Antigen-binding proteins

Abstract

The present disclosure relates to a ULBP6-binding protein that inhibits the interaction between ULBP6 and NKG2D, and further to a method of treating cancer with the ULBP6-binding protein. The present disclosure provides a ULBP6-binding protein comprising: a) any one or combination of CDRs selected from (i) CDRH1, CDRH2, CDRH3 from SEQ ID NOs: 16-19, and / or CDRL1, CDRL2, CDRL3 from SEQ ID NOs: 26-29, or (ii) a CDR variant of (i), said variant having one, two, or three amino acid modifications; or b) a VH region comprising a sequence at least 80% identical to any one of SEQ ID NOs: 21-24, and / or a VL region comprising a sequence at least 80% identical to any one of SEQ ID NOs: 31-34.

Description

[Technical field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to U.S. Provisional Application No. 63 / 336,336, filed April 29, 2022, which is incorporated herein by reference.

[0002] Reference to sequence listing An official copy of the sequence listing is submitted contemporaneously herewith as an ST26 format xml file with the filename "09402-014WO1.xml", creation date April 27, 2023, and size 171,756 bytes. The sequence listing submitted via EFS-Web is a part of the present specification and is incorporated herein by reference in its entirety.

[0003] Names of parties to the joint research agreement The claimed subject matter of this application was made by or on behalf of the parties to a joint research agreement, which was in effect on or before the date the claimed subject matter was made. The claimed subject matter was created as a result of activities undertaken within the scope of the joint research agreement. The parties to the joint research agreement include GlaxoSmithKline Intellectual Property (No. 3) Limited and 23andMe, Inc.

[0004] The joint research agreement was a written contract, grant, or cooperative agreement entered into by the parties for the performance of experimental, developmental, or research work in the field of the claimed subject matter.

[0005] FIELD OF THEINVENTION The present invention relates to the treatment of cancer. In particular, the present invention relates to ULBP6 binding proteins, including anti-ULBP6 antibodies, and their use in the treatment of cancer. [Background technology]

[0006] Effective treatment of cancer is an ongoing goal in the field of oncology. Immunotherapy, which enhances antitumor immune responses, is a powerful approach for cancer treatment. Several immuno-oncological antibodies are currently on the market, including monoclonal antibodies targeting PD-1 (e.g., OPDIVO / nivolumab and KEYTRUDA / pembrolizumab), PD-L1 (e.g., TECENTRIQ / atezolizumab, BAVENCIO / avelumab, IMFINZI / durvalumab), CTLA-4 (e.g., YERVOY / ipilimumab), and CD20 (e.g., RITUXAN / rituximab, ARZERRA / ofatumumab). Each of these antibody-based cancer immunotherapies may act via a different mechanism. The mechanisms are, for example, increasing the immune responsiveness of T cells, alleviating inhibitory checkpoints on tumor-specific T cells, inducing apoptosis, inducing enhanced antibody-dependent cellular cytotoxicity (ADCC), and / or inducing complement-dependent cytotoxicity (CDC).

[0007] Although recent advances in the field of immunotherapy have improved the treatment of cancer, many patients do not respond to conventional therapies. Thus, there remains a need for alternative or improved immunotherapeutic compositions and methods for more effectively treating cancer. Summary of the Invention

[0008] The present specification provides a ULBP6-binding protein, which comprises: a) any one or combination of CDRs selected from (i) CDRH1, CDRH2, CDRH3 from SEQ ID NOs: 16 to 19, and / or CDRL1, CDRL2, CDRL3 from SEQ ID NOs: 26 to 29, or (ii) a CDR variant of (i), said variant having 1, 2, or 3 amino acid modifications; or b) a VH region comprising a sequence at least 80% identical to any one of SEQ ID NOs: 21 to 24, and / or a VL region comprising a sequence at least 80% identical to any one of SEQ ID NOs: 31 to 34. The ULBP6 binding protein comprises a CDRH1 of SEQ ID NO: 36, 39, 42, or 45, a CDRH2 of SEQ ID NO: 37, 40, 43, or 46, a CDRH3 of SEQ ID NO: 38, 41, 44, or 47, a CDRL1 of SEQ ID NO: 51, 54, 57, or 60, a CDRL2 of SEQ ID NO: 52, 55, 58, or 61, and / or a CDRL3 of SEQ ID NO: 53, 56, 59, or 62, or a variant thereof.

[0009] Also provided herein is a ULBP6 binding protein, the ULBP6 binding protein comprising six CDRs selected from the group consisting of CDRH1 of SEQ ID NO: 36, CDRH2 of SEQ ID NO: 37, CDRH3 of SEQ ID NO: 38, CDRL1 of SEQ ID NO: 51, CDRL2 of SEQ ID NO: 52, and CDRL3 of SEQ ID NO: 53, CDRH1 of SEQ ID NO: 39, CDRH2 of SEQ ID NO: 40, CDRH3 of SEQ ID NO: 41, CDRL1 of SEQ ID NO: 54, CDRL2 of SEQ ID NO: 55, and CDRL3 of SEQ ID NO: 56. six CDRs selected from the group consisting of CDRH1 of SEQ ID NO:42, CDRH2 of SEQ ID NO:43, CDRH3 of SEQ ID NO:44, CDRL1 of SEQ ID NO:57, CDRL2 of SEQ ID NO:58, and CDRL3 of SEQ ID NO:59, or six CDRs selected from the group consisting of CDRH1 of SEQ ID NO:45, CDRH2 of SEQ ID NO:46, CDRH3 of SEQ ID NO:47, CDRL1 of SEQ ID NO:60, CDRL2 of SEQ ID NO:61, and CDRL3 of SEQ ID NO:62.

[0010] The present specification also provides a ULBP6-binding protein, which binds to ULBP6 and ULBP2 with a KD of 1 nM or less, and binds to ULBP5 with a KD of 1 nM or more.

[0011] Also provided herein is a ULBP6-binding protein that binds to human ULBP6 and competes for binding to said ULBP6 with a reference ULBP6-binding protein comprising a VH domain sequence of SEQ ID NO:21 and a VL domain sequence of SEQ ID NO:31, a reference ULBP6-binding protein comprising a VH domain sequence of SEQ ID NO:22 and a VL domain sequence of SEQ ID NO:32, a reference ULBP6-binding protein comprising a VH domain sequence of SEQ ID NO:16 and a VL domain sequence of SEQ ID NO:26, or a reference ULBP6-binding protein comprising a VH domain sequence of SEQ ID NO:17 and a VL domain sequence of SEQ ID NO:27.

[0012] Also provided herein is a nucleic acid sequence encoding one or both of the HC and LC of a ULBP6 binding protein provided herein. Also provided herein is an expression vector comprising a nucleic acid sequence provided herein. Also provided herein is a recombinant host cell comprising a nucleic acid sequence(s) or expression vector(s) provided herein. Also provided herein is a method for producing a ULBP6 binding protein, comprising culturing a recombinant host cell provided herein under conditions suitable for expression of said nucleic acid sequence(s) or vector(s), thereby producing a polypeptide comprising a ULBP6 binding protein provided herein. Also provided herein is a cell line engineered to express a ULBP6 binding protein provided herein. Also provided herein is a pharmaceutical composition comprising a ULBP6 binding protein as defined herein and a pharma- ceutically acceptable excipient.

[0013] Also provided herein is a method of treating cancer in a subject in need thereof, the method comprising administering to said subject a therapeutically effective amount of a ULBP6 binding protein or pharmaceutical composition as defined herein.Also provided herein is a ULBP6 binding protein and pharmaceutical composition as provided herein for use in treatment.Also provided herein is a ULBP6 binding protein and pharmaceutical composition as provided herein for use in treatment of cancer.Also provided herein is a ULBP6 antagonist for use in treatment of basal cell carcinoma.

[0014] Also provided herein are ULBP6 antibodies and fragments thereof, the ULBP6 antibodies and fragments thereof a) that bind to one or more residues of SEQ ID NO:125 and SEQ ID NO:127, b) that bind to one or more residues of SEQ ID NO:126 and SEQ ID NO:127, c) that bind to a conformational epitope of SEQ ID NO:125 and / or a conformational epitope of SEQ ID NO:127, d) that bind to a conformational epitope of SEQ ID NO:126 and / or a conformational epitope of SEQ ID NO:127, and e) that bind to one or more residues of SEQ ID NO:125, SEQ ID NO:126, or SEQ ID NO:127 by HDX-MS analysis of SEQ ID NO:124. f) binds to human ULBP6, resulting in a peptide derived from human ULBP6 SEQ ID NO:125, SEQ ID NO:126, or SEQ ID NO:127 that incorporates less deuterium than the corresponding peptide derived from unconjugated human ULBP6; g) binds to the NKG2D binding interface of human ULBP6; and / or h) interferes with the interaction of human ULBP6 to NKG2D by binding to one or more residues selected from E96, E154, Y184, and D189 of SEQ ID NO:124.

[0015] Also provided herein are ULBP6 antibodies and fragments thereof that a) bind to one or more residues of SEQ ID NO: 128, b) bind to a conformational epitope of SEQ ID NO: 128, c) protect one or more residues of SEQ ID NO: 128 from deuterium exchange in HDX-MS analysis of SEQ ID NO: 128, and / or bind to human ULBP6, resulting in a peptide derived from SEQ ID NO: 128 of human ULBP6 that is less susceptible to deuterium incorporation than the corresponding peptide derived from unconjugated human ULBP6. [Brief description of the drawings]

[0016] [Figure 1] Figure 1 shows a region association plot of the ULBP6 locus on chromosome 6 for basal cell carcinoma. The x-axis shows the location on chromosome 6 using GRCh38 / hg38 as the human reference genome construct. The left y-axis is the -log10(P-value) from a logistic regression test for association of genotype with the phenotype of interest. The plus sign (+) indicates a variant for which genotype was imputed. The circle sign (o) indicates a variant for which genotyping was used. The x sign (x) indicates an imputed coding variant, and the diamond sign (◇) indicates a genotype coding variant. The color reflects the linkage disequilibrium (calculated as r2) between the index SNP (shown in gray) and other nearby SNPs. The gray horizontal line represents the genome-wide significance threshold of 5x10-8. The credible set track displays the number and location of variants in the 99% credible set that are likely to contain the causal variant. The gene track displays genes within a locus with thick lines representing exons and thin lines representing introns.

[0017] [Diagram 2]FIG. 1 shows validation of the NKG2D detection assay with tool antibodies (top panel) and the ability of anti-ULBP6 antibodies (h7B3(E-) and h8E11(E-)) to reverse sULBP6_02-induced NKG2D blocking / internalization / downregulation in an NKG2D detection assay in stimulated PBMCs from a representative healthy donor compared to tool antibodies (1D4 and 6E1) and isotype controls (bottom panel).

[0018] [Figure 3A] FIG. 13 shows plots of validation of the tool antibodies used (top two panels).

[0019] [Figure 3B] FIG. 1 shows an overview of the properties of the tool antibodies used.

[0020] [Figure 3C] FIG. 1 shows the ability of anti-ULBP6 antibodies (h7B3, h8E11, and isotype control) to block the immunosuppression conferred by sULBP6_02 as measured by the level of IFNg secretion in a plate-binding MICA assay.

[0021] [Figure 3D] FIG. 1 shows the ability of anti-ULBP6 antibodies (h7B3, h1D4, h6E1, and isotype control) to block the immunosuppression conferred by sULBP6_02 as measured by the level of IFNg secretion in a plate-binding MICA assay.

[0022] [Figure 4] FIG. 1 shows the synergistic effect of NKG2D stimulation and activation of FcgRIIIa assay mediated by IFNg release from PBMCs upon activation of NKG2D alone, FcgRIIIa alone, or both simultaneously.

[0023] [Figure 5A]FIG. 1 shows a mixed cell assay (PBMC and tumor cells) in which stimulated PBMCs were plated on COV644 tumor cells in the presence of 50 nM sULBP6_02 and measured IFNg release to determine immune cell activation, and the ability of effector activity (E+) anti-ULBP6 to increase IFNg secretion in a dose-dependent manner by both neutralizing soluble ULBP6-mediated inhibition and promoting ADCC from surface ULBP6 expression on tumors.

[0024] [Figure 5B] FIG. 1 shows a mixed cell assay (PBMCs and tumor cells) in which stimulated PBMCs are plated on COV644 tumor cells in the presence of 50 nM sULBP6_02 and IFNg release is measured to determine immune cell activation; and a comparison of E+, E-, and isotype control antibodies for maximum potency of IFNg release from 200 nM antibody treatment.

[0025] [Figure 6] Figure 1 shows IFNg release in a mixed cell assay (PBMC and tumor cells) with PBMC stimulated on COV644 tumor cells with 50 nM soluble ULBP6. In the assay, increased IFNg release can be driven by blocking the inhibitory effect of sULBP6_02 or via ADCC-mediated activation from E+ molecules bound to the tumor cell surface. Data are shown for (upper panel) h7B3E+ and h7B3E- and (lower panel) h7B3E+, h7B3E-, h8E11E+, and h8E11E-. The top graph shows the titration of h7B3E- vs. E+, where the antibodies show similar EC50, but E+ shows higher maximum potency. Isotype controls are included at the highest dose of antibody for comparison. The bottom graph shows a comparison of four anti-ULBP6 antibodies (h7B3E- and E+ and h8E11E- and E+) in the same assay, where the antibodies all show similar EC50s, but the E+ antibody shows higher maximum potency. An isotype control is included at the highest dose of antibody for comparison.

[0026] [Figure 7A] Figure 10 demonstrates the ability of anti-ULBP6 antibodies (h7B3(E+) and h7B3(E-)) to rescue soluble ULBP6-induced depletion of Granzyme B in CD8 T cells (bottom two panels) in a mixed cell assay of stimulated PBMCs on COV644 tumor cells with 50 nM sULBP6_02, as measured using FACS from total PBMCs. In the assay, PBMC levels of Granzyme B and Ki-67 were measured by FACS following co-incubation.

[0027] [Figure 7B] FIG. 13 demonstrates the ability of anti-ULBP6 antibodies (h7B3(E+) and h7B3(E-)) to rescue the reduction of Ki-67 in CD8 T cells by soluble ULBP6 in a mixed cell assay of stimulated PBMCs on COV644 tumor cells with 50 nM sULBP6_02, as measured using FACS from total PBMCs. In the assay, PBMC levels of granzyme B and Ki-67 were measured by FACS following co-incubation.

[0028] [Figure 7C] FIG. 13 demonstrates the ability of anti-ULBP6 antibodies (h7B3(E+) and h7B3(E-)) to rescue soluble ULBP6-induced depletion of Granzyme B in NK cells (measured using FACS from total PBMCs) in a mixed cell assay of stimulated PBMCs on COV644 tumor cells with 50 nM sULBP6_02. In the assay, PBMC levels of Granzyme B and Ki-67 were measured by FACS following co-incubation.

[0029] [Figure 7D]FIG. 13 demonstrates the ability of anti-ULBP6 antibodies (h7B3(E+) and h7B3(E-)) to rescue the reduction of Ki-67 in NK cells by soluble ULBP6 in a mixed cell assay of stimulated PBMCs on COV644 tumor cells with 50 nM sULBP6_02, as measured using FACS from total PBMCs. In the assay, PBMC levels of granzyme B and Ki-67 were measured by FACS following co-incubation.

[0030] [Figure 8A] Figure 1 shows a mixed cell assay (PBMC and tumor cells) using pre-stimulated PBMC and COV644 tumor cells treated with 50 nM sULBP6_02, where the assay looks at IFNg release as a measure of immune cell activation (immune cell activation can be induced by blocking soluble ULBP6 by h7B3E- or E+, or by ADCC against h7B3E+). Dose titration of h7B3E+ vs. E- antibodies is shown, showing similar EC50s, but higher maximum potency for the E+ antibody. Maximum dose of antibody relative to isotype control is also shown.

[0031] [Figure 8B] Figure 1 shows a mixed cell assay (PBMC and tumor cells) using pre-stimulated PBMC and COV644 tumor cells treated with 50 nM sULBP6_02, where the assay looks at IFNg release as a measure of immune cell activation (immune cell activation can be induced by blocking soluble ULBP6 by h7B3E- or E+, or by ADCC against h7B3E+). Dose titration of h7B3E+ vs E++DE antibodies is shown, showing similar EC50s, but with increased maximum potency for E++DE. Maximum dose of antibody relative to isotype control is also shown.

[0032] [Figure 8C]Mixed cell assay (PBMC and tumor cells) using pre-stimulated PBMC and COV644 tumor cells treated with 50 nM sULBP6_02, where the assay looks at IFNg release as a measure of immune cell activation (immune cell activation can be induced by blocking soluble ULBP6 with h7B3E- or E+, or by ADCC against h7B3E+). Shown is the highest dose (100 nM) of all three h7B3 antibodies (E-, E+, E++DE) compared to isotype control.

[0033] [Figure 9A] Figure 1 shows strong responder donors in the INCUCYTE tumor killing assay setup. Reduction of GFP tumor cells with h7B3E- is shown, which is further reduced with h7B3E+ and even more so with h7B3E++ compared to isotype control. The right panel shows measurements of GFP at the final time point.

[0034] [Figure 9B] Figure 1 shows a moderately reactive donor in an INCUCYTE assay setup. Moderate GFP reduction with h7B3E- antibody is shown, which is further reduced with both E+ antibody ("h7B3E+") and E++DEh7B3 ("h7B3E++") antibody compared to the isotype control. The right panel shows GFP measurements at the final time point.

[0035] [Figure 9C] Figure 1 shows weak reacting donors in the INCUCYTE assay setup. Only h7B3E++DE ("h7B3E++") molecules show reduced GFP compared to isotype control. The right panel shows measurements of GFP at the final time point.

[0036] [Figure 10]Figure 1 shows IFNg levels from a mixed cell assay. In the mixed cell assay, stimulated PBMCs are incubated on COV644 cells in the presence of 50 nM sULBP6_02 and isotype or anti-ULBP6 antibodies (set doses of h7B3E++DE or h7B3E+Afu) for 24 hours. Both antibodies show similar EC50 curves and similar maximum potency of IFNg release.

[0037] [Figure 11A] FIG. 1 shows strong responding donors in an INCUCYTE assay setting and the reduction in GFP tumor cells observed in h7B3E+, which was further reduced to the same extent in h7B3E++DE ("h7B3E++") and h7B3E+Afu ("h7B3Afu").

[0038] [Figure 11B] FIG. 1 shows a moderately reactive donor in an INCUCYTE assay setup and the moderate GFP reduction observed with h7B3E+ antibodies, which was amplified for both h7B3E++DE ("h7B3E++") and h7B3E+Afu ("h7B3Afu") antibodies compared to the isotype control.

[0039] [Figure 12] Figure 1 shows the concentration-time profiles of h7B3(E+) and h8E11(E+) in cynomolgus monkeys after a 3 mg / kg intravenous dose administered by slow bolus. Data are plotted as mean and standard deviation (N=3M / group) on semi-log graphs using GraphPad PRISM 9.0. LLOQ=lower limit of quantification, and SD=standard deviation.

[0040] [Figure 13] Schematic representation of the crystal structure of h7B3Fab in complex with ULBP602. ULBP6 is colored black (towards the top of the structure), the h7B3Fab light chain is colored light grey (towards the left side of the structure), and its heavy chain is colored dark grey (towards the right side of the structure).

[0041] [Figure 14] FIG. 1 shows the structural epitope of h7B3 on huULBP602 as determined by the change in solvent accessible surface area of ​​ULBP602 residues between the free and bound states (Panel A). The epitope residues are colored black and shown as branches in the schematic representation of the structure of UBLP602. (Panel B) Surface view of ULBP6 with residues containing the epitope h7B3 epitope colored black. (Panel C) Structural binding site of NKG2D on ULBP6 as determined from the crystal structure of ULBP6 in complex with NKG2D (PDB ID: 4s0u). Residues interacting with NKG2D protomer A (shown as branches) are colored black and residues interacting with NKG2D protomer B are shown in grey in the schematic representation of the structure of UBLP602. (Panel D) Surface representation of ULBP6 with residues comprising the NKG2D protomer A binding side colored in black and residues forming the NKG2D protomer B binding site shown in grey.

[0042] [Figure 15] FIG. 1 shows ULBP6 phage binding to anti-HA antibodies.

[0043] [Figure 16] FIG. 1 shows the binding affinity of ULBP6 to hu7B3, hu8E1, hu1D4, hu6E1, and hu8E8.

[0044] [Figure 17] FIG. 1 shows binding affinity to the ULBP6 receptor.

[0045] [Figure 18] FIG. 1 shows the functional epitopes of 7B3 (Panels A and B), functional epitopes (Panels C and D), and functional epitopes of 6E11 (Panels E and F) identified herein.

[0046] [Figure 19]FIG. 131 peptides covering 87.7% of the ULBP6 construct sequence included in the evaluation of data from HDX experiments.

[0047] [Figure 20A] FIG. 1 shows the differential uptake rates for each detected peptide of amino acids 26 to 123 of ULBP6 from HDX experiments.

[0048] [Figure 20B] FIG. 1 shows the incorporation rate difference for each detected peptide of amino acids 114 to 226 of ULBP6 from HDX experiments.

[0049] [Figure 21] FIG. 1 shows the amino acids of ULBP6 (as underlined bold) that were determined by HDX experiments to be significantly protected by mAbs 1298, h7B3, h8E8, h1D4, h8E11, and h6E1.

[0050] [Figure 22] FIG. 1 shows residues of ULBP6 with significant protection by mAb1298 (panel A), h7B3(E+Afu) (panel B), h8E8(E+) (panel C), h1D4(E+) (panel D), h8E11(E+) (panel E), and h6E1(E+) (panel F) on the crystal structure of the NKG2D-ULBP6 complex (determined by HDX and shown in black). DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0051] As used herein, "ULBP6" refers to any UL16 binding protein 6. Alternative names for ULBP6 include RAET1L, retinoic acid early transcript 1L, retinoic acid early transcript 1L protein, and UL16 binding protein 6. ULBP6 is encoded by the RAET1L gene and is a molecule of the UL16 binding protein family. ULBP6 specifically binds to the activating receptor natural killer group 2D receptor (NKG2D). Like other NKG2D ligands, ULBP6 functions as a stress-induced modulator of the NKG2D pathway. ULBP6 can be expressed as a cell surface membrane-bound form. ULBP6 can be a soluble molecule, can be from the cell surface, or can be released from the cell surface. As used herein, "sULBP6" refers to non-surface bound, soluble, or released ULBP6. As used herein, "human ULBP6" refers to cell surface membrane-bound human ULBP6 or human sULBP6. Human ULBP6 is highly polymorphic, with two haplotypes (ULBP6_01 and ULBP6_02) being the most common haplotypes found in the population (Antoun et al., 2010, Hum Immunol, 71:610-20; Eagle et al., 2009, Eur J Immunol, 39:3207-3216). The amino acid sequences of human sULBP6 haplotypes ULBP6_01 and ULBP6_02 are shown in Table 6.

[0052] As used herein, "NKG2D" refers to the natural killer group 2D receptor of ULBP6 and other NKG2DL. NKG2D is an activating receptor encoded by the KLRK1 (killer cell lectin-like receptor K1) gene and expressed by natural killer (NK) cells and T cells. It plays an important role in anti-pathogen and anti-tumor responses, because binding of NKG2D leads to cell activation, cytokine production, and cytotoxic activity. In some examples, the NKG2D is human NKG2D (hNKG2D).

[0053] As used herein, "NKG2DL" and "NKG2D ligand" are used interchangeably and refer to protein ligands of NKG2D. Human NKG2DL (hNKG2DL) includes MICA, MICB, ULBP1 (RAET1), ULBP2 (RAET1H), ULBP3 (RAET1N), ULBP4 (RAET1E), ULBP5 (RAET1G), and ULBP6 (RAET1L). Some NKG2DLs may be soluble molecules or may be released from the cell surface. As used herein, "sNKG2DL" refers to non-surface bound, soluble, or released NKG2DL.

[0054] ULBP6 and other NKG2DLs bind to the activating receptor NKG2D expressed on NK cells and T cells (e.g., CD8 T cells). The expression of these NKG2D ligands is normally low under homeostatic conditions, but can be highly induced under conditions of cellular stress (e.g., by pathogen-infected or malignant cells). When expressed on the cell surface of stressed cells, ULBP6 can bind to NKG2D, thereby enhancing the immune surveillance of NK cell / T cell activation. When ULBP6 is in a released / soluble form, sULBP6 can act to block the interaction between NKG2D and other NKG2DLs, and thus act immunosuppressively by inhibiting NKG2D-mediated activation of immune cells. Therefore, blocking the interaction of sULBP6 with NKG2D using ULBP6-binding proteins is expected to increase NKG2D-mediated activation of immune cells. ULBP6-binding protein-induced NKG2D-mediated immune cell activation may result in the treatment of diseases such as cancer.

[0055] As used herein, the term "ULBP6-binding protein" refers to antibodies and other protein constructs (e.g., domains) that can bind to ULBP6. In some examples, the ULBP6 is human ULBP6. The terms "ULBP6-binding protein" and "antigen-binding protein" are used interchangeably herein. This does not include the natural cognate ligand or receptor (i.e., NKG2D).

[0056] The term "antibody" is used herein in the broadest sense to refer to molecules having immunoglobulin-like domains (e.g., IgG, IgM, IgA, IgD, or IgE) and includes monoclonal antibodies, recombinant antibodies, polyclonal antibodies, chimeric antibodies, human antibodies, humanized antibodies, multispecific antibodies such as bispecific antibodies, and heteroconjugate antibodies, single variable domains (such as, but not limited to, domain antibodies (DABs)), Fab, F(ab')2, Fv, disulfide-linked Fv, single chain Fv, disulfide-linked scFv, diabodies, TANDABS, and the like, as well as modifications or antigen-binding fragments of any of the above.

[0057] The terms complete, whole, or intact antibody, used interchangeably herein, refer to a heterotetrameric glycoprotein with an approximate molecular weight of 150,000 daltons. Intact antibodies are composed of two identical heavy chains (HC) and two identical light chains (LC) linked by covalent disulfide bonds. This H2L2 structure folds to form three functional domains, including two antigen-binding fragments known as "Fab" fragments and an "Fc" crystallizable fragment. The Fab fragment is composed of an amino-terminal variable domain, a variable heavy chain (VH) or a variable light chain (VL), and a carboxyl-terminal constant domain, CH1 (heavy chain) and CL (light chain). The Fc fragment is composed of two domains formed by dimerization of paired CH2 and CH3 regions. Fc can trigger effector functions by binding to receptors on immune cells or by binding to C1q, the first component of the classical complement pathway. The five classes of antibodies, IgM, IgA, IgG, IgE, and IgD, are defined by different heavy chain amino acid sequences, called μ, α, γ, ε, and δ, respectively, and each heavy chain can pair with either a K or a λ light chain. The majority of antibodies in serum belong to the IgG class, and there are four isotypes of human IgG (IgG1, IgG2, IgG3, and IgG4), whose sequences differ mainly in their hinge region.

[0058] Fully human antibodies can be obtained using a variety of methods, for example, using yeast-based libraries or transgenic animals (e.g., mice) that can produce a repertoire of human antibodies. Yeast-displayed human antibodies on the surface that bind to the antigen of interest can be selected using FACS (fluorescence-activated cell sorting)-based methods or by capture on beads using labeled antigen. Transgenic animals modified to express human immunoglobulin genes can be immunized with the antigen of interest and antigen-specific human antibodies isolated using B-cell sorting methods. Human antibodies generated using these techniques can then be characterized for desired properties such as affinity, developability, and selectivity.

[0059] Alternative antibody formats include alternative scaffolds in which one or more CDRs of an antigen binding protein may be placed onto a suitable non-immunoglobulin protein scaffold or framework, such as an affibody, an SpA scaffold, an LDL receptor class A domain, an avimer, or an EGF domain.

[0060] "Humanized antibody" refers to a type of engineered antibody, such that the antibody has CDRs derived from a non-human donor immunoglobulin, with the remaining immunoglobulin-derived portion of the molecule derived from one or more human immunoglobulin(s). Additionally, framework support residues may be modified to maintain binding affinity. Suitable human acceptor antibodies may be selected from the usual databases, e.g., the KABAT database, the Los Alamos database, and the Swiss Protein database, by homology to the nucleotide and amino acid sequences of the donor antibody. The acceptor antibody may be heterologous to the donor antibody, and in some embodiments may be a human antibody. The donor antibody may contribute the amino acid sequence of its variable regions, CDRs, or other functional fragments or analogs thereof to the first immunoglobulin partner. The donor may contribute antigen specificity to the altered immunoglobulin coding region and the resulting expressed altered antibody. A human antibody characterized by homology to the framework regions (amino acid basis) of the donor antibody may be suitable to provide the heavy chain constant region and / or the heavy chain variable framework region for inserting the donor CDRs. A suitable acceptor antibody capable of donating the heavy and / or light chain constant or variable framework regions can be selected in a similar manner. It should be noted that the heavy and light chains of the acceptor antibody do not have to be derived from the same acceptor antibody.

[0061] The ULBP6 binding protein described herein may be an antibody or an antigen-binding fragment thereof. The ULBP6 binding protein may be a humanized antibody or an antigen-binding fragment thereof. The ULBP6 binding protein may comprise a humanized VH region or a humanized heavy chain (HC) sequence, and / or a humanized VL region or a humanized light chain (LC) sequence. The humanized HC may be humanized using human germline sequences IGHV3-48, IGHV7-4-1, or IGHV1-2. The humanized HC may be humanized using human germline sequences IGHV3-48. The humanized LC may be humanized using human germline sequences IGKV1-33, IGKV3-11, or IGKV1-16. The humanized LC may be humanized using human germline sequences IGKV1-33. The humanized LC may be humanized using human germline sequences IGKV3-11.

