a series of graphs demonstrating that monospecific COBRAs containing two sdABD-HER2 (aHer2 hVIB1139) conditionally killed human or cyno-HER2 expressing tumor cell lines in a T cell-dependent cytotoxicity (TDCC) assay. D: SKOV3 cells overexpressing HER2 were tested with various fusion proteins. The fusion proteins tested were Pro1123 NCL (non-cleavable control), Pro1117 MMP9 (non-cleavable MMP9-containing COBRA), or Pro1117 MMP9cl (cleavable MMP9-containing COBRA), or Pro1060 Pro51 format (a positive control × CD3-positive control Pro51 format similar to anti-EGFR, as described in U.S. Patent No. 2020 / 0347132 and International Publication No. WO2020 / 181140), and Pro1069 AD (active domain only). The amino acid sequence of Pro1117 is shown in Figure 74 and SEQ ID NO: 493. [Figure 27E]This is a series of graphs demonstrating that monospecific COBRAs containing two sdABD-HER2 (aHer2 hVIB1139) conditionally killed human or cyno-HER2 expressing tumor cell lines in a T cell-dependent cell-mediated cytotoxicity (TDCC) assay. HT29 cells with low E:HER2 expression were tested with various fusion proteins. The fusion proteins tested were Pro1123 NCL (non-cleavable control), Pro1117 MMP9 (non-cleavable MMP9-containing COBRA), or Pro1117 MMP9cl (cleavable MMP9-containing COBRA), or Pro1060 Pro51 format (a positive control × CD3-positive control Pro51 format similar to anti-EGFR, as described in U.S. Patent No. 2020 / 0347132 and International Publication No. WO2020 / 181140), and Pro1069 AD (active domain only). The amino acid sequence of Pro1117 is shown in Figure 74 and SEQ ID NO. 493. [Figure 28A] This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-HER2 (aHer2 h1159) conditionally killed human or cyno-HER2 expressing tumor cell lines in a TDCC assay. A: Human HER2-Raji cells were tested with various fusion proteins. The fusion proteins tested were Pro1110 NCL, Pro1109 MMP9, Pro1109 MMP9cl, Pro1062 Pro51 (a positive control similar in form to anti-EGFR × CD3 positive control Pro51, as described in U.S. Patent No. 2020 / 0347132 and International Publication No. WO2020 / 181140), and Pro1071 AD. The amino acid sequence of Pro1109 is shown in Figure 73 and SEQ ID NO. 491. [Figure 28B]This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-HER2 (aHer2 h1159) conditionally killed human or cyno HER2-expressing tumor cell lines in a TDCC assay. B:Cyno HER2-Raji cells were tested with various fusion proteins. The fusion proteins tested were Pro1110 NCL, Pro1109 MMP9, Pro1109 MMP9cl, Pro1062 Pro51 (a positive control similar in form to anti-EGFR × CD3 positive control Pro51, as described in U.S. Patent No. 2020 / 0347132 and International Publication No. WO2020 / 181140), and Pro1071 AD. The amino acid sequence of Pro1109 is shown in Figure 73 and SEQ ID NO. 491. [Figure 28C] This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-HER2 (aHer2 h1159) conditionally killed human or cyno-HER2 expressing tumor cell lines in a TDCC assay. C:Raji cells were tested with various fusion proteins. The fusion proteins tested were Pro1110 NCL, Pro1109 MMP9, Pro1109 MMP9cl, Pro1062 Pro51 (a positive control similar in form to anti-EGFR × CD3 positive control Pro51, as described in U.S. Patent No. 2020 / 0347132 and International Publication No. WO2020 / 181140), and Pro1071 AD. The amino acid sequence of Pro1109 is shown in Figure 73 and SEQ ID NO. 491. [Figure 28D]This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-HER2 (aHer2 h1159) conditionally killed human or cyno-HER2 expressing tumor cell lines in a TDCC assay. D: SKOV3 cells highly expressing HER2 were tested with various fusion proteins. The fusion proteins tested were Pro1110 NCL, Pro1109 MMP9, Pro1109 MMP9cl, Pro1062 Pro51 (a positive control similar in form to anti-EGFR × CD3 positive control Pro51, as described in U.S. Patent No. 2020 / 0347132 and International Publication No. WO2020 / 181140), and Pro1071 AD. The amino acid sequence of Pro1109 is shown in Figure 73 and SEQ ID NO. 491. [Figure 28E] This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-HER2 (aHer2 h1159) conditionally killed human or cyno-HER2 expressing tumor cell lines in a TDCC assay. HT29 cells with low E:HER2 expression were tested with various fusion proteins. The fusion proteins tested were Pro1110 NCL, Pro1109 MMP9, Pro1109 MMP9cl, Pro1062 Pro51 (a positive control similar in form to anti-EGFR × CD3 positive control Pro51, as described in U.S. Patent No. 2020 / 0347132 and International Publication No. WO2020 / 181140), and Pro1071 AD. The amino acid sequence of Pro1109 is shown in Figure 73 and SEQ ID NO. 491. [Figure 29A]This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-HER2 (aHER2 h1162) conditionally killed human or cyno HER2-expressing tumor cell lines in a TDCC assay. A: Human HER2-Raji cells tested with various fusion proteins. B: Cyno HER2-Raji cells tested with various fusion proteins. C: Raji cells tested with various fusion proteins. D: SKOV3 cells with high HER2 expression tested with various fusion proteins. E: HT29 cells with low HER2 expression tested with various fusion proteins. The fusion proteins tested were Pro1112 NCL, Pro1111 MMP9, Pro1111 MMP9cl, Pro1064 Pro51, and Pro1073 AD. The amino acid sequence of Pro1111 is shown in Figure 73 and SEQ ID NO: 492. [Figure 29B] This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-HER2 (aHER2 h1162) conditionally killed human or cyno HER2-expressing tumor cell lines in a TDCC assay. A: Human HER2-Raji cells tested with various fusion proteins. B: Cyno HER2-Raji cells tested with various fusion proteins. C: Raji cells tested with various fusion proteins. D: SKOV3 cells with high HER2 expression tested with various fusion proteins. E: HT29 cells with low HER2 expression tested with various fusion proteins. The fusion proteins tested were Pro1112 NCL, Pro1111 MMP9, Pro1111 MMP9cl, Pro1064 Pro51, and Pro1073 AD. The amino acid sequence of Pro1111 is shown in Figure 73 and SEQ ID NO: 492. [Figure 29C]This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-HER2 (aHER2 h1162) conditionally killed human or cyno HER2-expressing tumor cell lines in a TDCC assay. A: Human HER2-Raji cells tested with various fusion proteins. B: Cyno HER2-Raji cells tested with various fusion proteins. C: Raji cells tested with various fusion proteins. D: SKOV3 cells with high HER2 expression tested with various fusion proteins. E: HT29 cells with low HER2 expression tested with various fusion proteins. The fusion proteins tested were Pro1112 NCL, Pro1111 MMP9, Pro1111 MMP9cl, Pro1064 Pro51, and Pro1073 AD. The amino acid sequence of Pro1111 is shown in Figure 73 and SEQ ID NO: 492. [Figure 29D] This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-HER2 (aHER2 h1162) conditionally killed human or cyno HER2-expressing tumor cell lines in a TDCC assay. A: Human HER2-Raji cells tested with various fusion proteins. B: Cyno HER2-Raji cells tested with various fusion proteins. C: Raji cells tested with various fusion proteins. D: SKOV3 cells with high HER2 expression tested with various fusion proteins. E: HT29 cells with low HER2 expression tested with various fusion proteins. The fusion proteins tested were Pro1112 NCL, Pro1111 MMP9, Pro1111 MMP9cl, Pro1064 Pro51, and Pro1073 AD. The amino acid sequence of Pro1111 is shown in Figure 73 and SEQ ID NO: 492. [Figure 29E]This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-HER2 (aHER2 h1162) conditionally killed human or cyno HER2-expressing tumor cell lines in a TDCC assay. A: Human HER2-Raji cells tested with various fusion proteins. B: Cyno HER2-Raji cells tested with various fusion proteins. C: Raji cells tested with various fusion proteins. D: SKOV3 cells with high HER2 expression tested with various fusion proteins. E: HT29 cells with low HER2 expression tested with various fusion proteins. The fusion proteins tested were Pro1112 NCL, Pro1111 MMP9, Pro1111 MMP9cl, Pro1064 Pro51, and Pro1073 AD. The amino acid sequence of Pro1111 is shown in Figure 73 and SEQ ID NO: 492. [Figure 30A] This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-HER2 (aHer2 h1156) conditionally killed human or cyno HER2-expressing tumor cell lines in a TDCC assay. A: Human HER2-Raji cells tested with various fusion proteins. B: Cyno HER2-Raji cells tested with various fusion proteins. C: Raji cells tested with various fusion proteins. D: SKOV3 cells with high HER2 expression tested with various fusion proteins. E: HT29 cells with low HER2 expression tested with various fusion proteins. The fusion proteins tested were Pro1124 NCL, Pro1118 MMP9, Pro 1118 MMP9cl, and Pro106 Pro51. The amino acid sequence of Pro1118 is shown in Figure 74 and SEQ ID NO: 494. [Figure 30B]This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-HER2 (aHer2 h1156) conditionally killed human or cyno HER2-expressing tumor cell lines in a TDCC assay. A: Human HER2-Raji cells tested with various fusion proteins. B: Cyno HER2-Raji cells tested with various fusion proteins. C: Raji cells tested with various fusion proteins. D: SKOV3 cells with high HER2 expression tested with various fusion proteins. E: HT29 cells with low HER2 expression tested with various fusion proteins. The fusion proteins tested were Pro1124 NCL, Pro1118 MMP9, Pro 1118 MMP9cl, and Pro106 Pro51. The amino acid sequence of Pro1118 is shown in Figure 74 and SEQ ID NO: 494. [Figure 30C] This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-HER2 (aHer2 h1156) conditionally killed human or cyno HER2-expressing tumor cell lines in a TDCC assay. A: Human HER2-Raji cells tested with various fusion proteins. B: Cyno HER2-Raji cells tested with various fusion proteins. C: Raji cells tested with various fusion proteins. D: SKOV3 cells with high HER2 expression tested with various fusion proteins. E: HT29 cells with low HER2 expression tested with various fusion proteins. The fusion proteins tested were Pro1124 NCL, Pro1118 MMP9, Pro 1118 MMP9cl, and Pro106 Pro51. The amino acid sequence of Pro1118 is shown in Figure 74 and SEQ ID NO: 494. [Figure 30D]This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-HER2 (aHer2 h1156) conditionally killed human or cyno HER2-expressing tumor cell lines in a TDCC assay. A: Human HER2-Raji cells tested with various fusion proteins. B: Cyno HER2-Raji cells tested with various fusion proteins. C: Raji cells tested with various fusion proteins. D: SKOV3 cells with high HER2 expression tested with various fusion proteins. E: HT29 cells with low HER2 expression tested with various fusion proteins. The fusion proteins tested were Pro1124 NCL, Pro1118 MMP9, Pro 1118 MMP9cl, and Pro106 Pro51. The amino acid sequence of Pro1118 is shown in Figure 74 and SEQ ID NO: 494. [Figure 30E] This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-HER2 (aHer2 h1156) conditionally killed human or cyno HER2-expressing tumor cell lines in a TDCC assay. A: Human HER2-Raji cells tested with various fusion proteins. B: Cyno HER2-Raji cells tested with various fusion proteins. C: Raji cells tested with various fusion proteins. D: SKOV3 cells with high HER2 expression tested with various fusion proteins. E: HT29 cells with low HER2 expression tested with various fusion proteins. The fusion proteins tested were Pro1124 NCL, Pro1118 MMP9, Pro 1118 MMP9cl, and Pro106 Pro51. The amino acid sequence of Pro1118 is shown in Figure 74 and SEQ ID NO: 494. [Figure 31A] This is a series of graphs showing the results for aHER2 Pro51 fusion proteins leading to Pro1043 VIB1139, Pro 1044 VIB1156, Pro1045 VIB1159, and Pro1047 VIB1162, which were selected to demonstrate good activity and cynomolgus monkey cross-reactivity in the TDCC assay. On the other hand, Pro1036 VIB1055 and Pro1038 VIB1059 showed low activity. [Figure 31B] This is a series of graphs showing the results for aHER2 Pro51 fusion proteins leading to Pro1043 VIB1139, Pro 1044 VIB1156, Pro1045 VIB1159, and Pro1047 VIB1162, which were selected to demonstrate good activity and cynomolgus monkey cross-reactivity in the TDCC assay. On the other hand, Pro1036 VIB1055 and Pro1038 VIB1059 showed low activity. [Figure 31C] This is a series of graphs showing the results for aHER2 Pro51 fusion proteins leading to Pro1043 VIB1139, Pro 1044 VIB1156, Pro1045 VIB1159, and Pro1047 VIB1162, which were selected to demonstrate good activity and cynomolgus monkey cross-reactivity in the TDCC assay. On the other hand, Pro1036 VIB1055 and Pro1038 VIB1059 showed low activity. [Figure 32] This graph shows that HER2 / MMP9 COBRA causes regression in established N87 xenografts. In this assay, Pro1118 was used at a dose of 100 ug / kg. [Figure 33] This graph shows that the HER2 / MMP9 COBRA PK matched mouse HER2 binding. In this assay, Pro1111 was used at a dose of 30 ug / kg. [Figure 34] This table shows epitope binning for various HER2 sdAbs. 100 nM competitive antibodies were tested with 333 nM saturated antibodies. The aHER2 antibodies tested were VIB1121, VIB1139, VIB1058, VIB1097, trastuzumab, VIB1156, VIB1160, VIB1159, and VIB1162. "B" indicates binding of a competing Ab, and "NB" indicates no binding of a competing Ab. [Figure 35]This table shows epitope binning for various HER2 sdAbs. 100 nM competitive antibodies were tested with 333 nM saturated antibodies. The antibodies tested were Pro1118, Pro1111, trastuzumab, and pertuzumab. "B" indicates binding of a competing Ab, and "NB" indicates no binding of a competing Ab. [Figure 36] This is a list of amino acid positions and sequences for epitope mapping of HER2 sdAb h1156 (SEQ ID NO: 503) and HER2 sdAb h1162 (SEQ ID NO: 511). [Figure 37] This table shows the affinity of HER2 sdAbs in Pro51 format. Various sdAb and fusion protein combinations were evaluated in humans, cynos, and mice. The combinations are as follows: 1055 and Pro1036, 1058 and Pro1037, 1059 and Pro1038, 1091 and Pro1039, 1092 and Pro1040, 1097 and Pro1041, 1121 and Pro1042, 1139 and Pro1043, 1156 and Pro1044, 1159 and Pro1045, 1160 and Pro1046, 1162 and Pro1047, h1058 and Pro1056, h1092 and Pro1057, h1097 and Pro1058, h1121 and Pro1059, h1139 and Pro1060, h1156 and Pro1061, h1159 and Pro1062, h1160 and Pro1063, and h1162 and Pro1064. [Figure 38A] This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-CA9 (aCA9 h407) conditionally killed human or cyno-CA9 expressing tumor cell lines in a TDCC assay. A: Human CA9-Raji cells were tested with various fusion proteins. The fusion proteins tested were Pro514 NCL, Pro518 MMP9, Pro518 MMP9cl, Pro511 Pro51, and Pro521 AD. The amino acid sequence of Pro518 is shown in Figure 10 AA and SEQ ID NO: 331. [Figure 38B]This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-CA9 (aCA9 h407) conditionally killed human or cyno CA9-expressing tumor cell lines in a TDCC assay. B:Cyno CA9-Raji cells were tested with various fusion proteins. The fusion proteins tested were Pro514 NCL, Pro518 MMP9, Pro518 MMP9cl, Pro511 Pro51, and Pro521 AD. The amino acid sequence of Pro518 is shown in Figure 10 AA and SEQ ID NO: 331. [Figure 38C] This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-CA9 (aCA9 h407) conditionally killed human or cyno-CA9 expressing tumor cell lines in a TDCC assay. C:HT29 cells were tested with various fusion proteins. The fusion proteins tested were Pro514 NCL, Pro518 MMP9, Pro518 MMP9cl, Pro511 Pro51, and Pro521 AD. The amino acid sequence of Pro518 is shown in Figure 10 AA and SEQ ID NO: 331. [Figure 39A] This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-CA9 (aCA9 h445) conditionally killed human or cyno-CA9 expressing tumor cell lines. A: Human CA9-Raji cells were tested with various fusion proteins. The fusion proteins tested were Pro515 NCL, Pro519 MMP9, Pro519 MMP9cl, and Pro512 Pro51. The amino acid sequence of Pro519 is shown in Figure 10 BB and SEQ ID NO: 332. [Figure 39B]This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-CA9 (aCA9 h445) conditionally killed human or cyno CA9-expressing tumor cell lines. B:Cyno CA9-Raji cells were tested with various fusion proteins. The fusion proteins tested were Pro515 NCL, Pro519 MMP9, Pro519 MMP9cl, and Pro512 Pro51. The amino acid sequence of Pro519 is shown in Figure 10 BB and SEQ ID NO: 332. [Figure 39C] This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-CA9 (aCA9 h445) conditionally killed human or cyno-CA9 expressing tumor cell lines. C:HT29 cells were tested with various fusion proteins. The fusion proteins tested were Pro515 NCL, Pro519 MMP9, Pro519 MMP9cl, and Pro512 Pro51. The amino acid sequence of Pro519 is shown in Figure 10 BB and SEQ ID NO: 332. [Figure 40A] This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-CA9 (aCA9 h456) conditionally killed human or cyno-CA9 expressing tumor cell lines in a TDCC assay. A: Human CA9-Raji cells were tested with various fusion proteins. The fusion proteins tested were Pro1095 NCL, Pro516 MMP9, Pro516 MMP9cl, and Pro509 Pro51. The amino acid sequence of Pro516 is shown in Figure 10 Z and SEQ ID NO: 329. [Figure 40B] This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-CA9 (aCA9 h456) conditionally killed human or cyno CA9-expressing tumor cell lines in a TDCC assay. B:Cyno CA9-Raji cells were tested with various fusion proteins. The fusion proteins tested were Pro1095 NCL, Pro516 MMP9, Pro516 MMP9cl, and Pro509 Pro51. The amino acid sequence of Pro516 is shown in Figure 10 Z and SEQ ID NO: 329. [Figure 40C] This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-CA9 (aCA9 h456) conditionally killed human or cyno-CA9 expressing tumor cell lines in a TDCC assay. ¹¹C:HT29 cells were tested with various fusion proteins. The fusion proteins tested were Pro1095 NCL, Pro516 MMP9, Pro516 MMP9cl, and Pro509 Pro51. The amino acid sequence of Pro516 is shown in Figure 10 Z and SEQ ID NO: 329. [Figure 41A] This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-CA9 (aCA9 h476) conditionally killed human or cyno-CA9 expressing tumor cell lines in a TDCC assay. A: Human CA9-Raji cells were tested with various fusion proteins. The fusion proteins tested were Pro513 NCL, Pro517 MMP9, Pro517 MMP9cl, Pro520 AD, and Pro510 Pro51. The amino acid sequence of Pro517 is shown in Figure 10 AA and SEQ ID NO: 330. [Figure 41B] This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-CA9 (aCA9 h476) conditionally killed human or cyno CA9-expressing tumor cell lines in a TDCC assay. B:Cyno CA9-Raji cells were tested with various fusion proteins. The fusion proteins tested were Pro513 NCL, Pro517 MMP9, Pro517 MMP9cl, Pro520 AD, and Pro510 Pro51. The amino acid sequence of Pro517 is shown in Figure 10 AA and SEQ ID NO: 330. [Figure 41C]This is a series of graphs demonstrating that monospecific COBRA containing two sdABD-CA9 (aCA9 h476) conditionally killed human or cyno-CA9 expressing tumor cell lines in a TDCC assay. C:HT29 cells were tested with various fusion proteins. The fusion proteins tested were Pro513 NCL, Pro517 MMP9, Pro517 MMP9cl, Pro520 AD, and Pro510 Pro51. The amino acid sequence of Pro517 is shown in Figure 10 AA and SEQ ID NO: 330. [Figure 42] This table shows the affinity of CA9 sdAbs in Pro51 format. Various sdAbs and sdAb-fusion protein combinations were evaluated in humans, cynos, and mice. The sdAbs were h407, h445, h456, h472, and h476, and the combinations were h445 with Pro512, h456 with Pro509, and h476 with Pro510. [Figure 43A] This is a series of graphs demonstrating that CA9 / MMP9 heterozygous COBRA caused regression in established tumor xenografts. The graphs show tumor SNU-16 in the presence of Pro513, a non-cleaving control, Pro517, and Pro518, all at a dose of 300 ug / kg. The graphs also show tumor 786-O in the presence of Pro513 and Pro517, all at a dose of 100 ug / kg. [Figure 43B] This is a series of graphs demonstrating that CA9 / MMP9 heterozygous COBRA caused regression in established tumor xenografts. The graphs show tumor SNU-16 in the presence of Pro513, a non-cleaving control, Pro517, and Pro518, all at a dose of 300 ug / kg. The graphs also show tumor 786-O in the presence of Pro513 and Pro517, all at a dose of 100 ug / kg. [Figure 44] This graph shows CA9 / MMP9 hetero-COBRA. The PK of Pro516 is consistent with its binding to the mouse CA9 protein. Pro517 and Pro516 were used at a dose of 100 ug / kg. [Figure 45A]This is a series of graphs demonstrating that EGFR / EpCAM hetero-COBRAs induced TDCC in Raji cells expressing one or both antigens. Raji parental cells (A), Raji-EGFR cells (B), Raji-EpCAM cells (C), and Raji-EGFR / EpCAM cells (D) were tested with monospecific COBRAs (Pro233 EGFR / EGFR monospecific COBRA) and Pro566 (EpCAM / EpCAM monospecific COBRA), and hetero-COBRAs (Pro624 EGFR / EpCAM hetero-COBRA) and Pro698 (EpCAM / EGFR hetero-COBRA). All COBRAs were pre-cleaved. [Figure 45B] This is a series of graphs demonstrating that EGFR / EpCAM hetero-COBRAs induced TDCC in Raji cells expressing one or both antigens. Raji parental cells (A), Raji-EGFR cells (B), Raji-EpCAM cells (C), and Raji-EGFR / EpCAM cells (D) were tested with monospecific COBRAs (Pro233 EGFR / EGFR monospecific COBRA) and Pro566 (EpCAM / EpCAM monospecific COBRA), and hetero-COBRAs (Pro624 EGFR / EpCAM hetero-COBRA) and Pro698 (EpCAM / EGFR hetero-COBRA). All COBRAs were pre-cleaved. [Figure 45C] This is a series of graphs demonstrating that EGFR / EpCAM hetero-COBRAs induced TDCC in Raji cells expressing one or both antigens. Raji parental cells (A), Raji-EGFR cells (B), Raji-EpCAM cells (C), and Raji-EGFR / EpCAM cells (D) were tested with monospecific COBRAs (Pro233 EGFR / EGFR monospecific COBRA) and Pro566 (EpCAM / EpCAM monospecific COBRA), and hetero-COBRAs (Pro624 EGFR / EpCAM hetero-COBRA) and Pro698 (EpCAM / EGFR hetero-COBRA). All COBRAs were pre-cleaved. [Figure 45D]This is a series of graphs demonstrating that EGFR / EpCAM hetero-COBRAs induced TDCC in Raji cells expressing one or both antigens. Raji parental cells (A), Raji-EGFR cells (B), Raji-EpCAM cells (C), and Raji-EGFR / EpCAM cells (D) were tested with monospecific COBRAs (Pro233 EGFR / EGFR monospecific COBRA) and Pro566 (EpCAM / EpCAM monospecific COBRA), and hetero-COBRAs (Pro624 EGFR / EpCAM hetero-COBRA) and Pro698 (EpCAM / EGFR hetero-COBRA). All COBRAs were pre-cleaved. [Figure 46A] This is a series of graphs demonstrating that EGFR / EpCAM heterocobras containing EGFR sdABD hD12 and EpCAM sdABD h665 induced TDCC in HT29 cells expressing both antigens. A: EGFR / EpCAM heterocobras were tested with Pro623 MMP9, cleaved Pro623, Pro625 NCL, and buffer as a control. B: EGFR / EpCAM heterocobras (EpCAM sdABD h665 / EGFR sdABD hD12) were tested with Pro698 MMP9, Pro698 MMP9cl, Pro699 NCL, and buffer as a control. C: EGFR / EpCAM heterocobras (hD12 / h665) were tested with Pro624 MMP9, Pro624 MMP9cl, Pro626 NCL, and buffer as a control. The amino acid sequence of Pro624 is shown in Figure 10 (W) and SEQ ID NO: 323. The amino acid sequence of Pro623 is shown in Figure 10 (X) and SEQ ID NO: 322. The amino acid sequence of Pro698 is shown in Figure 10 (X) and SEQ ID NO: 324. [Figure 46B]This is a series of graphs demonstrating that EGFR / EpCAM heterocobras containing EGFR sdABD hD12 and EpCAM sdABD h665 induced TDCC in HT29 cells expressing both antigens. A: EGFR / EpCAM heterocobras were tested with Pro623 MMP9, cleaved Pro623, Pro625 NCL, and buffer as a control. B: EGFR / EpCAM heterocobras (EpCAM sdABD h665 / EGFR sdABD hD12) were tested with Pro698 MMP9, Pro698 MMP9cl, Pro699 NCL, and buffer as a control. C: EGFR / EpCAM heterocobras (hD12 / h665) were tested with Pro624 MMP9, Pro624 MMP9cl, Pro626 NCL, and buffer as a control. The amino acid sequence of Pro624 is shown in Figure 10 (W) and SEQ ID NO: 323. The amino acid sequence of Pro623 is shown in Figure 10 (X) and SEQ ID NO: 322. The amino acid sequence of Pro698 is shown in Figure 10 (X) and SEQ ID NO: 324. [Figure 46C] This is a series of graphs demonstrating that EGFR / EpCAM heterocobras containing EGFR sdABD hD12 and EpCAM sdABD h665 induced TDCC in HT29 cells expressing both antigens. A: EGFR / EpCAM heterocobras were tested with Pro623 MMP9, cleaved Pro623, Pro625 NCL, and buffer as a control. B: EGFR / EpCAM heterocobras (EpCAM sdABD h665 / EGFR sdABD hD12) were tested with Pro698 MMP9, Pro698 MMP9cl, Pro699 NCL, and buffer as a control. C: EGFR / EpCAM heterocobras (hD12 / h665) were tested with Pro624 MMP9, Pro624 MMP9cl, Pro626 NCL, and buffer as a control. The amino acid sequence of Pro624 is shown in Figure 10 (W) and SEQ ID NO: 323. The amino acid sequence of Pro623 is shown in Figure 10 (X) and SEQ ID NO: 322. The amino acid sequence of Pro698 is shown in Figure 10 (X) and SEQ ID NO: 324. [Figure 47A] This is a series of graphs demonstrating that heterozygous EGFR / FOLR1 COBRAs induced TDCC in Raji cells expressing one or both antigens. Raji-EGFR cells (A), Raji-FOLR1 cells (B), and Raji-EGFR / FOLR1 cells (C) were tested with monospecific COBRAs: Pro233 (EGFR / EGFR) and Pro311 (FOLR1 / FOLR1), and heterozygous COBRAs: Pro421 (EGFR / FOLR1) and Pro420 (FOLR1 / EGFR). All COBRAs were pre-cleaved. The amino acid sequence of Pro420 is shown in Figure 9G and SEQ ID NO: 421. The amino acid sequence of Pro421 is shown in Figure 9G and SEQ ID NO: 422. The amino acid sequence of Pro233 is shown in Figure 9D and SEQ ID NO: 415. The amino acid sequence of Pro311 is shown in Figure 9D and SEQ ID NO: 416. [Figure 47B] This is a series of graphs demonstrating that heterozygous EGFR / FOLR1 COBRAs induced TDCC in Raji cells expressing one or both antigens. Raji-EGFR cells (A), Raji-FOLR1 cells (B), and Raji-EGFR / FOLR1 cells (C) were tested with monospecific COBRAs: Pro233 (EGFR / EGFR) and Pro311 (FOLR1 / FOLR1), and heterozygous COBRAs: Pro421 (EGFR / FOLR1) and Pro420 (FOLR1 / EGFR). All COBRAs were pre-cleaved. The amino acid sequence of Pro420 is shown in Figure 9G and SEQ ID NO: 421. The amino acid sequence of Pro421 is shown in Figure 9G and SEQ ID NO: 422. The amino acid sequence of Pro233 is shown in Figure 9D and SEQ ID NO: 415. The amino acid sequence of Pro311 is shown in Figure 9D and SEQ ID NO: 416. [Figure 47C]This is a series of graphs demonstrating that heterozygous EGFR / FOLR1 COBRAs induced TDCC in Raji cells expressing one or both antigens. Raji-EGFR cells (A), Raji-FOLR1 cells (B), and Raji-EGFR / FOLR1 cells (C) were tested with monospecific COBRAs: Pro233 (EGFR / EGFR) and Pro311 (FOLR1 / FOLR1), and heterozygous COBRAs: Pro421 (EGFR / FOLR1) and Pro420 (FOLR1 / EGFR). All COBRAs were pre-cleaved. The amino acid sequence of Pro420 is shown in Figure 9G and SEQ ID NO: 421. The amino acid sequence of Pro421 is shown in Figure 9G and SEQ ID NO: 422. The amino acid sequence of Pro233 is shown in Figure 9D and SEQ ID NO: 415. The amino acid sequence of Pro311 is shown in Figure 9D and SEQ ID NO: 416. [Figure 48A] This is a series of graphs demonstrating that aFOLR1 / aEGFR heterozygous COBRA containing EGFR D12 and FOLR1 h59-3 conditionally killed tumor cell lines expressing both FOLR1 and EGFR. A: H292 cells tested with monospecific COBRA: Pro214 NCL (EGFR D12), Pro186 MMP9 (EGFR D12), and Pro186 MMP9c1 (EGFR D12). B: H292 cells tested with monospecific COBRA: Pro303 NCL (FOLR1 h59-3), Pro312 MMP9 (FOLR1 h59-3), and Pro312 MMP9cl (FOLR1 h59-3). C:H292 cells were tested with hetero-COBRA:Pro550 NCL (EGFR D12 / FOLR1 h59-3), Pro551 MMP9 (EGFR D12 / FOLR1 h59-3), and Pro551 (MMP9cl EGFR D12 / FOLR1 h59-3). The amino acid sequence of Pro551 is shown in Figure 10V and SEQ ID NO: 320. [Figure 48B]This is a series of graphs demonstrating that aFOLR1 / aEGFR heterozygous COBRA containing EGFR D12 and FOLR1 h59-3 conditionally killed tumor cell lines expressing both FOLR1 and EGFR. A: H292 cells tested with monospecific COBRA: Pro214 NCL (EGFR D12), Pro186 MMP9 (EGFR D12), and Pro186 MMP9c1 (EGFR D12). B: H292 cells tested with monospecific COBRA: Pro303 NCL (FOLR1 h59-3), Pro312 MMP9 (FOLR1 h59-3), and Pro312 MMP9cl (FOLR1 h59-3). C:H292 cells were tested with hetero-COBRA:Pro550 NCL (EGFR D12 / FOLR1 h59-3), Pro551 MMP9 (EGFR D12 / FOLR1 h59-3), and Pro551 (MMP9cl EGFR D12 / FOLR1 h59-3). The amino acid sequence of Pro551 is shown in Figure 10V and SEQ ID NO: 320. [Figure 48C] This is a series of graphs demonstrating that aFOLR1 / aEGFR heterozygous COBRA containing EGFR D12 and FOLR1 h59-3 conditionally killed tumor cell lines expressing both FOLR1 and EGFR. A: H292 cells tested with monospecific COBRA: Pro214 NCL (EGFR D12), Pro186 MMP9 (EGFR D12), and Pro186 MMP9c1 (EGFR D12). B: H292 cells tested with monospecific COBRA: Pro303 NCL (FOLR1 h59-3), Pro312 MMP9 (FOLR1 h59-3), and Pro312 MMP9cl (FOLR1 h59-3). C:H292 cells were tested with hetero-COBRA:Pro550 NCL (EGFR D12 / FOLR1 h59-3), Pro551 MMP9 (EGFR D12 / FOLR1 h59-3), and Pro551 (MMP9cl EGFR D12 / FOLR1 h59-3). The amino acid sequence of Pro551 is shown in Figure 10V and SEQ ID NO: 320. [Figure 49A]This is a series of graphs demonstrating that aFOLR1(h77.2) / aEGFR(hD12) conditionally killed tumor cell lines expressing both FOLR1 and EGFR. A: H292 cells were tested with monospecific COBRA:Pro600 NCL(EGFR / EGFR), Pro233 MMP9 EGFR / EGFR, and Pro233 MMP9cl(EGFR / EGFR). B: H292 cells were tested with monospecific COBRA:Pro299 NCL FOLR1 / FOLR1, Pro311 MMP9(FOLR1 / FOLR1), and Pro311 MMP9cl(FOLR1 / FOLR1). C: H292 cells were tested with heterozygous COBRA:Pro420 MMP9(FOLR1 / EGFR) and Pro420 MMP9cl(FOLR1 / EGFR). D:H292 cells were tested with hetero-COBRA:Pro421 MMP9(EGFR / FOLR1) and Pro421 MMP9cl(EGFR / FOLR1). The amino acid sequence of Pro420 is shown in Figure 9G and SEQ ID NO: 421. The amino acid sequence of Pro421 is shown in Figure 9G and SEQ ID NO: 422. [Figure 49B]This is a series of graphs demonstrating that aFOLR1(h77.2) / aEGFR(hD12) conditionally killed tumor cell lines expressing both FOLR1 and EGFR. A: H292 cells were tested with monospecific COBRA:Pro600 NCL(EGFR / EGFR), Pro233 MMP9 EGFR / EGFR, and Pro233 MMP9cl(EGFR / EGFR). B: H292 cells were tested with monospecific COBRA:Pro299 NCL FOLR1 / FOLR1, Pro311 MMP9(FOLR1 / FOLR1), and Pro311 MMP9cl(FOLR1 / FOLR1). C: H292 cells were tested with heterozygous COBRA:Pro420 MMP9(FOLR1 / EGFR) and Pro420 MMP9cl(FOLR1 / EGFR). D:H292 cells were tested with hetero-COBRA:Pro421 MMP9(EGFR / FOLR1) and Pro421 MMP9cl(EGFR / FOLR1). The amino acid sequence of Pro420 is shown in Figure 9G and SEQ ID NO: 421. The amino acid sequence of Pro421 is shown in Figure 9G and SEQ ID NO: 422. [Figure 49C]This is a series of graphs demonstrating that aFOLR1(h77.2) / aEGFR(hD12) conditionally killed tumor cell lines expressing both FOLR1 and EGFR. A: H292 cells were tested with monospecific COBRA:Pro600 NCL(EGFR / EGFR), Pro233 MMP9 EGFR / EGFR, and Pro233 MMP9cl(EGFR / EGFR). B: H292 cells were tested with monospecific COBRA:Pro299 NCL FOLR1 / FOLR1, Pro311 MMP9(FOLR1 / FOLR1), and Pro311 MMP9cl(FOLR1 / FOLR1). C: H292 cells were tested with heterozygous COBRA:Pro420 MMP9(FOLR1 / EGFR) and Pro420 MMP9cl(FOLR1 / EGFR). D:H292 cells were tested with hetero-COBRA:Pro421 MMP9(EGFR / FOLR1) and Pro421 MMP9cl(EGFR / FOLR1). The amino acid sequence of Pro420 is shown in Figure 9G and SEQ ID NO: 421. The amino acid sequence of Pro421 is shown in Figure 9G and SEQ ID NO: 422. [Figure 49D]This is a series of graphs demonstrating that aFOLR1(h77.2) / aEGFR(hD12) conditionally killed tumor cell lines expressing both FOLR1 and EGFR. A: H292 cells were tested with monospecific COBRA:Pro600 NCL(EGFR / EGFR), Pro233 MMP9 EGFR / EGFR, and Pro233 MMP9cl(EGFR / EGFR). B: H292 cells were tested with monospecific COBRA:Pro299 NCL FOLR1 / FOLR1, Pro311 MMP9(FOLR1 / FOLR1), and Pro311 MMP9cl(FOLR1 / FOLR1). C: H292 cells were tested with heterozygous COBRA:Pro420 MMP9(FOLR1 / EGFR) and Pro420 MMP9cl(FOLR1 / EGFR). D:H292 cells were tested with hetero-COBRA:Pro421 MMP9(EGFR / FOLR1) and Pro421 MMP9cl(EGFR / FOLR1). The amino acid sequence of Pro420 is shown in Figure 9G and SEQ ID NO: 421. The amino acid sequence of Pro421 is shown in Figure 9G and SEQ ID NO: 422. [Figure 50] This table lists the affinity between EGFR / FOLR1 heterocobra molecules and Pro51-type molecules. [Figure 51A] This is a series of graphs demonstrating that Pro566 aEpCAM(h664) conditionally killed EpCAM-expressing EpCAM Raji transfectants and tumor cell lines. Trop2-Raji cells (A), EpCAM-Raji cells (B), SKOV3 cells (C), and HT29 cells (D) were all tested with Pro566 and cleaved Pro566 (Pro566cl). [Figure 51B] This is a series of graphs demonstrating that Pro566 aEpCAM(h664) conditionally killed EpCAM-expressing EpCAM Raji transfectants and tumor cell lines. Trop2-Raji cells (A), EpCAM-Raji cells (B), SKOV3 cells (C), and HT29 cells (D) were all tested