ANTIGEN BINDING PROTEINS SPECIFICALLY BINDING PRAME

Abstract

The present invention concerns antigen binding proteins directed against PRAME protein-derived antigens. The invention in particular provides antigen binding proteins which are specific for the tumor expressed antigen PRAME, wherein the tumor antigen comprises or consists of SEQ ID NO: 50 and is in a complex with a major histocompatibility complex (MHC) protein. The antigen binding proteins of the invention contain, in particular, the complementary determining regions (CDRs) of novel engineered T cell receptors (TCRs) that specifically bind to said PRAME peptide. The antigen binding proteins of the invention are for use in the diagnosis, treatment and prevention of PRAME expressing cancerous diseases. Further provided are nucleic acids encoding the antigen binding proteins of the invention, vectors comprising said nucleic acids, recombinant cells expressing the antigen binding proteins and pharmaceutical compositions comprising the antigen binding proteins of the invention.

Claims

1. An antigen binding protein specifically binding to a PRAME antigenic peptide that comprises or consists of the amino acid sequence SLLQHLIGL of SEQ ID NO: 50 and is in a complex with a major histocompatibility complex (MHC) protein, the antigen binding protein comprising (a) a first polypeptide comprising a variable domain VA comprising complementarity determining regions (CDRs) CDRa1, CDRa2 and CDRa3, wherein the CDRa1 comprises or consists of the amino acid sequence VKEFQD (SEQ ID NO: 16), or an amino acid sequence differing from SEQ ID NO: 16 by one, two or three amino acid mutations, optionally amino acid substitutions, and the CDRa3 comprises or consists of the amino acid sequence of ALYNNLDMR (SEQ ID NO: 33) or ALYNNYDMR (SEQ ID NO: 34), or an amino acid sequence differing from SEQ ID NO: 33 or SEQ ID NO: 34 by one, two or three, optionally one or two, amino acid mutations, optionally amino acid substitutions, and optionally the CDRa2 comprises or consists of the amino acid sequence FGPYGKE (SEQ ID NO: 32), or an amino acid sequence differing from SEQ ID NO: 32 by one, two or three amino acid mutations, optionally amino acid substitutions, and (b) a second polypeptide comprising a variable domain V.sub.B comprising CDRb1, CDRb2 and CDRb3, wherein the CDRb1 comprises or consists of the amino acid sequence SGHNS (SEQ ID NO: 10) or an amino acid sequence differing from SEQ ID NO: 10 by one or two amino acid mutations, optionally amino acid substitutions, and the CDRb3 comprises or consists of the amino acid sequence ASSX1GX2X3DX4QY (SEQ ID NO: 327), wherein X1 is P, A or T, X2 is A or S, X3 is T or I, and X4 is K or A, or an amino acid sequence differing from SEQ ID NO: 327 by one, two or three amino acid mutations, optionally amino acid substitutions, and optionally the CDRb2 comprises or consists of the amino acid sequence FQNTAV (SEQ ID NO: 36) or a CDRb2 amino acid sequence differing from SEQ ID NO: 36 by one, two, three, four, five or six amino acid mutations, optionally amino acid substitutions.

2. The antigen binding protein of claim 1, wherein said antigen binding protein specifically binds to a functional epitope comprising or consisting of at least 3, 4 or 5 amino acid positions selected from the group consisting of positions 3, 5, 6, 7 and 8, in particular 3, 5 and 7, of SEQ ID NO: 50, optionally to a functional epitope consisting of amino acid positions 3, 5 and 7, or 3, 5, 6 and 7, or 3, 5, 7 and 8, or 3, 5, 6, 7 and 8 of SEQ ID NO: 50, but optionally not amino acid positions 1 and 4 of SEQ ID NO: 50, or specifically binds to a functional epitope comprising or consisting of at least 6 or 7 amino acid positions selected from the group consisting of positions 1, 3, 4, 5, 6, 7 and 8 of SEQ ID NO: 50.

3. The antigen binding protein of claim 1, wherein said antigen binding protein does not significantly bind to at least 1, at least 2, at least 3, at least 4, at least 5, at least 10, at least 20 or all similar peptides selected from the group consisting of TMED9-001 (SEQ ID NO: 51), CAT-001 (SEQ ID NO: 52), DDX60L-001 (SEQ ID NO: 53), LRRC70-001 (SEQ ID NO: 54), PTPLB-001 (SEQ ID NO: 55), HDAC5-001 (SEQ ID NO: 56), VPS13B-002 (SEQ ID NO: 57), ZNF318-001 (SEQ ID NO: 58), CCDC51-001 (SEQ ID NO: 59), IFT17-003 (SEQ ID NO: 60), DIAPH1-004 (SEQ ID NO: 62), FADS2-001 (SEQ ID NO: 63), FRYL-003 (SEQ ID NO: 64), GIMAP8-001 (SEQ ID NO: 65), HSF1-001 (SEQ ID NO: 66), KNT-001 (SEQ ID NO: 67), MAU-001 (SEQ ID NO: 68), MCM4-001 (SEQ ID NO: 69), MPPE1-001 (SEQ ID NO: 71), MYO1B-002 (SEQ ID NO: 72), PRR12-001 (SEQ ID NO: 73), PTRF-003 (SEQ ID NO: 74), RASGRP1-001 (SEQ ID NO: 75), SMARCD1-001 (SEQ ID NO: 76), TGM2-001 (SEQ ID NO: 77), VAV1-001 (SEQ ID NO: 78), VIM-009 (SEQ ID NO: 317), FARSA-001 (SEQ ID NO: 306), ALOX15B-003 (SEQ ID NO: 304), FAM114A2-002 (SEQ ID NO: 305), GPR56-002 (SEQ ID NO: 307), IGHD-002 (SEQ ID NO: 308), NOMAP-3-0972 (SEQ ID NO: 309), NOMAP-3-1265 (SEQ ID NO: 310), NOMAP-3-1408 (SEQ ID NO: 311), NOMAP-3-1587 (SEQ ID NO: 312), NOMAP-3-1768 (SEQ ID NO: 313), NOMAP-5-0765 (SEQ ID NO: 314), PDCD10-004 (SEQ ID NO: 315), TSN-001 (SEQ ID NO: 316), ARMC9-002 (SEQ ID NO: 187), CLI-001 (SEQ ID NO: 188), COPG1-001 (SEQ ID NO: 190), COPS7A-001 (SEQ ID NO: 192), EIF-009 (SEQ ID NO: 194), EXT2-006 (SEQ ID NO: 196), LMNA-001 (SEQ ID NO: 198), PKM-005 (SEQ ID NO: 200), PSMB3-002 (SEQ ID NO: 202), RPL-007 (SEQ ID NO: 204), SPATS2L-003 (SEQ ID NO: 206), SYNE1-002 (SEQ ID NO: 208), TGM2-002 (SEQ ID NO: 210) and TPR-004 (SEQ ID NO: 212), in a complex with a MHC protein, optionally said antigen binding protein does not significantly bind to IFT17-003 (SEQ ID NO: 60) in a complex with a MHC protein.

4. The antigen binding protein of claim 1, wherein the antigen binding protein is multispecific, e.g. tetra-, tri- or bispecific, optionally bispecific, in particular said antigen binding protein is a bispecific TCR, a bispecific antibody or a bispecific TCR-antibody molecule.

5. The antigen binding protein of claim 1, wherein the first and the second polypeptide are comprised in a single polypeptide chain or two polypeptide chains, optionally wherein VA is comprised in a first polypeptide chain and VB is comprised in a second polypeptide chain; and/or the variable domain VA is a Vα or Vγ domain and the variable domain VB is a Vβ or Vδ domain.

6. The antigen binding protein of claim 1, wherein VA further comprises one or more framework regions, optionally all framework regions, selected from the group consisting of FR1-a, FR2-a, FR3-a and FR4-a, wherein FR1-a comprises or consists of the amino acid sequence of SEQ ID NO: 345 or SEQ ID NO: 346, or an amino acid sequence at least 85%, 90% or 95% identical to SEQ ID NO: 345, optionally comprising K or N, optionally K, at position 20 and/or L or M, optionally L, at position 2; FR2-a comprises or consists of the amino acid sequence of SEQ ID NO: 347 or SEQ ID NO: 348, or an amino acid sequence at least 85%, 90% or 95% identical to SEQ ID NO: 347, optionally comprising L, I or M, optionally L or I, at position 39, A or D, optionally A, at position 47, K or W, optionally K, at position 44, F or A, optionally F, at position 52 and/or Y or V, preferably Y, at position 55; FR3-a comprises or consists of the amino acid sequence of SEQ ID NO: 349 or an amino acid sequence at least 85%, 90% or 95% identical to SEQ ID NO: 349, optionally comprising T or K, optionally T, at position 92 and/or D or G, optionally D, at position 93; FR4-a comprises or consists of the amino acid sequence of SEQ ID NO: 350 or an amino acid sequence at least 85%, 90% or 95% identical to SEQ ID NO: 350; and VB further comprises one or more framework regions, optionally all framework regions, selected from the group consisting of FR1-b, FR2-b, FR3-b and FR4-b, wherein FR1-b comprises or consists of the amino acid sequence of SEQ ID NO: 351 or SEQ ID NO: 352 or an amino acid sequence at least 85%, 90% or 95% identical to SEQ ID NO: 351, optionally comprising H or N, optionally H, at position 10, E, L or K, preferably E, at position 11 and/or R or H, at position 22; FR2-b comprises or consists of the amino acid sequence of SEQ ID NO: 353 or an amino acid sequence at least 85%, 90% or 95% identical to SEQ ID NO: 353, optionally comprising R or K, optionally R, at position 43, E or Q, optionally E, at position 44, M or P, optionally P, at position 46, and/or R or Q, optionally Q, at position 48; FR3-b comprises or consists of the amino acid sequence of SEQ ID NO: 354 or SEQ ID NO: 355 or an amino acid sequence at least 85%, 90% or 95% identical to SEQ ID NO: 354, optionally comprising D, A, E, R, K Q, N or S, preferred optionally D, A, E, Q, N or S, optionally D or A, optionally D, at position 84; and FR4-b comprises or consists of the amino acid sequence of SEQ ID NO: 356 or an amino acid sequence at least 85%, 90% or 95% identical to SEQ ID NO: 356.

7. The antigen binding protein of claim 1, wherein VA comprises or consists of the amino acid sequence of SEQ ID NO: 132 or an amino acid sequence at least 85%, 90% or 95% identical to SEQ ID NO: 132, optionally comprising a CDRa1 of SEQ ID NO: 16, a CDRa2 of SEQ ID NO: 32 and a CDRa3 of SEQ ID NO: 33, SEQ ID NO: 34, or SEQ ID NO: 9, and further K or N, optionally K, at position 20, L, M, or I, optionally L or I, at position 39, K or W, optionally K, at position 44, F or A, optionally F, at position 52, Y or V, optionally Y, at position 55, T or K, optionally T, at position 92 and/or D or G, optionally D, at position 93, in particular VA comprises or consists of the amino acid sequence of SEQ ID NO: 132, SEQ ID NO: 129, SEQ ID NO: 137 or SEQ ID NO: 142; and VB comprises or consists of the amino acid sequence of SEQ ID NO: 134 or an amino acid sequence at least 85%, 90% or 95% identical to SEQ ID NO: 134, optionally comprising a CDRb1 of SEQ ID NO: 10, a CDRb2 of SEQ ID NO: 36, and a CDRb3 of SEQ ID NO: 48, SEQ ID NO: 49, SEQ ID NO: 47, SEQ ID NO: 281, SEQ ID NO: 292, SEQ ID NO: 294, SEQ ID NO: 297, SEQ ID NO: 298, SEQ ID NO: 301 or SEQ ID NO: 283, and further E, L or K, optionally E, at position 11, R or H at position 22, E or Q, optionally E, at position 44, P or M, optionally P, at position 46, Q or R, optionally Q, at position 48 and/or D, A, E, Q, N, or S, optionally D or A, at position 84, in particular V.sub.B comprises or consists of the amino acid sequence of SEQ ID NO: 134, SEQ ID NO: 130, SEQ ID NO: 135, SEQ ID NO: 136, SEQ ID NO: 138, SEQ ID NO: 139, SEQ ID NO: 140, SEQ ID NO: 141, SEQ ID NO: 144, SEQ ID NO: 145, SEQ ID NO: 146, SEQ ID NO: 147 or SEQ ID NO: 148.

8. The antigen binding protein of claim 1, further comprising an antibody light chain variable domain (V.sub.L) and an antibody heavy chain variable domain (V.sub.H), wherein optionally wherein VL and VH bind to an antigen selected from the group consisting of CD2, CD3, in particular CD37, CD36, and/or CD3E, CD4, CD5, CD7, CD8, CD10, CD11b, CD11c, CD14, CD16, CD18, CD22, CD25, CD28, CD32a, CD32b, CD33, CD41, CD41b, CD42a, CD42b, CD44, CD45RA, CD49, CD55, CD56, CD61, CD64, CD68, CD90, CD94, CD95, CD117, CD123, CD125, CD134, CD137, CD152, CD163, CD193, CD203c, CD235a, CD278, CD279, CD287, Nkp46, NKG2D, GITR, FcεRI, TCRα/β and TCRγ/δ, HLA-DR and 4-1 BB, or combinations thereof and/or bind to an effector cell, in particular a T cell or natural killer cell.

9. The antigen binding protein of claim 8, wherein the antigen binding protein comprises a first and a second polypeptide chain, and wherein the first polypeptide chain is represented by a formula [Ia]:
V1-L1-D1-L2-V2-L3-D2  [Ia], and the second polypeptide chain is represented by a formula [IIa]
V3-L4-D3-L5-V4-L6-D4  [IIa], wherein V1, V2, V3, and V4 are variable domains, wherein one of V1 to V4 is VA, one is VB, one is VL and one is VH; D1, D2, D3, and D4 are dimerization domains and may be present or absent, wherein D1 and D3, and D2 and D4, specifically bind to each other and at least one pair of D1 and D3, or D2 and D4 is present; and L1, L2, L3, L4, L5, and L6 are linkers, wherein L1 and L4 are present and L2, L3, L5, and L6 may be present or absent.

10. The antigen binding protein of claim 1, comprising a first polypeptide chain selected from SEQ ID NO: 100, 103, 105, 106, 111, 122, 126, 128, 151, 155, 156, 157, 158, 159, 166, 167, 169, 171, 173, 175, 177, 178, 179, 180, 181, 183, 189, 191, 193, 195, 197, 199, 201, 203, 205, 207, 209, 211, 213, 215, 217, 285, 291, 295, 299 and 303, and a second polypeptide chain selected from SEQ ID NO: 101, 102, 104, 107, 110, 119, 121, 131, 133, 143, 152, 160, 161, 162, 163, 164, 165, 168, 170, 172, 174, 176, 182, 184, 185, 186, 216, 218, 220, 222, 224, 228, 230, 232, 234, 236, 238, 240, 242, 244, 246, 248, 250, 252, 254, 256, 258, 260, 262, 264, 266, 268, 270, 272, 274, 276, 278, 282, 284, 296 or 300.

11. An isolated nucleic acid comprising a sequence encoding the antigen binding protein of claim 1.

12. A vector comprising the nucleic acid of claim 11.

13. A host cell comprising the vector of claim 12, optionally wherein the host cell is a lymphocyte, optionally a T lymphocyte or T lymphocyte progenitor cell, or a cell for recombinant expression, optionally a Chinese Hamster Ovary (CHO) cell or a yeast cell.

14. A pharmaceutical composition comprising the antigen binding protein of of claim 1 and a pharmaceutically acceptable carrier.

15. A method of producing the antigen binding protein according to claim 1, comprising a. providing a host cell, b. providing a genetic construct comprising a coding sequence encoding the antigen binding protein, c. introducing said genetic construct into said host cell, and d. expressing said genetic construct by said host cell, e. optionally further comprising the isolation and purification of the antigen binding protein from the host cell and, optionally, reconstitution of the antigen binding protein in a T cell.

16. A method of treating a patient who has cancer, comprising administering to the patient the pharmaceutical composition of claim 14, wherein said cancer is selected from the group of cancers consisting of acute myeloid leukemia, breast cancer, cholangiocellular carcinoma, gallbladder cancer, glioblastoma, hepatocellular carcinoma, head and neck squamous cell carcinoma, melanoma, amelanotic melanoma, non-Hodgkin lymphoma, non-small cell lung cancer adenocarcinoma, non-small cell lung cancer, squamous cell non-small cell lung cancer, ovarian cancer, esophageal cancer, renal cell carcinoma, small cell lung cancer, urinary bladder carcinoma, uterine and endometrial cancer, osteosarcoma, chronic lymphocytic leukemia, colorectal carcinoma, and synovial sarcoma.