[0062] The term "single variable domain" refers to a folded polypeptide domain comprising sequences characteristic of antibody variable domains. It thus includes complete antibody variable domains such as VH, VHH and VL, and modified antibody variable domains, for example in which one or more loops have been replaced with sequences not characteristic of antibody variable domains, or includes truncated antibody variable domains or antibody variable domains comprising N-terminal or C-terminal extensions, as well as folded fragments of variable domains that retain at least the binding activity and specificity of the full-length domain. A single variable domain as defined herein is capable of binding an antigen or epitope independently of the different variable regions or domains. A "domain antibody" or "DAB" can be considered the same as a human "single variable domain". The single variable domains may be human single variable domains, but also include single variable domains from other species, such as those from rodents (e.g. those disclosed in WO00 / 29004), nurse sharks and camelid VHHs, which are immunoglobulin single variable domain polypeptides derived from species such as camel, llama, alpaca, dromedary and guanaco. These species produce heavy chain-only antibodies that are naturally devoid of light chains. Such VHH domains can be humanized according to standard techniques available in the art, and such domains are considered "single variable domains".

[0063] Antigen-binding fragments can be produced by the arrangement of one or more CDRs on a non-antibody protein scaffold.As used herein, "protein scaffold" includes, but is not limited to, immunoglobulin (Ig) scaffolds, such as IgG scaffolds.Such scaffolds can be four-chain or two-chain antibodies, can include only the Fc region of an antibody, or can include one or more constant regions from an antibody, which constant regions can be of human or primate origin, or can be artificial chimeras of human constant regions and primate constant regions.

[0064] The protein scaffold can be an Ig scaffold, for example an IgG scaffold or an IgA scaffold. The IgG scaffold can include some or all of the domains of an intact antibody (i.e., CH1, CH2, CH3, VH, VL). The antigen-binding protein can include an IgG scaffold selected from IgG1, IgG2, IgG3, IgG4, or IgG4PE. For example, the scaffold can be an IgG1. The scaffold can consist of, include, or be a portion of the Fc region of an antibody.

[0065] The protein scaffold may be a derivative of a scaffold selected from the group consisting of CTLA-4, lipocalin, protein A derived molecules (e.g., the Z domain of protein A) (affibody, SpA), A domain (Avimer / Maxibody), heat shock proteins such as GroEl and GroES, transferrin (trans-bodies), ankyrin repeat proteins (DARpins), peptide aptamers, C-type lectin domains (tetranectin), human g-crystallin and human ubiquitin (affilin), PDZ domains, scorpion venom Kunitz-type domains of human protease inhibitors, and fibronectin / adnectin (which have been subjected to protein engineering to obtain binding to antigens other than the natural ligand, such as ULBP6).

[0066] The term multispecific antigen-binding protein refers to an antigen-binding protein that comprises at least two different antigen-binding sites.Each of such antigen-binding sites can bind to different epitopes that may be present on the same antigen or different antigens.Multispecific antigen-binding proteins can have specificity for multiple antigens, for example, two antigens, or three antigens, or four antigens.

[0067] Bispecifics can generally be classified as having symmetric or asymmetric structures. Bispecifics can have an Fc or can be fragment-based (lacking an Fc). Fragment-based bispecifics combine multiple antigen-binding antibody fragments in one molecule without an Fc region, e.g., Fab-scFv, Fab-scFv2, orthogonal Fab-Fab, Fab-Fv, tandem scFc (e.g., BiTE and BiKE molecules), diabodies, DART, TandAb, sc diabodies, tandem dAbs, etc.

[0068] Symmetric formats combine multiple binding specificities in a single polypeptide chain or single HL pair, including Fc fusion proteins in fragment-based formats, whereby antibody fragments are fused to a conventional antibody molecule. Examples of symmetric formats include DVD-Ig, TVD-Ig, CODV-Ig, (scFv)4-Fc, IgG-(scFv)2, tetravalent DART-Fc, F(ab)4 cross-Mab, IgG-HC-scFv, IgG-LC-scFv, mAb-dAb, etc.

[0069] Asymmetric formats preserve as closely as possible the native structure of natural antibodies by enforcing correct H-chain pairing and / or promoting H-chain heterodimerization during co-expression of three (when using common heavy or light chains) or four polypeptide chains, such as Triomab, Asymmetric Reengineered Technology Immunoglobulin (ART-Ig), CrossMab, Biclonics common light chain, ZW1 common light chain, DuoBody and Knob-into-hole (KiH), DuetMab, κλ body, Xmab, YBODY, HET-mAb, HET-Fab, DART-Fc, SEEDbody, mouse / rat chimeric IgG.

[0070] Bispecific formats may also include antibodies fused to non-Ig scaffolds such as affimab, finitomab, zybodies, and anticalin-IgG fusions, ImmTAC.

[0071] The term "chimeric antigen receptor" ("CAR") as used herein refers to an engineered receptor consisting of an extracellular antigen-binding domain (usually derived from a monoclonal antibody or a fragment thereof, e.g., the VH and VL domains in the form of an scFv), optionally a spacer region, a transmembrane region, and one or more intracellular effector domains. CARs are also called chimeric T cell receptors or chimeric immune receptors (CIRs). CARs are genetically transferred into hematopoietic cells, such as T cells, to redirect the specificity of the T cells to a desired cell surface antigen, resulting in CAR-T therapeutics.

[0072] As used herein, the term "spacer region" refers to an oligopeptide or polypeptide that functions to link the transmembrane domain to the target binding domain. This region may also be called the "hinge region" or "stalk region". The size of the spacer may vary depending on the location of the target epitope to maintain a certain distance (e.g., 14 nm) during CAR:target binding.

[0073] The term "transmembrane domain" as used herein refers to the portion of the CAR molecule that traverses the cell membrane.

[0074] As used herein, the term "intracellular effector domain" (also referred to as "signal transduction domain") refers to a domain in a CAR that is involved in intracellular signal transduction after the antigen binding domain binds to a target. The intracellular effector domain is involved in activating at least one of the normal effector functions of the immune cell in which the CAR is expressed. For example, the effector function of a T cell can be a cytolytic activity or a helper activity, including secretion of cytokines.

[0075] As will be apparent to one of skill in the art, the VH and / or VL domains disclosed herein can be incorporated into a CAR-T therapeutic, for example in the form of an scFv.

[0076] The antigen binding proteins described herein may exhibit cross-reactivity between human ULBP6 and ULBP6 from another species (e.g., cynomolgus ULBP2 / 6). It is noted that cynomolgus ULBP2 / 6 is so named because it is unclear whether the single cynomolgus orthologue is evolutionarily related to human ULBP2 or ULBP6. Thus, the cynomolgus orthologue is referred to herein as cynomolgus ULBP2 / 6 or cULBP2 / 6. The antigen binding proteins described herein may specifically bind human ULBP6 and cynomolgus ULBP2 / 6. This is particularly useful, as drug development typically requires testing lead drug candidates in non-human primate systems before testing the drug in humans. Providing a drug that can bind to human and monkey species allows for testing results in these systems and side-by-side comparison of data using the same drug. This avoids the cumbersome need to find a drug that acts on cynomolgus monkey ULBP6 and another that acts on human ULBP6, and also avoids the need to use non-identical drugs to compare results in humans and cynomolgus monkeys. Cross-reactivity between other species used in disease models, such as dogs or monkeys, especially monkeys, is also envisioned. In some cases, the binding affinity of the antigen-binding protein for at least cynomolgus monkey ULBP2 / 6 differs from that for human ULBP6 by no more than 2, 5, 10, 50, or 100-fold.

[0077] Affinity, also referred to as "binding affinity", is the binding strength at a single interaction site (i.e., the site of one molecule, e.g., an antigen-binding protein of the present invention) to another molecule (e.g., ULBP6) at a single binding site. The binding affinity of an antigen-binding protein to its target can be determined by equilibrium methods (e.g., enzyme-linked immunosorbent assay (ELISA) or radioimmunoassay (RIA)), or kinetics (e.g., BIACORE analysis). For example, the BIACORE method described in Example 7 can be used to measure binding affinity.

[0078] The equilibrium dissociation constant (KD) of the antigen-binding protein-ULBP6 interaction may be 100 nM or less, 10 nM or less, 5 nM or less, 2 nM or less, or 1 nM or less. Alternatively, the KD may be 5-10 nM, or 1-2 nM. The KD may be 1 pM-500 pM, or 500 pM-1 nM. The KD may be 100 pM-500 pM, or 100 pM-1 nM. The KD may be 100 pM-2 nM, or 100 pM-5 nM. As will be apparent to those skilled in the art, the smaller the KD value, the stronger the binding. The reciprocal of the KD (i.e., 1 / KD) is expressed in units of M. -1 As will be appreciated by those of skill in the art, the higher the KA value, the stronger the binding.

[0079] Human ULBP6 shows high sequence similarity with human ULBP2 and low level sequence similarity with ULBP5. The ULBP6 binding proteins described herein may bind to human ULBP6 and ULBP2 with a KD of 1 nM or less. Alternatively, the ULBP6 binding proteins described herein may bind to human ULBP6 with a KD of 500 pM or less and ULBP2 with a KD of 500 pM or more. Alternatively, the ULBP6 binding proteins described herein may bind to human ULBP6 with a KD of 300 pM or less and ULBP2 with a KD of 500 pM or more. The ULBP6 binding proteins described herein may bind to human ULBP6 with a KD of 1 nM or less and ULBP5 with a KD of 1 nM or more. Alternatively, the ULBP6 binding proteins may bind to ULBP6 with a KD of 500 pM or less and ULBP5 with a KD of 1 nM or more. Alternatively, the ULBP6 binding protein may bind to ULBP6 with a KD of 400 pM or less and to ULBP5 with a KD of 1 nM or more. Alternatively, the ULBP6 binding protein may bind to ULBP6 with a KD of 300 pM or less and to ULBP5 with a KD of 1 nM or more. Alternatively, the ULBP6 binding protein may bind to ULBP6 with a KD of 300 pM or less and to ULBP5 with a KD of 2 nM or more.

[0080] The dissociation rate constant (kd) or "off rate" represents the stability of the antigen binding protein-ULBP6 complex, i.e., the fraction of the complex that disintegrates per second. For example, 0.01 s -1 The kd corresponds to 1% of the complex disintegrating per second. The dissociation rate constant (kd) of the antigen-binding protein-ULBP6 interaction is 1×10 -3 s -1 Below, 1×10 -4 s -1 Below, 1×10 -5 s -1 or less, or 1×10 -6 s -1 Alternatively, kd can be less than or equal to 1×10 -5 s -1 ~1×10 -4 s -1 , or 1 × 10 -4 s -1 ~1×10 -3 s -1 Alternatively, kd can be 1×10 -3 s -1 ~1×10 -2 s -1 It could be. The association rate constant (ka) or "on rate" represents the rate of antigen binding protein-ULBP6 complex formation. The ka of the antigen binding protein-ULBP6 interaction is approximately 1.5x10 6 M -1 s -1 Alternatively, ka can be 1×10 6 M -1 s -1 ~1×10 5 M -1 s -1 Alternatively, ka can be 1×10 6 M -1 s -1 ~5x10 6 M -1 s -1 or 1×10 6 M -1 s -1 ~2x10 6 M -1 s -1 It could be.

[0081] By "isolated" is intended that a molecule, such as an antigen-binding protein or a nucleic acid, is removed from its previous environment. For example, the molecule can be purified to separate it from the materials in which it would normally occur in nature. For example, the mass of the molecule in a sample can be 95% of the total mass.

[0082] The ULBP6 binding proteins described herein may be encoded by nucleic acids, which may be isolated. The production of ULBP6 binding proteins, such as antibodies, may be accomplished in a cell or in vivo by delivering exogenous isolated nucleic acids encoding the heavy and light chains of the antibody. One or more nucleic acids encoding the heavy and light chains of the ULBP6 binding protein may be delivered to a subject in need. The heavy and light chains of the antibody may be delivered by the same or separate nucleic acids. The nucleic acid may be DNA or RNA. The nucleic acid may be mRNA. The nucleic acid encoding the ULBP6 binding protein may be modified or unmodified. The nucleic acid encoding the ULBP6 binding protein may include at least one chemical modification. The nucleic acid encoding the ULBP6 binding protein may be provided to a subject naked (i.e., without an encapsulating particle) or packaged (i.e., encapsulated in a liposome or polymer-based vehicle).

[0083] Also provided herein is a method for producing a ULBP6 binding protein in a cell, tissue, or organism, comprising contacting the cell, tissue, or organism with a composition comprising an isolated nucleic acid that comprises at least one chemical modification and that encodes the ULBP6 binding protein.Also provided herein is a method for producing a ULBP6 binding protein in a cell, tissue, or organism, comprising contacting the cell, tissue, or organism with a composition comprising a polynucleotide that comprises at least one chemical modification and that encodes the ULBP6 binding protein.

[0084] The term "expression vector" as used herein means an isolated nucleic acid that can be used to introduce a nucleic acid of interest into a cell, such as a eukaryotic or prokaryotic cell, or into a cell-free expression system, in which the nucleic acid sequence of interest is expressed as a peptide chain, such as a protein. Such expression vectors can be, for example, cosmids, plasmids, viral sequences, transposons, and linear nucleic acids that contain the nucleic acid of interest. When the expression vector is introduced into a cell or into a cell-free expression system (e.g., a reticulocyte lysate), the protein encoded by the nucleic acid of interest is produced by the transcription / translation machinery. Expression vectors within the scope of the present disclosure can provide the necessary elements for eukaryotic or prokaryotic expression, and include CMV promoter-driven vectors, e.g., pcDNA3.1, pCEP4, and derivatives thereof, viral promoter-driven vectors, such as baculovirus expression vectors, Drosophila expression vectors, and expression vectors driven by mammalian gene promoters, such as human Ig gene promoters. Other examples include prokaryotic expression vectors, e.g., T7 promoter-driven vectors such as pET41, lactose promoter-driven vectors, and arabinose gene promoter-driven vectors. Those skilled in the art will recognize many other suitable expression vectors and expression systems.

[0085] The term "recombinant host cell" as used herein refers to a cell that contains a nucleic acid sequence of interest that has been isolated and subsequently introduced into the cell. For example, the nucleic acid sequence of interest may be in an expression vector and the cell may be a prokaryotic or eukaryotic cell. Exemplary eukaryotic cells include, but are not limited to, mammalian cells, such as COS-1, COS-7, HEK293, BHK21, CHO, BSC-1, HepG2, 653, SP2 / 0, NS0, 293, HeLa, myeloma cells, lymphoma cells, or any derivatives thereof. The eukaryotic cell may be a HEK293, NS0, SP2 / 0, or CHO cell. E. coli is an exemplary prokaryotic cell. Recombinant cells according to the present disclosure may be generated by transfection, cell fusion, immortalization, or other procedures well known in the art. The nucleic acid of interest, such as an expression vector, transfected into the cell may be extrachromosomal or stably integrated into the chromosome of the cell.

[0086] "CDR" is defined as the amino acid sequence of the complementarity determining region of an antigen-binding protein. The CDRs may be the same as or overlap with the hypervariable regions (HVRs) of immunoglobulin heavy and light chains, depending on the method(s) used to determine the CDRs and HDRs. There are three heavy and three light chain CDRs (or CDR regions) in the variable portion of an immunoglobulin. Thus, as used herein, "CDR" generally refers to all three heavy chain CDRs, all three light chain CDRs, all heavy and light chain CDRs, or at least two CDRs.

[0087] As used herein, a hypervariable region can include extended or alternative hypervariable regions such as: L 24–36 or 24–34 (L1), 46–56 or 50–56 (L2), and 89–97 or 89–96 (L3) in domains, and V H26-35 or 30-35 (H1), 50-61, 50-65, or 49-65 (H2), and 93-102, 94-102, or 95-102 (H3) in domains. The variable domain residues are numbered for each of these definitions according to Kabat et al. (supra).

[0088] Throughout the specification, amino acid residues within variable domain sequences and variable domain regions within full-length antigen-binding sequences, e.g., within antibody heavy chain sequences or antibody light chain sequences, are numbered according to the Kabat numbering convention. Similarly, the terms "CDR", "CDRL1", "CDRL2", "CDRL3", "CDRH1", "CDRH2", "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., USDepartment of Health and Human Services, National Institutes of Health (1987).

[0089] As will be appreciated by those skilled in the art, there are alternative numbering conventions for amino acid residues in variable domain sequences and full-length antibody sequences. Alternative numbering conventions for CDR sequences also exist, as shown, for example, in Chothia et al. (1989) Nature 342: 877-883. The structure and protein folding of antigen-binding proteins may mean that other residues are considered part of the CDR sequences, and will be understood by those skilled in the art.

[0090] Other numbering conventions for CDR sequences available to those skilled in the art include the "AbM" convention (University of Bas) and the "contact" convention (University College Lonon). The CDR regions of SEQ ID NOs: 16-19 and 26-29 can be defined by any numbering convention, such as the Kabat, Chothia, AbM, and contact conventions.

[0091] Table 1 below shows one definition using the respective numbering rules for each CDR or binding unit. In Table 1, the Kabat numbering scheme is used to number the variable domain amino acid sequence. Please note that some of the definitions of CDRs may vary depending on the individual publication used. [Table 1]

[0092] Provided herein is a ULBP6 binding protein, which comprises any one or combination of CDRs selected from a) CDRH1 of SEQ ID NO: 16 or 21, CDRH2 of SEQ ID NO: 16 or 21, CDRH3 of SEQ ID NO: 16 or 21, CDRL1 of SEQ ID NO: 26 or 31, CDRL2 of SEQ ID NO: 26 or 31, and CDRL3 of SEQ ID NO: 26 or 31, or b) any one or combination of CDRs selected from CDRH1 of SEQ ID NO: 17 or 22, CDRH2 of SEQ ID NO: 17 or 22, CDRH3 of SEQ ID NO: 17 or 22, CDRL1 of SEQ ID NO: 27 or 32, CDRL2 of SEQ ID NO: 27 or 32, and CDRL3 of SEQ ID NO: 27 or 32. or c) any one or combination of CDRs selected from CDRH1 of SEQ ID NO: 18 or 23, CDRH2 of SEQ ID NO: 18 or 23, CDRH3 of SEQ ID NO: 18 or 23, CDRL1 of SEQ ID NO: 28 or 33, CDRL2 of SEQ ID NO: 28 or 33, and CDRL3 of SEQ ID NO: 28 or 33; or d) any one or combination of CDRs selected from CDRH1 of SEQ ID NO: 19 or 24, CDRH2 of SEQ ID NO: 19 or 24, CDRH3 of SEQ ID NO: 19 or 24, CDRL1 of SEQ ID NO: 29 or 34, CDRL2 of SEQ ID NO: 29 or 34, and CDRL3 of SEQ ID NO: 29 or 34.

[0093] Alternatively, the ULBP6 binding protein described herein comprises all six CDRs selected from a) CDRH1 of SEQ ID NO: 36, CDRH2 of SEQ ID NO: 37, CDRH3 of SEQ ID NO: 38, CDRL1 of SEQ ID NO: 51, CDRL2 of SEQ ID NO: 52, and CDRL3 of SEQ ID NO: 53; b) CDRH1 of SEQ ID NO: 39, CDRH2 of SEQ ID NO: 40, CDRH3 of SEQ ID NO: 41, CDRL1 of SEQ ID NO: 54, CDRL2 of SEQ ID NO: 55, and CDRL3 of SEQ ID NO: 56. It may comprise all six CDRs, c) all six CDRs selected from CDRH1 of SEQ ID NO: 42, CDRH2 of SEQ ID NO: 43, CDRH3 of SEQ ID NO: 44, CDRL1 of SEQ ID NO: 57, CDRL2 of SEQ ID NO: 58, and CDRL3 of SEQ ID NO: 59, or d) all six CDRs selected from CDRH1 of SEQ ID NO: 45, CDRH2 of SEQ ID NO: 46, CDRH3 of SEQ ID NO: 47, CDRL1 of SEQ ID NO: 60, CDRL2 of SEQ ID NO: 61, and CDRL3 of SEQ ID NO: 62.

[0094] The CDRs may be modified by substitution, deletion, or addition of at least one amino acid, where the mutant ULBP6 binding protein substantially retains one or more biological properties of the unmodified binding protein, which may be a biological property set forth herein, such as inhibition of binding of soluble ULBP6 to NKG2D.

[0095] As is clear, each of CDR H1, H2, H3, L1, L2, L3 can be modified alone or in combination with any other CDR, in any permutation or combination.CDR can be modified by substitution, deletion, or addition of up to 3 amino acids, for example, 1 or 2 amino acids, for example, 1 amino acid.Each modification of CDR, VH, VL, or other proteins provided herein can be conservative substitution.Usually, modification is substitution, particularly conservative substitution, for example as shown in Table 2 below.

[0096] [Table 2]

[0097] For example, in a variant CDR, the adjacent residues which comprise the CDR may be substituted with conservative amino acid residues as part of the alternative definition(s) (eg, Kabat or Chothia).

[0098] Such antigen binding proteins comprising the above-mentioned mutant CDRs may be referred to herein as "functional CDR mutants".

[0099] The ULBP6 binding proteins described herein comprise any one or combination of CDRs selected from (i) CDRH1, CDRH2, CDRH3 from SEQ ID NO: 16, and / or CDRL1, CDRL2, CDRL3 from SEQ ID NO: 26, or (ii) a CDR variant of (i), including a variant having one, two or three amino acid modifications, or a VH region comprising a sequence at least 80% identical to the sequence of SEQ ID NO: 21, and / or a VL region comprising a sequence at least 80% identical to the sequence of SEQ ID NO: 31.

[0100] Alternatively, the ULBP6 binding protein described herein comprises any one or combination of CDRs selected from (i) CDRH1, CDRH2, CDRH3 from SEQ ID NO: 17, and / or CDRL1, CDRL2, CDRL3 from SEQ ID NO: 27, or (ii) a CDR variant of (i), including a variant having one, two or three amino acid modifications, or a VH region comprising a sequence at least 80% identical to the sequence of SEQ ID NO: 22, and / or a VL region comprising a sequence at least 80% identical to the sequence of SEQ ID NO: 32.

[0101] Alternatively, the ULBP6 binding protein described herein comprises any one or combination of CDRs selected from (i) CDRH1, CDRH2, CDRH3 from SEQ ID NO: 18, and / or CDRL1, CDRL2, CDRL3 from SEQ ID NO: 28, or (ii) a CDR variant of (i), including a variant having one, two or three amino acid modifications, or a VH region comprising a sequence at least 80% identical to the sequence of SEQ ID NO: 23, and / or a VL region comprising a sequence at least 80% identical to the sequence of SEQ ID NO: 33.

[0102] Alternatively, the ULBP6 binding protein described herein comprises any one or combination of CDRs selected from (i) CDRH1, CDRH2, CDRH3 from SEQ ID NO: 19, and / or CDRL1, CDRL2, CDRL3 from SEQ ID NO: 29, or (ii) a CDR variant of (i), including a variant having one, two or three amino acid modifications, or a VH region comprising a sequence at least 80% identical to the sequence of SEQ ID NO: 24, and / or a VL region comprising a sequence at least 80% identical to the sequence of SEQ ID NO: 34.

[0103] The ULBP6 binding proteins described herein may comprise any one or combination of CDRs selected from CDRH1 of SEQ ID NO: 36, CDRH2 of SEQ ID NO: 37, and / or CDRH3 of SEQ ID NO: 38, CDRL1 of SEQ ID NO: 51, CDRL2 of SEQ ID NO: 52, and / or CDRL3 of SEQ ID NO: 53. Alternatively, the ULBP6 binding proteins described herein may comprise any one or combination of CDRs selected from CDRH1 of SEQ ID NO: 39, CDRH2 of SEQ ID NO: 40, and / or CDRH3 of SEQ ID NO: 41, CDRL1 of SEQ ID NO: 54, CDRL2 of SEQ ID NO: 55, and / or CDRL3 of SEQ ID NO: 56. Alternatively, the ULBP6 binding proteins described herein may comprise any one or combination of CDRs selected from CDRH1 of SEQ ID NO: 42, CDRH2 of SEQ ID NO: 43, and / or CDRH3 of SEQ ID NO: 44, CDRL1 of SEQ ID NO: 57, CDRL2 of SEQ ID NO: 58, and / or CDRL3 of SEQ ID NO: 59. Alternatively, the ULBP6 binding proteins described herein may comprise any one or combination of CDRs selected from CDRH1 of SEQ ID NO: 45, CDRH2 of SEQ ID NO: 46, and / or CDRH3 of SEQ ID NO: 47, CDRL1 of SEQ ID NO: 60, CDRL2 of SEQ ID NO: 61, and / or CDRL3 of SEQ ID NO: 62.

[0104] L1, L2, L3, H1, and H2 of a CDR tend to structurally exhibit one of a finite number of main-chain conformations. The particular canonical structural class of a CDR is defined by both the length and loop packing of the CDR, and by the residues in key positions in both the CDR and framework regions (which structurally determine the residues or SDRs). Martin and Thornton (1996; J Mol Biol 263:800-815) have produced an automatic method for defining canonical templates of "key residues". A cluster analysis is used to define a canonical class for a set of CDRs, and then the canonical templates are identified by analyzing buried hydrophobic, hydrogen-bonding residues, and conserved glycines and prolines. The CDRs of an antibody sequence can be assigned to a canonical class by comparing the sequence to templates of key residues and scoring each template using an identity or similarity matrix.

[0105] Examples of CDR standards are shown below in Table 3. The amino acid numbering used is Kabat.

[0106] [Table 3]

[0107] There may be multiple canonical positions of variant CDRs for each CDR, for each corresponding CDR, for each binding unit, for each heavy or light chain variable region, for each heavy or light chain, and for each antigen binding protein. Thus, any combination of substitutions may be present in the antigen binding proteins described herein (provided that the canonical structure of the CDRs is maintained such that the antigen binding protein is capable of specifically binding to ULBP6).

[0108] The particular canonical structural class of a CDR is defined by both the length and the loop packing of the CDR, and is determined by residues located at key positions in both the CDR and framework regions.

[0109] The term "epitope" as used herein refers to the portion of an antigen that contacts a specific binding domain of an antigen-binding protein, also known as a paratope. Epitopes can be linear or conformational / discontinuous. Conformational or discontinuous epitopes include amino acid residues that are separated by other sequences, i.e., that are not in a continuous sequence in the primary sequence of the antigen assembled by tertiary folding of the polypeptide chain. Such residues may come from different regions of the polypeptide chain, but are in close proximity in the three-dimensional structure of the antigen. In the case of multimeric antigens, conformational or discontinuous epitopes may include residues from different peptide chains. The specific residues contained within an epitope can be determined by the three-dimensional structure obtained through computer modeling programs or through methods known in the art, such as X-ray crystallography. Epitope mapping can be performed using a variety of techniques known to those skilled in the art. Exemplary methods include peptide-based approaches such as pepscan. According to peptide-based approaches, a set of overlapping peptides are screened for binding using techniques such as ELISA or by in vitro display of large libraries of peptide or protein variants, for example on phage. Detailed epitope information can be determined by structural analysis methods such as X-ray crystallography, solution nuclear magnetic resonance (NMR) spectroscopy, and cryo-electron microscopy (cryo-EM). Mutagenesis, such as alanine scanning, is an effective approach that uses analysis of binding loss for epitope mapping. Another method is the hydrogen / deuterium exchange (HDX) method combined with proteolysis and liquid chromatography-mass spectrometry (LC-MS) analysis to characterize discontinuous or conformational epitopes.

[0110] The term "antigen-binding site" refers to a site on an antigen-binding protein that can specifically bind to an antigen, which can be a single variable domain, e.g., a folded polypeptide domain containing sequences characteristic of antibody variable domains, or a paired VH / VL domain as can be present in a standard antibody. Single-chain Fv (ScFv) domains can also provide an antigen-binding site.

[0111] The competition between the ULBP6-binding protein described herein and a reference ULBP6-binding protein, e.g., a reference antibody, can be measured by one or more techniques known to those skilled in the art, such as ELISA, FMAT, surface plasmon resonance (SPR), or FORTEBIO OCTET biolayer interferometry (BLI). Such techniques may also be referred to as epitope binning. The competition assay may be performed by BLI. There are several possible reasons for this competition, including: the two proteins may bind to the same or overlapping epitopes, steric inhibition of binding may occur, or the binding of the first protein may induce a conformational change in the antigen that prevents or reduces the binding of the second protein. The ULBP6-binding protein may bind to human ULBP6 and compete with the reference ULBP6-binding protein for binding to human ULBP6. The reference ULBP6 binding protein may comprise (a) a VH region sequence of SEQ ID NO:21 and a VL region sequence of SEQ ID NO:31, or (b) a VH region sequence of SEQ ID NO:22 and a VL region sequence of SEQ ID NO:32, or (c) a VH region sequence of SEQ ID NO:16 and a VL region sequence of SEQ ID NO:26, or (d) a VH region sequence of SEQ ID NO:17 and a VL region sequence of SEQ ID NO:27.