with Pro566 and cleaved Pro566 (Pro566cl). [Figure 51C]This is a series of graphs demonstrating that Pro566 aEpCAM(h664) conditionally killed EpCAM-expressing EpCAM Raji transfectants and tumor cell lines. Trop2-Raji cells (A), EpCAM-Raji cells (B), SKOV3 cells (C), and HT29 cells (D) were all tested with Pro566 and cleaved Pro566 (Pro566cl). [Figure 51D] This is a series of graphs demonstrating that Pro566 aEpCAM(h664) conditionally killed EpCAM-expressing EpCAM Raji transfectants and tumor cell lines. Trop2-Raji cells (A), EpCAM-Raji cells (B), SKOV3 cells (C), and HT29 cells (D) were all tested with Pro566 and cleaved Pro566 (Pro566cl). [Figure 52A] This is a series of graphs demonstrating that Pro677 aTrop2(h557) conditionally kills Trop2-expressing Trop2 Raji transfectants and tumor cell lines. Trop2-Raji cells (A), EpCAM-Raji cells (B), SKOV3 cells (C), and HT9 cells (D) were all tested with Pro677 and cleaved Pro677 (Pro677cl.). [Figure 52B] This is a series of graphs demonstrating that Pro677 aTrop2(h557) conditionally kills Trop2-expressing Trop2 Raji transfectants and tumor cell lines. Trop2-Raji cells (A), EpCAM-Raji cells (B), SKOV3 cells (C), and HT9 cells (D) were all tested with Pro677 and cleaved Pro677 (Pro677cl.). [Figure 52C] This is a series of graphs demonstrating that Pro677 aTrop2(h557) conditionally kills Trop2-expressing Trop2 Raji transfectants and tumor cell lines. Trop2-Raji cells (A), EpCAM-Raji cells (B), SKOV3 cells (C), and HT9 cells (D) were all tested with Pro677 and cleaved Pro677 (Pro677cl.). [Figure 52D] This is a series of graphs demonstrating that Pro677 aTrop2(h557) conditionally kills Trop2-expressing Trop2 Raji transfectants and tumor cell lines. Trop2-Raji cells (A), EpCAM-Raji cells (B), SKOV3 cells (C), and HT9 cells (D) were all tested with Pro677 and cleaved Pro677 (Pro677cl.). [Figure 53A] This is a series of graphs demonstrating that Pro824 aEpCAM(h664) / aTROP2(h557) conditionally kills Raji transfectants and tumor cell lines that express both TROP2 and EpCAM. Trop2-Raji cells (A), EpCAM-Raji cells (B), SKOV3 cells (C), and HT29 cells (D) were all tested with Pro824 and cleaved Pro824 (Pro824cl.). [Figure 53B] This is a series of graphs demonstrating that Pro824 aEpCAM(h664) / aTROP2(h557) conditionally kills Raji transfectants and tumor cell lines that express both TROP2 and EpCAM. Trop2-Raji cells (A), EpCAM-Raji cells (B), SKOV3 cells (C), and HT29 cells (D) were all tested with Pro824 and cleaved Pro824 (Pro824cl.). [Figure 53C] This is a series of graphs demonstrating that Pro824 aEpCAM(h664) / aTROP2(h557) conditionally kills Raji transfectants and tumor cell lines that express both TROP2 and EpCAM. Trop2-Raji cells (A), EpCAM-Raji cells (B), SKOV3 cells (C), and HT29 cells (D) were all tested with Pro824 and cleaved Pro824 (Pro824cl.). [Figure 53D]This is a series of graphs demonstrating that Pro824 aEpCAM(h664) / aTROP2(h557) conditionally kills Raji transfectants and tumor cell lines that express both TROP2 and EpCAM. Trop2-Raji cells (A), EpCAM-Raji cells (B), SKOV3 cells (C), and HT29 cells (D) were all tested with Pro824 and cleaved Pro824 (Pro824cl.). [Figure 54A] This is a series of graphs demonstrating that Pro826 aTROP2(h557) / aEpCAM(h664) conditionally kills Raji transfectants and tumor cell lines that express both TROP2 and EpCAM. Trop2-Raji cells (A), EpCAM-Raji cells (B), SKOV3 cells (C), and HT29 cells (D) were all tested with Pro826 and cleaved Pro826 (Pro826cl). [Figure 54B] This is a series of graphs demonstrating that Pro826 aTROP2(h557) / aEpCAM(h664) conditionally kills Raji transfectants and tumor cell lines that express both TROP2 and EpCAM. Trop2-Raji cells (A), EpCAM-Raji cells (B), SKOV3 cells (C), and HT29 cells (D) were all tested with Pro826 and cleaved Pro826 (Pro826cl). [Figure 54C] This is a series of graphs demonstrating that Pro826 aTROP2(h557) / aEpCAM(h664) conditionally kills Raji transfectants and tumor cell lines that express both TROP2 and EpCAM. Trop2-Raji cells (A), EpCAM-Raji cells (B), SKOV3 cells (C), and HT29 cells (D) were all tested with Pro826 and cleaved Pro826 (Pro826cl). [Figure 54D]This is a series of graphs demonstrating that Pro826 aTROP2(h557) / aEpCAM(h664) conditionally kills Raji transfectants and tumor cell lines that express both TROP2 and EpCAM. Trop2-Raji cells (A), EpCAM-Raji cells (B), SKOV3 cells (C), and HT29 cells (D) were all tested with Pro826 and cleaved Pro826 (Pro826cl). [Figure 55A] This is a series of graphs showing that EpCAM, Trop2 COBRA, and hetero-COBRA all function well in BXPC3. A: BXPC3 cells tested with Pro569 NCL, Pro566 MMP9, and Pro566 MMP9cl. B: BXPC3 cells tested with Pro681 NCL, Pro677 MMP9, and Pro677 MMP9cl. C: BXPC3 cells tested with Pro825 NCL, Pro824 MMP9, and Pro824 MMP9cl. D: BXPC3 cells tested with Pro827 NCL, Pro826 MMP9, and Pro826 MMP9cl. [Figure 55B] This is a series of graphs showing that EpCAM, Trop2 COBRA, and hetero-COBRA all function well in BXPC3. A: BXPC3 cells tested with Pro569 NCL, Pro566 MMP9, and Pro566 MMP9cl. B: BXPC3 cells tested with Pro681 NCL, Pro677 MMP9, and Pro677 MMP9cl. C: BXPC3 cells tested with Pro825 NCL, Pro824 MMP9, and Pro824 MMP9cl. D: BXPC3 cells tested with Pro827 NCL, Pro826 MMP9, and Pro826 MMP9cl. [Figure 55C]This is a series of graphs showing that EpCAM, Trop2 COBRA, and hetero-COBRA all function well in BXPC3. A: BXPC3 cells tested with Pro569 NCL, Pro566 MMP9, and Pro566 MMP9cl. B: BXPC3 cells tested with Pro681 NCL, Pro677 MMP9, and Pro677 MMP9cl. C: BXPC3 cells tested with Pro825 NCL, Pro824 MMP9, and Pro824 MMP9cl. D: BXPC3 cells tested with Pro827 NCL, Pro826 MMP9, and Pro826 MMP9cl. [Figure 55D] This is a series of graphs showing that EpCAM, Trop2 COBRA, and hetero-COBRA all function well in BXPC3. A: BXPC3 cells tested with Pro569 NCL, Pro566 MMP9, and Pro566 MMP9cl. B: BXPC3 cells tested with Pro681 NCL, Pro677 MMP9, and Pro677 MMP9cl. C: BXPC3 cells tested with Pro825 NCL, Pro824 MMP9, and Pro824 MMP9cl. D: BXPC3 cells tested with Pro827 NCL, Pro826 MMP9, and Pro826 MMP9cl. [Figure 56A] This is a series of graphs showing that EpCAM, Trop2 COBRA, and hetero-COBRA all function well with HCT116. A: HCT116 cells (human colon cancer cell line) tested with Pro569 NCL, Pro566 MMP9, and Pro566 MMP9cl. B: HCT116 cells tested with Pro681 NCL, Pro677 MMP9, and Pro677 MMP9cl. C: HCT116 cells tested with Pro825 NCL, Pro824 MMP9, and Pro824 MMP9cl. D: HCT116 cells tested with Pro827 NCL, Pro826 MMP9, and Pro846 MMP9cl. [Figure 56B]This is a series of graphs showing that EpCAM, Trop2 COBRA, and hetero-COBRA all function well with HCT116. A: HCT116 cells (human colon cancer cell line) tested with Pro569 NCL, Pro566 MMP9, and Pro566 MMP9cl. B: HCT116 cells tested with Pro681 NCL, Pro677 MMP9, and Pro677 MMP9cl. C: HCT116 cells tested with Pro825 NCL, Pro824 MMP9, and Pro824 MMP9cl. D: HCT116 cells tested with Pro827 NCL, Pro826 MMP9, and Pro846 MMP9cl. [Figure 56C] This is a series of graphs showing that EpCAM, Trop2 COBRA, and hetero-COBRA all function well with HCT116. A: HCT116 cells (human colon cancer cell line) tested with Pro569 NCL, Pro566 MMP9, and Pro566 MMP9cl. B: HCT116 cells tested with Pro681 NCL, Pro677 MMP9, and Pro677 MMP9cl. C: HCT116 cells tested with Pro825 NCL, Pro824 MMP9, and Pro824 MMP9cl. D: HCT116 cells tested with Pro827 NCL, Pro826 MMP9, and Pro846 MMP9cl. [Figure 56D] This is a series of graphs showing that EpCAM, Trop2 COBRA, and hetero-COBRA all function well with HCT116. A: HCT116 cells (human colon cancer cell line) tested with Pro569 NCL, Pro566 MMP9, and Pro566 MMP9cl. B: HCT116 cells tested with Pro681 NCL, Pro677 MMP9, and Pro677 MMP9cl. C: HCT116 cells tested with Pro825 NCL, Pro824 MMP9, and Pro824 MMP9cl. D: HCT116 cells tested with Pro827 NCL, Pro826 MMP9, and Pro846 MMP9cl. [Figure 57A]This is a series of graphs showing that EpCAM, Trop2 COBRA, and hetero-COBRA all function well in SCC25. A: SCC25 cells were tested with Pro569 NCL, Pro566 MMP9, and Pro566 MMP9cl. The amino acid sequence of Pro566 is shown in Figure 10 F and SEQ ID NO: 289. B: SCC25 cells were tested with Pro681 NCL, Pro677 MMP9, and Pro677 MMP9cl. The amino acid sequence of Pro677 is shown in Figure 10 K and SEQ ID NO: 298. C: SCC25 cells were tested with Pro825 NCL, Pro824 MMP9, and Pro824 MMP9cl. The amino acid sequence of Pro824 is shown in Figure 12 Q and SEQ ID NO: 485. D: SCC25 cells were tested with Pro827 NCL, Pro826 MMP9, and Pro826 MMP9cl. The amino acid sequence of Pro826 is shown in Figure 12 Q and SEQ ID NO: 486. [Figure 57B] This is a series of graphs showing that EpCAM, Trop2 COBRA, and hetero-COBRA all function well in SCC25. A: SCC25 cells were tested with Pro569 NCL, Pro566 MMP9, and Pro566 MMP9cl. The amino acid sequence of Pro566 is shown in Figure 10 F and SEQ ID NO: 289. B: SCC25 cells were tested with Pro681 NCL, Pro677 MMP9, and Pro677 MMP9cl. The amino acid sequence of Pro677 is shown in Figure 10 K and SEQ ID NO: 298. C: SCC25 cells were tested with Pro825 NCL, Pro824 MMP9, and Pro824 MMP9cl. The amino acid sequence of Pro824 is shown in Figure 12 Q and SEQ ID NO: 485. D: SCC25 cells were tested with Pro827 NCL, Pro826 MMP9, and Pro826 MMP9cl. The amino acid sequence of Pro826 is shown in Figure 12 Q and SEQ ID NO: 486. [Figure 57C]This is a series of graphs showing that EpCAM, Trop2 COBRA, and hetero-COBRA all function well in SCC25. A: SCC25 cells were tested with Pro569 NCL, Pro566 MMP9, and Pro566 MMP9cl. The amino acid sequence of Pro566 is shown in Figure 10 F and SEQ ID NO: 289. B: SCC25 cells were tested with Pro681 NCL, Pro677 MMP9, and Pro677 MMP9cl. The amino acid sequence of Pro677 is shown in Figure 10 K and SEQ ID NO: 298. C: SCC25 cells were tested with Pro825 NCL, Pro824 MMP9, and Pro824 MMP9cl. The amino acid sequence of Pro824 is shown in Figure 12 Q and SEQ ID NO: 485. D: SCC25 cells were tested with Pro827 NCL, Pro826 MMP9, and Pro826 MMP9cl. The amino acid sequence of Pro826 is shown in Figure 12 Q and SEQ ID NO: 486. [Figure 57D] This is a series of graphs showing that EpCAM, Trop2 COBRA, and hetero-COBRA all function well in SCC25. A: SCC25 cells were tested with Pro569 NCL, Pro566 MMP9, and Pro566 MMP9cl. The amino acid sequence of Pro566 is shown in Figure 10 F and SEQ ID NO: 289. B: SCC25 cells were tested with Pro681 NCL, Pro677 MMP9, and Pro677 MMP9cl. The amino acid sequence of Pro677 is shown in Figure 10 K and SEQ ID NO: 298. C: SCC25 cells were tested with Pro825 NCL, Pro824 MMP9, and Pro824 MMP9cl. The amino acid sequence of Pro824 is shown in Figure 12 Q and SEQ ID NO: 485. D: SCC25 cells were tested with Pro827 NCL, Pro826 MMP9, and Pro826 MMP9cl. The amino acid sequence of Pro826 is shown in Figure 12 Q and SEQ ID NO: 486. [Figure 58A]This is a series of graphs demonstrating that B7H3 / EpCAM hetero-COBRAs induced TDCC in cells expressing one or both antigens. Raji parental cells (A), Raji-B7H3 cells (B), Raji-EpCAM cells (C), and Raji-B7H3 / EpCAM cells (D) were tested with monospecific COBRAs: Pro225 (B7H3 / B7H3) and Pro566 (EpCAM / EpCAM), and hetero-COBRAs: Pro656 (B7H3 / EpCAM) and Pro658 (EpCAM / B7H3). All COBRAs were pre-cleaved. The amino acid sequence of Pro225 is shown in Figure 10 DD and SEQ ID NO: 336. The amino acid sequence of Pro566 is shown in Figure 10 F and SEQ ID NO: 289. The amino acid sequence of Pro656 is shown in Figure 10 Y and SEQ ID NO: 326. The amino acid sequence of Pro516 is shown in Figure 10 Z and SEQ ID NO: 329. [Figure 58B] This is a series of graphs demonstrating that B7H3 / EpCAM hetero-COBRAs induced TDCC in cells expressing one or both antigens. Raji parental cells (A), Raji-B7H3 cells (B), Raji-EpCAM cells (C), and Raji-B7H3 / EpCAM cells (D) were tested with monospecific COBRAs: Pro225 (B7H3 / B7H3) and Pro566 (EpCAM / EpCAM), and hetero-COBRAs: Pro656 (B7H3 / EpCAM) and Pro658 (EpCAM / B7H3). All COBRAs were pre-cleaved. The amino acid sequence of Pro225 is shown in Figure 10 DD and SEQ ID NO: 336. The amino acid sequence of Pro566 is shown in Figure 10 F and SEQ ID NO: 289. The amino acid sequence of Pro656 is shown in Figure 10 Y and SEQ ID NO: 326. The amino acid sequence of Pro516 is shown in Figure 10 Z and SEQ ID NO: 329. [Figure 58C]This is a series of graphs demonstrating that B7H3 / EpCAM hetero-COBRAs induced TDCC in cells expressing one or both antigens. Raji parental cells (A), Raji-B7H3 cells (B), Raji-EpCAM cells (C), and Raji-B7H3 / EpCAM cells (D) were tested with monospecific COBRAs: Pro225 (B7H3 / B7H3) and Pro566 (EpCAM / EpCAM), and hetero-COBRAs: Pro656 (B7H3 / EpCAM) and Pro658 (EpCAM / B7H3). All COBRAs were pre-cleaved. The amino acid sequence of Pro225 is shown in Figure 10 DD and SEQ ID NO: 336. The amino acid sequence of Pro566 is shown in Figure 10 F and SEQ ID NO: 289. The amino acid sequence of Pro656 is shown in Figure 10 Y and SEQ ID NO: 326. The amino acid sequence of Pro516 is shown in Figure 10 Z and SEQ ID NO: 329. [Figure 58D] This is a series of graphs demonstrating that B7H3 / EpCAM hetero-COBRAs induced TDCC in cells expressing one or both antigens. Raji parental cells (A), Raji-B7H3 cells (B), Raji-EpCAM cells (C), and Raji-B7H3 / EpCAM cells (D) were tested with monospecific COBRAs: Pro225 (B7H3 / B7H3) and Pro566 (EpCAM / EpCAM), and hetero-COBRAs: Pro656 (B7H3 / EpCAM) and Pro658 (EpCAM / B7H3). All COBRAs were pre-cleaved. The amino acid sequence of Pro225 is shown in Figure 10 DD and SEQ ID NO: 336. The amino acid sequence of Pro566 is shown in Figure 10 F and SEQ ID NO: 289. The amino acid sequence of Pro656 is shown in Figure 10 Y and SEQ ID NO: 326. The amino acid sequence of Pro516 is shown in Figure 10 Z and SEQ ID NO: 329. [Figure 59A]This is a series of graphs showing the results for CRISPR knockout cells. HT29 cells (A), HT29-B7H3 KO cells (B), HT29-EpCAM KO cells (C), and HT29-B7H3 / EpCAM KO cells (D) were all tested with monospecific COBRA: Pro225 (B7H3 / B7H3) and Pro566 (EpCAM / EpCAM), and hetero-COBRA: Pro656 (B7H3 / EpCAM). All COBRAs were pre-cleaved. The amino acid sequence of Pro225 is shown in Figure 10 DD and SEQ ID NO: 336. The amino acid sequence of Pro566 is shown in Figure 10 F and SEQ ID NO: 289. The amino acid sequence of Pro656 is shown in Figure 10 Y and SEQ ID NO: 326. [Figure 59B] This is a series of graphs showing the results for CRISPR knockout cells. HT29 cells (A), HT29-B7H3 KO cells (B), HT29-EpCAM KO cells (C), and HT29-B7H3 / EpCAM KO cells (D) were all tested with monospecific COBRA: Pro225 (B7H3 / B7H3) and Pro566 (EpCAM / EpCAM), and hetero-COBRA: Pro656 (B7H3 / EpCAM). All COBRAs were pre-cleaved. The amino acid sequence of Pro225 is shown in Figure 10 DD and SEQ ID NO: 336. The amino acid sequence of Pro566 is shown in Figure 10 F and SEQ ID NO: 289. The amino acid sequence of Pro656 is shown in Figure 10 Y and SEQ ID NO: 326. [Figure 59C] This is a series of graphs showing the results for CRISPR knockout cells. HT29 cells (A), HT29-B7H3 KO cells (B), HT29-EpCAM KO cells (C), and HT29-B7H3 / EpCAM KO cells (D) were all tested with monospecific COBRA: Pro225 (B7H3 / B7H3) and Pro566 (EpCAM / EpCAM), and hetero-COBRA: Pro656 (B7H3 / EpCAM). All COBRAs were pre-cleaved. The amino acid sequence of Pro225 is shown in Figure 10 DD and SEQ ID NO: 336. The amino acid sequence of Pro566 is shown in Figure 10 F and SEQ ID NO: 289. The amino acid sequence of Pro656 is shown in Figure 10 Y and SEQ ID NO: 326. [Figure 59D]This is a series of graphs showing the results for CRISPR knockout cells. HT29 cells (A), HT29-B7H3 KO cells (B), HT29-EpCAM KO cells (C), and HT29-B7H3 / EpCAM KO cells (D) were all tested with monospecific COBRA: Pro225 (B7H3 / B7H3) and Pro566 (EpCAM / EpCAM), and hetero-COBRA: Pro656 (B7H3 / EpCAM). All COBRAs were pre-cleaved. The amino acid sequence of Pro225 is shown in Figure 10 DD and SEQ ID NO: 336. The amino acid sequence of Pro566 is shown in Figure 10 F and SEQ ID NO: 289. The amino acid sequence of Pro656 is shown in Figure 10 Y and SEQ ID NO: 326. [Figure 60A] This is a series of graphs demonstrating that EpCAM / B7H3 hetero-COBRA, containing aEpCAM sdABD(h664) and aB7H3 sdABD(hF7), conditionally killed tumor cell lines expressing both EpCAM and B7H3. A: IGROV cells tested with Pro295 NCL(B7H3), Pro225 MMP9(B7H3), and Pro225 MMP9cl(B7H3). B: IGROV cells tested with Pro568 NCL(EpCAM), Pro565 MMP9(EpCAM), and Pro565 MMP9cl(EpCAM). C: IGROV cells tested with Pro659 NCL(B7H3 / EpCAM), Pro655 MMP9(B7H3 / EpCAM), and Pro655 MMP9cl(B7H3 / EpCAM). D: IGROV cells were tested with Pro661 NCL (EpCAM / B7H3), Pro657 MMP9 (EpCAM / B7H3), and Pro657 MMP9cl (EpCAM / B7H3). The amino acid sequence of Pro655 is shown in Figure 10 X and SEQ ID NO: 325. The amino acid sequence of Pro657 is shown in Figure 10 Y and SEQ ID NO: 327. [Figure 60B]This is a series of graphs demonstrating that EpCAM / B7H3 hetero-COBRA, containing aEpCAM sdABD(h664) and aB7H3 sdABD(hF7), conditionally killed tumor cell lines expressing both EpCAM and B7H3. A: IGROV cells tested with Pro295 NCL(B7H3), Pro225 MMP9(B7H3), and Pro225 MMP9cl(B7H3). B: IGROV cells tested with Pro568 NCL(EpCAM), Pro565 MMP9(EpCAM), and Pro565 MMP9cl(EpCAM). C: IGROV cells tested with Pro659 NCL(B7H3 / EpCAM), Pro655 MMP9(B7H3 / EpCAM), and Pro655 MMP9cl(B7H3 / EpCAM). D: IGROV cells were tested with Pro661 NCL (EpCAM / B7H3), Pro657 MMP9 (EpCAM / B7H3), and Pro657 MMP9cl (EpCAM / B7H3). The amino acid sequence of Pro655 is shown in Figure 10 X and SEQ ID NO: 325. The amino acid sequence of Pro657 is shown in Figure 10 Y and SEQ ID NO: 327. [Figure 60C]This is a series of graphs demonstrating that EpCAM / B7H3 hetero-COBRA, containing aEpCAM sdABD(h664) and aB7H3 sdABD(hF7), conditionally killed tumor cell lines expressing both EpCAM and B7H3. A: IGROV cells tested with Pro295 NCL(B7H3), Pro225 MMP9(B7H3), and Pro225 MMP9cl(B7H3). B: IGROV cells tested with Pro568 NCL(EpCAM), Pro565 MMP9(EpCAM), and Pro565 MMP9cl(EpCAM). C: IGROV cells tested with Pro659 NCL(B7H3 / EpCAM), Pro655 MMP9(B7H3 / EpCAM), and Pro655 MMP9cl(B7H3 / EpCAM). D: IGROV cells were tested with Pro661 NCL (EpCAM / B7H3), Pro657 MMP9 (EpCAM / B7H3), and Pro657 MMP9cl (EpCAM / B7H3). The amino acid sequence of Pro655 is shown in Figure 10 X and SEQ ID NO: 325. The amino acid sequence of Pro657 is shown in Figure 10 Y and SEQ ID NO: 327. [Figure 60D]This is a series of graphs demonstrating that EpCAM / B7H3 hetero-COBRA, containing aEpCAM sdABD(h664) and aB7H3 sdABD(hF7), conditionally killed tumor cell lines expressing both EpCAM and B7H3. A: IGROV cells tested with Pro295 NCL(B7H3), Pro225 MMP9(B7H3), and Pro225 MMP9cl(B7H3). B: IGROV cells tested with Pro568 NCL(EpCAM), Pro565 MMP9(EpCAM), and Pro565 MMP9cl(EpCAM). C: IGROV cells tested with Pro659 NCL(B7H3 / EpCAM), Pro655 MMP9(B7H3 / EpCAM), and Pro655 MMP9cl(B7H3 / EpCAM). D: IGROV cells were tested with Pro661 NCL (EpCAM / B7H3), Pro657 MMP9 (EpCAM / B7H3), and Pro657 MMP9cl (EpCAM / B7H3). The amino acid sequence of Pro655 is shown in Figure 10 X and SEQ ID NO: 325. The amino acid sequence of Pro657 is shown in Figure 10 Y and SEQ ID NO: 327. [Figure 61A]This is a series of graphs demonstrating that aEpCAM / aB7H3 hetero-COBRA, containing aEpCAM sdABD(h665) and aB7H3 sdABD(hF7)sdABD, conditionally killed tumor cell lines expressing both EpCAM and B7H3. A: IGROV cells tested with Pro295 NCL(B7H3), Pro225 MMP9(B7H3), and Pro225 MMP9cl(B7H3). B: IGROV cells tested with Pro569 NCL(EpCAM), Pro566 MMP9(EpCAM), and Pro566 MMP9cl(EpCAM). C: IGROV cells tested with Pro660 NCL(B7H3 / EpCAM), Pro656 MMP9(B7H3 / EpCAM), and Pro656 MMP9cl(B7H3 / EpCAM). D: IGROV cells were tested with Pro662 NCL (EpCAM / B7H3), Pro658 MMP9 (EpCAM / B7H3), and Pro658 MMP9cl (EpCAM / B7H3). The amino acid sequence of Pro656 is shown in Figure 10 Y and SEQ ID NO: 326. The amino acid sequence of Pro658 is shown in Figure 10 Z and SEQ ID NO: 328. [Figure 61B]This is a series of graphs demonstrating that aEpCAM / aB7H3 hetero-COBRA, containing aEpCAM sdABD(h665) and aB7H3 sdABD(hF7)sdABD, conditionally killed tumor cell lines expressing both EpCAM and B7H3. A: IGROV cells tested with Pro295 NCL(B7H3), Pro225 MMP9(B7H3), and Pro225 MMP9cl(B7H3). B: IGROV cells tested with Pro569 NCL(EpCAM), Pro566 MMP9(EpCAM), and Pro566 MMP9cl(EpCAM). C: IGROV cells tested with Pro660 NCL(B7H3 / EpCAM), Pro656 MMP9(B7H3 / EpCAM), and Pro656 MMP9cl(B7H3 / EpCAM). D: IGROV cells were tested with Pro662 NCL (EpCAM / B7H3), Pro658 MMP9 (EpCAM / B7H3), and Pro658 MMP9cl (EpCAM / B7H3). The amino acid sequence of Pro656 is shown in Figure 10 Y and SEQ ID NO: 326. The amino acid sequence of Pro658 is shown in Figure 10 Z and SEQ ID NO: 328. [Figure 61C]This is a series of graphs demonstrating that aEpCAM / aB7H3 hetero-COBRA, containing aEpCAM sdABD(h665) and aB7H3 sdABD(hF7)sdABD, conditionally killed tumor cell lines expressing both EpCAM and B7H3. A: IGROV cells tested with Pro295 NCL(B7H3), Pro225 MMP9(B7H3), and Pro225 MMP9cl(B7H3). B: IGROV cells tested with Pro569 NCL(EpCAM), Pro566 MMP9(EpCAM), and Pro566 MMP9cl(EpCAM). C: IGROV cells tested with Pro660 NCL(B7H3 / EpCAM), Pro656 MMP9(B7H3 / EpCAM), and Pro656 MMP9cl(B7H3 / EpCAM). D: IGROV cells were tested with Pro662 NCL (EpCAM / B7H3), Pro658 MMP9 (EpCAM / B7H3), and Pro658 MMP9cl (EpCAM / B7H3). The amino acid sequence of Pro656 is shown in Figure 10 Y and SEQ ID NO: 326. The amino acid sequence of Pro658 is shown in Figure 10 Z and SEQ ID NO: 328. [Figure 61D]This is a series of graphs demonstrating that aEpCAM / aB7H3 hetero-COBRA, containing aEpCAM sdABD(h665) and aB7H3 sdABD(hF7)sdABD, conditionally killed tumor cell lines expressing both EpCAM and B7H3. A: IGROV cells tested with Pro295 NCL(B7H3), Pro225 MMP9(B7H3), and Pro225 MMP9cl(B7H3). B: IGROV cells tested with Pro569 NCL(EpCAM), Pro566 MMP9(EpCAM), and Pro566 MMP9cl(EpCAM). C: IGROV cells tested with Pro660 NCL(B7H3 / EpCAM), Pro656 MMP9(B7H3 / EpCAM), and Pro656 MMP9cl(B7H3 / EpCAM). D: IGROV cells were tested with Pro662 NCL (EpCAM / B7H3), Pro658 MMP9 (EpCAM / B7H3), and Pro658 MMP9cl (EpCAM / B7H3). The amino acid sequence of Pro656 is shown in Figure 10 Y and SEQ ID NO: 326. The amino acid sequence of Pro658 is shown in Figure 10 Z and SEQ ID NO: 328. [Figure 62A] This is a series of graphs demonstrating that aEpCAM / aB7H3 hetero-COBRA, containing aEpCAM sdABD(h664) and aB7H3 sdABD(hF7), conditionally killed tumor cell lines expressing both EpCAM and B7H3. A: H292 cells tested with Pro295 NCL(B7H3), Pro225 MMP9(B7H3), and Pro225 MMP9cl(B7H3). B: H292 cells tested with Pro568 NCL(EpCAM), Pro565 MMP9(EpCAM), and Pro565 MMP9cl(EpCAM). C: H292 cells tested with Pro659 NCL(B7H3 / EpCAM), Pro655 MMP9(B7H3 / EpCAM), and Pro655 MMP9cl(B7H3 / EpCAM). D:H292 cells were tested with Pro661 NCL (EpCAM / B7H3), Pro657 MMP9 (EpCAM / B7H3), and Pro657 MMP9cl (EpCAM / B7H3). [Figure 62B]This is a series of graphs demonstrating that aEpCAM / aB7H3 hetero-COBRA, containing aEpCAM sdABD(h664) and aB7H3 sdABD(hF7), conditionally killed tumor cell lines expressing both EpCAM and B7H3. A: H292 cells tested with Pro295 NCL(B7H3), Pro225 MMP9(B7H3), and Pro225 MMP9cl(B7H3). B: H292 cells tested with Pro568 NCL(EpCAM), Pro565 MMP9(EpCAM), and Pro565 MMP9cl(EpCAM). C: H292 cells tested with Pro659 NCL(B7H3 / EpCAM), Pro655 MMP9(B7H3 / EpCAM), and Pro655 MMP9cl(B7H3 / EpCAM). D:H292 cells were tested with Pro661 NCL (EpCAM / B7H3), Pro657 MMP9 (EpCAM / B7H3), and Pro657 MMP9cl (EpCAM / B7H3). [Figure 62C] This is a series of graphs demonstrating that aEpCAM / aB7H3 hetero-COBRA, containing aEpCAM sdABD(h664) and aB7H3 sdABD(hF7), conditionally killed tumor cell lines expressing both EpCAM and B7H3. A: H292 cells tested with Pro295 NCL(B7H3), Pro225 MMP9(B7H3), and Pro225 MMP9cl(B7H3). B: H292 cells tested with Pro568 NCL(EpCAM), Pro565 MMP9(EpCAM), and Pro565 MMP9cl(EpCAM). C: H292 cells tested with Pro659 NCL(B7H3 / EpCAM), Pro655 MMP9(B7H3 / EpCAM), and Pro655 MMP9cl(B7H3 / EpCAM). D:H292 cells were tested with Pro661 NCL (EpCAM / B7H3), Pro657 MMP9 (EpCAM / B7H3), and Pro657 MMP9cl (EpCAM / B7H3). [Figure 62D]This is a series of graphs demonstrating that aEpCAM / aB7H3 hetero-COBRA, containing aEpCAM sdABD(h664) and aB7H3 sdABD(hF7), conditionally killed tumor cell lines expressing both EpCAM and B7H3. A: H292 cells tested with Pro295 NCL(B7H3), Pro225 MMP9(B7H3), and Pro225 MMP9cl(B7H3). B: H292 cells tested with Pro568 NCL(EpCAM), Pro565 MMP9(EpCAM), and Pro565 MMP9cl(EpCAM). C: H292 cells tested with Pro659 NCL(B7H3 / EpCAM), Pro655 MMP9(B7H3 / EpCAM), and Pro655 MMP9cl(B7H3 / EpCAM). D:H292 cells were tested with Pro661 NCL (EpCAM / B7H3), Pro657 MMP9 (EpCAM / B7H3), and Pro657 MMP9cl (EpCAM / B7H3). [Figure 63A] This is a series of graphs demonstrating that aEpCAM / aB7H3 hetero-COBRA, containing aEpCAM sdABD(h665) and aB7H3 sdABD(hF7), conditionally killed tumor cell lines expressing both EpCAM and B7H3. A: H292 cells tested with Pro295 NCL(B7H3), Pro225 MMP9(B7H3), and Pro225 MMP9cl(B7H3). B: H292 cells tested with Pro569 NCL(EpCAM), Pro566 MMP9(EpCAM), and Pro566 MMP9cl(EpCAM). C: H292 cells tested with Pro660 NCL(B7H3 / EpCAM), Pro656 MMP9(B7H3 / EpCAM), and Pro656 MMP9cl(B7H3 / EpCAM). D:H292 cells were tested with Pro662 NCL (EpCAM / B7H3), Pro658 MMP9 (EpCAM / B7H3), and Pro658 MMP9cl (EpCAM / B7H3). [Figure 63B]This is a series of graphs demonstrating that aEpCAM / aB7H3 hetero-COBRA, containing aEpCAM sdABD(h665) and aB7H3 sdABD(hF7), conditionally killed tumor cell lines expressing both EpCAM and B7H3. A: H292 cells tested with Pro295 NCL(B7H3), Pro225 MMP9(B7H3), and Pro225 MMP9cl(B7H3). B: H292 cells tested with Pro569 NCL(EpCAM), Pro566 MMP9(EpCAM), and Pro566 MMP9cl(EpCAM). C: H292 cells tested with Pro660 NCL(B7H3 / EpCAM), Pro656 MMP9(B7H3 / EpCAM), and Pro656 MMP9cl(B7H3 / EpCAM). D:H292 cells were tested with Pro662 NCL (EpCAM / B7H3), Pro658 MMP9 (EpCAM / B7H3), and Pro658 MMP9cl (EpCAM / B7H3). [Figure 63C] This is a series of graphs demonstrating that aEpCAM / aB7H3 hetero-COBRA, containing aEpCAM sdABD(h665) and aB7H3 sdABD(hF7), conditionally killed tumor cell lines expressing both EpCAM and B7H3. A: H292 cells tested with Pro295 NCL(B7H3), Pro225 MMP9(B7H3), and Pro225 MMP9cl(B7H3). B: H292 cells tested with Pro569 NCL(EpCAM), Pro566 MMP9(EpCAM), and Pro566 MMP9cl(EpCAM). C: H292 cells tested with Pro660 NCL(B7H3 / EpCAM), Pro656 MMP9(B7H3 / EpCAM), and Pro656 MMP9cl(B7H3 / EpCAM). D:H292 cells were tested with Pro662 NCL (EpCAM / B7H3), Pro658 MMP9 (EpCAM / B7H3), and Pro658 MMP9cl (EpCAM / B7H3). [Figure 63D]This is a series of graphs demonstrating that aEpCAM / aB7H3 hetero-COBRA, containing aEpCAM sdABD(h665) and aB7H3 sdABD(hF7), conditionally killed tumor cell lines expressing both EpCAM and B7H3. A: H292 cells tested with Pro295 NCL(B7H3), Pro225 MMP9(B7H3), and Pro225 MMP9cl(B7H3). B: H292 cells tested with Pro569 NCL(EpCAM), Pro566 MMP9(EpCAM), and Pro566 MMP9cl(EpCAM). C: H292 cells tested with Pro660 NCL(B7H3 / EpCAM), Pro656 MMP9(B7H3 / EpCAM), and Pro656 MMP9cl(B7H3 / EpCAM). D:H292 cells were tested with Pro662 NCL (EpCAM / B7H3), Pro658 MMP9 (EpCAM / B7H3), and Pro658 MMP9cl (EpCAM / B7H3). [Figure 64A] This is a series of graphs showing the effects of T cell-dependent cytotoxicity (TDCC) on tumor cell lines. HT29 cells (A), U87-MG (EpCAM-negative) cells (B), Capan2 cells (C), and VCAP cells (D) were all tested with monospecific COBRA: Pro225 (B7H3 / B7H3) and Pro566 (EpCAM / EpCAM), and heterogeneous COBRA: Pro656 (B7H3 / EpCAM) and Pro658 (EpCAM / B7H3). All COBRA cells were pre-cleaved. [Figure 64B] This is a series of graphs showing the effects of T cell-dependent cytotoxicity (TDCC) on tumor cell lines. HT29 cells (A), U87-MG (EpCAM-negative) cells (B), Capan2 cells (C), and VCAP cells (D) were all tested with monospecific COBRA: Pro225 (B7H3 / B7H3) and Pro566 (EpCAM / EpCAM), and heterogeneous COBRA: Pro656 (B7H3 / EpCAM) and Pro658 (EpCAM / B7H3). All COBRA cells were pre-cleaved. [Figure 64C]This is a series of graphs showing the effects of T cell-dependent cytotoxicity (TDCC) on tumor cell lines. HT29 cells (A), U87-MG (EpCAM-negative) cells (B), Capan2 cells (C), and VCAP cells (D) were all tested with monospecific COBRA: Pro225 (B7H3 / B7H3) and Pro566 (EpCAM / EpCAM), and heterogeneous COBRA: Pro656 (B7H3 / EpCAM) and Pro658 (EpCAM / B7H3). All COBRA cells were pre-cleaved. [Figure 64D] This is a series of graphs showing the effects of T cell-dependent cytotoxicity (TDCC) on tumor cell lines. HT29 cells (A), U87-MG (EpCAM-negative) cells (B), Capan2 cells (C), and VCAP cells (D) were all tested with monospecific COBRA: Pro225 (B7H3 / B7H3) and Pro566 (EpCAM / EpCAM), and heterogeneous COBRA: Pro656 (B7H3 / EpCAM) and Pro658 (EpCAM / B7H3). All COBRA cells were pre-cleaved. [Figure 65] This graph shows Jurkat luciferase T cell activation in HT29 cells. HT29 cells were tested with monospecific COBRA: Pro225 (B7H3 / B7H3) and Pro566 (EpCAM / EpCAM), and heterogeneous COBRA: Pro656 (B7H3 / EpCAM) and Pro658 (EpCAM / B7H3). All COBRA cells were pre-cleaved. [Figure 66A] This series of graphs shows that hetero-COBRA activity against HT29 cells is less sensitive to inhibition by soluble antigens compared to monospecific COBRA. Cells were assayed with soluble EpCAM, soluble B7H3 4Ig, and monospecific COBRA: A Pro225 (B7H3 / B7H3) and B Pro566 (EpCAM / EpCAM), and hetero-COBRA: C Pro656 (B7H3 / EpCAM) and D Pro658 (EpCAM / B7H3) without antigen (control). Stronger inhibition was detected with monospecific COBRA. All COBRAs were pre-cleaved. [Figure 66B]This series of graphs shows that hetero-COBRA activity against HT29 cells is less sensitive to inhibition by soluble antigens compared to monospecific COBRA. Cells were assayed with soluble EpCAM, soluble B7H3 4Ig, and monospecific COBRA: A Pro225 (B7H3 / B7H3) and B Pro566 (EpCAM / EpCAM), and hetero-COBRA: C Pro656 (B7H3 / EpCAM) and D Pro658 (EpCAM / B7H3) without antigen (control). Stronger inhibition was detected with monospecific COBRA. All COBRAs were pre-cleaved. [Figure 66C] This series of graphs shows that hetero-COBRA activity against HT29 cells is less sensitive to inhibition by soluble antigens compared to monospecific COBRA. Cells were assayed with soluble EpCAM, soluble B7H3 4Ig, and monospecific COBRA: A Pro225 (B7H3 / B7H3) and B Pro566 (EpCAM / EpCAM), and hetero-COBRA: C Pro656 (B7H3 / EpCAM) and D Pro658 (EpCAM / B7H3) without antigen (control). Stronger inhibition was detected with monospecific COBRA. All COBRAs were pre-cleaved. [Figure 66D] This series of graphs shows that hetero-COBRA activity against HT29 cells is less sensitive to inhibition by soluble antigens compared to monospecific COBRA. Cells were assayed with soluble EpCAM, soluble B7H3 4Ig, and monospecific COBRA: A Pro225 (B7H3 / B7H3) and B Pro566 (EpCAM / EpCAM), and hetero-COBRA: C Pro656 (B7H3 / EpCAM) and D Pro658 (EpCAM / B7H3) without antigen (control). Stronger inhibition was detected with monospecific COBRA. All COBRAs were pre-cleaved. [Figure 67]This table shows the binding affinity of heteroCOBRA for B7H3 / EpCAM. Antigens containing only huB7H3-4Ig, huEpCAM, and huB7H3-4Ig and huEpCAM alone were assayed with heteroCOBRA Pro656 (B7H3 / EpCAM) and Pro658 (EpCAM / B7H3). [Figure 68] This graph shows the pharmacokinetics of various B7H3 / EpCAM heterocobra drugs. [Figure 69] This graph shows that B7H3 / EpCAM hetero-COBRA is active in a xenograft model derived from the mouse HT29 cell line. Hetero-COBRAa doses such as Pro660 NCL(B7H3 / EpCAM; 0.3 mg / kg), Pro656 MMP9(B7H3 / EpCAM; 0.01 mg / kg), Pro656 MMP9(B7H3 / EpCAM; 0.03 mg / kg), and Pro656 MMP9(B7H3 / EpCAM; 1 mg / kg) were administered at various time intervals. [Figure 70] This graph shows that B7H3 / EpCAM hetero-COBRA is active in a xenograft model derived from the mouse HT29 cell line. Hetero-COBRA doses, such as Pro662 NCL (EpCAM / B7H3; 0.1 mg / kg) and Pro658 MMP9 (EpCAM / B7H3; 0.1 mg / kg), were administered at various time intervals. [Figure 71] Further sequences of exemplary hetero-COBRA (dual-targeted COBRA) described herein are provided. [Figure 72] Further sequences of exemplary hetero-COBRA (dual-targeted COBRA) described herein are provided. [Figure 73] Further sequences of exemplary single-specific HER2 COBRAs (single-targeted COBRAs) are provided. [Figure 74] Further sequences of exemplary single-specific HER2 COBRAs (single-targeted COBRAs) are provided. [Figure 75]This specification provides the sequences of the humanized anti-EpCAM sdAb h664 and the humanized anti-HER2 sdAb h1139, h1156, h1159, and h1162 described herein. [Modes for carrying out the invention]

