Anti-sialyl-tn antigen-binding proteins

Abstract

The present application relates to antigen-binding proteins that bind Sialyl-Thomsen-nouveau antigen. The antigen-binding proteins find application in the treatment of diseases and disorders, such as cancer.

Description

[0001] Antigen-binding protein

[0002] This application claims priority from GB2413300.1 filed 10 September 2024, the contents and elements of which are herein incorporated by reference for all purposes.

[0003] FIELD OF THE INVENTION

[0004] The present invention relates to antigen-binding proteins that bind Sialyl-Thomsen-nouveau antigen. The antigenbinding proteins find application in the treatment of diseases and disorders, such as cancer.

[0005] BACKGROUND TO THE INVENTION

[0006] Cancer is a complex disease for which there is still significant unmet medical need. Evasion of the immune system is now widely recognised as a key hallmark of cancer progression and the interplay between the host immune system and the tumour has been an area of intense assessment in recent years. Tumour-infiltrating lymphocytes (TIL) have the capacity to directly kill tumour cells and there is emerging clinical evidence that patients with increased TIL numbers have improved prognosis. However, the tumour microenvironment (TME) is a hostile environment due, in part, to dysfunctional cytotoxic effector cells and the presence of aberrantly glycosylated immunosuppressive cell types. Together, these features may contribute to lack of response to checkpoint blockade in a subset of cancer patients.

[0007] Glycans are carbohydrate-based polymers that play a central role in fundamental aspects of cell and organismal biology including cellular interactions that involve both, normal and pathological processes. Added to protein and lipid backbones during their biosynthesis, glycans can be found as a dense layer on the cell surface. In contrast to protein synthesis, glycan synthesis is a template-independent process and glycan sequences are not directly encoded by single genes. Instead, glycans originate from the coordinated activity and relocation of several biosynthetic enzymes.

[0008] Due to its central role in cellular and organismal function, defects in glycosylation are important contributors to human disease. In tumour cells, glycans, displayed on the cell surface as part of glycoproteins and -lipids, are fundamentally altered, resulting in cells with distinct glycan matrices. As such, aberrant cell-surface glycosylation is recognised as a hallmark of cancer that can result in immunosuppression. Conversely, in normal cells with no disruptions to the expression or localisation of the enzymes that carry out glycosylation, there are very low levels of tumour-associated glycans. Commonly upregulated tumour-associated glycans, such as Thomsen-nouveau antigen (Tn) and Sialyl-Thomsen-nouveau antigen (Sialyl-Tn or STn), are associated with poor prognosis. The aberrant production of Tn and STn on tumour tissues versus healthy tissues makes these antigens excellent actionable targets for cancer therapies. The production of Tn and STn (hereinafter, sometimes collectively

[0009] 1

[0010] 008828600 referred to in this application as "Tumour-Associated Glycans" or "TAG") by tumour cells acts as a phenotypic marker of cancer cells and can control many aspects of tumour development. TAG can induce immunosuppressive signals through their interaction with glycan-binding receptors to alter immune cell signalling, differentiation, and cytokine responses towards anti-inflammatory or immunosuppressive phenotypes. Further, STn has been shown to inhibit the cytotoxicity of NK cells, induce TGF-|3 production, an immunosuppressive cytokine, by macrophages, and prevent maturation of antigen presenting DC cells resulting in tolerogenic tumours Ogata S, et al Cancer Res. 1992 Sep l;52(17):4741-6. PMID: 1511439. Takamiya R, et al Glycobiology. 2013 Feb;23(2):178-87. doi: 10.1093 / glycob / cwsl39. Epub 2012 Oct 3. PMID: 23035012 and Carrascal MA, et al Mol Oncol. 2014 May;8(3):753-65. doi: 10.1016 / j.molonc.2014.02.008. Epub 2014 Mar 6. PMID: 24656965; PMCID: PMC5528624. In such TMEs, there is a need to drive T-cell activation in the tumour at sites of immunosuppression, whilst overcoming this TAG induced suppression may render therapeutics designed to activate T cells less effective. Preclinical proof of concept for TAG blockade is provided by studies showing that removal of TAGs from the surface of mouse tumour cells re-sensitise tumours to checkpoint blockade. Glyco-biology is emerging as an important aspect of tumour immunobiology, reflected in the emerging competitive landscape where there are clinical efforts to develop antibody therapies to specifically target glycoimmunology pathways for cancer therapy e.g., E-602 Palleon Therapeutics and NEO-201 Precision Biologies. Nonetheless, there remains a need in the art for new anti-cancer therapeutics. The present invention has been devised in light of the above considerations.

[0011] SUMMARY OF THE INVENTION

[0012] The present inventors recognised that therapeutic approaches that combine T-cell engagers, or CD137 agonists, with glycan targeting may be beneficial in the treatment of cancer. To our knowledge, no therapeutic approaches that bind to these target types in combination exist in the art.

[0013] In order to allow simultaneous targeting of CD137 and Sialyl-Tn (STn), the present inventors initiated a development program for an antibody which binds STn. The anti-STn antibody was required to be human or humanised, to bind STn specifically and with high affinity, be stable, be free from developmental liabilities as a result of the presence of high-risk motifs, and be suitable for the preparation of bispecific antibodies binding both STn and comprising a second antigen-binding site for CD137 for the administration to human patients.

[0014] However, the development of an anti-STn antibody which met the above requirements proved to be fraught with unexpected difficulty and setbacks, not normally seen in the development of monoclonal antibodies.

[0015] Specifically, the present inventors initially attempted to generate antibodies that bind to STn by immunising rats and mice with two suitably glycosylated proteins, which were known immunogens (Example 1). Six IgGl clones

[0016] 2

[0017] 008828600 were identified which were shown to bind to cells expressing STn. Alanine scanning of the most promising clone was performed to guide humanisation of the identified antibodies. Despite extensive efforts, humanisation of this clone failed, as it was not possible to remove all high-risk motifs for developability within the heavy chain CDR sequences without associated loss of function. As a consequence, this approach for generating suitable anti-STn antibodies surprisingly failed to produce an anti-STn antibody suitable for the administration to human patients.

[0018] In view of the unexpected failure of the rat / mouse immunisation approach, the present inventors next tried to generate a suitable anti-STn antibody through selection from a naive Fab library (Example 2). However, despite multiple rounds of selection and using different Fab libraries, no anti-STn antibodies could be identified when the libraries were screened with one of the antigens which had been successfully used in the initial mouse / rat immunisations in Example 1. An alternative antigen was then used for screening of the libraries but only a limited number of low stability clones could be identified which were found not to be suitable for further development. Screening with a third antigen was then performed which was expected to yield Fab clones specific for STn. However, despite several Fab clones capable of binding STn being initially identified following a number of selection rounds, once these Fabs were expressed as complete antibody molecules comprising an antigen-binding site for human CD137 in the CH3 domain, the majority of the antibody molecules did not show monomeric stability. The limited number of stable clones were investigated further but showed no activity in a CD8+ T-cell activation assay, demonstrating that these molecules were not able to agonise CD137 in a TAG- dependent manner, indicating a lack of sufficient specificity for STn. Further selections of the libraries using cells expressing STn with counter selection for other glycans, did not yield any Fab clones capable of binding STn. Thus, despite the unusually extensive efforts by the inventors to identify anti-STn antibodies using a large number of different antigens, no anti-STn antibody suitable for further development as a human therapeutics could be identified using naive library selection.

[0019] In view of the failure of two standard approaches in the development of an anti-STn antibody, the inventors next attempted humanisation of a further mouse anti-STn antibody (Example 3). Although initial humanisation produced stable antibody molecules, the resulting molecules had lost binding to cell-expressed STn and showed only limited binding to synthetic STn. Consequently, back mutations were introduced in an attempt to restore binding activity. With exception of a single back mutation (W50Y), the backmutations either resulted in instable molecules or were not sufficient to restore binding. However, even with this successful back mutation, binding to STn was still significantly weaker than seen with the parent mouse antibody (28% Emax). Following further screening, an additional backmutation (T58I) was identified by the inventors which, in combination with mutation W50Y, was able to restore binding activity to a level comparable to that seen with the mouse antibody. The resulting antibody clone was then affinity matured (Example 4) and the lead clone further modified to

[0020] 3

[0021] 008828600 remove high risk motifs for developability of the antibody (Example 5). In contrast to Example 1, all high-risk motifs could be successfully removed by mutation, thus demonstrating the unpredictability associated with attempts to prepare humanised antibodies based on known mouse antibodies. When taken together with the unexpected failure of the extensive naive selection experiments performed by the inventors, as described in Example 1, this demonstrates the unusual difficulty associated with preparing the anti-STn antibodies described herein.

[0022] The anti-STn antibodies identified in Example 5 were further shown to be stable when an anti-CD137 binding site was introduced into the CH3 domain (Example 6) and were shown to be capable of activating T cells (Example 7), demonstrating that the antibody retained sufficient specificity for STn. Further deimmunisation mutations were then introduced to produce a molecule suitable for the administration to human patients (Example 8). A mouse surrogate of the human bispecific antibody was further tested in mouse tumour models and was shown to provide a significant survival benefit (Example 16).

[0023] Thus, in one embodiment of the invention as herein described, there is provided an antigen-binding protein that binds to STn. In another embodiment, there is provided an antigen-binding protein which binds to STn and Tn. In one such embodiment, the antigen-binding protein of the present invention binds to clustered STn.

[0024] In one embodiment of the invention as herein described, there is provided an antigen-binding protein which binds to STn and comprises:

[0025] CDRH1 of X1HAX2H (SEQ ID NO: 109), wherein Xi is alanine (A), aspartic acid (D) or glutamic acid (E), tryptophan (W) or tyrosine (Y); and X2 is isoleucine (I) or methionine (M);

[0026] CDRH2 of YISPGX3X4DX5KYSQKFQG (SEQ ID NO: 110), wherein X3is asparagine (N) or glutamine (Q); X4is alanine (A), aspartic acid (D) or glutamic acid (E), glycine (G), lysine (K), asparagine (N), glutamine (Q) or tyrosine (Y); and X5is isoleucine (I), leucine (L), glutamine (Q), threonine (T) or valine (V);

[0027] CDRH3 of SYYGXs (SEQ ID NO: 111), wherein Xs is glutamic acid (E) or histidine (H);

[0028] CDRL1 of RASX7NIYSNLA (SEQ ID NO: 112), wherein X7is glutamic acid (E) or glutamine (Q);

[0029] CDRL2 of AAXsXgLXioXu (SEQ ID NO: 113), wherein Xs is serine (S) or threonine (T); Xg is asparagine (N) or arginine (R); X10 is alanine (A) or glutamine (Q); and Xu is aspartic acid (D), leucine (L) or serine (S); and / or

[0030] CDRL3 of QHFX12GTPYT (SEQ ID NO: 114), wherein X12 is phenylalanine (F) or tryptophan (W); and wherein the CDR sequences are defined according to Kabat.

[0031] In one embodiment of the invention as herein described, the antigen-binding protein additionally binds to CD137.

[0032] 4

[0033] 008828600 In one embodiment of the invention as herein described, there is provided a nucleic acid molecule or set of nucleic acid molecules encoding the antigen-binding protein of the invention. In another embodiment, there is provided a vector or set of vectors comprising a nucleic acid or set of nucleic acid molecules encoding the antigen-binding protein of the invention as herein described. In one embodiment, there is provided a recombinant host cell comprising a nucleic acid molecule or set of nucleic acid molecules or a vector or set of vectors encoding the antigen-binding protein of the invention as herein described.

[0034] In one embodiment, there is provided a method of producing the antigen-binding protein of the invention as herein described comprising culturing a recombinant host cell under conditions for production of the antigen-binding protein. In a further embodiment, there is provided a method of further isolating and / or purifying the antigenbinding protein of the invention as herein described.

[0035] In one embodiment, there is provided an antigen-binding protein of the invention as herein described for use as a medicament.

[0036] In one embodiment, there is provided an antigen-binding protein of the invention as herein described for use in a method of treating cancer in an individual.

[0037] In one embodiment, there is provided a method of treating cancer in an individual, wherein the method comprises administering to the individual a therapeutically effective amount of the antigen-binding protein of the invention as herein described.

[0038] In one embodiment, there is provided a pharmaceutical composition comprising the antigen-binding protein of the invention as herein described and a pharmaceutically acceptable excipient.

[0039] In one embodiment, there is provided a pharmaceutical composition for use in the treatment of cancer in an individual, comprising the antigen-binding protein of the invention as herein described.

[0040] In one embodiment, there is provided the use of an antigen-binding protein of the invention as herein described for the manufacture of a medicament for treating cancer.

[0041] DESCRIPTION OF DRAWINGS / FIGURES

[0042] Figure 1: Binding of Humanised mAbm4 Variants to STn-BSA as assessed by ELISA. The mouse parent was tested in human IgGl form with mouse variable domains of mAbm4.

[0043] 5

[0044] 008828600 Figure 2: Binding of humanised variants to TAG-overexpressing HEK293 (HEK.STGGalNAcl.l) -cells transfected with STGGalNAcl.l for overexpression of TAG. The mouse parent was tested in human IgGl form with mouse variable domains of mAbm4. N = 1.

[0045] Figure 3: Binding of backmutated Variants of Humanised mAbm4 to HEK.STBGalNAcl.l cells. N = 1 with duplicates. Error bars indicate SD.

[0046] Figure 4: Binding of backmutated Variants of Humanised mAbm4 to HEK.STBGalNAcl.l cells. N = 1 with duplicates. Error bars indicate SD.

[0047] Figure 5: Binding to HEK.STBGAINAcl.l cells by humanised variants of mAbm4 containing the backmutation W50Y VH1 domains. N = 1 with duplicates. Error bars indicate SD.

[0048] Figure 6: Binding of humanised variant VH4 with additional W50Y and T58I backmutations to HEK.STBGAINAcl.l cells when paired with VL1 and VL2 variant light chains. N = 1 with duplicates. Error bars indicate SD.

[0049] Figure 7: Binding to HEK.STBGalNAcl.l cells by humanised variants containing backmutation W50Y and affinity maturation mutations R98Q and R98L. Mutations W50Y, R98Q, and R98L are located on the VH domain. All variants were paired with the humanised light chain variant VL1. N = 1 with duplicates. Error bars indicate SD. Figure 8: HEK.STBGalNAcl.l cell binding by high-risk motif mitigation variants. N = 1 with duplicates. Error bars indicate SD.

[0050] Figure 9: Binding of humanised variants VH4(W50Y, R98L) / VL1 and VH4(W50Y, R98QJ / VL1 in mAb2format with a CD137 Fcabs to HEK.STBGalNAcl.l. N = 1 with duplicates. Error bars indicate SD.

[0051] Figure 10: Representative CD8+T-cell activation by a humanised mAb2variant CD137-AA / TAG1 in the presence of HEK.STBGalNAcl.l cells. N = 1 with duplicates. Error bars indicate SD.

[0052] Figure 11: CD8+T-cell activation by a humanised mAb2variant CD137-AA / TAGla in the presence of HEK.STBGalNAcl.l cells. N = 1 with duplicates. Error bars indicate SD.

[0053] Figure 12: Cell binding of CD137-AA / TAGla and control compounds to HEK.STBGalNAcl.l. N = 3 with triplicates. Error bars indicate SD.

[0054] Figure 13: Cell binding of CD137-AA / TAGla and control compounds to OV90 cells, endogenously expressing STn at high levels. N = 4 with triplicates. Error bars indicate SD.

[0055] Figure 14: Mucl binding profile of CD137-AA / TAGla at the highest tested concentration of 15 pg / mL, as assessed on a microarray. Probe 1 to 31: Mucl glycopeptide probes, representing clustered and glycan motifs of Tn; Probe 32 to 63: Mucl glycopeptide probes, representing clustered and unclustered glycan motifs of STn; NC: negative control; PCI: NH2-PEG-BI control; PC2: Human IgG control; PC3: Mouse IgG control; PC4: Rabbit IgG control; M: Array Marker (SA-Cy3 + SA-Cy5). N = 1 with triplicates.

[0056] Figure 15: Functional Activity of CD137-AA / TAGla in primary CD8+T-cell activation assay in the presence of the STn-overexpressing cell line HEK.ST6GalNAcl.l. Meant SD, representative of n=5 shown. Error bars indicate SD.

[0057] 6

[0058] 008828600 Figure 16: Functional Activity of CD137-AA / TAGla in primary CD8+T-cell activation assays in presence of endogenous TAG+cell line OV90 (Ovarian adenocarcinoma cell line with reported endogenous expression of TAG at high levels). Mean ± SD, representative of n=3 shown.

[0059] Figure 17: Functional activity of CD137-AA / TAGla in primary CD8+T-cell activation assays in presence of endogenous TAG+cell Line SNU16 (Gastric carcinoma cell line with endogenous expression of TAG at moderate levels). Mean ± SD, representative of n=3 shown.

[0060] Figure 18 A-D: Tumour volume growth in mm3vs days post inoculation. Graphs show the mean tumour volume for each group, and the tumour volume for individual mice within each group. Error bars indicate SEM. ** indicates p < 0.01.

[0061] Figure 19 A-C: Tumour volume growth in mm3vs days post inoculation. Graphs show the mean tumour volume for each group, and the tumour volume for individual mice within each group. Error bars indicate SEM. *** indicates p <0.001.

[0062] Figure 20: Summary of median survival in days for each group and pairwise statistical analyses (log-rank) in CT26.STnLlF9 colorectal carcinoma tumour model grown subcutaneously in Balb / c mice treated with IgG FITC negative control, TAGla-IgGl or mCD137-AA / TAGla. ** indicates p < 0.01

[0063] Figure 21: Summary of median survival in days for each group and pairwise statistical analyses (log-rank) in CT26.STnLlF9 colorectal carcinoma tumour model grown subcutaneously in Balb / c mice treated with IgG FITC negative control or mCD137-AA / TAGla. * indicates p <0.05

[0064] Figure 22 A-C: Antibody-to-cell binding curves for mCD137-AA / TAGla surrogate and control molecules in the mouse induced-STn-expressing CT26.STnLlF9 colorectal carcinoma cell line, the wild type CT26.WT cell line and mouse splenocytes (negative control). EC50 table values are in nM. N=1

[0065] Figure 23: CD8+ T-cell activation assays showing mouse IL-2 release from co-culture of the induced-STn- expressing CT26.STnLlF9 colorectal carcinoma mouse cell line and DO11.10-mCD137 in response to test antibodies. N=l.

[0066] Figure 24 A and B: Immunophenotyping analysis of CD8 T-cell activation markers Ki67 and Granzyme B in peripheral blood after dosing mice with mCD137-AA / TAGla, TAGla-IgGl-AA or IgG 1 FITC negative control molecule. * p <0.05, ** p <0.01, *** p <0.001.

[0067] Figure 25: Figures 25A, B and C shows an alignment of the amino acid sequences of the VH domain of clones TAGla (SEQ ID NO: 39), TAG124 (SEQ ID NO: 56), TAG129 (SEQ ID NO: 57) and TAG130 (SEQ ID NO: 58). The consecutive numbering of the residues and the numbering of the residues according to the IMGT and Kabat numbering systems, as well as the locations of CDRH1, CDRH2 and CDRH3 according to both the IMGT and Kabat numbering systems, are indicated.

[0068] Figure 26: Figures 26A, B and C shows the amino acid sequence of the VL domain of clones TAGla, TAG124, TAG129 and TAG130 (SEQ ID NO: 40). The consecutive numbering of the residues and the numbering of the

[0069] 7

[0070] 008828600 residues according to the IMGT and Kabat numbering systems, as well as the locations of CDRL1, CDRL2 and CDRL3 according to both the IMGT and Kabat numbering systems, are indicated.

[0071] Figure 27: Glycan structures of motif AO and CO depicted graphically (AO (SEQ ID NO: 115); CO (SEQ ID NO: 116)) and textually (AO (SEQ ID NO: 126); CO (SEQ ID NO: 127)).

[0072] DETAILED DESCRIPTION OF THE INVENTION

[0073] The present invention relates to antigen-binding proteins which bind to certain tumour-associated glycans or "TAG". The term "TAG" as used herein however, refers to the commonly upregulated tumour-associated glycans Thomsen-nouveau antigen (Tn) (GalNAcal-Ser / Thr) and Sialyl-Thomsen-nouveau antigen (Sialyl-Tn or STn) (Neu5Aca2,6GalNAcal-Ser / Thr).

[0074] In one embodiment, there is provided an antigen-binding protein that binds specifically to STn. In one embodiment, there is provided an antigen-binding protein that binds specifically to STn and Tn. In another embodiment, there is provided an antigen-binding protein which preferentially binds to STn over Tn. In one such embodiment, the antigen-binding protein binds to clustered STn. In one such embodiment, the antigenbinding protein binds to clusters of at least three STn residues.

[0075] The term "antigen-binding protein", as used herein, refers to an isolated protein, an immunoglobulin, an antibody or antibody molecule, one or more antibody fragments (e.g., Fabs, etc.), and other, antibody-derived protein constructs, including antibody fusion proteins, such as those comprising domains (e.g., domain antibodies, etc.) which are capable of binding to the antigen. Included are alternative antibody formats, such as diabody, triabody, tetrabody, mini-antibody, and minibody. Also included are alternative scaffolds in which the one or more complementarity-determining regions (CDRs) of any molecules in accordance with the disclosure can be arranged onto a suitable non-immunoglobulin protein scaffold or skeleton, such as an affibody, a SpA scaffold, an LDL receptor class A domain, an avimer or an EGF domain. An antigen-binding protein also includes antigen-binding fragments of such antibodies or other molecules or constructs. Further, an antigen-binding protein of the invention comprises one or more heavy chain variable (VH) domains formatted or incorporated into a full-length antibody, a (Fab')2 fragment, a Fab fragment, a multispecific, e.g., bispecific, antibody molecule or a biparatopic antibody molecule, or an equivalent or derivative of any of the foregoing (such as a single-chain variable fragment (scFV), a tandem diabody (TandAb), etc.), when paired with an appropriate light chain. The antigen-binding protein comprises an antibody that is an immunoglobulin (Ig) G, such as an IgGl, lgG2, lgG3 or lgG4, or an IgM, IgA, IgE or IgD, or a modified variant thereof. The constant domains of the antibody heavy chain may be selected accordingly. The light chain constant domain may be a kappa or a lambda constant domain.