[0112] The ULBP6 binding proteins described herein may share an overlapping epitope with a reference ULBP6 binding protein comprising a VH region comprising SEQ ID NO:21 and a VL region comprising SEQ ID NO:31, or may share an overlapping epitope with a reference ULBP6 binding protein comprising a VH region comprising SEQ ID NO:22 and a VL region comprising SEQ ID NO:32. Alternatively, the ULBP6 binding proteins described herein may share an overlapping epitope with a reference ULBP6 binding protein comprising a HC sequence comprising SEQ ID NO:75 and a LC sequence comprising SEQ ID NO:105, or may share an overlapping epitope with a reference ULBP6 binding protein comprising a HC sequence comprising SEQ ID NO:76 and a LC sequence comprising SEQ ID NO:106. Alternatively, the ULBP6 binding proteins described herein may not share an overlapping epitope with a reference ULBP6 binding protein comprising a VH region comprising SEQ ID NO:25 and a VL region comprising SEQ ID NO:35, or may not share an overlapping epitope with a reference ULBP6 binding protein comprising a HC sequence comprising SEQ ID NO:79 and a LC sequence comprising SEQ ID NO:109.

[0113] The term "neutralizing" as used throughout the present specification means that the biological activity of ULBP6 in the presence of a ULBP6-binding protein described herein is reduced in vitro or in vivo compared to the biological activity of ULBP6 in the absence of the ULBP6-binding protein. Neutralization can be due to one or more of the following: blocking the binding of ULBP6 to its receptor, preventing the activation of its receptor by ULBP6, downregulating ULBP6 or its receptor, or affecting effector function. Neutralization can be determined or measured using one or more assays known to those skilled in the art or described herein. For example, the blocking assay method described in Example 7 can be used to evaluate the neutralizing ability of a ULBP6-binding protein.

[0114] The effect of the ULBP6 binding protein on the interaction between ULBP6 and NKG2D may be partial or total. A neutralizing ULBP6 binding protein may neutralize the activity of ULBP6 by at least 20%, 30%, 40%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 82%, 84%, 86%, 88%, 90%, 92%, 94%, 95%, 96%, 97%, 98%, 99%, or 100%, compared to the ULBP6-NKG2D interaction in the absence of the ULBP6 binding protein. The ULBP6 binding protein described herein may inhibit the interaction between human ULBP6 and human NKG2D by 70% or more, 75% or more, 80% or more, 85% or more, 90% or more, or 95% or more.

[0115] The ULBP6 binding protein described herein can inhibit the binding of sULBP6 to NKG2D. The ULBP6 binding protein can inhibit the binding of sULBP6 to NKG2D by 70% or more, 75% or more, 80% or more, 85% or more, 90% or more, or 95% or more. The ULBP6 binding protein described herein can block the binding of sULBP6 to NKG2D. In some examples, the ULBP6 binding protein can block the binding of sULBP6 to NKG2D by 70% or more, 75% or more, 80% or more, 90% or more, or 95% or more.

[0116] "Percent identity" or "% identity" between a query nucleic acid sequence and a nucleic acid sequence of interest is an "identity" value expressed as a percentage, which is calculated over the entire length of the query sequence using an appropriate algorithm (e.g., BLASTN, FASTA, Needleman-Wunsch, Smith-Waterman, LALIGN, or GenePAST / KERR) or software (e.g., DNASTAR Lasergene, GenomeQuest, EMBOSS needle, or EMBOSS infoalign) after pairwise global sequence alignment using an appropriate algorithm (e.g., Needleman-Wunsch or GenePAST / KERR) or software (e.g., DNASTAR Lasergene or GenePAST / KERR). Importantly, the query nucleic acid sequence can be represented by the nucleic acid sequence disclosed herein, particularly in one or more of the claims.

[0117] "Percent identity" or "% identity" between a query amino acid sequence and a subject amino acid sequence is an "identity" value expressed as a percentage, which is calculated over the entire length of the query sequence using an appropriate algorithm (e.g., BLASTP, FASTA, Needleman-Wunsch, Smith-Waterman, LALIGN, or GenePAST / KERR) or software (e.g., DNASTAR Lasergene, GenomeQuest, EMBOSS needle, or EMBOSS infoalign) after performing a pairwise global sequence alignment using an appropriate algorithm (e.g., Needleman-Wunsch or GenePAST / KERR) or software (e.g., DNASTAR Lasergene or GenePAST / KERR). Importantly, the query amino acid sequence can be represented by an amino acid sequence disclosed herein, particularly in one or more of the claims.

[0118] The query sequence may be 100% identical to the subject sequence, or may contain a certain integer or less of amino acid or nucleotide modification compared to the subject sequence such that the identity percentage is less than 100%. For example, the query sequence is at least 50%, 60%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, or 99% identical to the subject sequence. In the case of nucleic acid sequences, such modifications include deletion, substitution, or insertion of at least one nucleotide residue, which may occur at the 5' or 3' terminal position of the query sequence, or at any position between these terminal positions, either individually among the nucleotide residues in the query sequence, or among one or more contiguous groups within the query sequence. In the case of amino acid sequences, such modifications include the deletion, substitution (including conservative and non-conservative substitutions), or insertion of at least one amino acid residue, which may occur at the amino- or carboxy-terminal position of the query sequence, or anywhere between these terminal positions, either individually among the amino acid residues in the query sequence, or among one or more contiguous groups within the query sequence.

[0119] In the case of antibody sequences, the % identity can be determined over the entire length of the query sequence, including the CDRs. Alternatively, the % identity can exclude one or more or all of the CDRs. For example, all of the CDRs can be 100% identical to the subject sequence, and the % identity changes can be in the remainder of the query sequence, such as framework sequences, so that the CDR sequences are constant and intact.

[0120] The ULBP6 binding proteins described herein may comprise a CDRH3 that is 100% identical to any one of SEQ ID NOs: 38, 41, 44, or 47. The ULBP6 binding proteins described herein may comprise a CDRH3 that is 100% identical to SEQ ID NO: 38. The ULBP6 binding proteins described herein may comprise a CDRH3 that is 100% identical to SEQ ID NO: 41. The ULBP6 binding proteins described herein may comprise a CDRH3 that is 100% identical to SEQ ID NO: 44. The ULBP6 binding proteins described herein may comprise a CDRH3 that is 100% identical to SEQ ID NO: 47.

[0121] A ULBP6 binding protein described herein may comprise a CDRH1 that is 100% identical to SEQ ID NO: 36, a CDRH2 that is 100% identical to SEQ ID NO: 37, and / or a CDRH3 that is 100% identical to SEQ ID NO: 38, a CDRL1 that is 100% identical to SEQ ID NO: 51, a CDRL2 that is 100% identical to SEQ ID NO: 52, and / or a CDRL3 that is 100% identical to SEQ ID NO: 53. Alternatively, a ULBP6 binding protein described herein may comprise a CDRH1 that is 100% identical to SEQ ID NO: 39, a CDRH2 that is 100% identical to SEQ ID NO: 40, and / or a CDRH3 that is 100% identical to SEQ ID NO: 41, a CDRL1 that is 100% identical to SEQ ID NO: 54, a CDRL2 that is 100% identical to SEQ ID NO: 55, and / or a CDRL3 that is 100% identical to SEQ ID NO: 56. Alternatively, a ULBP6 binding protein described herein may comprise a CDRH1 that is 100% identical to SEQ ID NO: 42, a CDRH2 that is 100% identical to SEQ ID NO: 43, and / or a CDRH3 that is 100% identical to SEQ ID NO: 44, a CDRL1 that is 100% identical to SEQ ID NO: 57, a CDRL2 that is 100% identical to SEQ ID NO: 58, and / or a CDRL3 that is 100% identical to SEQ ID NO: 59. Alternatively, a ULBP6 binding protein described herein may comprise a CDRH1 that is 100% identical to SEQ ID NO: 45, a CDRH2 that is 100% identical to SEQ ID NO: 46, and / or a CDRH3 that is 100% identical to SEQ ID NO: 47, a CDRL1 that is 100% identical to SEQ ID NO: 60, a CDRL2 that is 100% identical to SEQ ID NO: 61, and / or a CDRL3 that is 100% identical to SEQ ID NO: 62.

[0122] A ULBP6 binding protein described herein may comprise a CDRH1 that is 100% identical to SEQ ID NO: 36, a CDRH2 that is 100% identical to SEQ ID NO: 37, a CDRH3 that is 100% identical to SEQ ID NO: 38, a CDRL1 that is 100% identical to SEQ ID NO: 51, a CDRL2 that is 100% identical to SEQ ID NO: 52, and a CDRL3 that is 100% identical to SEQ ID NO: 53. Alternatively, a ULBP6 binding protein described herein may comprise a CDRH1 that is 100% identical to SEQ ID NO: 39, a CDRH2 that is 100% identical to SEQ ID NO: 40, a CDRH3 that is 100% identical to SEQ ID NO: 41, a CDRL1 that is 100% identical to SEQ ID NO: 54, a CDRL2 that is 100% identical to SEQ ID NO: 55, and a CDRL3 that is 100% identical to SEQ ID NO: 56. Alternatively, a ULBP6 binding protein described herein may comprise a CDRH1 that is 100% identical to SEQ ID NO: 42, a CDRH2 that is 100% identical to SEQ ID NO: 43, a CDRH3 that is 100% identical to SEQ ID NO: 44, a CDRL1 that is 100% identical to SEQ ID NO: 57, a CDRL2 that is 100% identical to SEQ ID NO: 58, and a CDRL3 that is 100% identical to SEQ ID NO: 59. Alternatively, a ULBP6 binding protein described herein may comprise a CDRH1 that is 100% identical to SEQ ID NO: 45, a CDRH2 that is 100% identical to SEQ ID NO: 46, a CDRH3 that is 100% identical to SEQ ID NO: 47, a CDRL1 that is 100% identical to SEQ ID NO: 60, a CDRL2 that is 100% identical to SEQ ID NO: 61, and a CDRL3 that is 100% identical to SEQ ID NO: 62.

[0123] The ULBP6 binding proteins described herein may comprise a VH region that is at least 90% identical to SEQ ID NO:21 and a VL region that is at least 90% identical to SEQ ID NO:31. The ULBP6 binding proteins described herein may comprise a VH region that is at least 95% identical to SEQ ID NO:21 and a VL region that is at least 95% identical to SEQ ID NO:31. The ULBP6 binding proteins described herein may comprise a VH region that is at least 90% identical to SEQ ID NO:22 and a VL region that is at least 90% identical to SEQ ID NO:32. The ULBP6 binding proteins described herein may comprise a VH region that is at least 95% identical to SEQ ID NO:22 and a VL region that is at least 95% identical to SEQ ID NO:32. The ULBP6 binding proteins described herein may comprise a VH region that is at least 90% identical to SEQ ID NO:23 and a VL region that is at least 90% identical to SEQ ID NO:33. The ULBP6 binding proteins described herein may comprise a VH region that is at least 95% identical to SEQ ID NO:23 and a VL region that is at least 95% identical to SEQ ID NO:33. The ULBP6 binding proteins described herein may comprise a VH region that is at least 90% identical to SEQ ID NO: 24 and a VL region that is at least 90% identical to SEQ ID NO: 34. The ULBP6 binding proteins described herein may comprise a VH region that is at least 95% identical to SEQ ID NO: 24 and a VL region that is at least 95% identical to SEQ ID NO: 34.

[0124] The ULBP6 binding proteins described herein may comprise a VH region that is 100% identical to SEQ ID NO:21 and a VL region that is 100% identical to SEQ ID NO:31, or may comprise a VH region that is 100% identical to SEQ ID NO:22 and a VL region that is 100% identical to SEQ ID NO:32, or may comprise a VH region that is 100% identical to SEQ ID NO:23 and a VL region that is 100% identical to SEQ ID NO:33, or may comprise a VH region that is 100% identical to SEQ ID NO:24 and a VL region that is 100% identical to SEQ ID NO:34.

[0125] The ULBP6 binding proteins described herein may comprise a heavy chain (HC) sequence that is at least 90% identical to SEQ ID NO: 75 and / or a light chain (LC) sequence that is at least 90% identical to SEQ ID NO: 105. Alternatively, the ULBP6 binding proteins described herein may comprise a heavy chain (HC) sequence that is at least 95% identical to SEQ ID NO: 75 and / or a light chain (LC) sequence that is at least 95% identical to SEQ ID NO: 105. Alternatively, the ULBP6 binding proteins described herein may comprise a heavy chain (HC) sequence that is at least 90% identical to SEQ ID NO: 76 and / or a light chain (LC) sequence that is at least 90% identical to SEQ ID NO: 106. Alternatively, the ULBP6 binding proteins described herein may comprise a heavy chain (HC) sequence that is at least 95% identical to SEQ ID NO: 76 and / or a light chain (LC) sequence that is at least 95% identical to SEQ ID NO: 106.

[0126] The ULBP6 binding proteins described herein may comprise a heavy chain (HC) sequence that is 100% identical to SEQ ID NO: 75 and / or a light chain (LC) sequence that is 100% identical to SEQ ID NO: 105, or may comprise a heavy chain (HC) sequence that is 100% identical to SEQ ID NO: 76 and / or a light chain (LC) sequence that is 100% identical to SEQ ID NO: 106, or may comprise a heavy chain (HC) sequence that is 100% identical to SEQ ID NO: 77 and / or a light chain (LC) sequence that is 100% identical to SEQ ID NO: 107, or may comprise a heavy chain (HC) sequence that is 100% identical to SEQ ID NO: 78 and / or a light chain (LC) sequence that is 100% identical to SEQ ID NO: 108. The ULBP6 binding proteins described herein may comprise a heavy chain (HC) sequence that is 100% identical to SEQ ID NO: 75 and a light chain (LC) sequence that is 100% identical to SEQ ID NO: 105. Alternatively, a ULBP6 binding protein described herein may comprise a heavy chain (HC) sequence that is 100% identical to SEQ ID NO: 76 and a light chain (LC) sequence that is 100% identical to SEQ ID NO: 106. Alternatively, a ULBP6 binding protein described herein may comprise a heavy chain (HC) sequence that is 100% identical to SEQ ID NO: 77 and a light chain (LC) sequence that is 100% identical to SEQ ID NO: 107. Alternatively, a ULBP6 binding protein described herein may comprise a heavy chain (HC) sequence that is 100% identical to SEQ ID NO: 78 and a light chain (LC) sequence that is 100% identical to SEQ ID NO: 108.

[0127] The VH or VL (or HC or LC) sequence may be a variant sequence having up to 10 amino acid substitutions, additions, or deletions. For example, the variant sequence may have up to 9, 8, 7, 6, 5, 4, 3, 2, or 1 amino acid substitution(s), addition(s), or deletion(s).

[0128] The HC sequence can be a variant sequence having up to 40 amino acid substitutions, additions, or deletions. For example, the variant sequence can have up to 35, 30, 25, 20, 15, or 10 amino acid substitutions, additions, or deletions.

[0129] The LC sequence can be a variant sequence having up to 20 amino acid substitutions, additions, or deletions. For example, the variant sequence can have up to 15, 10, or 5 amino acid substitutions, additions, or deletions.

[0130] The sequence variations may exclude one or more or all of the CDRs, e.g., the CDRs are the same as the VH or VL (or HC or LC) sequence, and the variations are in the remainder of the VH or VL (or HC or LC) sequence, such that the CDR sequences remain constant and intact.

[0131] Usually the mutations are substitutions, particularly conservative substitutions, such as those shown in Table 2.

[0132] The mutant sequences substantially retain the biological properties of the unaltered protein, such as inhibition of binding of ULBP6 to NKG2D.

[0133] As will be appreciated by those skilled in the art, post-translational modifications may occur during the production of antigen-binding proteins such as antibodies in host cells. For example, post-translational modifications include cleavage of specific leader sequences, addition of various sugar moieties in various glycosylation patterns, non-enzymatic glycosylation, deamidation, oxidation, scrambling of disulfide bonds, and other cysteine ​​variants, such as free sulfhydryls, racemic disulfides, thioether and trisulfide bonds, isomerization, C-terminal lysine clipping, and N-terminal glutamine cyclization. The disclosure herein encompasses the use of antigen-binding proteins that have been subjected to or have been subjected to one or more post-translational modifications. Thus, the "antigen-binding protein" or "antibody" of the present invention includes the above-defined "antigen-binding protein" or "antibody", respectively, that has been subjected to post-translational modifications as described herein.

[0134] Glycation is a post-translational nonenzymatic chemical reaction between reducing sugars, such as glucose, and free amine groups in proteins, typically found at the epsilon amine of lysine side chains or at the N-terminus of proteins. Glycation can only occur in the presence of reducing sugars during production and storage.

[0135] Deamidation can occur during production and storage and is primarily an enzymatic reaction that converts asparagine (N) to iso-aspartic acid (iso-aspartate) and aspartic acid (aspartate) (D) in a ratio of about 3:1. The deamidation reaction is therefore associated with the isomerization of aspartic acid (D) to iso-aspartic acid. Both the deamidation of asparagine and the isomerization of aspartic acid involve the intermediate succinimide. To a much lesser extent, deamidation can occur at glutamine residues as well. Deamidation can occur in the CDRs, Fab (non-CDR regions), or Fc regions.

[0136] Oxidation can occur during production and storage (i.e., in the presence of oxidizing conditions) and results in covalent modification of proteins induced directly by reactive oxygen species or indirectly by reaction with secondary by-products of oxidative stress. Oxidation occurs primarily at methionine residues, but can also occur at tryptophan and free cysteine ​​residues. Oxidation can occur in the CDRs, Fab (non-CDR) regions, or Fc regions.

[0137] Disulfide bond scrambling can occur during production and basic storage conditions. Under certain circumstances, disulfide bonds can be cleaved or misformed, resulting in unpaired cysteine ​​residues (-SH). Such free (unpaired) sulfhydryls (-SH) can facilitate shuffling.

[0138] Thioether formation and racemization of disulfide bonds can occur under basic conditions during production or storage by beta-elimination of the disulfide bridge back to a cysteine ​​residue via dehydroalanine and a persulfide intermediate. Subsequent cross-linking of dehydroalanine with cysteine ​​leads to the formation of a thioether bond, or the free cysteine ​​residue can reform the disulfide bond with a mixture of D- and L-cysteine.

[0139] Trisulfides result from the insertion of a sulfur atom into a disulfide bond (Cys-SSS-Cys) and form due to the presence of hydrogen sulfide in the production cell culture.

[0140] N-terminal glutamine (Q) and glutamate (E) of the heavy and / or light chains can form pyroglutamic acid (pGlu) via cyclization. Most pGlu formation occurs in the production bioreactor, but it may form non-enzymatically depending on the pH and temperature of processing and storage conditions. Cyclization of the N-terminal Q or E is commonly observed in native human antibodies.

[0141] C-terminal lysine clipping is an enzymatic reaction catalyzed by carboxypeptidases and is commonly observed in recombinant and native human antibodies. Variations of this process include the removal of lysines from one or both heavy chains by cellular enzymes from the recombinant host cell. When administered to a human subject / patient, any remaining C-terminal lysines are likely to be removed. Pharmaceutical Compositions

[0142] The antigen binding proteins described herein may be incorporated into pharmaceutical compositions for use in the treatment of human diseases as described herein. The pharmaceutical compositions may comprise the antigen binding protein, optionally in combination with one or more pharma- ceutically acceptable carriers and / or excipients.

[0143] Such compositions may contain known pharma- ceutically acceptable carriers as required for good pharmaceutical practice.

[0144] The pharmaceutical composition may be administered by injection or continuous infusion (examples include, but are not limited to, intravenous, intraperitoneal, intradermal, subcutaneous, intramuscular, intraocular, and intraportal). The pharmaceutical composition may be suitable for intravenous administration. The pharmaceutical composition may be suitable for subcutaneous administration. The pharmaceutical composition may be suitable for topical administration (including, but not limited to, transdermal, inhalation, intranasal, or ocular administration) or enteral administration (including, but not limited to, oral, vaginal, or rectal administration).

[0145] The pharmaceutical composition may contain 0.5 mg to 10 g of ULBP6 binding protein, for example, 5 mg to 1 g of ULBP6 binding protein, or may contain 5 mg to 500 mg of ULBP6 binding protein, for example, 5 mg to 50 mg of ULBP6 binding protein.

[0146] Methods for preparing such pharmaceutical compositions are well known to those skilled in the art. Other additives may be added to the composition depending on the mode of administration and the particular protein used.

[0147] The effective dosage and treatment regimen for administering the antigen binding protein may depend on factors such as the age, weight, and health condition and disease of the patient being treated. Such factors are within the jurisdiction of the attending physician.

[0148] The pharmaceutical composition may be included in a kit that includes the antigen binding protein together with other agents and / or together with instructions for use. For convenience, the kit may include reagents in predetermined amounts together with instructions for use. The kit may also include a device used to administer the pharmaceutical composition.

[0149] The terms "individual," "subject," and "patient" are used herein. Typically, an individual of interest may be a subject, and a subject in need of treatment may be a patient. An individual, subject, or patient may be an animal. An individual, subject, or patient may be a mammal, such as a primate, e.g., a marmoset or monkey. An individual, subject, or patient may be a human.

[0150] The antigen binding proteins described herein can also be used in therapeutic methods. As will be appreciated by those skilled in the art, the treatments referred to herein refer to the treatment of established diseases. However, the antigen binding proteins disclosed herein may also be useful for the prevention of certain diseases, depending on the condition. The antigen binding proteins described herein are used in an amount effective for therapeutic, prophylactic, or prophylactic treatment. A therapeutically effective amount of the antigen binding proteins described herein is an amount effective for improving or alleviating one or more symptoms of a disease, or for preventing or treating a disease.

[0151] Antigen binding proteins can be prepared by any of a number of conventional techniques: for example, they can be purified from cells which naturally express them (e.g., antibodies can be purified from the hybridoma which produces them) or they can be produced in recombinant expression systems.

[0152] The ULBP6 binding protein described herein is provided for use in treatment. The ULBP6 binding protein is provided for use in treatment of cancer. Use of the ULBP6 binding protein in the manufacture of a medicament for treatment of cancer is also provided. Use of the ULBP6 binding protein in the manufacture of a medicament for treatment of cancer is also provided, the ULBP6 binding protein comprising a VH region comprising SEQ ID NO: 21 and a VL region comprising SEQ ID NO: 31. Use of the ULBP6 binding protein in the manufacture of a medicament for treatment of cancer is also provided, the ULBP6 binding protein comprising a VH region comprising SEQ ID NO: 22 and a VL region comprising SEQ ID NO: 32. Further provided is a pharmaceutical composition comprising 100 to 1400 mg of a ULBP6 binding protein. Further provided is a pharmaceutical composition comprising 100 to 1400 mg of a ULBP6 binding protein comprising a VH region comprising SEQ ID NO: 21 and a VL region comprising SEQ ID NO: 31. Also provided herein is a pharmaceutical composition comprising 100 to 1400 mg of a ULBP6 binding protein comprising an HC sequence comprising SEQ ID NO: 75 and an LC sequence comprising SEQ ID NO: 105.

[0153] Further provided herein is an isolated ULBP6 binding protein or a pharmaceutical composition comprising said isolated ULBP6 binding protein for the treatment of cancer. Manufacturing method

[0154] A number of different expression systems and purification methods can be used to produce antigen binding proteins. Generally, a host cell is transformed with a recombinant expression vector encoding the desired antigen binding protein. Depending on the expression system, the expression vector can be maintained by the host as a separate genetic element or can be integrated into the host chromosome. A wide range of host cells can be used, including prokaryotes (including gram-negative or gram-positive bacteria, e.g., Escherichia coli, Bacilli, Pseudomonas, Corynebacterium), eukaryotes including yeast (e.g., Saccharomyces cerevisiae, Pichia pastoris), fungi (e.g., Aspergillus), or higher eukaryotes including insect cells and cell lines of mammalian origin (e.g., CHO, NS0, PER.C6, HEK293, HeLa).

[0155] The host cell can be an isolated host cell. The host cell is not normally part of a multicellular organism (e.g., a plant or an animal). The host cell can be a non-human host cell.

[0156] Suitable cloning and expression vectors for use with bacterial, fungal, yeast, and mammalian host cells are known in the art.

[0157] The cells can be cultured under conditions that promote the expression of the antigen-binding protein using a variety of devices, such as shake flasks, spinner flasks, and bioreactors. The polypeptide can be recovered by conventional protein purification procedures. Protein purification procedures typically consist of a series of unit operations, including, for example, one or more filtration and / or chromatography processes developed to selectively enrich and isolate the antigen-binding protein. The purified antigen-binding protein can be formulated into a pharma- ceutically acceptable composition.

[0158] Fc engineering methods can be applied to modify the functional or pharmacokinetic properties of antibodies. Effector functions can be altered by making mutations in the Fc region, which increase or decrease binding to C1q or Fcγ receptors and alter CDC or ADCC activity, respectively. Modification of the glycosylation pattern of the antibody can also be performed to alter effector functions. The in vivo half-life of the antibody can be altered by making mutations that affect Fc binding to FcRn (neonatal Fc receptor).

[0159] The term "effector function" as used herein refers to one or more of the antibody-mediated effects, including antibody-dependent cell-mediated cytotoxicity (ADCC), antibody-mediated complement activation, including complement-dependent cytotoxicity (CDC), complement-dependent cell-mediated phagocytosis (CDCP), antibody-dependent complement-mediated cytolysis (ADCML), and Fc-mediated phagocytosis or antibody-dependent cellular phagocytosis (ADCP).

[0160] Interactions between the Fc region of an antigen binding protein or antibody and various Fc receptors (FcRs), including FcγRI (CD64), FcγRII (CD32), FcγRIII (CD16), FcRn, C1q, and type II Fc receptors, are thought to mediate the effector functions of the antigen binding protein or antibody. Important biological effects can be the result of effector functions. Typically, the ability to mediate effector functions requires that the antigen binding protein or antibody binds to an antigen, and not all antigen binding proteins or antibodies mediate all effector functions.

[0161] Effector function can be assessed in a number of ways, including, for example, assessing ADCC effector function of an antibody coated on a target cell mediated by natural killer (NK) cells via FcγRIII or by monocytes / macrophages via FcγRI, or assessing CDC effector function of an antibody coated on a target cell mediated by the complement cascade via C1q. For example, the antigen binding proteins described herein can be assessed for ADCC effector function in a natural killer cell assay.

[0162] The effect of the mutations on effector function (e.g., FcRn binding, FcγR and C1q binding, CDC, ADCML, ADCC, ADCP) can be assessed.

[0163] Throughout this specification, amino acid residues of the Fc region in an antibody sequence or full length antigen binding protein sequence are numbered according to the EU index numbering rules.

[0164] Human IgG1 constant regions containing certain mutations have been shown to enhance binding to Fc receptors. In some cases, such mutations have also been shown to enhance effector functions such as ADCC and CDC, as described below. The antigen binding proteins described herein may include any of the following mutations:

[0165] Enhanced CDC: Fc engineering can be used to enhance complement-based effector functions. For example (for IgG1), K326W / E333S, S267E / H268F / S324T, and IgG1 / IgG3 cross-subclasses can increase C1q binding, while E345R and E345R / E430G / S440Y result in preformed IgG hexamers.

[0166] Enhancement of ADCC: Fc engineering can be used to enhance ADCC. For example (for IgG1), F243L / R292P / Y300L / V305I / P396L, S239D / I332E, and S298A / E333A / K334A increase FcγRIIIa binding, S239D / I332E / A330L increase FcγRIIIa binding and decrease FcγRIIb binding, and G236A / S239D / I332E improve binding to FcγRIIa, improve the FcγRIIa / FcγRIIb binding ratio (activation / inhibition ratio), and enhance phagocytosis of antibody-coated target cells by macrophages. The ULBP6 binding protein described herein is mutated at positions 239 and 332 (e.g., S239D and I332E). The ULBP6 binding protein described herein may be mutated at three or more positions selected from positions 239, 332, and 330 (e.g., S239D, I332E, and A330L (EU index numbering)). The ULBP6 binding protein described herein may be an antibody or an antigen-binding fragment thereof, and the antibody or antigen-binding fragment thereof may be S239D / A330L / I332E, G236A / S ... and an altered Fc region comprising amino acid substitutions selected from L234Y / L235Q / G236W / S239M / H268D / D270E / S298A, L235V / F243L / R292P / Y300L / P396L, or F243L / R292P / Y300L / V305I / P396L. An asymmetric Fc in which one heavy chain contains L234Y / L235Q / G236W / S239M / H268D / D270E / S298A mutations and the opposite heavy chain contains D270E / K326D / A330M / K334E has increased affinity for FcγRIIIa F158 (low affinity allele) and FcγRIIIa V158 (high affinity allele) without increased binding affinity to the inhibitory FcγRIIb.

[0167] Enhancing ADCP: Fc engineering can be used to enhance ADCP, for example (with reference to IgG1), G236A / S239D / I332E increases FcγRIIa binding and increases FcγRIIIa binding.

[0168] Increased co-binding: Fc engineering can be used to increase co-binding to FcRs. For example (with reference to IgG1), S267E / L328F increases FcγRIIb binding, N325S / L328F increases FcγRIIa binding and decreases FcγRIIIa binding.

[0169] The antigen binding proteins described herein may comprise a heavy chain constant region with an altered glycosylation profile, such that the antigen binding protein has enhanced effector function, such as enhanced ADCC, enhanced CDC, or both enhanced ADCC and CDC. Examples of suitable methods for generating antigen binding proteins with altered glycosylation profiles are described in WO2003011878, WO2006014679, and EP1229125, all of which may be applied to the antigen binding proteins described herein.

[0170] The absence of α1,6 innermost fucose residues in the Fc glycan moiety on N297 of IgG1 antibodies can enhance affinity for FcγRIIIA. Thus, defucosylated or hypofucosylated monoclonal antibodies may have increased therapeutic efficacy compared to monoclonal antibodies with standard or higher fucosylation.