17. The antigen binding protein of claim 2, wherein said antigen binding protein specifically binds to a functional epitope comprising or consisting of at least 3, 4 or 5 amino acid positions selected from the group consisting of positions 3, 5, 6, 7 and 8 of SEQ ID NO: 50.

18. The antigen binding protein of claim 2, wherein said antigen binding protein specifically binds to a functional epitope comprising or consisting of at least 3, 4 or 5 amino acid positions selected from the group consisting of positions 3, 5 and 7, of SEQ ID NO: 50.

19. The antigen binding protein of claim 2, wherein said antigen binding protein specifically binds to a functional epitope comprising or consisting of at least 3, 4 or 5 amino acid positions selected from the group consisting of positions 3, 5 and 7, or 3, 5, 6 and 7, or 3, 5, 7 and 8, or 3, 5, 6, 7 and 8 of SEQ ID NO: 50, but not amino acid positions 1 and 4 of SEQ ID NO: 50.

20. The antigen binding protein of claim 2, wherein said antigen binding protein specifically binds to a functional epitope comprising or consisting of at least 6 or 7 amino acid positions selected from the group consisting of positions 1, 3, 4, 5, 6, 7 and 8 of SEQ ID NO: 50.

Description

BRIEF DESCRIPTION OF THE FIGURES

[0630] FIG. 1: Conversion of a TCR into stabilized scTCR via yeast surface display. ScTCR molecules displayed on the surface of transformed Saccharomyces cerevisiae EBY100 were stained with anti Myc-FITC antibody to determine expression level and PE-labeled HLA-A*02/PRAME-004 tetramer to investigate functional binding. The non-modified scTCR R11P3D3 (left panel, SEQ ID NO: 5) is compared to R11P1 D3_stabilised scTCR variant bearing nine stabilizing framework mutations and three single-point mutations in the CDRs (right panel, SEQ ID NO: 6), which was derived from the selection of the scTCR library.

[0631] FIG. 2: Affinity maturation of scTCR CDR1 alpha vie yeast surface display. Stabilized scTCRs comprising non-modified and maturated CDR1 alpha were stained with HLA-A*02/PRAME-004 monomer at a concentration of 10 nM. Counterstaining with a mix of HLA-A*02/SimPep tetramers, each applied at a concentration of 10 nM, containing peptides (SEQ ID NO: 51 to 59) with high sequence similarity to PRAME-004 (SEQ ID NO: 50). Stabilized scTCR R11P3D3SD (SEQ ID NO: 6) with non-modified alpha chain CDR1 sequence SSNFYN (SEQ ID NO: 13; bottom right panel) is compared to scTCR variants comprising the affinity maturated alpha chain CDR1 sequences VKEFQD, NKEFQD, TREFQD, NREFQD, TSEFQD, TKEFQD, VREFQD, TAEFQD, VSEFQD, VAEFQD, IKEFQN, VREFQN and TAEFQN (SEQ ID NOs 16 to 28), respectively. SSNFYN (SEQ ID NO 13) is the corresponding CDRa1 sequence of the stabilized scTCR R11P3D3SD.

[0632] FIG. 3: Binding of high affinity scTCR yeast clones to similar peptides. Yeast clones bearing stabilized scTCRs with maturated CDRs (SEQ ID NOs: 79 to 87 and 89 to 92) were stained with 100 nM HLA-A*02 monomers containing the PRAME-004 target peptide or one out of 7 similar peptides (SEQ ID NOs: 52 to 56 and 58 to 59).

[0633] FIG. 4: Binding of high affinity scTCR yeast clones to similar peptides. Yeast clones bearing stabilized scTCRs with maturated CDRs (SEQ ID NOs: 79 to 87 and 89 to 92) were stained with 100 nM HLA-A*02 monomers containing the PRAME-004 target peptide or one out of 19 similar peptides (SEQ ID NOs 51, 57, 60, 62 to 69 and 71 to 78). R16P1C10_CDR6_scTCR (SEQ ID NO 357) was added as reference, but only binding to PRAME-004 and IFT17-003 (SEQ ID NO 60) was assessed for this done.

[0634] FIG. 5: Binding motif determination with high affinity scTCR yeast clones. Yeast clones bearing stabilized scTCRs with maturated CDRs (SEQ ID NOs: 79, 80, 82, 83 and 85 to 87) were stained with PRAME-004 as well as with PRAME-004 peptide variants containing alanine substitutions (SEQ ID NOs 318 to 324) in the context of HLA-A*02 applied at concentrations of 10 nM, 3 nM, 1 nM and 0.3 nM.

[0635] FIG. 6: Similar peptide screening for soluble scTCR-Fab molecules. Binding to 14 similar peptides (SEQ ID NOs: 187, 188, 190, 192, 194, 196, 198, 200, 202, 204, 206, 208, 210 and 212) in the context of HLA-A*02 was analyzed at a concentration of 1 μM scTCR-Fab using bio-layer interferometry. Upper curve in each graph represents scTCR-Fab binding to the target HLA-A*02/PRAME-004 monomer.

[0636] FIG. 7: In vitro cytotoxicity of TCER® molecules on target-positive and target-negative tumor cell lines. PBMC from a healthy HLA-A*02-positive donor were incubated with either target-positive tumor cell line Hs695T (.circle-solid.) or target-negative, but HLA-A*02-positive tumor cell line T98G (∘), respectively, at a ratio of 1:10 in the presence of increasing TCER® concentrations. TCER®-induced cytotoxicity was quantified after 48 hours of co-culture by measurement of released LDH. Results for experiments assessing TPP-93 and TPP-79 are shown in the upper and lower panel, respectively.

[0637] FIG. 8: In vitro cytotoxicity of TCER® molecule TPP-105 on target-positive and target-negative tumor cell lines. PBMC from a healthy HLA-A*02-positive donor were incubated with either target-positive tumor cell line Hs695T (.circle-solid.) or target-negative, but HLA-A*02-positive tumor cell line T98G (∘), respectively, at a ratio of 1:10 in the presence of increasing concentrations of TPP-105. TCER®-induced cytotoxicity was quantified after 48 hours of co-culture by measurement of released LDH.

[0638] FIG. 9: Summary of cytotoxicity data of TCER® Slot III molecules. EC.sub.50 values of dose-response curves obtained in LDH-release assays were calculated utilizing non-linear 4-point curve fitting. For each assessed TCER®-molecule calculated EC.sub.50 values on target-positive tumor cell lines Hs695T (.circle-solid.), U2OS (◯), and target-negative but HLA-A*02-positive tumor cell line T98G (*) are depicted. Thereby, each symbol represents one assay utilizing PBMC derived from various HLA-A*02-positive donors. For TPP-8711T980, the EC.sub.50 is estimated, as T98G was not recognized by TPP-871.

[0639] FIG. 10: In vitro cytotoxicity of TCER® Slot III variants on T2 cells loaded with different concentrations of target peptide. Cytotoxicity was determined by quantifying LDH released into the supernatants. Human PBMC were used as effector cells at an E:T ratio of 5:1. Read out was performed after 48 h.

[0640] FIG. 11: Normal tissue cell safety analysis for selected TCER® Slot III variants. TCER®-mediated cytotoxicity against 5 different normal tissue cell types expressing HLA-A*02 was assessed in comparison to cytotoxicity directed against PRAME-004-positive Hs695T tumor cells. PBMCs from a healthy NIA-A*02+ donor were co-cultured at a ratio 10:1 with the normal tissue cells or Hs695T tumor cells (in triplicates) in a 1:1 mixture of the respective normal tissue cell medium (4, 10a or 13a) and T cell medium (LDH-AM) or in T cell medium alone. After 48 hours, lysis of normal tissue cells and Hs695T cells was assessed by measuring LDH release (LDH-Glo™ Kit, Promega).

[0641] FIG. 12: Normal tissue cell safety analysis for selected TCER® Slot IV variants. TCER®-mediated cytotoxicity against 10 different normal tissue cell types expressing HLA-A*02 was assessed in comparison to cytotoxicity directed against PRAME-004-positive Hs695T tumor cells. PBMCs from a healthy HLA-A*02+ donor were co-cultured at a ratio 10:1 with the normal tissue cells or Hs695T tumor cells (in triplicates) in a 1:1 mixture of the respective normal tissue cell medium (3, 4, 8a, 10a, 13a or 16a) and T cell medium (LDH-AM) or in T cell medium alone. After 48 hours, lysis of normal tissue cells and Hs695T cells was assessed by measuring LDH release (LDH-Glo™ Kit, Promega).

[0642] FIG. 13: Normal tissue cell safety analysis for selected TCER® Slot IV variants. TCER®-mediated cytotoxicity against 6 different normal tissue cell types expressing HLA-A*02 was assessed in comparison to cytotoxicity directed against PRAME-004-positive Hs695T tumor cells. PBMCs from a healthy HLA-A*02+ donor were co-cultured at a ratio 10:1 with the normal tissue cells or Hs695T tumor cells (in triplicates) in a 1:1 mixture of the respective normal tissue cell medium (10a, 13a or 16a) and T cell medium (LDH-AM). After 48 hours, lysis of normal tissue cells and Hs695T cells was assessed by measuring LDH release (LDH-Glo™, Kit, Promega).

EXAMPLES

Example 1: Single Chain TCR (scTCR Format)

Example 1.1: Generation of Stable scTCR

[0643] For the present invention, the TCR R11P3D3 (SEQ ID NOs: 1 and 2, full length) was converted into a single chain TCR construct (scTCR R11P3D3, SEQ ID NO: 5) using the variable alpha (SEQ ID NO: 3) and beta (SEQ ID NO: 4) domains and an appropriate glycine-serine linker sequence (SEQ ID NO: 61). For TCR maturation via yeast surface display, the DNA of the corresponding sequence was synthesized and transformed into Saccharomyces cerevisiae EBY100 (MATa AGA1::GAL1¬AGA1::URA3 ura3¬52 trp1 leu2¬detta200 his3¬delta200 pep4::HIS3 prbd1.6R can1 GAL) (ATCC® MYA¬4941™) together with a yeast display vector based on pCT302 (Boder and Wittrup, Methods Enzymol. 2000; 328:430-44:). The resulting fusion protein after homologous recombination in the yeast (SEQ ID NO: 325) contains a leader peptide at the N-terminus of the Aga2p protein (SEQ ID NO: 88) (Boder and Wittrup, Nat Biotechnol. 1997 Jun.; 15(6):553-7), the protein of interest, namely the scTCR R11P3D3 (SEQ ID NO: 5) or its variants and additional peptide tags (FLAG and Myc (SEQ ID NOs 99 and 288)) to determine the expression level of the fusion protein. Libraries of scTCR variants were generated via PCR using degenerate primers and the transformation of yeast cells was performed as described in WO 2018/091396 and resulted in up to 10.sup.9 yeast Bones per library.

[0644] The selection process for the yeast clones bearing mutant scTCR variants with improved binding to PRAME-004 in the context of HLA-A*02 was essentially performed as described in Smith et al. (Methods Mol Biol. 2015; 1319:95-141). Expression as determined by Myc tag-FITC staining and in particular functional binding via HLA-A*02/PRAME-004 tetramer staining was applied to select for most promising candidates (FIG. 1). The scTCR conversion by yeast surface display revealed nine framework mutations in combination with three single point CDR mutations, resulting in the stabilized scTCR R11P3D3SD (SEQ ID NO: 6) showing improved expression as well as HLA-A*02/PRAME-004 tetramer binding.

Example 1.2: Affinity Maturation of Stabilized scTCR, Binding Motif and Specificity Assessment

[0645] To generate scTCR molecules with higher binding affinity towards HLA-A*02/PRAME-004, all CDRs were maturated individually, using the previously identified stabilized scTCR R11P3D3SD (SEQ ID NO: 6). The CDR residues were randomized by using degenerate DNA oligo primers essentially as described previously (Smith et al, Methods Mol Biol. 2015; 1319:95-141). The resulting DNA libraries were transformed as described in example 1.

[0646] For the selection of affinity enhanced and specific R11P3D3SD scTCR variants, decreasing concentrations of HLA-A*02/PRAME-004 tetramer or monomer were used for each selection round. After four selection rounds, single scTCR clones were isolated and sequenced, resulting in a multitude of affinity maturated CDR sequences. As exemplarily shown for scTCR with maturated CDRa1 sequences (SEQ ID NOs: 16 to 28, FIG. 2), a strong improvement in binding of HLA-A*02/PRAME-004 monomers could also be demonstrated for scTCR with maturated CDRa2 and CDRb2 (SEQ ID NOs 29 to 32 and 35 to 45, Table 3). The selectivity of HLA-A*02/PRAME-004 binding was retained during maturation as confirmed by the low binding of the scTCR to a mix of HLA-A*02 tetramers containing peptides (similar peptides or SimPeps) with high degree of sequence similarity to PRAME-004 peptide (SEQ ID NO: 50). All selected scTCR maturation variants showed substantial staining with HLA-A*02/PRAME-004 monomers at a concentration of 10 nM, while the non-maturated stabilized scTCR R11P3D3SD as reference did not show staining (FIG. 2 and Table 3). Furthermore, binding of maturated scTCR to a mix of similar peptides, applied in a high avidity format of HLA-A*02 tetramers at a concentration of 10 nM, could not be detected or showed only low signals in comparison to HLA-A*02/PRAME-004 monomer binding, which confirms the capability of the scTCR maturation variants to bind the PRAME-004 target peptide in a highly specific manner.

TABLE-US-00004 TABLE 3 Binding data of yeast-bearing scTCRs with mutant CDR2s. Stabilized scTCR comprising non-modified and maturated CDR2 alpha and CDF-12 beta were stained with 10 nM HLA- A*2/PRAME-004 monomer and counterstained with a mix of HLA-A*02 tetramers, each applied at a concentration of 10 nM, containing peptides (similar peptides or SimPeps, SEQ ID NO: 51 to 59) with high sequence similarity to PRAME-004 (SEQ ID NO: 50). Yeast cells stained positive with Yeast cells stained positive with CDRa2 HLA-A*02/ HLA-A*02/ CDRb2 HLA-A*02/ HLA-A*02/ SEQ PRAME-004, SimPep, SEQ PRAME-004, SimPep, Sequence ID NO monomer tetramer mix Sequence ID NO monomer tetramer mix, FGPYGKE 32 61.0% 8.1% YQNTAV 37 66.9% 3.8% FGPYGRE 30 59.0% 6.6% YQNTAL 38 51.6% 3.3% FGPYGTE 31 64.5% 10.9% FQNTAT 39 57.4% 3.8% FGPYGVE 29 54.5% 5.7% MQNSAV 40 69.2% 4.2% MTSNGDE* 14 3.6% 3.3% FQNTAL 41 62.0% 5.5% MQNTAI 42 60.7% 4.6% LQNTAV 43 60.5% 3.3% MQNTAV 44 58.0% 4.4% YQNTAI 35 51.7% 2.9% FQNTAV 36 66.9% 3.3% FNNNEP* 15 1.9% 2.5% *corresponding CDR from R11P3D3SD_scTCR (SEQ ID NO: 6)

[0647] To further increase the affinity of scTCR clones, maturated CDRs identified in above-described CDR libraries were systematically combined in one DNA library and transformed into Saccharomyces cerevisiae EBY100 as described in example 1.1. This library was selected using HLA-A*02/PRAME-004 monomer and scTCR from single yeast clones were sequenced and analyzed regarding their binding towards HLA-A*02 monomers containing either the PRAME-004 target peptide or one peptide derived from the group of 26 peptides (similar peptides) sharing sequence similarities with PRAME-004 (SEQ ID NOs: 51-60, 62 to 69 and 71 to 78). All the selected high affinity scTCR variants (SEQ ID NOs 79 to 87 and 89 to 92) bound strongly to HLA-A*02/PRAME-004 monomer with binding EC.sub.50 values in the low nanomolar or sub-nanomolar range (Table 4), as calculated by non-linear 4-point curve fitting. With the exception of SMARCD1-001 (SEQ ID NO: 76) that provoked a binding signal slightly above background (FIG. 4), none of the scTCR variants (SEQ ID NOs 79 to 87 and 89 to 92) exhibited binding above background levels to any of the similar peptides (SEQ ID NOs: 51-60, 62 to 69 and 71 to 78) in the context of HLA-A*02 monomers applied at a concentration of 100 nM (FIG. 3, FIG. 4. Table 4). The presented data confirm the high binding specificity of the scTCR variants with combined CDR mutations whose binding properties were superior to the reference scTCR (R16P1C10 CDR6_scTCR, SEQ ID NO 357) that showed strong binding to IFT17-003 (SEQ ID NO 60) at a level indistinguishable from PRAME-004 binding (FIG. 4).