[0076] 8

[0077] 008828600 The term "antibody molecule", as used herein, refers to an immunoglobulin, which may be partly or wholly synthetically produced. The antibody molecule may be humanised. The antibody molecule is preferably a monoclonal antibody molecule. Examples of antibodies are the immunoglobulin isotypes, such IgG, and their isotypic subclasses, such as IgG 1, lgG2, lgG3 and lgG4, as well as fragments thereof. Unless the context requires otherwise, the term "antibody molecule", as used herein, is thus equivalent to "antibody molecule or fragment thereof".

[0078] In relation to binding the terms "specific" or "specifically" may refer to the situation in which the antibody molecule will not show any significant binding to molecules other than its specific binding partner(s), here STn and / or Tn, but preferably STn. The terms "specific" and "specifically" are also applicable where the antigenbinding protein is specific for particular epitopes, such as epitopes on STn, that are carried by a number of antigens, in which case the antigen-binding protein will be able to bind to the various antigens carrying the epitope.

[0079] The antigen-binding proteins described herein show a high level of specificity for the O-linked glycans STn and Tn and do not show any significant binding to other human O-linked glycans, or any binding at all to human N- linked glycans. Thus, in one embodiment, the antigen-binding protein does not bind, or does not show any significant binding, to any human O -linked glycan or N-linked glycan except STn and Tn.

[0080] In one embodiment, there is provided an antigen-binding protein according to the invention as herein described, wherein the antigen-binding protein comprises:

[0081] CDRH1 of X1HAX2H (SEQ ID NO: 109), wherein Xi is alanine (A), aspartic acid (D) or glutamic acid (E), tryptophan (W) or tyrosine (Y); and X2 is isoleucine (I) or methionine (M);

[0082] CDRH2 of YISPGX3X4DX5KYSQKFQG (SEQ ID NO: 110), wherein X3is asparagine (N) or glutamine (Q); X4is alanine (A), aspartic acid (D), glutamic acid (E), glycine (G), lysine (K), asparagine (N), glutamine (Q) or tyrosine (Y); and X5is isoleucine (I), leucine (L), glutamine (Q), threonine (T) or valine (V);

[0083] CDRH3 of SYYGXs (SEQ ID NO: 111), wherein Xs is glutamic acid (E) or histidine (H);

[0084] CDRL1 of RASX7NIYSNLA (SEQ ID NO: 112), wherein X7is glutamic acid (E) or glutamine (Q);

[0085] CDRL2 of AAXsXgLXioXu (SEQ ID NO: 113), wherein Xs is serine (S) or threonine (T); Xg is asparagine (N) or arginine (R); X10 is alanine (A) or glutamine (Q); and Xu is aspartic acid (D), leucine (L) or serine (S); and / or CDRL3 of QHFX12GTPYT (SEQ ID NO: 114), wherein X12 is phenylalanine (F) or tryptophan (W); and wherein the CDR sequences are defined according to Kabat.

[0086] 9

[0087] 008828600 In one embodiment of the invention as herein described, Xi is alanine(A). In another embodiment, X2 is isoleucine(l). In another embodiment, X3 is asparagine(N). In another embodiment, X4is alanine (A) or glutamic acid (E); for example, in one such embodiment, X4is glutamic acid (E). In another embodiment, X5is isoleucine (I), leucine (L), glutamine (Q) or valine (V); for example, in one such embodiment, X5is valine (V). In another embodiment, Xs is histidine (H). In another embodiment, X7is glutamic acid (E). In another embodiment, Xs is serine (S). In another embodiment, Xg is arginine (R). In another embodiment, Xw is glutamine (Q). In another embodiment, Xn is serine (S). In another embodiment, X12 is phenylalanine (F).

[0088] In one embodiment, the antigen-binding protein of the invention comprises the CDRH1 of SEQ ID NO:28, CDRH2 of YISPGNEDXiKYSQKFQG (SEQ ID NO: 117), wherein Xi is leucine (L), glutamine (Q), or valine (V), CDRH3 of SEQ ID NQ:30, CDRL1 of SEQ ID NO:31, CDRL2 of SEQ ID NO:32 and CDRL3 of SEQ ID NO:33.

[0089] In one embodiment, the antigen-binding protein of the invention comprises one or more of CDRH1 of SEQ ID NO:28, CDRH2 of SEQ ID NO:29, CDRH3 of SEQ ID NQ:30, CDRL1 of SEQ ID NO:31, CDRL2 of SEQ ID NO:32 and CDRL3 of SEQ ID NO:33.

[0090] In another embodiment, the antigen-binding protein of the invention comprises CDRH1 of SEQ ID NO:28, CDRH2 of SEQ ID NO:29 and CDRH3 of SEQ ID NQ:30.

[0091] In a further embodiment, the antigen-binding protein of the invention comprises CDRH1 of SEQ ID NO:28, CDRH2 of SEQ ID NO:29, CDRH3 of SEQ ID NQ:30, CDRL1 of SEQ ID NO:31, CDRL2 of SEQ ID NO:32 and CDRL3 of SEQ ID NO:33.

[0092] In one embodiment, the antigen-binding protein of the invention comprises one or more of CDRH1 of SEQ ID NO:62, CDRH2 of SEQ ID NO:63, CDRH3 of SEQ ID NO:64, CDRL1 of SEQ ID NO:65, CDRL2 of SEQ ID NO:66 and CDRL3 of SEQ ID NO:67.

[0093] In another embodiment, the antigen-binding protein of the invention comprises CDRH1 of SEQ ID NO:62, CDRH2 of SEQ ID NO:63 and CDRH3 of SEQ ID NO:64.

[0094] In a further embodiment, the antigen-binding protein of the invention comprises CDRH1 of SEQ ID NO:62, CDRH2 of SEQ ID NO:63, CDRH3 of SEQ ID NO:64, CDRL1 of SEQ ID NO:65, CDRL2 of SEQ ID NO:66 and CDRL3 of SEQ ID NO:67

[0095] 10

[0096] 008828600 In one embodiment, the antigen-binding protein of the invention comprises one or more of CDRH1 of SEQ ID NO:68, CDRH2 of SEQ ID NO:69, CDRH3 of SEQ ID NQ:70, CDRL1 of SEQ ID NO:71, CDRL2 of SEQ ID NO:72 and CDRL3 of SEQ ID NO:73.

[0097] In another embodiment, the antigen-binding protein of the invention comprises CDRH1 of SEQ ID NO:68, CDRH2 of SEQ ID NO:69 and CDRH3 of SEQ ID NQ:70.

[0098] In a further embodiment, the antigen-binding protein of the invention comprises CDRH1 of SEQ ID NO:68, CDRH2 of SEQ ID NO:69, CDRH3 of SEQ ID NQ:70, CDRL1 of SEQ ID NO:71, CDRL2 of SEQ ID NO:72 and CDRL3 of SEQ I D NO:73.

[0099] In one embodiment, the antigen-binding protein of the invention comprises one or more of CDRH1 of SEQ ID NO:74, CDRH2 of SEQ ID NO:75, CDRH3 of SEQ ID NO:76, CDRL1 of SEQ ID NO:77, CDRL2 of SEQ ID NO:78 and CDRL3 of SEQ ID NO:79.

[0100] In another embodiment, the antigen-binding protein of the invention comprises CDRH1 of SEQ ID NO:74, CDRH2 of SEQ ID NO:75 and CDRH3 of SEQ ID NO:76.

[0101] In a further embodiment, the antigen-binding protein of the invention comprises CDRH1 of SEQ ID NO:74, CDRH2 of SEQ ID NO:75, CDRH3 of SEQ ID NO:76, CDRL1 of SEQ ID NO:77, CDRL2 of SEQ ID NO:78 and CDRL3 of SEQ ID NO:79.

[0102] In one embodiment, there is provided an antigen-binding protein according to the invention as herein described, wherein the antigen-binding protein comprises a VH domain comprising:

[0103] CDRH1 of GYTFX1X2HA (SEQ I D NO: 118) wherein Xi is lysine (K) or threonine (T); and X2 is alanine (A), aspartic acid (D), glutamic acid (E), tryptophan (W) or tyrosine (Y);

[0104] CDRH2 of ISPGX3X4DX5 (SEQ ID NO: 119) wherein X3 is asparagine (N) or glutamine (Q); X4is alanine (A), aspartic acid (D), glutamic acid (E), glycine (G), lysine (K), asparagine (N), glutamine (Q) or tyrosine (Y); and X5is isoleucine (I), leucine (L), glutamine (Q), threonine (T) or valine (V);

[0105] CDRH3 of KXgSYYGX? (SEQ ID NO: 120) wherein Xs is leucine (L), glutamine (Q) or arginine (R); and X7is glutamic acid (E) or histidine (H);

[0106] CDRL1 of XgNIYSN (SEQ ID NO: 121) wherein Xg is glutamic acid (E) or glutamine (Q);

[0107] CDRL2 of AAXg wherein Xg is serine (S) or threonine (T); and

[0108] CDRL3 of QHFX10GTPYT (SEQ ID NO: 114) wherein X10 is phenylalanine (F) or tryptophan (W);

[0109] 11

[0110] 008828600 and wherein the VH domain comprises a tyrosine (Y) residue at position 55 and wherein the CDR sequences and numbering are defined according to the IMGT numbering scheme.

[0111] In a further embodiment Xi is threonine (T). In another embodiment, X2 is alanine (A). In another embodiment, X3 is asparagine (N) . In another embodiment, X4is alanine (A) or glutamic acid (E); for example, in one such embodiment, X4is glutamic acid (E). In another embodiment, X5is isoleucine (I), leucine (L), glutamine (Q) or valine (V); for example, in one such embodiment, X5is valine (V). In another embodiment, Xs is glutamine (Q).

[0112] In another embodiment, X7is histidine (H). In another embodiment, Xs is glutamic acid (E). In another embodiment, Xg is serine (S). In another embodiment, X10 is phenylalanine (F).

[0113] In one embodiment, the antigen-binding protein of the invention comprises the CDRH1 of SEQ ID NO:87, CDRH2 of ISPGNEDXi (SEQ ID NO: 122), wherein Xi is leucine (L), glutamine (Q), or valine (V), CDRH3 of SEQ ID NO:92, CDRL1 of SEQ ID NO:93, CDRL2 of SEQ ID NO:94 and CDRL3 of SEQ ID NO:95.

[0114] In one embodiment, the antigen-binding protein of the invention comprises one or more of CDRH1 of SEQ ID NO:87, CDRH2 of SEQ ID NO:88, CDRH3 of SEQ ID NO:92, CDRL1 of SEQ ID NO:93, CDRL2 of SEQ ID NO:94 and CDRL3 of SEQ ID NO:95.

[0115] In another embodiment, the antigen-binding protein of the invention comprises a VH domain comprising CDRH1 of SEQ ID NO:87, CDRH2 of SEQ ID NO:88 and CDRH3 of SEQ ID NO:92, and further comprising a tyrosine (Y) residue at position 55, wherein the numbering is defined according to IMGT.

[0116] In a further embodiment, the antigen-binding protein of the invention comprises a VH domain comprising CDRH1 of SEQ ID NO:87, CDRH2 of SEQ ID NO:88 and CDRH3 of SEQ ID NO:92, and a VL domain comprising CDRL1 of SEQ ID NO:93, CDRL2 of SEQ ID NO:94 and CDRL3 of SEQ ID NO:95, wherein the VH domain comprises a tyrosine (Y) residue at position 55, and wherein the numbering is defined according to IMGT.

[0117] In one embodiment, the antigen-binding protein of the invention comprises one or more of CDRH1 of SEQ ID NO:87, CDRH2 of SEQ ID NO:89, CDRH3 of SEQ ID NO:92, CDRL1 of SEQ ID NO:93, CDRL2 of SEQ ID NO:94 and CDRL3 of SEQ ID NO:95.

[0118] In another embodiment, the antigen-binding protein of the invention comprises a VH domain comprising CDRH1 of SEQ ID NO: 87, CDRH2 of SEQ ID NO:89 and CDRH3 of SEQ ID NO:92, and further comprising a tyrosine (Y) residue at position 55, wherein the numbering is defined according to IMGT.

[0119] 12

[0120] 008828600 In a further embodiment, the antigen-binding protein of the invention comprises a VH domain comprising CDRH1 of SEQ ID NO:87, CDRH2 of SEQ ID NO:89 and CDRH3 of SEQ ID NO:92, and a VL domain comprising CDRL1 of SEQ ID NO:93, CDRL2 of SEQ ID NO:94 and CDRL3 of SEQ ID NO:95, wherein the VH domain comprises a tyrosine (Y) residue at position 55, and wherein the numbering is defined according to IMGT.

[0121] In one embodiment, the antigen-binding protein of the invention comprises one or more of CDRH1 of SEQ ID NO:87, CDRH2 of SEQ ID NQ:90, CDRH3 of SEQ ID NO:92, CDRL1 of SEQ ID NO: 93, CDRL2 of SEQ ID NO:94 and CDRL3 of SEQ ID NO:95.

[0122] In another embodiment, the antigen-binding protein of the invention comprises a VH domain comprising CDRH1 of SEQ ID NO:87, CDRH2 of SEQ ID NQ:90 and CDRH3 of SEQ ID NO:92, and further comprising a tyrosine (Y) residue at position 55, wherein the numbering is defined according to IMGT.

[0123] In a further embodiment, the antigen-binding protein of the invention comprises a VH domain comprising CDRH1 of SEQ ID NO:87, CDRH2 of SEQ ID NQ:90 and CDRH3 of SEQ ID NO:92, and a VL domain comprising CDRL1 of SEQ ID NO:93, CDRL2 of SEQ ID NO:94 and CDRL3 of SEQ ID NO:95, wherein the VH domain comprises a tyrosine (Y) residue at position 55, and wherein the numbering is defined according to IMGT.

[0124] In one embodiment, the antigen-binding protein of the invention comprises one or more of CDRH1 of SEQ ID NO:87, CDRH2 of SEQ ID NO:91, CDRH3 of SEQ ID NO:92, CDRL1 of SEQ ID NO:93, CDRL2 of SEQ ID NO:94 and CDRL3 of SEQ ID NO:95.

[0125] In another embodiment, the antigen-binding protein of the invention comprises a VH domain comprising CDRH1 of SEQ ID NO:87, CDRH2 of SEQ ID NO:91 and CDRH3 of SEQ ID NO:92, and further comprising a tyrosine (Y) residue at position 55, wherein the numbering is defined according to IMGT.

[0126] In a further embodiment, the antigen-binding protein of the invention comprises a VH domain comprising CDRH1 of SEQ ID NO:87, CDRH2 of SEQ ID NO:91 and CDRH3 of SEQ ID NO:92, and a VL domain comprising CDRL1 of SEQ ID NO:93, CDRL2 of SEQ ID NO:94 and CDRL3 of SEQ ID NO:95, wherein the VH domain comprises a tyrosine (Y) residue at position 55, and wherein the numbering is defined according to IMGT.

[0127] The skilled person would have no difficulty in determining the sequences of the CDRs from the VH and VL domain sequences of the antigen-binding proteins set out above. The CDR sequences may, for example, be

[0128] 13

[0129] 008828600 determined according to Kabat (Kabat, E.A et al. (1991). Sequences of Proteins of Immunological Interest, 5th edit., NIH Publication no. 91-3242. U.S. Department of Health and Human Services) or the international ImMunoGeneTics information system (IMGT: Lefranc, M.-P. et al. Nucleic Acids Res. 43, D413-22 (2015)).

[0130] In one embodiment, the antigen-binding protein comprises a glutamine (Q) residue at position 94 of the VH domain according to the Kabat numbering scheme. In a further embodiment the antigen-binding protein additionally comprises an alanine (A) or valine (V) residue at position 67, an isoleucine (I) or leucine (L) residue at position 69, and an alanine (A) or threonine (T) residue at position 75 of the VH domain, wherein the numbering is defined according to Kabat. In yet a further embodiment, the antigen-binding protein comprises a valine (V) residue at position 67, an isoleucine (I) residue at position 69, and a threonine (T) residue at position 75 of the VH domain, wherein the numbering is defined according to Kabat.

[0131] In one embodiment, the antigen-binding protein comprises an isoleucine (I) or a methionine (M) residue at position 39 of the VH domain according to the IMGT numbering scheme. For example, in one such embodiment, the amino acid residue at position 39 (IMGT) of the VH domain is isoleucine (I). In a further embodiment, the antigen-binding protein additionally comprises a glycine (G) or arginine (R) residue at position 49, an isoleucine (I) or methionine (M) residue at position 53, a tryptophan (W) or tyrosine (Y) residue at position 55, an alanine (A) or valine (V) residue at position 76, an isoleucine (I) or leucine (L) residue at position 78, and an alanine (A) or threonine (T) residue at position 84 of the VH domain, wherein the numbering is defined according to the IMGT numbering scheme. In yet a further embodiment, the antigenbinding protein comprises a glycine (G) residue at position 49, a methionine (M) residue at position 53, a tyrosine (Y) residue at position 55, a valine (V) residue at position 76, an isoleucine (I) residue at position 78, and a threonine (T) residue at position 84 of the VH domain, wherein the numbering is defined according to IMGT.

[0132] In one embodiment, there is provided an antigen-binding protein of the invention as herein described, wherein the antigen-binding protein comprises a variable heavy chain sequence which comprises: CDRH1 of SEQ ID NO:28, CDRH2 of SEQ ID NO:29, CDRH3 of SEQ ID NQ:30, and further comprises a glutamine (Q) residue at position 94, wherein the numbering is defined according to Kabat. In a further embodiment, there is provided an antigenbinding protein of the invention as herein described, wherein the antigen-binding protein comprises a variable heavy chain sequence which comprises: CDRH1 of SEQ ID NO:28, CDRH2 of SEQ ID NO:29 and CDRH3 of SEQ ID NQ:30, and further comprises a glutamine (Q) residue at position 94, an alanine (A) or valine (V) residue at position 67, an isoleucine (I) or leucine (L) residue at position 69, and / or an alanine (A) or threonine (T) residue at position 75, wherein the numbering is defined according to Kabat. In a further embodiment, there is provided an antigenbinding protein of the invention as herein described, wherein the antigen-binding protein comprises a variable

[0133] 14

[0134] 008828600 heavy chain sequence which comprises: CDRH1 of SEQ ID NO:28, CDRH2 of SEQ ID NO:29, CDRH3 of SEQ ID NO:30, and further comprises a glutamine (Q) residue at position 94, a valine (V) residue at position 67, an isoleucine (I) residue at position 69, and a threonine (T) residue at position 75 wherein the numbering is defined according to Kabat.

[0135] In one embodiment, there is provided an antigen-binding protein of the invention as herein described wherein the antigen-binding protein comprises a variable heavy chain sequence according to SEQ ID NO: 39 and / or a variable light chain sequence according to SEQ ID NO: 40. In a further embodiment, there is provided an antigen-binding protein of the invention as herein described wherein the antigen-binding protein comprises a variable heavy chain sequence according to SEQ ID NO: 39 and a variable light chain sequence according to SEQ ID NO: 40.

[0136] In one embodiment, there is provided an antigen-binding protein of the invention as herein described wherein the antigen-binding protein comprises a heavy chain sequence according to SEQ ID NO: 41 and / or a light chain sequence according to SEQ ID NO: 43. In a further embodiment, the antigen-binding protein comprises a heavy chain sequence according to SEQ ID NO: 41 and a light chain sequence according to SEQ ID NO: 43.

[0137] In one embodiment, the antigen-binding protein comprises an isoleucine (I) or a methionine (M) residue at position 39 of the VH domain according to the IMGT numbering scheme. In another embodiment, the amino acid residue at position 39 (IMGT) of the VH domain is isoleucine (I). In a further embodiment, the VH domain of the antigen-binding protein additionally comprises a glycine or arginine residue at position 49, an isoleucine or methionine residue at position 53, a tryptophan or tyrosine residue at position 55, an alanine or valine residue at position 76, an isoleucine or leucine residue at position 78, and an alanine or threonine residue at position 84, wherein the numbering is defined according to IMGT. For example, the antigen-binding protein comprises a glycine residue at position 49, a methionine residue at position 53, a tyrosine residue at position 55, a valine residue at position 76, an isoleucine residue at position 78, and a threonine residue at position 84 of the VH domain according to the IMGT numbering scheme.

[0138] In one embodiment the antigen-binding protein comprises a VL domain comprising CDRL1 of SEQ ID NO: 93, CDRL2 of SEQ ID NO: 94 and CDRL3 of SEQ ID NOs: 95 wherein the CDR sequences are defined according to the IMGT numbering scheme.

[0139] In one embodiment, the antigen-binding protein comprises an asparagine (N) or arginine (R) residue at position 66, an alanine (A) or glutamine (Q) residue at position 68, and an aspartic acid (D), leucine (L) or serine (S) residue at position 69 of the VL domain according to the IMGT numbering scheme. In one embodiment, the antigen-binding protein comprises an arginine (R) residue at position 66, a glutamine (Q) residue at position 15

[0140] 008828600 68, and a serine (S) residue at position 69 of the VL domain according to the IMGT numbering scheme. In a further embodiment, the VL domain of the antigen-binding protein additionally comprises an isoleucine (I), leucine (L) or valine (V) residue at position 54, and a phenylalanine (F) or tyrosine (W) residue at position 87, wherein the numbering is defined according to IMGT. For example, the antigen-binding protein may additionally comprise a leucine (L) residue at position 54, and a tyrosine (Y) residue at position 87 of the VL domain according to the IMGT numbering scheme.

[0141] "CDRs" are defined as the complementarity determining region amino acid sequences of an antigen-binding protein. These are the hypervariable regions of immunoglobulin heavy and light chains. There are three heavy chain and three light chain CDRs (or CDR regions) in the variable portion of an immunoglobulin. Thus, "CDRs" as used herein refers to all three heavy chain CDRs, all three light chain CDRs, all heavy and light chain CDRs, or at least two CDRs.

[0142] Throughout this specification, amino acid residues in variable domain sequences and variable domain regions within full-length antigen-binding sequences, e.g. within an antibody heavy chain sequence or antibody light chain sequence, are numbered primarily according to the Kabat numbering convention. Similarly, the terms "CDR", "CDRL1", "CDRL2", "CDRL3", "CDRH1", "CDRH2" and "CDRH3" used in the Examples follow the Kabat numbering convention. For further information, see Kabat et al., Sequences of Proteins of Immunological Interest, 4th Ed., U.S. Department of Health and Human Services, National Institutes of Health (1987). IMGT numbering is also provided for certain embodiments.