[0171] Provided herein is an antigen binding protein comprising a chimeric heavy chain constant region. The antigen binding protein may comprise an IgG1 / IgG3 chimeric heavy chain constant region, such that the antigen binding protein has enhanced effector function, e.g., enhanced ADCC or enhanced CDC, or enhanced ADCC and CDC function. For example, the chimeric antigen binding protein of the present invention may comprise at least one CH2 domain from IgG3. The antigen binding protein may have one CH2 domain from IgG3 or both CH2 domains from IgG3. The chimeric antigen binding protein may comprise an IgG1 CH1 domain, an IgG3 CH2 domain, and an IgG3 CH3 domain. The chimeric antigen binding protein may comprise an IgG1 CH1 domain, an IgG3 CH2 domain, and an IgG3 CH3 domain, except for position 435, which is H (histidine).

[0172] The antigen binding protein may comprise an IgG1 CH1 domain and at least one CH2 domain from IgG3. The chimeric antigen binding protein may comprise an IgG1 CH1 domain and the following residues 274Q, 276K, 296F, 300F, and 339T, which correspond to IgG3 residues in the CH2 domain. The chimeric antigen binding protein may also comprise 356E, which corresponds to IgG3 residues in the CH3 domain. The antigen binding protein may also comprise one or more of the following residues 358M, 384S, 392N, 397M, 422I, 435R, and 436F, which correspond to IgG3 residues in the CH3 domain.

[0173] Also provided is a method of producing an antigen binding protein, the method comprising a) culturing a recombinant host cell comprising an expression vector comprising an isolated nucleic acid as described herein, the expression vector comprising a nucleic acid sequence encoding a chimeric Fc region having both amino acid residues of an IgG1 Fc region and amino acid residues of an IgG3 Fc region (e.g., as described above), the method further comprising b) recovering the antigen binding protein.

[0174] Such methods for the production of antigen binding proteins with chimeric heavy chain constant regions can be carried out, for example, using the COMPLEGENT technology system available from BioWa, Inc. (Princeton, NJ) and Kyowa Hakko Kirin Co, Ltd. The COMPLEGENT system comprises a recombinant host cell containing an expression vector in which a nucleic acid sequence encoding a chimeric Fc region having amino acid residues of both IgG1 and IgG3 Fc regions is expressed to produce an antigen binding protein with enhanced CDC activity, i.e., CDC activity is increased compared to an otherwise identical antigen binding protein lacking such chimeric Fc region, as described in WO2007011041 and US20070148165, each of which is incorporated herein by reference. Alternatively, CDC activity may be increased by introducing sequence-specific mutations into the Fc region of the IgG chain. Those skilled in the art will recognize other suitable systems.

[0175] Also provided is a method of producing the antigen binding proteins described herein, the method comprising a) culturing a recombinant host cell comprising an expression vector comprising an isolated nucleic acid described herein, wherein the FUT8 gene encoding alpha-1,6-fucosyltransferase is inactivated in the recombinant host cell, the method further comprising b) recovering the antigen binding protein.

[0176] Such methods for the production of antigen-binding proteins can be performed, for example, using the POTELLIGENT technology system available from BioWa, Inc. (Princeton, NJ), in which CHOK1SV cells lacking a functional copy of the FUT8 gene produce monoclonal antibodies with enhanced ADCC activity. The ADCC activity is increased compared to the same monoclonal antibodies produced in cells with a functional FUT8 gene, as described in US7214775, US6946292, WO0061739, and WO0231240 (all of which are incorporated herein by reference). Those skilled in the art will recognize other suitable systems.

[0177] The antigen binding proteins described herein can be produced in a host cell in which the FUT8 gene has been inactivated. The antigen binding proteins can be produced in - / -FUT8 host cells. The antigen binding proteins can be defucosylated at Asn297 (IgG1). The ULBP6 binding proteins described herein can be defucosylated. The ULBP6 binding protein can be an antibody comprising a heavy chain (HC) of SEQ ID NO: 75 and a light chain (LC) of SEQ ID NO: 105, which antibody is defucosylated. The ULBP6 binding protein can be an antibody comprising a heavy chain (HC) of SEQ ID NO: 76 and a light chain (LC) of SEQ ID NO: 106, which antibody is defucosylated.

[0178] Some isotypes of human constant regions, particularly IgG4 and IgG2 isotypes, may essentially lack the functions of a) classical pathway complement activation and / or b) ADCC. Various modifications to the heavy chain constant region of antigen-binding proteins may be made to alter effector function depending on the desired effector properties. IgG1 constant regions containing specific mutations that reduce binding to Fc receptors and reduce effector functions (e.g., ADCC and CDC) have been reported. Such modifications may be used alone, but may also be used alone or in combination with each other to alter effector function.

[0179] Provided herein is an antigen binding protein, which comprises a heavy chain constant region comprising both the mutated heavy chain constant region and the chimeric heavy chain constant region described individually above. For example, the antigen binding protein comprises at least one CH2 domain from IgG3 and one CH2 domain from IgG1, where the IgG1 CH2 domain has one or more mutations at positions selected from 239, 332, and 330 (e.g., the mutations can be selected from S239D, I332E, and A330L), such that the antigen binding protein has enhanced effector function, e.g., enhanced ADCC or enhanced CDC, or enhanced ADCC and enhanced CDC, compared to a comparable antigen binding protein having an IgG1 heavy chain constant region lacking said mutations. The IgG1 CH2 domain can have the mutations S239D and I332E. The IgG1 CH2 domain can have the mutations S239D, A330L, and I332E.

[0180] The ULBP6 binding proteins described herein may comprise a heavy chain (HC) sequence that is at least 90% identical to SEQ ID NO: 95 and / or a light chain (LC) sequence that is at least 90% identical to SEQ ID NO: 105. Alternatively, the ULBP6 binding proteins described herein may comprise a heavy chain (HC) sequence that is at least 95% identical to SEQ ID NO: 95 and / or a light chain (LC) sequence that is at least 95% identical to SEQ ID NO: 105. Alternatively, the ULBP6 binding proteins described herein may comprise a heavy chain (HC) sequence that is at least 90% identical to SEQ ID NO: 96 and / or a light chain (LC) sequence that is at least 90% identical to SEQ ID NO: 106. Alternatively, the ULBP6 binding proteins described herein may comprise a heavy chain (HC) sequence that is at least 95% identical to SEQ ID NO: 96 and / or a light chain (LC) sequence that is at least 95% identical to SEQ ID NO: 106.

[0181] The ULBP6 binding proteins described herein may comprise a heavy chain (HC) sequence that is 100% identical to SEQ ID NO: 95 and / or a light chain (LC) sequence that is 100% identical to SEQ ID NO: 105, may comprise a heavy chain (HC) sequence that is 100% identical to SEQ ID NO: 96 and / or a light chain (LC) sequence that is 100% identical to SEQ ID NO: 106, may comprise a heavy chain (HC) sequence that is 100% identical to SEQ ID NO: 97 and / or a light chain (LC) sequence that is 100% identical to SEQ ID NO: 107, or may comprise a heavy chain (HC) sequence that is 100% identical to SEQ ID NO: 98 and / or a light chain (LC) sequence that is 100% identical to SEQ ID NO: 108. Alternatively, the ULBP6 binding proteins described herein may comprise a heavy chain (HC) sequence that is 100% identical to SEQ ID NO: 95 and a light chain (LC) sequence that is 100% identical to SEQ ID NO: 105. Alternatively, a ULBP6 binding protein described herein may comprise a heavy chain (HC) sequence that is 100% identical to SEQ ID NO:96 and a light chain (LC) sequence that is 100% identical to SEQ ID NO:106.

[0182] Alternatively, there is provided an antigen binding protein comprising both a chimeric heavy chain constant region and a modified glycosylation profile as described herein. The antigen binding protein may have a modified glycosylation profile such that the ratio of fucose to mannose is 0.8:3 or less. The heavy chain constant region may comprise at least one CH2 domain from IgG3 and one CH2 domain from IgG1, and has a modified glycosylation profile such that the ratio of fucose to mannose is 0.8:3 or less, e.g., the antigen binding protein is defucosylated. The antigen binding protein may have an enhanced effector function, e.g., enhanced ADCC or enhanced CDC, or enhanced ADCC and enhanced CDC, compared to a comparable antigen binding protein having an IgG1 heavy chain constant region lacking said glycosylation profile.

[0183] Alternatively, the antigen binding protein has at least one IgG3 heavy chain CH2 domain and at least one heavy chain constant domain derived from an IgG1, wherein the CH2 domains of both IgGs are mutated according to the limitations described herein.

[0184] Provided herein are methods of producing the antigen binding proteins described herein, the methods comprising a) culturing a recombinant host cell comprising an expression vector comprising an isolated nucleic acid as described herein, where the expression vector comprises a nucleic acid sequence encoding a chimeric Fc domain having amino acid residues of both IgG1 and IgG3 Fc domains (e.g., as described above), and where the FUT8 gene encoding alpha-1,6-fucosyltransferase has been inactivated in the recombinant host cell, the method further comprising b) recovering the antigen binding protein.

[0185] Such methods for producing antigen binding proteins can be performed, for example, using the ACCRETAMAB technology system available from BioWa, Inc. (Princeton, NJ), which combines the POTELLIGENT and COMPLEGENT technology systems to produce antigen binding proteins with both enhanced ADCC activity and enhanced CDC activity compared to an otherwise identical monoclonal antibody that lacks a chimeric Fc domain and is fucosylated.

[0186] Also provided is an antigen binding protein comprising a mutated chimeric heavy chain constant region, the antigen binding protein having an altered glycosylation profile such that the antigen binding protein has an enhanced effector function, e.g., enhanced ADCC or enhanced CDC, or both enhanced ADCC and enhanced CDC. The mutations may be selected from positions 239, 332, and 330 (e.g., S239D, I332E, and A330L). The heavy chain constant region may comprise at least one CH2 domain from IgG3 and one CH1 domain from IgG1. The heavy chain constant region may have an altered glycosylation profile such that the fucose to mannose ratio is 0.8:3 or less, e.g., the antigen binding protein is defucosylated, whereby the antigen binding protein has an enhanced effector function compared to a comparable non-chimeric antigen binding protein without the mutations and altered glycosylation profile.

[0187] Provided herein is an antigen binding protein, which comprises a constant region such that it has reduced effector function, e.g., reduced ADCC and / or CDC. The heavy chain constant region may comprise a non-naturally functional constant region of IgG2 or IgG4 isotype or a mutated IgG1 constant region. Examples of suitable modifications are described in EP0307434. One example comprises substitutions with alanine at positions 235 and 237 (EU index numbering), i.e., L235A and G237A (commonly referred to as "LAGA" mutations). Another example comprises substitutions with alanine at positions 234 and 235 (EU index numbering), i.e., L234A and L235A (commonly referred to as "LALA" mutations). Further examples described in EP2691417 and US8969526 include P329G or P329R in combination with the LALA mutation to the Fcs of IgG1 (EU index numbering), and further include P329G or P329R in combination with S228P and L235E to the Fcs of IgG4 (EU index numbering).

[0188] Further modifications and mutations to reduce effector function may include deglycosylated N297A or N297Q or N297G (with reference to IgG1 unless otherwise stated), L235E, IgG4:F234A / L235A, and chimeric IgG2 / IgG4. IgG2:H268Q / V309L / A330S / P331S, and IgG2:V234A / G237A / P238S / H268A / V309L / A330S / P331S may reduce FcγR and C1q binding (US8961967). The ULBP6 binding proteins described herein may comprise a modified Fc region, where the modified Fc region comprises the amino acid substitutions N297A, N297Q, or N297G.

[0189] Other mutations that reduce effector function include L234F / L235E / P331S, a chimeric antibody made using CH1 and hinge regions from human IgG2 and CH2 and CH3 regions from human IgG4; IgG2m4, which is based on the IgG2 isotype with four key amino acid residue changes from IgG4 (H268Q, V309L, A330S, and P331S); IgG2σ, which contains V234A / G237A / P238S / H268A / V309L / A330S / P331S substitutions to eliminate affinity for Fcγ receptors and the C1q complement protein; These include G2m4 (H268Q / V309L / A330S / P331S, changes relative to IgG4), IgG4 (S228P / L234A / L235A), L234A / L235A(AA) in huIgG1, S228P / L234A / L235A in huIgG4, IgG1 (L234A / L235A / G237A / P238S / H268A / A330S / P331S), IgG4s1 (S228P / F234A / L235A / G237A / P238S), and IgG4s2 (S228P / F234A / L235A / DG236 / G237A / P238S (where D indicates a deletion).

[0190] The ULBP6 binding proteins described herein may comprise a heavy chain (HC) sequence that is at least 90% identical to SEQ ID NO: 85 and / or a light chain (LC) sequence that is at least 90% identical to SEQ ID NO: 105. Alternatively, the ULBP6 binding proteins described herein may comprise a heavy chain (HC) sequence that is at least 95% identical to SEQ ID NO: 85 and / or a light chain (LC) sequence that is at least 95% identical to SEQ ID NO: 105. The ULBP6 binding proteins described herein may comprise a heavy chain (HC) sequence that is at least 90% identical to SEQ ID NO: 86 and / or a light chain (LC) sequence that is at least 90% identical to SEQ ID NO: 106. Alternatively, the ULBP6 binding proteins described herein may comprise a heavy chain (HC) sequence that is at least 95% identical to SEQ ID NO: 86 and / or a light chain (LC) sequence that is at least 95% identical to SEQ ID NO: 106.

[0191] The ULBP6 binding proteins described herein may comprise a heavy chain (HC) sequence that is 100% identical to SEQ ID NO: 85 and / or a light chain (LC) sequence that is 100% identical to SEQ ID NO: 105, may comprise a heavy chain (HC) sequence that is 100% identical to SEQ ID NO: 86 and / or a light chain (LC) sequence that is 100% identical to SEQ ID NO: 106, may comprise a heavy chain (HC) sequence that is 100% identical to SEQ ID NO: 87 and / or a light chain (LC) sequence that is 100% identical to SEQ ID NO: 107, or may comprise a heavy chain (HC) sequence that is 100% identical to SEQ ID NO: 88 and / or a light chain (LC) sequence that is 100% identical to SEQ ID NO: 108. The ULBP6 binding proteins described herein may comprise a heavy chain (HC) sequence that is 100% identical to SEQ ID NO: 85 and a light chain (LC) sequence that is 100% identical to SEQ ID NO: 105. A ULBP6 binding protein described herein may comprise a heavy chain (HC) sequence that is 100% identical to SEQ ID NO:86 and a light chain (LC) sequence that is 100% identical to SEQ ID NO:106.

[0192] Half-life (t 1 / 2) refers to the time it takes for the serum concentration of an antigen-binding protein to reach half of its original value. The serum half-life of a protein can be measured by pharmacokinetic studies according to a method in which a radiolabeled protein is injected intravenously into mice and its plasma concentration is measured periodically as a function of time, for example from about 3 minutes to about 72 hours after injection. Other methods for pharmacokinetic analysis and measurement of the half-life of molecules are well known to those of skill in the art.

[0193] It has been reported that the long half-life of IgG antibodies depends on binding to FcRn.Therefore, substitutions that increase the binding affinity of IgG to FcRn at pH 6.0 while maintaining the pH dependency of the interaction with the target by engineering constant regions have been widely studied.The in vivo half-life of the antigen-binding proteins described herein can be altered by modifying the heavy chain constant domain or the FcRn-binding domain therein.

[0194] In adult mammals, FcRn, also known as the neonatal Fc receptor, plays an important role in maintaining serum antibody levels by acting as a protective receptor that binds antibodies of the IgG isotype and prevents their degradation. IgG molecules are endocytosed by endothelial cells and, if they bind to FcRn, are recycled out of the cell and returned to the circulation. In contrast, IgG molecules enter cells, do not bind to FcRn, and are targeted to the lysosomal pathway and degraded.

[0195] FcRn may be involved in both transcytosis across tissues and antibody clearance. Human IgG1 residues identified to directly interact with human FcRn include Ile253, Ser254, Lys288, Thr307, Gln311, Asn434, and His435. Mutations at any of these positions may increase serum half-life and / or alter effector properties of the antigen binding proteins of the invention.

[0196] The antigen binding proteins described herein may have amino acid modifications that increase the affinity of the constant domain or fragment thereof for FcRn. Increasing the half-life (i.e., serum half-life) of therapeutic and diagnostic IgG antibodies and other bioactive molecules has many advantages, including reducing the dosage and / or frequency of administration of these molecules. The antigen binding proteins may comprise all or a portion of an IgG constant domain (FcRn binding portion) with one or more of the following amino acid modifications:

[0197] For example, with respect to IgG1, M252Y / S254T / T256E (commonly referred to as the "YTE" mutation) and / or M428L / N434S (commonly referred to as the "LS" mutation) can increase FcRn binding at pH 6.0.

[0198] Half-life can also be extended by the following mutations (see, e.g., IgG1 and Kabat numbering): T250Q / M428L, V259I / V308F / M428L, N434A, and T307A / E380A / N434A.

[0199] Half-life and FcRn binding can also be extended by introducing the H433K and N434F mutations (commonly referred to as "HN" or "NHance" mutations) (for IgG1) (WO2006 / 130834).

[0200] WO00 / 42072 discloses polypeptides comprising variant Fc regions with altered FcRn binding affinity, the polypeptides comprising amino acid modifications at any one or more of amino acid positions 238, 252, 253, 254, 255, 256, 265, 272, 286, 288, 303, 305, 307, 309, 311, 312, 317, 340, 356, 360, 362, 376, 378, 380, 386, 388, 400, 413, 415, 424, 433, 434, 435, 436, 439 and 447 (EU index numbering) of the Fc region.

[0201] WO02 / 060919 discloses modified IgGs comprising an IgG constant domain which comprises one or more amino acid modifications compared to a wild-type IgG constant domain, which have an increased half-life compared to the half-life of an IgG having a wild-type IgG constant domain, wherein the one or more amino acid modifications are at one or more of positions 251, 253, 255, 285-290, 308-314, 385-389 and 428-435.

[0202] Shields et al. (2001, J Biol Chem;276:6591-604) used alanine scanning mutagenesis to modify residues in the Fc region of a human IgG1 antibody and then assessed binding to human FcRn. Positions that effectively blocked binding to FcRn when changed to alanine included I253, S254, H435, and Y436. Other positions such as E233-G236, R255, K288, L309, S415, and H433 showed less pronounced reductions in binding. Several amino acid positions showed improved FcRn binding when changed to alanine. Notable among these are P238, T256, E272, V305, T307, Q311, D312, K317, D376, E380, E382, S424, and N434. Many other amino acid positions showed either slight improvements in FcRn binding (D265, N286, V303, K360, Q362, and A378) or no changes (S239, K246, K248, D249, M252, E258, T260, S267, H268, S269, D270, K274, N276, Y278, D280, V282, E283, H285, T289, K290, R292, E293, E294, Q295, Y296, N297, S298, R301, N315, E318, K320, K321, K322, K323, K324, K325, K326, K327, K328, K329, K330, K331, K332, K333, K334, K335, K336, K337, K338, K339, K340, K341, K342, K343, K344, K345, K346, K347, K348, K349, K350, K351, K352, K353, K354, K355, K356, K357, K358, K360, K362, K363, K364, K365, K366, K367, K368, K369, K370, K374, K374, K376, K378, K380, K382, K383, K385, K389, K390, K392, K393, K394, 22, S324, K326, A327, P329, P331, E333, K334, T335, S337, K338, K340, Q342, R344, E345, Q345, Q347, R356, M358, T359, K360, N361, Y373, S375, S383, N384, Q386, E388, N389, N390, K392, L398, S400, D401, K414, R416, Q418, Q419, N421, V422, E430, T437, K439, S440, S442, S444, and K447).

[0203] The most significant effect in terms of improved FcRn binding was found with the combination mutants. At pH 6.0, the E380A / N434A mutant showed over 8-fold better binding to FcRn versus wild-type IgG1, compared to 2-fold for E380A and 3.5-fold for N434A. Adding T307A to this improved binding by 12-fold versus wild-type IgG1. The antigen binding proteins described herein may contain the E380A / N434A mutations and may have increased binding to FcRn.

[0204] Dall'Acqua et al. (2002, J Immunol.; 169:5171-80) described random mutagenesis and screening of a human IgG1 hinge-Fc fragment phage display library against mouse FcRn. They disclosed random mutagenesis of positions 251, 252, 254-256, 308, 309, 311, 312, 314, 385-387, 389, 428, 433, 434, and 436. A large improvement in the stability of the IgG1-human FcRn complex occurs when substituting residues located within a band spanning the Fc-FcRn interface (M252, S254, T256, H433, N434, and Y436) and, to a lesser extent, in the surrounding areas, such as V308, L309, Q311, G385, Q386, P387, and N389. The mutant with the highest affinity for human FcRn was obtained by combining the M252Y / S254T / T256E ("YTE") and H433K / N434F / Y436H mutations, and showed a 57-fold increase in affinity compared to wild-type IgG1. The in vivo behavior of such mutant human IgG1 showed an approximately 4-fold increase in serum half-life in cynomolgus monkeys compared to wild-type IgG1.

[0205] Thus, antigen binding proteins with optimized binding to FcRn are also provided. The antigen binding protein may comprise at least one amino acid modification in the Fc region of the antigen binding protein, the modification comprising 226, 227, 228, 230, 231, 233, 234, 239, 241, 243, 246, 250, 252, 256, 259, 264, 265, 267, 269, 270, 276, 284, 285, 288, 289, 290, 291, 292, 294, 297, 298, 299, 301, 302, 303, 305, 307, 308, 309, 311, 315, 317, 320, 322, 325, 327, 330, 332, 334, 335, 338, 340, 341, 342, 343, 344, 345, 346, 347, 348, 349, 350, 351, 352, 353, 354, 355, 356, 357, 358, 359, 360, 361, 362, 363, 364, 365, 366, 367, 368, 369, 370, 371, 372, 373, 374, 375, 376, 377, 378, 379, 380, 381, 382, ​​383, 384, 385, 42, 343, 345, 347, 350, 352, 354, 355, 356, 359, 360, 361, 362, 369, 370, 371, 375, 378, 380, 382, ​​384, 385, 386, 387, 389, 390, 392, 393, 394, 395, 396, 397, 398, 399, and at an amino acid position selected from the group consisting of 400, 401, 403, 404, 408, 411, 412, 414, 415, 416, 418, 419, 420, 421, 422, 424, 426, 428, 433, 434, 438, 439, 440, 443, 444, 445, 446, and 447.

[0206] Furthermore, various publications describe methods for obtaining physiologically active molecules with modified half-lives, which can be obtained by introducing FcRn-binding polypeptides into the molecule (WO97 / 43316, US5869046, US5747035, WO96 / 32478 and WO91 / 14438) or by fusing the molecule with antibodies that have a preserved FcRn-binding affinity but a significantly reduced affinity for other Fc receptors (WO99 / 43713) or with the FcRn-binding domain of an antibody (WO00 / 09560, US4703039).

[0207] Fc variants with enhanced FcRn affinity that improve both antibody cytotoxicity and half-life were identified in a pH 6.0 screen. Selected IgG variants can be generated as hypofucosylated molecules. The resulting variants show increased serum persistence in hFcRn mice as well as preserved enhanced ADCC. Exemplary variants (with reference to IgG1 and Kabat numbering) include P230T / V303A / K322R / N389T / F404L / N434S, P228R / N434S, Q311R / K334R / Q342E / N434Y, C226G / Q386R / N434Y, T307P / N38 9T / N434Y, P230S / N434S, P230T / V305A / T307A / A378V / L398P / N434S, P23OT / P387 S / N434S, P230Q / E269D / N434S, N276S / A378V / N434S, T307A / N315D / A330V / 382V / N 389T / N434Y, T256N / A378V / S383N / N434Y, N315D / A330V / N361D / A387V / N434Y, V2 59I / N315D / M428L / N434Y, P230S / N315D / M428L / N434Y, F241L / V264E / T307P / A37 Examples include 8V / H433R, T250A / N389K / N434Y, V305A / N315D / A330V / P395A / N434Y, V264E / Q386R / P396L / N434S / K439R, and E294del / T307P / N434Y (where "del" indicates a deletion).

[0208] Substitutions in the constant region can greatly improve the function of therapeutic IgG antibodies, but substitutions in strictly conserved constant regions carry the risk of immunogenicity in humans, and substitutions in highly diverse variable region sequences may be less immunogenic. Reports on the variable region include engineering CDR residues to improve binding affinity to antigens, and engineering CDR and framework residues to improve stability and reduce immunogenicity risk. Improved affinity to antigens can be achieved by affinity maturation using phage display or ribosome display of randomized libraries.

[0209] Improved stability can be reasonably obtained from sequence-based and structure-based rational design. Reduction of immunogenicity risk (deimmunization) can be achieved by various humanization methods and removal of potential T-cell epitopes, which can be predicted using in silico techniques or predicted by in vitro assays. Additionally, variable regions have been engineered to lower pI. Longer half-lives have been observed for such antibodies compared to wild-type antibodies, despite comparable FcRn binding. By engineering or selecting antibodies whose antigen binding is pH-dependent, antibody and / or antigen half-life can be modified, for example, the half-life of an IgG2 antibody can be shortened if antigen-mediated clearance mechanisms normally degrade antibodies that bind to antigen. Similarly, antigen:antibody complexes can affect antigen half-life by either extending the half-life by protecting the antigen from typical degradation processes or shortening the half-life via antibody-mediated degradation (target-mediated drug processing). An antibody may have a higher affinity for an antigen at pH 7.4 compared to endosomal pH (i.e., pH 5.5-6.0), such that the KD ratio at pH 5.5 / pH 7.4 or pH 6.0 / pH 7.4 is greater than or equal to 2. For example, to enhance the pharmacokinetic (PK) and pharmacodynamic (PD) properties of the antibody, it is possible to engineer pH-sensitive binding into the antibody by introducing histidine residues into the CDRs.

[0210] Recycling antibodies are engineered so that a single recycling antibody molecule can bind to an antigen multiple times (e.g., more than once). Recycling antibodies can dissociate from antigens under acidic conditions within cells. For example, recycling antibodies bound to membrane-bound antigens dissociate from the antigen in a pH-dependent manner. The dissociated recycling antibodies are recycled by FcRn while the antigen is transported to lysosomes and degraded. This mechanism allows recycling antibodies to repeatedly bind to antigens in plasma, reducing antibody clearance.

[0211] Sweeping antibodies incorporate two antibody engineering techniques. One is variable region engineering (the "pH switch" mentioned above) that allows the antibody to bind to antigen in plasma and dissociate from the antigen in endosomes (where the antigen is subsequently degraded in lysosomes). The other is by engineering the constant region to increase cellular uptake of the antibody-antigen complex into endosomes, mediated, for example, via FcRn, FcγRIIb, or possibly other surface receptors. Sweeping antibodies can therapeutically target soluble antigens and promote their clearance from circulation (Igawa T et al.,Sweeping antibody as a novel therapeutic antibody modality capable of eliminating soluble antigens from circulation. Immunol Rev.2016 Mar;270(1):132-51). Ikawa et al. developed a panel of Fc mutants with enhanced binding to FcRn. A panel of Fc mutants, including M252Y, V308P, and N434Y, have enhanced binding to FcRn, which, together with pH-dependent binding to the target soluble IL-6 receptor, enhances clearance of the target antigen compared to wild-type Fc antibody mutants (PLoS One.8(5):e63236,2013). FcγRIIb can be used to accelerate the rate of intracellular uptake of antibody-antigen complexes. Ikawa et al. developed an FcγRIIb sweeping antibody, in which the Fc region of a pH-dependent antibody is engineered to selectively increase human FcγRIIb binding, enhancing the rate of uptake of antibody-antigen complexes. An important role of this inhibitory receptor is the uptake of antibody-antigen complexes into liver endothelial cells (LSECs), and therefore, if cellular uptake of antibody-antigen complexes is mediated by human FcγRIIb, it may reduce antigen concentrations in the circulation. Using global mutagenesis of the Fc domain of human IgG, we identified an Fc variant (v12) containing the mutations E233D / G237D / P238D / H268D / P271G / A330R, which has selectively increased binding affinity for human FcγRIIb.Studies in human FcγRIIb transgenic mice revealed that the v12 variant accelerated hsIL-6R clearance more than a pH-dependent sweeping antibody with wild-type hIgG1 while maintaining comparable pharmacokinetics.

[0212] Also provided are immunoconjugates (also interchangeably referred to as "antibody-drug conjugates," "ADCs," or "antigen binding protein-drug conjugates") that comprise an antigen binding protein as described herein conjugated to one or more drugs (e.g., a cytotoxic drug, a chemotherapeutic agent, an immunotherapeutic agent, a growth inhibitory agent, a toxin (e.g., a protein toxin such as an enzymatically active toxin of bacterial, fungal, plant, or animal origin, or a fragment thereof), an antiviral agent, a radioisotope (e.g., a radioconjugate), an antibiotic, or a small interfering RNA (siRNA)).

[0213] Immunoconjugates have been used in the treatment of cancer for the local delivery of cytotoxic drugs, i.e. drugs that kill cells or inhibit cell growth or proliferation. Immunoconjugates allow targeted delivery of drug moieties to tumors in particular, allowing intracellular accumulation in tumors where systemic administration of unconjugated drugs may result in unacceptable levels of toxicity to normal cells. Immunoconjugates allow selective delivery of potent cytotoxic payloads to targeted cancer cells, resulting in improved efficacy, reduced systemic toxicity, and favorable pharmacokinetics (PK) / pharmacodynamics (PD) and biodistribution compared to traditional chemotherapy.