[0052] Introduction This 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 cause serious side effects by targeting normal tissues. Therefore, it is necessary to activate the binding ability of therapeutic molecules only in the vicinity of diseased tissue to avoid normal tissue interactions. Accordingly, this invention relates to a multivalent conditionally effective (MCE) protein having 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 T cell antigens such as CD3 under specific conditions. In addition, the MCE protein also contains one or more protease cleavage sites. This means that the therapeutic molecule is produced in a "prodrug"-like form in which the CD3-binding domain is inactive until exposed to the tumor environment. Since the tumor environment contains proteases, exposure to proteases causes the prodrug to cleave and become activated.

[0053] This is generally achieved by restricting the CD3 Fv of MCE to the inactive form discussed herein and using a protein containing a "pseudo" variable heavy domain and a "pseudo" variable light domain that target T cell antigens such as CD3. When TTA directs MCE toward the vicinity of the tumor, the MCE is consequently exposed to the protease. Upon cleavage, the active variable heavy domain and active light domain then pair up to form one or more active ABDs against CD3, thereby recruiting T cells to the tumor and delivering therapeutic effects.

[0054] Generally, the CD3-binding domain ("Fv") is a constrained form where the linker between the active variable heavy domain and the active variable light domain that traditionally form the Fv is too short for 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 form 1) or constrained and incleavable (CNCL used in form 2). Rather, in the prodrug (e.g., uncleaved) form, the prodrug polypeptide also contains a "pseudo-Fv domain." The pseudo-Fv domain contains a variable heavy domain and a variable light domain and has a standard framework region, but the CDR is "inert" or "inactive." The pseudo-Fv domain also has a constrained linker between the inert variable heavy domain and the inert variable light domain. Since neither the Fv domain nor the pseudo-Fv domain can self-assemble due to steric constraints, intramolecular assemblies exist that pair aVL and iVH and aVH and iVL depending on the affinity of their respective framework regions. However, the "inactive" CDR of the pseudodomain prevents the resulting ABD from binding to CD3, thereby preventing toxicity outside diseased tissues such as tumors. However, in the presence of proteases within or near tumors, the prodrug construct induces effective CD3 binding and produces an antitumor effect by cleaving, such that the pseudo-Fv domain is released from the surface and the "true" variable heavy and variable light domains become intermolecularly associable (e.g., two cleaved constructs come together). These constructs are generally referred to herein as conditionally bispecific reorientation-activated constructs or "COBRA®". The stability of the intramolecular assemblies has been demonstrated herein by conditioned experiments, and in the absence of proteases, the uncleaved constructs are inactive (e.g., no active CD3-binding domain is formed).

[0055] Interestingly, for the sake of simplicity, all these constructs are referred to herein as “constrained,” while further studies show that intramolecular assemblies are preferable even when one of the Fv domains is unconstrained, for example, when one of the domains can have a longer, more flexible linker. That is, as shown in Figures 37–39, intramolecular assemblies still occur when only one of the Fv domains is constrained, whether it is the domain with active VL and VH, or a pseudo-Fv domain (e.g., the uncleaved construct is inactive in the absence of protease cleavage). However, in this system, the protein exhibits better expression when both linkers are constrained. However, as will be understood by those skilled in the art, any of the Form 1, Form 2, or Form 4 constructs herein may have one of these Fv domains with an “unconstrained” or “flexible” linker. For ease of reference, the constructs are shown in the form where both Fv domains are constrained.

[0056] The constructs and forms of the present invention are variations of embodiments described in International Publication Nos. WO2017 / 156178, WO2019 / 051102, WO2020 / 181140, U.S. Patent Nos. 2019 / 0076524 and 2020 / 0347132, which are expressly incorporated herein in their entirety by reference. As shown in Figures 17-21 of International Publication No. WO2017 / 156178, the figures of International Publication No. WO2019 / 051102 and the figures of International Publication No. WO2020 / 181140, the prior constructs have the ability to isomerize by the presence of two sets of VH and VL domains in a single polypeptide to form both bivalent scFv and single-chain bispecific antibodies. Even after purification of each isoform, the bivalent constructs can still reach equilibrium with the bispecific antibody isoforms. Single-chain bispecific antibodies have the ability to bind to CD3 in the absence of protease cleavage, which reduces the usefulness of the construct.

[0057] To solve this problem, the present invention provides four distinct types of constructs to achieve this conditional activation. Prodrug activation can occur in one of four general ways, as generally shown in the figures. Figure 1 shows the “Form 1” mechanism. In this embodiment, the prodrug construct has two cleavage sites. One is located between the VH and vl domains of the restrictive Fv, thus allowing the two variable domains to freely associate, and the second cleavage site is located where the pseudo-Fv domain is released from the prodrug construct, leaving two molecules that associate by intrinsic self-assembly of the variable weight and variable light domains, each similarly possessing an antigen-binding domain against a tumor antigen, thus enabling the recruitment of T cells to the tumor site.

[0058] In an alternative embodiment, the prodrug construct is shown in the “Form 2” mechanism in Figure 2. In this embodiment, the domain linker between the active variable heavy chain and the active light chain is a restricted but non-cleavable linker (“CNCL”). In the prodrug form, the restricted pseudo-Fv domain inactive VH and VL associate with the restricted Fv domain VH and VL such that CD3 binding is absent. However, once cleavage occurs in the tumor environment, the two distinct activating proteins, each containing an active variable heavy and light domain, associate to form two anti-CD3 binding domains. This Form 2 has two target tumor antigen binding domains (“TTA-ABD”), which can be identical (e.g., “homo-COBRA”) or different (e.g., “hetero-COBRA”), as will be explained more fully below. If different, they may be directed to different tumor antigens, or they may be directed to the same tumor antigen but to different epitopes, as will be explained more fully below.

[0059] In addition to the "single-chain protein" COBRA form, which is considered in which all components are contained on a single amino acid sequence, there are also constructs that rely on two proteins acting in pairs, known as "half-COBRA," as shown in Figure 3. In this embodiment, each protein has one active variable domain and one inactive variable domain separated by a protease cleavage site. Each molecule contains a TTA-binding domain, which allows the molecule to bind to TTA. When exposed to a tumor protease, the inactive domain is cleaved, and the two active variable domains self-assemble to form an anti-CD3-binding domain.

[0060] Furthermore, the present invention also provides “Form 4” constructs, as shown in Figure 4. These are similar to the “Form 2” designs, except that a single ABD is used for TTA, and upon cleavage thereafter, two of the prodrug molecules form a tetravalent bispecific construct containing two active anti-CD3 domains, as further described below.

[0061] Therefore, the forms and structures of the present invention are used in the treatment of diseases.

[0062] definition To enable a more complete understanding of this application, several definitions are provided below. Such definitions are intended to encompass grammatical equivalents.

[0063] As used herein, “amino acid” and “amino acid identity” mean one of the 20 natural amino acids that may be present in a particular position, or any non-natural analog. In many embodiments, “amino acid” means one of the 20 natural amino acids. “Protein” as used herein means at least two covalently bonded amino acids, and includes proteins, polypeptides, oligopeptides, and peptides.

[0064] In this specification, "amino acid modification" means the substitution, insertion, and / or deletion of amino acids in a polypeptide sequence, or modification of a portion chemically bound to a protein. For example, the modification may be a modified carbohydrate or PEG structure bound to a protein. For clarity, unless otherwise specified, amino acid modification always refers to the amino acids encoded by DNA, such as the 20 amino acids that have codons in DNA and RNA. The preferred amino acid modification in this specification is substitution.