[0143] It will be apparent to those skilled in the art that there are alternative numbering conventions for amino acid residues in variable domain sequences and full-length antibody sequences. There are also alternative numbering conventions for CDR sequences, for example those set out in Chothia et al. (1989) Nature 342: 877-883. The structure and protein folding of the antigen-binding protein may mean that other residues are considered part of the CDR sequence and would be understood to be so by a skilled person. Other numbering conventions for CDR sequences available to a skilled person include "AbM" (University of Bath) and "contact" (University College London) methods.

[0144] Table A below represents one definition using each numbering convention for each CDR or binding unit. Kabat is used to number the variable domain amino acid sequence.

[0145] Table A

[0146] 16

[0147] 008828600

[0148] In one embodiment, the antigen-binding protein of the invention comprises a second antigen-binding site that binds to a second antigen; for example, in one embodiment, the antigen-binding protein is bispecific, trispecific or multispecific. In one such embodiment, the antigen-binding protein is bispecific. In a further embodiment, the antigen binding protein is a bispecific antibody. In one such embodiment, the bispecific antibody is an antibody molecule in mAb2™ format.

[0149] In one embodiment, the antigen-binding protein additionally binds to CD137. CD137 is also known as tumour necrosis factor receptor superfamily member 9 (TNFRSF9) or 4-1BB.

[0150] In one such embodiment, the second antigen-binding site is located in a constant domain of the heavy chain of the antigen-binding protein. In one embodiment, the constant domain may be a CL, CHI, CH2, CH3, or CH4 domain, for example the constant domain is a CHI, CH2, or CH3 domain. The antigen-binding protein may be partly, or wholly, synthetically produced. In one such embodiment, the constant domain is a CH3 domain or a CH2 domain. In one embodiment, the second antigen-binding site is located in the CH3 domain. In a further embodiment, the second antigen-binding site comprises a modified amino acid sequence in one or more of the AB, CD and EF structural loops of the constant domain. In one embodiment, the antigen-binding protein is a dimer of two (identical) polypeptide chains, each comprising a CH2 and a CH3 domain. In one such embodiment, the antigen-binding protein further comprises an immunoglobulin hinge region, or part thereof, at the N-terminus of the CH2 domain. Such a molecule is also referred to herein as an antigen-binding Fc fragment, or Fcab™.

[0151] The terms "Fcab" and "Fcab™"describe a proprietary immunoglobulin fragment, comprising a constant domain, which may be a CL, CHI, CH2, CH3, or CH4 domain, preferably a CHI, CH2, or CH3 domain, more preferably a CH2 or CH3 domain, most preferably a CH3 domain, and wherein said constant domain has been modified to create an antigen-binding site in the constant domain. The Fcab fragment may be partly, or wholly, synthetically produced. The Fcab fragment comprises a CH2 and a CH3 domain, wherein the CH2 or the CH3 domain, preferably the CH3 domain, comprises the antigen-binding site. The Fcab fragment is preferably a dimer of two polypeptide chains, 17

[0152] 008828600 which may or may not be identical, each chain comprising a CH2 domain and a CH3 domain, and at least one of the chains comprising an antigen-binding site engineered into its CH2 domain or CH3 domain, preferably its CH3 domain. In a preferred embodiment, the Fcab further comprises an immunoglobulin hinge region, or part thereof, at the N-terminus of the CH2 domain. Such a molecule is also referred to herein as an antigen-binding Fc fragment. In one embodiment the hinge region or truncated hinge region has the sequence of SEQ ID NO: 107 or 108.

[0153] Sequences which can be incorporated into a CH3 domain of the antigen-binding protein to impart binding to CD137 are disclosed in W02020011972, which is hereby incorporated by reference in its entirety.

[0154] In another embodiment, the AB structural loop comprises a sequence according to SEQ ID NO:81. In a further embodiment, the CD structural loop comprises a sequence according to SEQ ID NO:82. In a further embodiment, the EF structural loop comprises a sequence according to SEQ ID NO:83. In a further embodiment, the CH3 domain comprises a sequence according to SEQ ID NO: 84.

[0155] The CD loop sequence of the antibody molecule is preferably unmodified, i.e. wild type. The CD loop sequence therefore preferably has the sequence set forth in SEQ ID NO: 82. The CD loop sequence is preferably located at positions 43 to 78 of the CH3 domain. The wild-type nature of this sequence indicates that this sequence is not important for binding to CD137. Thus, in one embodiment, the antigen-binding protein further binds CD137 and comprises a CD137 antigen-binding site located in the CH3 domain of the antigen-binding protein, the CD137 antigen-binding site comprising SEQ ID NO: 81 in the AB structural loop and SEQ ID NO: 83 in the EF structural loop of the CH3 domain.

[0156] In one embodiment, the second antigen-binding site is provided by an Fcab for example a CD137 binding Fcab. The antigen-binding proteins of the invention such as the bispecific antibody or Mab2 require crosslinking in order to cluster and activate the CD137. Fey receptor-mediated crosslinking has the disadvantage that Fey receptors are found throughout the human body and therefore CD137 activation is not limited to a particular site and as such mutations in the CH2 domain of the Fcabs that reduce or abrogate Fey receptor binding are needed. Thus, in the absence of crosslinking through an agent other than Fey receptors, the antigen binding proteins of the invention do not exhibit CD137 agonist activity and thus are not expected to induce liver inflammation. The bispecific antibody molecules of the invention are capable of activating CD137 conditionally in the presence of STn antigen without the need for Fey receptor-mediated crosslinking as required by conventional antibody molecules. The crosslinking of the antigen-binding protein in turn leads to clustering and activation of CD137 on the T-cell surface. The agonistic activity is therefore dependent on both STn and CD137 being present. In other words, the agonistic activity is conditional on both antigens being present.

[0157] 18

[0158] 008828600 In one embodiment, the antigen-binding protein comprises a CH2 domain which has been modified to reduce or abrogate binding of the CH2 domain to Fey receptors. For example in one such embodiment, the antigen-binding protein does not bind to one or more Fey receptors. In one embodiment, the heavy chain of the antigen-binding protein comprises substitutions in the CH2 domain to abrogate the binding for example an alanine at position 234 and an alanine as position 327 (L234A / G327A) or an alanine, glycine or glutamine residue at position 297 (N297A, N297G or N297Q mutation) and an alanine or glycine residue at position 329 (P329A or P329G). In one embodiment, the antigen-binding proteins of the invention comprise substitutions at positions 234 and 235 of the IgGl CH2 domain where in the leucine residues are substituted for alanine (L234A / L235A).

[0159] In a further embodiment, the antigen-binding proteins of the invention comprise a CH2 domain according to SEQ ID NO: 85. In a further embodiment, the antigen-binding proteins of the invention comprise an Fcab sequence according to SEQ ID NO: 21 or 22 or 23. In one such embodiment, the antigen-binding proteins of the invention comprise an Fcab sequence according to SEQ ID NO:21.

[0160] In one embodiment, the antigen-binding protein comprises a heavy chain sequence according to SEQ ID NO: 49 and a light chain sequence according to SEQ ID NO: 43.

[0161] In one embodiment, the antigen-binding protein comprises a heavy chain sequence according to SEQ ID NO:96 and a light chain sequence according to SEQ ID NO:48.

[0162] In one embodiment, the antigen-binding protein comprises a heavy chain sequence according to SEQ ID NO:97 and a light chain sequence according to SEQ ID NO:48.

[0163] In one embodiment, the antigen-binding protein comprises a heavy chain sequence according to SEQ ID NO:98 and a light chain sequence according to SEQ ID NO:48.

[0164] In one embodiment, the antigen-binding protein comprises a heavy chain sequence according to SEQ ID NO:99 and a light chain sequence according to SEQ ID NO:48.

[0165] In one embodiment, the antigen-binding protein comprises a heavy chain sequence according to SEQ ID NQ:100 and a light chain sequence according to SEQ ID NO:48.

[0166] 19

[0167] 008828600 In one embodiment, the antigen-binding protein comprises a heavy chain sequence according to SEQ ID NO:101 and a light chain sequence according to SEQ ID NO:48.

[0168] In one embodiment, the antigen-binding protein comprises a heavy chain sequence according to SEQ ID NO:102 and a light chain sequence according to SEQ ID NO:48.

[0169] In one embodiment, the antigen-binding protein comprises a heavy chain sequence according to SEQ ID NO:103 and a light chain sequence according to SEQ ID NO:48.

[0170] In one embodiment, the antigen-binding protein comprises a heavy chain sequence according to SEQ ID NO:104 and a light chain sequence according to SEQ ID NO:48.

[0171] In one embodiment, the antigen-binding protein comprises a heavy chain sequence according to SEQ ID NO:105 and a light chain sequence according to SEQ ID NO:48.

[0172] In one embodiment, the antigen-binding protein comprises a heavy chain sequence according to SEQ ID NO:106 and a light chain sequence according to SEQ ID NO:48.

[0173] In one embodiment, at least one, e.g., one ortwo, of the heavy chains of the antigen-binding proteins of the present invention comprises a lysine residue as the final amino acid residue at the C-terminal end of the heavy chain sequence, otherwise known as a C-terminal lysine residue.

[0174] In one embodiment, the antigen-binding protein binds to Sialyl-Tn with a binding affinity (KD) of 1 mM or less, 100 nM or less, 10 nM or less, 2 nM or less or 1 nM or less. Alternatively, the KD may be between 1 and 5 nM; or between 1 and 2 nM. For example, in one embodiment, the binding affinity (KD) is 1.2nM.

[0175] In one embodiment, the antigen-binding protein additionally binds human CD137, for example human and cynomolgus CD137, for example dimeric human and cynomolgus CD137. In one embodiment, the antigen-binding protein binds to dimeric human CD137 with an affinity (KD) stronger than 60 nM, 50 nM, 40 nM, 30 nM, 20 nM, 10 nM, 5 nM, 4 nM, 3 nM, or 2 nM, In one embodiment, the antigen binding protein binds to dimeric CD137 with a higher affinity than monomeric CD137. In one such embodiment, the antigen binding protein binds to dimeric CD137 with an affinity which is at least 50-fold, 60-fold, 70-fold, 80-fold, 90-fold, 100-fold, 110-fold, 120-fold, 130- fold, 140-fold, 150-fold, 160-fold, 170-fold or 200-fold higher than the affinity of the antigen-binding protein for monomeric CD137.

[0176] 20

[0177] 008828600 Affinity as referred to herein may refer to the strength of the binding interaction between an antibody molecule and its cognate antigen as measured by KD. As would be readily apparent to the skilled person, where the antibody molecule is capable of forming multiple binding interactions with an antigen (e.g. where the antibody molecule is capable of binding the antigen bivalently and, optionally, the antigen is dimeric) the affinity, as measured by KD, may also be influenced by avidity, whereby avidity refers to the overall strength of an antibody-antigen complex.

[0178] It is important to note when considering affinity measurements for TNFR agonists, such as the CD137 Fcab described here, that a higher affinity does not always translate into a more desirable biological outcome. It has been shown that low, rather than high, affinity delivered greater activity by driving increased clustering of receptors. Furthermore, whilst intrinsically strong agonistic anti-CD137 antibodies can activate CD137 in the absence of exogenous crosslinking by Fey receptors, balancing agonist activity with the strength of crosslinking interaction may provide optimal function.

[0179] The ability of an antigen-binding protein to activate T cells may be measured using a T-cell activation assay. T cells release IL-2 on activation. A T-cell activation assay may therefore measure IL-2 release to determine the level of T- cell activation induced by the antigen-binding protein. For example, the ability of the antigen-binding protein to activate T cells may be determined by measuring the concentration of the antigen-binding protein required to achieve half-maximal release of IL-2 by the T cells in a T-cell activation assay when the antigen-binding protein is crosslinked. This is referred to as the EC50. A lower EC50 indicates that a lower concentration of the antibody molecule or specific binding member is needed to achieve half-maximal release of IL-2 by the T cells in the T-cell activation assay, and thus that the antigen-binding protein has a higher T-cell activation activity. The specific antigen-binding protein may be crosslinked using an anti-CH2 antibody, for example.

[0180] In one embodiment, the antigen-binding protein binds to both STn and CD137 and has an EC50 in a T-cell activation assay of 5 nM or less, 4 nM or less, 3 nM or less, 2 nM or less, 1 nM or less, or 0.5 nM or less.

[0181] In one such embodiment, the antigen binding protein has a heavy chain sequence according to SEQ ID NO:49 and a lights chain sequence according to SEQ ID NO:50 and an EC50 of 2nM or less for example InM or less for example an EC50 of 0.6nM.

[0182] In addition, or alternatively, the ability of an antigen-binding protein to activate T cells may be determined by measuring the maximum concentration of IL-2 released by the T cells in a T-cell activation assay in the presence of the antigen-binding protein when it is crosslinked.

[0183] Multispecific, e.g., bispecific, antigen-binding proteins of the invention comprise a Fab region specific for STn and a CD137 Fcab region which preferably bind to their respective targets bivalently. This is advantageous, as the

[0184] 21

[0185] 008828600 bivalent binding of both targets makes the bridging between the T cell expressing CD137 and the TAG antigen for example Sialyl-Tn, more stable and thereby extends the time during which the T cell is localised at a particular site, such as a tumour microenvironment, and can act on the tumour. This is different to the vast majority of conventional bispecific antibody formats which are heterodimeric and bind each target antigen monovalently via one Fab arm. Such a monovalent interaction is expected to be not only less stable but in many cases is insufficient to induce clustering of TNF receptors such as CD137 in the first place.

[0186] In one embodiment, of the invention the antigen-binding sites for Sialyl-Tn and CD137 are both contained within the antigen-binding protein structure itself. In particular, the antigen-binding protein does not require other proteins to be fused to it via linkers or other means to result in a molecule that binds bivalently to both of its targets. This has a number of advantages. Specifically, the antigen-binding proteins can be produced using methods similar to those employed for the production of standard antibodies, as they do not comprise any additional fused portions. The structure is also expected to result in improved stability, as linkers may degrade over time, resulting in a heterogeneous population of molecules. Cleavage / degradation of the linker could take place prior to administration or after administration of the therapeutic to the patient (e.g. through enzymatic cleavage or the in vivo pH of the patient), thereby resulting in a reduction of its effectiveness whilst circulating in the patient. As there are no linkers in the antigen-binding proteins of the invention, they are expected to retain the same number of binding sites both before and after administration. Additionally, the rigid positioning and / or close proximity of the CD137 antigenbinding sites, which results from the rigid structure of the Fcab molecules of the invention, is advantageous for inducing CD137 clustering as compared with molecules where the CD137 binding site is not integral to the antibody structure but provided e.g. by binding moieties attached to e.g. an antibody molecule, or part thereof, via flexible linkers. Furthermore, the structure of the antigen-binding proteins is also preferred from the perspective of immunogenicity of the molecules, as the introduction of fused proteins or linkers or both may induce immunogenicity when the molecules are administered to a patient, resulting in reduced effectiveness of the therapeutic.

[0187] In one embodiment of the invention as herein described, there is provided a nucleic acid molecule or set of nucleic acid molecules encoding the antigen-binding protein of the invention. In another embodiment, there is provided a vector or set of vectors comprising a nucleic acid or set of nucleic acid molecules encoding the antigen-binding protein of the invention as herein described. In one embodiment, there is provided a recombinant host cell comprising a nucleic acid molecule or set of nucleic acid molecules or a vector or set of vectors encoding the antigen-binding protein of the invention as herein described.

[0188] In one embodiment, there is provided a method of producing the antigen-binding protein of the invention as herein described comprising culturing a recombinant host cell under conditions for production of the antigen-binding

[0189] 22

[0190] 008828600 protein. In a further embodiment, there is provided a method of further isolating and / or purifying the antigenbinding protein of the invention as herein described.

[0191] In one embodiment, there is provided an antigen-binding protein of the invention as herein described for use as a medicament. In one embodiment, there is provided an antigen-binding protein of the invention as herein described for use in a method of treating cancer in an individual. The individual may be a patient, preferably a human patient.

[0192] Treatment may be any treatment or therapy in which some desired therapeutic effect is achieved, for example, the inhibition or delay of the progress of the condition, and includes a reduction in the rate of progress, a halt in the rate of progress, amelioration of the condition, cure or remission (whether partial or total) of the condition, preventing, ameliorating, delaying, abating or arresting one or more symptoms and / or signs of the condition or prolonging survival of an individual or patient beyond that expected in the absence of treatment.

[0193] Treatment as a prophylactic measure (i.e. prophylaxis) is also included. For example, an individual susceptible to or at risk of the occurrence or re-occurrence of a disease such as cancer may be treated as described herein. Such treatment may prevent or delay the occurrence or re-occurrence of the disease in the individual.

[0194] In one embodiment, there is provided a method of treating cancer in an individual, wherein the method comprises administering to the individual a therapeutically effective amount of the antigen-binding protein of the invention as herein described.

[0195] In one embodiment, there is provided a pharmaceutical composition comprising the antigen-binding protein of the invention as herein described and a pharmaceutically acceptable excipient.

[0196] In one embodiment, there is provided a pharmaceutical composition for use in the treatment of cancer in an individual, comprising the antigen-binding protein of the invention as herein described.

[0197] In one embodiment, there is provided the use of an antigen-binding protein of the invention as herein described for the manufacture of a medicament for treating cancer.

[0198] The skilled person will appreciate that, upon production of an antigen-binding protein, such as an antibody in a host cell, post-translational modifications may occur. For example, this may include the cleavage of certain leader sequences, the addition of various sugar moieties in various glycosylation patterns, non-enzymatic glycation, deamidation, oxidation, disulphide bond scrambling and other cysteine variants such as free sulfhydryls, racemised disulphides, thioethers and trisulphide bonds, isomerisation, C-terminal lysine clipping, and N-terminal glutamine cyclisation. The present invention encompasses the use of antigen-binding proteins that have been subjected to, or have undergone, one or more post-translational modifications. Thus an "antigen-binding protein"

[0199] 23

[0200] 008828600 or "antibody" of the invention includes an "antigen-binding protein" or "antibody", respectively, as defined earlier that has undergone a post-translational modification such as described herein.

[0201] Glycation is a post-translational non-enzymatic chemical reaction between a reducing sugar, such as glucose, and a free amine group in the protein, and is typically observed at the epsilon amine of lysine side chains or at the N- terminus of the protein. Glycation can occur during production and storage only in the presence of reducing sugars.

[0202] Deamidation can occur during production and storage, is an enzymatic reaction primarily converting asparagine (N ) to iso-aspartic acid (iso-aspartate) and aspartic acid (aspartate) (D) at approximately 3:1 ratio. This deamidation reaction is therefore related to isomerization of aspartate (D) to iso-aspartate. The deamidation of asparagine and the isomerisation of aspartate, both involve the intermediate succinimide. To a much lesser degree, deamidation can occur with glutamine residues in a similar manner. Deamidation can occur in a CDR, in a Fab (non-CDR region), or in the Fc region.

[0203] Oxidation can occur during production and storage (i.e., in the presence of oxidising conditions) and results in a covalent modification of a protein, induced either directly by reactive oxygen species or indirectly by reaction with secondary by-products of oxidative stress. Oxidation happens primarily with methionine residues, but may occur at tryptophan and free cysteine residues. Oxidation can occur in a CDR, in a Fab (non-CDR) region, or in the Fc region.

[0204] Disulphide bond scrambling can occur during production and basic storage conditions. Under certain circumstances, disulphide bonds can break or form incorrectly, resulting in unpaired cysteine residues (-SH). These free (unpaired) sulfhydryls (-SH) can promote shuffling.

[0205] The formation of a thioether and racemisation of a disulphide bond can occur under basic conditions, in production or storage, through a beta elimination of disulphide bridges back to cysteine residues via a dehydroalanine and persulphide intermediate. Subsequent crosslinking of dehydroalanine and cysteine results in the formation of a thioether bond or the free cysteine residues can reform a disulphide bond with a mixture of D- and L-cysteine.

[0206] Trisulphides result from insertion of a sulphur atom into a disulphide bond (Cys-S-S-S-Cys) and are formed due to the presence of hydrogen sulphide in production cell culture.

[0207] N-terminal glutamine (Q) and glutamate (glutamic acid) (E) residues in the heavy chain and / or light chain is likely to form pyroglutamate (pGlu) via cyclisation. Most pGlu formation happens in the production bioreactor, but it can be formed non-enzymatically, depending on pH and temperature of processing and storage conditions.

[0208] Cyclisation of N-terminal Q or E is commonly observed in natural human antibodies. C-terminal lysine clipping is an enzymatic reaction catalysed by carboxypeptidases, and is commonly observed in recombinant and natural

[0209] 24

[0210] 008828600 human antibodies. Variants of this process include removal of lysine from one or both heavy chains due to cellular enzymes from the recombinant host cell.

[0211] EXAMPLES le 1 - - Immunisation of mice and rats with STn-KLH and BSMse

[0212] In order to generate antibodies that bind to STn, rats and mice were immunised with suitably glycosylated proteins. Bovine submaxillary mucin (BSM) is naturally highly glycosylated with multiple glycans, including STn, and therefore represents native STn production. Synthetic STn-conjugated keyhole limpet haemocyanin (STn- KLH) has previously been developed as a cancer vaccine (Theratope™) and has been shown to generate an immune response. BSM and STn-KLH were therefore considered suitable immunogens for the generation of anti-STn antibodies. BSM, STn-KLH or a mix of the two were injected intraperitoneally into six rats and nine mice, and into the footpad of a further six rats. Three immunisations were performed followed by serum sampling at day 40. ELISAs were performed against the two immunogens, and additionally against STn-conjugated bovine serum albumin (STn-BSA), to confirm the specificity of binding to STn. STn-BSA is a synthetic construct of the glycan STn conjugated to bovine serum albumin (BSA) and was included in the ELISA-based screening to confirm carrier-independent binding to STn.

[0213] Serum from animals immunised with STn-KLH recognised STn-KLH and STn-BSA but not BSM, whereas serum from animals immunised with BSM recognised BSM and STn-BSA but not STn-KLH. Only serum from mice injected with the immunogen mix recognised all antigens tested. The mouse that gave the most positive results in the ELISA screen was injected with a further antigen boost three days before fusion of B cells, obtained from the immunised mouse, with myeloma cells for screening of antigen-specific antibody-secreting hybridoma cells.