[0214] Both polyclonal and monoclonal antibodies can be useful in these strategies. Drugs used in such methods include daunomycin, doxorubicin, methotrexate, and vindesine. Toxins used in antibody-toxin conjugates include bacterial toxins such as diphtheria toxin, plant toxins such as ricin, small molecule toxins such as geldanamycin, maytansinoids (EP1391213), and calicheamicin.

[0215] An immunoconjugate can include an antigen binding protein, such as an antibody, and a drug, such as a chemotherapeutic agent, a toxin, etc. The drug can be modified (e.g., via standard synthetic chemistry) to allow for the drug to be chemically coupled (e.g., to contain a reactive means that allows for chemical coupling) to a reactive end of a linker that attaches the drug to the antigen binding protein.

[0216] Also provided is a method of treating cancer in a subject in need thereof, comprising administering to said subject a therapeutically effective amount of a ULBP6 binding protein or pharmaceutical composition as defined herein. The subject may be an animal or a human. The subject may be a human. In some embodiments, the subject may be a patient.

[0217] Cancers that can be treated include adenocarcinoma, basal cell carcinoma, squamous cell carcinoma, adenosquamous carcinoma, carcinosarcoma, melanoma, adrenal carcinoma, adrenal cortical carcinoma, pheochromocytoma, breast cancer, ductal carcinoma in situ, lobular carcinoma, inflammatory breast cancer, invasive ductal carcinoma, Paget's disease of the nipple, papillary carcinoma, medullary carcinoma, breast cancer, anal cancer, cloacal carcinoma, rectal melanoma, appendix cancer, appendix neuroendocrine tumor, appendix mucinous cystadenocarcinoma, colonic adenocarcinoma, signet ring cell carcinoma, goblet cell carcinoma / adenoendocrine carcinoma, cholangiocarcinoma, intrahepatic cholangiocarcinoma, extrahepatic cholangiocarcinoma, extrahepatic cholangiocarcinoma, perihilar cholangiocarcinoma, distal extrahepatic cholangiocarcinoma, colorectal cancer, mucinous adenocarcinoma, gastrointestinal carcinoid tumor, gastrointestinal stromal tumor tumor, primary colorectal lymphoma, leiomyosarcoma, esophageal cancer, esophageal squamous cell carcinoma, small cell carcinoma, leiomyoma, gallbladder cancer, nonpapillary adenocarcinoma, papillary adenocarcinoma, gastric cancer, gastric adenocarcinoma, liver cancer, hepatocellular carcinoma, fibromembranous carcinoma, angiosarcoma, lymphangiosarcoma, angiosarcoma, hepatoblastoma, pancreatic cancer, pancreatic adenocarcinoma, ductal adenocarcinoma, acinar adenocarcinoma, acinar cell carcinoma, gelatinous carcinoma, giant cell tumor, hepatoid carcinoma, mucinous cystic neoplasm, pancreatoblastoma, serous cystadenoma, intraductal papillary mucinous neoplasm, pancreatic neuroendocrine tumor, gastrinoma, insulinoma, glucagonoma, vipoma, somatostatinoma, pancreatic polypeptide-secreting tumor (PPoma), small intestine cancer, eye cancer, intraocular melanoma tumor, intraocular lymphoma, intraocular retinoblastoma, conjunctival melanoma, eyelid cancer, sebaceous gland carcinoma, lacrimal gland tumor, malignant mixed epithelial tumor, adenoid cystic carcinoma, bladder cancer, urothelial carcinoma, kidney cancer, renal cell carcinoma (RCC), clear cell RCC, papillary RCC, chromophobe cell RCC, collecting duct RCC, multilocular cystic RCC, renal mucinous tubular cell and spindle cell carcinoma, tubulocystic RCC, thyroidoid follicular RCC, acquired cystic kidney disease-associated RCC, RCC with t(6;11) translocation (TFEB), hybrid oncocytoma / chromophobe cell RCC, Wilms' tumor, penile cancer, prostate cancer, castration-resistant prostate cancer, transitional cell carcinoma, Testicular cancer, Seminoma, Classical seminoma, Spermatocyte seminoma, Nonseminoma, Germ cell tumor, Yolk sac carcinoma, Choriocarcinoma, Teratoma, Leydig cell tumor, Sertoli cell tumor, Rete testis carcinoma, Urethral cancer, Extracranial germ cell tumor, Germinoma, Gonadoblastoma, Mixed germ cell tumor, Extragonadal germ cell tumor, Endodermal sinus tumor, Cervical cancer, Cervical squamous cell carcinoma, Endometrial cancer, Ovarian cancer, Fallopian tube cancer, Epithelial carcinoma, Dysgerminoma, Sex cord stromal tumor, Gestational trophoblastic tumor, Primary peritoneal cancer, Uterine sarcoma, Uterine papillary serous carcinoma, Vaginal cancer, Clear cell adenocarcinoma, Vulvar cancer, Verrucous carcinoma, Head and neck cancer, Head and neck squamous cell carcinoma, Pharyngeal cancer, Hypopharyngeal cancer, Nasopharyngeal cancer,Oropharyngeal cancer, non-keratinizing squamous cell carcinoma, undifferentiated carcinoma, laryngeal cancer, oral cancer, oral cavity cancer, mucoepidermoid carcinoma, paranasal sinus and nasal cavity cancer, nasal neuroblastoma, salivary gland cancer, epithelial-myoepithelial carcinoma, parathyroid cancer, thyroid cancer, papillary thyroid cancer, follicular thyroid cancer, Hürthle cell carcinoma, medullary thyroid cancer, anaplastic thyroid cancer, paraganglioma, carotid paraganglioma, carotid paratympanic paraganglioma, vagus paraganglioma, leukemia, acute lymphoblastic leukemia, T lymphoblastic leukemia, precursor B-cell lymphoblastic leukemia, acute myeloid leukemia, acute promyelocytic leukemia, acute myelomonocytic leukemia, acute monocytic leukemia, acute megakaryoblastic leukemia, erythroblastic leukemia, Leukemia, chronic lymphocytic leukemia, B-cell chronic lymphocytic leukemia, B-cell prolymphocytic leukemia, T-cell prolymphocytic leukemia, large granular lymphocytic leukemia, T-cell large granular lymphocytic leukemia, NK-cell granular lymphocytic leukemia, hairy cell leukemia, chronic myelogenous leukemia, chronic myelomonocytic leukemia, chronic neutrophilic leukemia, chronic eosinophilic leukemia, plasma cell leukemia, lymphoma, Hodgkin lymphoma, classical Hodgkin lymphoma, nodular sclerosis classical Hodgkin lymphoma, mixed cell classical Hodgkin lymphoma, lymphocyte-rich classical Hodgkin lymphoma, lymphocyte-depleted classical Hodgkin lymphoma lymphoma, nodal lymphocyte predominant Hodgkin lymphoma, non-Hodgkin lymphoma, diffuse large B-cell lymphoma, primary mediastinal B-cell lymphoma, primary effusion lymphoma, T-cell / histiocyte-rich large B-cell lymphoma, lymphoplasmacytic lymphoma, lymphoblastic lymphoma, small lymphocytic lymphoma, double-hit / triple-hit lymphoma, Burkitt lymphoma, Burkitt-like lymphoma, small noncleaved cell lymphoma, follicular lymphoma, follicular large cell lymphoma, immunoblastic lymphoma, intravascular large cell lymphoma, primary splenic lymphoma, anaplastic large cell lymphoma, mantle cell lymphoma , marginal zone lymphoma (MZL), extranodal MZL, nodal MZL, splenic MZL, splenic MZL with villous lymphocytes, peripheral T-cell lymphoma, angioimmunoblastic T-cell lymphoma, adult T-cell lymphoma / leukemia, extranodal NK / T-cell lymphoma, enteropathy-associated T-cell lymphoma, hepatosplenic T-cell lymphoma, subcutaneous panniculitis-like T-cell lymphoma, unspecified T-cell non-Hodgkin's lymphoma, gamma / delta T-cell lymphoma, mucosa-associated lymphoid tissue lymphoma, post-transplant lymphoproliferative disorder, HIV-associated lymphoma, Langerhans cell histiocytosis, multiple myeloma, smoldering multiple myeloma, active multiple myeloma,Plasmacytoma, Isolated plasmacytoma of bone, Extramedullary plasmacytoma, Primary amyloidosis, Myelodysplastic syndrome, Refractory anemia, Refractory anemia due to ringed sideroblasts, Refractory anemia due to excess blasts, Refractory anemia due to excess blasts in transformation, Myeloproliferative neoplasms, Polycythemia vera, Essential thrombocythemia, Myelofibrosis, Systemic mastocytosis, Bone cancer, Ewing's sarcoma, Osteosarcoma, Intramedullary osteosarcoma, Parosteal osteosarcoma, Extraskeletal osteosarcoma, Malignant fibrous osteocytoma, Notochord chondrosarcoma, classical chordoma, chondroitin chordoma, dedifferentiated chordoma, chondrosarcoma, conventional chondrosarcoma, clear cell chondrosarcoma, myxoid chondrosarcoma, mesenchymal chondrosarcoma, dedifferentiated chondrosarcoma, rhabdomyosarcoma, embryonal rhabdomyosarcoma, alveolar rhabdomyosarcoma, botryoid rhabdomyosarcoma, pleomorphic rhabdomyosarcoma, sarcoma of soft tissue, extraskeletal sarcoma, dermatofibrosarcoma protuberans, epithelioid sarcoma, Kaposi's sarcoma, liposarcoma, malignant peripheral nerve sheath tumor, fibrosarcoma, myxosarcoma, synovial tumor, brain cancer, undifferentiated astrocytic tumor alveolar cyst, glioblastoma, glioblastoma multiforme, meningioma, pituitary carcinoma, schwannoma, oligodendroglioma, ependymoma, medulloblastoma, astrocytoma, brain stem glioma, atypical teratoid / rhabdoid tumor, pinealoma, neuroblastoma, primary CNS lymphoma, primitive neuroectodermal tumor, diffuse intrinsic pontine glioma, lung cancer, non-small cell lung cancer (NSCLC), NSCLC undifferentiated, small cell lung carcinoma, lung squamous cell carcinoma, pleuropulmonary blastoma, bronchial carcinoma, malignant mesothelioma, malignant pleural mesothelioma, malignant These include peritoneal mesothelioma, thymoma, thymic carcinoma, keratoacanthoma, sebaceous gland carcinoma, sweat gland carcinoma, apocrine gland carcinoma, eccrine gland carcinoma, clear cell eccrine gland carcinoma, Merkel cell carcinoma, cutaneous T-cell lymphoma, mycosis fungoides, Sézary syndrome, chondroid syringoma, HPV-associated cancer, tumors containing transformed cells, tumors containing precancerous cells, precancerous hyperplasia, precancerous metaplasia, precancerous dysplasia, carcinoma in situ, mixed tumors, malignant mixed tumors, and composite cancers.

[0218] Cancers that can be treated include those with high tumor mutation count (TMB), defective DNA mismatch repair system (dMMR), microsatellite instability high (MSI-H), microsatellite instability low (MSI-L), microsatellite mutations elevated in alternative tetranucleotide repeats (EMAST), microsatellite stable (MSS) cancers, polymerase delta (POLD) mutations, polymerase epsilon (POLE) mutations, or homologous recombination repair deficient (HRD) cancers.

[0219] Cancers that can be treated include breast cancers defined by expression profiling (triple negative breast cancer, HER2 positive breast cancer, luminal A breast cancer, luminal B breast cancer, normal breast cancer) or breast cancers with BRCA1 or BRCA2 mutations.

[0220] The cancer to be treated may express cell surface ULBP6 or sULBP6. sULBP6 released from cancer cells may create an immunosuppressive environment. Thus, the ULBP6 binding protein described herein may be effective in treating cancers that express sULBP6. Thus, cancers that can be treated include cancers that have been shown to express sULBP6. Cancers that can be treated include head and neck cancer, cervical cancer, bladder cancer, lung cancer, colorectal cancer, pancreatic cancer, or glioblastoma. Cancers that can be treated include squamous cell carcinoma. The cancer to be treated may be head and neck squamous cell carcinoma (HNSCC). The cancer to be treated may be lung cancer. The cancer to be treated may be lung cancer. The lung cancer may be NSCLC.

[0221] Genetic data described in Example 1 show that single nucleotide polymorphisms in the ULBP6 locus are associated with susceptibility to basal cell carcinoma diagnosis in humans. Thus, ULBP6 antagonists may be useful for treating basal cell carcinoma. The ULBP6 binding proteins described herein may be useful for treating basal cell carcinoma.

[0222] The ULBP6 binding protein described herein may bind to cell surface ULBP6. For example, the ULBP6 binding protein may bind to ULBP6 expressed (endogenously or otherwise) on the surface of tumor cells or tumor cell lines, such as CHO UBLP6-expressing cells, COV644 cells, or PANC02.13 cells. The half-maximal effective concentration (EC50) of the ULBP6 binding protein that binds to cell surface ULBP6 can be determined using the FACS binding assay described in Example 7. The ULBP6 binding protein may bind to PANC02.13 cells and COV644 cells with an EC50 of 2 nM or less, or 1 nM or less.

[0223] The ULBP6 binding proteins described herein can inhibit the binding of NKG2D-Fc to cell surface ULBP6. In some examples, the ULBP6 binding proteins can inhibit the binding of NKG2D-Fc to cell surface ULBP6 with an IC50 of 1 nM or less, 0.5 nM or less, 0.25 nM or less, or 0.2 nM or less.

[0224] Soluble ULBP6 can reduce NKG2D-mediated cell activation by binding to NKG2D, thereby inhibiting the binding of other NKG2DLs to NKG2D. sULBP6 can reduce NKG2D-mediated cell activation by binding to NKG2D, thereby inducing NKG2D internalization, or can block NKG2D-mediated activation. sULBP6 can reduce NKG2D-mediated cell activation by binding to NKG2D, thereby inducing downregulation of NKG2D, or can block NKG2D-mediated activation. The ULBP6 binding protein described herein can increase NKG2D-mediated cell activation by inhibiting sULBP6-induced NKG2D blocking. Alternatively, the ULBP6 binding protein can increase NKG2D-mediated cell activation by inhibiting sULBP6-induced NKG2D internalization. Alternatively, ULBP6-binding proteins may increase NKG2D-mediated cell activation by inhibiting sULBP6-induced NKG2D downregulation.

[0225] In some embodiments, the ULBP6 binding proteins provided herein may target tumor immune response evasion mechanisms. Immune evasion through MHC loss may lead to vulnerability to NK-mediated killing. This may be driven by antigen immunogenicity, such as novel antigens derived from viral and / or tumor genetic alterations. Immune evasion may thus occur via soluble NKG2D ligands (e.g., ULBP6). In some embodiments, the ULBP6 binding proteins provided herein may interfere with such immune evasion. In some instances, tumor expression of ULBP2 / 5 / 6+ secondary NKG2D ligands may trigger drug activity (e.g., affinity-prioritized MICA / B) in NK-infiltrated tumors. In some embodiments, soluble ULBPs may be a readout from ULBP2 / 5 / 6+ tumor burden, and thus may be an alternative to IHC in some cases.

[0226] The ULBP6 binding proteins described herein can inhibit the binding of sULBP6 to NKG2D, thereby increasing the activation of NKG2D through interaction with other NKG2DLs such as MICA. The assay described in Example 9 can be used to measure MICA-mediated cell activation.

[0227] In addition to inhibiting the binding of sULBP6 to NKG2D, the ULBP6 binding protein described herein may also include a modified Fc region to enhance ADCC. The dual mechanisms of 1) inhibition of sULBP6 binding to NKG2D and 2) Fc receptor-mediated killing of ULBP6-expressing tumor cells may synergize to effectively treat cancer. ULBP6 binding protein enhanced by ADCC may synergistically increase the levels of cytokines (e.g., IFNy) secreted by peripheral blood mononuclear cells (PBMCs) co-cultured with ULBP6-expressing tumor cells compared to ULBP6 binding protein lacking Fc effector function. ULBP6 binding protein may inhibit the binding of sULBP6 to NKG2D, thereby increasing tumor cell killing in vitro.

[0228] The ULBP6 binding protein described herein can bind to human ULBP6 and interfere with the interaction of ULBP6 with NKG2D. The NKG2D binding interface of ULBP6 includes residues of ULBP6 that can interact with residues from NKG2D. The ULBP6 binding protein can interfere with the interaction of ULBP6 with NKG2D by interacting with residues contained within or proximal to the NKG2D binding interface of ULBP6. The HDX assay described in Example 17 can be used to characterize the interaction of ULBP6 with anti-ULBP6 antibodies. The ULBP6 binding protein can bind to one or more residues contained within alpha helix 1 (SEQ ID NO: 125) of human ULBP6 and / or one or more residues contained within alpha helix 2 (SEQ ID NO: 126) of human ULBP6. The ULBP6 binding protein can bind to one or more residues contained within SEQ ID NO: 128 and / or SEQ ID NO: 129 of human ULBP6. The ULBP6 binding protein may bind to a conformational epitope contained within alpha helix 1 (SEQ ID NO: 125) and / or alpha helix 2 (SEQ ID NO: 126) of human ULBP6. The ULBP6 binding protein may bind to a conformational epitope contained within SEQ ID NO: 128 and / or SEQ ID NO: 129 of human ULBP6. The ULBP6 binding protein may protect one or more residues contained within alpha helix 1 (SEQ ID NO: 125) and / or alpha helix 2 (SEQ ID NO: 126) of human ULBP6 from deuterium exchange in HDX-MS analysis. The ULBP6 binding protein may protect one or more residues contained within SEQ ID NO: 128 and / or SEQ ID NO: 129 of human ULBP6 from deuterium exchange in HDX-MS analysis. The ULBP6 binding protein can bind to human ULBP6 and result in a peptide derived from alpha helix 1 (SEQ ID NO: 125) and / or alpha helix 2 (SEQ ID NO: 126) of human ULBP6 that is less susceptible to deuterium incorporation than the corresponding peptide derived from uncomplexed human ULBP6.The ULBP6-binding protein may bind to human ULBP6, resulting in a peptide derived from SEQ ID NO: 128 and / or SEQ ID NO: 129 of human ULBP6 that incorporates less deuterium than the corresponding peptide derived from uncomplexed human ULBP6. The ULBP6-binding protein may bind to the NKG2D-binding interface of human ULBP6. The ULBP6-binding protein may interfere with the interaction of human ULBP6 to NKG2D by binding to one or more residues selected from E96, E154, Y184, and D189 located within the NKG2D-binding interface of human ULBP6.

[0229] The ULBP6 binding proteins described herein may bind to one or more residues contained within alpha helix 1 (SEQ ID NO: 125) of human ULBP6, one or more residues contained within alpha helix 2 (SEQ ID NO: 126) of human ULBP6, and / or one or more residues contained within SEQ ID NO: 127 of human ULBP6. The ULBP6 binding proteins may bind to one or more residues contained within SEQ ID NO: 128, SEQ ID NO: 129, and / or SEQ ID NO: 127 of human ULBP6. The ULBP6 binding proteins may bind to a conformational epitope contained within alpha helix 1 (SEQ ID NO: 125), alpha helix 2 (SEQ ID NO: 126), and / or SEQ ID NO: 127 of human ULBP6. The ULBP6 binding proteins may bind to a conformational epitope contained within SEQ ID NO: 128, SEQ ID NO: 129, and / or SEQ ID NO: 127 of human ULBP6. The ULBP6 binding protein may protect one or more residues contained within alpha helix 1 (SEQ ID NO: 125), alpha helix 2 (SEQ ID NO: 126), and / or SEQ ID NO: 127 of human ULBP6 from deuterium exchange in HDX-MS analysis. The ULBP6 binding protein may protect one or more residues contained within SEQ ID NO: 128, SEQ ID NO: 129, and / or SEQ ID NO: 127 of human ULBP6 from deuterium exchange in HDX-MS analysis. The ULBP6 binding protein may bind to human ULBP6, resulting in peptides derived from alpha helix 1 (SEQ ID NO: 125), alpha helix 2 (SEQ ID NO: 126), and / or SEQ ID NO: 127 of human ULBP6 that incorporate less deuterium than the corresponding peptides derived from uncomplexed human ULBP6. The ULBP6 binding protein may bind to human ULBP6, resulting in peptides derived from SEQ ID NO: 128, SEQ ID NO: 129, and / or SEQ ID NO: 127 of human ULBP6 that incorporate less deuterium than the corresponding peptides derived from uncomplexed human ULBP6.

[0230] Also provided herein are antibodies and antigen-binding fragments thereof that bind to UL-16 binding protein 6 (ULBP6), the antibodies and antigen-binding fragments thereof comprising a variable heavy chain (VH) sequence having a heavy chain complementarity determining region 1 (CDRH1) sequence, a heavy chain complementarity determining region 2 (CDRH2) sequence, and a heavy chain complementarity determining region 3 (CDRH3) sequence, and a variable light chain (VL) sequence having a light chain complementarity determining region 1 (CDRL1) sequence, a light chain complementarity determining region 2 (CDRL2) sequence, and a light chain complementarity determining region 3 (CDRL3) sequence; wherein CDRH1 is selected from the group consisting of SEQ ID NOs: 36, 39, 42, and 45, CDRH2 is selected from the group consisting of SEQ ID NOs: 37, 40, 43, and 46, CDRH3 is selected from the group consisting of SEQ ID NOs: 38, 41, 44, and 47, CDRL1 is selected from the group consisting of SEQ ID NOs: 51, 54, 57, and 60, CDRL2 is selected from the group consisting of SEQ ID NOs: 52, 55, 58, and 61, or CDRL3 is selected from the group consisting of SEQ ID NOs: 53, 56, 59, or 62.

[0231] Also provided herein are antibodies and antigen-binding fragments thereof, in which CDRH1 is SEQ ID NO:36, CDRH2 is SEQ ID NO:37, CDRH3 is SEQ ID NO:38, CDRL1 is SEQ ID NO:51, CDRL2 is SEQ ID NO:52, and CDRL3 is SEQ ID NO:53, or CDRH1 is SEQ ID NO:39, CDRH2 is SEQ ID NO:40, CDRH3 is SEQ ID NO:41, CDRL1 is SEQ ID NO:54, CDRL2 is SEQ ID NO:55, and CDRL3 is SEQ ID NO:56, or CDRH1 is SEQ ID NO:42, CDRH2 is SEQ ID NO:43, CDRH3 is SEQ ID NO:44, CDRL1 is SEQ ID NO:57, CDRL2 is SEQ ID NO:58, and CDRL3 is SEQ ID NO:59, or CDRH1 is SEQ ID NO:45, CDRH2 is SEQ ID NO:46, CDRH3 is SEQ ID NO:47, CDRL1 is SEQ ID NO:60, CDRL2 is SEQ ID NO:61, and CDRL3 is SEQ ID NO:62.

[0232] Also provided herein are antibodies and antigen-binding fragments thereof, wherein the VH region is at least 80% identical to SEQ ID NO:21 and the VL region is at least 80% identical to SEQ ID NO:31; or the VH region is at least 80% identical to SEQ ID NO:22 and the VL region is at least 80% identical to SEQ ID NO:32; or the VH region is at least 80% identical to SEQ ID NO:23 and the VL region is at least 80% identical to SEQ ID NO:33; or the VH region is at least 80% identical to SEQ ID NO:24 and the VL region is at least 80% identical to SEQ ID NO:34.

[0233] Also provided herein are antibodies and antigen-binding fragments thereof, wherein the VH region is at least 80% identical to SEQ ID NO:16 and the VL region is at least 80% identical to SEQ ID NO:26, or the VH region is at least 80% identical to SEQ ID NO:17 and the VL region is at least 80% identical to SEQ ID NO:27, or the VH region is at least 80% identical to SEQ ID NO:18 and the VL region is at least 80% identical to SEQ ID NO:28, or the VH region is at least 80% identical to SEQ ID NO:19 and the VL region is at least 80% identical to SEQ ID NO:29.

[0234] Also provided herein are antibodies and antigen-binding fragments thereof, in which the binding protein comprises a heavy chain (HC) sequence that is at least 80% identical to any one of SEQ ID NOs: 75-78, 85-88, or 95-98, and / or a light chain (LC) sequence that is at least 80% identical to any one of SEQ ID NOs: 105-108.

[0235] Also provided herein are antibodies and antigen-binding fragments thereof, which bind to ULBP6 and ULBP2 with a KD of 1 nM or less, and bind to ULBP5 with a KD of 1 nM or more.

[0236] Also provided herein are antibodies and antigen-binding fragments thereof that inhibit the interaction between human ULBP6 and human NKG2D by 80% or more.

[0237] Also provided herein are antibodies and antigen-binding fragments thereof that bind to human ULBP6 protein or a fragment thereof and compete for binding to human ULBP6 protein or a fragment thereof with a reference ULBP6 binding protein or a fragment thereof comprising a VH region sequence of SEQ ID NO:21 and a VL region sequence of SEQ ID NO:31 or comprising a VH region sequence of SEQ ID NO:22 and a VL region sequence of SEQ ID NO:32.

[0238] Also provided herein are antibodies and antigen-binding fragments thereof that inhibit the binding of soluble ULBP6 to NKG2D.

[0239] Also provided herein are antibodies and antigen-binding fragments thereof that inhibit the binding of soluble ULBP6 to NKG2D by 80% or more.

[0240] Also provided herein are antibodies and antigen-binding fragments thereof that comprise an altered Fc region.

[0241] Also provided herein are antibodies and antigen-binding fragments thereof comprising a modified Fc region, the modified Fc region comprising the amino acid substitutions S239D and I332E (numbered according to the EU index).

[0242] Also provided herein are antibodies and antigen-binding fragments thereof that are defucosylated.

[0243] Also provided herein are antibodies and antigen-binding fragments thereof, in which the modified Fc region comprises a set of amino acid substitutions selected from the group consisting of S239D / A330L / I332E, G236A / S239D / A330L / I332E, L235V / F243L / R292P / Y300L / P396L, and F243L / R292P / Y300L / V305I / P396L (numbered according to the EU index).

[0244] Also provided herein are nucleic acid sequences encoding one or both of the heavy chain (HC) and light chain (LC) of the antibodies or antigen-binding fragments thereof provided herein.

[0245] Also provided herein are expression vectors, which contain nucleic acids of the antibodies and antigen-binding fragments thereof provided herein.

[0246] Also provided herein are recombinant host cells, which contain the nucleic acid sequences of the antibodies and antigen-binding fragments thereof provided herein.

[0247] Also provided herein is a method for producing an antibody or antigen-binding fragment thereof provided herein, the method comprising culturing a recombinant host cell under conditions suitable for expression of the nucleic acid sequence or vector contained therein, thereby producing a polypeptide comprising the antibody or antigen-binding fragment thereof.

[0248] Also provided herein are cell lines engineered to express the antibodies and / or antigen-binding fragments thereof provided herein.

[0249] Also provided herein are pharmaceutical compositions comprising one or more of the antibodies or antigen-binding fragments thereof provided herein, and a pharma- ceutically acceptable excipient.

[0250] Also provided herein is a method of treating cancer in a subject in need thereof, the method comprising administering to the subject a therapeutically effective amount of an antibody or antigen-binding fragment thereof provided herein, or a therapeutically effective amount of a pharmaceutical composition provided herein.

[0251] Also provided herein are methods of treating cancer, wherein the cancer is basal cell carcinoma, head and neck cancer, cervical cancer, bladder cancer, lung cancer, colorectal cancer, pancreatic cancer, or glioblastoma. In some such methods, the subject is a human.

[0252] Some of the methods provided herein may further include confirming that the subject expresses soluble ULBP6.

[0253] Also provided herein are kits comprising one or more of the antibodies or antigen-binding fragments thereof provided herein and instructions for administering the one or more antibodies or antigen-binding fragments thereof to a subject with cancer.

[0254] Some such kits may be for subjects with head and neck cancer, cervical cancer, bladder cancer, lung cancer, colorectal cancer, pancreatic cancer, or glioblastoma.

[0255] Also provided herein are ULBP6 antagonists for the treatment of basal cell carcinoma.

[0256] Also provided herein is a ULB6 binding protein and fragments thereof, the ULB6 binding protein and fragments thereof comprising: a) HVRH1 of SEQ ID NO: 130, HVRH2 of SEQ ID NO: 131, HVRH3 of SEQ ID NO: 132, HVRL1 of SEQ ID NO: 145, HVRL2 of SEQ ID NO: 146, and HVRL3 of SEQ ID NO: 147; b) HVRH1 of SEQ ID NO: 133, HVRH2 of SEQ ID NO: 134, HVRH3 of SEQ ID NO: 135, HVRL1 of SEQ ID NO: 148, HVRL2 of SEQ ID NO: 149, and HVRL3 of SEQ ID NO: 150; or c) HVRH1 of SEQ ID NO: 136, HVRH2 of SEQ ID NO: 137, HVRL1 of SEQ ID NO: 149, and HVRL3 of SEQ ID NO: 150. 37, HVRH2 of SEQ ID NO: 138, HVRL1 of SEQ ID NO: 151, HVRL2 of SEQ ID NO: 152, and HVRL3 of SEQ ID NO: 153, or d) HVRH1 of SEQ ID NO: 139, HVRH2 of SEQ ID NO: 140, HVRH3 of SEQ ID NO: 141, HVRL1 of SEQ ID NO: 154, HVRL2 of SEQ ID NO: 155, and HVRL3 of SEQ ID NO: 156, or HVRH1 of SEQ ID NO: 142, HVRH2 of SEQ ID NO: 143, HVRH3 of SEQ ID NO: 144, HVRL1 of SEQ ID NO: 157, HVRL2 of SEQ ID NO: 158, and HVRL3 of SEQ ID NO: 159. EXAMPLES

[0257] Various features and embodiments of the present disclosure are illustrated in the following representative examples, which are intended to be illustrative and not limiting. As one skilled in the art will readily appreciate, the specific examples are merely illustrative of the invention, which is more fully described in the claims that follow. It should be understood that the embodiments and features described in this application are interchangeable and combinable with at least other such embodiments and features contained herein.