[0065] "Amino acid substitution" or "substitution" as used herein means replacing an amino acid at a specific position in a parent polypeptide sequence with a different amino acid. In particular, in some embodiments, the substitution is for a non-native amino acid at a specific position in either a living organism or any organism. For clarity, a protein that has been engineered to change the nucleic acid coding sequence but not the starting amino acid (for example, replacing CGG (which codes for arginine) with CGA (which also codes for arginine) to increase its expression level 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 that protein has the same amino acid at the specific position where it started, it is not an amino acid substitution.

[0066] "Amino acid insertion" or "insertion," as used herein, means adding an amino acid sequence to a specific position in a parent polypeptide sequence.

[0067] "Amino acid deletion" or "deletion," as used herein, means the removal of an amino acid sequence at a specific position in the parent polypeptide sequence.

[0068] The polypeptide of the present invention specifically binds to CD3 and a target tumor antigen (TTA), such as a target cell receptor outlined herein. "Specific binding", or "specifically binds to" or "specific for" a particular antigen or epitope, means 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 generally a molecule of similar structure that does not have binding activity. For example, specific binding can be measured by competition with a control molecule similar to the target.

[0069] Specific binding to a particular antigen or epitope can be demonstrated, for example, by an antibody having at least about 10 , -7 M, at least about 10 -5 M, 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 M, or more KD (KD means the dissociation rate of a specific antibody-antigen interaction). 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 more greater than that of a control molecule.

[0070] Similarly, specific binding to a particular antigen or epitope can be demonstrated, for example, by an antibody having a KA or Ka (KA or Ka means the association rate of a specific 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 greater than that of a control. Binding affinity is usually measured using a Biacore assay or Octet, as is well known in the art.

[0071] "Parent polypeptide" or "precursor polypeptide" (including Fc parent or Fc precursor) means, as used herein, a polypeptide that is subsequently modified to produce a variant. The parent polypeptide may be a natural polypeptide, or a variant or modified form of a natural polypeptide. The parent polypeptide may refer to the polypeptide itself, a composition containing the parent polypeptide, or an amino acid sequence encoding the parent polypeptide. Therefore, "parent Fc polypeptide" means, as used herein, an unmodified Fc polypeptide that is modified to produce a variant, and "parent antibody" means, as used herein, an unmodified antibody that is modified to produce a variant antibody.

[0072] As used herein, "position" means a location within a protein sequence. Positions may be numbered sequentially or according to an established system, such as the EU index for antibody numbering.

[0073] As used herein, "target antigen" means a molecule to which the variable region of any antibody specifically binds. The target antigen may be a protein, carbohydrate, lipid, or other compound. Various preferred and exemplary target antigens are described herein.

[0074] As used herein, "target cells" means cells that express a target antigen. Generally, for the purposes of the present invention, target cells are either tumor cells that express TTA or T cells that express the CD3 antigen.

[0075] "Fv," "Fv domain," or "Fv region" as used herein generally refers to a polypeptide containing VL and VH domains of an antigen-binding domain from an antibody. Fv domains typically form an "antigen-binding domain" or "ABD" as considered herein, if they contain active VH and VL domains (in some cases, Fv containing a restrictive linker is used, thereby preventing the formation of an active ABD before cleavage). As discussed below, Fv domains can be organized in many ways of the present invention and can be "active" or "inactive" in scFv form, restrictive Fv form, pseudo-Fv form, etc. In the present invention, it should be understood that in some cases, an Fv domain consists of VH and VL domains on a single polypeptide chain, as shown in Figures 1 and 2, but has a restrictive linker so as not to form an intramolecular ABD. In these embodiments, the two active ABDs are formed after cleavage. In some cases, an Fv domain consists of VH and VL domains, one of which is inactive, thereby preventing the formation of an intermolecular ABD only after cleavage. As will be discussed below, Fv domains can be organized in many ways according to the present invention and can be “active” or “inactive” in the form of scFv, restrictive Fv, pseudo-Fv, etc. In addition, as will be discussed herein, Fv domains containing VH and VL can be ABDs / ABDs, and other ABDs that do not contain VH and VL domains can be formed using sdABDs.

[0076] "Variable domain" as used herein means a region of immunoglobulin containing one or more Ig domains substantially encoded by any of the Vκ, Vλ, and / or VH genes that constitute the kappa immunoglobulin locus, lambda immunoglobulin locus, and heavy chain immunoglobulin locus, respectively. In some cases, a single variable domain such as sdFv (also referred to herein as sdABD) may be used.

[0077] In embodiments utilizing both variable weight (VH) and variable light (VL) domains, each VH and VL consists of three hypervariable regions ("complementarity-determining regions," or "CDRs") and four "framework regions," or "FRs," arranged from the amino terminus to the carboxyl terminus in the following order: FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4. Therefore, 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 will be described in more detail herein, the vhFR and vlFR regions self-assemble to form the Fv domain. Generally, the prodrug form of the present invention contains a "restrictive Fv domain" in which the VH and VL domains cannot self-associate, as well as a "pseudo-Fv domain" that does not form an antigen-binding domain when the CDR self-associates.

[0078] The hypervariable region confers antigen-binding specificity and typically consists of amino acid residues around 24-34 (LCDR1, where "L" represents the light chain), 50-56 (LCDR2), and 89-97 (LCDR3) in the light chain variable region, and around 31-35B (HCDR1, where "H" represents the 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 This includes Health, Bethesda, Md. (1991), and / or the amino acid residues that form the hypervariable loop (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.

[0079] As those skilled in the art will understand, the exact numbering and arrangement of CDRs may differ among 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. Therefore, the present disclosure of each variable heavy chain region is a disclosure of vhCDRs (e.g., vhCDR1, vhCDR2, and vhCDR3), and the present disclosure of each variable light chain region is a disclosure of vlCDRs (e.g., vlCDR1, vlCDR2, and vlCDR3).

[0080] For available comparisons of CDR numbering, please refer to Lafranc et al., Dev.Comp.Immunol.27(1):55-77(2003). [Table 1]

[0081] Throughout this specification, when referring to residues in the variable domain (approximately residues 1-107 in the light chain variable region and residues 1-113 in the heavy chain variable region), the Kabat numbering system is generally used, and the EU numbering system (e.g., Kabat et al., supra (1991)) is used for the Fc region.

[0082] The present invention provides a number of different CDR sets. In this case, a “complete CDR set” in the context of anti-CD3 components includes three variable light chain CDRs and three variable heavy chain CDRs, e.g., vlCDR1, vlCDR2, vlCDR3, vhCDR1, vhCDR2, and vhCDR3. As those skilled in the art will understand, each set of CDRs, VHCDR and VLCDR, can bind to the antigen both individually and as a set. For example, in the restrictive Fv domain, vhCDR can bind to CD3, and vlCDR can bind to CD3, but in the restricted form, vhCDR and vlCDR cannot bind to CD3.

[0083] In the context of single-domain ABDs ("sdABDs"), as commonly used herein for binding to target tumor antigens (TTAs), the CDR set consists of only three CDRs. These are sometimes referred to in the art as "VHH" domains.

[0084] Each of these CDRs may be part of a larger variable light domain or variable weight domain. In addition, as will be more fully outlined herein, variable weight 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 form and arrangement of the parts herein.

[0085] CDRs contribute to the formation of antigen-binding sites, or more specifically, epitope-binding sites. An "epitope" refers to a determinant that interacts with a specific antigen-binding site in a variable region known as a paratope. Epitopes are aggregates of molecules, such as amino acids or sugar side chains, and typically possess specific charge properties in addition to specific structural properties. A single antigen may have two or more epitopes.

[0086] The epitope may include amino acid residues directly involved in binding (also called the major antigenic components of the epitope), and other amino acid residues not directly involved in binding, such as amino acid residues that are effectively inhibited by peptides that specifically bind to the antigen (i.e., those amino acid residues are within the footprint of the peptides that specifically bind to the antigen).

[0087] Epitopes can be either conformational or linear. Conformational epitopes are generated by spatially aligned amino acids derived from different segments of a linear polypeptide chain. Linear epitopes are epitopes generated by adjacent amino acid residues within a polypeptide chain. Conformational and non-conformational epitopes can be distinguished by the fact that binding to the former is lost in the presence of a denaturing solvent, while binding to the latter is not.

[0088] Epitopes typically have a unique spatial conformation and contain at least three, more commonly, at least five, or eight to ten amino acids. Antibodies that recognize the same epitope can be verified by a simple immunoassay, such as "binning," which demonstrates the ability of one antibody to inhibit the binding of the other antibody to its target antigen. As outlined below, the present invention includes not only the antigen-binding domains and antibodies listed herein, but also antigen-binding domains and antibodies that compete for binding to the epitopes to which the listed antigen-binding domains bind.

[0089] The variable heavy domain and variable light domain of the present invention may be "active" or "inactive".

[0090] As used herein, “inactive VH” (“iVH”) and “inactive VL” (“iVL”) refer to components of a pseudo-Fv domain. “Inactive VH” (“iVH”) and “inactive VL” (“iVL”), when paired with their homogeneous VL partner or homogeneous VH partner, respectively, form a resulting VH / VL pair that does not specifically bind to antigens to which “active” VH or “active” VL can bind, but if the antigen binds to a similar VL or similar VH, the similar VL or similar VH is not “inactive.” Exemplary “inactive VH” and “inactive VL” domains are formed by mutations in the wild-type VH or VL sequence, as will be more fully outlined below. Exemplary mutations reside within the CDR1, CDR2, or CDR3 of the VH or VL. Exemplary mutations include forming an “inactive VH” domain or an “inactive VL” domain by placing a domain linker within CDR2. In contrast, "active VH" or "active VL" is a VH or VL that, when paired with its "active" allogeneic partner, i.e., VL or VH, can specifically bind to its target antigen. Therefore, it should be understood that a pseudo-Fv can be a VH / iVL pair, an iVH / VL pair, or an iVH / iVL pair.

[0091] 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 a single member of an Fv-binding pair (i.e., VH or VL) (a sequence that, in conjunction with its homogeneous partner, can specifically bind to CD3); and (b) when referring to a homogeneous pair of sequences capable of specifically binding to CD- (i.e., VH and VL). Exemplary “active” VH, VL, or VH / VL pairs are wild-type sequences or parent sequences.

[0092] "CD-x" refers to a differentiation antigen (CD) protein. In exemplary embodiments, CD-x is selected from those CD proteins that play a role in recruiting or activating T cells in subjects administered with the polypeptide construct of the present invention. In exemplary embodiments, CD-x is CD3, and its sequence is shown in Figure 7.

[0093] In relation to the present invention, the term "binding domain" refers to a domain that (specifically) binds to / interacts with / recognizes a target molecule (antigen), such as any target epitope or target site on EGFR and CD3, respectively. The structure and function of the target antigen-binding domain (which recognizes EGFR), and preferably also the structure and / or function of the CD3-binding domain (which recognizes CD3), are based on the structure and / or function of an antibody, such as a full-length antibody or a complete immunoglobulin molecule containing sdABD. According to the present invention, the target antigen-binding domain is typically characterized by the presence of three CDRs that bind to the target tumor antigen (although there are no corresponding light chain CDRs, these are commonly referred to as variable heavy domains in the art). Alternatively, ABD for TTA may contain three light chain CDRs (i.e., CDR1, CDR2, and CDR3 in the VL region) and / or three heavy chain CDRs (i.e., CDR1, CDR2, and CDR3 in the VH region). The CD3-binding domain also preferably contains at least the minimum structural requirements of the antibody that enable target binding. More preferably, the CD3-binding domain comprises at least three light chain CDRs (i.e., CDR1, CDR2, and CDR3 in the VL region) and / or three heavy chain CDRs (i.e., CDR1, CDR2, and CDR3 in the VH region). In exemplary embodiments, it is assumed that the target antigen and / or CD3-binding domain can be constructed or obtained using phage display or library screening.

[0094] As used herein, "domain" refers to a protein sequence having the function outlined herein. Examples of domains in the present invention include tumor target antigen-binding domains (TTA domains), variable weight domains, variable light domains, scFv domains, linker domains, and half-life extension domains.

[0095] In this specification, "domain linker" means an amino acid sequence that links two domains outlined herein. Domain linkers may be cleavable linkers, restricted cleavable linkers, non-cleavable linkers, restricted non-cleavable linkers, scFv linkers, etc.

[0096] A “cleavable linker” ("CL"), as used herein, means an amino acid sequence that can be cleaved by a protease, preferably a human protease, in the disease tissue outlined herein. Cleavable linkers are typically at least 3 amino acids long, and depending on the required flexibility, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or more amino acids are useful in the present invention. Figures 6 and 7 show many cleavable linker sequences.

[0097] In this specification, "non-cleavable linker" ("NCL") means an amino acid sequence that cannot be cleaved by human proteases under normal physiological conditions.

[0098] A “constrained cleavage linker” (“CCL”), as used herein, means a short polypeptide containing a protease cleavage site (as defined herein) that links two domains outlined herein, such that the two domains cannot significantly interact with each other after they are located on different polypeptide chains, for example, until after cleavage. When a CCL binds to the VH and VL domains as defined herein, VH and VL cannot self-assemble to form a functional Fv before cleavage due to intermolecular steric hindrance (although they can assemble intermolecularly into a pseudo-Fv domain). Upon cleavage by the relevant protease, VH and VL can assemble to form an intermolecularly active antigen-binding domain. Typically, CCLs have an amino acid length of less than 10, with 9, 8, 7, 6, 5, and 4 amino acids being useful in the present invention. Generally, the protease cleavage site is at least 4+ amino acid long to confer sufficient specificity, as shown in Figure 6.

[0099] A “constrained non-cleavable linker” ("CNCL"), as used herein, means a short polypeptide that connects two domains 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.

[0100] In this specification, a "constrained Fv domain" refers to an Fv domain containing an active variable heavy domain and an active variable light domain that are covalently linked to a restrictive linker outlined herein, in a state where the active variable heavy domain and the active variable light domain interact intramolecularly to prevent the formation of an active Fv that binds to an antigen such as CD3. Therefore, although a restricted Fv domain is similar to an scFv, it is a domain that cannot bind to an antigen due to the presence of a restrictive linker (although they can assemble with inactive variable domains intermolecularly to form a pseudo-Fv domain).

[0101] In this specification, "pseudo-Fv domain" means a domain containing a pseudo- or inactive variable heavy domain or a pseudo- or inactive variable light domain linked using a domain linker (which may be cleavable, restrictive, non-cleavable, or unrestrictive). The iVH domain and iVL domain of a pseudo-Fv domain do not bind to human antigens, whether associated with each other (iVH / iVL) or with active VH or active VL. Therefore, the iVH / iVL Fv domain, iVH / VL Fv domain, and iVL / VH Fv domain do not bind to human proteins to a measurable extent, and as a result, these domains are inactive in the human body.

[0102] In this specification, "single-stranded Fv," or "scFv," refers to a variable weight (VH) domain covalently bonded to a variable light (VL) domain, which is used to form scFv or an scFv domain using the domain linkers discussed herein. The scFv domain can be oriented either N-terminus or C-terminus (VH-linker-VL or VL-linker-VH).

[0103] In this specification, “single-domain Fv,” “sdFv,” or “sdABD” generally refers to an antigen-binding domain having only three CDRs, based on camel antibody technology. See Protein Engineering 9(7):1129-35 (1994), Rev Mol Biotech 74:277-302 (2001), and Ann Rev Biochem 82:775-97 (2013). As outlined herein, two common types of sdABD are used herein: sdABD that binds to TTA and is annotated as such (generic names include sdABD-TTA, or sdABD-EGFR for those that bind to EGFR, sdABD-FOLR1 for those that bind to FOLR1, etc.), and sdABD that binds to HSA ("sdABD-HSA" or "sdABD(1 / 2)").

[0104] The "protease cleavage site" refers to the amino acid sequence that a protease recognizes and cleaves. Suitable protease cleavage sites are outlined below and shown in Figures 7 and 6.

[0105] As used herein, "protease cleavage domain" means "protease cleavage site," and any linkers between individual protease cleavage sites and between protease cleavage sites(s), as well as other functional components of the construct of the present invention (e.g., V H , V L This refers to a peptide sequence incorporating an iVH, iVL, target antigen-binding domain(s), half-life extension domain(s), etc. As outlined herein, the protease cleavage domain may also contain additional amino acids as needed, for example, to provide flexibility.

[0106] The terms "COBRA®" and "conditional bispecific redirection activation" refer to a bispecific, conditionally active 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 specific conditions. Examples of T cell antigens include, but are not limited to, CD3. The term "half-COBRA®" refers to a conditionally active protein that can bind to a T cell antigen if, upon aggregation on the surface of a target-expressing cell, its variable heavy chain becomes capable of associating with the variable light chain of another half-COBRA® (complementary half-COBRA®).

[0107] Modes for carrying out the invention I. The fusion protein of the present invention The fusion protein of the present invention has several different components linked together in various ways, which are generally referred to herein as domains. Some of the domains are binding domains, each binding to a target antigen (e.g., TTA or CD3). When binding domains bind to two or more antigens, those binding domains 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 the first αTTA binds to EGFR, the second αTTA binds to EpCAM, and an anti-CD3 binding domain is present, this is a “triplespecific” molecule. Similarly, the addition of an anti-HSA binding domain to this construct makes it “quadrispecific,” as shown in Figure 3B.

[0108] As will be understood by those skilled in the art, the proteins of the present invention can have different valencies and may be multispecific. That is, the proteins of the present invention can bind to two or more binding sites for a target; for example, Pro186 is divalent with respect to EGFR.

[0109] The protein of the present invention may include a CD3 antigen-binding domain, a tumor target antigen-binding domain, a half-life extension domain, a linker, etc., arranged in various configurations outlined herein.

[0110] A. CD3 antigen-binding domain The specificity of the T cell response is mediated by the recognition of antigens (presented in relation to the major histocompatibility complex, MHC) by the T cell receptor complex. As part of the T cell receptor complex, CD3 is a protein complex present on the cell surface, containing CD3γ (gamma) chains, CD3δ (delta) chains, 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 form the TCR complex. Clustering of CD3 on T cells by Fv domains and other factors that bind to CD3 leads to T cell activation that is similar to the involvement of the T cell receptor but is independent of the clonal specificity.

[0111] However, as is well known in the art, CD3 activation causes many toxic side effects. Therefore, the present invention relates to enabling effective CD3 binding of the polypeptide of the present invention only in the presence of tumor cells in which a specific protease is present, allowing the prodrug polypeptide to be cleaved and an active CD3-binding domain to be obtained. Accordingly, in the present invention, the binding of the anti-CD3 Fv domain to CD3 is controlled by a protease-cleaving domain that restricts the binding of the CD3 Fv domain to CD3 only to the microenvironment of disease cells or tissues with high protease levels, such as the tumor microenvironment described herein.

[0112] Accordingly, the present invention provides two sets of VH and VL domains: an active set (VH and VL) and an inactive set (inactive VH and inactive VL), all four of which are present in a prodrug construct. The construct is configured such that the VH and VL sets cannot self-associate, but rather associate with inactive partners, such as iVH and VL and iVL and VH as shown herein.

[0113] 1. Activated anti-CD3 variable weight and variable light domains Many suitable active CDR sets and / or VH and VL domains are known in the art and useful for the present invention. For example, the CDR and / or VH and VL domains are well known anti-CD3 antibodies, such as muromonab-CD3 (OKT3), otelixizumab (TRX4), teplizumab (MGA031), bicilizumab (Nuvion), SP34 or I2C, TR-66 or X35-3, VIT3, BMA030 (BW264 / 56), CLB-T3 / 3, C It originates from 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, among others.

[0114] In one embodiment, the VH and VL sequences that form an active Fv domain that binds to human CD3 are shown in Figures 7A and 7B. As shown herein, these active VH ("aVH") and active VL ("aVL") domains can be used in different configurations and forms 1, 2, 3, and 4.

[0115] 2. Inactive anti-CD3 variable weight and variable light domains The inactive iVH and iVL domains contain a “regular” framework region (FR) that enables association, thereby allowing the inactive variable domain to associate with the active variable domain and inactivate the pair, for example, preventing it from binding to CD3.

[0116] As those skilled in the art will understand, there are several "inactive" variable domains useful for the present invention. Basically, any variable domain having a human framework region that enables self-assembly with other variable domains can be used, regardless of which amino acid is located at the CDR position of the variable region. For simplicity, although strictly speaking inactive variable domains do not confer binding ability, inactive domains are said to contain a CDR.

[0117] As will be appreciated in the art, generating an inert VH or VL domain is generally straightforward and can be done in a variety of ways. In some embodiments, generating an inert variable domain generally involves modifying one or more of the CDRs of an active Fv by altering 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 of the 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 unrelated antibodies (e.g., those directed against proteins of different organisms).

[0118] In some cases, only one of the CDRs within the variable region may be modified to render it inert, while other embodiments involve modification of 1, 2, 3, 4, 5 or 6 CDRs.

[0119] In some cases, the inert domain can be engineered in a prodrug format to promote selective binding and to favor (e.g., over intermolecular pairing) the formation of intramolecular iVH-VL and VH-iVL domains prior to cleavage. See, e.g., Igawa et al., Protein Eng. Des. Selection 23(8):667-677 (2010), which is hereby incorporated by reference in its entirety, particularly with respect to interfacial residue amino acid substitutions.

[0120] In certain embodiments, the CD3 binding domain of the polypeptide constructs described herein not only exhibits strong CD3 binding affinity for human CD3, but also exhibits good cross-reactivity with the corresponding cynomolgus CD3 protein. In some cases, the CD3 binding domain of the polypeptide construct cross-reacts with cynomolgus-derived CD3. In certain cases, the human:cynomolgus KD ratio for CD3 is 5 to 0.2.

[0121] 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 antibodies, polyclonal antibodies, recombinant antibodies, human antibodies, and humanized antibodies. In some cases, it is beneficial for the CD3-binding domain to originate from the same species in which the antigen-binding protein is ultimately used. For example, when used in humans, it may be advantageous to include human or humanized residues derived from the antigen-binding domain of the antibody or antibody fragment in the CD3-binding domain of the antigen-binding protein.

[0122] Therefore, in one embodiment, the antigen-binding domain includes a humanization-binding domain or a human-binding domain. In one embodiment, the humanized anti-CD3-binding domain or human anti-CD3-binding domain includes 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 the humanized anti-CD3-binding domain 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 the humanized anti-CD3-binding domain or human anti-CD3-binding domain described herein, for example, the humanized anti-CD3-binding domain or human anti-CD3-binding domain includes one or more, e.g., all three LC CDRs, and one or more, e.g., all three HC CDRs.

[0123] In some embodiments, the humanized anti-CD3 binding domain or human anti-CD3 binding domain includes a CD3-specific humanized light chain variable region or human light chain variable region, the CD3-specific light chain variable region containing a human light chain CDR or non-human light chain CDR within the human light chain framework region. In certain examples, the light chain framework region is a lambda (λ) light chain framework. In other examples, the light chain framework region is a kappa (κ) light chain framework.

[0124] In some embodiments, one or more CD3-binding domains are humanized or fully human. In some embodiments, one or more activated CD3-binding domains have KD bindings of 1000 nM or less to CD3 on CD3-expressing cells. In some embodiments, one or more activated CD3-binding domains have KD bindings of 100 nM or less to CD3 on CD3-expressing cells. In some embodiments, one or more activated CD3-binding domains have KD bindings of 10 nM or less to CD3 on CD3-expressing cells. In some embodiments, one or more CD3-binding domains are cross-reactive with cynomolgus monkey CD3. In some embodiments, one or more CD3-binding domains include amino acid sequences provided herein.

[0125] In some embodiments, the humanized anti-CD3 binding domain or human anti-CD3 binding domain includes a CD3-specific humanized heavy chain variable region or human heavy chain variable region, the CD3-specific heavy chain variable region including a human heavy chain CDR or non-human heavy chain CDR within the human heavy chain framework region.

[0126] In some embodiments, the anti-CD3 binding domain is an Fv comprising the light and heavy chains of the amino acid sequences 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) of the amino acid sequence of the light chain variable region provided herein, but with 30, 20, or 10 or fewer modifications (e.g., substitutions), 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) of the amino acid sequence of the heavy chain variable region provided herein, but with 30, 20, or 10 or fewer modifications (e.g., substitutions), 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 the light chain variable region comprising the amino acid sequence described herein is bound to the heavy chain variable region comprising the amino acid sequence described herein via an scFv linker. The light chain variable region and heavy chain variable region of 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.

[0127] In some embodiments, the CD3-binding domain of the antigen-binding protein has a KD affinity of ≤1000 nM, ≤100 nM, ≤50 nM, ≤20 nM, ≤10 nM, ≤5 nM, ≤1 nM, or ≤0.5 nM for CD3 on CD3-expressing cells. In some embodiments, the CD3-binding domain of the antigen-binding protein has a KD affinity of ≤1000 nM, ≤100 nM, ≤50 nM, ≤20 nM, ≤10 nM, ≤5 nM, ≤1 nM, or ≤0.5 nM for CD3ε. 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.

[0128] For example, as is well known in the art, the binding affinity to CD3 may be measured by the binding ability of the antigen-binding protein itself or its CD3-binding domain to CD3, such as when it is coated on an assay plate, presented on the surface of bacterial cells, or in solution (generally using the Biacore assay or Octet assay). The binding activity of the antigen-binding protein itself or its CD3-binding domain to CD3 may also be assayed by immobilizing the ligand (e.g., CD3) or the antigen-binding protein itself or its CD3-binding domain on beads, a substrate, cells, etc. The drug may be added to a suitable buffer and binding partner and cultured at any temperature for a period of time. After washing to remove unbound substances, the binding protein may be detached using, for example, SDS or a high-pH buffer, and then analyzed, for example, by surface plasmon resonance (SPR).

[0129] In many embodiments, preferred active and inactive binding domains are those shown in Figure 7. Figure 7 shows one active VH and VL, as well as three inactive VHi and three inactive VLi, inactivated in different ways.

[0130] As shown in Figure 7, particularly useful pairs of active anti-CD3 VL and VH domains include a VL comprising vlCDR1 of SEQ ID NO: 255, vlCDR2 of SEQ ID NO: 256, and vlCDR3 of SEQ ID NO: 257, and a VH comprising vhCDR1 of SEQ ID NO: 271, vhCDR2 of SEQ ID NO: 272, and vhCDR3 of SEQ ID NO: 273.

[0131] As shown in Figure 7, a particularly useful pair of active anti-CD3 VL and VH domains is the VL of SEQ ID NO: 254 and the VH of SEQ ID NO: 270.

[0132] B. Antigen-binding domain for tumor target antigens In addition to the CD3 and half-life extension domains described herein, the polypeptide constructs described herein also include target domains that bind to one or more target antigens, or to one or more regions on a single target antigen. In this specification, it is intended that the polypeptide constructs of the present invention are cleaved, for example, in a disease-specific microenvironment or in the blood of a target at the protease cleavage domain, and that each target antigen-binding domain binds to a target antigen on a target cell, thereby activating the CD3-binding domain to bind to T cells. Generally, TTA-binding domains can bind to their targets before protease cleavage, and therefore they can "wait" on the target cell to be activated as T cell induction. At least one target antigen is involved in and / or associated with a disease, disorder, or condition. Exemplary target antigens include those related to proliferative disorders, neoplastic diseases, inflammatory diseases, immunological disorders, autoimmune diseases, infectious diseases, viral diseases, allergic reactions, parasitic reactions, graft-versus-host diseases, or host-versus-graft diseases. In some embodiments, the target antigen is a tumor antigen expressed on tumor cells. Alternatively, in some embodiments, the target antigen is related to a pathogen, such as a virus or bacteria. At least one target antigen may also be related to healthy tissue.

[0133] In some embodiments, the target antigen is a cell surface molecule, such as a protein, lipid, or polysaccharide. In some embodiments, the target antigen is located on tumor cells, virus-infected cells, bacterial-infected cells, damaged red blood cells, arterial plaque cells, or fibrous tissue cells.

[0134] A preferred embodiment of the present invention utilizes sdABD as the targeting domain. These are preferred over scFv ABD because the addition of other VH and VL domains to the construct of the present invention can complicate the form of the pseudo-Fv domain.

[0135] In some embodiments, the prodrug constructs of the present invention generally utilize a single TTA-binding domain, as shown in Figure 3A as an sdABD-TTA pair and in Figure 4 as a “Form 4” configuration. Figure 4 shows the use of a single anti-EGFR ABD, but other TTA-binding domains may be used.

[0136] In some embodiments, particularly in the Form 1 and Form 2 constructs, the prodrug constructs of the present invention utilize two TTA ABDs, also preferably in the sdABD-TTA form. When dual targeting domains are used, they can bind to the same epitope of the same TTA. For example, many of the constructs herein, as considered herein, utilize two identical targeting domains. In some embodiments, two targeting domains that bind to different epitopes of the same TTA may be used, for example, as shown in Figure 7, two EGFR sdABDs bind to different epitopes on human EGFR. In some embodiments, the two targeting domains bind to different TTAs, as fully described below.

[0137] The polypeptide constructs envisioned 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, LyPD3, B7H3, CA9, Trop2, and FOLR1. These can be combined in various ways, as described below.

[0138] (a) EGFR sdABD As shown in A of FIG. 5, there are a number of particularly useful sdABDs that bind to human EGFR, referred to herein as "αEGFR", "aEGFR", "sdABD-EGFR", "EGFR sdABD", "EGFR sdAbs", "EGFR ABD" or "EGFRABD".