[0214] Following hybridoma production, an ELISA screen against STn-BSA and BSM identified nine clones that recognised both STn-BSA and BSM. Once expanded,, these clones were also confirmed to bind to STn-KLH. Limited dilutions confirmed that seven of the nine investigated clones contained subclones specific for both STn- BSA and BSM. Supernatants from these seven subclones were tested for binding to HEK cells with high levels of surface STn (HEK.ST6GalNAcl.l cells).

[0215] HEK.STBGalNAcl.l cells were produced by inserting the human STBGalNAcl.l gene into a pcDNA5FRT vector and transformed into Flp-ln T-REx 293 cells using Lipofectamine 2000. The overexpression of STBGalNAcl.l is reported to result in high surface levels of STn. The resulting cell line was named HEK.STBGalNAcl.l. Production of STn on these cells was confirmed by flow cytometry. An empty vector transfection was performed as a negative control, resulting in the cell line HEK.EV.

[0216] 25

[0217] 008828600 Six of the subclones showed significant binding to the HEK.ST6GalNAcl.l cells but no binding to STn-negative HEK293 cells (unmodified Flp-ln T-Rex 293 cells). These six subclones were sequenced and found to have unique VH and VL sequences. The VH and VL domain pairs of each of the six subclones were subcloned into a pTT5 expression vector to generate six STn-specific antibody molecules of human IgGl format, each having a heavy chain composed of a mouse VH domain and human IgGl CHI, CH2 and CH3 domains and hinge region, and a light chain composed of a mouse VL domain and a human IgGl kappa (K) constant domain. These antibody sequences were expressed in HEK cells and binding of the resulting antibodies to HEK.ST6GalNAcl.l cells was confirmed.

[0218] Three of the six IgGl antibody clones (CL4, CL5 and CL9) showed high binding affinity to HEK.ST6GalNAcl.l cells. No binding to the STn-negative HEK293 cells could be observed. These three clones exhibited a high similarity in sequence. Clone CL4 was found to bind with high specificity to STn when tested for binding to 94 common human O-glycans. Clone CL5 bound weakly to disialyl-T antigen in addition to binding strongly to STn. Clone CL9 bound weakly to sialyl-T antigen and disialyl-T antigen in addition to binding strongly to STn. All three clones bound to STn in both its serine- and threonine-linked forms. No binding of these clones to further glycans could be detected.

[0219] Table 1: EC5o of HEK.STGGalNAcl.l cell line binding for clones derived from mice immunisations.

[0220] An alanine scan was performed across the variable heavy CDRs and Vernier regions of the CL4 antibody to assess mutational robustness and aid in humanisation attempts. Amino acid residues at positions in these regions were individually mutated to an alanine residue and binding to STn was assessed by measuring binding to HEK.ST6GalNAcl.l cells.

[0221] All mutations and variants are numbered according to the consecutive numbering of the parent sequence.

[0222] Residues T31, Y32, W33, M34, H35, E50, 151, T53, S54, N55, G56, N59, Y60, N61, F64, K65, S99, D100, Y101, G102, Y103, R105, F107, Y109 of the heavy chain variable domain of clone CL4 were identified as being involved

[0223] 008828600 in binding. Unexpectedly, the D104A mutation led to a 2-fold increase in response compared to the parent G1 / CL4 antibody. Results of the alanine scan are shown in Table 2.

[0224] Table 2: Binding to HEK.STGGalNAcl.l cells by alanine scan variants of CL4 tested at 1 nM. MFI: Median fluorescent intensity. Fold reduction: Fold reduction in MFI compared to parental sequence CL4.

[0225] Attempts were made to humanise the CL4 antibody using the information from the alanine scan. Three humanised variable heavy chain variants (CL4H1-3) and four humanised variable light chain variants (CL4L1-4)

[0226] 27

[0227] 008828600 were designed based on the IGHV1-2*O2:IGHJ1*O1 and IGKV3-ll*01:IGKJ2*01 germline sequences respectively with increasing similarity to the human sequences.

[0228] The parental VL sequence and the CL4L1 and CL4L2 humanised sequences were paired with the parental VH and the CL4H1-CL4H3 sequences and expressed in HEK cells.

[0229] The CL4L1 and CL4L2 combinations with the parental VH and the CL4Hlsequences did not express at all and so were not progressed any further.

[0230] CL4H2 and CL4H3 were expressed in combination with the parental VL and CL4L3 and CL4L4.

[0231] Although two variant combinations (parental VH / CL4L3 and parental VH / CL4L4) exhibited some minimal binding, all other variants tested showed a loss of binding to HEK.ST6GalNAcl.l cells so from this data it was concluded that the loss of binding stemmed from mutations in both the VH and VL variants.

[0232] Backmutations to the parental sequence (A61N, Q62E, Q65K, G66S, R67K, V68A and I70L) were introduced as single mutations or in combination with the S55T mutation to CL4H2 and expressed as antibodies with CL4L3. These mutations were unable to restore binding to HEK.STGGalNAcl.l cells. The CL4H2 region containing V37I, M48I, G49A, S55T, V68A, I70L, S75A, T76S, S77N, R85S, R87T and D89E was named CL4H5 and further mutations were introduced as single or double mutations in an attempt to restore binding (T55H, I48M, A49G, A68V, L69I, S85R, T87R, E89D, A75S, S76T, S76I, N77S, N77A). T55H, A68V and S76T partially restored binding, however binding as measured at a concentration of 2.5 nM was still approximately 2-fold lower than that seen with the parental CL4 sequence. The addition of these further mutations also increased the predicted immunogenicity of the antibody.

[0233] Table 3: Predicted MHC-II immunogenicity scores for humanised and backmutated variants of clone CL4. Scores are based on the number of predicted MHC-II binding peptides for the set of alleles.

[0234] 28

[0235] 008828600 The CL4 variants comprised several high-risk developability motifs within the heavy chain CDR sequences. One tryptophan oxidation motif (W33), two asparagine deamidation motifs (N52T53 and N55G56) and one N-linked glycosylation motif (N52T53S54N55) were identified to be of high risk in terms of developability and required mitigation. The tryptophan oxidation motif (W33) could not be replaced with other residues without loss of function. This result was in line with a previous alanine scan where the respective position was identified to be essential for binding activity. The N-linked glycosylation motif and two asparagine deamidation motifs, as identified in the HCDR2 could be addressed by mutating N52 and N55 during the humanisation process, removing these liabilities but not without loss of function and reduced binding activity. Given the presence of the potential tryptophan oxidation site in the HCDR1 and the reduced binding seen with humanised variants of CL4, development of this lineage of anti-STn Fabs was terminated, thus resulting in an unexpected failure of this approach.

[0236] Example 2 - Antibody generation - selection from naive Fab libraries against natural and synthetic STn antigens

[0237] Due to the difficulty in producing suitable antibodies from immunisation, an alternative approach using naive library selections was attempted. Selections from two naive Fab-phage libraries against BSM were performed, in which each library was subjected individually to a selection campaign. A total of four rounds of selections were performed against BSM directly immobilised onto MaxiSorp plates. No variants binding to STn could be identified in the following phage ELISA screen where binding was tested towards STn-BSA.

[0238] Selections were also performed against a further antigen, human mucin-1 (hIX / IUC-l), which is a human glycoprotein reported to display TAG. However, only a limited number of variants that bound to STn-BSA were identified by phage ELISA, and these variants were found to be of-low stability when tested in full IgGl format.

[0239] Selections were then performed against a further antigen, STn conjugated to BSA (STn-BSA), which would ensure STn to be the sole glycan present on the antigen. This selection strategy aimed for increased identification of variants binding to STn. Each library was subjected individually to a selection campaign. A total of three rounds of selections were performed against STn-BSA directly immobilised onto MaxiSorp plates, followed by a single round of selection against either STn-BSA or sole STn without carrier protein immobilised onto MaxiSorp plates. Several variants that bound to STn-BSA were identified in the subsequent ELISA screening where binding was tested towards STn-BSA. VH and VL domains of these variants were subcloned into a pTT5 expression vector, to generate bispecific (mAb2format) antibody molecules of human IgGl format that consisted of the anti-human CD137 Fcab (CD137-AA), containing a human CD137 binding site engineered into its CH3 domain, a LALA mutation in the CH2 domain and the VH and VL domains of TAG binding variants However, only a limited number of variants exhibited monomeric stability when tested in this antibody format. When tested in an in

[0240] 29

[0241] 008828600 vitro CD8+ T-cell activation assay for the ability to agonise CD137 in a TAG dependent manner, none of these variants showed TAG-dependent agonism indicating a lack of sufficient glycan specificity.

[0242] Selections were then performed against HEK.ST6GalNAcl.l cells with counter-selection against STn-negative HEK293 cells, which would allow for binding of natively expressed STn whilst counter-selecting against a large diversity of glycans and glycoproteins on STn-negative cell surface. No variants binding to STn could be identified in the subsequent ELISA screening where binding was tested towards STn-BSA.

[0243] In summary, variants obtained by selection from naive Fab libraries against natural and synthetic STn antigens exhibited overall poor biophysical properties and lacked sufficient glycan specificity. A variant, suitable for further sequence optimisation and development, could not be identified.

[0244] Example 3 - Antibody generation - Humanisation

[0245] A number of mouse antibodies which bound to TAG antigens were then analysed to determine which bound with suitably high affinity and specificity for further development.

[0246] Table 4: Summary of binding to serine and threonine-linked STn for tested mouse antibodies at 10 pg / mL.

[0247] Mouse antibody mAbm4 was the strongest and most specific binder and was selected for humanisation.

[0248] To identify residues of mAbm4 essential for binding to TAG, an alanine scan was performed across all six CDRs and neighbouring flanking regions.

[0249] For the VH domain, residue H32 in CDRH1, S52, N55 and D57 in CDRH2 and S99, Y100, and Y101 in CDRH3 were identified to be essential for maintained binding to TAG.

[0250] 008828600 Mutating residues G54, 158 (in CDRH2) and K97 (CDRH3 by IMGT definition or the Vernier region by Kabat definition) resulted in reduced binding. Within the VL domain, residues L54 (CDRL2 by Kabat or Vernier region by IMGT definition) F91, and W92 (in CDRL2) appeared to be essential for maintained binding to TAG. Mutating position P95 resulted in reduced binding.

[0251] Humanisation was accomplished through the grafting of the parental mAbm4 antibody CDRs sequences into selected human acceptor framework regions. The preferred acceptor framework regions were selected from the set of human germline sequences based on sequence similarity and structural considerations. For the variable light chain (VL) domain of mAbm4Athe human germline sequence IGKV1-NL1*O1 / IGKV1-39*O1 was selected as acceptor framework region. For the variable heavy chain domain (VH) of mAbm4xthe human germline sequence IGHV1-3*O1 was selected as acceptor framework region.

[0252] Each humanised VH domain was paired with each humanised VL domain, resulting in 16 humanised variants. All variants expressed well and, except for the combinations of the humanised variant VH1 with VL3 and VL4, exhibited favourable biophysical properties, as assessed by size exclusion chromatography (SEC) and hydrophobic interaction chromatography (HIC).

[0253] Expressed variants were further tested for maintained binding activity towards STn. Considerably reduced binding was observed for all variants against synthetic STn-BSA (Figure 1). This observation was confirmed in cell- based assays where binding to STn on HEK.ST6GalNAcl.l cells (Figure 2) was lost. As a result, backmutations were introduced in an attempt to restore binding activity.

[0254] A total of seven selected residues M34I, V37A, W50Y, T58I S61N, Q62E and Q65K were chosen and individually back mutated to the non-humanised parent sequence mAbm4 and tested for restored binding. Individual backmutations were introduced and tested in the context of the least stringently humanised variant VH1 (SEQ ID NO:13) paired with the least stringently humanised light chain variable domain variant VL1 (SEQ ID NO:17). All resulting variants expressed well. However, introduction of backmutation V37A, Q62E or Q65K resulted in reduced monomeric stability, as assessed by SEC. Variants were tested for binding to HEK.ST6GalNAcl.l cells. As seen in Figure 3, binding activity could be partially restored by back mutating the tryptophan at position 50 to tyrosine however, binding still remained weaker compared to the parent murine antibody mAbm4 with an approximately 3-fold increase in EC5o and 3.5-fold lower Emax (Table 5). All other tested backmutations alone were not sufficient to restore binding to TAG-producing cells at concentrations of up to 300 nM (Figures 3 and 4).

[0255] 31

[0256] 008828600 Table 5: EC50 and Emax values obtained for binding to HEK.ST6GalNAcl.l cells by back mutated variant VH1- W50Y. Potency can be measured by EC50, the amount of drug required to produce 50% maximal effect. Efficacy (Emax) is the maximum effect which can be expected from this drug (i.e. when this magnitude of effect is reached, increasing the dose will not produce a greater magnitude of effect).

[0257] As the backmutation W50Y was the only backmutation tested that was able to restore binding activity towards HEK.STGGalNAcl.l cells, albeit only partially, achieving 28% Emax relative to the mouse parental antibody, it was therefore chosen to be tested in the remaining variants and in combination with additional mutations.

[0258] The W50Y backmutation was able to restore binding to HEK.STGGalNAcl.l cells when paired with all four humanised VL variants (Figure 5). While pairing with humanised VL variants VL1 (SEQ ID NO:17) and VL2 (SEQ ID NO:18) exhibited similar binding, pairings with variants VL3 (SEQ ID NO:19) and VL4 (SEQ ID NQ:20) resulted in considerably reduced binding with an approximately 10-fold increase in EC50 (Table 6). Strongest binding signals for the backmutation W50Y were observed for the humanised VH variant VH4 (SEQ ID NO:16) paired with VL1 (SEQ ID NO:17).

[0259] Table 6: Binding of humanised variants of parent mAbm4 antibody, each containing W50Y backmutation in heavy chain variable domain VH1 or VH4, to HEK.STGGalNAcl.l cells

[0260] VH4 (SEQ ID NO:16) with additional backmutation W50Y was furthertested in combination with backmutation T58I. The resulting variant was paired with the light chain variants VL1 (SEQ ID NO:17) and VL2 (SEQ ID NO:18) and tested for binding to HEK.STGGalNACl.l cells. This combination further increased binding affinity to STn, with EC50 and Emax values comparable to the mouse parental antibody (Figure 6 and Table 7). No difference in binding activity could be observed between pairing with VL1 or VL2.

[0261] 32

[0262] 008828600 Table 7: EC50 and Emaxfor binding to STn-expressing cells by VH4 with W50Y and T58I paired with VL1 and VL2.

[0263] Example 4 - Generation of affinity matured sequences

[0264] Libraries based on the mouse parental sequence containing one, two orthree random mutations in the CDR regions of VH or VL domain of the mouse parent sequence mAbm4 were created using NNK primers. Each library was subjected individually to a selection campaign, comprising a counter-selection against STn-negative cells for maintained specificity, followed by off-rate selection using STn-BSA. Following the off-rate selection, surviving variants were subjected to a second round of selection for the recovery of specific binders. This round consisted of a counter-selection against STn-negative cells, followed by a selection against HEK.STGGalNAcl.l cells under reduced stringency. Selected clones were then screened for binding to recombinant antigen and sequences were analysed.

[0265] Affinity maturation by off-rate selection, followed by a recovery round under low stringency, yielded successful enrichment of phage variants over background. A total of 285 randomly picked clones were subsequently screened by ELISA for binding to STn-BSA and BSM. Several variants with signals for binding to both antigens above those obtained from background and the parental clone mAbm4 were identified. Notably, stronger signals for binding to BSM were observed. Sanger sequencing of clones with improved binding yielded six positional mutations. These mutations were selected for further characterisation by firstly introducing them into the complete IgGl sequence of antibody Gl-AA / mAbm4 (VH / VL of mAbm4 on an IgGl human backbone comprising LALA mutations). VH mutations were made at T30K, R98L, R98Q, H103E and VL mutations at N53R and W92F.

[0266] Mutations identified in the affinity maturation selections were cloned into the Gl-AA / mAbm4 antibody and expressed in HEK293 cells. All variants yielded high expression and exhibited favourable biophysical properties, as assessed by SEC and HIC.

[0267] All variants bound to HEK.STBGalNAcl.l cell with high affinity and specificity. Mutations at positions R98L, R98Q, and H103E, as well as N53R and W92F, resulted in improved Emaxvalues compared to the parent mAbm4 antibody when tested for binding to HEK.STBGalNAcl.l cells (Table 8). Mutation T30K provided no improvement in binding compared to the parent antibody, and negative epistatic effects when combined with other affinity maturation mutations.

[0268] 33

[0269] 008828600 None of the variant antibodies were observed to bind to STn-negative HEK293 cells.

[0270] Table 8: Binding affinity of mAbm4 variants to HEK.ST6GalNACl.l cells.

[0271] Mutations from Table 8 with positive effects on affinity were then combined with the humanised variants (including those which contained additional backmutations) to determine if they improved binding on the humanised variants. All the tested variants expressed well and exhibited favourable biophysical properties as assessed by SEC and HIC. Introduction of single heavy chain affinity maturation mutation R98Q or R98L to the humanised and back-mutated variant VH1(W5OY) improved binding to HEK.ST6GalNAcl.l cells to levels comparable to those observed for the parent mAbm4 (Figure 7), in terms of both EC5o and Emax. Binding affinities were further increased by introducing the respective mutations, W50Y + R98Q, and / or W50 + R98L, into the humanised variant VH4, with Emaxvalues surpassing those observed for the parent mAbm4. The humanised variant VH4, containing the backmutation W50Y and affinity maturation mutation R98Q or R98L, when paired with the humanised light chain variant VL1, exhibited the strongest binding and was therefore considered for further sequence optimisation. See Table 9.

[0272] Table 9: Binding to HEK.STBGalNAcl.l cells by humanised variants containing backmutation W50Y and affinity maturation mutations R98Q or R98L.

[0273] 34

[0274] 008828600

[0275] Example 5 - Sequence Optimisation

[0276] Sequences were investigated for presence of liability motifs that can cause issues with antibody developability. Severity of developability risk was determined based on the nature of the liability, location within antigenbinding loops or framework, and the representation of germline versus non-germline residues.

[0277] Several high-risk motifs were identified within the sequence of the mouse parental mAbm4. One exposed aspartate isomerisation motif (D31H32) was found in the heavy chain CDR1 (CDRH1). A further Aspartate isomerisation motif (D56D57) and one aspartate deamidation motif (N55D56) were found within the CDRH2. In the light chain sequence of mAbm4, one aspartate isomerisation motif (D56G57) was found within CDR2 and one tryptophan oxidation motif (W92) within CDR3.

[0278] Humanisation of mAbm4 had also resulted in the introduction of two further high-risk motifs - an exposed methionine oxidation motif (M34) and an aspartate deamidation motif (D57T58), both on the VH domain.

[0279] As all of these identified motifs were deemed to be high risk in terms of developability, their removal or mitigation was considered necessary.

[0280] The two motifs introduced through the humanisation process (M34 and D57T58) were corrected by backmutations to the mouse sequence, M34I and T58I.

[0281] Two high-risk motifs, identified in the VL domain of mAbm4, could be removed by humanisation and affinity maturation. The aspartate isomerisation motif (D56G57), as present in CDRL2, was successfully removed in humanised VL variant VL1 by mutation D56S. The tryptophan oxidation motif (W92), as present in CDRL3, could be removed by introduction of the affinity maturation mutation W92F.

[0282] To address the two aspartate isomerisation motifs, D31H32 present in CDRH1, and D56D57 present in CDRH2, site saturation mutagenesis of the respective positions in the VH was performed. Mutations that introduced new high-risk motifs were excluded from characterisation. The remaining variants were assessed for maintained binding to TAG.

[0283] Mutations at position D31 were well accepted, with mutation D31E, D31W, and D31Y maintaining binding to on HEK.ST6GalNAcl.l cells compared to the parent mAbm4 antibody (Figure 8). Further, in previous alanine scanning experiments mutation D31A maintained binding compared to the parent mAbm4 antibody. Based on

[0284] 35

[0285] 008828600 binding and biophysical properties D31A was selected as the mitigating mutation in all subsequent clones. In contrast to position D31, position H32 did not tolerate any of the investigated mutations and was therefore not changed in subsequent variants. Position D56 in CDRH2, tolerated several mutations, with D56E, D56G, D56K, D56N, D56Q, and D56Y maintaining binding compared to the parent mAbm4 along with D56A tested in a previous alanine scan experiment. All investigated mutations at position D57 exhibited significant losses in binding and was not changed in subsequent variants.

[0286] All high-risk motifs, other than those introduced by humanisation, could be mitigated by the mutations D31A and D56A or D56E. Mutations (D31A and D56E) were selected for introduction into humanised variant VH4 with backmutation W50Y and mutations identified by affinity maturation.

[0287] Example 6 - Generation and screening of CD137 / TAG mAb2variants

[0288] VH and VL domains of the variants were subcloned into a pTT5 expression vector, to generate bispecific (mAb2format) CD137 / TAG antibody molecules of human IgGl format that comprised the anti-human CD137 Fcab (CD137-AA) (SEQ ID NO:21), containing a human CD137 binding site engineered into its CH3 domain, the LALA mutation in its CH2 domain, and the VH and VL domains of the anti-TAG variant clones. The sequences were characterised for favourable biophysical properties, binding to HEK.ST6GalNAcl.l cells, and conditional activation of CD8+T cells in the presence of HEK.ST6GalNAcl.l.

[0289] Sequence-optimised and humanised variants of mAbm4 in bispecific (mAb2) format expressed well and exhibited favourable biophysical properties, as assessed by SEC and HIC. The humanised heavy chain variant VH4 in bispecific format, when containing either the R98L or R98Q affinity-maturation mutation in addition to the back-mutation W50Y, exhibited similar levels of binding to STn on HEK.STGGalNAcl.l, as compared to the parental mouse monospecific anti-TAG mAb in human IgGl format (Figure 9). Therefore, it can be concluded that the ability of these humanised, bispecific (mAb2format) antibodies to bind to TAG-producing cells was not compromised by the additional CD137 binding sites present in their CH3 domains.