[0648] A set of selected high affinity scTCRs from yeast surface display was further examined regarding their functional epitope on the target peptide in context of the HLA-A*02 presentation, called binding motif. This was addressed by single alanine substitutions of positions 1, 3, 4, 5, 6, 7 and 8 of the PRAME-004 target peptide (SEQ ID NOs 318 to 324) and assessment of binding of scTCR-bearing yeast cells to the respective PRAME-004 peptide variants in context of HLA-A*02. Four concentrations (10 nM, 3 nM, 1 nM, 0.3 nM) of HLA-A*02 monomers with PRAME-004 or the respective alanine-substituted peptides were used to stain the high affinity scTCR-bearing yeast cells and revealed a broad binding motif for all scTCR variants with strong recognition of positions 3, 5 and 7 as confirmed by the lack of staining signals at all tested monomer concentrations. For positions 6 and 8 of the PRAME-004 peptide, a contribution to the binding motif can be assumed since alanine replacements at these positions significantly reduced the staining signals, even if this was observed with lower stringency than for the positions 3, 5 and 7. For positions 1 and 4 of the PRAME-004 peptide, only a marginal or no contribution to the binding motif could be determined since alanine substitutions resulted in staining intensities nary comparable to those observed with the PRAME-004 target peptide (FIG. 5 and Table 4).

[0649] For further analysis, the five scTCR clones R11P3D3SDA7_A02_scTCR (SEQ ID NO: 79), R11P3D3SDA7_A09_scTCR (SEQ ID NO: 82), R11P3D3SDA7_A10_scTCR (SEQ ID NO: 83), R11P3D3SDA7_B03_scTCR (SEQ ID NO: 85) and R11P3D3SDA7_B06_scTCR (SEQ ID NO: 87) were subject to conversion into scTCR-Fab bispecific format in order to determine further protein features (see following example).

TABLE-US-00005 TABLE 4 Binding data of yeast-bearing scTCR and soluble scTCR-Fab molecules and respective variable chain sequences. For  scTCR-bearing yeast cells, binding towards HLA-A*02/PRAME-004 monomers is presented as EC.sub.50 values and binding  towards 26 similar peptides (SEQ ID NOs 51 to 60. 62-69 and 71 to 79) in context of 100 nM  HLA-A*02 monomer is presented as number of peptides showing no binding. Binding motif positions constituting the functional epitope of PRAME-004 were determined by alanine scanning and positions with strong and  medium (positions in brackets) impact on scTCR binding are indicated. Five soluble scTCR-Fab molecules  (TPP-70 to TPP-74) were assessed for binding affinity (K.sub.D) towards HLA-A*02/PRAME-004 monomer and for binding towards a set of 14 similar peptides (see example 2). n.d. not determined scTCR on yeast cells soluble EC50  scTCR-Fab for HLA- K.sub.D Simi- FRa FRb A*02/ for HLA- lar Se- mutations mutations PRAME- Similar Binding A*02/ pep- quence (compared (compared 004 peptides motif PRAME-004 tides Molecule ID to parental to parental binding without pos- binding without name NOs TCR) CDRa1 CDRa2 CDRa3 TCR) CDRb1 CDRb2 CDRb3 [nM] binding itions nM] binding R11P3D3SD_ 6 W44K, A52F, SSNFYN MTSNGDE ALYNNNDMR L11E, Q44E, SGHNS FNNNEP ASSPGSTDTQY n.d. n.d. n.d. n.d. n.d. stablized V55Y, K92T, M46P, R48Q scTCR G93D R11P3D3SDA7_ 79, W44K, A52F TREFQD FGPYGVE ALYNNNDMR L11E, Q44E, SGHMS FQNTAV ASSPGSTDTQY 0.53 25/26 3,5,7 11.7 14/14 A02_scTCR 93 V55Y, K92T  M46P, R48Q (6.8) and TPP-70 and  G93D 94 R11P3D3SDA7_ 80 W44K, A52F,  TKEFQD FGPYGVE ALYNNNDMR L11E, Q44E, SGHNS FQNTAV ASSPGATDTQY 0.28 25/26 3,5,7 n.d. n.d. A05_scTCR V55Y, K92T, M46P, R48Q (6.8) G93D R11P3D3SDA7_ 81 W44K, A52F,  TREFQD FGPYGKE ALYNNNDMR L11E, R22H, SGHNS FQNTAV ASSPGSTDTQY 0.33 25/26 n.d. n.d. n.d. A06_scTCR V55Y, K92T, Q44E, M46P, G93D R48Q R11P3D3SDA7_ 82, W44K, A52F,  TKEFQD FGPYGRE ALYNNNDMR L11E, Q44E, SGHNS FQNTAV ASSPGATDTQY 0.29 25/26 3,5,7 11.1 14/14 A09_scTCR 93 V55Y, K92T, M46P, R48Q (6.8) and TPP-71 and  G93D 95 R11P3D3SDA7_ 83, W44K, A52F,  SSNFYN FGPYGVE ALYNNNDMR L11E, Q44E, SGHNS FNSETV ASSPGATDTQY 0.4 25/26 3,5,7  4.38 14/14 A10_scTCR 93 V55Y, K92T, M46P, R48Q (6.8) and T99-72 and  G93D 96 R11P3D3SDA7_ 84 W44K, A52F,  NKEFQ FGPYGVE ALYNNNDMR L11E, Q44E, SGHNS YQNTAV ASSPGATDTQY 0.24 25/26 n.d. n.d. n.d. B01_scTCR V55Y, K92T, D M46P, R48Q G93D R11P3D3SDA7_ 85, W44K, A52F,  NKEFG FGPYGTE ALYNNNDMR L11E, Q44E, SGHNS FQNTAV ASSPGSTDTQY 0.31 25/26 3,5,7 12.5 14/14 B03_scTCR 93 V55Y, K92T, D M46P, R48Q (6.8) and TPP-73 and  G93D 97 R11P3D3SDA7_ 86 W44K, A52F,  SSNFYN FGPYGKE ALYNNNDMR L11E, R22H, SGHNS YQNTAI ASSPGSTDTQY 2.26 25/26 3,6,7 n.d. n.d. B04_scTCR V55Y, K92T, Q44E, M46P, (6.8) G93D R48Q R11P3D3SDA7_ 87, W44K, A52F,  VKEFQD FGPYGKE ALYNNNDMR L11E, Q44E, SGHNS FQNTAV ASSPGATDTQY 0.81 25/26 3,5,7  6.41 14/14 B06_scTCR 93 V55Y, K92T, M46P, R48Q (6.8) and TPP-74 and  G93D 98 R11P3D3SDA7_ 89 W44K, A52F,  VKEFQD FGPYGKE ALYNNNDMR H10N, R22H, SGHNS FNSETV ASSPGSTDTQY 1.42 25/26 n.d. n.d. n.d. F11_scTCR V55Y, K92T, L11E, Q44E, G93D M46P, R48Q R11P3D3SDA7_ 90 W44K, A47D,  NKEFQ FGPYGRE ALYNNNDMR L11E, R43K, SGHNS YQNTAV ASSPGATDTQY 0.65 25/26 n.d. n.d. n.d. G11_scTCR A52F, V55Y, D Q44E, M46P, K92T, G93D R48Q R11P3D3SDA7_ 91 W44K, A52F,  TREFQD FGPYGTE ALYNNNDMR L11E, Q44E, SGHNS YQNTAV ASSSGATDTQY 0.67 26/26 n.d. n.d. n.d. H08_scTCR V55Y, K92T, M46P, R48Q G93D R11P3D3SDA7_ 92 L39M, W44K, TKEFQD FGPYGVE ALYNNNDMR L11E, Q44E, SGHNS FQNTAV ASSPGSTDTQY 0.91 25/26 n.d. n.d. n.d. H09_scTCR A52F, V55Y, M46P, R48Q K92T, G93D

Example 2: Production and Characterization of Soluble scTCR-Fab Molecules

[0650] TCRs consisting of Valpha and Vbeta domains were designed, produced and tested in a single-chain (scTCR) format coupled to a Fab-fragment of a humanized UCHT1-antibody (Table 5 and Table 18). Vectors for the expression of recombinant proteins were designed as mono-cistronic, controlled by HCMV-derived promoter elements, pUC19-derivatives. Plasmid DNA was amplified in E. coli according to standard culture methods and subsequently purified using commercial-available kits (Macherey & Nagel). Purified plasmid DNA was used for transient transfection of CHO cells. Transfected CHO-cells were cultured for 10-11 days at 32° C. to 37° C.

[0651] Conditioned cell supernatant was cleared by filtration (0.22 μm) utilizing Sartoclear Dynamic. Lab Filter Aid (Sartorius). Bispecific molecules were purified using an Äkta Pure 25 L FPLC system (GE Lifesciences) equipped to perform affinity and size-exclusion chromatography in line. Affinity chromatography was performed on protein L columns (GE Lifesciences) following standard affinity chromatographic protocols. Size exclusion chromatography was performed directly after elution (pH 2.8) from the affinity column to obtain highly pure monomeric protein using Superdex 200 pg 16/600 columns (GE Lifesciences) following standard protocols. Protein concentrations were determined on a NanoDrop system (Thermo Scientific) using calculated extinction coefficients according to predicted protein sequences. Concentration was adjusted, if needed, by using Vivaspin devices (Sartorius). Finally, purified molecules were stored in phosphate-buffered saline at concentrations of about 1 mg/mL at temperatures of 2-8° C. Final product yield was calculated after completed purification and formulation.

[0652] Quality of purified bispecific molecules was determined by HPLC-SEC on MabPac SEC-1 columns (5 μm, 4×300 mm) running in 50 mM sodium-phosphate pH 6.8 containing 300 mM NaCl within a Vanquish uHPLC-System.

[0653] Stress stability testing was performed by incubation of the molecules formulated in PBS for up to two weeks at 40° C. Integrity, aggregate-content as well as monomer-recovery was analyzed by HPLC-SEC analyses as described above.

TABLE-US-00006 TABLE 5 Summary of productivity and stress stability data obtained for scTCR-Fab molecules. Monomer Final Monomer content after product content after 14 days scTCR-Fab yield production at 40° C. variant (mg/L) (%) (%) TPP-70 14.3 97.12 87.82 TPP-71 10.0 85.87 64.15 TPP-72 51.4 98.21 48.41 TPP-73 50.4 98.33 92.76 TPP-74 78.0 98.89 95.82

[0654] scTCR-Fab molecules TPP-70-TPP-74 were analyzed for their binding affinity to HLA-A*02 monomers containing the PRAME-004 target peptide via bio-layer interferometry. Measurements were performed on an Octet RED384 system using settings recommended by the manufacturer. Assays were run at a sensor offset of 3 mm and an acquisition rate of 5 Hz. Binding kinetics were measured at 30° C., and 1000 rpm shake speed using PBS, 0.05% Tween-20, 0.1% BSA as buffer. His-tagged HLA-A*02/PRAME-004 monomers were loaded onto HIS1K biosensors prior to analyzing serial dilutions of the scTCR-Fab molecules. Data evaluation was done using Octet Data Analysis HT Software. Strong binding affinities were determined for the scTCR-Fab molecules with K.sub.D values ranging from 4 nM to 12 nM (Table 4). Furthermore, the scTCR-Fab variants were screened for binding to 14 similar peptides (SEC ID NOs: 187, 188, 190, 192, 194, 196, 198, 200, 202, 204, 206, 208, 210 and 212). Screening with similar peptides was performed by bio-layer interferometry essentially as described above analyzing a high concentration of scTCR-Fab molecules of 1 μM to allow detection of weak binding signals. None of the maturated scTCR variants showed binding to any of the tested similar peptides (FIG. 6). The scTCR from TPP-74 was used for generation of bispecific molecules with alternative formats, such as the TCER® format.

Example 3: T Cell Engaging Receptor (TCER®) Format

Example 3.1: Production and Characterization of Soluble scTCR in Bispecific TCER® Format

[0655] For construction of TCER® molecules, DNA-sequences coding for V.sub.H and V.sub.L, derived from either hUCHT1 (Var17), a newly humanized version of the anti-CD3 antibody UCHT1, BMA031 (V36), a humanized antibody binding to TCR/CD3 complex, or the anti-CD3 antibody ID4 as well as sequences coding for and Valpha and Vbeta and respective linkers were obtained by gene synthesis. Resulting DNA-sequences were cloned in frame into expression vectors coding for hinge region, CH2 and CH3 domain derived from human IgG1 (Accession #: P01851. The CH2 and CH3 domains were engineered to contain different mutations (including N2970 mutation) to ablate binding to Fc gamma receptors and complement and to incorporate a knob-into-hole structure into CH3 domains with an additional interchain disulfide bond stabilization. Production, purification and characterization of TCER® molecules (Table 6, Table 18) was performed as outlined in example 2.

TABLE-US-00007 TABLE 6 Summary of productivity and stress stability data obtained for TCER ® molecules. Final product Monomer (%) TCER ® Va, Vb yield Monomer after 14 days variant (SEQ ID NO) Recruiter (mg/L) (%) at 40° C. TPP-93 129, 130 UCHT1-V17 18.8 94.49 n/a TPP-79 129, 130 BMA031(V36) 66.2 99.47 n/a TPP-105 129, 130 ID4 54.2 98.50 97.91

[0656] Functionality of TCER® molecules, with respect to killing of an HLA-A*02-positive tumor cell line presenting PRAME-004 target peptide on their cell surface (e.g. Hs695T), was assessed in LDH-release assays. In addition, an HLA-A*02-positive but PRAME-004-negative tumor cell line (e.g. T98G) was assessed to characterize unspecific or off-target activity of the TCER® variants. Tumor cell lines were co-incubated with PBMC from a healthy HLA-A*02-positive donor at a ratio of 1:10 in the presence of increasing TCER® concentrations. TCER®-induced cytotoxicity was quantified after 48 hours of co-culture by measurement of released LDH. EC.sub.50 values of dose-response curves were calculated utilizing non-linear 4-point curve fitting. Results representative for 3 TCER® molecules (Table 6, Table 18) are shown in FIG. 7 and FIG. 8. The results demonstrate that all 3 TCER® molecules utilizing different recruiting antibody domains are functional and induce T cell-mediated cytotoxicity in a strictly PRAME-004 dependent manner.

Example 3.2: Slot I

[0657] TCER® molecules were constructed utilizing VH and VL domains derived from hUCHT1 (Var17) or BMA031 (V36) as well as Valpha and Vbeta as described above (example 3.1). Production, purification and characterization of TCER® molecules (Table 7, Table 18) was performed as outlined in example 2.

TABLE-US-00008 TABLE 7 Summary of productivity and stress stability data obtained for TCER ® molecules of slot I. Final product Monomer (%) TCER ® yield Monomer after 14 days variant Recruiter (mg/L) (%) at 40° C. TPP-106 UCHT1-V17 2.92 96.97 94.11 TPP-108 UCHT1-V17 4.30 95.44 94.10 TPP-109 BMA031(V36) 34.00 97.8 93.82 TPP-110 BMA031(V36) 50.00 97.12 92.70 TPP-111 BMA031(V36) 61.30 98.04 94.46 TPP-112 UCHT1-V17 2.47 96.75 92.71 TPP-113 UCHT1-V17 2.24 97.79 95.95 TPP-114 UCHT1-V17 2.64 97.68 95.37 TPP-115 UCHT1-V17 1.80 97.84 94.15 TPP-116 UCHT1-V17 3.26 97.54 94.13 TPP-117 UCHT1-V17 3.02 97.29 94.33 TPP-118 UGHT1-V17 2.13 98.09 95.11 TPP-119 UCHT1-V17 3.04 97.56 95.18 TPP-120 UCHT1-V17 2.58 97.57 94.52 TPP-121 UCHT1-V17 2.74 97.92 92.80 TPP-122 UCHT1-V17 3.22 96.9 92.77 TPP-123 UCHT1-V17 2.48 97.16 92.62 TPP-124 UCHT1-V17 2.68 96.38 90.73 TPP-125 UCHT1-V17 2.48 96.56 92.33 TPP-126 UCHT1-V17 1.76 96.71 90.62 TPP-127 UCHT1-V17 2.68 96.37 90.95 TPP-128 UCHT1-V17 1.81 97.25 90.44 TPP-129 UCHT1-V17 1.47 96.94 89.55

[0658] TCER® Slot I variants TPP-106, TPP-108-TPP-129 were analyzed for their binding affinity to the target peptide-HLA complex (HLA-A*02/PRAME-004) via bio-layer interferometry. Measurements were performed on an Octet RED384 system as described above. Strong binding affinities were determined with K.sub.D values ranging from 3 nM to 10 nM (Table 8). These data show the additional affinity-improving effects of TCR mutations bA84D and aN114V, while mutations bT115L/K, bL11E, bP46M, bQ48R, aN20K do not seem to affect binding affinity. Furthermore, binding affinities were determined for three selected similar peptides serving as potential off-target peptides in the context of HLA-A*02 and K.sub.D windows were calculated compared to binding of the target peptide-HLA. Strongest TCER® binding to similar peptides was observed for GIMAP8-001 with K.sub.D windows ranging from 26- to 168-fold. K.sub.D windows of more than 25-fold already provide good therapeutic windows.