[0139] In some embodiments, sdABD-EGFR (e.g., sdABD-αEGFR1) has sdCDR1 of SEQ ID NO: 2, sdCDR2 of SEQ ID NO: 3, and sdCDR3 of SEQ ID NO: 4. In some cases, sdABD-EGFR has the amino acid sequence of SEQ ID NO: 1 as shown in A of FIG. 5.

[0140] In some embodiments, sdABD-EGFR (e.g., sdABD-αEGFR2) has sdCDR1 of SEQ ID NO: 6, sdCDR2 of SEQ ID NO: 7, and sdCDR3 of SEQ ID NO: 8. In some cases, sdABD-EGFR has the amino acid sequence of SEQ ID NO: 5 as shown in A of FIG. 5.

[0141] In some embodiments, sdABD-EGFR (e.g., sdABD-hαEGFR1) has sdCDR1 of SEQ ID NO: 10, sdCDR2 of SEQ ID NO: 11, and sdCDR3 of SEQ ID NO: 12. In some cases, sdABD-EGFR has the amino acid sequence of SEQ ID NO: 9 as shown in A of FIG. 5.

[0142] In some embodiments, sdABD-EGFR (e.g., sdABD-aEGFR2a) has sdCDR1 of SEQ ID NO: 14, sdCDR2 of SEQ ID NO: 15, and sdCDR3 of SEQ ID NO: 16. In some cases, sdABD-EGFR has the amino acid sequence of SEQ ID NO: 13 as shown in A of FIG. 5.

[0143] In some embodiments, sdABD-EGFR (e.g., sdABD-hαEGFR2d) has sdCDR1 of SEQ ID NO: 18, sdCDR2 of SEQ ID NO: 19, and sdCDR3 of SEQ ID NO: 20. In some cases, sdABD-EGFR has the amino acid sequence of SEQ ID NO: 17, as shown in Figure 5A.

[0144] (b) EpCAM sdABD As shown in Figures 5D to E, there are many particularly useful sdABDs that bind to human EpCAM, referred to herein as "αEpCAM," "aEpCAM," "sdABD-EpCAM," "EpCAM sdABD," "EpCAM sdAb," "EpCAM ABD," or "EpCAMABD."

[0145] In some embodiments, sdABD-EpCAM (e.g., sdABD-EpCAM h13) has sdCDR1 of SEQ ID NO: 62, sdCDR2 of SEQ ID NO: 63, and sdCDR3 of SEQ ID NO: 64. In some cases, sdABD-EpCAM has the amino acid sequence of SEQ ID NO: 61, as shown in Figure 5D.

[0146] In some embodiments, sdABD-EpCAM (e.g., sdABD-EpCAM h23) has sdCDR1 of SEQ ID NO: 66, sdCDR2 of SEQ ID NO: 67, and sdCDR3 of SEQ ID NO: 68. In some cases, sdABD-EpCAM has the amino acid sequence of SEQ ID NO: 65, as shown in Figure 5D.

[0147] In some embodiments, sdABD-EpCAM (e.g., sdABD-EpCAM hVIB665) has sdCDR1 of SEQ ID NO: 70, sdCDR2 of SEQ ID NO: 71, and sdCDR3 of SEQ ID NO: 72. In some cases, sdABD-EpCAM has the amino acid sequence of SEQ ID NO: 69, as shown in Figure 5E. In contrast to h13 and h23 EpCAM sdABD, it should be noted that hVIB665 (also referred to as "acEpCAM hVIB665") binds to both cleaved and uncleaved forms of EpCAM (which are known to undergo cleavage in vivo).

[0148] In some embodiments, sdABD-EpCAM (e.g., sdABD-EpCAM hVIB666) has sdCDR1 of SEQ ID NO: 74, sdCDR2 of SEQ ID NO: 75, and sdCDR3 of SEQ ID NO: 76. In some cases, sdABD-EpCAM has the amino acid sequence of SEQ ID NO: 73, as shown in Figure 5E. In contrast to h13 and h23 EpCAM sdABD, it should be noted that hVIB666 (also referred to as "acEpCAM hVIB666") binds to both cleaved and uncleaved forms of EpCAM (which are known to undergo cleavage in vivo).

[0149] In some embodiments, sdABD-EpCAM (e.g., humanized EpCAM sdAb) has sdCDR1 of SEQ ID NO: 496, sdCDR2 of SEQ ID NO: 497, and sdCDR3 of SEQ ID NO: 498. In some cases, sdABD-EpCAM has the amino acid sequence of SEQ ID NO: 495, as shown in Figure 75.

[0150] (c)B7H3 sdABD As shown in Figures 5B to 5D, there are many particularly useful sdABDs that bind to human B7H3, referred to herein as "αB7H3", "aB7H3", "sdABD-B7H3", "B7H3 sdAb", "B7H3 ABD", "B7H3ABD", or "B7H3-ABD".

[0151] In one useful embodiment, sdABD-B7H3 (e.g., sdABD-B7H3 hF7) has sdCDR1 of SEQ ID NO: 34, sdCDR2 of SEQ ID NO: 35, and sdCDR3 of SEQ ID NO: 36. In some cases, sdABD-B7H3 has the amino acid sequence of SEQ ID NO: 33, as shown in Figure 5B.

[0152] In one useful embodiment, sdABD-B7H3 (e.g., sdABD-B7H3 hF12) has sdCDR1 of SEQ ID NO: 38, sdCDR2 of SEQ ID NO: 39, and sdCDR3 of SEQ ID NO: 40. In some cases, sdABD-B7H3 has the amino acid sequence of SEQ ID NO: 37, as shown in Figure 5C.

[0153] In one useful embodiment, sdABD-B7H3 (e.g., sdABD-B7H3 hF12(N57Q)) has sdCDR1 of SEQ ID NO: 42, sdCDR2 of SEQ ID NO: 43, and sdCDR3 of SEQ ID NO: 44. In some cases, sdABD-B7H3 has the amino acid sequence of SEQ ID NO: 41, as shown in Figure 5C. In contrast to hF7 and hF12 B7H3 sdABD, the amino acid substitution N57Q removes the glycosylation site.

[0154] In one useful embodiment, sdABD-B7H3 (e.g., sdABD-B7H3 HF12(N57E)) has sdCDR1 of SEQ ID NO: 46, sdCDR2 of SEQ ID NO: 47, and sdCDR3 of SEQ ID NO: 48. In some cases, sdABD-B7H3 has the amino acid sequence of SEQ ID NO: 45, as shown in Figure 5C. In contrast to hF7 and hF12 B7H3 sdABD, the amino acid substitution N57E removes the glycosylation site.

[0155] In one useful embodiment, sdABD-B7H3 (e.g., sdABD-B7H3 hF12(N57D)) has sdCDR1 of SEQ ID NO: 50, sdCDR2 of SEQ ID NO: 51, and sdCDR3 of SEQ ID NO: 52. In some cases, sdABD-B7H3 has the amino acid sequence of SEQ ID NO: 49, as shown in Figure 5B. In contrast to hF7 and hF12 B7H3 sdABD, the amino acid substitution N57D removes the glycosylation site.

[0156] In one useful embodiment, sdABD-B7H3 (e.g., sdABD-B7H3 hF12(S59A)) has sdCDR1 of SEQ ID NO: 54, sdCDR2 of SEQ ID NO: 55, and sdCDR3 of SEQ ID NO: 56. In some cases, sdABD-B7H3 has the amino acid sequence of SEQ ID NO: 53, as shown in Figure 5D. In contrast to hF7 and hF12 B7H3 sdABD, the amino acid substitution S59A removes the glycosylation site.

[0157] In one useful embodiment, sdABD-B7H3 (e.g., sdABD-B7H3 hF12(S59Y)) has sdCDR1 of SEQ ID NO: 58, sdCDR2 of SEQ ID NO: 59, and sdCDR3 of SEQ ID NO: 60. In some cases, sdABD-B7H3 has the amino acid sequence of SEQ ID NO: 57, as shown in Figure 5D. In contrast to hF7 and hF12 B7H3 sdABD, the amino acid substitution NS59Y removes the glycosylation site.

[0158] (d)FOLR1 sdABD As shown in Figure 5B, there are many particularly useful sdABDs that bind to human FOLR1, referred to herein as "αFOLR1", "aFOLR1", "sdABD-FOLR1", "FOLR1 sdAb", "sdABD FOLR1", "FOLR1 ABD", "FOLR1ABD", or "FOLR1-ABD".

[0159] In one useful embodiment, sdABD-FOLR1 (e.g., sdABD-FOLR1 h77-2) has sdCDR1 of SEQ ID NO: 22, sdCDR2 of SEQ ID NO: 23, and sdCDR3 of SEQ ID NO: 24. In some cases, sdABD-FOLR1 has the amino acid sequence of SEQ ID NO: 21, as shown in Figure 5B.

[0160] In one useful embodiment, sdABD-FOLR1 (e.g., sdABD-FOLR1 h59.3) has sdCDR1 of SEQ ID NO: 26, sdCDR2 of SEQ ID NO: 27, and sdCDR3 of SEQ ID NO: 28. In some cases, sdABD-FOLR1 has the amino acid sequence of SEQ ID NO: 25, as shown in Figure 5B.

[0161] In one useful embodiment, sdABD-FOLR1 (e.g., sdABD-FOLR1 h22-4) has sdCDR1 of SEQ ID NO: 30, sdCDR2 of SEQ ID NO: 31, and sdCDR3 of SEQ ID NO: 32. In some cases, sdABD-FOLR1 has the amino acid sequence of SEQ ID NO: 29, as shown in Figure 5B.

[0162] (e)Trop2 sdABD As shown in Figure 5E, there are many particularly useful sdABDs that bind to human Trop2, referred to herein as "αTrop2," "aTrop2," "sdABD-Trop2," "sdABD Trop2," "Trop2 sdAb," "Trop2ABD," or "Trop2-ABD."

[0163] In one useful embodiment, sdABD-Trop2 (e.g., sdABD-Trop2 hVIB557) has sdCDR1 of SEQ ID NO: 78, sdCDR2 of SEQ ID NO: 79, and sdCDR3 of SEQ ID NO: 80. In some cases, sdABD-Trop2 has the amino acid sequence of SEQ ID NO: 77, as shown in Figure 5E.

[0164] In some embodiments, sdABD-Trop2 (e.g., sdABD-Trop2 hVIB565) has sdCDR1 of SEQ ID NO: 82, sdCDR2 of SEQ ID NO: 83, and sdCDR3 of SEQ ID NO: 84. In some cases, sdABD-Trop2 has the amino acid sequence of SEQ ID NO: 81, as shown in Figure 5E.

[0165] In some embodiments, sdABD-Trop2 (e.g., sdABD-Trop2 hVIB575) has sdCDR1 of SEQ ID NO: 86, sdCDR2 of SEQ ID NO: 87, and sdCDR3 of SEQ ID NO: 88. In some cases, sdABD-Trop2 has the amino acid sequence of SEQ ID NO: 85, as shown in Figure 5F.

[0166] In some embodiments, sdABD-Trop2 (e.g., sdABD-Trop2 hVIB578) has sdCDR1 of SEQ ID NO: 90, sdCDR2 of SEQ ID NO: 91, and sdCDR3 of SEQ ID NO: 92. In some cases, sdABD-Trop2 has the amino acid sequence of SEQ ID NO: 89, as shown in Figure 5F.

[0167] In some embodiments, sdABD-Trop2 (e.g., sdABD-Trop2 hVIB609) has sdCDR1 of SEQ ID NO: 94, sdCDR2 of SEQ ID NO: 95, and sdCDR3 of SEQ ID NO: 96. In some cases, sdABD-Trop2 has the amino acid sequence of SEQ ID NO: 93, as shown in Figure 5F.

[0168] In some embodiments, sdABD-Trop2 (e.g., sdABD-Trop2 hVIB619) has sdCDR1 of SEQ ID NO: 98, sdCDR2 of SEQ ID NO: 99, and sdCDR3 of SEQ ID NO: 100. In some cases, sdABD-Trop2 has the amino acid sequence of SEQ ID NO: 97, as shown in Figure 5F.

[0169] (f)CA9 sdABD As shown in Figures 5F to G, there are many particularly useful sdABDs that bind to human CA9, referred to herein as "αCA9," "aCA9," "sdABD-CA9," "sdABD CA9," "CA9 sdAb," "CA9 ABD," or "CA9-ABD."

[0170] In some embodiments, sdABD-CA9 (e.g., sdABD-CA9 hVIB456) has sdCDR1 of SEQ ID NO: 102, sdCDR2 of SEQ ID NO: 103, and sdCDR3 of SEQ ID NO: 104. In some cases, sdABD-Trop2 has the amino acid sequence of SEQ ID NO: 101, as shown in Figure 5F.

[0171] In some embodiments, sdABD-CA9 (e.g., sdABD-CA9 hVIB476) has sdCDR1 of SEQ ID NO: 106, sdCDR2 of SEQ ID NO: 107, and sdCDR3 of SEQ ID NO: 108. In some cases, sdABD-Trop2 has the amino acid sequence of SEQ ID NO: 105, as shown in Figure 5G.

[0172] In some embodiments, sdABD-CA9 (e.g., sdABD-CA9 hVIB407) has sdCDR1 of SEQ ID NO: 110, sdCDR2 of SEQ ID NO: 111, and sdCDR3 of SEQ ID NO: 112. In some cases, sdABD-Trop2 has the amino acid sequence of SEQ ID NO: 109, as shown in Figure 5G.

[0173] In some embodiments, sdABD-CA9 (e.g., sdABD-CA9 hVIB445) has sdCDR1 of SEQ ID NO: 114, sdCDR2 of SEQ ID NO: 115, and sdCDR3 of SEQ ID NO: 116. In some cases, sdABD-Trop2 has the amino acid sequence of SEQ ID NO: 113, as shown in Figure 5G.

[0174] (g)LyPD3 sdABD As shown in Figures 5G-H, there are many particularly useful sdABDs that bind to human LyPD3, referred to herein as "αLyPD3," "sdABD-LyPD3," "sdABDs LyPD3," "LyPD3 sdAb," "LyPD3 ABD," "LyPD3ABD," or "LyPD3-ABD."

[0175] In one useful embodiment, sdABD-LyPD3 (e.g., sdABD-LyPD3 h787) has sdCDR1 of SEQ ID NO: 118, sdCDR2 of SEQ ID NO: 119, and sdCDR3 of SEQ ID NO: 120. In some cases, sdABD-LyPD3 has the amino acid sequence of SEQ ID NO: 117, as shown in Figure 5G.

[0176] In one useful embodiment, sdABD-LyPD3 (e.g., sdABD-LyPD3 h790) has sdCDR1 of SEQ ID NO: 122, sdCDR2 of SEQ ID NO: 123, and sdCDR3 of SEQ ID NO: 124. In some cases, sdABD-LyPD3 has the amino acid sequence of SEQ ID NO: 121, as shown in Figure 5G.

[0177] In one useful embodiment, sdABD-LyPD3 (e.g., sdABD-LyPD3 H804) has sdCDR1 of SEQ ID NO: 126, sdCDR2 of SEQ ID NO: 127, and sdCDR3 of SEQ ID NO: 128. In some cases, sdABD-LyPD3 has the amino acid sequence of SEQ ID NO: 125, as shown in Figure 5H.

[0178] In one useful embodiment, sdABD-LyPD3 (e.g., sdABD-LyPD3 h773) has sdCDR1 of SEQ ID NO: 130, sdCDR2 of SEQ ID NO: 131, and sdCDR3 of SEQ ID NO: 132. In some cases, sdABD-LyPD3 has the amino acid sequence of SEQ ID NO: 129, as shown in Figure 5H.

[0179] In one useful embodiment, sdABD-LyPD3 (e.g., sdABD-LyPD3 h840) has sdCDR1 of SEQ ID NO: 134, sdCDR2 of SEQ ID NO: 135, and sdCDR3 of SEQ ID NO: 136. In some cases, sdABD-LyPD3 has the amino acid sequence of SEQ ID NO: 133, as shown in Figure 5H.

[0180] In one useful embodiment, sdABD-LyPD3 (e.g., sdABD-LyPD3 h885) has sdCDR1 of SEQ ID NO: 138, sdCDR2 of SEQ ID NO: 139, and sdCDR3 of SEQ ID NO: 140. In some cases, sdABD-LyPD3 has the amino acid sequence of SEQ ID NO: 137, as shown in Figure 5H.

[0181] (h)HER2 sdABD As shown in Figures 5H-M, there are many particularly useful sdABDs that bind to human HER2, referred to herein as "αHER2," "aHER2," "sdABD-HER2," "sdABD HER2," "HER2 sdAb," "HER2 ABD," "HER2ABD," or "HER2-ABD."

[0182] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1054) has sdCDR1 of SEQ ID NO: 142, sdCDR2 of SEQ ID NO: 143, and sdCDR3 of SEQ ID NO: 144. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 141, as shown in Figure 5H.

[0183] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1055) has sdCDR1 of SEQ ID NO: 146, sdCDR2 of SEQ ID NO: 147, and sdCDR3 of SEQ ID NO: 148. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 145, as shown in Figure 5I.

[0184] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1058) has sdCDR1 of SEQ ID NO: 150, sdCDR2 of SEQ ID NO: 151, and sdCDR3 of SEQ ID NO: 153. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 149, as shown in Figure 5I.

[0185] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1059) has sdCDR1 of SEQ ID NO: 154, sdCDR2 of SEQ ID NO: 155, and sdCDR3 of SEQ ID NO: 156. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 153, as shown in Figure 5I.

[0186] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1065) has sdCDR1 of SEQ ID NO: 158, sdCDR2 of SEQ ID NO: 159, and sdCDR3 of SEQ ID NO: 160. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 157, as shown in Figure 5I.

[0187] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1090) has sdCDR1 of SEQ ID NO: 162, sdCDR2 of SEQ ID NO: 163, and sdCDR3 of SEQ ID NO: 164. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 161, as shown in Figure 5I.

[0188] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1191) has sdCDR1 of SEQ ID NO: 166, sdCDR2 of SEQ ID NO: 167, and sdCDR3 of SEQ ID NO: 168. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 165, as shown in J of Figure 5.

[0189] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1092) has sdCDR1 of SEQ ID NO: 170, sdCDR2 of SEQ ID NO: 171, and sdCDR3 of SEQ ID NO: 172. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 169, as shown in J of Figure 5.

[0190] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1097) has sdCDR1 of SEQ ID NO: 174, sdCDR2 of SEQ ID NO: 175, and sdCDR3 of SEQ ID NO: 176. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 173, as shown in J of Figure 5.

[0191] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1118) has sdCDR1 of SEQ ID NO: 178, sdCDR2 of SEQ ID NO: 179, and sdCDR3 of SEQ ID NO: 180. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 177, as shown in J of Figure 5.

[0192] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1121) has sdCDR1 of SEQ ID NO: 182, sdCDR2 of SEQ ID NO: 183, and sdCDR3 of SEQ ID NO: 184. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 181, as shown in J of Figure 5.

[0193] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1134) has sdCDR1 of SEQ ID NO: 186, sdCDR2 of SEQ ID NO: 187, and sdCDR3 of SEQ ID NO: 188. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 185, as shown in K of Figure 5.

[0194] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1138) has sdCDR1 of SEQ ID NO: 190, sdCDR2 of SEQ ID NO: 191, and sdCDR3 of SEQ ID NO: 192. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 189, as shown in K of Figure 5.

[0195] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1139) has sdCDR1 of SEQ ID NO: 194, sdCDR2 of SEQ ID NO: 195, and sdCDR3 of SEQ ID NO: 196. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 193, as shown in K of Figure 5.

[0196] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1140) has sdCDR1 of SEQ ID NO: 198, sdCDR2 of SEQ ID NO: 199, and sdCDR3 of SEQ ID NO: 200. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 197, as shown in K of Figure 5.

[0197] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1145) has sdCDR1 of SEQ ID NO: 202, sdCDR2 of SEQ ID NO: 203, and sdCDR3 of SEQ ID NO: 204. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 201, as shown in K of Figure 5.

[0198] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1146) has sdCDR1 of SEQ ID NO: 206, sdCDR2 of SEQ ID NO: 207, and sdCDR3 of SEQ ID NO: 203. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 205, as shown in L of Figure 5.

[0199] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1149) has sdCDR1 of SEQ ID NO: 210, sdCDR2 of SEQ ID NO: 211, and sdCDR3 of SEQ ID NO: 212. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 209, as shown in L of Figure 5.

[0200] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1150) has sdCDR1 of SEQ ID NO: 214, sdCDR2 of SEQ ID NO: 215, and sdCDR3 of SEQ ID NO: 216. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 213, as shown in L of Figure 5.

[0201] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1156) has sdCDR1 of SEQ ID NO: 218, sdCDR2 of SEQ ID NO: 219, and sdCDR3 of SEQ ID NO: 220. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 217, as shown in L of Figure 5.

[0202] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1158) has sdCDR1 of SEQ ID NO: 222, sdCDR2 of SEQ ID NO: 223, and sdCDR3 of SEQ ID NO: 224. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 221, as shown in L of Figure 5.

[0203] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1159) has sdCDR1 of SEQ ID NO: 226, sdCDR2 of SEQ ID NO: 227, and sdCDR3 of SEQ ID NO: 228. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 225, as shown in M ​​of Figure 5.

[0204] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1160) has sdCDR1 of SEQ ID NO: 230, sdCDR2 of SEQ ID NO: 231, and sdCDR3 of SEQ ID NO: 232. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 229, as shown in M ​​of Figure 5.

[0205] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1161) has sdCDR1 of SEQ ID NO: 234, sdCDR2 of SEQ ID NO: 235, and sdCDR3 of SEQ ID NO: 236. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 233, as shown in M ​​in Figure 5.

[0206] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1162) has sdCDR1 of SEQ ID NO: 238, sdCDR2 of SEQ ID NO: 239, and sdCDR3 of SEQ ID NO: 240. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 237, as shown in M ​​of Figure 5.

[0207] In some embodiments, sdABD-HER2 (e.g., sdABD-HER2 1163) has sdCDR1 of SEQ ID NO: 242, sdCDR2 of SEQ ID NO: 243, and sdCDR3 of SEQ ID NO: 244. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 241, as shown in M ​​of Figure 5.

[0208] In some embodiments, sdABD-HER2 (e.g., humanized aHER2 sdAb h1130) has sdCDR1 of SEQ ID NO: 500, sdCDR2 of SEQ ID NO: 501, and sdCDR3 of SEQ ID NO: 502. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 499, as shown in Figure 75.

[0209] In some embodiments, sdABD-HER2 (e.g., humanized aHER2 sdAb h1156) has sdCDR1 of SEQ ID NO: 504, sdCDR2 of SEQ ID NO: 505, and sdCDR3 of SEQ ID NO: 506. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 503, as shown in Figure 75. Epitope mapping revealed that humanized aHER2 sdAb h1156 binds to amino acid sequence WK at amino acid positions 147-148, amino acid sequence LALTL (SEQ ID NO: 515) at amino acid positions 157-161, and amino acid sequence TRTVC (SEQ ID NO: 516) at amino acid positions 194-198 of the HER2 protein (Figure 36).

[0210] In some embodiments, sdABD-HER2 (e.g., humanized aHER2 sdAb h1159) has sdCDR1 of SEQ ID NO: 508, sdCDR2 of SEQ ID NO: 509, and sdCDR3 of SEQ ID NO: 510. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 507, as shown in Figure 75.

[0211] In some embodiments, sdABD-HER2 (e.g., humanized aHER2 sdAb h1162) has sdCDR1 of SEQ ID NO: 512, sdCDR2 of SEQ ID NO: 513, and sdCDR3 of SEQ ID NO: 514. In some cases, sdABD-HER2 has the amino acid sequence of SEQ ID NO: 511, as shown in Figure 75. Epitope mapping revealed that humanized aHER2 sdAb h1162 binds to the amino acid sequence QLTFRNPHQALL at amino acid positions 462-472 of the HER2 protein (Figure 36).

[0212] In some embodiments, the protein before cleavage of the protease cleavage domain is less than approximately 100 kDa. In some embodiments, the protein after cleavage of the protease cleavage domain is approximately 25 to approximately 75 kDa. In some embodiments, the protein before cleavage has a size exceeding the renal threshold for first-pass clearance. In some embodiments, the protein before cleavage has an elimination half-life of at least approximately 50 hours. In some embodiments, the protein before cleavage has an elimination half-life of at least approximately 100 hours. In some embodiments, the protein has increased tissue penetration compared to IgG for the same target antigen. In some embodiments, the protein has increased tissue distribution compared to IgG for the same target antigen.

[0213] C. Half-life extension domain The MCE protein of the present invention (which in this case is also referred to herein as the "COBRA®" protein or construct) optionally comprises a half-life extension domain. Such domains include, but are not limited to, HSA-binding domains, Fc domains, small molecules, and other half-life extension domains known in the art.

[0214] Human serum albumin (HSA) (molecular mass approximately 67 kDa) is the most abundant protein in plasma, present at approximately 50 mg / mL (600 μM), and has a half-life of approximately 20 days in humans. HSA plays a role in maintaining plasma pH, contributing to colloidal blood pressure, functioning as a carrier for many metabolites and fatty acids, and acting as a major drug transport protein in plasma.

[0215] Non-covalent association with albumin extends the elimination half-life of short-lived proteins. For example, recombinant fusion of the albumin-binding domain to a Fab fragment resulted in a 25–58-fold reduction in in vivo clearance and a 26–37-fold extension of half-life when administered intravenously to mice and rabbits, respectively, compared to administration of the Fab fragment alone. In another example, when insulin was acylated with a fatty acid to promote association with albumin, a sustained effect was observed when administered subcutaneously to rabbits or pigs. Together, these experiments demonstrated a link between albumin binding and sustained action.

[0216] In many embodiments, the half-life extension 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 "sdABD(1 / 2)" to distinguish these binding domains from sdABD for TTA. Particularly useful sdABD(1 / 2) are shown in Figure 6.

[0217] In some embodiments, sdABD-HSA (e.g., sdABD-HSA(10GE)) has sdCDR1 of SEQ ID NO: 246, sdCDR2 of SEQ ID NO: 247, and sdCDR3 of SEQ ID NO: 248. In some embodiments, sdABD-HSA has the amino acid sequence of SEQ ID NO: 245. In certain embodiments, sdABD-HSA (e.g., sdABD-HSA with a histidine (His) tag) has sdCDR1 of SEQ ID NO: 250, sdCDR2 of SEQ ID NO: 251, and sdCDR3 of SEQ ID NO: 252. In some embodiments, sdABD-HSA has the amino acid sequence of SEQ ID NO: 249.

[0218] The half-life extension domain of an antigen-binding protein provides modified pharmacodynamics and pharmacokinetics of the antigen-binding protein itself. As described above, the half-life extension domain extends the elimination half-life. The half-life extension domain also modifies the pharmacological properties of the antigen-binding protein, including modifications to its tissue distribution, penetration, and diffusion. In some embodiments, the half-life extension domain provides improved tissue (including tumor) targeting, tissue penetration, tissue distribution, diffusion within tissue, and enhanced efficacy compared to proteins without the half-life extension domain. In one embodiment, the therapy effectively and efficiently utilizes a reduced amount of antigen-binding protein, resulting in reduced side effects, such as reduced cytotoxicity to non-tumor cells.

[0219] Furthermore, the characteristics of half-life extension domains, such as HSA-binding domains, include the binding affinity of the HSA-binding domain to HSA. The aforementioned affinity of the HSA-binding domain may be selected to target a specific elimination half-life in a particular polypeptide construct. Thus, in some embodiments, the HSA-binding domain has high binding affinity. In other embodiments, the HSA-binding domain has moderate binding affinity. In yet another embodiment, the HSA-binding domain has low or slight binding affinity. Exemplary binding affinities include KD concentrations of ≤10 nM (high), 10 nM to 100 nM (medium), and >100 nM (low). As described above, the binding affinity to HSA is determined by known methods such as surface plasmon resonance (SPR).

[0220] D. Protease cleavage site The protein compositions of the present invention, in particular the prodrug constructs, generally contain one or more protease cleavage sites located within a cleavable linker, as outlined herein.

[0221] As described herein, the prodrug constructs of the present invention include at least one protease cleavage site comprising an amino acid sequence cleaved by at least one protease. In some cases, the MCE proteins described herein include 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. As will be more fully considered 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., and for example, some constructs may utilize three sites for two different proteases, etc.

[0222] The amino acid sequence of the protease cleavage site is determined by the target protease. As is well known in this field, there are many human proteases that exist in the body and may be involved in disease states.

[0223] Proteases are known to be secreted by certain diseased cells and tissues, such as tumor cells or cancer cells that generate a protease-rich microenvironment. In some cases, the target blood is protease-rich. In some cases, cells surrounding the 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 are present in the target blood; for example, proteases that target amino acid sequences are present in bacterial peptides. Except for the protease-rich microenvironment of target cells or tissues, antigen-binding proteins do not bind to T cells. This characteristic allows for additional specificity of targeted therapeutic agents such as antigen-binding proteins.

[0224] Proteases are proteins that, in some cases, cleave other proteins in a sequence-specific manner. Examples of proteases include serine protease, cysteine ​​protease, aspartate protease, threonine protease, glutamate protease, metalloprotease, asparagine peptide lyase, serum protease, cathepsin (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, stromericin, factor XA, chymotrypsin-like protease, trypsin-like protease, elastase-like protease, subtilisin-like protease, actinidine, bromelain, calpain, and caspase (e.g., caspase 3). Examples include, but are not limited to, Mir1-CP, papain, HIV-1 protease, HSV protease, CMV protease, chymosin, renin, pepsin, matryptase, regmine, plasmmepsin, nepenthesin, metalloexopeptidase, metalloendopeptidase, matrix metalloproteinase (MMP), MMP1, MMP2, MMP3, MMP8, MMP9, MMP13, MMP11, MMP14, meprine, 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).