[0258] Example 1: Genome-wide association study in individuals with basal cell carcinoma

[0259] A genome-wide association study (GWAS) was conducted for individuals diagnosed with basal cell carcinoma. The accepted genome-wide statistical threshold was P<5.0E10. -8 We identified a genetic region (6q25.1) within the ULBP6 locus reaching

[0260] The inventors have identified a common genetic association between basal cell carcinoma and immune-related disease diagnosis in the 6q25.1 region, where multiple NKG2D ligands are located: ULBP6 (also known as RAET1L), ULBP1, ULBP2, ULBP3, ULBP4 (RAET1E), and ULBP5 (RAET1G). The index mutations are rs9478354 (dbSNP build 153 identifier) ​​located on chromosome 6 at position 150013230 (in human genome build GRCh38 / hg38) (basal cell carcinoma; A / G, AF [allele frequency of allele G] = 0.55 or = 0.965, P value = 7.2E-24), rs10666240 (dbSNP build 153 identifier) ​​located on chromosome 6 at position 150036968 (in human genome build GRCH38 / hg38), and rs20666240 (dbSNP build 153 identifier) ​​located on chromosome 6 at position 150036968 (in human genome build GRCH38 / hg38). The most common mutations were rs5006620 (dbSNP Assembly 153 identifier) ​​(alopecia universalis; C / T, AF [allelic frequency of allele T] = 0.45 or = 0.790, P value = 1.3E-21), located on chromosome 6 at position 150015529 (in human genome assembly GRCH38 / hg38) (Figure 1). These mutations are located in the terminal intergenic region of ULBP6.

[0261] Figure 1 is a regional association plot of ULBP6 on chromosome 6 for basal cell carcinoma. The feature analyzed is a diagnosis of basal cell carcinoma, where "cases" are individuals diagnosed with basal cell carcinoma and "controls" are individuals without a diagnosis of basal cell carcinoma. The x-axis indicates the physical location on the human genome construct (GRCh38 / hg38) and the left y-axis indicates the -log10 of the P-value associated with the phenotype of interest. Each point on the depicted plot represents a genetic variant tested for association in that region. The grey horizontal line indicates the genome-wide significance threshold of 5x10 -8 The human genes in this region are illustrated in the lower panel. These GWAS data indicate that the indicated loci (gene regions) are associated with susceptibility to basal cell carcinoma diagnosis in humans.

[0262] Example 2: ULBP6 Expression in Cancer Patients

[0263] To determine whether human ULBP6 and other NKG2DLs are expressed by tumor cells and therefore could be therapeutic targets in cancer, we took two complementary approaches: in the first approach, we investigated transcript expression by analysis of The Cancer Genome Atlas (TCGA) and single-cell RNA sequencing data, and in the second approach, we measured the levels of NKG2DL proteins in the serum of patients with various cancer types.

[0264] NKG2D Ligand Expression on Tumor Cells

[0265] Expression data from multiple tumor types from the TCGA program were analyzed for ULBP2 and ULBP6 expression. Overall, ULBP2 appears to be widely expressed across the cancer types analyzed, while ULBP6 was found to be most highly expressed in squamous cell carcinomas, including HNSC, CESC, LUSC, and ESCA (squamous cell carcinoma lesions), compared to non-squamous cell carcinomas. Both ULBP6_01 and ULBP6_02 haplotypes of ULBP6 were found to be expressed in tumor samples.

[0266] Additionally, single-cell RNA sequencing data from Puram et al. (PMID:29198524) were analyzed for ULBP2 and ULBP6 expression, suggesting that these ligands are specifically expressed in tumor cells.

[0267] Detection of NKG2DL in patient serum

[0268] To determine the levels of NKG2DL released into serum (ie, shed / soluble NKG2DL, "sNKG2DL"), ELISA and Luminex assays were performed to measure the presence of NKG2DL in patient serum samples.

[0269] Materials and Methods:

[0270] sNKG2DL was measured in the serum of 158 patients and in nine cancer types. Serum samples were collected from individuals with cervical, colorectal, bladder, lung, melanoma, pancreatic, head and neck, esophageal, and glioblastoma, as well as from individuals without cancer (healthy donors). ULBP1, ULBP2 / ULBP6, ULBP3, and MICB levels were measured in serum samples using ELISA kits from R&D systems according to the manufacturer's instructions. The ELISA does not distinguish between ULBP2 and ULBP6, as the detection antibody binds to both ULBP6 and ULBP2. MICA levels were measured in serum samples using the Luminex MICA assay according to the manufacturer's instructions. Each serum sample was screened twice (i.e., the method was repeated).

[0271] The sNKG2DL concentrations from healthy donors were averaged and used to calculate the cutoff value for determining sNKG2DL elevation in serum samples from cancer patients. The cutoff value for each sNKG2DL tested was set as the mean value of the sNKG2DL concentrations from healthy donors plus 3 standard deviations. Each sNKG2DL level from tumor samples was compared to the cutoff value. If the values ​​in both replicates of a serum sample were higher than the calculated cutoff value for that sNKG2DL, it was determined to be an elevated sNKG2DL level.

[0272] result:

[0273] As shown in Table 4, serum-released ULBP2 / 6 (sULBP2 / 6) was elevated in 22.2% of patients compared to control samples from healthy donors, while other sNKG2DLs (including ULBP1 and ULBP3 and MICA / B) were elevated in <5.2% of patients compared to control samples from healthy donors. These results indicate that sULBP2 / 6 was the most upregulated sNKG2DL in the tested sera from cancer patients.

[0274] [Table 4]

[0275] Example 3: Binding of human sNKG2DL to NKG2D

[0276] NKG2DLs show variation in primary amino acid sequence, structure, expression profile, and binding affinity to NKG2D, which may lead to functional differences between membrane-bound and soluble forms in interacting with NKG2D. Example 2 showed that sULBP2 and / or sULBP6 protein is commonly increased in the serum of cancer patients. To rank the relative binding affinities of sULBP2, sULBP6, and other human sNKG2D ligands to the hNKG2D receptor, surface plasmon resonance (SPR) assays were performed using a BIACORE 8K (Cytiva) instrument as described below.

[0277] Materials and Methods:

[0278] To demonstrate binding of human sNKG2D ligands to hNKG2D, SPR assays were performed at 37° C. using various human sNKG2D ligands, including sMICA, sMICB, sULBP1, sULBP2, sULBP3, sULBP4, sULBP5, sULBP6_01, and sULBP6_02. Briefly, 0.5 μg / ml hNKG2D-Fc in HBS-P buffer (0.01 M HEPES, pH 7.4, 0.15 M NaCl, 0.005% surfactant P20) was applied to a CYTIVA sensor chip Protein A at a flow rate of 30 μL / min to achieve approximately 50 response units in the second flow cell (FC2). FC1 was used for reference. For kinetic measurements, hNKG2D ligands were serially diluted 3-fold in HBS-P buffer (0.01 M HEPES, pH 7.4, 0.15 M NaCl, 0.005% surfactant P20) from high (50 nM for ULBP6_02, 2 μM for other sNKG2DLs) to low (0.07 nM for ULBP6_02, 2.74 nM for other sNKG2DLs) concentrations and injected at 37° C. (flow rate: 30 μL / min). Sensorgrams were recorded and evaluated by BIACORE 8K evaluation software (version 1.1.1.7442) after reference and buffer subtraction. Association rates (k on ) and dissociation rate (k off ) was calculated using a simple one-to-one Langmuir binding model. The equilibrium dissociation constant (KD) was k off / k on The ratios were calculated as follows (summarized in Table 5).

[0279] [Table 5]

[0280] result:

[0281] No binding of sULBP4 to NKG2D was observed, whereas the binding affinities of the remaining sNKG2D ligands ranged from 594 nM to 4.8 nM. The strongest binding affinity to NKG2D was observed for sULBP6. The binding affinity of haplotype 02 (sULBP6_02) was 4.8 nM, and that of haplotype 01 (sULBP6_01) was 41 nM (Table 5). For sULBP2, a much lower affinity of 220 nM to NKG2D was observed.

[0282] The relatively high affinity of the soluble form of ULBP6 for NKG2D shown herein, and the widespread presence of sULBP2 / ULBP6 in the serum of cancer patients (Example 2), indicate that inhibition of the sULBP6-NKG2D interaction using anti-ULBP6 binding proteins may provide advances in cancer therapy.

[0283] Example 4: Production of Polypeptides

[0284] This example describes the preparation of ULBP6 antigen for the generation of anti-ULBP6 antibodies. This example also includes the generation of other sNKG2DLs, including human sULBP1, sULBP2, sULBP3, sULBP4, sULBP5, sMICA, and sMICB, which are used as antigens in screening for anti-ULBP6 antibodies of the present disclosure. This example also includes the preparation of NKG2D and Fc tag polypeptides. Note that cULBP5 and cULBP2 / 6 refer to the soluble orthologs of cynomolgus monkeys. Due to the high level of sequence similarity between human ULBP2 and ULBP6, it is unclear whether the single cynomolgus monkey ortholog is evolutionarily related to human ULBP2 or ULBP6. Thus, the cynomolgus monkey ortholog is referred to herein as cULBP2 / 6.

[0285] The sequences of human sULBP1-6, cULBP5, and cULBP2 / 6 polypeptide forms are shown in Table 6, and the sequences of human NKG2D (hNKG2D) or cynomolgus NKG2D (cNKG2D), additional sNKG2DLs (sMICA and sMICB), and Fc polypeptide forms are shown in Table 7.

[0286] [Table 6-1] [Table 6-2] [Table 6-3]

[0287] [Table 7-1] [Table 7-2] [Table 7-3] [Table 7-4] [Table 7-5]

[0288] Example 5: Generation and selection of anti-ULBP6 antibodies

[0289] This example demonstrates the generation of anti-ULBP6 antibodies using a mouse hybridoma method.

[0290] immunization

[0291] Multi-strain mice were immunized with in-house produced hULBP6-01 (SEQ ID NO: 6) and hULBP6-02 (SEQ ID NO: 7). SIGMA adjuvant system (MILLIPORE SIGMA, St. Louis, MO), Freund's adjuvant complete (MILLIPORE SIGMA, St. Louis, MO), TITERMAX Gold adjuvant (MILLIPORE SIGMA, St. Louis, MO), GERBU adjuvant (V-Biognostics), TLR cocktail (INVIVOGEN) were used for immunization. Antibody titers were measured by ELISA. Mice that showed sufficient antibody titers were given a final boost without adjuvant. Fusions were performed 3 days after the final boost. Hybridomas were generated, selected and cloned for mouse antibody expression.

[0292] Example 6: Screening and characterization of anti-ULBP6 antibodies

[0293] This example describes the methods used to screen and select anti-ULBP6 antibodies that effectively block the interaction between soluble ULBP6 and NKG2D.

[0294] A. Enzyme-Linked Immunosorbent Assay (ELISA)

[0295] To detect single hybridoma clones secreting ULBP6-specific antibodies, direct ELISA was performed after 7 days of culturing the picked hybridoma clones.

[0296] Materials and Methods:

[0297] Briefly, 96-well MAXISORP flat-bottom plates (THERMO SCIENTIFIC, Cat. No. 439454) were coated with 1 μg / mL of sULBP6_01, sULBP6_02, and sULBP2 in PBS (50 μL / well) overnight at 4° C. The wells were washed twice with tap water and blocked with blocking solution (pH 7.4, 200 μL / well) containing 1% bovine serum albumin (BSA) in phosphate-buffered saline (PBS) for 1 h at room temperature. The wells were then washed again and 50 μL / well of culture supernatant from selected hybridomas was added and incubated for 1 h at room temperature. After washing, a secondary anti-mouse IgG antibody conjugated to horseradish peroxidase (HRP, Jackson Immuno115-035-164) was used for detection. The plates were incubated at room temperature for 45 min. Tetramethylbenzidine (TMB) solution (0.12 mg / mL TMB with 0.04% hydrogen peroxide in 25 mM NaH2PO4) was used as substrate. The reaction was stopped after 5 min with 2N H2SO4. Optical density (OD) was measured at 450 nm.

[0298] result:

[0299] From the primary direct ELISA, 191 clones (binders) were selected and confirmatory indirect ELISAs were performed yielding a total of 90 binders which were further tested using FACS binding and single point blocking ELISA assays as described below.

[0300] B. Flow cytometry cell binding assay

[0301] To determine whether the 90 ULBP6-specific binder antibody clones identified in the primary indirect ELISA assay also bind to cell surface-expressed NKG2DL, a flow cytometry cell-binding assay was performed.

[0302] Materials and Methods:

[0303] CHO-S cells were transiently transfected to overexpress full-length membrane-bound human ULBP6_01, ULBP6_02, ULBP1, ULBP2, ULBP3, and ULBP5. CHO-S cells were harvested by centrifugation at 1200 RPM for 15 minutes. Cell pellets were resuspended in 10 mL of FACS buffer (1x PBS containing 1% BSA) and counted. Cells were then plated at 1x10 in 96-well plates (Corning 3365). 6 Cells / well were seeded. 50 μL of supernatant was transferred from the 96-well plate used to culture hybridoma cells to the 96-well plate containing the target-expressing cells using a 96-head liquid handler. The supernatant was incubated with the cells for 1 hour at 4° C. The cells were then washed twice with FACS buffer, incubated with a secondary anti-mouse IgG antibody conjugated to APC (5 μg / mL, Jackson Immuno, 115-136-146) for 45 minutes at 4° C., and then washed twice with FACS buffer. After washing, the cells were resuspended in 100 μl of FACS buffer for flow cytometry analysis.

[0304] result:

[0305] Of the 90 clones tested in flow cytometry binding assays, 7 / 90 showed binding to CHO-S-ULBP1, 89 / 90 showed binding to CHO-S-ULBP2, 2 / 90 showed binding to CHO-S-ULBP3, 34 / 90 showed binding to CHO-S-ULBP5, 88 / 90 showed binding to CHO-S-ULBP6_01, and 87 / 90 showed binding to CHO-S-ULBP6_02. All 90 clones were then tested in single point biochemical blocking assays.

[0306] C. Single-point biochemical blocking assay

[0307] To detect antibody clones that block the binding of human soluble ULBP6 to human NKG2D, we performed biochemical blocking assays. In these experiments, we used sULBP6_02, the ULBP6 haplotype with the highest affinity for NKG2D.

[0308] Materials and Methods:

[0309] 96-well MAXISORP flat-bottom plates (THERMO SCIENTIFIC, Cat. No. 439454) were coated with 100 μL / well of neutravidin (1 μg / mL) in PBS overnight at 4° C. The wells were washed twice with tap water and blocked with blocking solution (pH 7.4, 200 μL / well used) containing 1% bovine serum albumin (BSA) in phosphate-buffered saline (PBS) for 1 h at room temperature. The wells were washed twice again and biotinylated sULBP6_02 was added at 1 μg / ml and the plate was incubated for 1 h at room temperature. 60 μl of undiluted hybridoma supernatant and 60 μl of NKG2D-Fc (10 nM) were preincubated and the mixed solution was added to the plate for 1 h at room temperature. After washing, a secondary anti-human IgG Fc antibody conjugated to horseradish peroxidase (HRP, Jackson Immuno. 109-035-098) was used for detection. The secondary antibody solution was incubated for 45 min at room temperature. Tetramethylbenzidine (TMB) solution (0.12 mg / mL TMB with 0.04% hydrogen peroxide in 25 mM NaH2PO4) was used as substrate. The reaction was stopped after 5 min with 2N H2SO4. The optical density (OD) was measured at 450 nm. Mouse anti-ULBP2 / 5 / 6 antibody (R&D SYSTEMS, MAB1298-100) was used at 5 μg / ml as a positive control. Culture medium was used as a negative control. The ELISA signal of the negative control was defined as 0% blocking activity. Mouse anti-hULBP2 / 5 / 6 antibody (R&D SYSTEMS, MAB1298-100) showed 72% blocking activity.

[0310] result:

[0311] Of the 90 clones tested in a single point biochemical blocking ELISA, 68 / 90 showed blocking of sULBP6_02 binding to NKG2D. Clones 7B3 and 8E11 showed 100% blocking activity. The 25 antibodies with the highest blocking activity were then tested for cross-reactivity in cynomolgus monkeys using cULBP2 / 6 from Ensembl v99.

[0312] D. Cynomolgus monkey cross-reactivity

[0313] To determine cross-reactivity to cULBP2 / 6 proteins, binding to both human and cynomolgus monkey sULBP6 proteins was tested by assessing binding of mouse hybridoma supernatants with the OCTET R-8 BLI label-free detection system.

[0314] Materials and Methods:

[0315] 300 μl of each hybridoma supernatant was pipetted into an Octet 96-well with 300 nM of both human ULBP6 and cULBP2 / 6 proteins in 1×PBST (1×PBS, 0.05% Tween®-20). A stable baseline for 60 seconds was measured in 1×PBST buffer using an anti-mouse Fc capture (AMC) biosensor, followed by capturing mouse antibodies from the hybridoma supernatant for 150-300 seconds. After another stable baseline was recorded for 60 seconds, the biosensor was immersed in either human soluble ULBP6 or cynomolgus ULBP2 / 6 to measure the interaction / association of captured mouse antibodies from the hybridoma supernatant with sULBP6, followed by a 300 second dissociation step in 1×PBST.

[0316] result:

[0317] 25 blocking antibodies from the single-point blocking ELISA were selected for cross-reactivity testing. Of these 25 antibodies, 10 antibodies showed good cross-reactivity to cynomolgus ULBP6. 7B3 showed the strongest binding to cULBP2 / 6. 2 / 10 blocking antibodies (7B3 and 8E11) showed very strong blocking and good cross-reactivity in the single-point blocking assay, so they were directly selected to be tested in functional assays, while the remaining 8 / 10 blocking antibodies were further tested in three-point biochemical blocking assays.

[0318] E. Three-point biochemical blocking assay

[0319] Eight hybridoma clones were selected based on clone type, blocking activity using single point undiluted supernatant, human / cynomolgus cross-reactivity, and binding affinity, and these eight clones were tested in a three-point blocking assay for further ranking.

[0320] Materials and Methods:

[0321] 96-well MAXISORP flat-bottom plates (THERMO SCIENTIFIC, Cat. No. 439454) were coated with 100 μL / well of hNKG2D-Fc (2 μg / mL) in PBS overnight at 4° C. The wells were washed twice with tap water and blocked with blocking solution (pH 7.4, 200 μL / well used) containing 1% bovine serum albumin (BSA) in phosphate-buffered saline (PBS) for 1 h at room temperature. Mouse IgG concentrations of the supernatants were standardized to 1 nM, 0.1 nM, and 0.01 nM. 60 μl of standardized hybridoma supernatant and 60 μl of biotinylated ULBP6_02 (1 nM) were pre-incubated together for 1 h at room temperature, and the mixed solution was added to the coated plate for 15 min. After washing, secondary streptavidin poly-HRP (THERMO SCIENTIFIC, N200) was added to the wells and incubated for 45 min at room temperature. Tetramethylbenzidine (TMB) solution (0.12 mg / mL TMB with 0.04% hydrogen peroxide in 25 mM NaH2PO4) was used as substrate. The reaction was stopped after 5 min with 2N H2SO4. Optical density (OD) was measured at 450 nm. Purified potent blocking monoclonal antibody 7B3 was used as positive control and blank medium was used as negative control for the assay.

[0322] result:

[0323] Of eight clones tested in the three-point biochemical blocking assay, four clones (6E1, 1D4, 4B1, and 5B8) showed greater than 50% blocking activity.

[0324] F. Narrowing the selection of clones

[0325] Based on ELISA / FACS binding, cynomolgus cross-reactivity, blocking efficiency in blocking ELISA, and clonotype information, four clones were selected for cloning, expression, and purification as recombinant humanized antibodies (7B3, 8E11, 6E1, 1D4). One non-blocking antibody (8E8) was selected as a negative control.

[0326] These clones were then humanized to obtain four functional blocking human antibody clones (h7B3, h8E11, h6E1, and h1D4) and one strong binding but non-blocking human antibody clone (h8E8). The germline sequences (HC / LC) used for humanization were h7B3 (HC:IGHV3-48, LC:IGKV1-33), h8E11 (HC:IGHV3-48, LC:IGKV3-11), h1D4 (HC:IGHV7-4-1, LC:IGKV1-33), and h6E1 (HC:IGHV1-2, LC:IGKV1-16). These blocking antibodies were cloned into pRK vectors (both hlgG1 and mIgG2a backbones), expressed in Expi293 cells, purified, and characterized. The variable heavy (VH) and variable light (VL) sequences of the heavy (HC) and light (LC, kappa light, KC) chains of the non-humanized and humanized antibody clones are shown in Table 8. The complementarity determining region (CDR) sequences of the VH and VL sequences of the HC and LC of the selected murine and humanized anti-ULBP6 antibody clones are shown in Table 9 (according to Kabat numbering). The sequences of the CH1 domains with the fragment crystallizable (Fc) region including the heavy chain domains CH2 and CH3 that can be modularly combined with the VH sequences of the clones (Table 8) are shown (Table 10). The Fc sequences shown include (a) effector-positive sequences (E+), which can be considered as wild-type (WT) Fc sequences and are fucosylated unless otherwise indicated; (b) effector-negative sequences (E-), which contain an N297G mutation (EU numbering) that abolishes effector function (mutations shown in bold); and (c) effector-enhanced sequences (referred to herein as "E++" or "E++DE" or "E++(DE)"), which contain S239D and I332E mutations (abbreviated herein as DE) that enhance effector function and are fucosylated unless otherwise indicated (mutations shown in bold). In some examples, the wild-type Fc sequence (e.g., E+) can be defucosylated (Afu) such that the oligosaccharides in the Fc chain do not contain fucose sugar units. Such defucosylation of the wild-type Fc (e.g., E+Afu) results in substantially similar enhanced effector function as that of the E++ Fc sequence.The constant light domain (CL) sequences (Table 11) that can be modularly combined with the VL sequences (Table 8) of the clones are shown. The full length sequences of the HC and LC of the non-humanized and humanized clones contemplated and / or generated herein are shown in Table 12. In Table 12, the variable chains are underlined and the CDR sequences are shown in bold. However, in some embodiments, the C-terminal lysine (Lys447) of the Fc region may or may not be present. Unless otherwise stated herein, the numbering of amino acid residues in the Fc region or constant region is according to the EU numbering system, also referred to as the EU index, as described in Kabat et al., Sequences of Proteins of Immunological Interest, 5th Ed. Public Health Service, National Institutes of Health, Bethesda, MD, 1991. For example, an antibody lacking the C-terminal lysine may have the sequence of SEQ ID NO: 160 or SEQ ID NO: 161. The full length nucleic acid sequences for the HC and LC of the humanized clones contemplated and / or generated herein are shown in Table 13.

[0327] [Table 8-1] [Table 8-2] [Table 8-3]

[0328] [Table 9-1] [Table 9-2]

[0329] [Table 10]

[0330]

Table 11

[0331]

Table 12-1

Table 12-2

Table 12-3

Table 12-4

Table 12-5

Table 12-6

Table 12-7

Table 12-8

Table 12-9

Table 12-10

Table 12-11

Table 12-12

Table 12-13

Table 12-14

Table 12-15

[0332] [Table 13-1] [Table 13-2] [Table 13-3] [Table 13-4] [Table 13-5] [Table 13-6]

[0333] Example 7: Characterization of humanized antibody candidates

[0334] A. SPR analysis of humanized antibodies

[0335] To determine the binding affinity of recombinant humanized h7B3, h8E11, h1D4, and h6E1 (blocking antibodies) and h8E8 (non-blocking antibody) to human and cynomolgus soluble ULBPs, SPR measurements were performed using a BIACORE 8K instrument as follows.

[0336] Materials and Methods:

[0337] BIACORE was performed at 37° C. with various human and cynomolgus NKG2D ligands including soluble hULBP2, hULBP6_01, hULBP6_02, cULBP5, and cULBP2 / 6 (ENSEMBL v99). Briefly, recombinant h7B3, h8E11, h1D4, h6E1, and h8E8 at 0.5 μg / ml in HBS-P buffer (0.01 M HEPES, pH 7.4, 0.15 M NaCl, 0.005% surfactant P20) were applied to a CYTIVA sensor chip Protein A at a flow rate of 30 μL / min to achieve approximately 50 response units in the second flow cell (FC2). The first flow cell (FC1) was kept as a reference. For kinetic measurements, hNKG2D and cNKG2D ligands were serially diluted 3-fold in HBS-P buffer (0.01 M HEPES, pH 7.4, 0.15 M NaCl, 0.005% surfactant P20) from high (10 nM or 50 nM) to low (0.01 nM or 0.07 nM) concentrations, and the dilutions were injected at 37° C. (flow rate: 30 μL / min). Sensorgrams were recorded and subjected to reference and buffer subtraction before evaluation using the BIACORE 8K evaluation software (version 1.1.1.7442). Association rates (kon) and dissociation rates (koff) were calculated using a simple 1:1 Langmuir binding model. Equilibrium dissociation constants (KD) were calculated as the ratio koff / kon (summarized in Table 14).

[0338] result:

[0339] Humanized h7B3, h8E11, h1D4, and h6E1 showed tight binding to both human and cynomolgus ULBPs (Table 14). h7B3 showed the highest binding strength, followed by h8E11, h1D4, and h6E1.

[0340] [Table 14-1] [Table 14-2]

[0341] B. Blocking ELISA of humanized clones

[0342] To determine and compare the blocking potency of humanized h7B3, h8E11, h1D4, and h6E1 between human sULBP6 and hNKG2D, a blocking ELISA was performed.

[0343] Materials and Methods:

[0344] 96-well plates were coated with 100 μl of hNKG2D (1 ug / ml) per well and incubated overnight at 4° C. The plates were then blocked with 300 μl / well of blocking buffer (1×PBS, 1% BSA, and 0.05% Tween®-20) for 1 hour at room temperature. The plates were then washed with washing buffer (1×PBS and 0.05% Tween®-20). Each antibody was then titrated 3-fold from an initial concentration of 300 nM to a final concentration of 0.005 nM, and each dose was mixed with biotinylated human sULBP6 at 1 μg / ml (pre-calculated EC80) for 1 hour at room temperature. The mixtures were incubated on the coated hNKG2D plates for 15 minutes at room temperature. Plates were washed and binding of biotinylated human sULBP6 to hNKG2D was detected using 100 μl / well of streptavidin-HRP (1:10,000 dilution) for 1 h at room temperature, followed by addition of 100 μl / well of TMB and washing for 15 min at room temperature. Absorbance was measured at 450 nm.

[0345] result:

[0346] h7B3 and h8E11 were IC 50 The IC values ​​were 4.47 nM and 4.32 nM, respectively, for the hNKG2D-hULBP6_02 interaction. 50 These antibodies were found to be the most potent blocking antibodies of these interactions, with values ​​of 1.52 nM and 1.59 nM, respectively (Table 15).

[0347] [Table 15]

[0348] C. Epitope Binning

[0349] To determine whether the humanized antibodies bind to the same or different epitopes on human sULBP6, epitope binning experiments were performed using Octet.

[0350] Materials and Methods:

[0351] 100 μM of each antibody was diluted in 300 μl of 1×PBST buffer (1×PBS and 0.05% Tween®-20) and pipetted into different wells of the Octet plate. An anti-human IgG Fc capture (AHC) biosensor in 1×PBST was immersed in the first humanized antibody (e.g., h7B3) for 150-300 seconds to capture the antibody. The biosensor was then immersed in 1×PBST buffer for 60 seconds to reach a stable baseline, followed by immersing the same biosensor in human sULBP6 for an additional 150-300 seconds to promote the interaction of the first humanized antibody with sULBP6. The biosensor was then immersed in a second antibody (e.g., h8E11) to measure the blocking of the second antibody to sULBP6 by the first antibody. This was repeated for each combination of two antibodies selected from h7B3, h8E8, h8E11, h1D4, and h6E1.

[0352] result:

[0353] The four blocking antibodies can be grouped into two different epitope bins (Table 16). h7B3 shares an overlapping epitope with h8E11, h1D4 shares an overlapping epitope with h6E1, and h8E8 does not share an overlapping epitope with any of the other antibodies tested.

[0354] [Table 16]

[0355] D. Characterization of Humanized Anti-ULBP6 Antibodies

[0356] Flow cytometry analysis was performed to measure binding of h7B3, h8E11, h1D4, and h6E1 to surface ULBP6 expressed on tumor cells. Binding of humanized antibodies to cell surface ULBP6 on PANC02.13 cells (expressing ULPB6_02) and COV644 cells (expressing ULPB6_01) was measured as summarized in Table 16. Blocking of NKG2D-Fc binding to cell surface ULBP6 on COV644 cells (expressing ULBP6_01) by humanized antibodies was measured as summarized in Table 17.