TABLE-US-00009 TABLE 8 K.sub.D values for binding of TCER ® Slot I variants to HLA-A*02/PRAME-004 and K.sub.D windows for three selected similar peptides serving as potential off-target peptides as measured via bio-layer interferometry. TCER ® PRAME-004 K.sub.D(GIMAP8-001)/ K.sub.D(SMARCD1-001)/ K.sub.D(MYO1B-002)/ variant Recruiter K.sub.D (M) K.sub.D(PRAME-004) K.sub.D(PRAME-004) K.sub.D(PRAME-004) TPP-108 UCHT1-V17 1.03E−08 168 no binding no binding TPP-112 UCHT1-V17 4.68E−09 39 380 no binding TPP-106 UCHT1-V17 4.08E−09 42 272 no binding TPP-110 BMA031(V36) 1.33E−08 Not analyzed Not analyzed Not analyzed TPP-111 BMA031(V36) 4.98E−09 Not analyzed Not analyzed Not analyzed TPP-109 BMA031(V36) 4.45E−09 Not analyzed Not analyzed Not analyzed TPP-113 UCHT1-V17 5.24E−09 33 322 no binding TPP-114 UCHT1-V17 5.68E−09 37 225 no binding TPP-115 UCHT1-V17 5.06E−09 38 221 no binding TPP-116 UCHT1-V17 5.18E−09 31 205 no binding TPP-117 UCHT1-V17 3.42E−09 34 41 no binding TPP-118 UCHT1-V17 3.29E−09 49 51 no binding TPP-119 UCHT1-V17 4.57E−09 30 213 no binding TPP-120 UCHT1-V17 5.49E−09 28 324 no binding TPP-121 UCHT1-V17 5.41E−09 26 98 no binding TPP-122 UCHT1-V17 4.43E−09 31 174 no binding TPP-123 UGHT1-V17 3.63E−09 28 33 no binding TPP-124 UCHT1-V17 3.43E−09 30 32 no binding TPP-125 UCHT1-V17 5.98E−09 18 248 no binding TPP-126 UCHT1-V17 5.37E−09 41 221 no binding TPP-127 UGHT1-V17 5.24E−09 34 195 no binding TPP-128 UCHT1-V17 3.75E−09 40 52 no binding TPP-129 UCHT1-V17 3.05E−09 40 47 no binding

Example 3.3: Slot II

[0659] Further TCER® molecules were constructed utilizing VH and VL domains derived from BMA031 (V36) or ID4 as well as Valpha and Vbeta as described above (example 3.1). Production, purification and characterization of the respective TCER® molecules (Table 9, Table 18) was performed as outlined in example 2 whereby all ID4-based molecules were purified using MAbSelect SuRE columns (GE Lifesciences).

TABLE-US-00010 TABLE 9 Summary of productivity and stress stability data obtained for TCER ® molecules of slot II. Final product Monomer (%) yield Monomer after 14 days Protein Recruiter (mg/L) (%) at 40° C. TPP-207 BMA031(V36) 31.8 98.92 95.22 TPP-208 BMA031(V36) n/a 96.96 92.61 TPP-209 BMA031(V36) 32.2 98.87 94.79 TPP-210 BMA031(V36) 19.6 98.15 92.35 TPP-211 BMA031(V36) 44.8 98.60 96.35 TPP-212 BMA031(V36) 34.4 97.66 98.53 TPP-213 BMA031(V36) 53.2 98.12 92.45 TPP-214 BMA031(V36) 45.2 98.26 92.08 TPP-215 BMA031(V36) 33.8 99.21 95.15 TPP-216 BMA031(V36) 4.5 96.53 85.24 TPP-217 BMA031(V36) 26.0 98.16 93.87 TPP-218 BMA031(V36) 19.8 98.24 94.49 TPP-219 ID4 >22.8 71.07 36.49 TPP-220 ID4 21.8 98.36 94.94 TPP-221 ID4 49.2 97.80 96.51 TPP-222 ID4 45.4 98.23 95.79 TPP-227 ID4 48.2 97.60 93.67 TPP-228 ID4 12.1 97.55 94.30 TPP-229 ID4 45.6 97.22 96.99 TPP-230 ID4 47.4 97.29 97.07

[0660] TCER® Slot II variants TPP-207-TPP-222 and TPP-227-TPP-230 were analyzed for their binding affinity to the target peptide-HLA complex (HLA-A*02/PRAME-004) via bio-layer interferometry. Measurements were performed on an Octet RED384 system as described above. Strong binding affinities were determined with K.sub.D values ranging from 1 nM to 7 nM (Table 10). Higher binding affinities were observed for the Identical TCR variants (i.e. Identical V.sub.A and V.sub.B) in combination with the ID4 recruiter when compared to combination with the BMA031 (V36) recruiter (TPP-219-TPP-222 vs. TPP-211-TPP-214). As observed for the TCER® molecules from Slot I (example 3.2), the affinity-improving effects of TCR mutations bA84D and aN114Y could be confirmed for the TCER® variants generated in Slot II, while again no effects on affinity were found for the mutations bT115L/K, bP46M, bQ48R, aN20K.

[0661] TCR binding motifs were assessed for selected TCER® molecules. To determine binding motifs, affinities were measured for the target peptide-HLA complex (HLA-A*02/PRAME-004) as well as for complexes with PRAME-004 variants carrying alanine-substitutions at peptide positions 1, 3, 4, 5, 6, 7 or 8. Affinity measurements were performed on an Octet RED384 or HTX system as described above. PRAME-004 positions were considered to be part of the TCR binding motif if an at least 2-fold reduction in binding affinity or signal (measured for the highest concentration analyzed) was detected for the alanine-substituted peptide variants. All TCER® variants showed broad binding motifs recognizing at least four peptide PRAME-004 positions (Table 10).

TABLE-US-00011 TABLE 10 K.sub.D values for binding of TCER ® Slot II variants to HLA-A*02/PRAME-004 and binding motif determination according to K.sub.D windows for Ala-substituted PRAME-004 peptide variants as measured via bio-layer interferometry. For the A5 peptide, the K.sub.D window was set to 100-fold since no to very low binding precluded affinity determination. PRAME-004 Fold K.sub.D window TCER ® PRAME-004 K.sub.D (M), Binding (Ala/PRAME-004) variant Recruiter K.sub.D (M) for motif motif A1 A3 A4 A5 A6 A7 A8 TPP-207 BMA031(V36) 4.33E−09 TPP-208 BMA031(V36) 3.40E−09 TPP-209 BMA031(V36) 3.29E−09 5.88E−09 -x3-5678x 1.1 16.0 1.2 100.0 4.3 33.4 2.4 TPP-210 BMA031(V36) 3.41E−09 TPP-211 BMA031(V36) 4.53E−09 TPP-212 BMA031(V36) 2.86E−09 TPP-213 BMA031(V36) 4.55E−09 4.92E−09 -x3-5678x 1.1 13.4 1.3 100.0 4.3 32.3 2.6 TPP-214 BMA031(V36) 3.29E−09 2.76E−09 -x3-5-78x 1.3  3.0 1.2 100.0 2.0  5.4 2.2 TPP-215 BMA031(V36) 4.65E−09 TPP-216 BMA031(V36) 3.38E−09 TPP-217 BMA031(V36) 4.22E−09 TPP-218 BMA031(V36) 2.51E−09 TPP-219 ID4 3.40E−09 TPP-220 ID4 1.85E−09 TPP-221 ID4 2.28E−09 2.61E−09 -x3-5678x 1.1 11.0 1.2 100.0 4.1 24.2 3.2 TPP-222 ID4 1.47E−09 1.30E−09 -x3-5678x 1.4  2.9 1.2 100.0 2.2  5.5 2.1 TPP-227 ID4 6.89E−09 TPP-228 ID4 3.46E−09 TPP-229 ID4 6.48E−09 TPP-230 ID4 2.93E−09 2.63E−09 -x3-5678x 1.3 13.0 1.9 100.0 3.9 26.7 3.3

Example 3.4: Slot IIa

[0662] Based on the data generated for the previous TCER® variants (example 3.3), new variants were generated carrying systematic substitutions of selected TCR amino acid positions for which a positive effect on protein properties or binding properties could be detected in previous experiments. Production, purification and characterization of the respective TCER® molecules (Table 11 and Table 18) was performed as outlined in example 3.3. Productivity and stress stability data are summarized in Table 11.

TABLE-US-00012 TABLE 11 Summary of productivity and stress stability data obtained for TCER ® molecules of slot IIa. Final product Monomer (%) TCER ® yield Monomer after 14 days variant Recruiter (mg/L) (%) at 40° C. TPP-235 BMA031(V36) 40.4 98.12 96.16 TPP-236 BMA031(V36) 48.5 98.34 98.08 TPP-237 BMA031(V36) 55.0 97.98 98.21 TPP-238 BMA031(V36) 37.8 98.21 98.15 TPP-239 BMA031(V36) 27.4 98.19 97.22 TPP-240 BMA031(V36) 44.2 98.68 95.72 TPP-241 BMA031(V36) 42.8 98.45 98.02 TPP-242 BMA031(V36) 23.6 98.82 98.54 TPP-243 BMA031(V36) 44.8 98.81 98.10 TPP-244 BMA031(V36) 22.6 98.21 98.27 TPP-245 BMA031(V36) 59.2 98.81 98.32 TPP-246 BMA031(V36) 4.7 92.20 79.35 TPP-247 BMA031(V36) 2.7 93.80 82.82 TPP-248 BMA031(V36) 2.4 92.07 80.49 TPP-249 BMA031(V36) 3.0 92.38 81.45 TPP-250 BMA031(V36) 3.8 93.10 79.11 TPP-252 BMA031(V36) 5.6 93.86 80.14 TPP-253 BMA031(V36) 3.7 94.86 86.09 TPP-254 BMA031(V36) 3.0 94.66 81.85 TPP-255 BMA031(V36) 12.0 92.40 82.01 TPP-256 BMA031(V36) 12.5 97.34 92.67 TPP-257 BMA031(V36) 8.2 95.27 85.31 TPP-258 BMA031(V36) 5.1 96.50 84.32 TPP-259 BMA031(V36) 2.4 97.31 88.55 TPP-260 BMA031(V36) 2.6 96.69 86.45 TPP-261 BMA031(V36) 7.9 97.37 91.72 TPP-262 BMA031(V36) 6.6 96.71 91.53 TPP-263 BMA031(V36) 3.6 93.72 86.61 TPP-264 BMA031(V36) 3.3 93.25 82.35 TPP-265 BMA031(V36) 9.9 91.87 83.48 TPP-266 BMA031(V36) 8.6 95.67 90.72 TPP-267 BMA031(V36) 6.0 94.51 85.97 TPP-266 BMA031(V36) 0.9 93.64 87.21 TPP-269 BMA031(V36) 0 n/a n/a TPP-270 BMA031(V36) 1.7 97.30 91.83 TPP-271 BMA031(V36) 2.2 95.13 87.69 TPP-272 BMA031(V36) 2.9 95.16 87.63 TPP-220 ID4 5.9 97.36 94.81 TPP-273 ID4 5.2 97.77 92.43 TPP-274 ID4 2.6 97.11 95.06 TPP-275 ID4 2.2 96.47 94.08 TPP-276 ID4 1.8 97.02 95.39 TPP-277 ID4 2.7 96.84 94.89 TPP-279 ID4 5.4 98.03 95.9

[0663] TCER® Slot IIa variants TPP-235-250, -252-268, -270-277, -279 were analyzed for their binding affinity to the target peptide-HLA complex (HLA-A*02/PRAME-004) via bio-layer interferometry. Measurements were performed on an Octet RED384 or HTX system as described above. Strong binding affinities were found with K.sub.D values ranging from 2 nM to 15 nM (Table 12). For position bA84, amino acid substitutions showed that bA84D is the most preferred substitution. At position aN114, alternative amino acid substitutions with affinities comparable to aN114Y were found, such as A, H, I and L. Alternatives to bT115K/L with comparable affinities were identified and included R, A. I and V. Introducing the mutation bA110S slightly reduced the affinities of the respective variants.

[0664] Binding motifs were assessed for selected TCER® variants. To determine binding motifs, affinities were measured for the target peptide-HLA complex (HLA-A*02/PRAME-004) as well as for complexes with PRAME-004 variants carrying alanine substitutions at peptide positions 1, 3, 4, 5, 6, 7 or 8 as described above. PRAME-004 positions were considered to be part of the TCR binding motif if an at least 2-fold reduction in binding affinity or signal (measured for the highest concentration analyzed) was detected for the alanine-substituted peptide variants. All tested TCER® variants showed broad binding motifs recognizing at least three peptide positions (Table 12).

[0665] In addition to binding motifs, the binding specificity of selected TCER® Slot II and IIa variants was further analyzed by bio-layer interferometry for binding to a set of 16 similar peptides potentially serving as off-target peptides. Measurements were performed on an Octet HTX system basically as described above. For the analysis, peptide-HLA complexes comprising the PRAME-004 target peptide, individual peptides out of a set of similar peptides or a control peptide were loaded onto HIS1K biosensors and binding of the TCER® variants was analyzed at a high TCER® concentration of 1 μM. The response signal at the end of a 5 min association phase was used to calculate the relative binding signal of the similar peptides in comparison to the PRAME-004 target peptide for selected TCER® variants (Table 13). Under these conditions, even binding events with very low affinity, which can be described as non significant (e.g. binding with a K.sub.D that is increased by a factor of ≥25, ≥30, ≥40, ≥50, ≥75, or 2100, compared to the K.sub.D for binding to the PRAME-004 peptide:MHC complex), will be detected. Among the 16 analyzed similar peptides, 11 peptides did not show any binding to any of the selected TCER® variants. Binding with lower signals compared to PRAME-004 was detected for five of the 16 similar peptides and four of these peptides were analyzed in more detail for TCER® Slot III variants such as measuring K.sub.D windows compared to the PRAME-004 target peptide.