[0225] Some suitable proteases and protease cleavage sequences are shown in Figure 8A-D. In some embodiments, one of the fusion proteins described herein includes a cleavage linker containing a protease cleavage domain sequence shown in any one of SEQ ID NOs: 339-408 and SEQ ID NOs: 532-535.

[0226] E. Linker As discussed herein, the different domains of the present invention are generally linked together using amino acid linkers, which can provide functionality including flexibility or inflexibility (e.g., steric hindrance), as well as the ability to be cleaved using in situ proteases. These linkers can be classified in many ways.

[0227] The present invention provides “domain linkers” used to link two or more domains (e.g., VH and VL, a target tumor antigen binding domain to VH or VL (TTABD, sometimes referred to herein as “αTTA” (as opposed to “anti-TTA”)), a half-life extension domain to another component, etc.). Domain linkers may be, for example, non-cleavable (NCL), cleavable ("CL"), restrictive and cleavable (CCL), and restrictive and non-cleavable (CNCL).

[0228] 1. Non-fissurable linker In some embodiments, domain linkers are non-cleavable. Generally, these can be one of two types: non-cleavable and flexible, allowing the "upstream" and "downstream" components of the construct's linker to self-assemble within the molecule in a particular manner, or non-cleavable and restrictive, preventing the two components separated by the linker from self-assembling within the molecule. However, in the latter case, it should be noted that while the two component domains separated by the non-cleavable restrictive linker do not self-assemble within the molecule, other intramolecular components do self-assemble to form a pseudo-Fv domain.

[0229] (a) Non-cleaving but flexible linker In this embodiment, the linker is generally used to bind to a domain and maintain its functionality by having a longer, flexible domain that is not cleaved by situ proteases in the patient. Examples of internal non-cleavable linkers suitable for linking to a domain in the polypeptide of the present invention include, but are not limited to, (GS)n, (GGS)n, (GGGS)n [SEQ ID NO: 518], (GGSG)n [SEQ ID NO: 519], (GGSGG)n [SEQ ID NO: 520], or (GGGGS)n [SEQ ID NO: 521], where n is 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10. In some embodiments, the linker length may be about 15 amino acids.

[0230] (b) Non-cleaving and restrictive linker In some cases, the linker is a "restrictive 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 within the molecule. For example, in Pro186, the active VH and active VL are separated by an 8-amino acid ("8-mer" or "8mer") which does not allow VH and VL to self-assemble into the active antigen-binding domain. In some embodiments, the linker is still flexible, e.g., (GGGS)n (n=2). In other embodiments, although generally less preferred, a more rigid linker may be used, such as one containing proline or a bulky amino acid.

[0231] 2. Splitting linker All prodrug constructs herein contain at least one cleavage linker. Thus, in one embodiment, the domain linker is cleavage (CL) and is sometimes referred to herein as a “protease cleavage domain” (“PCD”). In this embodiment, the CL contains a protease cleavage site as outlined herein and shown in Figures 8A–D. In some cases, the CL contains only a protease cleavage site. Optionally, depending on the length of the cleavage recognition site, several additional linked amino acids may be present at either the N-terminus or C-terminus or both of the CL; for example, there may be 1, 2, 3, 4, or 5 amino acids at either the N-terminus or C-terminus or both of the cleavage site. Thus, cleavage linkers can also be constrained (e.g., 8mer) or flexible.

[0232] Of particular interest in this invention are the MMP9 cleavage linker and the meprin cleavage linker, and especially the MMP9-constrained cleavage linker and the meprin-constrained cleavage linker.

[0233] II. Domain of the Invention The present invention provides many different forms of the prodrug polypeptide of the present invention. The present invention provides restrictive Fv domains and restrictive pseudo-Fv domains. In addition, the present invention provides polyvalent conditionally effective ("MCE") proteins that contain two Fv domains but are non-isomerized constructs. As outlined herein, these may be non-isomerized cleavable or non-isomerized non-cleavable forms, but all constructs contain at least one protease cleavage domain.

[0234] Importantly, both of these domains (Fv domain and pseudo-Fv domain) are referred to as “constrained” in this specification, meaning that only one of them needs to be constrained, as discussed above and as shown in Figures 1-5. However, generally, when both linkers are constrained, the protein exhibits better expression.

[0235] Those skilled in the art will understand, with respect to forms 1, 2, and 4, the limitations of the present invention and the four possibilities for the N-terminus-C-terminus order of the pseudo-Fv domain (without indicating the linker): aVH-aVL and iVL-iVH, aVH-aVL and iVH-iVL, aVL-aVH and iVL-iVH, and aVL-aVH and iVH-iVH. All four have been tested and all four are active, but the first order, aVH-aVL and iVL-iVH, shows better expression than the other three. Therefore, the description herein is generally given in this aVH-aVL and iVL-iVH form, but all disclosure herein also includes other orders of these domains.

[0236] In general, it should be noted that the N-terminus to C-terminus order of the full-length constructs of the present invention is based on aVH-aVL and iVL-iVH orientations.

[0237] In addition, it is known in the art that immunogenicity in humans may exist from the C-terminal sequences of specific ABDs. Therefore, histidine tags (either His6 or His10) may be used in general, especially when the C-terminus of a construct is terminated with sdABD (e.g., many of the sdABD-HSA domains in the constructs). Many or most of the sequences herein were generated using the His6 C-terminal tag for purification reasons, but these sequences may also be used to reduce immunogenicity in humans, as shown by Holland et al., DOI10.1007 / s10875-013-9915-0 and International Publication WO2013 / 024059.

[0238] A. Constraining Fv domain The present invention provides a restrictive Fv domain comprising active VH and active VL domains covalently linked using a restrictive linker (which may be cleavable (forms 1 and 3) or non-cleavable (forms 2 and 4), as outlined herein). The restrictive linker prevents intramolecular association between VH and VL in the absence of cleavage. Thus, a restrictive Fv domain generally comprises a set of six CDRs contained within a variable domain, where vhCDR1, vhCDR2, and vhCDR3 of VH bind to human CD3, and vlCDR1, vCDR2, and vlCDR3 of VL bind to human CD3; however, in prodrug form (e.g., non-cleavable), VH and VL cannot sterically associate to form an active binding domain and rather prefer to pair intramolecularly with a pseudo-Fv.

[0239] The restrictive Fv domain may include active VH and active VL (aVH and aVL) or inactive VH and VL (iVH and iVL, in which case it is a restrictive pseudo-Fv domain), as described herein, or a combination thereof.

[0240] As will be understood by those skilled in the art, the order of VH and VL in the restrictive Fv domain can be either VH-linker-VL or VL-linker-VH (from N-terminus to C-terminus).

[0241] As outlined herein, with respect to a Form 1 construct, the restrictive Fv domain may include VH and VL linked using a cleavable linker, as shown in Figure 1. In this embodiment, the restrictive Fv domain has the structure (from N-terminus to C-terminus) vhFR1-vhCDR1-vhFR2-vhCDR2-vhFR3-vhCDR3-vhFR4-CCL-vlFR1-vlCDR1-vlFR2-vlCDR2-vlFR3-vlCDR3-vlFR4. Generally, restrictive Fv domains contain active VH and VL domains (for example, capable of binding to CD3 when associated), and therefore have the structure (from N-terminus to C-terminus) vhFR1-avhCDR1-vhFR2-avhCDR2-vhFR3-avhCDR3-vhFR4-CCL-vlFR1-avlCDR1-vlFR2-avlCDR2-vlFR3-avlCDR3-vlFR4.

[0242] As outlined herein, with respect to a Type 2 construct, the restrictive Fv domain may include VH and VL linked using a non-cleavable linker. In this embodiment, the restrictive Fv domain has the structure (from N-terminus to C-terminus) vhFR1-vhCDR1-vhFR2-vhCDR2-vhFR3-vhCDR3-vhFR4-CNCL-vlFR1-vlCDR1-vlFR2-vlCDR2-vlFR3-vlCDR3-vlFR4. Generally, restrictive Fv domains contain active VH and VL domains (for example, capable of binding to CD3 when associated), and therefore have the structure (from N-terminus to C-terminus) vhFR1-avhCDR1-vhFR2-avhCDR2-vhFR3-avhCDR3-vhFR4-CNCL-vlFR1-avlCDR1-vlFR2-avlCDR2-vlFR3-avlCDR3-vlFR4.

[0243] A constrained non-cleavable Fv domain having a VH having SEQ ID NO: 270, a VL having SEQ ID NO: 254, and a domain linker having SEQ ID NO: 287 is particularly used in the present invention.

[0244] B. Constraining pseudo-Fv domain The present invention provides a constrained pseudo-Fv domain comprising inactive or pseudo-iVH and iVL domains covalently linked using a constrained linker (which may be cleavable or non-cleavable, as outlined herein). The constrained linker prevents intramolecular association between iVH and iVL in the absence of cleavage. Thus, the constrained pseudo-Fv domain generally comprises iVH and iVL having a framework region that allows for the association of iVH and iVL (when in an 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 to CD3.

[0245] The restrictive pseudo-Fv domain includes inactive VH and VL (iVH and iVL).

[0246] As will be understood by those skilled in the art, the order of VH and VL in the restrictive pseudo-Fv domain can be either VH-linker-VL or VL-linker-VH (from N-terminus to C-terminus).

[0247] As outlined herein, restrictive pseudo-Fv domains may include iVHs and iVLs linked using non-cleavable linkers such as those shown in Forms 1, 2, and 4, or using cleavable linkers such as those shown in Form 3.

[0248] Generally, restrictive Fv domains contain inactive VH and VL domains (for example, capable of binding to CD3 when associated), and therefore have the structure (from N-terminus to C-terminus) vhFR1-ivlCDR1-vhFR2-ivlCDR2-vhFR3-ivlCDR3-vhFR4-CNCL-vlFR1-ivhCDR1-vlFR2-ivhCDR2-vlFR3-ivhCDR3-vlFR4.

[0249] Particularly used in the present invention are constrained non-cleavable pseudo-Fv domains having (i) an iVH having sequence number 274 (αCD3 VHi), sequence number 278 (αCD3 VHi2), or sequence number 282 (αCD3 VHiGL4), (ii) an iVL having sequence number 258 (αCD3 VLi), sequence number 262 (αCD3 VLi2), or sequence number 266 (αCD3 VLiGL), and (iii) a domain linker having sequence number 287. In some embodiments, the constrained non-cleavable pseudo-Fv domain includes (i) an iVH (αCD3 VHi) having the amino acid sequence of sequence number 274, (ii) an iVL (αCD3 VLi) having the amino acid sequence of sequence number 258, and (iii) a domain linker having the amino acid sequence of sequence number 287. In some embodiments, the restrictive non-cleavable pseudo-Fv domain includes (i) iVH(αCD3 VHi2) having the amino acid sequence of SEQ ID NO: 278, (ii) iVL(αCD3 VLi2) having the amino acid sequence of SEQ ID NO: 262, and (iii) a domain linker having the amino acid sequence of SEQ ID NO: 287. In some embodiments, the restrictive non-cleavable pseudo-Fv domain includes (i) iVH(αCD3 VHi2GL4) having the amino acid sequence of SEQ ID NO: 282, (ii) iVL(αCD3 VLi2GL) having the amino acid sequence of SEQ ID NO: 266, and (iii) a domain linker having the amino acid sequence of SEQ ID NO: 287.

[0250] III. Forms of the present invention As discussed herein, the prodrug constructs of the present invention can take many different forms, including a cleavable form having a double TTA-binding domain, a non-cleavable form having a double TTA-binding domain (one of which may have the same or different TTA-binding domains), and a non-cleavable form having a single targeting domain.

[0251] A. "Form 2" structures As shown in Figure 2, the present invention provides a non-isomerized non-cleavable form. In this embodiment, since an activated cleavage site is present in the prodrug construct, it is understood that "non-cleavable" applies only to the linking of restrictive Fv domains. In this embodiment, the restrictive Fv domain includes VH and VL domains linked using a restrictive non-cleavable linker, and the restrictive pseudo-Fv domain uses a restrictive non-cleavable linker.

[0252] As will be understood by those skilled in the art, the order of VH and VL in either a restrictive Fv domain or a restrictive pseudo-Fv domain can be either VH-linker-VL or VL-linker-VH (from N-terminus to C-terminus).

[0253] The present invention provides a prodrug protein comprising (sdABD-TTA1)-domain linker-constraining Fv domain-domain linker-(sdABD-TTA2)-cleavable linker-constraining pseudo-Fv domain-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus.

[0254] As will be understood by those skilled in the art, the order of VH and VL in either a restrictive Fv domain or a restrictive pseudo-Fv domain can be either VH-linker-VL or VL-linker-VH (from N-terminus to C-terminus).

[0255] Therefore, in one embodiment, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus.

[0256] Therefore, in one embodiment, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVH-CNCL-iVL-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus.

[0257] Therefore, in one embodiment, the prodrug protein contains (sdABD-TTA1)-domain linker-aVL-CNCL-aVH-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus.

[0258] Therefore, in one embodiment, the prodrug protein contains (sdABD-TTA1)-domain linker-aVL-CNCL-aVH-domain linker-(sdABD-TTA2)-CL-iVH-CNCL-iVL-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus.

[0259] In some embodiments, the prodrug proteins described herein are provided in corresponding sequences shown as Figures including C-V in Figure 9 and sequence numbers 413-452, representing exemplary proteins Pro225, Pro226, Pro233, Pro311, Pro312, Pro313, Pro246, Pro256, Pro420, Pro421, Pro393, Pro394, Pro395, Pro396, Pro429, Pro430, Pro431, Pro258, Pro221, Pro222, Pro223, Pro224, Pro254, Pro255, Pro262, Pro356, Pro359, Pro364, Pro388, Pro429, Pro430, Pro431, Pro432, Pro448, Pro449, Pro450, Pro451, Pro479, Pro480, and Pro495. In some embodiments, the prodrug proteins described herein are (i) the modified proteins Pro601, Pro602, V3, V4, Pro664, Pro665, Pro667, Pro694, Pro695, Pro565, Pro566, Pro567, Pro727, Pro728, Pro729, Pro730, Pro731, Pro676, Pro677, Pro678, Pro679, Pro808, Pro819, Pro621, Pro622, Pro640, Pro641, Pro642, Pro643, Pro744, Pro746, Pro108, (ii) Exemplary protein aLyPD3 h787 COBRA, aLyPD3 h790 COBRA, aLyPD3 h804 COBRA, aLyPD3 h773 COBRA, aLyPD3 h840 COBRA, aLyPD3 h885 COBRA, aHER21054 COBRA, aHER2 1055 COBRA, aHER2 1058 COBRA, aHER2 COBRA, aHER2 105 1090 COBRA, aHER2 1091 COBRA, aHER2 1092 COBRA, aHER2 1097 COBRA, aHER2 1118 COBRA, aHER2 1121 COBRA, aHER2 1134 COBRA, aHER2 1138 COBRA, aHER2 1139 COBRA, aHER2 1140 COBRA, aHER2 1145 COBRA, aHER2 1150 COBRA, aHER2 1156 COBRA, aHER2 1158 COBRA, aHER2 1159 COBRA, aHER2 1160 COBRA, aHER2 1161 (iii) The figures are provided with the following: (iii) the corresponding sequences shown as A-Q in Figure 12 and sequence numbers 453-486, representing COBRA, aHER2 1162 COBRA, aHER2 1163 COBRA, Pro824 and Pro826; (iii) the sequences shown as Figures 71-74 and corresponding sequence numbers 487-494, representing exemplary proteins Pro751, Pro752, Pro824, Pro826, Pro1109, Pro1111, Pro1117 and Pro1118.

[0260] 1. Single-targeting type 2 construct "Single-specific COBRA" In some embodiments, both αTTA domains bind to the same tumor target antigen (TTA). Therefore, in some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In this embodiment, aVH, aVL, iVH, and iVL have sequences shown in Figure 7A-B. In this embodiment, the two targeting domains bind to the same TTA, which may be EGFR, EpCAM, FOLR1, Trop2, CA9, B7H3, LyPD3, or HER2, whose sequences are shown in Figure 5A-M.

[0261] In some embodiments, sdABD-TTA1 is selected from the group consisting of sdABD-B7H3, sdABD-CA9, sdABD-EGFR, sdABD-EpCAM, sdABD-FOLR1, sdABD-HER2, sdABD-LyPD3, and sdABD-Trop2. In some embodiments, sdABD-TTA2 is selected from the group consisting of sdABD-B7H3, sdABD-CA9, sdABD-EGFR, sdABD-EpCAM, sdABD-FOLR1, sdABD-HER2, sdABD-LyPD3, and sdABD-Trop2. In some embodiments, sdABD-TTA1 and sdABD-TTA2 bind to the same target antigen. In some embodiments, sdABD-TTA1 and sdABD-TTA2 bind to the same target antigen, but at different positions. In some embodiments, sdABD-TTA1 and sdABD-TTA2 bind to the same target antigen at the same position. In some embodiments, sdABD-TTA1 and sdABD-TTA2 have the same amino acid sequence. Any sequence of sdABD described herein may be the sequence of sdABD-TTA1, sdABD-TTA2, or both. In some embodiments, sdCDR1, sdCDR2, and sdCDR3 of sdABD-TTA1 are the same as sdCDR1, sdCDR2, and sdCDR3 of sdABD-TTA2, respectively.

[0262] In some embodiments, exemplary monospecific COBRAs (also referred to as monotumor antigen-targeted COBRAs) bind to tumor target antigens selected from the group consisting of B7H3, CA9, EGFR, EpCAM, FOLR1, HER2, LyPD3, and Trop2. In some embodiments, the monospecific COBRA comprises one of the sequences of sdABD provided herein, including the figure and formal sequence listing. In some embodiments, the monospecific COBRA that binds to B7H3 (e.g., human B7H3) is Pro225 (SEQ ID NO: 413) and Pro226 (SEQ ID NO: 414), as shown in Figure 9 C; Pro601 (SEQ ID NO: 522), Pro602 (SEQ ID NO: 523), V3 (SEQ ID NO: 524), V4 (SEQ ID NO: 525), Pro664 (SEQ ID NO: 526), ​​Pro665 (SEQ ID NO: 527), Pro667 (SEQ ID NO: 528), Pro694 (SEQ ID NO: 528). It includes any of the fusion proteins in Figure 10 A-E, such as 529) and Pro695 (SEQ ID NO: 530); O-Q in Figure 10, such as Pro640 (SEQ ID NO: 306), Pro641 (SEQ ID NO: 307), Pro642 (SEQ ID NO: 308), Pro643 (SEQ ID NO: 309), Pro774 (SEQ ID NO: 310), and Pro746 (SEQ ID NO: 311); and DD-EE in Figure 10, such as Pro225 (SEQ ID NO: 336) and Pro817 (SEQ ID NO: 338).

[0263] In some embodiments, the single-specific COBRA that binds to CA9 (e.g., human CA9) includes any of the fusion proteins Z-BB in Figure 10, such as Pro516 (SEQ ID NO: 329), Pro517 (SEQ ID NO: 330), Pro518 (SEQ ID NO: 331), and Pro519 (SEQ ID NO: 332).

[0264] In some embodiments, the single-specific COBRA that binds to EGFR (e.g., human EGFR) includes any of the S-T fusion proteins in Figure 10, such as Pro396 (SEQ ID NO: 314), Pro476 (SEQ ID NO: 315), Pro706 (SEQ ID NO: 316), and Pro709 (SEQ ID NO: 317).

[0265] In some embodiments, the single-specific COBRA that binds to EpCAM (e.g., human EpCAM) includes any of the fusion proteins F-J in Figure 10, such as Pro565 (SEQ ID NO: 288), Pro566 (SEQ ID NO: 289), Pro567 (SEQ ID NO: 290), Pro727 (SEQ ID NO: 292), Pro728 (SEQ ID NO: 293), Pro729 (SEQ ID NO: 294), Pro730 (SEQ ID NO: 295), and Pro731 (SEQ ID NO: 296).

[0266] In some embodiments, the single-specific COBRA that binds to FOLR1 (e.g., human FOLR1) includes any of the fusion proteins shown in Figure 9 D and E, such as Pro311 (SEQ ID NO: 416), Pro312 (SEQ ID NO: 417), and Pro313 (SEQ ID NO: 418).

[0267] In some embodiments, the single-specific COBRA that binds to HER2 (e.g., human HER2) includes one of the fusion proteins shown in Figures 12, D-P, such as any one of SEQ ID NOs. 459-484, as well as Pro1109 (SEQ ID NOs. 491), Pro1111 (SEQ ID NOs. 492), Pro1117 (SEQ ID NOs. 493), and Pro1118 (SEQ ID NOs. 494), such as Pro1109 (SEQ ID NOs. 491), Pro1111 (SEQ ID NOs. 492), Pro1117 (SEQ ID NOs. 493), and Pro1118 (SEQ ID NOs. 494).

[0268] In some embodiments, the single-specific COBRA that binds to LyPD3 (e.g., human LyPD3) includes one of the fusion proteins A-C in Figure 12, such as any of SEQ ID NOs. 453-458.

[0269] In some embodiments, the single-specific COBRA that binds to Trop2 (e.g., human Trop2) includes any of the fusion proteins J-M in Figure 10, such as Pro676 (SEQ ID NO: 297), Pro677 (SEQ ID NO: 298), Pro678 (SEQ ID NO: 299), Pro679 (SEQ ID NO: 300), Pro808 (SEQ ID NO: 301), and Pro819 (SEQ ID NO: 302).

[0270] 2. Dual-targeting type 2 construct "HeteroCOBRA" In some embodiments, each of the αTTA domains binds to a different tumor target. Therefore, in some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In this embodiment, aVH, aVL, iVH, and iVL have the sequences shown in Figures 5A and 5B. In this embodiment, the two targeting domains bind to different TTAs.

[0271] In Form 2, preferred dual tumor antigen targeting constructs (sometimes referred to herein as “heterospecific COBRA” or “heteroCOBRA”) include combinations targeting EGFR and EpCAM, EGFR and Trop2, EGFR and FOLR1, EGFR and B7H3, EGFR and LyPD3, EGFR and HER2, EpCAM and FOLR1, EpCAM and B7H3, EpCAM and Trop2, EpCAM and LyPD3, EpCAM and HER2, FOLR1 and B7H3, FOLR1 and HER2, FOLR1 and Trop2, FOLR1 and LyPD3, B7H3 and HER2, B7H3 and Trop2, B7H3 and LyPD3, HER2 and Trop2, HER2 and LyPD3, and Trop2 and LyPD3. These may be discussed herein as constructs such as “EGFR X EpCAM”.

[0272] In some embodiments, sdABD-TTA1 is selected from the group consisting of sdABD-B7H3, sdABD-CA9, sdABD-EGFR, sdABD-EpCAM, sdABD-FOLR1, sdABD-HER2, sdABD-LyPD3, and sdABD-Trop2. In some embodiments, sdABD-TTA2 is selected from the group consisting of sdABD-B7H3, sdABD-CA9, sdABD-EGFR, sdABD-EpCAM, sdABD-FOLR1, sdABD-HER2, sdABD-LyPD3, and sdABD-Trop2. In some embodiments, sdABD-TTA1 and sdABD-TTA2 bind to different target antigens.

[0273] In some embodiments, sdABD-TTA1 is sdABD-B7H3, and sdABD-TTA2 is selected from the group consisting of sdABD-CA9, sdABD-EGFR, sdABD-EpCAM, sdABD-FOLR1, sdABD-HER2, sdABD-LyPD3, and sdABD-Trop2. In some embodiments, sdABD-TTA1 is sdABD-EGFR, and sdABD-TTA2 is selected from the group consisting of sdABD-B7H3, sdABD-CA9, sdABD-EpCAM, sdABD-FOLR1, sdABD-HER2, sdABD-LyPD3, and sdABD-Trop2. In some embodiments, sdABD-TTA1 is sdABD-FOLR1, and sdABD-TTA2 is selected from the group consisting of sdABD-B7H3, sdABD-CA9, sdABD-EGFR, sdABD-EpCAM, sdABD-HER2, sdABD-LyPD3, and sdABD-Trop2. In some embodiments, sdABD-TTA1 is sdABD-HER2, and sdABD-TTA2 is selected from the group consisting of sdABD-B7H3, sdABD-CA9, sdABD-EGFR, sdABD-EpCAM, sdABD-FOLR1, sdABD-LyPD3, and sdABD-Trop2. In some embodiments, sdABD-TTA1 is sdABD-LyPD3, and sdABD-TTA2 is selected from the group consisting of sdABD-B7H3, sdABD-CA9, sdABD-EGFR, sdABD-EpCAM, sdABD-FOLR1, sdABD-HER2, and sdABD-Trop2.In some embodiments, sdABD-TTA1 is sdABD-Trop2, and sdABD-TTA2 is selected from the group consisting of sdABD-B7H3, sdABD-CA9, sdABD-EGFR, sdABD-EpCAM, sdABD-FOLR1, sdABD-HER2, and sdABD-LyPD3. Any sequence of sdABD-TTA described herein, such as the sequences of sdABD-B7H3, sdABD-CA9, sdABD-EGFR, sdABD-EpCAM, sdABD-FOLR1, sdABD-HER2, sdABD-LyPD3, and sdABD-Trop2, can be used in dual-targeted form 2 constructs or hetero-COBRA.

[0274] In some embodiments, sdABD-TTA1 is selected from the group consisting of sdABD-CA9, sdABD-EGFR, sdABD-EpCAM, sdABD-FOLR1, sdABD-HER2, sdABD-LyPD3, and sdABD-Trop2, and sdABD-TTA2 is sdABD-B7H3. In some embodiments, sdABD-TTA1 is selected from the group consisting of sdABD-B7H3, sdABD-EGFR, sdABD-EpCAM, sdABD-FOLR1, sdABD-HER2, sdABD-LyPD3, and sdABD-Trop2, and sdABD-TTA2 is sdABD-CA9. In some embodiments, sdABD-TTA1 is selected from the group consisting of sdABD-B7H3, sdABD-CA9, sdABD-EpCAM, sdABD-FOLR1, sdABD-HER2, sdABD-LyPD3, and sdABD-Trop2, and sdABD-TTA2 is sdABD-EGFR. In some embodiments, sdABD-TTA1 is selected from the group consisting of sdABD-B7H3, sdABD-CA9, sdABD-EGFR, sdABD-FOLR1, sdABD-HER2, sdABD-LyPD3, and sdABD-Trop2, and sdABD-TTA2 is sdABD-EpCAM. In some embodiments, sdABD-TTA1 is selected from the group consisting of sdABD-B7H3, sdABD-CA9, sdABD-EGFR, sdABD-EpCAM, sdABD-HER2, sdABD-LyPD3, and sdABD-Trop2, and sdABD-TTA2 is sdABD-FOLR1. In some embodiments, sdABD-TTA1 is selected from the group consisting of sdABD-B7H3, sdABD-CA9, sdABD-EGFR, sdABD-EpCAM, sdABD-FOLR1, sdABD-LyPD3, and sdABD-Trop2, and sdABD-TTA2 is sdABD-HER2. In some embodiments, sdABD-TTA1 is selected from the group consisting of sdABD-B7H3, sdABD-CA9, sdABD-EGFR, sdABD-EpCAM, sdABD-FOLR1, sdABD-HER2, and sdABD-Trop2, and sdABD-TTA2 is sdABD-LyPD3.In some embodiments, sdABD-TTA1 is selected from the group consisting of sdABD-B7H3, sdABD-CA9, sdABD-EGFR, sdABD-EpCAM, sdABD-FOLR1, sdABD-HER2, and sdABD-LyPD3, and sdABD-TTA2 is sdABD-Trop2. Any sequence of sdABD-TTA described herein, such as the sequences of sdABD-B7H3, sdABD-CA9, sdABD-EGFR, sdABD-EpCAM, sdABD-HER2, sdABD-LyPD3, and sdABD-Trop2, can be used in dual-targeted form 2 constructs or hetero-COBRA.

[0275] a. EGFR × EpCAM In some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In this embodiment, aVH, aVL, iVH, and iVL have the sequences shown in Figures 7A and 7B. In this embodiment, the two targeting domains bind to EGFR and EpCAM, and sdABD-TTA has the sequences in Figures 5A, 5D, and 7E, as well as the sequences and corresponding sequence listings provided therein. In this embodiment, some combinations of EGFR sdABD and EpCAM sdABD include: [Table 2]

[0276] In this case, "either orientation" means that either EpCAM sdABD is at the N-terminus of EGFR sdABD in the construct of the present invention, or EpCAM sdABD is at its C-terminus.

[0277] b.EGFR×FOLR1 In some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In some embodiments, aVH, aVL, iVH, and iVL have the sequences shown in Figures 7A and 7B. In some embodiments, the two targeting domains bind to EGFR and FOLR1, and sdABD-TTA has the sequences shown in Figures 5A and 7B, as well as the sequences and corresponding sequence listings provided therein. In some embodiments, some combinations of EGFR sdAb and FOLR1 sdAb include: [Table 3]

[0278] In this case, "either orientation" means that either FOLR1 sdABD is at the N-terminus of EGFR sdABD in the construct of the present invention, or FOLR1 sdABD is at its C-terminus.