[0290] Example 7 - Testing Humanised and Sequence-Optimised Variants in CD8+T-cell Activation Assays

[0291] The humanised bispecific (mAb2format) variant antibodies (VH4 / VL1) containing all the various backmutations were assessed for conditional activation of CD8+T cells, as assessed by release of human IL-2 (hlL-2) in the presence of TAG-producing cells and plate-bound anti-CD3E antibody (as described in Example 12). The respective variants exhibited potent activation of CD8+T cells in the presence of HEK.ST6GalNAcl.lcells with EC5o values consistently in the sub-nanomolar range.

[0292] 36

[0293] 008828600 Introduction of affinity maturation mutation (W92F) into the humanised VL domain (VL1) and introduction of the sequence-optimisation mutations (D31A and D56E) into the humanised VH domain (VH4), in combination with backmutation W50Y and affinity maturation mutation R98Q, yielded a bispecific (mAb2format) antibody with favourable biophysical properties. This variant, termed CD137-AA / TAG1, mediated similar CD8+T-cell activation levels in the presence of HEK.ST6GalNAcl.l cells, as compared to the parent mAbm4 in mAb2form, as assessed by CD8+T-cell activation assays (Figure 10).

[0294] No activation of T cells was observed for any of the antibodies tested in the presence of STn-negative HEK293 cells.

[0295] Example 8 - Deimmunisation of Variants

[0296] The immunogenicity risk of the humanised variants was assessed, following which four deimmunising mutations were introduced into the VL domain. Mutations, V48I, T52S, and A55Q represented mutations towards the human germline acceptor sequence IGKV1-NL1*O1 / IGKV1-39*O1 and were predicted to disrupt the binding pattern towards MHC-II. The fourth mutation, N53R, as initially identified by affinity maturation, was predicted to further reduce immunogenicity by disrupting identified T-cell epitopes. For the VH domain, the germline mutation M34I was introduced for reduced immunogenicity (due to an increased propensity for binding MHC- II) and removal of a high-risk motif for potential methionine oxidation. Mutation T58V was introduced for removal of a high-risk motif for potential aspartate isomerisation. This mutation was favoured over backmutation T58I for reduced immunogenicity (again to mitigate potential binding to MHC-II).

[0297] The resulting variant expressed well and exhibited favourable biophysical properties, as assessed by SEC and HIC. This variant, termed TAGla (SEQ ID NO:39), bound to STn produced on HEK293 cells and endogenous TAG+SNU16 cells. No binding to TAG-negative cells was observed, confirming binding specificity towards TAG. The bispecific (mAb2format) variant CD137-AA / TAGla mediated improved and potent CD8+T-cell activation levels in the presence of HEK.STGGalNAcl.lcells, as compared to the parent mAbm4 in bispecific (mAb2) format (Figure 11). No activation of T cells was observed in the presence of TAG-negative HEK293 cells. As a result, the Fab region of this TAGla variant was deemed to be suitable for use in mAb2-format bispecific antibody molecules.

[0298] Example 9 - Binding Kinetics Determination

[0299] Kinetics of CD137-AA / TAGla binding to its targets were determined in solution to targets in soluble format (by SPR, Example 9a), on the surface of the cells - to high levels of STn on the HEK.ST6GalNAcl.lcell line and two cell lines

[0300] 37

[0301] 008828600 derived from cancer patients, OV90 (high level of TAG, Example 9c) and SNU16 (medium level of TAG Example 9d). Additionally, binding at acidic pH mimicking TME conditions were determined using the OV90 cell line (Example 9e).

[0302] Example 9a - Kinetics of CD137-AA / TAGla binding dimeric human CD137 in solution

[0303] Approximately 3000 Response Units (RU) of anti-human Fab antibody was immobilised on Biacore Sensor S chip CM5 on flow cells 1 and 2 using Human Fab Capture kit following the manufacturer's instructions. CD137- AA / TAGla, CD137 / HelD1.3 and TAGla-IgGl, diluted to 10 - 20 pg / mL in 1 x HBS-EP+ (running buffer), were injected individually on flow cell 2 at 30 pL / minute to achieve a final response of approximately 200 RU.

[0304] Human dimeric CD137-mFc was prepared at six different concentrations: 27 nM, 9 nM, 3 nM, 1 nM, 0.33 nM, and 0.11 nM (diluted in running buffer). Buffer controls (0 nM) were included twice at the beginning and end of each assay. The antigen was injected on flow cells 1 and 2 for 240 seconds at a flow rate of 75 pL / minute and then allowed to dissociate in running buffer for 600 seconds. Regeneration was carried out for 60 seconds at 30 pL / minute with 10 mM Glycine-HCI pH 2.1. Binding of analyte at each concentration was measured twice to account for time-dependent variability. Subtracted data (flow cell 2 - flow cell 1) were analysed and affinities were determined with the BIAevaluation 2.0 software using the binding curves 0 to 9 nM of CD137 antigens and 1:1 binding model for fitting of the resulting response curves. The experiment was repeated once, and the obtained affinity values were averaged to determine the reported affinity.

[0305] The binding affinity and kinetics of CD137 binding are shown in Table 10. CD137-AA / TAGla had a binding affinity (KD) for dimeric human CD137 (CD137-mFc) of 1.2 nM (± 0.3 nM), and CD137 / Mock mAb2Ctrl had a binding affinity (KD) for dimeric human CD137 (CD137-mFc) of 1.3 nM (± 0.3 nM) demonstrating a lack of effect of VH / VL domains on CD137 binding. No binding could be observed for the anti-TAG mAb Ctrl (TAGla-IgGl-AA) to human CD137.

[0306] Table 10: Binding Affinity and Kinetics of CD137-AA / TAGla, CD137 / HelD1.3, and TAGla-IgGl-AA for Human CD137-mFc. The indicated rates and standard errors represent the average value obtained from two experiments.

[0307] 008828600 Example 9b - Binding of CD137-AA / TAGla to STn on HEK.ST6GalNAcl.l cells and QV90 cells

[0308] HEK.STGGalNAcl.l and HEK.EV cells were cultured in DMEM, supplemented with 10% heat-inactivated FBS and 100 pg / mL Hygromycin B. Prior to the experiment, cells were dissociated from the culture flask in 3 mL Trypsin- EDTA (0.25%) for 5 minutes at 37°C, followed by neutralisation in 8 mL culture media.

[0309] OV90 cells, reported to endogenously express TAG at high levels, were cultured in a 1:1 ratio of Mediuml99 and MCDB 105, supplemented with 15% heat-inactivated FBS and 10 U / mL Penicillin-Streptomycin. Prior to the experiment, cells were detached using 5 mL cell dissociation buffer For both cell types, cells were added at 0.5-1.0 x 105cells per well to a 96-well plate, washed once in FACS buffer and then incubated with antibody diluted in FACS buffer at 4°C for 3 hours. Cells were then washed with FACS buffer twice and incubated with secondary antibody anti-human IgG Fc, conjugated with A488, diluted 1:1000 in FACS buffer at 4°C for 1 hour. Cells were then washed twice with FACS buffer and samples read on the iQue Flow Cytometer. The data was analysed using ForeCyt software.

[0310] Exported data was analysed using GraphPad Prism (version 9). Before plotting the data and fitting a nonlinear regression curve, using the model 'Log(Agonist) vs Response (3-parameter)', the concentration values were transformed (X = Log(X)).

[0311] CD137-AA / TAGla was able to bind STn on HEK.STBGalNAcl.l cells, with an average EC5o value of 0.8 ± 0.3 nM across the three repeats (Table 11 and Figure 12). No binding towards the corresponding STn-negative cell line HEK.EV could be observed for any tested variant at a drug concentration of 1 pM, showing the specificity of the binding.

[0312] CD137-AA / TAGla was able to bind to OV90 cells with an average EC5o value of 2.1 ± 1.6 nM across the three repeats (Table 12 and Figure 13).

[0313] A similar experiment was carried out using SNU16 cells (reported to endogenously produce TAG at moderate levels), and also demonstrated that CD137-AA / TAGla was able to bind, with an appropriate EC5o value however, data not shown.

[0314] Table 11: Summary of binding affinities to STn overexpressing HEK.ST6GalNAcl.l cells. EC50 represents the drug concentration achieving half maximum response.

[0315] 39

[0316] 008828600 Table 12: Summary of binding affinities to endogenous TAG-expressing OV90 cells. EC50 represents the drug concentration achieving half maximum response.

[0317] Example 9c - Kinetics of CD137-AA / TAGla binding TAG on QV90 live adherent cells and HEK.ST6GalNAcl.lcells with high TAG levels

[0318] TAG positive OV90 cell line is an epithelial-like cell isolated from the ovary of a female patient with malignant papillary serous adenocarcinoma with naturally high levels of TAG on the cell surface. TAG-positive OV90 and STn-negative cells (HEK.EV) were immobilised on a culture treated petri dish by seeding cells into a tilted culture dish and incubation at 37 °C, 5% CO2. At approximately 80% confluency, cell-specific media was replaced with CO2-independent media, containing 10% FBS, and fluorescent signal measured in a Ligand Tracer Green at room temperature for determination of baseline. CD137-AA / TAGla and control molecules were fluorescently labelled using Alexa Fluor 488™ Antibody Labelling Kit according to the manufacturer's instructions. Following measurement of baseline, labelled CD137-AA / TAGla or control molecule was added to the media to a final concentration of 3 nM, and the change in fluorescent signal monitored for approximately two hours for association. The association was continued for a second analyte concentration of 9 nM for another two hours. In case of the endogenous cell line OV90, a third association step was measured at a concentration of 27 nM for two hours. Following association, the media was exchanged to fresh CO2- independent media, including 10% FBS, and the labelled antibody allowed to dissociate over several hours.

[0319] Subtracted data (immobilised cells - Blank) were analysed and affinities were determined with the LigandTracer Evaluation software using the 1:1 binding model for fitting of the resulting response curves. The experiment was repeated once, and the obtained affinity values were averaged to determine the reported affinity.

[0320] Experiment was also repeated with HEK.ST6GalNAcl.lcells. CD137-AA / TAGla bound to HEK.ST6GalNAcl.lcells with an average KDof 0.57 ± 0.13 nM across two repeated experiments and to

[0321] 40

[0322] 008828600 endogenously TAG-producing OV90 cells with an average KDof 2.49 ± 1.09 nM across two repeated experiments. No binding to HEK.EV cells could be observed.

[0323] Example 9d - Kinetics of CD137-AA / TAGla binding TAG on live suspension cell line SNU16 - medium level of TAG production

[0324] SNU16 is a TAG+suspension cell line exhibiting epithelial morphology that was isolated in 1987 from ascites derived from a 33-year-old, female, Asian, stomach cancer patient. These cells were immobilised on a petri dish according to the manufacturer's instructions. Following overnight incubation at 37 °C, 5% CO2, cell-specific media was replaced with CCh-independent media, containing 10% FBS, and fluorescent signal measured on a LigandTracer Green at room temperature for determination of baseline. CD137-AA / TAG1A and control molecules were fluorescently labelled using Alexa Fluor 488™ Antibody Labelling Kit according to manufacturer's instructions. Following the measurement of the baseline, labelled CD137-AA / TAG1A or a control molecule was added to the media to a final concentration of 3 nM, and the change in fluorescent signal was monitored for approximately two hours for association. The association was continued for a second analyte concentration of 9 nM for another two hours. A third association step was measured at a concentration of 27 nM for two hours. Following association, the media was exchanged to fresh CO2- independent media, including 10% FBS, and the labelled antibody allowed to dissociate over several hours.

[0325] Subtracted data (immobilised cells - Blank) were analysed and affinities were determined with the LigandTracer Evaluation software using the 1:1 binding model for fitting of the resulting response curves. The experiment was repeated once, and the obtained affinity values were averaged to determine the reported affinity. CD137-AA / TAG1A bound to SNU16 cells with an average KDof 1.38 ± 0.54 nM across two repeated experiments.

[0326] Example 9e - Kinetics of CD137-AA / TAGla towards TAG on live adherent cell line at acidic pH mimicking TME

[0327] TAG-producing cells were immobilized on a culture treated petri dish by seeding cells into a tilted culture dish and incubation at 37 °C, 5% CO2. At approximately 80% confluency, cell-specific media was replaced, and cells washed with CCh-independent media, containing 10% FBS and adjusted to pH 5.8 (assay media). Fresh assay media (3ml) was added, and fluorescent signal measured in a LigandTracer Green at room temperature for determination of baseline. CD137-AA / TAGla and control molecules were fluorescently labelled using Alexa Fluor 488™ Antibody Labelling Kit according to the manufacturer's instructions. Following measurement of baseline, labelled CD137-AA / TAGla or control molecule was added to the media to a final concentration of 3 nM, and the change in fluorescent signal monitored for approximately one hours for association. The

[0328] 41

[0329] 008828600 association was continued for a second analyte concentration of 9 nM for another hour. Following association, the media was exchanged to fresh COj-independent media, including 10% FBS and adjusted to pH 5.8, and the labelled antibody allowed to dissociate for approximately two hours.

[0330] Subtracted data (immobilised cells - Blank) were analysed and affinities were determined with the LigandTracer Evaluation software using the 1:1 binding model for fitting of the resulting response curves. The experiment was repeated once, and the obtained affinity values were averaged to determine the reported affinity. The resulting kinetic rates were compared to those obtained under physiological pH Table 13). No decrease in binding affinity to TAG on live cells could be observed for FS822 under acidic pH (pH 5.8) as compared to physiological pH (pH 7.4). Based on these results, a potential loss of binding to TAG under acidic conditions, as present in tumour environments, is not expected.

[0331] Table 13: Summary of Binding Kinetics of CD137-AA / TAGla to TAG on Live TAG-overexpressing Cells at pH 7.4 and 5.8. The indicated binding rates were obtained from a 1:1 binding model.

[0332] Example 10 - Simultaneous binding of CD137-AA / TAGla to CD137 and TAG

[0333] Simultaneous binding of CD137-AA / TAGla to CD137 and TAG was assessed on a LigandTracer Green. TAG- producing cells HEK.STGGalNAcl.l were immobilized on a culture treated petri dish by seeding cells into a tilted culture dish and incubation at 37 °C, 5% CO2. At approximately 80% confluency, cell-specific media was replaced with COj-independent media, containing 10% FBS, and fluorescent signal measured in a LigandTracer Green at room temperature for determination of baseline. CD137-AA / TAGla, dimeric human CD137 (hCD137- mFc), and anti-TAG mAb (Gl-AA / TAGla) were fluorescently labelled using Alexa Fluor 488™ Antibody Labelling Kit according to the manufacturer's instructions. Following measurement of baseline, labelled CD137- AA / TAGla or anti-TAG mAb was added to the media to a final concentration of 50 nM, and the change in fluorescent signal monitored for approximately two hours for association. The media was replaced for fresh CO2-independent media + 10% FBS only or for fresh media containing fluorescently labelled CD137 at 50 nM. The change in fluorescent signal was monitored for approximately one hour for association. Following association, the media was exchanged to fresh COj-independent media, including 10% FBS, and the labelled antibody and CD137 allowed to dissociate over several hours.

[0334] 008828600 Subtracted data (immobilised cells - Blank) were analysed with the LigandTracer Evaluation software. Spikes were removed from the response curves and obtained signals normalised and aligned to the association of CD137-AA / TAGla or anti-TAG mAb, and the resulting response curves overlayed. Simultaneous binding of CD137-AA / TAG1A to CD137 and TAG could be observed.

[0335] The experiment was repeated once to confirm observed binding and the results supported the finding that binding to both antigens simultaneously was needed for T-cell-directed activity and is the mode of action of CD137-AA / TAG1A.

[0336] Example 11 - Glycan specificity

[0337] The glycan specificity of CD137-AA / TAGla to human O-linked and N-linked glycans was assessed on three different microarrays of glycans. Short Mucl peptides were used as carriers on the arrays. Different glycans were linked synthetically to serine and threonine on the Mucl peptides. After blocking the microarray with Glycan Array Blocking Buffer (GABB) at room temperature for 30 minutes, CD137-AA / TAGla (0.02-15 ug / ml) and control antibodies were added to the microarray and incubated at room temperature for 1 hour. The microarray was then washed and incubated with the detection anti-human IgG Fc Cy3 secondary antibody (10 pg / mL) at room temperature for 1 hour. After washing, the microarray was scanned and the resulting images were analysed using Mapix.

[0338] Example 11a - Glycan specificity of CD137-AA / TAGla towards human O-glycans

[0339] The glycan specificity of CD137-AA / TAGla towards human O-glycans was assessed on a microarray of 94 common human O-glycans (01-094). CD137-AA / TAGla was observed to bind strongly to STn. The observed binding was strong for both serine- and threonine-linked STn, with favoured binding towards serine-linked STn. Residual binding could be detected for serine-linked Tn antigen at the highest tested antibody concentration. Faint nonspecific binding was observed for probes 040-048 at the highest tested antibody concentration. Non-specific binding was no longer observable at low antibody concentration. No binding to further O-linked glycans could be identified.

[0340] Example lib - Glycan specificity of CD137-AA / TAGla towards human N-glycans

[0341] The glycan specificity of CD137-AA / TAGla towards human N-glycans was assessed on a microarray of 114 common human N-glycans (N1-N114). No binding could be observed to any of the N-glycan probes by CD137- AA / TAGla.

[0342] 43

[0343] 008828600 The respective positive and negative controls performed as expected.

[0344] Example 11c - Determination of glycan motifs favoured for binding by CD137-AA / TAGla

[0345] Binding to different glycan motifs by CD137-AA / TAGla was determined on a microarray of different clustered and unclustered STn and Tn antigens on a Mucl peptide. CD137-AA / TAGla was observed to bind strongly to STn bearing glycopeptides (relative binding EC50 to STn motif AO = 1.13 nM). Binding to Tn bearing glycopeptides was also observed for CD137-AA / TAGla (Figure 14; relative binding EC50 to Tn motif CO = 114.11 nM). (The STn motif AO and Tn motif CO on the Mucl peptide are as shown in Figure 27). However, with decreasing concentration of CD137-AA / TAGla, preference for interaction with STn remained strong, while binding to Tn was no longer observable at low concentrations. Binding of CD137-AA / TAGla strongly favoured clustered STn, with highest binding signals observed for three or more clustered residues. The respective binding signals remained strong, even at lower concentrations of CD137-AA / TAGla. Further, CD137-AA / TAGla exhibited differential binding to clustered STn based on the underlying AA sequence, however this limited data set prevents making any conclusions at to the cause of this variation. Binding in the case of a single STn residues favoured serine-linked STn over threonine-linked STn. However, both glycoforms exhibited strong binding at higher concentrations of CD137-AA / TAGla. Tn was only bound by CD137- AA / TAGla in clustered form, with three or more Tn residues being displayed on the carrier peptide. Binding signals of CD137-AA / TAGla for binding clustered Tn decreased at lower concentrations of analyte.

[0346] The respective control assay performed as expected. The anti-human IgG detection antibody interacted with the positive control probe containing human IgG (PC2) and cross-reacted with mouse IgG (PC3) and rabbit IgG (PC4).

[0347] CD137-AA / TAGla bound strongly to STn residues, with residual binding detected for Tn, when sampled against unclustered glycans. No specific binding to further O-linked glycans and N-linked glycans could be observed. As a result, of 94 common human O-linked and 114 common human N-linked glycans tested, only the tumour- associated glycans STn and Tn were bound by CD137-AA / TAGla. CD137-AA / TAGla can therefore be assumed to be highly specific towards STn and Tn. The favouring of clustered over unclustered glycan residues may provide opportunity for further layers of safety by preventing interaction of sporadically occurring truncated glycans, such as Tn or STn, by CD137-AA / TAGla in healthy tissue.

[0348] Example 12 - Functional activity of CD137-AA / TAGla in primary CD8+T-cell activation assays in the presence of STn-producing cells.

[0349] 44

[0350] 008828600 Primary CD8+T cells were isolated from PBMCs obtained from leukocyte cones of healthy donors, using negative separation through magnetic columns (manufacturer's protocol CD8+T Cell Isolation Kit, human, Miltenyi Biotec). Anti-CD3e antibody (clone UCHT1), used to induce surface expression of CD137, was coated on plates overnight. STn+cells and CD8+T cells were then added in the presence of respective drugs and cocultured for three days. CD137-AA / TAGla activated anti-CD3 co-stimulated CD8+T cells to release IL-2, as measured by quantitative ELISA of the cell supernatant (manufacturer's protocol, Human IL-2 ELISA Ready- SET-Go), in the presence of HEK.ST6GalNAcl.l cells which produce STn as described in Example 10 (Figure 15) and not in the presence of HEK.EV cells that do not produce STn. These observations confirmed the requirement for STn on the cell surface for CD137-directed activation of CD8+T cells by CD137-AA / TAGla. No CD8+T-cell activation could be observed for the combination of TAGla-IgGl-AA with CD137-AA / HelD1.3, confirming the requirement of both binding specificities, directed against STn and CD137, to be present on a single molecular entity for the conditional activation of T cells in the presence of STn-positive cells. Further, no CD8+T-cell activation was observed for CD137-AA / HelD1.3 alone in the presence of STn-positive cells, confirming the lack of CD137 agonising activity in the absence of a TAG-binding moiety (Figure 15). Levels of CD137-AA / TAGla-mediated CD8+T-cell activation in the presence of STn-producing cells had below one nanomolar EC50 values (EC5o = 0.6 nM) calculated from five repeated experiments, using the same donor. No response was seen with the STn negative cell line HEK.EV.

[0351] In conclusion, CD137-AA / TAGla required crosslinking by TAG-expressing cells for CD8+ T-cell activity. Furthermore, both binding specificities, directed against TAG and CD137, needed to be present on a single molecular entity for conditional activation of CD8 T cells in the presence of STn-positive cells.

[0352] Example 13 - Functional activity of CD137-AA / TAGla in primary CD8+T-cell activation assays across multiple donors

[0353] CD137-AA / TAGla activated anti-CD3 co-stimulated CD8+T cells in the presence of endogenous TAG+cells (OV90 and SNU16) in samples from multiple donors, as evidenced by IL-2 cytokine production in quantitative ELISA (method in Example 12). No CD8+T-cell activation was observed for the combination of TAGla-IgGl-AA and CD137-AA / HelD1.3, in the presence of endogenous TAG+cell lines. This observation was consistent across all tested donors and confirmed the requirement of both binding specificities, directed against TAG and CD137, to be present on a single molecular entity for conditional activation of T cells in the presence of TAG- positive cells.