TABLE-US-00013 TABLE 12 K.sub.D values for binding of TCER ® Slot II variants to HLA-A*02/PRAME-004 and binding motif determination according to K.sub.D windows for Ala-substituted PRAME-004 peptide variants as measured via bio-layer interferometry. For the A5 peptide, the K.sub.D window was set to 100-fold since no to low binding precluded affinity determination. PRAME-004 Fold K.sub.D window TCER ® PRAME-004 K.sub.D (M) Binding (Ala/PRAME-004) variant Recruiter K.sub.D (M) for motif motif A1 A3 A4 A5 A6 A7 A8 TPP-246 BMA031(V36) 5.19E−09 TPP-247 BMA031(V36) 8.94E−09 TPP-248 BMA031(V36) 1.46E−08 TPP-249 BMA031(V36) 6.69E−09 TPP-250 BMA031(V36) 6.38E−09 TPP-252 BMA031(V36) 6.30E−09 TPP-220 ID4 1.92E−09 TPP-273 ID4 2.78E−09 TPP-274 ID4 4.61E−09 TPP-275 ID4 7.21E−09 TPP-276 ID4 2.93E−09 TPP-277 ID4 3.71E−09 TPP-279 ID4 2.18E−09 TPP-212 BMA031(V36) 3.38E−09 3.48E−09 -x3-5-7-x 1.1 2.5 1.0 100.0 1.8 4.6 1.9 TPP-235 BMA031(V36) 3.65E−09 TPP-236 BMA031(V36) 6.01E−09 TPP-237 BMA031(V36) 4.46E−09 TPP-238 BMA031(V36) 4.74E−09 TPP-239 BMA031(V36) 2.60E−09 3.44E−09 -x3-5-7-x 1.1 4.1 1.0 100.0 2.0 7.8 1.9 TPP-240 BMA031(V36) 3.48E−09 TPP-241 BMA031(V36) 3.38E−09 3.84E−09 -x3-567-x 1.0 6.7 1.0 100.0 2.1 13.8  2.0 TPP-242 BMA031(V36) 5.23E−09 TPP-243 BMA031(V36) 4.05E−09 TPP-244 BMA031(V36) 4.90E−09 TPP-245 BMA031(V36) 4.41E−09 TPP-253 BMA031(V36) 3.43E−09 TPP-254 BMA031(V36) 3.69E−09 TPP-255 BMA031(V36) 6.13E−09 TPP-256 BMA031(V36) 3.12E−09 4.08E−09 -x3-5-7-x 1.0 2.9 0.9 100.0 1.8 6.4 1.8 TPP-257 BMA031(V36) 3.52E−09 TPP-258 BMA031(V36) 4.79E−09 TPP-259 BMA031(V36) 4.80E−09 TPP-260 BMA031(V36) 4.31E−09 TPP-261 BMA031(V36) 3.45E−09 TPP-262 BMA031(V36) 3.29E−09 4.18E−09 -x3-5-7-x 1.0 3.3 0.8 100.0 1.8 6.5 1.7 TPP-263 BMA031(V36) 3.87E−09 TPP-264 BMA031(V36) 7.39E−09 TPP-265 BMA031(V36) 6.72E−09 TPP-266 BMA031(V36) 3.81E−09 4.57E−09 -x3-5678x 1.1 8.1 1.3 100.0 2.7 10.3  2.3 TPP-267 BMA031(V36) 4.78E−09 TPP-268 BMA031(V36) 6.00E−09 TPP-270 BMA031(V36) 5.74E−09 TPP-271 BMA031(V36) 4.08E−09 TPP-272 BMA031(V36) 4.11E−09 5.70E−09 -x3-5678x 1.2 6.9 1.1 100.0 2.4 9.8 2.6

TABLE-US-00014 TABLE 13 Relative binding signals for similar peptides (in percent of signals detected for PRAME-004 target peptide) of selected TCER ® Slot II and IIa variants as measured via bio-layer interferometry. TPP-214 TPP-239 TPP-241 TPP-256 TPP-266 Recruiter: TPP-230 Recruiter: Recruiter: Recruiter: Recruiter: BMA031 Recruiter: BMA031 BMA031 BMA031 BMA031 Peptide (V36) ID4 (V36) (V36) (V36) (V36) PRAME-004 100 100 100 100 100 100 SMARCD1-001 82 60 65 60 49 19 GIMAP8-001 70 46 55 56 38 −3 FARSA-001 69 35 49 72 17 −5 NOMAP-3-1408 46 11 24 25 7 −12 VIM-009 41 10 28 24 10 11 buffer control 0 1 0 0 1 0 FAM114A2-002 −11 −7 −5 −4 −3 −6 PDCD10-004 −12 −10 −14 −14 −14 −13 NOMAP-5-0765 −14 −12 −18 −16 −17 −18 IGHD-002 −15 −12 −15 −15 −10 −15 TSN-001 −16 −12 −17 −18 −17 −18 NOMAP-3-1587 −16 −14 −16 −17 −18 −18 DDX5-001 (negative control) −17 −13 −16 −17 −17 −16 ALOX15B-003 −18 −15 −15 −19 −14 −17 NOMAP-3-1768 −18 −16 −19 −19 −21 −19 GPR56-002 −18 −14 −19 −19 −17 −19 NOMAP-3-1265 −18 −13 −16 −20 −15 −20 NOMAP-3-0972 −22 −17 −22 −23 −20 −23

Example 3.5: Slot III

[0666] Further TCER® were constructed utilizing VH and VL domains derived from BMA031 (V36) or modified variants (A02 and D01) thereof, or 104 as well as Valpha and Vbeta as described above (example 3.1). An additional TCER® molecule based on the UCHT1-V17 recruiting antibody (TPP-1109) was generated as reference. DNA constructs coding for the respective molecules were generated as outlined above. Resulting plasmids were used for transfection of CHO-S cells by electroporation (MaxCyte) for transient expression and production of TCER® variants (Table 14 and Table 18). Purification, formulation and initial characterization of molecules was performed as outlined above in example 3.3.

TABLE-US-00015 TABLE 14 Summary of productivity and stress stability data obtained for TCER ® molecules of slot III. Final product Monomer (%) TCER ® Va, Vb yield Monomer after 14 days variant (SEQ ID NO) Recruiter (mg/L) (%) at 40° C. TPP-230 132, 135 ID4 73.8 98.83 95.13 TPP-871 137, 135 ID4 80.0 98.92 97.33 TPP-222 132, 134 ID4 70.6 98.80 97.46 TPP-872 137, 134 ID4 62.5 98.77 97.87 TPP-214 132, 134 BMA31(V36) 36.2 97.94 94.98 TPP-876 137, 134 ID4 36.9 97.94 92.28 TPP-666 132, 136 BMA31(V36)A02 49.7 97.59 93.11 TPP-879 137, 134 BMA31(V36)A02 43.5 92.98 90.42 TPP-891 137, 134 BMA31(V36)D01 40.0 98.18 94.94 TPP-669 132, 136 BMA31(V36)D01 72.9 97.83 94.66 TPP-894 132, 135 BMA31(V36)D01 40.2 97.45 93.11 TPP-1109 (CDR6) UCHT1-V17 13.6 98.10 92.62

[0667] Potency of TCER® molecules with respect to killing of HLA-A*02-positive tumor cell lines presenting different levels of PRAME-004 target peptide on their cell surface, was assessed in LDH-release assays. In addition, an HLA-A*02-positive but PRAME-004-negative tumor cell line (e.g. T98G) was assessed to characterize unspecific or off-target activity of the TCER® variants. Tumor cell lines were co-incubated with PBMC effectors derived from healthy HLA-A*02-positive donors at a ratio of 1:10 and in the presence of increasing TCER® concentrations. TCER®-induced cytotoxicity was quantified after 48 hours of co-culture by measurement of released LDH. EC.sub.50 values of dose-response curves were calculated utilizing non-linear 4-point curve fitting. EC.sub.50 values for two PRAME-004-positive tumor cell lines (Hs695T and U2OS) and a PRAME-004-negative tumor cell line (T98G) were determined in different experiments with different PBMC donors and are graphically summarized in FIG. 9.

[0668] TCER® Slot III variants TPP-214, -222, -230, -666, -669, -871, -872, -876, -879, -891, 894 were analyzed for their binding affinity to the target peptide-HLA complex (HLA-A*02/PRAME-004) via bio-layer interferometry. Measurements were performed on an Octet HTX system at 30° C. Assays were run at a sensor offset of 3 mm and an acquisition rate of 5 Hz on HIS1K biosensors in 16-channel mode using PBS, 0.05% Tween-20, 0.1% BSA as assay buffer. The following assay step sequence was repeated to measure all binding affinities: regeneration (5 s. 10 mM glycine pH 1.5)/neutralization (5 s, assay buffer; one regeneration cycle consists of four repeats of regeneration/neutralization), baseline (60 s, assay buffer), loading (120 s, 10 μg/ml peptide-HLA), baseline (120 s, assay buffer), association (300 s, twofold serial dilution of TCER's ranging from 100 nM to 1.56 nM or 50 nM to 0.78 nM, assay buffer as reference), dissociation (300 s, assay buffer). Data evaluation was done using Octet Data Analysis HT Software. Reference sensor subtraction was performed to subtract potential dissociation of peptide-HLA loaded onto the biosensor (via a biosensor loaded with peptide-HLA measured in buffer). Data traces were aligned to baseline (average of the last 5 s), inter-step correction was done to the dissociation step, Savitzky-Golay filtering was applied and curves were fitted globally using a 1:1 binding model (with Rmax unlinked by sensor). Strong binding affinities were found with K.sub.D values ranging from 2 nM to 5 nM (Table 15). Furthermore, binding affinities were determined for four previously identified potential off-target peptides and K.sub.D windows were calculated compared to binding of the target peptide-HLA. Measurements were performed on an Octet RED384 or HTX system at 30° C. Assays were run at a sensor offset of 3 mm and an acquisition rate of 5 Hz on HIS1K biosensors in 16-channel mode using PBS, 0.05% Tween-20, 0.1% BSA as assay buffer. The following assay step sequence was repeated to measure all binding affinities: regeneration (5 s, 10 mM glycine pH 1.5)/neutralization (5 s, assay buffer; one regeneration cycle consists of four repeats of regeneration/neutralization), baseline (60 s, assay buffer), loading (120 s, 10 μg/ml peptide-HLA), baseline (120 s, assay buffer), association (300 s, twofold serial dilution of TCER® ranging from 500 nM to 7.81 nM, assay buffer as reference), dissociation (300 s, assay buffer). Data evaluation was done using Octet Data Analysis HT Software. Reference sensor subtraction was performed to subtract potential dissociation of peptide-HLA loaded onto the biosensor (via a biosensor loaded with the respective peptide-HLA measured in buffer). Data traces were aligned to baseline (average of the last 5 s), inter-step correction was done to the dissociation step, Savitzky-Golay filtering was applied and curves were fitted globally using a 1:1 binding model (with Rmax unlinked by sensor). Overall, considerable weaker binding to the potential off-target peptides compared to target peptide was found for all variants showing windows of at least 60-fold to even no binding at all. NOMAP-3-1408 was not selected for K.sub.D determination, despite showing relative binding signals comparable to VIM-009 (Table 13). For VIM-009, the smallest measured K.sub.D windows were >100-fold (Table 15). Thus, binding to VIM-009 is not relevant and affinity determination of NOMAP-3-1408 binding was not considered necessary based on its binding signals comparable to VIM-009. For one interaction, a K.sub.D window of 50-fold was calculated. However, for this interaction and also several others, the Rmax value calculated by the fitting algorithm was too low, so that the interaction is assumed to be weaker than calculated and thus the window larger. Respective interactions are indicated in Table 15. To further analyze specificity of the different variants, binding motifs were determined by measuring the affinities for the target peptide-HLA complex as well as for the alanine-substituted variants for positions 1, 3, 4, 5, 6, 7, 8. Measurements were performed on an Octet HTX system at 30° C. Assays were run at a sensor offset of 3 mm and an acquisition rate of 5 Hz on HIS1K biosensors in 16- or 8-channel mode using PBS, 0.05% Tween-20, 0.1% BSA as assay buffer. The following assay step sequence was repeated to measure all binding affinities: regeneration (5 s, 10 mM glycine pH 1.5)/neutralization (5 s, assay buffer; one regeneration cycle consists of four repeats of regeneration/neutralization), baseline (60 s, assay buffer), loading (120 s, 10 μg/ml peptide-HLA), baseline (120 s, assay buffer), association (150 s, twofold serial dilution of TCER® ranging from 400 nM to 6.25 nM, assay buffer as reference), dissociation (300 s, assay buffer). Data evaluation was done using Octet Data Analysis HT Software. Reference sensor subtraction was performed to subtract potential dissociation of peptide-HLA loaded onto the biosensor (via a biosensor loaded with the respective peptide-HLA measured in buffer). Data traces were aligned to baseline (average of the last 5 s), inter-step correction was done to the dissociation step, Savitzky-Golay filtering was applied and curves were fitted globally using a 1:1 binding model (with Rmax unlinked by sensor). A position was considered part of the binding motif for an at least 2-fold reduction in affinity or binding signal (measured for the highest concentration analyzed). All tested TCER® variants showed broad binding motifs recognizing at least four and up to all analyzed peptide positions (Table 16). Positive effects on the binding motif were observed for bA84, aN114L and bA110S/bT115A, which is in accordance with previous data. For comparison, the binding motif of an alternative PRAME-004-targeting TCER® reference molecule (TPP-1109) was analyzed. This TCER® recognized positions 5-8 of the peptide and thus binding is limited to this peptide stretch, while positions recognized by TCER® Slot III variants are more evenly distributed throughout the whole peptide.

[0669] TCER® Slot III variants TPP-214, -222, -230, -666, -669, -871, -872, -876, -879, -891, 894 were additionally characterized for their ability to kill T2 cells loaded with varying levels of target peptide. After loading of the T2 cells with the respective concentrations of PRAME-004 for 2 h, peptide-loaded T2 cells were co-cultured with human PBMCs at an E:T ratio of 5:1 in the presence of increasing concentrations of TCER® variants for 48 h. Levels of LDH released into the supernatant were quantified using CytoTox 96 Non-Radioactive Cytotoxicity Assay Kit (Promega). All TCER® variants showed potent killing of PRAME-004-loaded T2 cells with subpicomolar EC.sub.50 values at a peptide loading concentration of 10 nM (FIG. 10, Table 17). EC.sub.50 values increased for decreasing PRAME-004 loading levels. However, even at a very low PRAME-004 loading concentration of 10 pM, killing was induced by all TCER® variants, except for TPP-214.

TABLE-US-00016 TABLE 15 K.sub.D values for binding to HLA-A*02/PRAME-004 and K.sub.D windows of four selected off-target peptides measured via bio-layer interferometry for TCER ® Slot III variants. Va, Vb TCER ® (SEQ PRAME-004 K.sub.D FARSA-001/ K.sub.D GIMAP8-001/ K.sub.D SMARCD1-001/ K.sub.D VIM-009/ variant Recruiter ID NO) K.sub.D (M) K.sub.D PRAME-004 K.sub.D PRAME-004 K.sub.D PRAME-004 K.sub.D PRAME-004 TPP-230 ID4 132, 135 3.05E−09 — .sup. 120.sup.1 .sup. 130.sup.1 — TPP-871 ID4 137, 135 2.89E−09 — — — — TPP-222 ID4 132, 134 1.56E−09 118   69  74 .sup. 112.sup.1 TPP-872 ID4 137, 134 1.60E−09 95 103 .sup. 119.sup.1 2153  TPP-214 BMA031(V36) 132, 134 2.43E−09 216   59  66 389 TPP-876 BMA031(V36)A02 137, 134 2.43E−09 86  80 267 .sup. 160.sup.1 TPP-666 BMA031(V36)A02 132, 136 3.37E−09 507  142 121 171 TPP-879 BMA031(V36)A02 132, 135 4.55E−09 — — — — TPP-891 BMA031(V36)D01 137, 134 2.34E−09 76  85 254 .sup. 146.sup.1 TPP-669 BMA031(V36)D01 132, 136 3.65E−09 .sup. 83.sup.1 .sup.  50.sup.1  84 165 TPP-894 BMA031(V36)D01 132, 135 5.18E−09 — — — — .sup.1K.sub.D windows are expected to be higher than the values given in the table (calculated Rmax values for these interactions are too low due to overall low binding signals).