[0279] c.EGFR×B7H3 In some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In this embodiment, aVH, aVL, iVH, and iVL have the sequences shown in Figure 7A-B. In some embodiments, the two targeting domains bind to EGFR and B7H3, and sdABD-TTA has the sequences shown in Figure 5A-D, as well as the sequences and corresponding sequence listings provided therein. In some embodiments, some combinations of EGFR sdABD and B7H3 sdABD include: [Table 4]

[0280] In this case, "either orientation" means that either B7H3 sdABD is at the N-terminus of EGFR sdABD in the construct of the present invention, or B7H3 sdABD is at its C-terminus.

[0281] d.EGFR×Trop2 In some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In some embodiments, aVH, aVL, iVH, and iVL have the sequences shown in Figure 7. In some embodiments, the two targeting domains bind to EGFR and Trop2, and sdABD-TTA has the sequences A, E, and F in Figure 5, as well as the sequences and corresponding sequence listings provided therein. In this embodiment, some combinations of EGFR sdABD and Trop2 sdABD include: [Table 5]

[0282] In this case, "either orientation" means that either the Trop2 sdABD is at the N-terminus of the EGFR sdABD in the construct of the present invention, or the Trop2 sdABD is at its C-terminus.

[0283] e.EGFR×LyPD3 In some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In some embodiments, aVH, aVL, iVH, and iVL have the sequences shown in Figures 7A and 7B. In some embodiments, the two targeting domains bind to EGFR and LyPD3, and sdABD-TTA has the sequences in Figures 5A, 5G, and 5H, as well as the sequences and corresponding sequence listings provided therein. In this embodiment, some combinations of EGFR sdABD and Trop2 sdABD include: [Table 6]

[0284] In this case, "either orientation" means that either the Trop2 sdABD is at the N-terminus of the EGFR sdABD in the construct of the present invention, or the Trop2 sdABD is at its C-terminus.

[0285] f.EGFR×HER2 In some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In some embodiments, aVH, aVL, iVH, and iVL have the sequences shown in Figure 7A-B. In some embodiments, the two targeting domains bind to EGFR and HER2, and sdABD-TTA has the sequences in Figure 5A, H-M and Figure 75, as well as the sequences and corresponding sequence listings provided therein. In this embodiment, some combinations of EGFR sdABD and HER2 sdABD include: [Table 7-1] [Table 7-2]

[0286] In this case, "either orientation" means that HER2 sdABD is either at the N-terminus or the C-terminus of EGFR sdABD in the construct of the present invention.

[0287] g.EpCAM×FOLR1 In some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In some embodiments, aVH, aVL, iVH, and iVL have the sequences shown in Figures 7A and 7B. In some embodiments, the two targeting domains bind to EpCAM and FOLR1, and sdABD-TTA has the sequences shown in Figures 5B, 5D, 7E and 75, as well as the sequences and corresponding sequence listings provided therein. In this embodiment, some combinations of EpCAM sdABD and FOLR1 sdABD include: [Table 8]

[0288] In this case, "either orientation" means that either EpCAM sdABD is at the N-terminus of FOLR1 sdABD in the construct of the present invention, or EpCAM sdABD is at its C-terminus.

[0289] h.EpCAM×B7H3 In some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In some embodiments, aVH, aVL, iVH, and iVL have the sequences shown in Figures 7A-B. In some embodiments, the two targeting domains bind to EpCAM and B7H3, and sdABD-TTA has the sequences shown in Figures 5B-E and Figure 75, as well as the sequences and corresponding sequence listings provided therein. In this embodiment, some combinations of EpCAM sdABD and B7H3 sdABD include: [Table 9]

[0290] In this case, "either orientation" means that either B7H3 sdABD is at the N-terminus of EpCAM sdABD in the construct of the present invention, or B7H3 sdABD is at its C-terminus.

[0291] i.EpCAM×Trop2 In some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In some embodiments, aVH, aVL, iVH, and iVL have the sequences shown in Figures 7A and 7B. In some embodiments, the two targeting domains bind to EpCAM and Trop2, and sdABD-TTA has the sequences shown in Figures 5D, 5E, 5F and 75, as well as the sequences and corresponding sequence listings provided therein. In this embodiment, preferred combinations of EpCAM sdABD and Trop2 sdABD include: [Table 10]

[0292] In this case, "either orientation" means that either the Trop2 sdABD is at the N-terminus of the EpCAM sdABD in the construct of the present invention, or the Trop2 sdABD is at its C-terminus.

[0293] j.EpCAM×LyPD3 In some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In some embodiments, aVH, aVL, iVH, and iVL have the sequences shown in Figures 7A and 7B. In some embodiments, the two targeting domains bind to EpCAM and LyPD3, and sdABD-TTA has the sequences shown in Figures 5D, 5E, 5H and 75, as well as the sequences and corresponding sequence listings provided therein. In this embodiment, some combinations of LyPD3 sdABD and EpCAM sdABD include: [Table 11]

[0294] In this case, "either orientation" means that LyPD3 sdABD is either at the N-terminus or C-terminus of EpCAM sdABD in the construct of the present invention.

[0295] k.EpCAM×HER2 In some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In some embodiments, aVH, aVL, iVH, and iVL have the sequences shown in Figure 7A-B. In some embodiments, the two targeting domains bind to HER2 and EpCAM, and sdABD-TTA has the sequences in Figure 5D, E, H-M and Figure 75, as well as the sequences and corresponding sequence listings provided therein. In this embodiment, some combinations of HER2 sdABD and EpCAM sdABD include: [Table 12-1] [Table 12-2]

[0296] In this case, "EO" means "either orientation," meaning that LyPD3 sdABD is either at the N-terminus or C-terminus of EpCAM sdABD in the construct of the present invention.

[0297] l.FOLR1×B7H3 In some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In some embodiments, aVH, aVL, iVH, and iVL have the sequences shown in Figure 7A-B. In some embodiments, the two targeting domains bind to FOLR1 and B7H3, and sdABD-TTA has the sequences shown in Figure 5B-D, as well as the sequences and corresponding sequence listings provided therein. In this embodiment, some combinations of FOLR1 sdABD and B7H3 sdABD include: [Table 13]

[0298] In this case, "either orientation" means that B7H3 sdABD is either at the N-terminus of FOLR1 sdABD or at its C-terminus in the construct of the present invention.

[0299] m.FOLR1×HER2 In some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In some embodiments, aVH, aVL, iVH, and iVL have the sequences shown in Figure 7A-B. In some embodiments, the two targeting domains bind to FOLR1 and HER2, and sdABD-TTA has the sequences in Figure 5B, H-M and Figure 75, as well as the sequences and corresponding sequence listings provided therein. In this embodiment, some combinations of FOLR1 sdABD and HER2 sdABD include: [Table 14]

[0300] In this case, "either orientation" means that either HER2 sdABD is at the N-terminus of FOLR1 sdABD in the construct of the present invention, or HER2 sdABD is at its C-terminus.

[0301] n.FOLR1×Trop2 In some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In some embodiments, aVH, aVL, iVH, and iVL have the sequences shown in Figures 7A and 7B. In some embodiments, the two targeting domains bind to FOLR1 and Trop2, and sdABD-TTA has the sequences shown in Figures 5B, 5E, and 5F, as well as the sequences and corresponding sequence listings provided therein. In this embodiment, preferred combinations of FOLR1 sdABD and Trop2 sdABD include: [Table 15]

[0302] In this case, "either orientation" means that either Trop2 sdABD is at the N-terminus of FOLR1 sdABD in the construct of the present invention, or Trop2 sdABD is at its C-terminus.

[0303] o.FOLR1×LyPD3 In some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In some embodiments, aVH, aVL, iVH, and iVL have the sequences shown in Figures 7A and 7B. In some embodiments, the two targeting domains bind to FOLR1 and LyPD3, and sdABD-TTA has the sequences in Figures 5B, 5G, and 5H, as well as the sequences and corresponding sequence listings provided therein. In this embodiment, some combinations of FOLR1 sdABD and LyPD3 sdABD include: [Table 16]

[0304] In this case, "either orientation" means that either LyPD3 sdABD is at the N-terminus of FOLR1 sdABD in the construct of the present invention, or LyPD3 sdABD is at its C-terminus.

[0305] p.B7H3×HER2 In some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In some embodiments, aVH, aVL, iVH, and iVL have the sequences shown in Figure 7A-B. In some embodiments, the two targeting domains bind to B7H3 and HER2, and sdABD-TTA has the sequences in Figure 5C, D, H-M and Figure 75, as well as the sequences and corresponding sequence listings provided therein. In this embodiment, some combinations of B7H3 sdABD and HER2 sdABD include: [Table 17-1] [Table 17-2] [Table 17-3] [Table 17-4]

[0306] In this case, "either orientation" means that either HER2 sdABD is at the N-terminus of B7H3 sdABD in the construct of the present invention, or HER2 sdABD is at its C-terminus.

[0307] q.B7H3×Trop2 In some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In some embodiments, aVH, aVL, iVH, and iVL have the sequences shown in Figure 7A-B. In some embodiments, the two targeting domains are bound to B7H3 and Trop2, and sdABD-TTA has the sequences shown in Figure 5B-F, as well as the sequences and corresponding sequence listings provided therein. In this embodiment, some combinations of B7H3 sdABD and Trop2 sdABD include: [Table 18]

[0308] In this case, "either orientation" means that either Trop2 sdABD is at the N-terminus of B7H3 sdABD in the construct of the present invention, or Trop2 sdABD is at its C-terminus.

[0309] r.B7H3×LyPD3 In some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In some embodiments, aVH, aVL, iVH, and iVL have the sequences shown in Figure 7A-B. In some embodiments, the two targeting domains bind to B7H3 and LyPD3, and sdABD-TTA has the sequences in Figure 5B-D, G, and H, as well as the sequences and corresponding sequence listings provided therein. In this embodiment, some combinations of B7H3 sdABD and LyPD3 sdABD include: [Table 19]

[0310] In this case, "either orientation" means that either LyPD3 sdABD is at the N-terminus of B7H3 sdABD in the construct of the present invention, or LyPD3 sdABD is at its C-terminus.

[0311] s.HER2×Trop2 In some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In some embodiments, aVH, aVL, iVH, and iVL have the sequences shown in Figure 7A-B. In some embodiments, the two targeting domains bind to HER2 and Trop2, and sdABD-TTA has the sequences in Figure 5E, F, H-M and Figure 75, as well as the sequences and corresponding sequence listings provided therein. In this embodiment, some combinations of HER2 sdABD and Trop2 sdABD include: [Table 20-1] [Table 20-2]

[0312] In this case, "either orientation" means that either HER2 sdABD is at the N-terminus of Trop2 sdABD in the construct of the present invention, or HER2 sdABD is at its C-terminus.

[0313] t.HER2×LyPD3 In some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In some embodiments, aVH, aVL, iVH, and iVL have the sequences shown in Figure 7A-B. In some embodiments, the two targeting domains bind to HER2 and LyPD3, and sdABD-TTA has the sequences in Figure 5G, H-M and Figure 75, as well as the sequences and corresponding sequence listings provided therein. In this embodiment, some combinations of HER2 sdABD and LyPD3 sdABD include: [Table 21-1] [Table 21-2]

[0314] In this case, "either orientation" means that HER2 sdABD is either at the N-terminus or the C-terminus of Trop2 sdABD in the construct of the present invention.

[0315] u.Trop2×LyPD3 In some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In some embodiments, aVH, aVL, iVH, and iVL have the sequences shown in Figures 7A and 7B. In some embodiments, the two targeting domains bind to Trop2 and LyPD3, and sdABD-TTA has the sequences shown in Figures 5E, 5G and 7H, as well as the sequences and corresponding sequence listings provided therein. In this embodiment, some combinations of Trop2 sdABD and LyPD3 sdABD include: [Table 22]

[0316] In this case, "either orientation" means that either Trop2 sdABD is at the N-terminus of LyPD3 sdABD in the construct of the present invention, or Trop2 sdABD is at its C-terminus.

[0317] In some embodiments, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CNCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus. In this embodiment, aVH, aVL, iVH, and iVL have sequences shown in Figure 7A-B. In some embodiments, two targeting mains bind to the same TTA, which may be EGFR, FOLR1, B7H3, CA9, Trop2, LyPD3, HER2, or EpCAM, whose sequences are shown in Figure 5A-M and Figure 75, CCL and CL are selected from linkers cleaved by MMP9 or meprin, and sdABD(1 / 2) has sequence number 249.

[0318] In Form 2, the preferred domain linker is Sequence ID No. 287 (which also functions as the preferred constrained non-cleavage linker).

[0319] B. Splitting form with dual targeting The present invention provides a non-isomerized cleavage form of type "Form 1" as shown in Figure 1. In this embodiment, the restrictive Fv domain comprises VH and VL domains linked using a restrictive cleavage linker, while the restrictive pseudo-Fv domain uses a restrictive non-cleavage linker. For ease of discussion, both of these are referred to herein as "restricted," but as discussed above and as shown in Figures 37, 38, and 39 of International Publication WO2019 / 051102, only one of these needs to be restricted; however, generally, the protein has better expression when both linkers are restricted.

[0320] All constructs of Form 1 (and other forms) also have a cleavable linker (CL) that is cleaved by human tumor proteases.

[0321] The present invention provides a prodrug protein comprising (sdABD-TTA1)-domain linker-constraining Fv domain-domain linker-(sdABD-TTA2)-CL-constraining pseudo-Fv domain-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus.

[0322] As will be understood by those skilled in the art, the order of VH and VL in either a restrictive Fv domain or a restrictive pseudo-Fv domain can be either VH-linker-VL or VL-linker-VH (from N-terminus to C-terminus).

[0323] Therefore, in one embodiment, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CCL-aVL-domain linker-(sdABD-TTA2)-CL-iVL-CNCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus.

[0324] Therefore, in one embodiment, the prodrug protein contains (sdABD-TTA1)-domain linker-aVH-CCL-aVL-domain linker-(sdABD-TTA2)-CL-iVH-CCL-iVL-domain linker-sdABD-HSA from the N-terminus to the C-terminus.

[0325] Therefore, in one embodiment, the prodrug protein contains (sdABD-TTA1)-domain linker-aVL-CCL-aVH-domain linker-(sdABD-TTA2)-CL-iVL-CCL-iVH-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus.

[0326] Therefore, in one embodiment, the prodrug protein contains (sdABD-TTA1)-domain linker-aVL-CCL-aVH-domain linker-(sdABD-TTA2)-CL-iVH-CCL-iVL-domain linker-(sdABD-HSA) from the N-terminus to the C-terminus.

[0327] 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). In this embodiment, aVH, aVL, iVH, and iVL have the sequences shown in Figures 7A and 7B.

[0328] In some embodiments, the prodrug construct comprises sdABD(TTA1)-domain linker-aVH-CCL-aVL-domain linker-sdABD(TTA2)-CL-iVL-CNCL-iVH-sdABD(1 / 2). In this embodiment, aVH, aVL, iVH, and iVL have sequences shown in Figures 7A-B. In this embodiment, the two targeting domains bind to the same TTA, which may be EGFR, EpCAM, FOLR1, Trop2, CA9, LyPD3, HER2, or B7H3, whose sequences are shown in Figures 5A-M and Figure 75.

[0329] In some embodiments, the prodrug construct comprises sdABD(TTA1)-domain linker-aVH-CCL-aVL-domain linker-sdABD(TTA2)-CL-iVL-CNCL-iVH-sdABD(1 / 2). In this embodiment, aVH, aVL, iVH, and iVL have the sequences shown in Figures 7A and 7B. In this embodiment, the two targeting domains bind to different TTAs.

[0330] In some embodiments, the prodrug construct comprises sdABD(TTA1)-domain linker-aVH-CCL-aVL-domain linker-sdABD(TTA2)-CL-iVL-CNCL-iVH-sdABD(1 / 2). In this embodiment, aVH, aVL, iVH, and iVL have the sequences shown in Figures 7A-B. In this embodiment, the two targeting domains bind to EGFR and EpCAM, and sdABD-TTA has the sequences shown in Figures 5A-M and Figure 75.

[0331] In some embodiments, the prodrug construct comprises sdABD(TTA1)-domain linker-aVH-CCL-aVL-domain linker-sdABD(TTA2)-CL-iVL-CNCL-iVH-sdABD(1 / 2). In this embodiment, aVH, aVL, iVH, and iVL have the sequences shown in Figures 7A-B. In this embodiment, the two targeting domains bind to EGFR and FOLR1, and sdABD-TTA has the sequences shown in Figures 5A-M and Figure 75.

[0332] In some embodiments, the prodrug construct comprises sdABD(TTA1)-domain linker-aVH-CCL-aVL-domain linker-sdABD(TTA2)-CL-iVL-CNCL-iVH-sdABD(1 / 2). In this embodiment, aVH, aVL, iVH, and iVL have the sequences shown in Figures 7A-B. In this embodiment, the two targeting domains bind to EGFR and B7H3, and sdABD-TTA has the sequences shown in Figures 5A-M and Figure 75.

[0333] In some embodiments, the prodrug construct comprises sdABD(TTA1)-domain linker-aVH-CCL-aVL-domain linker-sdABD(TTA2)-CL-iVL-CNCL-iVH-sdABD(1 / 2). In this embodiment, aVH, aVL, iVH, and iVL have the sequences shown in Figures 7A-B. In this embodiment, the two targeting domains bind to EpCAM and FOLR1, and sdABD-TTA has the sequences shown in Figures 5A-M and Figure 75.

[0334] In some embodiments, the prodrug construct comprises sdABD(TTA1)-domain linker-aVH-CCL-aVL-domain linker-sdABD(TTA2)-CL-iVL-CNCL-iVH-sdABD(1 / 2). In this embodiment, aVH, aVL, iVH, and iVL have the sequences shown in Figures 7A-B. In this embodiment, the two targeting domains bind to EpCAM and B7H3, and sdABD-TTA has the sequences shown in Figures 5A-M and Figure 75.

[0335] In some embodiments, the prodrug construct comprises sdABD(TTA1)-domain linker-aVH-CCL-aVL-domain linker-sdABD(TTA2)-CL-iVL-CNCL-iVH-sdABD(1 / 2). In this embodiment, aVH, aVL, iVH, and iVL have the sequences shown in Figures 7A and 7B. In this embodiment, the two targeting domains bind to B7H3 and FOLR1, and sdABD-TTA has the sequences shown in Figures 5A to 7M and Figure 75.

[0336] In some embodiments, the prodrug construct comprises sdABD(TTA1)-domain linker-aVH-CCL-aVL-domain linker-sdABD(TTA2)-CL-iVL-CNCL-iVH-sdABD(1 / 2). In this embodiment, aVH, aVL, iVH, and iVL have the sequences shown in Figure 7. In this embodiment, two targeted mains bind to the same TTA, which may be EGFR, FOLR1, B7H3, Trop2, CA9, LyPD3, HER2, or EpCAM, whose sequences are shown in Figure 5, CCL and CL are selected from linkers cleaved by MMP9 or meprin, and sdABD(1 / 2) has sequence number 245 or sequence number 249.

[0337] In Form 1, the preferred domain linker is Sequence ID No. 287 (which also functions as the preferred constrained non-cleavage linker).

[0338] C. Single TTA structure As shown in Figure 4, a “Form 4” construct, similar to the Form 2 construct but lacking a second TTA ABD, is also included in the compositions of the present invention. In this embodiment, since an activation cleavage site is present in the prodrug construct, it is understood that “non-cleavage” applies only to the linking of restrictive Fv domains. In this embodiment, the restrictive Fv domain includes VH and VL domains linked using a restrictive non-cleavageable linker, and the restrictive pseudo-Fv domain uses a restrictive non-cleavageable linker.

[0339] As will be understood by those skilled in the art, the order of VH and VL in either a restrictive Fv domain or a restrictive pseudo-Fv domain can be either VH-linker-VL or VL-linker-VH (from N-terminus to C-terminus).

[0340] The present invention provides a prodrug protein comprising, from N-terminus to C-terminus, a (sdABD-TTA)-domain linker-constraining Fv domain-cleavage linker-(sdABD-HSA)-constraining pseudo-Fv domain. Note that in all constructs of this form, sdABD-HSA generally does not have His6, but may contain it.

[0341] As will be understood by those skilled in the art, the order of VH and VL in either a restrictive Fv domain or a restrictive pseudo-Fv domain can be either VH-linker-VL or VL-linker-VH (from N-terminus to C-terminus).

[0342] Therefore, in one embodiment, the prodrug protein contains (sdABD-TTA)-domain linker-aVH-CNCL-aVL-CL-(sdABD-HSA)-domain linker-iVL-CNCL-iVH from the N-terminus to the C-terminus.

[0343] Therefore, in one embodiment, the prodrug protein contains (sdABD-TTA)-domain linker-aVH-CNCL-aVL-CL-(sdABD-HSA)-domain linker-iVH-CNCL-iVL from the N-terminus to the C-terminus.

[0344] Therefore, in one embodiment, the prodrug protein contains (sdABD-TTA)-domain linker-aVL-CNCL-aVH-CL-(sdABD-HSA)-domain linker-iVH-CNCL-iVL from the N-terminus to the C-terminus.

[0345] Therefore, in one embodiment, the prodrug protein contains (sdABD-TTA)-domain linker-aVL-CNCL-aVH-CL-(sdABD-HSA)-domain linker-iVL-CNCL-iVH from the N-terminus to the C-terminus.

[0346] Therefore, in one embodiment, the prodrug protein contains (sdABD-TTA)-domain linker-aVH-CNCL-aVL-CL-(sdABD-HSA)-domain linker-iVL-CNCL-iVH from the N-terminus to the C-terminus. In this embodiment, aVH, aVL, iVH, and iVL have sequences shown in Figures 7A-B. In this embodiment, the targeting domain binds to TTA, which can be EGFR, EpCAM, FOLR1, Trop2, CA9, LyPD3, HER2, or B7H3, the sequences of which are shown in Figures 5A-M and Figure 75.

[0347] D. Two protein compositions In some embodiments, the composition of the present invention comprises two distinct molecules, sometimes referred to as "semi-COBRA®" or "semi-construction," which, in the absence of cleavage, associate intramolecularly to form a pseudo-Fv. In the presence of a protease, the cleavage site is cleaved, releasing an inactive variable domain, and the protein pair then forms an active antigen-binding domain for CD3, generally as shown in Figure 3.

[0348] A key aspect of the design of the semiconstruct is that the activatable variable domain and sdABD-TTA remain together after cleavage, so that the two cleaved regions can be held together by tumor antigen receptors on the tumor surface and subsequently form an activatable anti-CD3 binding domain.

[0349] There are two different general form 3 constructs: one in which each member of the pair has a single sdABD-TTA (Figure 3A), and another in which each has two different sdABD-TTAs for different TTAs (Figure 3B).

[0350] 1. Semi-COBRA™ construct (Form 3A) having a single TTA-binding domain. In some embodiments, the first half-COBRA(trademark) has sdABD(TTA1)-domain linker-aVH-CL-iVL-domain linker-sdABD(1 / 2) from the N-terminus to the C-terminus, and the second half-COBRA(trademark) has sdABD(1 / 2)-domain linker-iVH-CL-aVL-domain linker-sdABD(TTA2). In this embodiment, aVH, aVL, iVH, iVL and sdABD(1 / 2) have the sequences shown in Figures 6 and 7, and sdABD-TTAa binds to human EGFR, EpCAM, Trop2, CA9, LyPD3, HER2, OLR1 and / or B7H3 and has the sequences shown in Figures 5A-M and 75.

[0351] 2. Semi-COBRA™ structure with dual TTA ABD In some embodiments, the paired prodrug constructs may have two sdABD-TTA binding domains per construct, as shown in Figure 3B. In this embodiment, the first member of the pair comprises sdABD-TTA1-domain linker-sdABD-TTA2-domain linker-aVH-CL-iVL-domain linker-sdABD(HAS) from N-terminus to C-terminus, and the second member comprises sdABD-TTA1-domain linker-sdABD-TTA2-aVL-CL-iVH-domain linker-(sdABD-HSA) from N-terminus to C-terminus.

[0352] While the two sdABD-TTAs on each member of a pair are different, generally speaking, both members (half-COBRA(trademark)) have the same two sdABD-TTAs, for example, both having EGFR and FOLR1 or EGFR and B7H3.

[0353] In some embodiments, the two sdABD-TTAs are selected from those shown in Figures 5A-M and Figure 75.

[0354] IV. Method for preparing the composition of the present invention The prodrug compositions of the present invention are prepared as will be generally understood by those skilled in the art and as outlined below.

[0355] The present invention provides nucleic acid compositions encoding the prodrug compositions of the present invention. As will be understood by those skilled in the art, the nucleic acid composition is determined by the form of the prodrug polypeptide. Thus, if the form requires two amino acid sequences, such as the “Form 3” construct, the two nucleic acid sequences can be incorporated into one or more expression vectors for expression. Similarly, prodrug constructs that are single polypeptides (Forms 1, 2, and 4) require a single nucleic acid in a single expression vector for production.

[0356] As is known in the art, nucleic acids encoding the components of the present invention can be incorporated into an expression vector, as is known in the art and depending on the host cell used to produce the prodrug composition of the present invention. Generally, nucleic acids are operably linked to any number of regulatory elements (promoters, origins of replication, selectable markers, ribosome binding sites, inducers, etc.). The expression vector can be extrachromosomal or embedded.

[0357] Subsequently, the nucleic acids and / or expression vectors of the present invention are converted into any number of different types of host cells, as well as those well known in the art, including mammalian cells, bacteria, yeast, insects, and / or fungal cells, in mammalian cells (e.g., CHO cells, 293 cells) used in many embodiments.

[0358] The prodrug composition of the present invention is prepared by culturing host cells containing an expression vector (or more) as is well known in the art. Once prepared, a conventional antibody purification step is performed, including a protein A affinity chromatography step and / or an ion exchange chromatography step.

[0359] V. Formulation and administration of the prodrug composition of the present invention Formulations of prodrug compositions used in accordance with the present invention are prepared for storage in the form of lyophilized formulations or aqueous solutions by mixing a prodrug of desired purity (a single protein in the case of forms 1, 2, and 4, and two proteins in the case of form 3) with any pharmaceutically acceptable carrier, excipient, or stabilizer (generally as outlined in Remington's Pharmaceutical Sciences 16th edition, Osol, A. Ed.

[1980] ).

[0360] The prodrug composition of the present invention is administered to a subject according to known methods, such as intravenous administration as a bolus or by continuous infusion over a period of time.

[0361] The prodrug compositions of the present invention are useful for the treatment of cancer. A method for treating cancer in a patient, comprising any of the described prodrug compositions, is provided herein. Prodrug compositions for use as pharmaceuticals are described herein. A pharmaceutical composition for treating cancer, comprising any of the described prodrug compositions, is provided. A pharmaceutical composition comprising any of the described prodrug compositions is provided for treating cancer in a patient who needs it. A described prodrug composition for use in a treatment or method for treating cancer is provided. A prodrug composition described herein for treating cancer in a patient who needs it is provided. The use of prodrug compositions in the manufacture of drugs for the treatment of cancer is provided.

[0362] VI. Exemplary Embodiments The present invention provides many different protein compositions for the treatment of cancer. Accordingly, in one embodiment, the present invention provides a first single-domain antigen-binding domain (sdABD) (sdABD-TTA) that binds to a human tumor target antigen (TTA) from the N-terminus to the C-terminus, a) a domain linker, c) i) a variable heavy domain including vhCDR1, vhCDR2 and vhCDR3, ii) a restrictive non-cleaving linker (CNCL), and iii) a restrictive Fv domain including a variable light domain including vlCDR1, vlCDR2 and vlCDR3, d) a second domain linker, e) a second sdABD-TTA, f) a cleaving linker (CL), and g) i) The present invention provides a “Form 2” protein comprising: ii) a pseudo-light variable domain, a restrictive non-cleavable linker (CNCL), and iii) a restrictive pseudo-Fv domain including a pseudo-heavy variable domain; h) a third domain linker; and i) a third sdABD that binds to human serum albumin, wherein the variable heavy domain and variable light domain are capable of binding to human CD3, but the restrictive Fv domain does not bind to CD3; the variable heavy domain and pseudo-variable light domain associate intramolecularly to form an inactive Fv; and the variable light domain and pseudo-variable heavy domain associate intramolecularly to form an inactive Fv. In some embodiments, the human tumor target antigen is B7H3.

[0363] In a further embodiment, the present invention comprises a first single-domain antigen-binding domain (sdABD) (sdABD-TTA) that binds to a human tumor target antigen (TTA), comprising sdFR1-sdCDR1-sdFR2-sdCDR2-sdFR3-sdCDR3-sdFR4 from the N-terminus to the C-terminus, a) a first domain linker, and c) i) vhFR1-vhCDR1-vhFR2-vhCDR2-vhFR3-vhCDR3 -vhFR4-containing variable heavy domain, ii) constrained non-cleavable linker (CNCL), and iii) constrained Fv domain containing a variable light domain including vlFR1-vlCDR1-vlFR2-vlCDR2-vlFR3-vlCDR3-vlFR4, d) second domain linker, e) second sdABD-TTA, f) cleavable linker (CL), g) i) sdFR1-sdCDR1-sdFR2-sd The present invention provides a protein comprising: ii) a pseudo-light variable domain containing CDR2-sdFR3-sdCDR3-sdFR4; ii) a restrictive non-cleavable linker (CNCL); and iii) a restrictive pseudo-Fv domain containing a pseudo-heavy variable domain containing vlFR1-vlCDR1-vlFR2-vlCDR2-vlFR3-vlCDR3-vlFR4; h) a third domain linker; and i) a third sdABD containing sdFR1-sdCDR1-sdFR2-sdCDR2-sdFR3-sdCDR3-sdFR4 that binds to human serum albumin. The variable heavy domain and variable light domain can bind to human CD3, but the restrictive Fv domain does not bind to CD3. The variable heavy domain and pseudo-variable light domain associate intramolecularly to form an inactive Fv, and the variable light domain and pseudo-variable heavy domain associate intramolecularly to form an inactive Fv. In some embodiments, the human tumor target antigen is B7H3.