[0357] Material / Method:

[0358] Binding of humanized anti-ULBP6 antibodies to cell surface ULBP6 on cancer cells with endogenous target expression was compared using a FACS assay. Cells were detached from the surface of culture flasks using ACCUTASE cell detachment solution, stained with LIVE / DEAD staining dye, and incubated with serially diluted antibodies h7B3, h8E11, h1D4, and h6E1 at concentrations ranging from 83.2 pM to 65 nM for 25 min at 4°C. Bound antibodies were detected with anti-hFc-APC antibody (BIOLEGEND Cat. No. 40936) at a 1:100 dilution. Binding of humanized anti-ULBP6 antibodies to cell surface ULBP6 was analyzed by FACS using a Cytoflex LX instrument. Geometric means were determined and EC 50 was used to calculate.

[0359] To measure blocking of NKG2D-mFc binding to ULBP6, COV644 cells were incubated with hNKG2D-mFc alone or with serially diluted anti-ULBP6 blocking antibodies selected from h7B3, h8E11, h1D4, and h6E1. Binding of human NKG2D-mFc to surface ULBP6 was detected by flow cytometry assay using anti-mFc-APC antibody. Geometric means were determined and IC 50 The values ​​were calculated respectively.

[0360] result

[0361] Binding to cell surface ULBP6: EC10 exhibited by h7B3, h8E11, h1D4, and h6E1 50 The values ​​were 1.97 nM (PANC02.13) and 1.93 nM (COV644), 1.28 nM (PANC02.13) and 1.29 nM (COV644), 0.72 nM (PANC02.13) and 1.4 nM (COV644), 0.44 nM (PANC02.13) and 0.36 nM (COV644), respectively, calculated as the average of two replicates, as shown in Table 17.

[0362] Blocking NKG2D-Fc binding to ULBP6 on the surface of COV-644 cells: IC of h7B3, h8E11, h1D4, and h6E1 50 The values ​​were 0.17 nM, 0.11 nM, 0.22 nM, and 0.2 nM, respectively, and were determined as the average of two replicates, as shown in Table 17.

[0363] In conclusion, this example demonstrated the biochemical properties of four humanized anti-ULBP6 antibodies that a) inhibit the interaction between soluble ULBP6 and NKG2D and b) bind to cell surface expressed ULBP6 on tumor cell lines.

[0364] [Table 17]

[0365] Example 8: Reversal of sULBP6-induced NKG2D blocking / internalization by anti-ULBP6 antibodies

[0366] Initial data (not shown) revealed that soluble ULBP6 competes with anti-NKG2D monoclonal antibodies for binding to NKG2D expressed by NKG2D-expressing cells. In particular, haplotype sULBP6_02 competed with anti-NKG2D monoclonal antibodies for NKG2D binding more strongly than sULBP6_01. These data suggest that soluble ULBP6 may a) interfere with the interaction of NKG2D with other NKG2DLs and / or b) induce internalization of NKG2D on NKG2D-expressing cells.

[0367] To measure the effect of anti-ULBP6 antibodies on soluble ULBP6 inhibition of NKG2D blocking / internalization, an NKG2D detection assay was performed on stimulated PBMCs from healthy donors. Soluble ULBP6_02 was used in the assay because it was found to bind NKG2D with the highest affinity (see Example 2) and more potently inhibit the binding of anti-NKG2D monoclonal antibodies compared to ULBP6_01.

[0368] Materials and Methods:

[0369] In this assay, stimulated PBMCs (20 ng / ml IL-2 and IL-15 in X-VIVO15 medium for 36 hours) were incubated as follows. (a) incubated without soluble ULBP6_02, (b) incubated with soluble ULBP6_02 (50 nM), or (c) incubated with soluble ULBP6_02 (50 nM) and anti-ULBP6 antibodies (h7B3(E-), h1D4(E-), h6E1(E-), h8E11(E-), or isotype control antibody (isotype (E-)), each in assay medium of RPMI containing 10% FBS. After 1 hour of incubation, cells were harvested and stained with anti-NKG2D-APC antibody to quantify surface expression of NKG2D on various PBMC populations. 69D.12_315 is a tool antibody that binds weakly to ULBP6. 8E8 is a control antibody that binds but does not block the interaction of NKG2D with ULBP6.

[0370] result:

[0371] The assay was first validated with tool antibodies (Figure 2, top panel). At high concentrations, isotype antibodies did not alter NKG2D MFI in the presence of sULBP6. The weak binding 69D.12_315 control antibody showed a modest reversal of sULBP6-induced NKG2D blocking / downregulation in a dose-dependent manner, whereas the high affinity 7B3 antibody showed a stronger increase in NKG2D MFI. The non-blocking 8E8 antibody reduced the detectable NKG2D MFI compared to the isotype control by stabilizing the interaction between ULBP6 and NKG2D.

[0372] The anti-ULBP6 antibodies tested reversed sULBP6-induced NKG2D blocking / downregulation in a dose-dependent manner (Figure 2, lower panel). h7B3(E-) and h8E11(E-) showed reactivation of NKG2D at lower doses (left shift of the curve) compared to the other antibodies tested. These results are representative of data from PBMCs derived from three different healthy donors.

[0373] These data indicate that humanized anti-ULBP6 antibodies, particularly 7B3 and 8E11, neutralize the inhibitory effect of soluble ULBP6, thus allowing NKG2D to interact with other NKG2DLs.

[0374] Example 9: MICA-mediated immune activation of PBMCs

[0375] In the above examples, it was shown that humanized anti-ULBP6 antibodies reduce the inhibitory effect of sULBP6 on cell surface NKG2D. By neutralizing sULBP6, anti-ULBP6 antibodies may enable activation of NKG2D via alternative NKG2DLs such as MICA.

[0376] A PBMC interferon gamma (IFNg) secretion assay was performed to measure the effect of anti-ULBP6 antibodies on reactivating MICA-mediated immune activation via neutralization of soluble ULBP6.

[0377] Materials and Methods:

[0378] In this assay, stimulated PBMCs from healthy donors (with IL-2 and IL-15 at 20 ng / ml in X-VIVO15 medium for 36 hours) were incubated on MICA-coated plate surfaces in the presence of soluble ULBP6_02 (50 nM) with anti-ULBP6 antibodies (h7B3(E-), h1D4(E-), h6E1(E-), and h8E11(E-)) at concentrations ranging from 0.39 nM to 200 nM in RPMI assay medium with 10% FBS, or with an isotype control antibody (Iso(E-)) at 200 nM in RPMI assay medium with 10% FBS. After 24 hours of incubation, cell-free assay medium was collected and IFNg levels were analyzed using the Luminex method. 69D.12_315 is a tool antibody that binds weakly to ULBP6, and h8E8 is a control antibody that binds to ULBP6 but does not block the interaction of NKG2D with ULBP6.

[0379] result:

[0380] PBMC activation was measured via IFNg secretion. The assay was first validated with tool antibodies (Figures 3A and 3B). The h8E8 antibody, as expected, was unable to reverse the inhibitory effect of sULBP6 at all concentrations tested. The h7B3 antibody dose-dependently reversed the blocking of sULBP6-induced MICA-mediated PBMC activation. The 69D.12_315 tool antibody showed a very modest contrast in blocking of sULBP6-induced MICA-mediated PBMC activation. This correlated well with its weak blocking and binding to ULBP6 (Figures 3A and 3B). In summary, the sensitivity of the assay was able to distinguish between the ineffective non-blocking antibody h8E8, the weak ULBP6-binding antibody 69D12.315, which showed partial activity, and the most effective strong ULBP6-binding and blocking antibody h7B3 (Figures 3A and 3B). h7B3(E-), h1D4(E-), h6E1(E-), and h8E11(E-) showed comparable efficacy in blocking sULBP6 in the assay (Figures 3C and 3D). These results from donor 1307 are representative of data from PBMCs of three independent healthy donors.

[0381] Thus, the anti-ULBP6 antibodies described herein can reactivate immune cell responses by neutralizing the inhibitory effect of soluble ULBP6 on NKG2D responses to NKG2DL.

[0382] Example 10: Synergistic effect of NKG2D and FcgRIIIa activation

[0383] In addition to the ULBP6 blocking effect of enhancing NKG2D activation described herein, anti-ULBP6 antibodies can also bind to immune cells expressing Fc receptors, for example, by binding to surface ULBP6. An IFNg secretion assay was performed to determine whether NKG2D activation is synergistic with FcgRIIIa activation by measuring IFNg secretion from PBMCs.

[0384] Materials and Methods:

[0385] For IFNg secretion assay, PBMCs were pre-stimulated with IL-2 and IL-15 (20 ng / ml each) in X-VIVO15 for 36 hours. 96-well plates were coated overnight at 4°C with anti-lysozyme Fc E+ ("Ab E+") or anti-lysozyme Fc E- ("Ab E-") antibodies with or without duplicate ULBP6_02 (referred to as ULBP6 in Figure 4) (fused to the molecule). They were plated in PBS and plate-coated in assay wells at concentrations ranging from 25 nM to 400 nM coating. This artificial fusion molecule in plate-bound form can model cell surface NKG2D ligand(s) and can result in activation of NKG2D. On the day of the assay, plates were washed, blocked with BSA, and pre-stimulated PBMCs were added to the plates and incubated at 37°C and 5% CO2 for 24 hours. The assay was performed in the absence of sULBP6. Cell-free assay medium was collected and IFNg levels were measured using either a Luminex IFNg kit or an Invitrogen ELISA kit (catalog number 88-7316-88).

[0386] result:

[0387] Plate-bound effectorless (E-) anti-lysozyme antibody ("AbE-") did not affect the modulation of IFNg secretion at any concentration and was therefore considered background (Figure 4). Plate-bound anti-lysozyme antibody (E+) ("AbE+") increased IFNg secretion in a dose-dependent manner above background levels (Figure 4). NKG2D activation via ULBP6 fused to effectorless (E-) anti-lysozyme antibody (graph ULBP6 / AbE-) increased IFNg secretion in a dose-dependent manner above background levels (Figure 4). The combination of ULBP6 with Fc effector-positive anti-lysozyme antibody (E+) (ULBP6 / AbE+) resulted in significantly higher IFNg secretion compared to the effects of ULBP6 / AbE- or anti-lysozyme antibody E+ ("AbE+"), respectively. A synergistic effect of NKG2D and FcgRIIIa activation was observed. Simultaneous co-activation induced greater IFNg release than the sum of activation of NKG2D alone and activation of FcgRIIIa alone. This mechanistic knowledge supports the use of anti-ULBP6 antibodies with active Fc function when such synergistic effects are desired.

[0388] Example 11: Immune activation in mixed tumor cell and PBMC assays

[0389] To measure the ability of anti-ULBP6 antibodies to block soluble ULBP6-mediated inhibition of IFNg secretion,

[0390] A mixed cell assay of tumor cells and PBMCs was performed. A comparison between the efficacy of the E+ and E- versions of the antibody was also determined.

[0391] A. Antibodies neutralize sULBP6-mediated inhibition of IFNg secretion in mixed cell assays of tumor cells and PBMCs.

[0392] Materials and Methods:

[0393] For the mixed cell assay using COV644 cancer cells (expressing shed ULBP6 and surface ULBP6), PBMCs from healthy donors were pre-stimulated with IL-2 and IL-15 (20ng / ml each) in X-VIVO15 medium for 36 hours. 24 hours prior to the experiment, 60,000 COV644 cells were seeded in 96-well plates. On the day of the assay, the floating stimulated PBMCs were collected, counted, and added to the COV644 cells in a 2:1 ratio. In addition, antibodies (h7B3, h8E11, h1D4, and h6E1 at dose settings of 200nM to 0.5nM) and 50nM of soluble recombinant sULBP6_02 were added to the plate. The assay was performed in RPMI medium containing 10% FBS. After 24 hours of co-incubation, the supernatants were collected and the levels of IFNg were measured using Luminex. For each antibody, a titration curve and maximum potency values ​​were determined.

[0394] result:

[0395] As shown in Figure 5A, all of the anti-ULBP6 antibodies neutralized sULBP6-mediated inhibition of IFNg secretion in a mixed cell assay of tumor cells and PBMCs. Although there were differences in the ability of these antibodies to increase IFNg in the assay, all responded in a dose-dependent manner. Based on the highest dose of antibody used (200 nM) (maximal efficacy), the E+ molecules were determined to be superior to the E- and isotype controls (Figure 5B).

[0396] B. Full titration of antibodies (E+ vs. E-) yields similar EC 50 It was revealed that the maximum activity of E+ molecules was greater than that of E- molecules.

[0397] A second PBMC mixed cell assay was performed as described in part A of this example to assess the EC 50 and maximum activity was determined.

[0398] Materials and Methods:

[0399] For the mixed cell assay with COV644 cancer cells (expressing shed ULBP6 and surface ULBP6), PBMCs from healthy donors were pre-stimulated with IL-2 and IL-15 (20ng / ml each) in X-VIVO15 medium for 36 hours. 24 hours prior to the experiment, 60,000 COV644 cells were seeded in a 96-well plate. On the day of the assay, the floating stimulated PBMCs were collected, counted, and added to the COV644 cells in a 2:1 ratio. In addition, antibodies (h7B3 (E+ and E-), h8E11 (E+ and E-) at dose settings of 0.05nM to 200nM) and 50nM recombinant sULBP6_02 were added to the plate. The assay was performed in RPMI medium containing 10% FBS. After 24 hours of co-incubation, the supernatants were collected and the levels of IFNg were measured using Luminex. EC 50 Curves were determined for h7B3(E-) and h7B3(E+).

[0400] result:

[0401] Both h7B3E- and h8E11E- antibodies showed a dose-dependent increase in IFNg levels compared to isotype control by neutralizing the inhibitory effect of soluble ULBP6 (Figure 6, upper panel). Both E+ antibodies showed higher maximum potency compared to E- versions of the same antibodies. This may be due to ADCC through tumor surface binding of E+ antibodies. h7B3 showed higher maximum potency than h8E11 in comparison of E- and E+ antibodies. Maximum potency levels of IFNg were further increased for the E+ versions of these antibodies (Figure 6, lower panel). H7B3E- showed similar EC 50 had value.

[0402] Example 12: Cytotoxicity and proliferation in tumor-immune mixed cell assays

[0403] In addition to IFNg secretion, the level of immune cell activation in the presence of tumor cells and the anti-ULBP6 antibodies described herein can be further determined by measuring immune cell surface expression of markers of cytotoxicity and proliferation. A tumor-immune mixed cell assay was performed using PBMCs from healthy donors. PBMCs were incubated with tumor cells and 500 nM sULBP6_02 in the presence or absence of anti-ULBP6 antibodies, and staining for granzyme B and Ki-67 was measured by flow cytometry. The results show the effect of anti-ULBP6 antibodies on NK and T cell cytotoxicity and proliferation in the presence of soluble ULBP6.

[0404] Materials and Methods:

[0405] PBMCs from healthy donors were stimulated for 36 hours in X-VIVO15 medium containing IL-2 and IL-15 (20 ng / ml each). Floating cells were collected and replated (100,000 cells per well) on COV644 cancer cells (60,000 cells per well). 50 nM sULBP6-02 was added to each well along with 100 nM of one of the following antibodies: 7B3(E-), 7B3(E+), or isotype (E+) (isotype control). Assays were performed in RPMI medium containing 10% FBS. After 24, 48, or 72 hours of incubation, cells were harvested from the wells, stained with anti-granzyme B or anti-Ki-67 antibodies, and gated on CD8 T cell and NK cell populations (defined by FACS antibodies CD3, CD8, and CD56).

[0406] result:

[0407] Granzyme B and Ki-67 were generally decreased over time in both CD8 and NK cell populations in the presence of 50 nM sULBP6_02 (isotype antibody). Anti-ULBP6 antibodies (h7B3(E+) and h7B3(E-)) were able to rescue these decreases, with the E+ antibodies generally showing stronger potency than the E- antibodies (Figures 7A, 7B, 7C, and 7D).

[0408] Example 13: Blocking activity of effector positive antibodies

[0409] To increase ADCC, anti-ULBP6 antibodies with the ADCC-enhancing mutations S239D / I332E were prepared. In this example, the ULBP6 blocking activity and FcG receptor activation of E-, E+, and E++ candidate anti-ULBP6 antibodies were compared in a mixed cell assay using COV644 cancer cells and PBMCs.

[0410] Materials and Methods:

[0411] For the mixed cell assay with COV644 cancer cells, PBMCs from healthy donors were pre-stimulated with IL-2 and IL-15 (20ng / ml each) in X-VIVO15 medium for 36 hours. 24 hours before the experiment, 60,000 COV644 cells were seeded in a 96-well plate. On the day of the assay, the floating stimulated PBMCs were collected, counted, and added to the COV644 cells in a 2:1 ratio. In addition, antibodies (various concentrations of h7B3(E-), h7B3(E+), and h7B3(E++)(DE)) and 50nM of soluble recombinant sULBP6_02 were added to the plate. The assay was performed in RPMI medium containing 10% FBS. After 24 hours of co-incubation, the supernatants were collected and the levels of IFNg were measured using Luminex.

[0412] result:

[0413] h7B3(E-), h7B3(E+), and h7B3(E++)DE each showed a dose-dependent increase in IFNg in this assay due to inhibition of soluble ULBP6 (Figures 8A and 8B). Maximal potency of IFNg in this assay was increased in h7B3(E++)(DE) compared to h7B3(E-) and h7B3(E+) (Figure 8C).

[0414] Example 14: Immune cell killing assay (E-, E+, E++ antibodies)

[0415] To test whether the anti-ULBP6 antibodies described herein can enhance tumor cell killing, an INCUCYTE immune cell killing assay was performed using PBMCs from eight healthy donors to detect tumor cell killing by immune cells.

[0416] Materials and Methods:

[0417] PBMCs from eight healthy donors were stimulated with IL-2 and IL-15 (20 ng / ml each) in X-VIVO15 medium for 36 hours. One day before setting up the assay, 3,000 COV644-GFP cancer cells (COV644 cancer cell line engineered to express GFP) were seeded into a 96-well plate suitable for the INCUCYTE assay. On the day of the assay, PBMCs were seeded on top of COV644 cells at a ratio of 25:1 in the presence of 50 nM human sULBP6_02 protein. In addition, isotype or anti-ULBP6 antibodies (100 nM) were also included in the experiment. The humanized antibodies tested in this assay included h7B3(E-) (ULBP6 neutralizer with no FcgR binding), h7B3(E+) (ULBP6 neutralizer with FcgR binding), and h7B3(E++) (ULBP6 neutralizer with enhanced FcgR binding). The assay was performed in RPMI medium with 10% FBS. The assay plates were placed in an INCUCYTE instrument and GFP levels (per well) were measured over the 5-day assay. The decrease in GFP signal between conditions was interpreted as indicating fewer tumor cells present, likely due in part to tumor cell killing via CD8 T cells and NK cells activated by the ULBP6 antibody.

[0418] result:

[0419] The h7B3 antibody efficiently blocked the immunosuppressive activity of soluble ULBP6, thereby increasing immune cell activation and tumor cell killing. h7B3E+ further enhanced tumor cell killing potency compared to h7B3E-. h7B3E++(DE) further enhanced tumor cell killing potency compared to h7B3E+. A total of eight healthy donors were screened in this assay. As shown in Figure 9A, four out of eight donors responded strongly to h7B3. As shown in Figure 9B, three out of eight donors were defined as moderate responders to h7B3. As shown in Figure 9C, one out of eight donors was identified as a poor responder to h7B3. Table 18 summarizes the strength of the response from each donor, showing all eight donors evaluated in this INCUCYTE assay, and how well they responded to each anti-ULBP6 antibody in their ability to kill tumor cells.

[0420] [Table 18]

[0421] Example 15: Effector-enhanced and defucosylated antibodies

[0422] This example demonstrates the testing of defucosylated anti-ULBP6 antibodies in the mixed immune cell-tumor cell assay and the INCUCYTE killing assay described above.

[0423] A. IFNg expression in mixed cell assays

[0424] In the mixed cell assay described above, the effects on IFNg expression will be compared between h7B3(E+), h7B3(E++)(DE), and h7B3(E+) (defucosylated (Afu) "h7B3E+Afu") antibodies.

[0425] Materials and Methods:

[0426] PBMCs were stimulated for 36 hours with X-VIVO15 medium containing IL-2 and IL-15 (20 ng / ml each). Floating cells were collected and replated (100,000 cells per well) on COV644 cancer cells (60,000 cells per well). 50 nM of sULBP6-02 was added to each well along with various doses of h7B3(E+), h7B3(E++)(DE mutant), or h7B3(E+)(Afu) ("h7B3E+Afu") or isotype E+ (isotype control) antibodies, ranging from 0.05 nM to 200 nM. Assays were performed in RPMI medium containing 10% FBS. Assay supernatants were harvested after 24 hours and assessed for IFNg levels.

[0427] result:

[0428] The 7B3(E+) antibody showed a significant increase in IFNg compared to the isotype control. This was further increased by both the h7B3(E++)DE and h7B3(E+) defucosylated ("h7B3E+Afu") molecules (Figure 10). There was little to no difference in both maximum potency and curve shape between the h7B3(E++)DE and h7B3(E+) defucosylated ("h7B3E+Afu") molecules in this assay. The h7B3(E+) defucosylated molecule ("h7B3E+Afu") is equivalent to the 7B3(E++)DE molecule in promoting IFNg secretion.

[0429] B. Tumor cell killing

[0430] INCUCYTE killing assays were performed on 4 healthy donors to compare h7B3E+, h7B3E++(DE), and h7B3E+Afu for their ability to increase tumor cell killing.

[0431] Materials and Methods:

[0432] PBMCs from four healthy donors were stimulated with IL-2 and IL-15 (20ng / ml each) in X-VIVO15 medium for 36 hours. One day before setting up the assay, 3,000 COV644-GFP cancer cells (COV644 cancer cell line engineered to express GFP) were seeded into a 96-well plate suitable for the INCUCYTE assay. On the day of the assay, PBMCs were seeded on COV644 cells at a ratio of 25:1 in the presence of 50nM sULBP6_02 protein. In addition, isotype or anti-ULBP6 antibodies were also included in the experiment. The assay was performed in RPMI medium containing 10% FBS. The humanized antibodies tested in this assay included h7B3(E+) (ULBP6 neutralizer with no FcgR binding), h7B3E++(DE) (ULBP6 neutralizer with enhanced FcgR binding), and h7B3Afu(E+) (ULBP6 neutralizer with enhanced FcgR binding through defucosylation). The assay plates were placed in the INCUCYTE instrument and GFP levels were measured over the 5-day assay. The lower GFP signal between conditions may be due to the presence of fewer tumor cells, in part due to increased tumor cell killing via CD8 T cells and NK cells due to increased immune activity driven by the ULBP6 antibody.

[0433] result:

[0434] The h7B3E+defucosylated (Afu) antibody was comparable to h7B3E++(DE) in the tumor killing assay, with some variability between donors. The h7B3E++(DE) and h7B3E+Afu ("h7B3Afu") molecules had lower average GFP per well, which may indicate fewer tumors per well, possibly due to increased tumor cell killing compared to the h7B3E+ and isotype controls. Two of the four donors showed strong responses to the h7B3E+_Afu ("h7B3Afu") and h7B3E++DE antibodies (Figure 11A), and two donors showed moderate responses (Figure 11B).

[0435] Example 16: Pharmacokinetic properties of anti-ULBP6 antibodies in cynomolgus monkeys

[0436] The objective of this study was to determine the pharmacokinetic properties of anti-ULBP6 antibodies in the pharmacologically relevant species, cynomolgus monkeys (Macaca fascicularis, Mf), and to screen the variable regions of h7B3(E+) and h8E11(E+) for sequence libraries that may induce poor pharmacokinetic behavior.

[0437] Materials and Methods:

[0438] Male (M) cynomolgus monkeys (naive, 3-4 years old, 3.0-3.3 kg, originating from China) were administered a single 3 mg / kg dose of h7B3(E+) (PUR4377) or h8E11(E+) (PUR4379) by slow intravenous bolus (5 min infusion). Approximately 12 whole blood samples (later processed to serum) were collected from each animal starting prior to dosing (pre-dose) and ending 28 days after dosing (post-dose). Serum samples were analyzed for the amount of h7B3(E+) or h8E11(E+) using a qualified ELISA with a lower limit of quantification (LLOQ) of 0.015 μg / mL. Group mean (N=3) concentration-time profiles of h7B3(E+) and h8E11(E+) are shown on semi-log plots (GraphPad PRISM 9.0). Mean PK parameters estimated by noncompartmental analysis (NCA) are also listed (Phoenix WinNonlin 8.2).

[0439] result:

[0440] The anti-ULBP6 antibodies h7B3(E+) and h8E11(E+) had typical IgG PK profiles with linear elimination (Figure 12). The mean half-lives were long, ranging from 11.8 to 13.3 days for h7B3(E+) and h8E11(E+), respectively (Table 19), consistent with other IgG therapeutics that exhibit linear PK in monkeys (Betts et al., 2018, mAbs, 10(5);751-764). Other PK properties, including clearance and volume of distribution, were within the typical range for therapeutic IgG in monkeys (Table 19). There was no difference in exposure between h7B3(E+) and h8E11(E+), as assessed by Cmax and AUClast, confirming the absence of obvious sequence liability between the two clones (Table 19). There was no evidence of target-mediated pharmacokinetics (TMDD) at the dose levels tested, suggesting that 3 mg / kg may be a target saturating dose in healthy monkeys (Figure 12).

[0441] [Table 19]

[0442] Example 17: Structural mapping of the h7B3 binding site on ULBP602

[0443] Materials and Methods:

[0444] For crystallization, a human ULBP602 construct spanning residues 26–218 was generated. This construct included a penta-His tag at the C-terminus of the protein for purification purposes. Complex formation, crystallization, data collection, model solving, and model building and refinement were then performed by Proteros Biostructures GmbH, Germany. A complex of ULBP6 with h7B3Fab was formed by mixing both proteins in a 1.2:1 ratio and incubating at room temperature for 2.5 h. The complex was purified by size-exclusion chromatography using a HiLoad16 / 60 Superdex75 preparative grade column pre-equilibrated with 20 mM HEPES, pH 7.0, 100 mM NaCl.

[0445] h7B3-Fab ULBP602 co-crystals were obtained under the following crystallization conditions: 0.1 ul of h7B3-Fab ULBP602 complex at 26.72 mg / ml in 20 mM HEPES, pH 7.0, 100 mM NaCl was mixed with 0.1 ul of a buffer containing 10% (v / v) glycerol, 0.1 M MES (pH 6.9), 5% (w / v) PEG1000, 30% (w / v) PEG600. The resulting droplets were incubated at 4°C.

[0446] The resulting cocrystal was flash frozen in liquid nitrogen. Diffraction data for the cocrystal of h7B3-Fab and ULBP602 were acquired at the Swiss Light Source (SLS, Villigen, Switzerland) on beamline PXII-X10SA. Diffraction data collection was performed at 100 Kelvin using a wavelength of 0.9998 Å and a Dectris EIGNER2 Si 16M detector. Data were processed using the autoROC XDS and autoPROC AIMLESS programs (Table 20).

[0447] [Table 20]

[0448] Phase information for determining the structure was obtained by molecular replacement using the phaser of the CCP4 (https: / / www.ccp4.ac.uk / ) software package and ULBP6 (4S0U) as a template. Subsequent model building and refinement were carried out according to standard protocols using COOT (https: / / www2.mrc-lmb.cam.ac.uk / personal / pemsley / coot / ) and the software package CCP4, respectively (Table 21).

[0449] [Table 21]

[0450] For the free R-factor calculation, approximately 5.0% of the measured reflections were excluded from the refinement procedure.

[0451] result

[0452] The crystal structure of h7B3Fab was solved at a resolution of 2.31 Å (Figure 13). The structure of ULBP602 has a resolution of 0.580 Å over 1098 atoms. 2 The structure is very similar to the previously solved crystal structure of UBLP6 in complex with NKG2D (PDB entry 4s0u), which has an rmsd of 1.2 nm. The h7B3Fab binding of ULBP6 is 1267 Å on ULBP6. 2 Using the change in solvent accessible surface area (SASA) between the free ULBP6 structure and that bound in complex, we identified 32 residues that constitute the structural epitope (Table 22). NKG2D is a dimeric protein composed of two protomers (A and B) (Zuo et al., 2017). The h7B3 epitope on ULBP602 overlaps with most of the binding site of the A protomer of NKG2D and overlaps with part of the binding site of the B protomer of NKG2D (Figure 14).

[0453] In addition to the SASA-based epitope, we also determined an extended structural epitope defined by all residues of ULBP602 that are within 4 Å of any residue in h7B3. This epitope is composed of 54 residues (Table 22).

[0454] [Table 22-1] [Table 22-2] [Table 22-3]

[0455] Example 18: Identification of hULBP6 epitope residues involved in binding of h7B3, h8E11, and h6E1

[0456] To identify the epitope residues involved in the binding of fully human clones h7B3, h8E11, and h6E1 to hULBP6, functional epitope mapping using phage display was used. Briefly, we generated human ULBP602 (residues R26 to G218) with a mouse IgG signal peptide and a C-terminal His tag, followed by functional epitope mapping using phage display, which involved displaying ULBP6 on phage, generating a ULBP6 phage mutagenesis library, screening the ULBP6 phage mutagenesis library, and performing NGS sequencing and data analysis.