[0354] Levels of CD137-AA / TAGla-mediated CD8+T-cell activation in the presence of endogenous cell lines, OV90 and SNU16, were consistent across donors (Figures 16 and 17). For the high expressing TAG+OV90 cells, the corresponding Emax values for CD137-AA / TAGla were consistently above those observed for the tested CD137

[0355] 45

[0356] 008828600 agonist Control molecules A and B which represent strong and weak CD137 agonists respectively. Emaxvalues for CD137-AA / TAGla were approximately 3-fold higher than those observed for Control A and 4-fold higher than those observed for Control B (when control B is crosslinked with an anti-human IgG CH2 antibody (MKlA6-lgGl) (Table 14). The corresponding EC50 values for CD137-AA / TAGla were within low single-digit nanomolar range (average EC50 = 1.5 nM) (Table 14).

[0357] For the TAG+SNU16 cells that have medium levels of TAG, the corresponding Emaxvalues for CD137-AA / TAGla were consistently similar to those observed for the tested CD137 agonist Control B (again crosslinked by anti- CH2 mAb MKlA6-lgGl) and approximately half the value of those observed for the CD137 agonist Control A (Table 14). The EC5o values for CD137-AA / TAGla were within low single digit nanomolar range (average EC5o = 2.0 nM) (Table 14), and the average EC5o and Emaxwere calculated from three repeated experiments, using a different donor for each experiment. As with STn-producing HEK cells, no CD8+T-cell activation could be observed for the combination of TAGla-IgGl-AA with CD137-AA / HelD1.3, with either OV90 or SNU16 TAG+cell lines confirming the requirement of both binding specificities, directed against TAG and CD137, to be present on a single molecular entity for conditional activation of T cells in the presence of physiological TAG (Figures 16 and 17). Consequently, the activity seen with CD137-AA / TAGla demonstrates the ability of CD8+T cells to respond to CD137 agonism when the antibody is crosslinked by binding to physiologically-produced TAG. The maximal response (Emax) was markedly higher with the CD137-AA / TAGla compared to the Controls (Table 14) using the OV90 cells for crosslinking and comparable to Controls using the SNU16 cells. The level of activation was dependent on the density of TAG on the surface of the cells. Potent CD8+T-cell activation by CD137-AA / TAGla was observed in the presence of endogenous TAG+cells. CD137-AA / TAGla-mediated CD8+T-cell potency as shown by the EC50 values was comparable to the Controls. These observations were consistent across multiple donors.

[0358] In conclusion, CD137-AA / TAGla was only able to activate primary human CD8+T cells in the presence of crosslinking cells that express TAG, with maximal response levels higher than CD137 agonist Controls. Despite the EC50 values of the two controls appearing more potent than CD137-AA / TAGla, we hypothesize that the TAG conditionality of the CD137 agonism of CD137-AA / TAGla has the additional advantage of managing localised tumour-targeted CD137 activation that avoids systemic toxicity, which is a known safety concern of the existing CD137 therapeutics.

[0359] Table 14: Summary of CD8+T-cell activity (IL-2 release) in the Primary CD8+T-Cell Activation Assay across multiple donors in the presence of the endogenous TAG+cell lines OV90 and SNU16.

[0360] 46

[0361] 008828600

[0362] Example 14 - Developing a mouse tumour model that expresses STn

[0363] A number of commonly-used mouse tumour cell lines that produce tumour models were screened for endogenous production of STn. However, cell lines with sufficient levels of STn could not be identified. The CT26.WT (ATCC #CRL-2638) colorectal carcinoma cells have a very low level of endogenous STn production. This cell line was engineered to over-express the ST6GalNAcl.l enzyme that sialylates Tn into STn. Using the human ST6GalNAcl.l gene (function is conserved between mouse and humans) subcloned into the pcDNA3.1 Geneticin resistance vector, the parental CT26.WT cells were transfected using Lipofectamine 3000 and grown in 10% FBS supplemented RPMI 1640 Medium GlutaMax with 750 ug / ml of Geneticin for 3-4 weeks. The cells were then single cell plated into 96 well plates in the same media and screened for STn on the surface of the cells. The clone with the highest STn levels was named CT26.STnLlF9, representing the STn-producing CT26 cell line. After checking levels of STn did not alter over 4 weeks of cell culture (split twice a week), the cells were inoculated into Balb / c mice to check for a sufficient tumour growth rate and to determine whether ex vivo STn production was still present. A number of mouse tumour cell lines have been used to engineer the ST6GalNAcl.l overexpression to drive STn expression however, it has proven difficult to screen out engineered clones that have high level of STn expression that are maintained through passages in vitro and in vivo. The CT26.STnLlF9 cells developed into tumours at a rate similar to other mouse cancer models and the ex vivo tumours retained membrane bound STn that was higher than tumours developed from the parental CT26.WT cells. Also, the CT26.STnLlF9 ex vivo STn level was comparable to its in vitro counterpart.

[0364] The mouse CD137-AA / TAGla surrogate molecule and TAG mAb (TAGla-IgGl-AA) were then assessed for binding to the engineered CT26.STnLlF9 STn mouse cell line, alongside CT26.WT cell line and mouse splenocyte cells (negative for STn). Non-specific binding was assessed by comparing to the control IgGl FITC negative control, which does not bind either target (CD137 or STn).

[0365] 008828600 CT26.STnLlF9 cells were grown in 10% Hi-FBS (Gibco #2452396RP) supplemented RPMI 1640 Medium GlutaMax (Gibco #61965-026) with the addition of 750ug / ml of Geneticin (Gibco #10131-035) or without Geneticin for the CT26.WT cells. Cells were detached using cell dissociation buffer (Gibco #A7906-500G) and washed in DPBS (Gibco #14190-094) before plating at lxlO5cells per well in a 96-well plate (Costar #734- 2328). Splenocytes were thawed from -80°C and washed before plating as described. Cells were then incubated with viability dye (L / D e780 Invitrogen #65-0864-14, 1:1000 dilution) and washed before incubating in O-lOOnM primary antibodies (mouse CD137-AA / TAGla, mouse CD137 / TAGla-lgGl, TAG benchmark - Gl / E8v2 or IgG control - IgGl FITC Negative control; in-house production) and FACs buffer (0.5% BSA in PBS) for 30 minutes at 4°C. Cells were washed with PBS and incubated in secondary antibody (goat anti-human anti- Fc AF488; Jackson ImmunoResearch #109-546-098) for 30 minutes at 4°C in the dark. Final washes in PBS were completed before analysing the samples on a flow cytometer (BD Fortessa R647800E6099). Data was analysed using FlowJo™ software (FlowJo vlO.9.0) to compare the STn expression across the cell lines.

[0366] The mouse-CD137-AA / TAGla surrogate was found to bind CT26.STnLlF9 cells in this experiment with a EC5o of 5.5 nM compared to the CT26.WT cells with an EC5o of 1200 nM, indicating there is increased binding to STn in the ST6GalNAcl.l over-expressing mouse cells compared to a cell line with almost no STn expression. To include a truly negative control, we show that splenocytes do not have measurable binding to our STn- targeting molecules (Figure 22).

[0367] Example 15 - mouse-CD137-AA / TAGla activates T cells in co-culture with STn-expressing mouse cells

[0368] To assess whether the bi-specific targeting of mouse-CD137-AA / TAGla surrogate mAb2confers immune cell activating function, the mAb2was incubated in a co-culture of mouse CT26.STnLlF9 cells with DO11.10 cells engineered to overexpress mouse CD137 (DO11.10-mCD137).

[0369] DO11.10-mCD137 T cells were cultured in DMEM (Gibco #61965-026) supplemented with 10% Hi-FBS (Gibco, #2452396RP), ImM sodium Pyruvate (Life Technologies #11360-070) and lug / ml Puromycin (Invitrogen, #ant- pr-1). Round-bottomed 96-well plate were coated with anti-mouse CD3 at lOOng / ml at 4°C overnight. Cancer cell lines in assay media (RPMI + 1% FBS) were added to the plates and incubated overnight at 37°C. Next day, DO11.10-mCD137 cells and the appropriate dilution of each antibody were added to the cancer cells and incubated for a further 24 hours. The supernatant is then collected and used in a mouse-IL2 detecting ELISA kit (Life Technologies #10547781) as per manufacturer's instructions to quantify the amount of I L-2 released in response to the mAb2binding and activation.

[0370] 008828600 When the STn-positive CT26.STnLlF9 cell line was co-cultured with DO11.10-mCD137 T cells, increased mouse IL-2 was detected in the supernatant when mouse-CD137-AA / TAGla surrogate was added. These results suggest that mouse-CD137-AA / TAGla molecule activates T cells in co-culture with STn-producing cells (Figure 23). Neither TAGla-IgGl-AA nor mCD137 / FITC alone induced IL-2 release, demonstrating the TAG-dependent conditional CD137 agonism induced by the engagement to both targets of the mouse-CD137-AA / TAGla surrogate molecule.

[0371] Example 16 - Efficacy studies to show mouse-CD137-AA / TAGla surrogate decreases tumour growth compared to IgG control

[0372] As the mouse-CD137-AA / TAGla surrogate molecule was able to activate DO11.10 T cells expressing mouse CD137 in vitro, the activity of the molecule was investigated in vivo.

[0373] Balb / c female mice (Charles River) aged 8-10 weeks and weighing 18-30g each were rested for three weeks prior to the study start. All animals were micro-chipped and given a unique identifier. Each cohort had 25 mice and each animal was inoculated with 0.5xl06of the engineered CT26.STnLlF9 colorectal carcinoma cells (a low STn-expressing cell line; Example 14) subcutaneously into the left flank in lOOpI DMEM.

[0374] The mouse-CD137-AA / TAGla surrogate molecule or IgG 1 FITC negative control were administered intraperitoneally into mice at a dose of 200pg (at lmg / ml in DPBS containing lOOnM Arginine and 0.05% Tween80) on days 13, 15 and 17 following tumour inoculation.

[0375] Tumour volume measurements were taken with calipers to determine the longest axis and the shortest axis of the tumour. The formula L x (S2) / 2 was used to calculate the tumour volume, where L = longest axis; S = shortest axis.

[0376] Tumours measuring equal to or below 62.5mm3were considered to be a non-established tumour (Table 15). This threshold was chosen as this size limit allows for any remaining inoculation material that did not develop into a tumour. Mice with established tumours that remitted to the size equal to or below the threshold for an established tumour, in response to treatments were regarded as complete responders.

[0377] 49

[0378] 008828600 Table 15: Summary of complete responders and mice without an established tumour at end of study 1 and 2 (referred to as the repeat).

[0379] All mice treated with IgG 1 FITC negative control had an established tumour by the end of the study. Of mice treated with mouse-CD137-AA / TAGla surrogate molecule, 32% did not have an established tumour at the end of the study. The tumour growth measurements in mm3for each animal and the mean tumour volume of each group over time are shown in Figure 18. The mean tumour size in the mouse-CD137-AA / TAGla group was significantly reduced compared to the IgG 1 FITC negative control groups (p = 0.00136; Linear Mixed Effect Model analysis). The reduction in mean tumour volume and the increase in mice without an established tumour in the group treated with mouse-CD137-AA / TAGla surrogate molecule, suggests this molecule has tumour growth inhibition activity (Figure 18).

[0380] This study was repeated to confirm the tumour growth inhibition activity of the mouse-CD137-AA / TAGla surrogate molecule (Figure 19; p = 0.0003 Linear Mixed Effect Model analysis). Similarly, the group of mice treated with the mouse-CD137-AA / TAGla surrogate molecule had more mice at the end of study without an established tumour and had three complete responders (Table 15). Of mice treated with IgGl FITC negative control, 3 mice (12%) did not have an established tumour by the end of the study. Of mice treated with mouse- CD137-AA / TAGla surrogate molecule, 40% did not have an established tumour at the end of the study.

[0381] Survival analysis of the two mouse studies (Figure 20, 21 and Table 16) show that mouse-CD137-AA / TAGla induces a significant survival benefit compared to IgGl FITC negative control.

[0382] Table 16: Survival analysis, showing a summary of median survival in days for each group and pairwise statistical analyses (log-rank) in CT26.STnLlF9 colorectal carcinoma tumour model grown subcutaneously in Balb / c mice treated with IgG FITC, TAGla-IgGl-AA or mCD137-AA / TAGla from study 1 and 2 (referred to as the repeat)

[0383] 50

[0384] 008828600

[0385] A repeat study was performed with immunophenotyping analysis of peripheral blood collected on days 3, 5, 7 and 11 post first dose (day 13 following tumour inoculation), to look for markers of T-cell activation associated with the tested molecules. Figure 24 shows that CD8+T cells in the blood of mice treated with mCD137- AA / TAGla, when compared to IgG FITC negative control treated mice have a significant increase in Ki67, a marker of proliferation, on day 7 and 11 post first dose (p=0.0007 and p=0.0001, respectively; Dunn's multiple comparisons test after Kruskal-Wallis / ANOVA). The functional activation marker Granzyme B is also significantly increased in CD8+T cells on day 7 and 11 post first dose (p=0.0053 and p=0.0002, respectively; Dunn's multiple comparisons test after Kruskal-Wallis / ANOVA), compared to the IgG FITC negative control group. This, alongside the increased Ki67 marker, suggests that CD8+T cells are activated and functional in response to treatment with mCD137-AA / TAGla, compared to control molecules, likely leading to the reduced tumour volume and increased survival also associated with mCD137-AA / TAGla in the previous in vivo efficacy studies.

[0386] Example 17 - Secondary mechanism of action - Prophetic example

[0387] Through engagement with receptors such as siglecs expressed on a range of immune cells, TAG has been reported to induce changes in these immune cells skewing them towards a more immunosuppressive phenotype. This can change the expression of receptors and cytokines by cells within the tumour microenvironment and contribute to promoting tumour growth.

[0388] CD137-AA / TAGla binds to TAG through its Fab arms to localize CD137 agonism to the site of the tumour. In doing so, it is hypothesized that CD137-AA / TAGla binding to TAG molecules on the surface of tumour cells may sterically block the interaction of TAG molecules with their receptors on immune cells, and therefore may reduce the immunosuppressive tumour microenvironment that TAG can contribute to. We therefore hypothesize that by binding to TAG in the tumour, CD137-AA / TAGla could exert an anti-tumour effect both by inducing CD137 mediated T-cell activation, and also by blocking TAG-receptor binding and reducing immunosuppressive signalling.

[0389] 008828600 We predict that such a potential secondary mechanism of action could be demonstrated in cell-based assays where TAG expressing tumour cells such as STn expressing tumour cells influence the phenotype and / or activity of receptor expressing immune cells. In such a situation, blocking of TAG with CD137-AA / TAGla, even in the absence of CD137 expression, would be predicted to reduce the immune suppression induced by TAG.

[0390] One example of an assay where this may be detectable may involve co-culturing dendritic cells, which are known to express siglecs, with cancer cells that express TAG on the cell surface. We would anticipate that TAG-siglec interactions would change the maturation state and / or cytokine production profile of the DCs in a way that could be readily quantified by standard experimental techniques such as flow cytometry and ELISA. Addition of CD137-AA / TAGla to this co-culture would be anticipated to reduce these changes in a dose-dependent manner, as CD137-AA / TAGla binds to TAG on the tumour cells and prevents it interacting with siglecs on the dendritic cell.

[0391] Another example of an assay to investigate TAG blocking may involve NK cells, which also express siglecs. NK cells are known to kill certain tumour cell lines, but this killing is reduced in the presence of proteins carrying STn. We hypothesize that addition of CD137-AA / TAGla to a co-culture of NK cells and an NK susceptible tumour cell line in the presence of STn carrying proteins would reduce the effect of the STn carrying protein on the levels of cancer cell killing seen.

[0392] 52

[0393] 008828600 SEQUENCE LISTING

[0394] 53

[0395] 008828600

[0396] 54

[0397] 008828600

[0398] 55

[0399] 008828600

[0400]

[0401] 56

[0402] 008828600

[0403]

[0404] 57

[0405] 008828600

[0406] SEQ ID NO:1: Amino acid sequence of CL4H (CL4 mouse heavy chain)

[0407] QVQLQQPGAELVRPGVSVKLSCRASGYTITTYWMHWIKQRPDQGLERIAEI NTSNGGTNYNEKFKSKATLTVDKASNTAYME

[0408] LSSLTSEDSAVYYCARSDYGYDRAFAYWGQGTLVTVSA

[0409] SEQ ID NO:2: Amino acid sequence of CL4H1

[0410] QVQLVQSGAEVKKPGASVKVSCKASGYTITTYWMHWIRQAPDQGLERMGEIATSNGGTNYNQKFQGRATLTVDKSISTAYM

[0411] ELSRLRSDDTAVYYCARSDYGYARAFAYWGQGTLVTVSS

[0412] SEQ ID NO:3: Amino acid sequence of CL4H2

[0413] QVQLVQSGAEVKKPGASVKVSCKASGYTITTYWMHWVRQAPGQGLERMGEIATSSGGTNYAQKFQGRVTITVDKSTSTAYMEL

[0414] SRLRSDDTAVYYCARSDYGYARAFAYWGQGTLVTVSS

[0415] SEQ ID NO:4: Amino acid sequence of CL4H3

[0416] QVQLVQSGAEVKKPGASVKVSCKASGYTITTYWMHWVRQAPGQGLEWMGEISTSSGGTNYAQKFQGRVTMTVDKSTSTAYME

[0417] LSRLRSDDTAVYYCARSDYGYARAFAYWGQGTLVTVS

[0418] SEQ ID NO:5: Amino acid sequence of CL4H5

[0419] QVQLVQSGAEVKKPGASVKVSCKASGYTITTYWMHWIRQAPGQGLERIAEIATSTGGTNYAQKFQGRATLTVDKASNTAYMELSS

[0420] LTSEDTAVYYCARSDYGYARAFAYWGQGTLVTVSS

[0421] SEQ ID NO:6: Amino acid sequence of CL4L (CL4 mouse light chain)

[0422] QIVLTQSPAIMSASLGERVTMTCTASSSVSSSYLHWYQQKPGSSPKLWIYSTSNLASGVPPRFSGSGSGTSYSLTISSIEAEDAATYYC

[0423] HQYHRSPTFGGGSKLEIR

[0424] SEQ ID NO:7: Amino acid sequence of CL4L1

[0425] EIVLTQSPATLSLSLGERATLSCTASSSVSSSYLHWYQQKPGQAPRLWIYSTSNLASGVPARFSGSGSGTDYTLTISSIEPEDFAVYYCH

[0426] QYHRSPTFGQGTKLEIK

[0427] SEQ ID NO:8: Amino acid sequence of CL4L2

[0428] EIVLTQSPATLSLSLGERATLSCTASSSVSSSYLHWYQQKPGQAPRLWIYSTSNLATGIPARFSGSGSGTDYTLTISSIEPEDFAVYYCH

[0429] QYHRSPTFGQGTKLEIK

[0430] SEQ ID NO:9: Amino acid sequence of CL4L3

[0431] 58

[0432] 008828600 EIVLTQSPATLSLSPGERATLSCRASSSVSSSYLHWYQQKPGQAPRLWIYSTSNRATGIPARFSGSGSGTDYTLTISSLEPEDFAVYYCH

[0433] QYHRSPTFGQGTKLEIK

[0434] SEQ ID NO:10: Amino acid sequence of CL4L4

[0435] EIVLTQSPATLSLSPGERATLSCRASSSVSSSYLHWYQQKPGQAPRLLIYSTSNRATGIPARFSGSGSGTDFTLTISSLEPEDFAVYYCH

[0436] QYHRSPTFGQGTKLEIK

[0437] SEQ ID NO:11: Amino acid sequence of mAbm4 VH domain

[0438] QVQLQQSDAELVKPGASVKISCKASGYTFTDHAIHWAKQKPEQGLEWIGYISPGNDDIKYNEKFKGKATLTADKSSSTAYMQL

[0439] NSLTSEDSAVYFCKRSYYGHWGQGTTLTVSS

[0440] SEQ ID NO:12: Amino acid sequence of mAbm4 VL domain

[0441] DIQMTQSPASLSVSVGETVTITCRASENIYSNLAWYQQKQGKSPQLLVYAATNLADGVPSRFSGSGSGTQYSLKI NSLQSEDFG

[0442] SYYCQHFWGTPYTFGGGTRLEIK

[0443] SEQ ID NO:13: Amino acid sequence of mAbm4 VH1

[0444] QVQLVQSGAEVKKPGASVKVSCKASGYTFTDHAMHWVRQAPGQRLEWIGWISPGNDDTKYSQKFQGRATLTADKSASTAY

[0445] MELSSLRSEDTAVYYCKRSYYGHWGQGTLVTVSS

[0446] SEQ ID NO:14: Amino acid sequence of mAbm4 VH2

[0447] QVQLVQSGAEVKKPGASVKVSCKASGYTFTEHAMHWVRQAPGQRLEWIGWISPGNDDTKYSQKFQGRATLTADKSASTAY

[0448] MELSSLRSEDTAVYYCKRSYYGHWGQGTLVTVSS

[0449] SEQ ID NO:15: Amino acid sequence of mAbm4 VH3

[0450] QVQLVQSGAEVKKPGASVKVSCKASGYTFTDHAMHWVRQAPGQGLEWMGWISPGNDDTKYSQKFQGRVTITADKSASTA

[0451] YMELSSLRSEDTAVYYCKRSYYGHWGQGTLVTVSS

[0452] SEQ ID NO:16: Amino acid sequence of mAbm4 VH4

[0453] QVQLVQSGAEVKKPGASVKVSCKASGYTFTDHAMHWVRQAPGQGLEWMGWISPGNDDTKYSQKFQGRVTITADKSTSTA

[0454] YMELSSLRSEDTAVYYCKRSYYGHWGQGTLVTVSS

[0455] SEQ ID NO:17: Amino acid sequence of mAbm4 VL1

[0456] DIQMTQSPSSVSASVGDRVTITCRASENIYSNLAWYQQKPGKAPKLLVYAATNLASGVPSRFSGSGSGTDYTLTISSLQPEDFATYYC

[0457] QH FWGTPYTFGQGTKLEI K

[0458] 59

[0459] 008828600 SEQ ID N0:18: Amino acid sequence of mAbm4 VL2

[0460] DIQMTQSPSSVSASVGDRVTITCRASENIYSNLAWYQQKPGKAPKLLVYAATNLADGVPSRFSGSGSGTDYTLTISSLQPEDFAT

[0461] YYCQHFWGTPYTFGQGTKLEIK

[0462] SEQ ID NO:19: Amino acid sequence of mAbm4 VL3

[0463] DIQMTQSPSSVSASVGDRVTITCRASENIYSNLAWYQQKPGKAPKLLIYAATNLADGVPSRFSGSGSGTDFTLTISSLQPEDFAT

[0464] YYCQHFWGTPYTFGQGTKLEIK

[0465] SEQ ID NQ:20: Amino acid sequence of mAbm4 VL4

[0466] DIQMTQSPSSVSASVGDRVTITCRASQNIYSNLAWYQQKPGKAPKLLIYAATNLASGVPSRFSGSGSGTDFTLTISSLQPEDFAT

[0467] YYCQHFWGTPYTFGQGTKLEIK

[0468] SEQ ID NO:21: Amino acid sequence of CD137-AA (Fcab sequence with LALA mutation, without truncated hinge region)