TABLE-US-00017 TABLE 16 K.sub.D values for binding to HLA-A*02/PRAME-004 and K.sub.D windows of Ala-substituted peptide variants for binding motif determination measured via bio-layer interferometry for TCER ® Slot III variants. For position 5, a threshold of 100 is given for the K.sub.D window. Recognition of this position is at least 100-fold. Va, Vb TCER ® (SEQ PRAME-004 Binding K.sub.D Ala/target variant Recruiter ID NO) K.sub.D, motif (M) motif A1 A3 A4 A5 A6 A7 A8 TPP-230 ID4 132, 135 3.03E−09 -x3-5678x 1.2 12.2 1.7 100.0 3.9 25.5 3.0 TPP-871 ID4 137, 135 2.47E−09 1x345678x 2.5 39.3 4.7 100.0 16.5 89.9 8.3 TPP-222 ID4 132, 134 1.50E−09 -x3-5-78x 1.1 2.3 0.9 100.0 1.8 4.3 1.8 TPP-872 ID4 137, 134 1.48E−09 -x3-5678x 1.1 7.6 1.1 100.0 3.0 17.5 2.7 TPP-214 BMA031(V36) 132, 134 3.17E−09 -x3-5-78x 0.9 2.1 0.8 100.0 1.6 4.6 1.7 TPP-876 BMA031(V36)A02 137, 134 2.87E−09 -x3-567-x 1.0 6.8 1.0 100.0 2.3 13.9 2.0 TPP-666 BMA031(V36)A02 132, 136 3.84E−09 -x3-5678x 1.1 7.9 1.2 100.0 2.6 9.7 2.1 TPP-879 BMA031V36)A02 132, 135 6.15E−09 -x3-5678x 1.1 12.5 1.6 100.0 3.5 27.5 2.6 TPP-891 BMA031(V36)D01 137, 134 2.80E−09 -x:3-5678x 1.0 7.2 1.1 100.0 2.6 14.7 2.3 TPP-669 BMA031(V36)D01 132, 136 3.28E−09 -x3-5678x 1.1 9.1 1.2 100.0 2.5 11.0 2.4 TPP-894 BMA031(V36)D01 132, 135 6.04E−09 -x3-5678x 1.2 14.9 1.9 100.0 3.8 26.4 2.8 TPP-1109 UCHT1-V17 (CDR6) 2.47E−09 -x-5678x 0.9 0.8 1.2 49.0 7.9 55.7 4.1

TABLE-US-00018 TABLE 17 In vitro cytotoxicity of TCER ® Slot III variants on PRAME-004-loaded T2 cells. T2 cells were co- cultured with human PBMCs at an E:T ratio of 5:1 for 48 h. PRAME-004 loading concentrations are indicated. Ec.sub.50 values and cytotoxicity levels in the plateau (Top) were calculated using non-linear 4-point curve fitting. Va, Vb 10 nM 1 nM 100 pM 10 pM TCER ® (SEQ PRAME-004 PRAME-004 PRAME-004 PRAME-004 variant Recruiter ID NO) EC.sub.50 [pM] Top EC.sub.50 [pM] Top EC.sub.50 [pM] Top EC.sub.50 [pM] Top TPP-230 ID4 132, 135 0.09 109 0.9 139 23.2.sup.1 179 145 80 TPP-871 ID4 137, 135 0.13 109 1.6 143 76.5.sup.1 90 361 76 TPP-222 ID4 132, 134 complete killing 109 complete killing 78  2.8.sup.1 127 58 90 TPP-872 ID4 137, 134 complete killing 109 complete killing 151  4.3.sup.1 84 49 74 TPP-876 BMA031(V36)A02 137, 134 0.16 111 2.0 113 24.4 100 539 40 TPP-666 BMA031(V36)A02 132, 136 0.15 113 2.4 113 39.8 100 182 35 TPP-879 BMA031(V36)A02 132, 135 0.54 106 6.2 109 94.4 117 1070 39 TPP-214 BMA031(V36) 132, 134 0.22 108 5.0 109 92.8 102 no killing 20 TPP-891 BMA031(V36)D01 137, 134 0.19 120 2.2 112 54.0 125 611 45 TPP-669 BMA031(V36)D01 132, 136 0.22 124 3.2 108 84.0 126 246 31 TPP-894 BMA031(V36)D01 132, 135 0.87 108 9.9 115 226.0 129 1084 44 TPP-214 BMA031(V36) 132, 134 0.26 121 5.4 111 105.4 99 no killing 23 .sup.1High variability within replicates do not allow for reliable EC.sub.50 calculation.

TABLE-US-00019 TABLE 18 Bispecific molecules α- β- chain chain SEQ ID SEQ ID ID NO NO TPP-70 93 94 TPP-71 93 95 TPP-72 93 96 TPP-73 93 97 TPP-74 93 98 TPP-93 100 101 TPP-79 103 102 TPP-105 105 104 TPP-106 106 107 TPP-108 106 101 TPP-109 111 110 TPP-110 111 102 TPP-111 103 110 TPP-112 100 107 TPP-113 100 119 TPP-114 100 121 TPP-115 122 121 TPP-116 106 121 TPP-117 126 121 TPP-118 128 121 TPP-119 100 131 TPP-120 100 133 TPP-121 122 133 TPP-122 106 133 TPP-123 126 133 TPP-124 128 133 TPP-125 100 143 TPP-126 122 143 TPP-127 106 143 TPP-128 126 143 TPP-129 128 143 TPP-207 103 152 TPP-208 155 152 TPP-209 157 152 TPP-210 159 152 TPP-211 103 160 TPP-212 155 162 TPP-213 157 162 TPP-214 167 160 TPP-215 169 168 TPP-216 171 168 TPP-217 173 168 TPP-218 167 168 TPP-219 177 176 TPP-220 179 176 TPP-221 181 176 TPP-222 183 176 TPP-226 159 184 TPP-227 105 186 TPP-228 189 186 TPP-229 191 186 TPP-230 193 186 TPP-235 195 160 TPP-236 197 160 TPP-237 199 160 TPP-238 201 160 TPP-239 203 160 TPP-240 205 160 TPP-241 207 160 TPP-242 209 160 TPP-243 211 160 TPP-244 213 160 TPP-245 215 160 TPP-246 217 216 TPP-247 217 218 TPP-248 217 220 TPP-249 217 222 TPP-250 217 224 TPP-252 217 228 TPP-253 217 230 TPP-254 217 232 TPP-255 217 234 TPP-256 217 236 TPP-257 217 238 TPP-258 217 240 TPP-259 217 242 TPP-260 217 244 TPP-261 217 246 TPP-262 217 248 TPP-263 217 250 TPP-264 217 252 TPP-265 217 254 TPP-266 217 256 TPP-267 217 258 TPP-268 217 260 TPP-269 217 262 TPP-270 217 264 TPP-271 217 266 TPP-272 155 268 TPP-273 189 270 TPP-274 189 272 TPP-275 189 274 TPP-276 189 276 TPP-277 189 278 TPP-279 189 282 TPP-666 285 284 TPP-669 291 284 TPP-871 295 186 TPP-872 295 296 TPP-876 299 162 TPP-879 285 300 TPP-891 303 162 TPP-892 303 284 TPP-894 291 300 TPP-1292 151 284 TPP-1293 156 162 TPP-1294 158 284 TPP-1295 158 300 TPP-1296 303 161 TPP-1297 303 163 TPP-1298 291 164 TPP-1300 291 165 TPP-1301 166 300 TPP-1302 291 170 TPP-1303 291 172 TPP-1304 291 174 TPP-1305 166 170 TPP-1306 166 172 TPP-1307 166 174 TPP-1308 291 182 TPP-1309 291 185 TPP-1332 175 186 TPP-1333 178 186 TPP-1334 180 186
In table 18, except for TPP-70, TPP-71, TPP-72, TPP-73 and TPP74, the term “α-chain” refers to a polypeptide chain comprising a V.sub.α, i.e. a variable domain derived from a TCR α-chain. The term “β-chain” refers to a polypeptide chain comprising a V.sub.β, i.e. a variable domain derived from a TCR β-chain. For TPP-70, TPP-71, TPP-72, TPP-73 and TPP74, the “α-chain” does not comprise any TCR derived variable domains, but the “β-chain” comprises two TCR-derived variable domains, one derived from a TCR α-chain and one derived from a TCR β-chain.

Example 3.6: Safety Assessment for Selected TCER® Slot III Candidates

[0670] The safety profile of the TCER® molecules TPP-230, TPP-666, TPP-871 and TPP-891 (Tables 14-18) was assessed in killing experiments with astrocytes and cardiomyocytes (derived from induced pluripotent stem cells) as well as aortic endothelial cells, mesenchymal stem cells and tracheal smooth muscle cells. FIG. 11 shows the results of co-cultures of above normal cell types (all expressing HLA-A*02) with PBMC effector cells from a healthy HLA-A*02+ donor at a ratio of 1:10 (target cells:effector cells) in presence of increasing TCER® concentrations. The cells were co-cultured in a 1:1 mixture of the respective normal tissue cell medium and T cell medium or in T cell medium alone (LDH-AM). After 48 h of co-culture, supernatants were harvested and TCER® induced normal tissue cell lysis was assessed by measuring LDH release with the LDH-Glow Kit (Promega). To determine a safety window, the TCER® molecules were co-incubated in an identical setup with the PRAME-004-positive tumor cell line Hs695T in the respective 1:1 mixture of normal tissue cell medium and T cell medium followed by the assessment of LDH release.

[0671] As shown in FIG. 11, no cytotoxicity against normal tissue cells was observed with TPP-230 and TPP-871 even at the highest TCER® concentration of 100 nM. For TPP-666 and TPP-891 some normal tissue cell lysis was observed at 100 nM TCER® concentration but no lysis was detected at 10 nM. When compared to Hs695T tumor cells that showed pronounced lysis at 100 pM for all tested TCER® molecules and for some molecules even lysis at 10 pM concentration, the normal tissue cell lysis at 100 nM concentration indicates a safety window of 1,000-fold (TPP-666 and TPP-891) or more (TPP-230 and TPP-871).

Example 3.7: Slot IV

[0672] Further TCER® were constructed utilizing V.sub.H and V.sub.L domains derived from BMA031 (V36) or modified variants (A02 and D01) thereof, or 104 as well as Valpha and Vbeta as described above (example 3.1). DNA constructs coding for the respective molecules were generated as outlined above. Resulting plasmids were used for transfection of CHO-S cells by electroporation (MaxCyte) for transient expression and production of TCER® variants (Table 20 and Table 18). Purification, formulation and initial characterization of molecules was performed as outlined above in example 3.3.

TABLE-US-00020 TABLE 20 Summary of productivity and stress stability data obtained for TCER ® molecules of slot IV. α-chain, β- Final product Monomer (%) TCER ® chain yield Monomer after 14 days variant (SEQ ID NO) Recruiter (mg/L) (%) at 40° C. TPP-1292 151, 284 BMA031(V36)A02_H90Y 42.9 97.53 93.46 TPP-1294 158, 284 BMA031(V36)D01_H90Y 39.8 97.78 90.61 TPP-1295 158, 300 BMA031(V36)D01_H90Y 56.5 94.89 91.49 TPP-1296 303, 161 BMA031(V36)D01 50.7 79.21 75.17 TPP-1297 303, 163 BMA031(V36)D01 41.3 94.12 86.77 TPP-1298 291, 164 BMA031(V36)D01 68.1 94.41 89.7 TPP-1300 291, 165 BMA031(V36)D01 43.9 96.81 87.5 TPP-1301 166, 300 BMA031(V36)D01 73.7 94.57 90.89 TPP-1302 291, 170 BMA031(V36)D01 67.3 83.48 79.58 TPP-1303 291, 172 BMA031(V36)D01 48.5 74.95 71.03 TPP-1304 291, 174 BMA031(V36)D01 55.0 95.13 88.87 TPP-1305 166, 170 BMA031(V36)D01 51.6 81.55 77.75 TPP-1306 166, 172 BMA031(V36)D01 71.7 86.37 81.18 TPP-1307 166, 174 BMA031(V36)D01 60.7 95.93 88.16 TPP-1308 291, 182 BMA031(V36)D01 61.9 92.28 87.98 TPP-1309 291, 185 BMA031(V36)D01 74.8 98.98 91.11 TPP-1332 175, 186 ID4 variant 0 n/a n/a TPP-1333 178, 186 ID4 variant 61.1 98.52 95.51 TPP-1334 180, 186 ID4 variant 61.4 98.42 95.94
In table 20, the term “α-chain” refers to a polypeptide chain comprising a V.sub.α, i.e. a variable domain derived from a TCR α-chain. The term “β-chain” refers to a polypeptide chain comprising a V.sub.β, i.e. a variable domain derived from a TCR β-chain.

[0673] Potency of TCER® molecules with respect to killing of HLA-A*02-positive tumor cell lines presenting different levels of PRAME-004 target peptide on their cell surface, was assessed in LDH-release assays. In addition, an HLA-A*02-positive but PRAME-004-negative tumor cell line (e.g. T98G) was assessed to characterize unspecific or off-target activity of the TCER® variants. Tumor cell lines were co-incubated with PBMC effectors derived from healthy HLA-A*02-positive donors at a ratio of 1:10 and in the presence of increasing TCER® concentrations. TCER®-induced cytotoxicity was quantified after 48 hours of co-culture by measurement of released LDH. EC.sub.50 values of dose-response curves were calculated utilizing non-linear 4-point curve fitting. EC.sub.50 values for a PRAME-004-positive tumor cell lines U2OS and a PRAME-004-negative tumor cell line (T98G) were determined In different experiments with different PBMC donors and are summarized in table 21.

TABLE-US-00021 TABLE 24 Summary of LDH-release assay data obtained for TCER ® molecules of slot IV. TCER ® EC50 [pM] for EC50 [pM] for EC50 [pM] for EC50 [pM] for variant HBC-1005 vs U2OS HBC-1005 vs T98G HBC-848 vs U2OS HBC-848 vs T98G TPP-1292 66 22,659 547 77,267 TPP-1294 99 69,150 431 >100,000 TPP-1295 150 >100,000 663 >100,000 TPP-1297 2,526 >100,000 4,096 >100,000 TPP-1298 48 37,953 249 >100,000 TPP-1300 186 >100,000 811 >100,000 TPP-1301 240 >100,000 979 >100,000 TPP-1304 7125 >100,000 13,686 >100,000 TPP-1307 8,056 >100,000 >100,000 >100,000 TPP-1333 226 >100,000 719 >100,000 TPP-1334 217 >100,000 829 >100,000

[0674] TCER® Slot IV variants TPP-1292, -1294 to -1298, -1300 to -1309, -1333, -1334 were analyzed for their binding affinity to the target peptide-HLA complex (HLA-A*02/PRAME-004) via bio-layer interferometry. Measurements were performed on an Octet HTX system at 30° C. Assays were run at a sensor offset of 3 mm and an acquisition rate of 5 Hz on HIS1K biosensors in 16-channel mode using PBS, 0.05% Tween-20, 0.1% BSA as assay buffer. The following assay step sequence was repeated to measure all binding affinities: regeneration (5 s, 10 mM glycine pH 1.5)/neutralization (5 s, assay buffer; one regeneration cycle consists of four repeats of regeneration/neutralization), baseline (60 s, assay buffer), loading (120 s, 10 μg/mi peptide-HLA), baseline (120 s, assay buffer), association (300 s, twofold serial dilution of TCER's ranging from 100 nM to 1.56 nM or 50 nM to 0.78 nM, assay buffer as reference), dissociation (300 s, assay buffer). Data evaluation was done using Octet Data Analysis HT Software. Reference sensor subtraction was performed to subtract potential dissociation of peptide-HLA loaded onto the biosensor (via a biosensor loaded with peptide-HLA measured in buffer). Data traces were aligned to baseline (average of the last 5 s), inter-step correction was done to the dissociation step, Savitzky-Golay filtering was applied and curves were fitted globally using a 1:1 binding model (with Rmax unlinked by sensor). Strong binding affinities were found with K.sub.D values ranging from 2 nM to 15 nM (Table 22). Furthermore, binding affinities were determined for two previously identified potential off-target peptides and K.sub.D windows were calculated compared to binding of the target peptide-HLA. Measurements were performed on an Octet RED384 or HTX system at 30° C. Assays were run at a sensor offset of 3 mm and an acquisition rate of 5 Hz on HIS 1K biosensors in 16-channel mode using PBS, 0.05% Tween-20, 0.1% BSA as assay buffer. The following assay step sequence was repeated to measure all binding affinities: regeneration (5 s, 10 mM glycine pH 1.5)/neutralization (5 s, assay buffer; one regeneration cycle consists of four repeats of regeneration/neutralization), baseline (60 s, assay buffer), loading (120 s, 10 μg/mi peptide-HLA), baseline (120 s, assay buffer), association (300 s, twofold serial dilution of TCER® ranging from 500 nM to 7.81 nM, assay buffer as reference), dissociation (300 s, assay buffer). Data evaluation was done using Octet Data Analysis HT Software. Reference sensor subtraction was performed to subtract potential dissociation of peptide-HLA loaded onto the biosensor (via a biosensor loaded with the respective peptide-HLA measured in buffer). Data traces were aligned to baseline (average of the last 5 s), inter-step correction was done to the dissociation step, Savitzky-Golay filtering was applied and curves were fitted globally using a 1:1 binding model (with Rmax unlinked by sensor). Overall, considerable weaker binding to the potential off-target peptides compared to target peptide was found for all variants showing windows of at least 10-fold to even no binding at all. Respective interactions are indicated in Table 22. To further analyze specificity of the variants TPP-1294, -1295, -1298, -1333, -1334, binding motifs were determined by measuring the affinities for the target peptide-HLA complex as well as for the alanine-substituted variants for positions 1, 3, 4, 5, 6, 7, 8. Measurements were performed on an Octet HTX system at 30° C. Assays were run at a sensor offset of 3 mm and an acquisition rate of 5 Hz on HIS1K biosensors in 16- or 8-channel mode using PBS, 0.05% Tween-20, 0.1% BSA as assay buffer. The following assay step sequence was repeated to measure all binding affinities: regeneration (5 s, 10 mM glycine pH 1.5)/neutralization (5 s, assay buffer; one regeneration cycle consists of four repeats of regeneration/neutralization), baseline (60 s, assay buffer), loading (120 s, 10 μg/ml peptide-HLA), baseline (120 s, assay buffer), association (150 s, twofold serial dilution of TCER® ranging from 400 nM to 6.25 nM, assay buffer as reference), dissociation (300 s, assay buffer). Data evaluation was done using Octet Data Analysis HT Software. Reference sensor subtraction was performed to subtract potential dissociation of peptide-HLA loaded onto the biosensor (via a biosensor loaded with the respective peptide-HLA measured in buffer). Data traces were aligned to baseline (average of the last 5 s), inter-step correction was done to the dissociation step, Savitzky-Golay filtering was applied and curves were fitted globally using a 1:1 binding model (with Rmax unlinked by sensor). A position was considered part of the binding motif for an at least 2-fold reduction in affinity or binding signal (measured for the highest concentration analyzed). All tested TCER® variants showed broad binding motifs recognizing at least five and up to all analyzed peptide positions (Table 23).