[0364] In some embodiments of the form 2 protein, the variable weight domain is located at the N-terminus of the variable light domain, and the pseudo-variable light domain is located at the N-terminus of the pseudo-variable weight domain. In some embodiments, the variable weight domain is located at the N-terminus of the variable light domain, and the pseudo-variable light domain is located at the C-terminus of the pseudo-variable weight domain. In some embodiments, the variable weight domain is located at the C-terminus of the variable light domain, and the pseudo-variable light domain is located at the N-terminus of the pseudo-variable weight domain. In some embodiments, the variable weight domain is located at the C-terminus of the variable light domain, and the pseudo-variable light domain is located at the C-terminus of the pseudo-variable weight domain.

[0365] In some embodiments of the form 2 protein, the first sdABDTTA and the second sdABDTTA are the same. In some embodiments, the first sdABDTTA and the second sdABDTTA are different. In these embodiments, sdABD-TTA is selected from those shown in Figure 7, including SEQ ID NOs: 1, 5, 9, 13, 17, 21, 25, 29, 33, 37, 41, 45, 49, 53, 57, 61, 65, 69, 73, 77, 81, 85, 89, 93, 97, 101, 105, 109, and 113.

[0366] In some embodiments of the form 2 protein, the pseudo-heavy chain variable domain of the restrictive pseudo-Fv domain is as shown in Figure 7, as in Sequence ID No. 146(V Hi ), Sequence ID 150(V Hi2 ) and sequence number 154 (VHiGL4) are selected from the group. In some embodiments, the pseudo-light variable domain of the restrictive pseudo-Fv domain is sequence number 130 (V) as shown in Figure 7. Li ), Sequence ID 134(V Li2 ) and Sequence ID 138(V LiGL Selected from the group of ).

[0367] In a further embodiment, from the N-terminus to the C-terminus, (a) a first sdABD-TTA, (b) a first domain linker, (c) a restrictive Fv domain including (i) a first variable heavy domain containing vhCDR1, vhCDR2 and vhCDR3 and (ii) a restrictive non-cleavable linker (CCL) and (iii) a first variable light domain containing vlCDR1, vlCDR2 and vlCDR3, (d) a second domain linker, (e) a second sdABD-TTA, (f) a cleavable linker (CL), (g) an (i) first pseudo-variable light domain and (ii) a restrictive non-cleavable linker (CCL), A “Form 1” protein is provided, comprising (iii) a restrictive pseudo-Fv domain including (nCl) and a first pseudo-variable weight domain, (h) a third domain linker, and (i) a third sdABD that binds to human serum albumin, wherein the first variable weight domain and the first variable light domain can bind to human CD3, but the restrictive Fv domain does not bind to CD3, the first variable weight 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 weight domain associate intramolecularly to form an inactive Fv. In a further embodiment, from the N-terminus to the C-terminus, (a) a single-domain antigen-binding domain (sdABD) (sdABD-TTA) that binds to human tumor target antigens (TTA), (b) a first domain linker, (c) a restrictive Fv domain comprising (i) a first variable heavy domain including vhCDR1, vhCDR2 and vhCDR3, (ii) a restrictive non-cleaving linker (CNCL), and (iii) a first variable light domain including vlCDR1, vlCDR2 and vlCDR3, (d) a cleaving linker (CL), (e) a second sdABD that binds to human serum albumin, and (f) a domain A "Form 4" protein is provided, comprising a linker, (g)(i) a first pseudo-variable light domain, (ii) a restrictive non-cleavable linker (CNCL), and (iii) a restrictive pseudo-Fv domain including a first pseudo-variable heavy domain, wherein the first variable heavy domain and the first variable light domain can bind to human CD3, but the restrictive pseudo-Fv domain does not bind to CD3, 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.

[0368] In further embodiments of the Form 1, Form 2, and Form 4 proteins described above, the first variable weight domain is located at the N-terminus of the first variable light domain, and the pseudo-light variable domain is located at the N-terminus of the pseudo-variable weight domain.

[0369] In further embodiments of the Form 1, Form 2, and Form 4 proteins described above, the first variable weight domain is located at the N-terminus of the first variable light domain, and the pseudo-variable weight domain is located at the N-terminus of the pseudo-variable light domain.

[0370] In further embodiments of the Form 1, Form 2, and Form 4 proteins described above, the first variable light domain is located at the N-terminus of the first variable heavy domain, and the pseudo-light variable domain is located at the N-terminus of the pseudo-variable heavy domain.

[0371] In further embodiments of the Form 1, Form 2, and Form 4 proteins described above, the first variable light domain is located at the N-terminus of the first variable heavy domain, and the pseudo-variable heavy domain is located at the N-terminus of the pseudo-variable light domain.

[0372] In an additional embodiment, the present invention provides form 1 and form 2 proteins in which the first and second TTAs are the same. In a further embodiment, the present invention provides form 1 and form 2 proteins in which the first and second TTAs are different.

[0373] In an additional embodiment, the present invention provides Form 1, Form 2, and Form 4 proteins in which the first and second TTAs are selected from EGFR, EpCAM, FOLR1, Trop2, ca9, and B7H3. These sequences may be selected from the group consisting of SEQ ID NOs: 1, 5, 9, 13, 17, 21, 25, 29, 33, 37, 41, 45, 49, 53, 57, 61, 65, 69, 73, 77, 81, 85, 89, 93, 97, 101, 105, 109, and 113.

[0374] In a further embodiment, the present invention provides proteins of form 1, form 2, and form 4, wherein the half-life extension domains include sequence number 117 (aHSA(10GE)) and sequence number 121 (aHSA with HIS tag).

[0375] In an additional embodiment, the present invention provides Form 1, Form 2, and Form 4 proteins in which 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, kalecreiin 7, matryptase, and thrombin, or others shown in Figure 6.

[0376] In further embodiments, the present invention relates to Pro186, Pro225, Pro226, Pro233, Pro262, Pro311, Pro312, Pro313, Pro356, Pro359, Pro364, Pro388, Pro448, Pro449, Pro450, Pro451, Pro495, Pro246, Pro254, Pro255, Pro256, Pro420, Pro421, Pro432, Pro479, Pro480, Pro187, Pro221, Pro222, Pro223, Pro224, Pro393, Pro394, Pro395, Pro396, Pro429, Pro430, Pro431, Pro601, Pro602, V3 and V4, Pro664, Pro665, Pro667, Pro694, P We provide proteins selected from the group consisting of ro695, Pro565, Pro566, Pro567, Pro727, Pro728, Pro729, Pro730, Pro731, Pro676, Pro677, Pro678, Pro679, Pro808, Pro819, Pro621, Pro622, Pro640, Pro641, Pro642, Pro643, Pro744, Pro746, Pro638, Pro639, Pro396, Pro476, Pro706, Pro709, Pro470, Pro471, Pro551, Pro552, Pro623, Pro624, Pro698, Pro655, Pro656, Pro657, Pro658, Pro516, Pro517, Pro518, and Pro519.

[0377] In additional embodiments, the present invention provides nucleic acids encoding the form 1, form 2, or form 4 proteins described herein, as well as expression vectors and host cells comprising the protein-encoding nucleic acids.

[0378] In a further embodiment, the present invention provides a method for producing the protein of the present invention and a method for treating patients who require it.

[0379] In an additional embodiment, the present invention comprises a) i) a first sdABD-TTA from the N-terminus to the C-terminus, ii) a first domain linker, iii) a pseudo-Fv domain from the N-terminus to the C-terminus comprising 1) a variable heavy chain including vhCDR1, vhCDR2 and vhCDR3, 2) a cleavable linker and 3) a first pseudo-variable light domain including iVLCDR1, iVLCDR2 and iVLCDR3, iv) a second domain linker, and v) a first sdABD-HSA. The protein comprises: a) a third sdABD that binds to a human tumor target antigen from the N-terminus to the C-terminus; ii) a third domain linker; iii) a pseudo-Fv domain that includes a first pseudo-variable heavy domain containing 1) a variable light chain containing VLCDR1, VLCDR2, and VLCDR3, 2) a cleavable linker, and 3) iVHCDR1, iVHCDR2, and iVHCDR3, from the N-terminus to the C-terminus; iv) a fourth domain linker; and v) sdABD-HSA. The present invention provides a composition comprising a "form 3A" pair of prodrug proteins comprising a second protein, wherein the first variable heavy domain and the first variable light domain can bind to human CD3 when associated, the first variable heavy domain and the first pseudo-variable light domain intermolecularly associate to form an inactive Fv, and 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 NOs: 1, 5, 9, 13, 17, 21, 25, 29, 33, 37, 41, 45, 49, 53, 57, 61, 65, 69, 73, 77, 81, 85, 89, 93, 97, 101, 105, 109, and 113.

[0380] In a further embodiment, the present invention comprises a) a first protein comprising i) a first sdABD-TTA from the N-terminus to the C-terminus, ii) a first domain linker, iii) a second sdABD-TTA, iv) a second domain linker, iii) a pseudo-Fv domain comprising 1) a variable heavy chain including vhCDR1, vhCDR2 and vhCDR3 from the N-terminus to the C-terminus, 2) a cleavable linker and 3) a first pseudo-variable light domain including iVLCDR1, iVLCDR2 and iVLCDR3, iv) a third domain linker, and v) a first protein comprising sdABD-HSA, and a) a) a third sdABD-TTA from the N-terminus to the C-terminus, ii) a fourth domain linker, iii) a fourth sdABD-TTA, and iv) a fifth domain linker. iii) provides a composition comprising a "form 3B" pair of prodrug proteins comprising 1) a variable light chain containing VLCDR1, VLCDR2 and VLCDR3 from the N-terminus to the C-terminus, 2) a cleavable linker and 3) a pseudo-Fv domain containing a first pseudo-variable heavy domain containing iVHCDR1, iVHCDR2 and iVHCDR3, iv) a sixth domain linker and v) a second protein containing sdABD-HSA, wherein the first variable heavy domain and the first variable light domain can bind to human CD3 when associated, the first variable heavy domain and the first pseudo-variable light domain intermolecularly associate to form an inactive Fv, and the first variable light domain and the first pseudo-variable heavy domain intermolecularly associate to form an inactive Fv.

[0381] In an additional embodiment, the form 3A and form 3B proteins have an sdABD-HSA having SEQ ID NO: 117 or SEQ ID NO: 121. In a further embodiment, the form 3A and form 3B proteins have an sdABD-TTA that binds to a TTA selected from EGFR, EpCAM, Trop2, CA9, 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, SEQ ID NO: 41, SEQ ID NO: 45, SEQ ID NO: 49, SEQ ID NO: 53, SEQ ID NO: 57, SEQ ID NO: 61, SEQ ID NO: 65, SEQ ID NO: 69, SEQ ID NO: 73, SEQ ID NO: 77, SEQ ID NO: 81, SEQ ID NO: 85, SEQ ID NO: 89, SEQ ID NO: 93, SEQ ID NO: 97, SEQ ID NO: 101, SEQ ID NO: 105, SEQ ID NO: 109 and SEQ ID NO: 113.

[0382] In an additional embodiment, the present invention provides an sdABD that binds to human Trop2 having a sequence selected from SEQ ID NOs. 77, 81, 85, 89, and 93. In a further embodiment, the present invention provides an sdABD that binds to human B7H3 having a sequence selected from SEQ ID NOs. 41, 45, 49, 53, and 57. In an additional embodiment, the present invention provides an sdABD that binds to human CA9 having a sequence selected from SEQ ID NOs. 101, 105, 109, and 113. In a further embodiment, the present invention provides an sdABD that binds to human EpCAM having a sequence selected from SEQ ID NOs. 69 and 73.

[0383] In some embodiments, from the N-terminus to the C-terminus, (a) a first sdABD (sdABD-TTA) that binds to a tumor target antigen, (b) a first domain linker, (c) a restrictive Fv domain including (i) a first variable heavy domain containing vhCDR1, vhCDR2 and vhCDR3, (ii) a restrictive non-cleavage linker (CNCL), and (iii) a first variable light domain containing vlCDR1, vlCDR2 and vlCDR3, (d) a second domain linker, (e) a second sdABD-TTA, (f) a cleavage linker (CL), (g) an (i) first pseudo-variable light domain, (ii) a non-cleavage linker (NCL), and (iii) a first pseudo-variable heavy domain. A fusion protein is provided comprising (h) a restrictive pseudo-Fv domain, and (i) a third sdABD (sdABD-HSA) that binds to human serum albumin, wherein the first variable weight domain and the first variable light domain can bind to human CD3, but the restrictive Fv domain does not bind to CD3, the first variable weight domain and the first pseudo-variable light domain associate intramolecularly to form an inactive Fv, and the first sdABD-TTA and the second sdABD-TTA bind to the same TTA selected from the group consisting of B7H3, CA9, EGFR, EpCAM, FOLR1, HER2, LyPD3, and Trop2. In some embodiments, the first and / or second sdABD-TTA may be any sdABD-TTA disclosed herein.

[0384] In some embodiments, from the N-terminus to the C-terminus, (a) a first sdABD (sdABD-TTA) that binds to a tumor target antigen, (b) a first domain linker, (c) a restrictive Fv domain comprising (i) a first variable heavy domain containing vhCDR1, vhCDR2 and vhCDR3, (ii) a restrictive non-cleaving linker (CNCL), and (iii) a first variable light domain containing vlCDR1, vlCDR2 and vlCDR3, (d) a second domain linker, (e) a second sdABD-TTA, (f) a cleaving linker (CL), (g) a restrictive pseudo-Fv domain comprising (i) a first pseudo-variable light domain, (ii) a non-cleaving linker (NCL), and (iii) a first pseudo-variable heavy domain, (h) a third domain linker, and (i) A fusion protein is provided comprising a third sdABD (sdABD-HSA) that binds to human serum albumin, wherein the first variable weight domain and the first variable light domain can bind to human CD3, but the restrictive Fv domain does not bind to CD3, the first variable weight domain and the first pseudo-variable light domain associate intramolecularly to form an inactive Fv, the first sdABD-TTA binds to a TTA selected from the group consisting of B7H3, CA9, EGFR, EpCAM, FOLR1, HER2, LyPD3 and Trop2, and the second sdABD-TTA binds to a different TTA selected from the group consisting of B7H3, CA9, EGFR, EpCAM, FOLR1, HER2, LyPD3 and Trop2. In some embodiments, the first and / or second sdABD-TTA may be any sdABD-TTA disclosed herein.

[0385] Sequence IDs 288 (Pro565), 289 (Pro566), 290 (Pro567), 292 (Pro727), 293 (Pro728), 294 (Pro729), 295 (Pro730), 296 (Pro731), 297 (Pro676), 298 (Pro677), 299 (Pro678), 300 (Pro679), 301 (Pro808), 302 (Pro819), 304 (Pro621), 30 5 (Pro622), SEQ ID NO: 306 (Pro640), SEQ ID NO: 307 (Pro641), SEQ ID NO: 308 (Pro642), SEQ ID NO: 309 (Pro643), SEQ ID NO: 310 (Pro744), SEQ ID NO: 311 (Pro746), SEQ ID NO: 312 (Pro108), SEQ ID NO: 313 (Pro109), SEQ ID NO: 314 (Pro396), SEQ ID NO: 315 (Pro476), SEQ ID NO: 316 (Pro706), SEQ ID NO: 317 (Pro709), SEQ ID NO: 318 (Pro470), SEQ ID NO: 319 (Pro471), SEQ ID NO: 320 (Pro55 1) Sequence numbers 321 (Pro552), 322 (Pro623), 323 (Pro624), 324 (Pro698), 325 (Pro655), 326 (Pro656), 327 (Pro657), 328 (Pro658), 329 (Pro516), 330 (Pro517), 331 (Pro518), 332 (Pro519), 333 (Pro513), 336 (Pro225), 338 (Pro817), sequence number Sequence numbers 416 (Pro311), 417 (Pro312), 418 (Pro313), 419 (Pro246), 420 (Pro256), 421 (Pro420), 422 (Pro421), 487 (Pro751), 488 (Pro752), 489 (Pro824), and 490 (Pro826), 522 (Pro601), 523 (Pro602), 524 (V3), 525 (V4), 526 (Pro664),A fusion protein containing an amino acid sequence selected from the group consisting of SEQ ID NOs: 527 (Pro665), 528 (Pro667), 529 (Pro694), 530 (Pro695), and 531 (Pro565).

[0386] In a further embodiment, the present invention provides a nucleic acid composition comprising a first nucleic acid encoding a first protein member of a prodrug pair, a second nucleic acid encoding a second protein member of the pair, and an expression vector and a host cell containing the nucleic acids. For example, this disclosure provides the following embodiments. [1] A fusion protein, wherein the fusion protein extends from the N-terminus to the C-terminus. a) A first sdABD (sdABD-HER2) that binds to HER2, b) The first domain linker, c)i) A first variable weight domain including vhCDR1, vhCDR2 and vhCDR3, ii) Constrained non-cleavable linkers (CNCLs) and iii) A restrictive Fv domain including a first variable light domain including vlCDR1, vlCDR2 and vlCDR3, d) The second domain linker, e) The second sdABD-HER2, f) Splitting linker (CL), g)i) The first pseudo-variable light domain, ii) Non-cleavable linker (NCL) and iii) A restrictive pseudo-Fv domain including the first pseudo-variable heavy domain, h) The third domain linker, i) comprising a third sdABD (sdABD-HSA) that binds to human serum albumin, Here, the first variable weight domain and the first variable light domain of the restrictive Fv domain can bind to human CD3, but the restrictive pseudo-Fv domain does not bind to CD3. The first variable heavy domain and the first pseudo-variable light domain associate intramolecularly to form an inactive Fv. The fusion protein wherein the first variable light domain and the first pseudo-variable heavy domain associate intramolecularly to form an inactive Fv. [2] The first sdABD-HER2 and / or the second sdABD-HER2 are a) sdCDR1 of sequence number 194, sdCDR2 of sequence number 195, and sdCDR3 of sequence number 196, b) sdCDR1 of sequence number 218, sdCDR2 of sequence number 219, and sdCDR3 of sequence number 220, c) sdCDR1 of sequence number 226, sdCDR2 of sequence number 227, and sdCDR3 of sequence number 228, d) sdCDR1 of sequence number 238, sdCDR2 of sequence number 239, and sdCDR3 of sequence number 240, e) sdCDR1 of sequence number 142, sdCDR2 of sequence number 143, and sdCDR3 of sequence number 144, f) sdCDR1 of sequence number 146, sdCDR2 of sequence number 147, and sdCDR3 of sequence number 148, g) sdCDR1 of SEQ ID NO: 150, sdCDR2 of SEQ ID NO: 151, and sdCDR3 of SEQ ID NO: 152 h) sdCDR1 of sequence number 154, sdCDR2 of sequence number 155, and sdCDR3 of sequence number 156, i) sdCDR1 of sequence number 158, sdCDR2 of sequence number 159, and sdCDR3 of sequence number 160, j) sdCDR1 of sequence number 162, sdCDR2 of sequence number 163, and sdCDR3 of sequence number 164, k) sdCDR1 of sequence number 166, sdCDR2 of sequence number 167, and sdCDR3 of sequence number 168, l) sdCDR1 of sequence number 170, sdCDR2 of sequence number 171, and sdCDR3 of sequence number 172, m) sdCDR1 of sequence number 174, sdCDR2 of sequence number 175, and sdCDR3 of sequence number 176, n) sdCDR1 of sequence number 178, sdCDR2 of sequence number 179, and sdCDR3 of sequence number 180, o) sdCDR1 of sequence number 182, sdCDR2 of sequence number 183, and sdCDR3 of sequence number 184, p) sdCDR1 of sequence number 186, sdCDR2 of sequence number 187, and sdCDR3 of sequence number 188, q) sdCDR1 of sequence number 190, sdCDR2 of sequence number 191, and sdCDR3 of sequence number 192, r) sdCDR1 of sequence number 194, sdCDR2 of sequence number 195, and sdCDR3 of sequence number 196, s) sdCDR1 of sequence number 198, sdCDR2 of sequence number 199, and sdCDR3 of sequence number 200, t) sdCDR1 of sequence number 202, sdCDR2 of sequence number 203, and sdCDR3 of sequence number 204, u) sdCDR1 of sequence number 206, sdCDR2 of sequence number 207, and sdCDR3 of sequence number 203, v) sdCDR1 of sequence number 210, sdCDR2 of sequence number 211, and sdCDR3 of sequence number 212, w) sdCDR1 of sequence number 214, sdCDR2 of sequence number 215, and sdCDR3 of sequence number 216, x) sdCDR1 of sequence number 218, sdCDR2 of sequence number 219, and sdCDR3 of sequence number 220, y) sdCDR1 of sequence number 222, sdCDR2 of sequence number 223, and sdCDR3 of sequence number 224, z) sdCDR1 of sequence number 226, sdCDR2 of sequence number 227, and sdCDR3 of sequence number 228, aa) sdCDR1 of sequence number 230, sdCDR2 of sequence number 231, and sdCDR3 of sequence number 232, ab) sdCDR1 of sequence number 234, sdCDR2 of sequence number 235, and sdCDR3 of sequence number 236, ac) sdCDR1 of sequence number 238, sdCDR2 of sequence number 239, and sdCDR3 of sequence number 240, ad) sdCDR1 of sequence number 242, sdCDR2 of sequence number 243, and sdCDR3 of sequence number 244, ae) sdCDR1 of sequence number 500, sdCDR2 of sequence number 501, and sdCDR3 of sequence number 502, af) sdCDR1 of sequence number 504, sdCDR2 of sequence number 505, and sdCDR3 of sequence number 506, ag) sdCDR1 of sequence number 508, sdCDR2 of sequence number 509, and sdCDR3 of sequence number 510, and ah) The fusion protein according to 1, having an amino acid sequence comprising a set of CDRs selected from the group consisting of sdCDR1 of SEQ ID NO: 512, sdCDR2 of SEQ ID NO: 513, and sdCDR3 of SEQ ID NO: 5. [3] The fusion protein according to 1, wherein the first sdABD-HER2 and / or the second sdABD-HER2 contains an amino acid sequence selected from the group consisting of SEQ ID NOs: 193, 217, 225, 237, 141, 145, 149, 153, 157, 161, 165, 169, 173, 177, 181, 185, 189, 197, 201, 205, 209, 213, 221, 229, 233, 241, 499, 503, 507, and 511. [4] The fusion protein described in any one of the above 1 to 3, wherein the first sdABD-HER2 and the second sdABD-HER2 are the same. [5] The fusion protein described in 1 or 3 above, wherein the first sdABD-HER2 and the second sdABD-HER2 are different. [6] The fusion protein according to any one of claims 1 to 5, wherein the first variable weight domain is located at the N-terminus of the first variable light domain, and the pseudo-variable light domain is located at the N-terminus of the pseudo-variable weight domain. [7] The fusion protein according to any one of claims 1 to 5, wherein the first variable weight domain is located at the N-terminus of the first variable light domain, and the pseudo-variable weight domain is located at the N-terminus of the pseudo-variable light domain. [8] The fusion protein according to any one of claims 1 to 5, wherein the first variable light domain is located at the N-terminus of the first variable heavy domain, and the pseudo-variable light domain is located at the N-terminus of the pseudo-variable heavy domain. [9] The fusion protein according to any one of claims 1 to 5, wherein the first variable light domain is located at the N-terminus of the first variable heavy domain, and the pseudo-variable heavy domain is located at the N-terminus of the pseudo-variable light domain.

[10] The third sdABD (sdABD-HSA) that binds to HSA is, (a)(i) a set of CDRs selected from the group consisting of sdCDR1 of sequence number 246, sdCDR2 of sequence number 247, and sdCDR3 of sequence number 248, and (ii) sdCDR1 of sequence number 250, sdCDR2 of sequence number 251, and sdCDR3 of sequence number 252, or (b) A fusion protein according to any one of items 1 to 9, having an amino acid sequence that includes an amino acid sequence selected from the group consisting of SEQ ID NO: 245 and SEQ ID NO: 249.

[11] The cleavable linker comprises a cleavage domain sequence selected from the group consisting of SEQ ID NOs: 339-408 and SEQ ID NOs: 532-535, wherein the fusion protein is according to any one of claims 1 to 10.

[12] The cleavable linker is c...

Claims

1. A single-domain antigen-binding domain (sdABD) (sdABD-HER2) that binds to human HER2, comprising a set of CDR amino acid sequences including sdCDR1 of SEQ ID NO: 504, sdCDR2 of SEQ ID NO: 505, and sdCDR3 of SEQ ID NO:

506.

2. The sdABD according to claim 1, comprising the amino acid sequence of SEQ ID NO:

503.

3. A fusion protein, wherein the fusion protein extends from the N-terminus to the C-terminus. a) A first sdABD (sdABD-HER2) that binds to HER2, b) The first domain linker, c) i) A first variable heavy domain including vhCDR1, vhCDR2 and vhCDR3, ii) Constrained non-cleavable linker (CNCL) and iii) A restrictive Fv domain including a first variable light domain including vlcDr1, vlcDr2 and vlcDr3, d) The second domain linker, e) The second sdABD-HER2, f) Fracture linker (CL), g) i) First pseudo-variable light domain, ii) Non-cleavable linker (NCL) and iii) A restrictive pseudo-Fv domain including a first pseudo-variable heavy domain, h) The third domain linker, i) comprising a third sdABD (sdABD-HSA) that binds to human serum albumin, Here, the first variable weight domain and the first variable light domain of the restrictive Fv domain can bind to human CD3, but the restrictive pseudo-Fv domain does not bind to CD3. The first variable heavy domain and the first pseudo-variable light domain associate intramolecularly to form an inactive Fv. The first variable light domain and the first pseudo-variable heavy domain associate intramolecularly to form an inactive Fv, and The first sdABD-HER2 and / or the second sdABD-HER2 is the sdABD described in claim 1 or 2. The aforementioned fusion protein.

4. The fusion protein according to claim 3, wherein the first sdABD-HER2 and the second sdABD-HER2 are the same, or the first sdABD-HER2 and the second sdABD-HER2 are different.

5. The first variable heavy domain is located at the N-terminus of the first variable light domain, and the pseudo-variable light domain is located at the N-terminus of the pseudo-variable heavy domain, The first variable heavy domain is located at the N-terminus of the first variable light domain, and the pseudo-variable heavy domain is located at the N-terminus of the pseudo-variable light domain, The first variable light domain is located at the N-terminus of the first variable heavy domain, and the pseudo-variable light domain is located at the N-terminus of the pseudo-variable heavy domain, or The first variable light domain is located at the N-terminus of the first variable heavy domain, and the pseudo-variable heavy domain is located at the N-terminus of the pseudo-variable light domain. The fusion protein according to claim 3 or 4.

6. The third sdABD (sdABD-HSA) that binds to the HSA is, (a) (i) a set of CDRs selected from the group consisting of sdCDR1 of sequence number 246, sdCDR2 of sequence number 247, and sdCDR3 of sequence number 248, and (ii) sdCDR1 of sequence number 250, sdCDR2 of sequence number 251, and sdCDR3 of sequence number 252, or (b) The fusion protein according to any one of claims 3 to 5, having an amino acid sequence comprising an amino acid sequence selected from the group consisting of SEQ ID NO: 245 and SEQ ID NO:

249.

7. The fusion protein according to any one of claims 3 to 6, wherein the cleavable linker includes a cleavage domain sequence selected from the group consisting of SEQ ID NOs: 339 to 408 and SEQ ID NOs: 532 to 535.

8. The cleavable linker is cleaved by a human protease selected from the group consisting of MMP2, MMP9, meprin A, meprin B, cathepsin S, capthepsin K, cathepsin L, granzyme B, uPA, kalecryein 7, matryptase, and thrombin, according to any one of claims 3 to 7.

9. The fusion protein according to any one of claims 3 to 8, wherein the fusion protein has an amino acid sequence selected from the group consisting of SEQ ID NOs: 459 to 484 and SEQ ID NOs: 491 to 494.

10. A nucleic acid encoding a fusion protein according to any one of claims 3 to 9 or sdABD according to claim 1 or 2.

11. An expression vector comprising the nucleic acid described in claim 10.

12. A host cell comprising the expression vector described in claim 11.

13. A method for producing the fusion protein or the sdABD, comprising: (i) culturing the host cell described in claim 12 under conditions in which the fusion protein described in any one of claims 3 to 9 or the sdABD described in claim 1 or 2 is expressed; and (ii) recovering the fusion protein or the sdABD.

14. A pharmaceutical composition comprising a fusion protein according to any one of claims 3 to 9 or a single-domain antigen-binding domain (sdABD) according to claim 1 or 2.

15. The pharmaceutical composition according to claim 14, further comprising a pharmaceutically acceptable carrier or excipient.

16. A pharmaceutical composition according to claim 14 or 15 for treating a target cancer.

Images