[0469] APEAAGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLN

[0470] GKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELPYI IPPYNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLD

[0471] SDGSFFLYSKLTVGADRWLEGNVFSCSVMHEALHNHYTQKSLSLSPG

[0472] SEQ ID NO:22: Amino acid sequence of CD137 Fcab08 (Fcab sequence with LALA mutation, with truncated hinge region)

[0473] TCPPCPAPEAAGGPSVFLFPPKPKDTLMISRTPEVTCVWDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLH

[0474] QDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELNPPYLFSNQVSLTCLVKGFYPSDIAVEWESNGQPENNYK

[0475] TTPPVLDSDGSFFLYSKLTVDYWRWLEGNVFSCSVMHEALHNHYTQKSLSLSPG

[0476] SEQ ID NO:23: Amino acid sequence of CD137 Fcabl7 (Fcab sequence with LALA mutation, with truncated hinge region)

[0477] TCPPCPAPEAAGGPSVFLFPPKPKDTLMISRTPEVTCVWDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLH

[0478] QDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELNPPYLFSNQVSLTCLVKGFYPSDIAVEWESNGQPENNYK

[0479] TTPPVLDSDGSFFLYSKLTVYHWRWLEGNVFSCSVMHEALHNHYTQKSLSLSPG

[0480] SEQ ID NO:24: Amino acid sequence of IgGl FITC negative control heavy chain

[0481] EVKLDETGGGLVQPGRPMKLSCVASGFTFSDYWMNWVRQSPEKGLEWVAQIRNKPYNYETYYSDSVKGRFTISRDDSKSSVYLQ

[0482] MNNLRVEDMGIYYCTGSYYGMDYWGQGTSVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGV

[0483] HTFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRT

[0484] PEVTCWVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKG

[0485] 60

[0486] 008828600 QPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSV MHEALHNHYTQKSLSLSPGK

[0487] SEQ ID NO:25: Amino acid sequence of IgGl FITC negative control light chain

[0488] DVVMTQTPLSLPVSLGDQASISCRSSQSLVHSNGNTYLRWYLQKPGQSPKVLIYKVSNRFSGVPDRFSGSGSGTDFTLKISRVEAED

[0489] LGVYFCSQSTHVPWTFGGGTKLEI KRTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDS KDSTYSLSSTLTLSKADYEKHKVYACEVT

[0490] SEQ ID NO:26: Amino acid sequence of CD137 / HelD1.3 heavy chain (with LALA)

[0491] QVQLQESGPGLVRPSQTLSLTCTVSGSTFSGYGVNWVRQPPGRGLEWIGMIWGDGNTDYNSALKSRVTMLVDTSKNQFSLRLSS

[0492] VTAADTAVYYCARERDYRLDYWGQGSLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFP

[0493] AVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLMISRTPEV

[0494] TCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPR

[0495] EPQVYTLPPSRDELPYIIPPYNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVGADRWLEGNVFSCSV MHEALHNHYTQKSLSLSPG

[0496] SEQ ID NO:27: Amino acid sequence of CD137 / HelD1.3 light chain

[0497] DIQMTQSPASLSASVGETVTITCRASGNIHNYLAWYQQKQGKSPQLLVYNAKTLADGVPSRFSGSGSGTQYSLKI NSLQPEDFGSYY

[0498] CQHFWSTPRTFGGGTKLEIKRTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTY

[0499] SLSSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC

[0500] SEQ ID NO:28: Amino acid sequence of TAGla CDRH1 (Kabat)

[0501] AHAIH

[0502] SEQ ID NO:29: Amino acid sequence of TAGla CDRH2 (Kabat)

[0503] YISPGNEDVKYSQKFQG

[0504] SEQ ID NQ:30: Amino acid sequence of TAGla CDRH3 (Kabat)

[0505] SYYGH

[0506] SEQ ID NO:31: Amino acid sequence of TAGla CDRL1 (Kabat)

[0507] RASENIYSNLA

[0508] SEQ ID NO:32: Amino acid sequence of TAGla CDRL2 (Kabat)

[0509] AASRLQS

[0510] 61

[0511] 008828600 SEQ ID NO:33: Amino acid sequence of TAGla CDRL3 (Kabat)

[0512] QHFFGTPYT

[0513] SEQ ID NO:34: Amino acid sequence of TAG1 VH domain (VH4 with D31A / W50Y / D56E / R98Q)

[0514] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAMHWVRQAPGQGLEWMGYISPGNEDTKYSQKFQGRVTITADKSTSTAYMEL SSLRSEDTAVYYCKQSYYGHWGQGTLVTVSS

[0515] SEQ ID NO:35: Amino acid sequence of TAG1 VL domain (VL1 with W92F)

[0516] DIQMTQSPSSVSASVGDRVTITCRASENIYSNLAWYQQKPGKAPKLLVYAATNLASGVPSRFSGSGSGTDYTLTISSLQPEDFATYYC QHFFGTPYTFGQGTKLEIK

[0517] SEQ ID NO:36: Amino acid sequence of TAG1 heavy chain (TAGl-IgGl)

[0518] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAMHWVRQAPGQGLEWMGYISPGNEDTKYSQKFQGRVTITADKSTSTAYMEL

[0519] SSLRSEDTAVYYCKQSYYGHWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPA

[0520] VLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTC

[0521] VVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREP QVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVM HEAL

[0522] HNHYTQKSLSLSPG

[0523] SEQ ID NO:37: Amino acid sequence of TAG1 heavy chain with LALA (TAGla-IgGl-AA)

[0524] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAMHWVRQAPGQGLEWMGYISPGNEDTKYSQKFQGRVTITADKSTSTAYMEL

[0525] SSLRSEDTAVYYCKQSYYGHWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPA

[0526] VLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLMISRTPEVT

[0527] CVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPRE

[0528] PQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHE ALHNHYTQKSLSLSPG

[0529] SEQ ID NO:38: Amino acid sequence of TAG1 light chain

[0530] DIQMTQSPSSVSASVGDRVTITCRASENIYSNLAWYQQKPGKAPKLLVYAATNLASGVPSRFSGSGSGTDYTLTISSLQPEDFATYYC

[0531] QHFFGTPYTFGQGTKLEIKRTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSL SSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC

[0532] SEQ ID NO:39: Amino acid sequence of TAGla VH domain (VH4 D31A / M34I / W50Y / D56E / T58V / R98Q)

[0533] 62

[0534] 008828600 QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAIHWVRQAPGQGLEWMGYISPGNEDVKYSQKFQGRVTITADKSTSTAYMELS SLRSEDTAVYYCKQSYYGHWGQGTLVTVSS

[0535] SEQ ID NQ:40: Amino acid sequence of TAGla VL domain (VL1 V48I / T52S / N53R / A55Q / W92F)

[0536] DIQMTQSPSSVSASVGDRVTITCRASENIYSNLAWYQQKPGKAPKLLLYAASRLQSGVPSRFSGSGSGTDYTLTISSLQPEDFATYYC QHFFGTPYTFGQGTKLEIK

[0537] SEQ ID NO:41: Amino acid sequence of TAGla heavy chain (TAGla-IgGl)

[0538] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAIHWVRQAPGQGLEWMGYISPGNEDVKYSQKFQGRVTITADKSTSTAYM

[0539] ELSSLRSEDTAVYYCKQSYYGHWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVH

[0540] TFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMIS

[0541] RTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTIS

[0542] KAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQG NVFSCSVMHEALHNHYTQKSLSLSPG

[0543] SEQ ID NO:42: Amino acid sequence of TAGla heavy chain with LALA (TAGla-IgGl-AA)

[0544] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAIHWVRQAPGQGLEWMGYISPGNEDVKYSQKFQGRVTITADKSTSTAYM

[0545] ELSSLRSEDTAVYYCKQSYYGHWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVH

[0546] TFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLMIS

[0547] RTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTIS

[0548] KAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQG NVFSCSVMHEALHNHYTQKSLSLSPG

[0549] SEQ ID NO:43: Amino acid sequence of TAGla light chain

[0550] DIQMTQSPSSVSASVGDRVTITCRASENIYSNLAWYQQKPGKAPKLLLYAASRLQSGVPSRFSGSGSGTDYTLTISSLQPEDFAT

[0551] YYCQHFFGTPYTFGQGTKLEIKRTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSK

[0552] DSTYSLSSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC

[0553] SEQ ID NO:44: Amino acid sequence of mCD137 heavy chain

[0554] TCPPCPAPEAAGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLT

[0555] VLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDEPYWSYVSLTCLVKGFYPSDIAVEWESNGQPENNY

[0556] KTTPPVLDSDGSFFLYSKLTVMNYRWELGNVFSCSVMHEALHNHYTQKSLSLSPG

[0557] SEQ ID NO:45: Amino acid sequence of mCD137 / FITC bispecific control heavy chain

[0558] 63

[0559] 008828600 EVKLDETGGGLVQPGRPMKLSCVASGFTFSDYWMNWVRQSPEKGLEWVAQIRNKPYNYETYYSDSVKGRFTISRDDSKSS

[0560] VYLQM NNLRVEDMGIYYCTGSYYGMDYWGQGTSVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWN

[0561] SGALTSGVHTFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPEAAGGPSVFLF

[0562] PPKPKDTLM ISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVS

[0563] NKALPAPIEKTISKAKGQPREPQVYTLPPSRDEPYWSYVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYS

[0564] KLTVM NYRWELGNVFSCSVMHEALHNHYTQKSLSLSPGK

[0565] SEQ ID NO:46: Amino acid sequence of mCD137 / FITC bispecific control light chain

[0566] DVVMTQTPLSLPVSLGDQASISCRSSQSLVHSNGNTYLRWYLQKPGQSPKVLIYKVSNRFSGVPDRFSGSGSGTDFTLKISRVEAED

[0567] LGVYFCSQSTHVPWTFGGGTKLEI KRTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDS KDSTYSLSSTLTLSKADYEKHKVYACEVTH

[0568] SEQ ID NO:47: Amino acid sequence of mCD137 / TAGla heavy chain

[0569] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAIHWVRQAPGQGLEWMGYISPGNEDVKYSQKFQGRVTITADKSTSTAYMELS

[0570] SLRSEDTAVYYCKQSYYGHWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAV

[0571] LQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLMISRTPEVTC

[0572] VVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREP

[0573] QVYTLPPSRDEPYWSYVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVMNYRWELGNVFSCSVMH EAL HNHYTQKSLSLSPGK

[0574] SEQ ID NO:48: Amino acid sequence of mCD137 / TAGla light chain

[0575] DIQMTQSPSSVSASVGDRVTITCRASENIYSNLAWYQQKPGKAPKLLLYAASRLQSGVPSRFSGSGSGTDYTLTISSLQPEDFATYYC

[0576] QHFFGTPYTFGQGTKLEIKRTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSL

[0577] SSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC

[0578] SEQ ID NO:49: Amino acid sequence of CD137-AA / TAGla heavy chain

[0579] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAIHWVRQAPGQGLEWMGYISPGNEDVKYSQKFQGRVTITADKSTSTAYMELS

[0580] SLRSEDTAVYYCKQSYYGHWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAV

[0581] LQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLMISRTPEVTC

[0582] VVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREP

[0583] QVYTLPPSRDELPYIIPPYNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVGADRWLEGNVFSCSVM HEALHNHYTQKSLSLSPG

[0584] SEQ ID NQ:50: Amino acid sequence of CD137-AA / TAGla light chain

[0585] 64

[0586] 008828600 DIQMTQSPSSVSASVGDRVTITCRASENIYSNLAWYQQKPGKAPKLLLYAASRLQSGVPSRFSGSGSGTDYTLTISSLQPEDFATYYC

[0587] QHFFGTPYTFGQGTKLEIKRTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSL

[0588] SSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC

[0589] SEQ ID N0:51: Amino acid sequence of Control A - heavy chain

[0590] QVQLQQWGAGLLKPSETLSLTCAVYGGSFSGYYWSWIRQSPEKGLEWIGEINHGGYVTYNPSLESRVTISVDTSKNQFSLKLSSVTA

[0591] ADTAVYYCARDYGPGNYDWYFDLWGRGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVH

[0592] TFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLM ISRTP

[0593] EVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQ

[0594] PREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVM HEALHNHYTQKSLSLSPGK

[0595] SEQ ID NO:52: Amino acid sequence of Control A - light chain

[0596] EIVLTQSPATLSLSPGERATLSCRASQSVSSYLAWYQQKPGQAPRLLIYDASNRATGIPARFSGSGSGTDFTLTISSLEPEDFAVYYCQ

[0597] QRSNWPPALTFGGGTKVEIKRTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDST

[0598] YSLSSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC

[0599] SEQ ID NO:53: Amino acid sequence of Control B - heavy chain

[0600] EVQLVQSGAEVKKPGESLRISCKGSGYSFSTYWISWVRQMPGKGLEWMGKIYPGDSYTNYSPSFQGQVTISADKSISTAYLQWSSL

[0601] KASDTAMYYCARGYGIFDYWGQGTLVTVSSASTKGPSVFPLAPCSRSTSESTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAV

[0602] LQSSGLYSLSSVVTVPSSNFGTQTYTCNVDHKPSNTKVDKTVERKCCVECPPCPAPPVAGPSVFLFPPKPKDTLMISRTPEVTCVVVD

[0603] VSHEDPEVQFNWYVDGVEVHNAKTKPREEQFNSTFRVVSVLTVVHQDWLNGKEYKCKVSNKGLPAPIEKTISKTKGQPREPQVYT

[0604] LPPSREEMTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPMLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALH N HYTQKSLSLSPGK

[0605] SEQ ID NO:54: Amino acid sequence of Control B - light chain

[0606] SYELTQPPSVSVSPGQTASITCSGDNIGDQYAHWYQQKPGQSPVLVIYQDKNRPSGIPERFSGSNSGNTATLTISGTQAMDEADYY

[0607] CATYTGFGSLAVFGGGTKLTVLGQPKAAPSVTLFPPSSEELQANKATLVCLISDFYPGAVTVAWKADSSPVKAGVETTTPSKQSNNK YAASSYLSLTPEQWKSHRSYSCQVTHEGSTVEKTVAPTECS

[0608] SEQ ID NO:55: Amino acid sequence MK1A6 heavy chain (anti-CH2 mAb)

[0609] EVQLVESGGGLVKPGGSLKLSCAASGFTFSDYYMYWVRQTPEKRLEWVATISDGGDYTYYADSVKGRFTISRDNAKNNLYLQMSS

[0610] LKSEDTAMFFCAREYYVSSFAYWGQGTLVTVSAAKTTPPSVYPLAPGSAAQTNSMVTLGCLVKGYFPEPVTVTWNSGSLSSGVHT

[0611] FPAVLQSDLYTLSSSVTVPSSTWPSETVTCNVAHPASSTKVDKKIVPRDCGCKPCICTVPEVSSVFIFPPKPKDVLTITLTPKVTCVVVD

[0612] ISKDDPEVQFSWFVDDVEVHTAQTQPREEQFNSTFRSVSELPIMHQDWLNGKEFKCRVNSAAFPAPIEKTISKTKGRPKAPQVYTI

[0613] 65

[0614] 008828600 PPPKEQMAKDKVSLTCMITDFFPEDITVEWQWNGQPAENYKNTQPIMDTDGSYFVYSKLNVQKSNWEAGNTFTCSVLHEGLHN HHTEKSLSHSPGK

[0615] SEQ ID NO. 56: Amino acid sequence of TAG124 VH domain

[0616] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAIHWVRQAPGQGLEWMGYISPGNEDIKYSQKFQGRVTITADKSTSTAYMELSS LRSEDTAVYYCKQSYYGHWGQGTLVTVSS

[0617] SEQ ID NO. 57: Amino acid sequence of TAG129 VH domain

[0618] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAIHWVRQAPGQGLEWMGYISPGNEDLKYSQKFQGRVTITADKSTSTAYMELS SLRSEDTAVYYCKQSYYGHWGQGTLVTVSS

[0619] SEQ ID NO. 58: Amino acid sequence of TAG130 VH domain

[0620] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAIHWVRQAPGQGLEWMGYISPGNEDQKYSQKFQGRVTITADKSTSTAYM ELSSLRSEDTAVYYCKQSYYGHWGQGTLVTVSS

[0621] SEQ ID NO:59: Amino acid sequence of TAG124 heavy chain

[0622] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAIHWVRQAPGQGLEWMGYISPGNEDIKYSQKFQGRVTITADKSTSTAYMELSS

[0623] LRSEDTAVYYCKQSYYGHWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVL

[0624] QSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVV

[0625] VDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQV

[0626] YTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHN HYTQKSLSLSPG

[0627] SEQ ID NQ:60: Amino acid sequence of TAG129 heavy chain

[0628] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAIHWVRQAPGQGLEWMGYISPGNEDLKYSQKFQGRVTITADKSTSTAYMELS

[0629] SLRSEDTAVYYCKQSYYGHWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAV

[0630] LQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCV

[0631] VVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQ

[0632] VYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALH N HYTQKSLSLSPG

[0633] SEQ ID NO:61: Amino acid sequence of TAG130 heavy chain

[0634] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAIHWVRQAPGQGLEWMGYISPGNEDQKYSQKFQGRVTITADKSTSTAYMELS

[0635] SLRSEDTAVYYCKQSYYGHWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAV

[0636] LQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCV

[0637] 66

[0638] 008828600 VVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQ

[0639] VYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALH NHYTQKSLSLSPG

[0640] SEQ ID NO:62: Amino acid sequence of TAG124 CDRH1 (Kabat)

[0641] AHAIH

[0642] SEQ ID NO:63: Amino acid sequence of TAG124 CDRH2 (Kabat)

[0643] YISPGNEDIKYSQKFQG

[0644] SEQ ID NO:64: Amino acid sequence of TAG124 CDRH3 (Kabat)

[0645] SYYGH

[0646] SEQ ID NO:65: Amino acid sequence of TAG124 CDRL1 (Kabat)

[0647] RASENIYSNLA

[0648] SEQ ID NO:66: Amino acid sequence of TAG124 CDRL2 (Kabat)

[0649] AASRLQS

[0650] SEQ ID NO:67: Amino acid sequence of TAG124 CDRL3 (Kabat)

[0651] QHFFGTPYT

[0652] SEQ ID NO:68: Amino acid sequence of TAG129 CDRH1 (Kabat)

[0653] AHAIH

[0654] SEQ ID NO:69: Amino acid sequence of TAG129 CDRH2 (Kabat)

[0655] YISPGNEDLKYSQKFQG

[0656] SEQ ID NQ:70: Amino acid sequence of TAG129 CDRH3 (Kabat)

[0657] SYYGH

[0658] SEQ ID NO:71: Amino acid sequence of TAG129 CDRL1 (Kabat)

[0659] RASENIYSNLA

[0660] SEQ ID NO:72: Amino acid sequence of TAG129 CDRL2 (Kabat)

[0661] 67

[0662] 008828600 AASRLQS

[0663] SEQ ID NO:73: Amino acid sequence of TAG129 CDRL3 (Kabat)

[0664] QHFFGTPYT

[0665] SEQ ID NO:74: Amino acid sequence of TAG130 CDRH1 (Kabat)

[0666] AHAIH

[0667] SEQ ID NO:75: Amino acid sequence of TAG130 CDRH2 (Kabat)

[0668] YISPGNEDQKYSQKFQG

[0669] SEQ ID NO:76: Amino acid sequence of TAG130 CDRH3 (Kabat)

[0670] SYYGH

[0671] SEQ ID NO:77: Amino acid sequence of TAG130 CDRL1 (Kabat)

[0672] RASENIYSNLA

[0673] SEQ ID NO:78: Amino acid sequence of TAG130 CDRL2 (Kabat)

[0674] AASRLQS

[0675] SEQ ID NO:79: Amino acid sequence of TAG130 CDRL3 (Kabat)

[0676] QHFFGTPYT

[0677] SEQ ID NQ:80 Amino acid sequence of MK1A6 Light chain (anti-CH2 mAb)

[0678] DVVVTQTPLSLPVSFGDQVSISCRSSQSLANSYGNTYLSWYLHKPGQSPQLLIYGVSNRFSGVPDRFSGSGSGTDFTLKISTIKPE

[0679] DLGMYYCLQGTHQPYTFGGGTKLEIKRADAAPTVSIFPPSSEQLTSGGASWCFLNNFYPKDINVKWKIDGSERQNGVLNSWT

[0680] DQDSKDSTYSMSSTLTLTKDEYERHNSYTCE

[0681] SEQ ID NO: 81 - Amino acid sequence within the CH3 domain AB loop of CD137-AA Fcab

[0682] PYI IPPY

[0683] SEQ ID NO: 82 - Amino acid sequence within the CH3 domain CD loop of CD137-AA Fcab SNGQPENNY

[0684] SEQ ID NO: 83 Amino acid sequence within the CH3 domain EF loop of CD137-AA Fcab

[0685] 68

[0686] 008828600 GADRWLE

[0687] SEQ ID NO: 84 - Amino acid sequence of CH3 domain of CD137-AA Fcab

[0688] GQPREPQVYTLPPSRDELPYIIPPYNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVGADRWLEG

[0689] NVFSCSVMHEALHNHYTQKSLSLSPG

[0690] SEQ ID NO:85 - Amino acid sequence of CH2 domain of CD137-AA Fcab

[0691] APEAAGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLN

[0692] GKEYKCKVSNKALPAPIEKTISKAK

[0693] SEQ ID NO:86 - Amino acid sequence of mCD137-AA heavy chain

[0694] EVKLDETGGGLVQPGRPMKLSCVASGFTFSDYWMNWVRQSPEKGLEWVAQIRNKPYNYETYYSDSVKGRFTISRDDSKSSVY

[0695] LQMNNLRVEDMGIYYCTGSYYGMDYWGQGTSVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGAL

[0696] TSGVHTFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKK

[0697] SEQ ID NO:87 - Amino acid sequence of CDRH1 ( I MGT) of TAGla, TAG124, TAG129 and TAG130

[0698] GYTFTAHA

[0699] SEQ ID NO:88 - Amino acid sequence of CDRH2 (IMGT) of TAGla

[0700] ISPGNEDV

[0701] SEQ ID NO:89 - Amino acid sequence of CDRH2 (IMGT) of TAG124

[0702] ISPGNEDI

[0703] SEQ ID NQ:90 - Amino acid sequence of CDRH2 (IMGT) of TAG129

[0704] ISPGNEDL

[0705] SEQ ID NO:91 - Amino acid sequence of CDRH2 (IMGT) of TAG130

[0706] ISPGNEDQ

[0707] SEQ ID NO:92 - Amino acid sequence of CDRH3 (IMGT) of TAGla, TAG124, TAG129 and TAG130

[0708] KQSYYGH

[0709] SEQ ID NO:93— Amino acid sequence of CDRL1 (IMGT) of TAGla, TAG124, TAG129 and TAG130

[0710] ENIYSN

[0711] 69

[0712] 008828600 SEQ ID NO:94 - Amino acid sequence of CDRL2(IMGT) of TAGla, TAG124, TAG129 and TAG130

[0713] AAS

[0714] SEQ ID NO:95 - Amino acid sequence of CDRL3 ( I MGT) of TAGla, TAG124, TAG129 and TAG130

[0715] QHFFGTPYT

[0716] SEQ ID NO:96 - Amino acid sequence of CD137-AA / TAG124 Heavy chain

[0717] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAIHWVRQAPGQGLEWMGYISPGNEDIKYSQKFQGRVTITADKSTSTAYME

[0718] LSSLRSEDTAVYYCKQSYYGHWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHT

[0719] FPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLMISR

[0720] TPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISK

[0721] AKGQPREPQVYTLPPSRDELPYIIPPYNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVGADRWL EGNVFSCSVMHEALHNHYTQKSLSLSPG

[0722] SEQ ID NO:97 - Amino acid sequence of CD137-AA / TAG129 Heavy chain

[0723] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAIHWVRQAPGQGLEWMGYISPGNEDLKYSQKFQGRVTITADKSTSTAYM

[0724] ELSSLRSEDTAVYYCKQSYYGHWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVH

[0725] TFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLMIS

[0726] RTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTIS

[0727] KAKGQPREPQVYTLPPSRDELPYIIPPYNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVGADRW LEGNVFSCSVMHEALHNHYTQKSLSLSPG

[0728] SEQ ID NO:98 - Amino acid sequence of CD137-AA / TAG130 Heavy chain

[0729] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAIHWVRQAPGQGLEWMGYISPGNEDQKYSQKFQGRVTITADKSTSTAYM

[0730] ELSSLRSEDTAVYYCKQSYYGHWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVH

[0731] TFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLMIS

[0732] RTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTIS

[0733] KAKGQPREPQVYTLPPSRDELPYIIPPYNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVGADRW LEGNVFSCSVMHEALHNHYTQKSLSLSPG

[0734] SEQ ID NO:99 - Amino acid sequence of CD137 FcabO8 / TAGla Heavy chain

[0735] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAIHWVRQAPGQGLEWMGYISPGNEDVKYSQKFQGRVTITADKSTSTAYM

[0736] ELSSLRSEDTAVYYCKQSYYGHWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVH

[0737] TFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLMIS

[0738] 70

[0739] 008828600 RTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTIS

[0740] KAKGQPREPQVYTLPPSRDELNPPYLFSNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDYWR WLEGNVFSCSVMHEALHNHYTQKSLSLSPG

[0741] SEQ ID N0:100 - Amino acid sequence of CD137 FcabO8 / TAG124 Heavy chain

[0742] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAIHWVRQAPGQGLEWMGYISPGNEDIKYSQKFQGRVTITADKSTSTAYME

[0743] LSSLRSEDTAVYYCKQSYYGHWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHT

[0744] FPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLMISR

[0745] TPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISK

[0746] AKGQPREPQVYTLPPSRDELNPPYLFSNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDYWR WLEGNVFSCSVMHEALHNHYTQKSLSLSPG

[0747] SEQ ID NQ:101 - Amino acid sequence of CD137 FcabO8 / TAG129 Heavy chain

[0748] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAIHWVRQAPGQGLEWMGYISPGNEDLKYSQKFQGRVTITADKSTSTAYM

[0749] ELSSLRSEDTAVYYCKQSYYGHWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVH

[0750] TFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLMIS

[0751] RTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTIS

[0752] KAKGQPREPQVYTLPPSRDELNPPYLFSNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDYWR WLEGNVFSCSVMHEALHNHYTQKSLSLSPG

[0753] SEQ ID NQ:102 - Amino acid sequence of CD137 Fcab08 / TAG130 Heavy chain

[0754] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAIHWVRQAPGQGLEWMGYISPGNEDQKYSQKFQGRVTITADKSTSTAYM

[0755] ELSSLRSEDTAVYYCKQSYYGHWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVH

[0756] TFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLMIS

[0757] RTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTIS

[0758] KAKGQPREPQVYTLPPSRDELNPPYLFSNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDYWR WLEGNVFSCSVMHEALHNHYTQKSLSLSPG

[0759] SEQ ID NQ:103 - Amino acid sequence of CD137 Fcabl7 / TAGla Heavy chain

[0760] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAIHWVRQAPGQGLEWMGYISPGNEDVKYSQKFQGRVTITADKSTSTAYM

[0761] ELSSLRSEDTAVYYCKQSYYGHWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVH

[0762] TFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLMIS

[0763] RTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTIS

[0764] KAKGQPREPQVYTLPPSRDELNPPYLFSNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVYHWR WLEGNVFSCSVMHEALHNHYTQKSLSLSPG

[0765] 71

[0766] 008828600 SEQ ID NQ:104 - Amino acid sequence of CD137 Fcabl7 / TAG124 Heavy chain

[0767] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAIHWVRQAPGQGLEWMGYISPGNEDIKYSQKFQGRVTITADKSTSTAYME

[0768] LSSLRSEDTAVYYCKQSYYGHWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHT

[0769] FPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLMISR

[0770] TPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISK

[0771] AKGQPREPQVYTLPPSRDELNPPYLFSNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVYHWR WLEGNVFSCSVMHEALHNHYTQKSLSLSPG

[0772] SEQ ID NQ:105 - Amino acid sequence of CD137 Fcabl7 / TAG129 Heavy chain

[0773] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAIHWVRQAPGQGLEWMGYISPGNEDLKYSQKFQGRVTITADKSTSTAYM

[0774] ELSSLRSEDTAVYYCKQSYYGHWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVH

[0775] TFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLMIS

[0776] RTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTIS

[0777] KAKGQPREPQVYTLPPSRDELNPPYLFSNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVYHWR WLEGNVFSCSVMHEALHNHYTQKSLSLSPG

[0778] SEQ ID NQ:106 - Amino acid sequence of CD137 Fcabl7 / TAG130 Heavy chain

[0779] QVQLVQSGAEVKKPGASVKVSCKASGYTFTAHAIHWVRQAPGQGLEWMGYISPGNEDQKYSQKFQGRVTITADKSTSTAYM

[0780] ELSSLRSEDTAVYYCKQSYYGHWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVH

[0781] TFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPEAAGGPSVFLFPPKPKDTLMIS

[0782] RTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTIS

[0783] KAKGQPREPQVYTLPPSRDELNPPYLFSNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVYHWR

[0784] WLEGNVFSCSVMHEALHNHYTQKSLSLSPG

[0785] SEQ ID NQ:107 - Amino acid sequence of human IgGl hinge region

[0786] EPKSCDKTHTCPPCP

[0787] SEQ ID NQ:108 - Amino acid sequence of truncated human IgGl hinge region

[0788] TCPPCP

[0789] SEQ ID NQ:109 - Amino acid consensus sequence (Kabat) of CDRH1

[0790] X1HAX2H wherein Xi is alanine (A), aspartic acid (D) or glutamic acid ( E), tryptophan (W) or tyrosine (Y); and X2 is isoleucine

[0791] (I) or methionine (M)

[0792] 72

[0793] 008828600 SEQ ID NO:110 - Amino acid consensus sequence (Kabat) of CDRH2

[0794] YISPGX3X4DX5KYSQKFQG wherein X3 is asparagine (N) or glutamine (Q); X4is alanine (A), aspartic acid (D) or glutamic acid ( E), glycine (G), lysine (K), asparagine (N), glutamine (Q) or tyrosine (Y); and X5is isoleucine (I), leucine (L), glutamine (Q), threonine (T) or valine (V)

[0795] SEQ ID NO:111 - Amino acid consensus sequence (Kabat) of CDRH3

[0796] SYYGXs wherein Xs is glutamic acid (E) or histidine (H)

[0797] SEQ ID NO:112 - Amino acid consensus sequence (Kabat) of CDRL1

[0798] RASX7NIYSNLA wherein X7is glutamic acid (E) or glutamine (Q)

[0799] SEQ ID NO:113 - Amino acid consensus sequence (Kabat) of CDRL2

[0800] AAXsXgLXioXu wherein Xs is serine (S) or threonine (T); Xg is asparagine (N) or arginine (R); X10 is alanine (A) or glutamine (Q); and Xu is aspartic acid (D), leucine (L) or serine (S)

[0801] SEQ ID NO:114 - Amino acid consensus sequence (Kabat / IMGT) of CDRL3

[0802] QHFX12GTPYT wherein X12 is phenylalanine (F) or tryptophan (W)

[0803] SEQ ID NO:115 - AO motif (graphical representation in Figure 27)

[0804] PPAHGVTSAPDTRPAPGSTAPPA wherein S (18) and T (19) each comprise an O-linked polysaccharide chain glycosylation (Neu5Aca2,6GalNAcal), and wherein T (7), S (8) and T (12) each independently optionally comprise an O-linked monosaccharide glycosylation.

[0805] SEQ ID NO:116 - CO motif (graphical representation in Figure 27)

[0806] PPAHGVTSAPDTRPAPGSTAPPA wherein S (18) and T (19) each comprise an O-linked GalNAcal monosaccharide glycosylation, and wherein T (12) optionally comprises an O-linked monosaccharide glycosylation.

[0807] 73

[0808] 008828600 SEQ ID N0:117 - Amino acid consensus sequence (Kabat) of CDRH2

[0809] YISPGNEDXiKYSQKFQG wherein Xi is leucine ( L), glutamine (Q), or valine (V)

[0810] SEQ ID NO:118 - Amino acid consensus sequence (IIX / IGT) of CDRH1

[0811] GYTFX1X2HA wherein Xi is lysine (K) or threonine (T); and X2 is alanine (A), aspartic acid (D), glutamic acid (E), tryptophan (W) or tyrosine (Y)

[0812] SEQ ID NO:119 - Amino acid consensus sequence (IIX / IGT) of CDRH2

[0813] ISPGX3X4DX5 wherein X3 is asparagine (N) or glutamine (Q); X4is alanine (A), aspartic acid (D), glutamic acid (E), glycine (G), lysine (K), asparagine (N), glutamine (Q) or tyrosine (Y); and X5is isoleucine (I), leucine (L), glutamine (Q), threonine (T) or valine (V)

[0814] SEQ ID 1X10:120 - Amino acid consensus sequence (IIX / IGT) of CDRH3

[0815] KX6SYYGX7wherein Xs is leucine ( L), glutamine (Q) or arginine (R); and X7is glutamic acid (E) or histidine (H)

[0816] SEQ ID 1X10:121 - Amino acid consensus sequence (IIX / IGT) of CDRL1

[0817] XgNIYSN wherein Xg is glutamic acid (E) or glutamine (Q)

[0818] SEQ ID 1X10:122 - Amino acid consensus sequence (IIX / IGT) of CDRH2

[0819] ISPGNEDXi wherein Xi is leucine ( L), glutamine (Q), or valine (V)

[0820] SEQ ID 1X10:123: Amino acid sequence of CD137-AA (Fcab sequence with LALA mutation, with truncated hinge region)

[0821] TCPPCPAPEAAGGPSVFLFPPKPKDTLMISRTPEVTCWVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLH

[0822] QDWLIXIGKEYKCKVSIXIKALPAPIEKTISKAKGQPREPQVYTLPPSRDELPYIIPPYIXIQVSLTCLVKGFYPSDIAVEWESIXIGQPEIXIIXIYKT TPPVLDSDGSFFLYSKLTVGADRWLEGIXIVFSCSVMHEALHIXIHYTQKSLSLSPG

[0823] SEQ I D 1X10:124: Amino acid sequence of CD137 Fcab08 (Fcab sequence with LALA mutation, without truncated hinge region)

[0824] 74

[0825] 008828600 APEAAGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLN

[0826] GKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELNPPYLFSNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVL

[0827] DSDGSFFLYSKLTVDYWRWLEGNVFSCSVMHEALHNHYTQKSLSLSPG SEQ I D NO:125: Amino acid sequence of CD137 Fcabl7 (Fcab sequence with LALA mutation, without truncated hinge region) APEAAGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLN GKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELNPPYLFSNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVL DSDGSFFLYSKLTVYHWRWLEGNVFSCSVMHEALHNHYTQKSLSLSPG

[0828] SEQ ID NO:126 - AO motif (textual representation in Figure 27) PPAHGVTSAPDTRPAPGSTAPPA wherein S (18) and T (19) each comprise an O-linked polysaccharide chain glycosylation (Neu5Aca2,6GalNAcal). SEQ ID NO:127 - CO motif (textual representation in Figure 27) DTRPAPGST wherein S (8) and T (9) each comprise an O-linked GalNAcal monosaccharide glycosylation

[0829] 75

[0830] 008828600

Claims

1. CLAIMS1. An antigen-binding protein that binds to sialyl-Tn, wherein the antigen-binding protein comprises:CDRH1 of X1HAX2H (SEQ ID NO: 109), wherein Xi is alanine (A), aspartic acid (D) or glutamic acid (E), tryptophan (W) or tyrosine (Y); and X2 is isoleucine (I) or methionine (M);CDRH2 of YISPGX3X4DX5KYSQKFQG (SEQ ID NO: 110), wherein X3is asparagine (N) or glutamine (Q); X4is alanine (A), aspartic acid (D) or glutamic acid (E), glycine (G), lysine (K), asparagine (N), glutamine (Q) or tyrosine (Y); and X5is isoleucine (I), leucine (L), glutamine (Q), threonine (T) or valine (V);CDRH3 of SYYGX6(SEQ ID NO: 111), wherein Xs is glutamic acid (E) or histidine (H);CDRL1 of RASX7NIYSNLA (SEQ ID NO: 112), wherein X? is glutamic acid (E) or glutamine (Q);CDRL2 of AAXgXgLXioXu (SEQ ID NO: 113), wherein Xg is serine (S) or threonine (T); Xg is asparagine (N) or arginine (R); X10 is alanine (A) or glutamine (Q); and Xu is aspartic acid (D), leucine (L) or serine (S); andCDRL3 of QHFX12GTPYT (SEQ ID NO: 114), wherein X12 is phenylalanine (F) or tryptophan (W); and wherein the CDR sequences are defined according to Kabat.

2. The antigen-binding protein according to claim 1, wherein Xi is alanine.

3. The antigen-binding protein according to claim 1 or claim 2, wherein X2 is isoleucine (I).

4. The antigen-binding protein according to claim 3, wherein the antigen-binding protein comprises a CDRH1 according to SEQ ID NO:28.

5. The antigen-binding protein according to any one of claims 1 to 4, wherein X3 is asparagine (N).

6. The antigen-binding protein according to any one of claims 1 to 5, wherein X4is alanine (A) or glutamic acid(E).

7. The antigen-binding protein according to claim 6, wherein X4is glutamic acid (E).

8. The antigen-binding protein according to any one of claims 1 to 7, wherein X5is isoleucine (I), leucine (L), glutamine (Q) or valine (V).

9. The antigen-binding protein according to claim 8, wherein X5is valine (V).7600882860010. The antigen-binding protein according to claim 9, wherein the antigen-binding protein comprises a CDRH2 sequence according to SEQ ID NO:29.

11. The antigen-binding protein according to any one of claims 1 to 10, wherein the antigen-binding protein comprises a CDRH3 according to SEQ ID NO:30.

12. The antigen-binding protein according to any one of claims 1 to 11, wherein the antigen-binding protein comprises a CDRL1 according to SEQ ID NO:31.

13. The antigen-binding protein according to any one of claims 1 to 12, wherein Xg is serine (S).

14. The antigen-binding protein according to any one of claims 1 to 13, wherein Xg is arginine (R).

15. The antigen-binding protein according to any of claims 1 to 14, wherein Xio is glutamine (Q).

16. The antigen-binding protein according to any one of claim 1 to 15, wherein the antigen-binding protein comprises a CDRL2 according to SEQ ID NO: 32.

17. The antigen-binding protein according to any one of claims 1 to 16, wherein the antigen-binding protein comprises a CDRL3 according to SEQ ID NO:33.

18. The antigen-binding protein according to any of claims 1 to 17, wherein the antigen-binding protein comprises a CDRH1 according to SEQ ID NO:28, CDRH2 according to SEQ ID NO:29, CDRH3 according to SEQ ID NQ:30, CDRL1 according to SEQ ID NO:31, CDRL2 according to SEQ ID NO:32 and CDRL3 according to SEQ ID NO:33.

19. The antigen-binding protein according to any of claims 1 to 18, wherein the antigen-binding protein comprises a VH domain comprising a glutamine (Q) residue at position 94, wherein the numbering is defined according to Kabat.

20. The antigen-binding protein according to claim 19, wherein the VH domain further comprises an alanine (A) or valine (V) residue at position 67, an isoleucine (I) or leucine (L) residue at position 69, and / or an alanine (A) or threonine (T) residue at position 75, wherein the numbering is defined according to Kabat.

21. The antigen-binding protein according to claim 20, wherein the VH domain comprises a valine (V) residue at position 67, an isoleucine (I) residue at position 69, and a threonine (T) residue at position 75.7700882860022. The antigen-binding protein according to any one of claims 1 to 21, wherein the antigen-binding protein comprises a VH domain according to SEQ ID NO: 39 and / or a VL domain according to SEQ ID NO: 40.

23. The antigen-binding protein according to claim 22, wherein the antigen-binding protein has a VH domain according to SEQ ID NO: 39 and a VL domain according to SEQ ID NO: 40.

24. The antigen-binding protein according to claim 23, wherein the antigen-binding protein comprises a heavy chain sequence according to SEQ ID NO: 41 and / or a light chain sequence according to SEQ ID NO: 43.

25. The antigen-binding protein according to claim 24, wherein the antigen-binding protein comprises a heavy chain sequence according to SEQ ID NO: 41 and a light chain comprises the sequence according to SEQ ID NO: 43.

26. The antigen-binding protein according to any one of claims 1 to 25, wherein the antigen-binding protein comprises a second antigen-binding site that binds to a second antigen.

27. The antigen-binding protein according to claim 26, wherein the second antigen-binding site is located in a constant domain of the heavy chain.

28. The antigen-binding protein according to claim 27, wherein the constant domain is a CH3 domain.

29. The antigen-binding protein according to any of claims 26 to 28, wherein the second antigen-binding site binds to CD137.

30. The antigen-binding protein according to any one of claims 27 to 29, wherein the second antigen-binding site comprises a modified amino acid sequence in one or more of the AB, CD and EF structural loops of the constant domain.

31. The antigen-binding protein according to claim 30, wherein the constant domain is a CH3 domain, and wherein the AB structural loop comprises the sequence according to SEQ ID NO:81 and the EF structural loop comprises the sequence according to SEQ ID NO: 83.

32. The antigen-binding protein according to claim 31, wherein the CH3 domain comprises the sequence according to SEQ ID NO: 84.7800882860033. The antigen-binding protein according to any one of claims 1 to 32, wherein the antigen-binding protein comprises a CH2 domain which has been modified to reduce or abrogate binding of the CH2 domain to Fey receptors.

34. The antigen-binding protein according to any one of claims 1 to 33, wherein the antigen-binding protein does not bind to one or more Fey receptors.

35. The antigen-binding protein according to claim 33 or claim 34, wherein the CH2 domain has the sequence according to SEQ ID NO: 85.

36. The antigen-binding protein according to any one of claims 1 to 24 and 26 to 35, wherein the antigenbinding protein comprises a heavy chain comprising the sequence according to SEQ ID NO: 49 and a light chain comprising the sequence according to SEQ ID NO: 43.

37. The antigen-binding protein according to any one of claims 1 to 36, wherein the antigen-binding protein comprises a CH3 domain, and wherein the CH3 domain comprises a C-terminal lysine residue.

38. A nucleic acid molecule or set of nucleic acid molecules encoding an antigen-binding protein according to any one of claims 1 to 37.

39. A vector or set of vectors comprising the nucleic acid or set of nucleic acid molecules according to claim 38.

40. A recombinant host cell comprising the nucleic acid molecule or set of nucleic acid molecules according to claim 38, or the vector or set of vectors according to claim 39.

41. A method of producing the antigen-binding protein according to any one of claims 1 to 37 comprising culturing the recombinant host cell of claim 40 under conditions for production of the antigen-binding protein.

42. The method according to claim 41, further comprising isolating and / or purifying the antigen-binding protein.

43. The antigen-binding protein according to any one of claims 1 to 37 for use as a medicament.7900882860044. The antigen-binding protein according to any one of claims 1 to 37 for use in a method of treating cancer in an individual.

45. A method of treating cancer in an individual, wherein the method comprises administering to the individual a therapeutically effective amount of the antigen-binding protein according to any one of claims 1 to 37.

46. A pharmaceutical composition comprising the antigen-binding protein according to any one of claims 1 to 37 and a pharmaceutically acceptable excipient.

47. A pharmaceutical composition for use in the treatment of cancer in an individual, comprising the antigenbinding protein according to any one of claims 1 to 37.

48. Use of the antigen-binding protein according to any one of claims 1 to 37 for the manufacture of a medicament for treating cancer.80008828600

Images