TABLE-US-00022 TABLE 22 K.sub.D values for binding to HLA-A*02/PRAME-004 and K.sub.D windows of two selected off-target peptides measured via bio- layer interferometry for TCER ® Slot IV variants. TCER ® K.sub.D IFIT-001/ K.sub.D MCMB-002/ variant PRAME-004 K.sub.D (M) K.sub.D PRAME-004 K.sub.D PRAME-004 TPP-1292 2.55E−09 29.5 18.6 TPP-1294 3.06E−09 30.7 20.4 TPP-1295 3.39E−09 45.2 28.6 TPP-1298 2.47E−09 24.1 17.2 TPP-1300 3.90E−09 20.6 20.7 TPP-1301 5.77E−09 33.6 16.8 TPP-1302 3.92E−09 26.4 16.1 TPP-1303 4.58E−09 23.0 17.6 TPP-1304 2.74E−08 >100 >100 TPP-1305 5.19E−09 23.8 13.7 TPP-1306 5.20E−09 47.2 23.3 TPP-1307 3.97E−08 >100 >100 TPP-1308 1.54E−08 83.4 76.7 TPP-1309 1.33E−08 38.8 9.9 TPP-1333 2.94E−09 27.3 16.0 TPP-1334 2.48E−09 26.7 18.0

TABLE-US-00023 TABLE 23 K.sub.D values for binding to HLA-A*02/PRAME-004 and K.sub.D windows of Ala-substituted peptide variants for binding motif determination measured via bio-layer interferometry for TCER ® Slot IV variants. For position 5, a threshold of 100 is given for the K.sub.D window. Recognition of this position is at least 100-fold. TCER ® PRAME-004 Binding K.sub.D Ala/target variant K.sub.D, motif (M) motif A1 A3 A4 A5 A6 A7 A8 TPP-1294 4.35E−09 -x3-5678x 1.6 10.6 2.0 92.4 3.6 13.8 3.3 TPP-1295 3.87E−09 1x345678x 2.2 21.8 2.8 20.7 5.2 35.3 5.0 TPP-1298 2.87E−09 -x3-5678x 1.4 10.3 1.6 100.0 2.9 9.6 2.8 TPP-1333 2.60E−09 -x3-5678x 1.4 12.8 2.0 100.0 3.9 21.0 3.7 TPP-1334 3.09E−09 -x3-5678x 1.1 9.2 1.6 100.0 3.1 15.9 2.6

Example 3.8: Safety Assessment for Selected TCER® Slot IV Candidates

[0675] The safety profile of the TCER® molecules TPP-1294, TPP-1295, TPP-1298, TPP-1333 and TPP-1334 (Tables 18 and 20-23) was assessed in killing experiments with astrocytes, GABAergic neurons and cardiomyocytes (derived from induced pluripotent stem cells; iHA, iHN and iHCM, respectively) as well as pulmonary fibroblasts (HPF), cardiac microvascular endothelial cells (HCMEC), dermal microvascular endothelial cells (HDMEC), aortic endothelial cells (HAoEC), coronary artery smooth muscle cells (HCASMC), renal cortical epithelial cells (HRCEpC) and tracheal smooth muscle cells (HTSMC). Furthermore, a bridging molecule TPP-891 was tested together with other molecules TPP-214 and TPP-669 from earlier slots. FIGS. 12 and 13 show the results of co-cultures of above normal cell types (all expressing HLA-A*02) with PBMC effector cells from a healthy HLA-A*02+ donor at a ratio of 1:10 (target cells:effector cells) in presence of increasing TCER® concentrations. The cells were co-cultured in a 1:1 mixture of the respective normal tissue cell medium and T cell medium or in T cell medium alone (LDH-AM). After 48 h of co-culture, supernatants were harvested and TCER®-induced normal tissue cell lysis was assessed by measuring LDH release with the LDH-Glo™ Kit (Promega). To determine a safety window, the TCER® molecules were co-incubated in an identical setup with the PRAME-004-positive tumor cell line Hs695T in the respective 1:1 mixture of normal tissue cell medium and T cell medium followed by the assessment of LDH release.

[0676] As shown in FIGS. 12 and 13, no cytotoxicity against normal tissue cells was observed for any of the tested molecules until a concentration of 10 nM TCER®. At a concentration of 100 nM only the bridging and reference molecules TPP-891, TPP-669 and TPP-214 show a somehow increased cytotoxicity level above background. The only exception is TPP-1294 in iPSC-derived astrocytes with elevated cytotoxicity exclusively at 100 nM. When compared to Hs695T tumor cells that showed pronounced lysis at 100 pM for all tested TCER® molecules and for some molecules even lysis at 10 pM concentration, the normal tissue cell lysis at 100 nM concentration indicates a safety window of 1,000-fold (TPP-1294) or more (TPP-1295, TPP-1298, TPP-1334 and TPP-1335).

Example 4: Detection of PRAME Peptide on Primary Tissues by Mass Spectrometry

[0677] For the identification and relative quantitation of HLA ligands by mass spectrometry, HLA molecules from shock-frozen tissue samples were purified and HLA-associated peptides were isolated. The isolated peptides were separated, and sequences were identified by online nano-electrospray-ionization (nanoESI) liquid chromatography-mass spectrometry (LC-MS) experiments. Since the peptides were directly identified as ligands of HLA molecules of primary tumors, these results provide direct evidence for the natural processing and presentation of the identified peptides on the primary cancer tissue. The acquired LC-MS data are subsequently processed and quantified using a proprietary label-free quantitation data analysis pipeline, combining algorithms for sequence identification, spectral clustering, ion counting, retention time alignment, charge state deconvolution and normalization. Resulting target detection frequencies are depicted herein below in Table 19.

TABLE-US-00024 TABLE 19 Peptide detection frequency in tumor samples. The target detection frequency is indicated as + (>0%), ++ (>10%), +++ (>30%), or ++++ (>50%). Target detection Entity frequency acute myeloid leukemia (AML) + breast cancer (BRCA) ++ cholanglocellular carcinoma (CCC) + chronic lymphocytic leukemia (CLL) + colorectal carcinoma (CRC) + gallbladder cancer (GBC) ++ glioblastoma (GBM) + hepatocellular carcinoma (HCC) + head and neck squamous cell carcinoma (HNSCC) + melanoma (MEL) ++++ non-Hodgkin lymphoma (NHL) + non-small cell lung cancer adenocarcinoma + (NSCLCadeno) NSCLC samples that cannot unambiguously be assigned ++ to NSCLC adeno or NSCLCsquam (NSCLCother) squamous cell non-small cell lung cancer ++ (NSCLCsquam) ovarian cancer (OC) +++ esophageal cancer (OSCAR) + renal cell carcinoma (RCC) ++ small cell lung cancer (SCLC) ++ urinary bladder carcinoma (UBC) + uterine and endometrial cancer (UEC) ++++

ITEMS

[0678] 1. An antigen binding protein specifically binding to a PRAME antigenic peptide that comprises or consists of the amino acid sequence SLLQHLIGL of SEQ ID NO: 50 and is in a complex with a major histocompatibility complex (MHC) protein, the antigen binding protein comprising [0679] (a) a first polypeptide comprising a variable domain V.sub.A comprising complementarity determining regions (CDRs) CDRa1. CDRa2 and CDRa3, wherein the CORa1 comprises or consists of the amino acid sequence VKEFQD (SEQ ID NO: 16), or an amino acid sequence differing from SEQ ID NO: 16 by one, two or three amino acid mutations, preferably amino acid substitutions, and the CDRa3 comprises or consists of the amino acid sequence of ALYNNLDMR (SEQ ID NO: 33) or ALYNNYDMR (SEQ ID NO: 34), or an amino acid sequence differing from SEQ ID NO: 33 or SEQ ID NO: 34 by one, two or three, preferably one or two, amino acid mutations, preferably amino acid substitutions, and [0680] (b) a second polypeptide comprising a variable domain V.sub.B comprising CDRb1, CDRb2 and CDRb3, wherein [0681] the CDRb1 comprises or consists of the amino acid sequence SGHNS (SEQ ID NO: 10) or an amino acid sequence differing from SEQ ID NO: 10 by one or two amino acid mutations, preferably amino acid substitutions, and [0682] the CDRb3 comprises or consists of the amino acid sequence ASSX.sub.1GX.sub.2X.sub.3DX.sub.4QY (SEQ ID NO: 327), wherein X.sub.1 is P, A or T, X.sub.2 is A or S, X.sub.3 is T or I, and X.sub.4 is K or A, or an amino acid sequence differing from SEQ ID NO: 327 by one, two or three amino acid mutations, preferably amino acid substitutions. [0683] 2. The antigen binding protein of item 1, wherein [0684] (a) the CDRa2 comprises or consists of the amino acid sequence FGPYGKE (SEQ ID NO: 32), or an amino acid sequence differing from SEQ ID NO: 32 by one, two or three amino acid mutations, preferably amino acid substitutions, and/or [0685] (b) the CDRb2 comprises or consists of the amino acid sequence FQNTAV (SEQ ID NO: 36) or a CDRb2 amino acid sequence differing from SEQ ID NO: 36 by one, two, three, four, five or six amino acid mutations, preferably amino acid substitutions. [0686] 3. The antigen binding protein of item 1 or 2, wherein [0687] Position 27 of CDRa1 according to IMGT is V or is substituted by an amino acid selected from L, I, M, F, A, T, N, Q, H, E, D and S, particularly selected from T, N, S and I, [0688] Position 28 of CDRa1 according to IMGT is K or is substituted by an amino acid selected from R, Q, H, N, A, V, S, G, L, I and T, particularly selected from R. A and S, [0689] Position 38 of CORa1 according to IMGT is D or is substituted by an amino acid selected from E, N, Q, H, K and R, particularly N, [0690] Position 64 of CDRa2 according to IMGT is K or is substituted by an amino acid selected from R, Q, H, N, T, V, A, L, I, M and F, particularly selected from R, T and V, [0691] Position 114 of CDRa3 according to IMGT is L or Y or is substituted by an amino acid selected from M, W, H, Q, A, I, K, R, V, D, E, F and N particularly selected from H, Q, A, I, K, R, V, D, E, F and N, more particularly selected from H, Q, A and I, [0692] Position 56 of CDRb2 according to IMGT is F or is substituted by an amino acid selected from Y, M, L, W, H, V, I and A, particularly selected from Y, M and L, [0693] Position 57 of CDRb2 according to IMGT is Q or is substituted by an amino acid selected from N, R, D, E, Q, H, K and K, particularly N, with the proviso that the amino acid at position 57 is not N when the amino acid at position 63 is T or S, [0694] Position 58 of CDRb2 according to IMGT is N or is substituted by an amino acid selected from Q, H, D, K, R, S and T, particularly S. [0695] Position 63 of CDRb2 according to IMGT is T or is substituted by an amino acid selected from S, V, A, D, Q and E, particularly selected from S and E, with the proviso that the amino acid at position 63 is not T or S when the amino add at position 57 is N, [0696] Position 64 of CDRb2 according to IMGT is A or is substituted by an amino acid selected from V, L, I, S, G and T, particularly T, [0697] Position 65 of CDRb2 according to IMGT is V or is substituted by an amino acid selected from L, I, M, A, T, F and S, particularly selected from I, L and T, [0698] Position 108 of CDRb3 according to IMGT is P, A or T or is substituted by an amino acid selected from V, L, I, S, G, R, K, N and Q, particularly selected from R and S, with the proviso that the amino acid at position 108 is not N when the amino acid at position 110 is T or S. [0699] Position 110 of CDRb3 according to IMGT is A or S or is substituted by an amino acid selected from V, L, I, G, T and C, particularly T, with the proviso that the amino acid at position 110 is not T or S when the amino acid at position 108 is N, [0700] Position 113 of CDRb3 according to IMGT is T or I or is substituted by an amino acid selected from V, L, and G, and [0701] Position 115 of CDRb3 according to IMGT is T, K or A or is substituted by an amino acid selected from G, L, I, V, R, Q, N, Y, H, E and F, particularly selected from L, I, V, R, Q, N, Y, H, E and F, more particularly from L, I, V and R. [0702] 4. The antigen binding protein of any one of items 1 to 3, wherein said antigen binding protein specifically binds to the amino acid sequence of SEQ ID NO: 50 in a complex with a MHC protein, in particular a HLA protein, more particularly HLA-A, even more particularly HLA-A*02. [0703] 5. The antigen binding protein of any one of items 1 to 4, wherein said antigen binding protein specifically binds to a functional epitope comprising or consisting of at least 3, 4 or 5 amino acid positions selected from the group consisting of positions 3, 5, 6, 7 and 8, in particular 3, 5 and 7, of SEQ ID NO: 50, preferably to a functional epitope consisting of amino acid positions 3, 5 and 7, or 3, 5, 6 and 7, or 3, 5, 7 and 8, or 3, 5, 6, 7 and 8 of SEQ ID NO: 50, but preferably not amino acid positions 1 and 4 of SEQ ID NO: 50. [0704] 6. The antigen binding protein of any one of items 1 to 4, wherein said antigen binding protein specifically binds to a functional epitope comprising or consisting of at least 6 or 7 amino acid positions selected from the group consisting of positions 1, 3, 4, 5, 6, 7 and 8 of SEQ ID NO: 50. [0705] 7. The antigen binding protein of any one of items 1 to 6, wherein said antigen binding protein binds to a complex of said PRAME antigenic peptide and a MHC protein, in particular a HLA protein, more particularly HLA-A, even more particularly HLA-A*02, with a K.sub.D of 5100 nM, 550 nM. 510 nM, preferably ≤5 nM. [0706] 8. The antigen binding protein of any one of items 1 to 7, wherein said antigen binding protein does not significantly bind to at least 1, at least 2, at least 3, at least 4, at least 5, at least 10, at least 20 or all similar peptides selected from the group consisting of TMED9-001 (SEQ ID NO: 51), CAT-001 (SEQ ID NO: 52), DDX60L-001 (SEQ ID NO: 53), LRRC70-001 (SEQ ID NO: 54), PTPLB-001 (SEQ ID NO: 55), HDAC5-001 (SEQ ID NO: 56), VPS138-002 (SEQ ID NO: 57), ZNF318.001 (SEQ ID NO: 58), CCDC51-001 (SEQ ID NO: 59), IFT17-003 (SEQ ID NO: 60), DIAPH1-004 (SEQ ID NO: 62), FADS2-001 (SEQ ID NO: 63), FRYL-003 (SEQ ID NO: 64), GIMAP8-001 (SEQ ID NO: 65), HSF1-001 (SEQ ID NO: 66), KNT-001 (SEQ ID NO: 67), MAU-001 (SEQ ID NO: 68), MCM4-001 (SEQ ID NO: 69), MPPE1-001 (SEQ ID NO: 71), MYO1 &002 (SEQ ID NO: 72), PRR12-001 (SEQ ID NO: 73), PTRF-003 (SEQ ID NO: 74), RASGRP1-001 (SEQ ID NO: 75), SMARCD1-001 (SEQ ID NO: 76), TGM2-001 (SEQ ID NO: 77), VAV1-001 (SEQ ID NO: 78), VIM-009 (SEQ ID NO: 317), FARSA-001 (SEQ ID NO: 306), ALOX158-003 (SEQ ID NO: 304), FAM114A2-002 (SEQ ID NO: 305), GPR56-002 (SEQ ID NO: 307), IGHD-002 (SEQ ID NO: 308), NOMAP-3-0972 (SEQ ID NO: 309), NOMAP-3-1265 (SEQ ID NO: 310), NOMAP-3-1408 (SEQ ID NO: 311), NOMAP-3-1587 (SEQ ID NO: 312), NOMAP-3-1768 (SEQ ID NO: 313), NOMAP-5-0765 (SEQ ID NO: 314), PDCD10-004 (SEQ ID NO: 315), TSN-001 (SEQ ID NO: 316), ARMC9-002 (SEQ ID NO: 187), CLI-001 (SEQ ID NO: 188), COPG1-001 (SEQ ID NO: 190), COPS7A-001 (SEQ ID NO: 192), EIF-009 (SEQ ID NO: 194), EXT2-006 (SEQ ID NO: 196), LMNA-001 (SEQ ID NO: 198), PKM-005 (SEQ ID NO: 200), PSMB3-002 (SEQ ID NO: 202), RPL-007 (SEQ ID NO: 204). SPATS2L-003 (SEQ ID NO: 206). SYNE1-002 (SEQ ID NO: 208), TGM2-002 (SEQ ID NO: 210) and TPR-004 (SEQ ID NO: 212), in a complex with a MHC protein, preferably said antigen binding protein does not significantly bind to IFT17-003 (SEQ ID NO: 60) in a complex with a MHC protein. [0707] 9. The antigen binding protein of any one of items 1 to 8, wherein the antigen binding protein is multispecific, e.g. tetra-, tri- or bispecific, preferably bispecific, in particular said antigen binding protein is a bispecific TCR, a bispecific antibody or a bispecific TCR-antibody molecule. [0708] 10. The antigen binding protein of any one of items 1 to 9, wherein the first and the second polypeptide are comprised in a single polypeptide chain or two polypeptide chains, preferably wherein V.sub.A is comprised in a first polypeptide chain and V.sub.B is comprised in a second polypeptide chain. [0709] 11. The antigen binding protein of any one of items 1 to 10, wherein V.sub.A further comprises one or more framework regions, preferably all framework regions, selected from the group consisting of FR 1-a, FR2-a, FR3-a and FR4-a, wherein [0710] FR1-a comprises or consists of the amino acid sequence of SEQ ID NO: 345 or SEQ ID NO: 346, or an amino acid sequence at least 85%, 90% or 95% identical to SEQ ID NO: 345, preferably comprising K or N, more preferably K, at position 20 and/or L or M more preferably L, at position 2; [0711] FR2-a comprises or consists of the amino acid sequence of SEQ ID NO: 347 or SEQ ID NO: 348, or an amino acid sequence at least 85%, 90% or 95% identical to SEQ ID NO: 347, preferably comprising L, I or M, more preferably L or I, at position 39, A or D, more preferably A, at position 47, K or W, preferably K, at position 44, F or A, preferably F.sub.c at position 52 and/or Y or V, preferably Y, at position 55; [0712] FR3-a comprises or consists of the amino acid sequence of SEQ ID NO: 349 or an amino acid sequence at least 85%, 90% or 95% identical to SEQ ID NO: 349, preferably comprising T or K, more preferably T, at position 92 and/or D or G, preferably D, at position 93; [0713] FR4-a comprises or consists of the amino acid sequence of SEQ ID NO: 350 or an amino acid sequence at least 85%, 90% or 95% identical to SEQ ID NO: 350; and V.sub.B further comprises one or more framework regions, preferably all framework regions, selected from the group consisting of FR1-b, FR2-b, FR3-b and FR4-b, wherein [0714] FR1-b comprises or consists of the amino acid sequence of SEQ ID NO: 351 or SEQ ID NO: 352 or an amino acid sequence at least 85%, 90% or 95% identical to SEQ ID NO: 351, preferably comprising H or N, more preferably H, at position 10, E, L or K, preferably E, at position 11 and/or R or H, at position 22; [0715] FR2-b comprises or consists of the amino acid sequence of SEQ ID NO: 353 or an amino acid sequence at least 85%, 90% or 95% identical to SEQ ID NO: 353, preferably comprising R or K, more preferably R, at position 43, E or Q, preferably E, at position 44, M or P, more preferably P, at position 46, and/or R or Q, more preferably Q, at position 48; [0716] FR3-b comprises or consists of the amino acid sequence of SEQ ID NO: 354 or SEQ ID NO: 355 or an amino acid sequence at least 85%, 90% or 95% identical to SEQ ID NO: 354, preferably comprising D, A, E. R, K Q. N or S, more preferred D, A, E, Q, N or S, more preferably D or A, even more preferably D, at position 84; and [0717] FR4-b comprises or consists of the amino acid sequence of SEQ ID NO: 356 or an amino acid sequence at least 85%, 90% or 95% identical to SEQ ID NO: 356. [0718] 12. The antigen binding protein of any one of items 1 to 11, wherein [0719] V.sub.A comprises or consists of the amino acid sequence of SEQ ID NO: 132 or an amino acid sequence at least 85%, 90% or 95% identical to SEQ ID NO: 132, preferably comprising a CDRa1 of SEQ ID NO: 16, a CDRa2 of SEQ ID NO: 32 and a CDRa3 of SEQ ID NO: 33, SEQ ID NO: 34, or SEQ ID NO: 9, and further K or N, preferably K, at position 20, L, M, or I, preferably L or I, at position 39, K or W, preferably K, at position 44, F or A, preferably F, at position 52, Y or V, preferably Y, at position 55, T or K, preferably T, at position 92 and/or D or G, preferably D, at position 93; and [0720] V.sub.B comprises or consists of the amino acid sequence of SEQ ID NO: 134 or an amino acid sequence at least 85%, 90% or 95% identical to SEQ ID NO: 134, preferably comprising a CDRb1 of SEQ ID NO: 10, a CDRb2 of SEQ ID NO: 36, and a CDRb3 of SEQ ID NO: 48, SEQ ID NO: 49, SEQ ID NO: 47, SEQ ID NO: 281, SEQ ID NO: 292. SEQ ID NO: 294, SEQ ID NO: 297, SEQ ID NO: 298, SEQ ID NO: 301 or SEQ ID NO: 283, and further E, L or K, preferably E, at position 11, R or H at position 22. E or Q, preferably E, at position 44, P or M, preferably P, at position 46, Q or R, preferably Q, at position 48 and/or D, A, E, Cr, N, or S, preferably D or A, at position 84. [0721] 13. The antigen binding protein of any one of items 1 to 12, wherein [0722] V.sub.A comprises or consists of the amino acid sequence of SEQ ID NO: 132, SEQ ID NO: 129, SEQ ID NO: 137 or SEQ ID NO: 142, and [0723] V.sub.B comprises or consists of the amino acid sequence of SEQ ID NO: 134, SEQ ID NO: 130, SEQ ID NO: 135, SEQ ID NO: 136, SEQ ID NO: 138, SEQ ID NO: 139, SEQ ID NO: 140, SEQ ID NO: 141, SEQ ID NO: 144, SEQ ID NO: 145. SEQ ID NO: 146, SEQ ID NO: 147 or SEQ ID NO:148. [0724] 14. The antigen binding protein of any of items 1 to 13, further comprising one or more of the following: [0725] (i) one or more further antigen binding sites; [0726] (ii) a transmembrane region, optionally including a cytoplasmic signalling region; [0727] (iii) a diagnostic agent; [0728] (iv) a therapeutic agent; and [0729] (v) PK modifying moiety. [0730] 15. The antigen binding protein of any one of items 1 to 14, further comprising an antibody light chain variable domain (V.sub.L) and an antibody heavy chain variable domain (V.sub.H). [0731] 16. The antigen binding protein of item 15, wherein VL and VH bind to an antigen selected from the group consisting of CD2, CD3, in particular CD3γ, CD345, and/or CD3e, CD4, CD5, CD7, CD8, CD10, CD11b, CD11c, CD14, CD16, CD18, CD22, CD25, CD28, CD32a, CD32b, CD33, CD41, CD41b, CD42a, CD42b, CD44, CD45RA, CD49, CD55, CD56, CD61, CD64, CD68, CD90, CD94, CD95, CD117, CD123, CD125, CD134, CD137, CD152, CD163, CD193, CD203c, CD235a, CD278, CD279, CD287, Nkp46, NKG2D, GITR, F.sub.cεRI, TCRα/β and TCRγ/δ, HLA-DR and 4-1 BB, or combinations thereof and/or bind to an effector cell, in particular a T cell or natural killer cell (NK cell). [0732] 17. The antigen binding protein of item 15 or 16, wherein the antigen binding protein comprises a first and a second polypeptide chain. [0733] wherein [0734] the first polypeptide chain is represented by a formula [Ia]:

V.sub.1-L.sub.1-D.sub.1-L.sub.2-V.sub.2-L.sub.3-D.sub.2  [Ia], [0735] and the second polypeptide chain is represented by a formula [IIa]

V.sub.3-L.sub.4-D.sub.3-L.sub.5-V.sub.4-L.sub.6-D.sub.4  [IIa], [0736] wherein [0737] V.sub.1, V.sub.2, V.sub.3, and V.sub.4 are variable domains, wherein one of V.sub.1 to V.sub.4 is V.sub.A, one is V.sub.B, one is V.sub.L and one is V.sub.H; [0738] D.sub.1, D.sub.2, D.sub.3, and D.sub.4 are dimerization domains and may be present or absent, wherein D.sub.1 and D.sub.3, and D.sub.2 and 04, specifically bind to each other and at least one pair of D.sub.1 and D.sub.3, or D.sub.2 and D.sub.4 is present; and [0739] L.sub.1, L.sub.2, L.sub.3, L.sub.4, L.sub.5, and L.sub.6 are linkers, wherein L.sub.1 and L.sub.4 are present and L.sub.2, L.sub.3, L.sub.5, and L.sub.6 may be present or absent. [0740] 18. The antigen binding protein of any of items 15 to 17, wherein the antigen binding protein comprises a first and a second polypeptide chain, [0741] wherein [0742] the first polypeptide chain is represented by a formula [Ib]:

V.sub.1-L.sub.1-V.sub.2-L.sub.3-D.sub.2  [Ib], [0743] and the second polypeptide chain is represented by a formula [IIb]:

V.sub.3-L.sub.4-V.sub.4-L.sub.6-D.sub.4  [IIb], [0744] wherein [0745] V.sub.1, V.sub.2, V.sub.3, V.sub.4, are variable domains, preferably wherein one of V.sub.1 and V.sub.2 is V.sub.A, one of V.sub.3 and V.sub.4 is V.sub.A and of the remaining two variable domains one is V.sub.L and the other is V.sub.H; [0746] D.sub.2 and D.sub.4 are dimerization domains, preferably F.sub.c-domains; and [0747] L.sub.1, L.sub.3, L.sub.4 and L.sub.6 are linkers, wherein L.sub.3, and L.sub.6 may be present or absent. [0748] 19. The antigen binding protein of item 17 or 18, wherein [0749] (1) V.sub.1 is V.sub.H, V.sub.2 is V.sub.B, V.sub.3 is V.sub.A, and V.sub.4 is V.sub.L: [0750] (2) V.sub.1 is V.sub.B, V.sub.2 is V.sub.H, V.sub.3 is V.sub.L, and V.sub.4 is V.sub.A; [0751] (3) V.sub.1 is V.sub.B, V.sub.2 is V.sub.L, V.sub.3 is V.sub.H, and V.sub.4 is V.sub.A; [0752] (4) V.sub.1 is V.sub.L, V.sub.2 is V.sub.B, V.sub.3 is V.sub.A, and V.sub.4 is V.sub.H; [0753] (5) V.sub.1 is V.sub.H, V.sub.2 is V.sub.B,V.sub.3 is V.sub.1, and V.sub.4 is V.sub.A; [0754] (6) V.sub.1 is V.sub.B, V.sub.2 is V.sub.H, V.sub.3 is V.sub.A, and V.sub.4 is V.sub.L; [0755] (7) V.sub.1 is V.sub.L, V.sub.2 is V.sub.B, V.sub.3 is V.sub.H, and V.sub.4 is V.sub.A; [0756] (8) V.sub.1 is V.sub.B, V.sub.2 is V.sub.1, V.sub.3 is V.sub.A, and V.sub.4 is V.sub.H: [0757] (9) V.sub.1 is V.sub.H, V.sub.2 is V.sub.L, V.sub.3 is V.sub.A, and V.sub.4 is V.sub.B; [0758] (10) V.sub.L is V.sub.L, V.sub.2 is V.sub.H, V.sub.3 is V.sub.A, and V.sub.4 is V.sub.B; [0759] (11) V.sub.1 is V.sub.H, V.sub.2 is V.sub.L, V.sub.3 is V.sub.B, and V.sub.4 is V.sub.A: or [0760] (12) V.sub.L is V.sub.L, V.sub.2 is V.sub.H, V.sub.3 is V.sub.B, and V.sub.4 is V.sub.A. [0761] 20. The antigen binding protein of any one of items 1 to 19, comprising [0762] a first polypeptide chain selected from SEQ ID NO: 100, 103, 105, 106, 111, 122, 126, 128, 151, 155, 156,157, 158, 159, 166, 167, 169, 171, 173, 175, 177, 178, 179, 180, 181, 183, 189, 191,193, 195, 197, 199, 201, 203, 205, 207, 209, 211, 213, 215, 217, 285, 291, 295, 299 and 303, and [0763] a second polypeptide chain selected from SEC) ID NO: 101, 102, 104, 107, 110, 119, 121, 131, 133, 143, 152,160, 161, 162, 163, 164, 165, 168, 170, 172, 174, 176, 182, 184, 185, 186, 216, 218, 220, 222, 224, 228, 230, 232, 234, 236, 238, 240, 242, 244, 246, 248, 250, 252, 254, 256, 258, 260, 262, 264, 266, 268, 270, 272, 274, 276, 278, 282, 284, 296 or 300. [0764] 21. The antigen binding protein of any one of items 1 to 13, wherein V.sub.A is comprised in a TCR α- or γ-chain: and/or V.sub.B is comprised in a TCR β- or α-chain. [0765] 22. An isolated nucleic acid comprising a sequence encoding the antigen binding protein of any one of items 1 to 21. [0766] 23. A vector comprising the nucleic acid of item 22. [0767] 24. A host cell comprising the antigen binding protein of any one of items 1 to 21, or the nucleic acid of item 22, or the vector of item 23. [0768] 25. The host cell of item 24, wherein the host cell is [0769] a lymphocyte, preferably a T lymphocyte or T lymphocyte progenitor cell, for example a CD4 or CD8 positive T cell or [0770] a cell for recombinant expression, such as a Chinese Hamster Ovary (CHO) cell or a yeast cell. [0771] 26. A pharmaceutical composition comprising the antigen binding protein of any one of items 1 to 21, the nucleic acid of item 22, the vector of item 23, or the host cell of item 24 or 25 and a pharmaceutically acceptable carrier. [0772] 27. A method of producing the antigen binding protein according to any one of items 1 to 21, comprising [0773] a. providing a host cell, [0774] b. providing a genetic construct comprising a coding sequence encoding the antigen binding protein of any of items 1 to 21, [0775] c. introducing said genetic construct into said host cell, and [0776] d. expressing said genetic construct by said host cell. [0777] 28. The method of item 27, further comprising the isolation and purification of the antigen binding protein from the host cell and, optionally, reconstitution of the antigen binding protein in a T cell. [0778] 29. The antigen binding protein of any one of items 1 to 21, the nucleic acid of item 22, the vector of item 23, the host cell of item 24 or 25, or the pharmaceutical composition of item 26 for use in medicine. [0779] 29. The antigen binding protein of any one of items 1 to 21, the nucleic acid of item 22 or the vector of item 23, the host cell of item 24 or 25 or the pharmaceutical composition of item 26 for use in the diagnosis, prevention, and/or treatment of a proliferative disease, such as cancer, wherein said cancer is selected from the group of cancers consisting of acute myeloid leukemia, breast cancer, cholangiocellular carcinoma, gallbladder cancer, glioblastoma, hepatocellular carcinoma, head and neck squamous cell carcinoma, melanoma, amelanotic melanoma, non-Hodgkin lymphoma, non-small cell lung cancer adenocarcinoma, non-small cell lung cancer, squamous cell non-small cell lung cancer, ovarian cancer, esophageal cancer, renal cell carcinoma, small cell lung cancer, urinary bladder carcinoma, uterine and endometrial cancer, osteosarcoma, chronic lymphocytic leukemia, colorectal carcinoma, and synovial sarcoma.