[0457] Materials and Methods:

[0458] Human ULBP602 (residues R26 to G218) with a mouse IgG signal peptide and a C-terminal His tag was transiently expressed in 293 cells. Secreted protein was purified from cell culture supernatant by Ni-NTA chromatography. A Ni Excel 10 mL, XK16 / 5 column was equilibrated with loading buffer PBS, pH 7.4. A stepwise gradient of 4%, 50%, 100% PBS, pH 7.4, 500 mM imidazole was used for elution. Eluted protein was further purified by size-exclusion chromatography (Superdex, 60L XK15 / 30 with PBSk, pH 7.4 as buffer). Heavy and light chains containing a C-terminal His tag of Fab 7B3 were cloned into two separate vectors and co-transfected into ExpiCHO cells. Fab was formed by heavy and light chains and purified from cell culture supernatant by Ni-NTA chromatography. For purification, clarified supernatant medium from expression cultures was loaded onto a Ni-NTA column equilibrated with PBS. After loading, the column was washed with PBS and 20 mM imidazole. The eluted protein was further purified by size-exclusion chromatography (using 16 / 600 Hiload Superdex200, 2x PBS, pH 6.5). To generate the ULBP6-Fab h7B3 complex, a molar excess of ULBP6 (1.2:1 molar ratio) was incubated with Fab h7B3 in PBS, pH 7.4 for 2.5 hours, followed by size-exclusion chromatography on a HiLoad 16 / 60 Superdex 75 preparative grade column pre-equilibrated with 20 mM HEPES, pH 7.0, 100 nM NaCl.

[0459] For ULBP6 display on phage, ULBP6 phage production was followed by phage purification, ELISA assay, ULBP6 display on phage, and ULBP6 folding on phage as described below.

[0460] A phagemid encoding human ULBP6-HA-pIII was designed to display the ULBP6 protein on the phage surface. The vector was constructed by fusing the ULBP6 gene to the N-terminal domain of the M13 minor coating protein (gIII). An epitope tag (HA tag, amino acid sequence: YPYDVPDYA) was added between the ULBP6 gene and the pIII gene. XL-1Blue competent cells (Agilent Technologies, Cat. No. 200249) were chemically transformed with the phagemid and spread on a carbenicillin-50 agar plate with overnight incubation at 37°C. One colony was then inoculated into 5 mL of 2YT broth (TEKNOVA, Cat. No. Y0167) for growth. The culture was incubated with shaking at 37°C until the OD600 reached 0.3 / mL. 50 μL of M13K07 helper phage (New England Biolabs, Catalog No. N0315S) was added to the culture. After shaking at 37° C. for 1 hour, 5 mL of the culture was spread into 50 mL of 2YT broth supplemented with 50 μg / mL carbenicillin and 50 μg / mL kanamycin. The spread culture was incubated at 37° C. for 4 hours, followed by incubation at 30° C. overnight.

[0461] Phage cultures were harvested by centrifugation at 8000g for 10 min at 4°C. The supernatant was collected for phage purification. Phages were precipitated from the supernatant with 2.5M NaCl, 20% PEG solution (TEKNOVA, Cat. No. P4137) and resuspended in 1x PBS buffer (137mM NaCl, 2.7mM KCl, 4.3mM Na2HPO4, 1.47mM KH2PO4, pH 7.4). The PEG precipitation process was repeated to increase purity. The final phage pellet was resuspended in 1x PBS buffer to a concentration of 10.0 OD / mL. OD was read at 268nm on a Nanodrop (ThermoFisher Scientific). The purified phage was stored at 4°C.

[0462] MaxiSorp plates (ThermoFisher Scientific, Cat. No. 437915) were coated overnight with 2 μg / mL of antibodies or recombinant protein receptors (e.g., anti-HA antibody, anti-ULBP6 antibody, and ULBP6 receptor) and then blocked with 1% bovine serum albumin (BSA, ThermoFisher Scientific, Cat. No. PI37520) in 1×PBS buffer for 30 min. Phages were diluted to 1.0 OD / mL in 120 μL of ELISA buffer (0.5% BSA, 0.05% Tween®-20 in 1×PBS) in 96-well plates (ThermoFisher Scientific, Cat. No. 156545). Three-fold serial dilutions were performed. 65 μL of diluted phage solution was added to each corresponding well of the blocked MaxiSorp plate and incubated at room temperature for 15 min with shaking. Horseradish peroxidase (HRP)-labeled anti-M13 antibody (Sino Biological, Cat. No. 11973-MM05T-H) diluted 1:2500 in ELISA buffer was then added at 100 μL / well as secondary reagent. The ELISA plate was developed using TMB solution (ThermoFisher Scientific, Cat. No. 34022) and the reaction was stopped by adding 50 μL of 2M H2SO4 (VWR, Cat. No. BDH7259-1). The absorbance was read at 450 nm in a VersaMax microplate reader (Molecular Devices).

[0463] The display of ULBP6 on the phage was confirmed by examining the expression of the HA tag in an ELISA assay. ULBP6 phage binding to anti-HA antibodies (shown in Figure 15) revealed the expression of the HA tag and the display of ULBP6 on the phage. Furthermore, non-specific phage binding to the anti-HA antibodies was excluded by a negative control using an irrelevant phage.

[0464] The folding of ULBP6 on the phage was verified by examining phage binding to recombinant ULBP6 or ULBP6 lead antibodies discovered by the ULBP6 receptor NKG2D (Table 23) in an ELISA assay. ULBP6 phages showed binding affinity to hu7B3, hu8E1, hu1D4, hu6E1, and hu8E8, as well as the ULBP6 receptor (Figures 16 and 17). This result demonstrated that ULBP6 was properly folded on the phage.

[0465] [Table 23]

[0466] A ULBP6 phage library was engineered for functional epitope mapping by generating random mutagenesis across the gene encoding ULBP6. Mutagenesis was performed by error-prone PCR. Error-prone PCR products were subcloned into phagemids and transformed into XL-1Blue cells to generate a mutagenized ULBP6 phage library.

[0467] Error-prone PCR was performed using GeneMorph II Random Mutagenesis Kit (Aglient, Cat. No. 200550). Mutation frequency was controlled by adjusting the amount of initial target DNA and the number of PCR cycles. To determine functional residues in antibody binding, the desired mutation frequency was one amino acid change (1-2 nucleotide changes) per gene. To achieve the desired mutation frequency, the error-prone PCR mixture was set up as shown in Table 24. The forward and reverse primers used were as shown in Table 25.

[0468] [Table 24]

[0469] [Table 25]

[0470] To amplify DNA in a thermal cycler, the program was set up as shown in Table 26.

[0471] [Table 26]

[0472] Error-prone PCR products were purified using a QIAquick Gel Extraction kit (Qiagen, Cat. No. 28706X4). The purified PCR products and phagemid vectors were digested with restriction enzymes Mlul (New England Biolabs, Cat. No. R3198) and Xhol (New England Biolabs, Cat. No. r0146S), followed by ligation reactions. The ligation products were transformed into XL-1Blue competent cells. After transformation, XL-1Blue cells were plated on carbenicillin-50 agar plates and incubated overnight at 37°C. 20–30 colonies were picked for Sanger sequencing (ELIM Biopharmaceuticals) and analyzed for mutation frequency. The error-prone PCR protocol described above was used to randomly generate unbiased mutations across the amplified region. The mutation frequency was 1.2 bp / gene.

[0473] To generate the mutagenized phage library, the error-prone PCR products were transformed into XL-1Blue competent cells by electroporation. The cells were harvested in 25 mL of SoC medium (ThermoFisher Scientific, Cat. No. 15544034) at 37°C. 250 μL of M13K07 helper phage was added to the culture. After 1 h incubation at 37°C, 25 mL of the culture was spread into 60 mL of 2YT broth containing 50 μg / mL carbenicillin and 50 μg / mL kanamycin. The spread culture was incubated at 37°C for 4 h, followed by overnight incubation at 30°C. The phage library was harvested by centrifugation of the culture at 8000g for 10 min at 4°C and purified according to the phage purification method described above. The purified ULBP6 phage library was stored at 4°C. The cell pellet was saved for NGS sequencing.

[0474] The ULBP6 phage library was panned twice with lead antibodies (7B3, 6E1, and 8E11) by a first round of plate panning and a second round of solution panning. Enrichment was measured by phage titration to assess panning performance.

[0475] MaxiSorp plates were coated overnight at 4° C. with lead antibodies (7B3, 6E1, and 8E11) at 5 μg / mL in 1×PBS (100 μL / well, 8 wells for each antibody). Two wells were coated with 1×PBS buffer as background. The next day, plates were washed six times using 1×PBS (1×PBS, 0.1% Tween® 20 surfactant) and blocked with 1% BSA in 1×PBS buffer (65 μL / well) for 30 minutes at room temperature. Then, 40 μL of 1×PBS containing 1% Tween®-20 was added to each well and incubated for another 30 minutes at room temperature. Plates were washed six times with 1×PBST. Mutagenized phage library was diluted to 2.5 OD / mL and blocked with 1×PBS buffer containing 1% BSA and 0.05% Tween®-20 for 30 minutes at room temperature with shaking. The blocked phage library was added to the blocked plate at 100 μL / well and incubated at room temperature for 2 hours with shaking. The plate was gently washed 10 times with 1×PBST buffer. Bound phages were eluted from the plate using 0.1 M HCl (100 μL / well) and neutralized with 1 M Tris base and 1% BSA. The eluted phages were titrated to determine enrichment (Table 27) and amplified using XL-1Blue cells. The selected phage library was purified and stored at 4° C. The cell pellet was saved for NGS sequencing.

[0476] [Table 27]

[0477] MaxiSorp plates were coated overnight at 4° C. with 100 μL / well of neutravidin (ThermoFisher Scientific, Catalog No. 31000) at 5 μg / mL in 1×PBS. The next day, lead antibodies (7B3, 6E1, and 8E11) were first biotinylated using the EZ-Link™ Sulfo-NHS-LC-Biotinylation Kit (ThermoFisher Scientific, Catalog No. 21435). Phage libraries were diluted in 600 μL of SuperBlock (ThermoFisher Scientific, Catalog No. 37515) supplemented with 0.05% Tween®-20 and incubated at room temperature for 30 minutes. The blocked phage libraries were then incubated with 100 nM, 20 nM, or 0 nM (background) biotinylated lead antibodies, followed by incubation at room temperature for 2 hours with shaking. The MaxiSorp plate was washed 6 times with 1x PBST and blocked with SuperBlock (65 μL / well) for 30 min at room temperature. Then, 40 μL of 1x PBS with 1% Tween®-20 was added to each well and incubated for another 30 min at room temperature. The plate was washed 6 times with 1x PBST. The phage / antigen solution was diluted 6-fold in Superblock and added to the Maxisorp plate at 100 μL / mL and incubated for 15 min at room temperature with shaking. The plate was gently washed 10 times with 1x PBST buffer. Bound phages were eluted from the plate using 0.1 M HCl (100 μL / well) and neutralized with 1 M Tris base and 1% BSA. The eluted phages were titrated to measure enrichment (Table 28) and amplified using XL-1Blue cells. The selected phage library was purified and stored at 4°C. The cell pellet was saved for NGS sequencing.

[0478] [Table 28]

[0479] For sequencing, the ULBP6 mutagenized phagemid drive library was split into two amplicons. Amplicon 1 spans positions 29-142 of the protein N-terminus, while amplicon 2 covers the C-terminus of the protein including amino acids 101-218. All amino acid positions are numbered according to Uniport entry Q5VY80 ULBP6_HUMAN. Phagemid libraries were extracted from phage pellets using Monarch® Plasmid Miniprep Kit (New England Biolabs, Cat. No. T1010S). For each amplicon, NGS samples were prepared by two PCR cycles for adapters and barcodes. Amplicon PCR was performed with Q5® High-Fidelity 2X Master Mix (New England Biolabs, Cat. No. M0492S) using the primers shown in Table 29.

[0480] [Table 29]

[0481] The amplicon PCR reaction settings are shown in Table 30, and the thermal cycle settings are shown as Table 31. The amplicon PCR products were purified with 2% E-Gel EX Size Select (Invitrogen, Cat. No. G661012). Sequencing adapters were further added using Nextera XT DNA Library Preparation Kit (Illumina, Cat. No. FC-131-1024), and the reaction products were purified with AMPure XP beads (ThermoFisher Scientific, Cat. No. NC9933872). The prepared samples were loaded with 25% PhiX control (Illumina, FC-110-3001) and sequenced on an Illumina MiSeq NGS sequencer.

[0482] [Table 30]

[0483] [Table 31]

[0484] Sequencing results from each panning of 6E1, 7B3, and 8E11 were first analyzed separately (see Table 4 for sample details). In a first step, paired reads were pairwise aligned using the Flash Pairwise Aligner. Samples sequenced in sequencing run 1 (Table 4) were aligned, and alignments were restricted to reads with 100% identical read overlap to minimize the effect of sequencing errors in subsequent analysis. Due to slightly lower read quality in sequencing run 2, 1% mismatches in the overlapping regions were allowed. Sequencing samples and their statistics are shown in Table 32.

[0485] [Table 32-1] [Table 32-2]

[0486] Next, a position-specific weight matrix (PWM) was generated for each amino acid sample. It contains the frequency of each of the 20 amino acids at all positions analyzed. The assembled paired-end reads were translated into amino acid sequences. The occurrences of all 20 amino acids were then counted for each protein position across all reads. The analysis was restricted to the section of each amplicon covered by both R1 and R2. An analysis strategy was chosen to minimize the effect of sequencing errors on the position-specific weight matrix. Thus, the PWM covers the following regions of ULBP6: amplicon 1: 43-122, amplicon 2: 118-202. Using the PWM, the mutation frequency at each position of the PWM was calculated by dividing the sum of the number of all mutations at a given position by the total number of reads covering that position. Fx =m i x / r x In the formula, F x is the mutation frequency at position x, and m i x is the mutation number of mutation i at position x, and r x is the total number of reads at position x. The mutation frequencies were then used to calculate the log2 enrichment ratio (ER). ER = log2(S x / P x ) In the formula, S x is the frequency of mutations in the selected library at amino acid position x, and P x is the frequency of the mutation in the preselection library at amino acid position x.

[0487] To identify residues of functional epitopes, a linear regression analysis between ERs from different epitope bins was performed and the ER values ​​obtained for h7B3 were compared with those obtained for h6E1 and for h8E11 with those obtained for h6E1. The idea is that positions where mutations do not affect antibody binding or affect it indirectly (e.g. because they generally destabilize the antigen folding) should have small residuals, whereas mutations that directly affect the binding of one of the antibodies will have large residuals. Positions with residuals larger than 0.5 were considered to be part of the functional epitope.

[0488] result:

[0489] For h7B3, four ULBP6 positions were identified as having residuals greater than 0.5 when comparing ER to h6E1 (Table 33).

[0490] [Table 33]

[0491] For h8E11, two ULBP6 positions were identified as having residuals greater than 0.5 when comparing ER to h6E1 (Table 34).

[0492] [Table 34]

[0493] For h6E1, seven ULBP6 positions were identified as having residuals greater than 0.5 when comparing the ER with h7B3 and h8E11 (Table 35).

[0494] [Table 35]

[0495] Of the four residues identified as the functional epitope of 7B3, three, 183H, 184Y, and 189D, are surface exposed and located in the NKG2D protomer A binding site, as shown in Figure 18, panels A and B. These three residues also lie within the structural epitope identified by the crystal structure of 7B3 in complex with ULBP6.

[0496] The two residues identified as the functional epitope of 8E11 are also located within the NKG2D protomer A binding site on ULBP6, as shown in Figure 18, panels C and D. The functional epitope of 8E11 overlaps with that of 7B3, since they both share residue 189D within the epitope.

[0497] In contrast to the functional epitopes of 7B3 and 8E1, the functional epitope of 6E11 has different residues, two of which are located in the NKG2D protomer B binding site, as shown in Figure 18, panels E and F.

[0498] Example 19: ULBP6-Antibody Interactions by HDX

[0499] The interactions between human ULBP6 and six different monoclonal antibodies (mAbs) were assessed using hydrogen-deuterium exchange mass spectrometry (HDX).

[0500] Materials and Methods:

[0501] The ULBP6 protein used in this example had the following sequence before processing, as shown in Table 36:

[0502] [Table 36-1]

[0503] Residues 20-212 of SEQ ID NO:123 correspond to residues 26-218 of the native human ULBP6_02 sequence (Uniprot:Q5VY80) shown herein as SEQ ID NO:7.

[0504] h7B3(E+Afu), h8E11(E+), h1D4(E+), h6E1(E+), h8E8(E+), and mab1298 (mIgG2a, R&D SYSTEMS, clone number 165903) were tested and are referred to as 7B3, 8E11, 1D4, 6E1, 8E8, and 1298 in Figures 21A, 21B, and 22, respectively.

[0505] Buffers and solvents used in the HDX experiments described here were as shown in Table 36.

[0506] [Table 36-2]

[0507] Liquid handling, deuterium labeling, online digestion, and LCMS analysis were performed on an automated HDX-MS system equipped with a LEAP H / DX PAL™ robot (LEAP-TRAJAN), an ACQUITY M-Class UPLC (WATERS), an HDX Manager (WATERS), and a CYCLIC mass spectrometer (WATERS). Software packages used for data acquisition and interpretation included MASSLYNX v4.2 (WATERS), Chronos V4.11.0-beta.24 (AXELSEMRAU), ProteinLynxGlobal SERVER (PLGS) v3.0.2 (WATERS), and HDEXAMINER v3.3.0 (LEAP-TRAJAN).

[0508] All samples were kept at 0°C prior to HDX analysis. Samples for HDX were prepared using an automated LEAP-TRAJAN liquid handling workstation. Samples for peptide mapping were prepared as 40 pmol ULBP6 protein stock solution in 5 μl MOPS-H2O buffer and diluted with 55 μl MOPS-H2O to a final volume of 60 μl. Prepared samples were labeled in 5 μl MOPS-H2O buffer as 40 pmol ULBP6 protein (apo reference) or as 40 pmol ULBP6 protein complexed with 0.75 molar equivalents of monoclonal antibodies (mAb1298, h7B3, h8E8, h1D4, h8E11, and h6E1, respectively), diluted with 55 μl MOPS-D2O to a final volume of 60 μl, and incubated at 20°C for 20, 200, 2000, or 20,000 seconds.

[0509] For peptide mapping, 55 μl of the reaction mixture was quenched with pre-chilled quench buffer (1:1) for 1 min (labeling experiments) or 1–20 min (peptide mapping) at 0° C. Then, 105 μl of the reaction mixture was injected onto an immobilized digestion column (Nepenthesin 2, Affipro part number AP-PC-004) and kept at 15° C.

[0510] During digestion, released peptides were captured on a guard column (2.1 x 5 mm, WATERS VANGUARD BEH C18 1.8 μm 130 Å guard column, part number 186003975) maintained at 0.5° C. Digestion / capture was carried out for 4 minutes in solvent A. Flow rates are as shown in Table 37.

[0511] [Table 37]

[0512] The captured peptides were then eluted from the guard column onto an analytical column (1.0x50mm ULP BEH C18, part number 186002344) at 0.5°C in solvent A according to the initial conditions and analytical gradient shown in Table 38.

[0513] [Table 38]

[0514] During elution, mass spectra were recorded in either HDMSE ion mobility acquisition mode (for peptide mapping experiments) or HDMS ion mobility mode (for labeling experiments) for 2.0-15.0 min. Four replicates were recorded for each of the peptide mapping time course and labeling time course.

[0515] Raw peptide mapping data were processed with PLGS. PLGS output from multiple replicates was further optimized for sequence coverage, data quality, and reproducibility by an in-house implementation of the selection algorithm LARS (Sorensen & Salbo, 2018). The LARS procedure was applied after hard filter criteria including PLGS peptide score >6.5, peptide length ≥25, and mass error ≥7 ppm.

[0516] Using HDEXAMINER, raw data from each labeling time course were analyzed against the LARS peptide map to determine update (ΔFU) values ​​for all peptides. Peptides that showed potential errors upon visual inspection (e.g., overlaps, low signal-to-noise ratio, bimodal isotope distribution, or high carryover) were excluded. The final HDXAMINER state files were uploaded to an in-house HDX database and visualization tool for interpretation.

[0517] result:

[0518] A final set of 131 peptides covering 87.7% of the total ULBP6 construct sequence, shown herein as SEQ ID NO: 123 (96.3% of the native portion of the ULBP6_02 sequence, shown herein as SEQ ID NO: 7), were included in the evaluation, as shown in FIG. 19.

[0519] To distinguish monoclonal antibody-induced site-specific protection from global protection at the peptide level, a cutoff value of one standard deviation was applied. All residues of any peptide above this threshold were considered significantly protected. The incorporation rate difference for each peptide detected is shown in Figure 20A (amino acids 26-123 of ULBP6) and Figure 20B (amino acids 114-226 of ULBP6). ULBP6 peptides determined to be significantly protected by mAb1298, h7B3, h8E8, h1D4, h8E11, and h6E1 are identified in Figure 21 as underlined and bold. The crystal structure of the NKG2D-ULBP6 complex is shown in Figure 22, panels A-F. Accordingly, residues in the ULBP6 peptide with significant protection by mAb1298 (panel A), h7B3(E+Afu) (panel B), h8E8(E+) (panel C), h1D4(E+) (panel D), h8E11(E+) (panel E), and h6E1(E+) (panel F), as determined by HDX, are shown in black.

[0520] According to the significantly protected residues, the antibodies were determined to be of three classes of HDX protection patterns, as shown in Table 39.

[0521] [Table 39]

[0522] Significant protection of ULBP6 residues 35 / 36–58 / 59 was provided by class 1 antibodies. Lower but significant levels of protection of ULBP6 residues 35 / 36–58 / 59 were provided by class 2 antibodies, with stronger protection of ULBP6 residues 81 / 82–96, 150–157, and 181–190. Significant protection of ULBP6 residues 150–169 was provided by class 3 antibodies.

[0523] Considering the location of the above patterns in the crystal structure of the NKG3D / ULBP6 complex, the HDX data were compatible with the following binding site:

[0524] The HDX protection pattern of class 1 antibodies was restricted to β-strands 1 and 2, including the intervening connecting loop, as shown in Figure 22, panels A (mAb1298), D (h1D4(E+)), and F (h6E1(E+)), consistent with these monoclonal antibodies targeting β-sheets distant from the NKG2D binding site without disrupting the NDG2D / ULBP6 complex.

[0525] Amino acid segments with strong protection patterns as described herein are shown in Table 40.

[0526] [Table 40]

[0527] Strong protection patterns of class 2 antibodies were observed on the first α-helix (SEQ ID NO: 126 and / or SEQ ID NO: 127), the second α-helix (SEQ ID NO: 127), and the intervening loop (SEQ ID NO: 128). An additional weaker protection pattern on the opposite side of the aforementioned α-helix, β-strand 2 (residues 36-58 of ULBP6), was on the border of significance. As seen in Figure 20A and Figure 20B, the peptide spanning residues 35 / 36-58 / 59 may be allosteric and does not necessarily indicate direct contact by the antibody on these residues. Overall, the protection patterns of class 2 antibodies were consistent with binding to α-helices located at the NKG2D-ULBP6 complex interface. This suggests that class 2 antibodies can disrupt the interaction between NKG2D and ULBP6.

[0528] Class 3 antibodies showed a very strong HDX protection pattern for residues 150 to 169. Because the dominant protective features are located on the flanks of the NKG2D-ULBP6 complex, it was not predicted whether this antibody would disrupt this complex.

[0529] Example 20: In vivo tumor model for evaluating the activity of anti-ULBP6 antibodies

[0530] The examples illustrate selected in vivo tumor model studies of the functional activity of one or more anti-ULBP6 antibodies provided herein, but may not include all in vivo studies and / or results that may demonstrate the functional activity or other properties of the ULBP6 binding proteins provided herein.

[0531] Materials and Methods

[0532] The anti-ULBP6 antibodies provided herein can be prepared for administration to a subject. For example, the anti-ULBP6 antibodies can be prepared as a dose of about or exactly 0.1 mg / kg, 1 mg / kg, 10 mg / kg, 100 mg / kg, or 1000 mg / kg, or within a range between any two of the above doses. In some embodiments, the anti-ULBP6 antibodies can be prepared such that a predetermined volume (e.g., at least about 1 mL / kg, not more than about 40 mL / kg, or between 1 mL / kg and 40 mL / kg) can be administered at one or more doses (e.g., doses shown herein or other doses confirmed to be effective). In some embodiments, doses of unknown efficacy can be administered. A control preparation can be prepared that can substantially match the buffer and concentration of the test article or its effect. In some embodiments, a control preparation can be identical to a preparation containing an anti-ULBP6 antibody, but the properties of the added antibody include no physiologically significant binding to ULBP6, no physiologically significant binding to FcyR, no physiologically significant binding to both ULBP6 and FcyR, or the absence of any antibody.

[0533] The mouse tumor model can be used to examine the functional activity of one or more anti-ULBP6 antibodies. Briefly, the mouse tumor model can be engineered to express human ULBP6 or injected with mouse or human hULBP6-expressing tumor cells. The mouse tumor model can host human immune cells or retain its own immune system.

[0534] Anti-ULBP6 antibody preparations or control preparations can be administered to mouse tumor models. For example, anti-ULBP6 can be administered, for example, in one or more doses as indicated above. The preparations can be administered to animals, for example, as several doses, one or more times per week. For example, animals can be sacrificed at different time points with different tumor sizes, at which time tumor, lymphatic tissue, blood, and / or other samples can be taken. In some embodiments, tumor size can be monitored, for example, using calipers or by one or more suitable imaging-based techniques.

[0535] Collected blood and / or tissue samples can be processed to extract informative phenotypic readouts of the biomarkers and to prepare slides for IHC or other pathology evaluation (e.g., for DNA and / or RNA).

[0536] The processed samples can be evaluated to determine the presence, absence, or levels of immune cells and their relative state (eg, activation, proliferation, and / or cytotoxicity).

[0537] result:

[0538] The inventors expect that upon treatment with anti-ULBP6 in vivo, the cytotoxicity of CD8 T cells and NK cells against tumor cells will be reactivated, and as a result, a change in the phenotype of CD8 T cells and NK cells and / or an inhibition of tumor growth will be detectable.

[0539] The results are described in this section for the mice depicted herein, but are also anticipated for other subjects (e.g., rats, rabbits, monkeys, cats, dogs, hamsters, humans) that have ULBP6-expressing tumors and / or metastases and are administered preparations containing anti-ULBP6 antibodies.

[0540] Tumor growth (e.g., measured as change in tumor volume over time by an accepted method) in a mouse or group of mice administered a preparation comprising an anti-ULBP6 antibody may be slowed by at least about 25% slower, at least about 50% slower, at least about 60% slower, at least about 70% slower, at least about 80% slower, at least about 90% slower, at least about 95% slower, or at least about 99% slower, or within a range between any two of the foregoing values, compared to tumor growth in a mouse or group of mice administered a preparation comprising a control antibody. The slowing of tumor growth may be dose-dependent, for example, for the dose ranges set forth herein.

[0541] Tumor size may be reduced by at least about 25%, at least about 50%, at least about 60%, at least about 70%, at least about 80%, at least about 90%, at least about 95%, or at least about 99%, or a range between any two of the foregoing values, compared to the start of treatment in a mouse or group of mice administered a preparation comprising an anti-ULBP6 antibody. The reduction in tumor size may be dose-dependent, for example, for the dose ranges set forth herein.

[0542] In a mouse or group of mice administered a preparation comprising an anti-ULBP6 antibody, metastasis may be reduced by at least about 20%, at least about 50%, at least about 60%, at least about 70%, at least about 80%, at least about 90%, at least about 95%, or at least about 99%, or a range between any two of the foregoing values, compared to metastasis in the same mouse or group of mice prior to administration of the preparation.

[0543] Activation of one or more types of immune cells in mice administered a preparation comprising an anti-ULBP6 antibody can be increased by at least about 20%, at least about 50%, at least about 60%, at least about 70%, at least about 80%, at least about 90%, at least about 95%, at least about 99%, or a range between any two of the foregoing values, in a mouse or group of mice administered a preparation comprising an anti-ULBP6 antibody compared to a control mouse or group of mice.

Claims

1. ULBP6 binding protein, a) comprising any one or combination of CDRs selected from (i) CDRH1, CDRH2, CDRH3 from SEQ ID NOs: 16-19, and / or CDRL1, CDRL2, CDRL3 from SEQ ID NOs: 26-29, or (ii) a CDR variant of (i), said CDR variant having one, two or three amino acid modifications; or b) The ULBP6 binding protein, comprising a VH region comprising a sequence at least 80% identical to any one of SEQ ID NOs: 21 to 24, and / or a VL region comprising a sequence at least 80% identical to any one of SEQ ID NOs: 31 to 34.

2. a) the CDRs of (i) CDRH1 of SEQ ID NO: 36, 39, 42, or 45, CDRH2 of SEQ ID NO: 37, 40, 43, or 46, CDRH3 of SEQ ID NO: 38, 41, 44, or 47, CDRL1 of SEQ ID NO: 51, 54, 57, or 60, CDRL2 of SEQ ID NO: 52, 55, 58, or 61, and 2. The ULBP6 binding protein of claim 1, wherein the CDRL3 is selected from the group consisting of SEQ ID NO: 53, 56, 59, or 62.

3. all six CDRs are present in said binding protein; and or selected from CDRH1 of SEQ ID NO: 36, CDRH2 of SEQ ID NO: 37, CDRH3 of SEQ ID NO: 38, CDRL1 of SEQ ID NO: 51, CDRL2 of SEQ ID NO: 52, and CDRL3 of SEQ ID NO: 53; or selected from CDRH1 of SEQ ID NO:39, CDRH2 of SEQ ID NO:40, CDRH3 of SEQ ID NO:41, CDRL1 of SEQ ID NO:54, CDRL2 of SEQ ID NO:55, and CDRL3 of SEQ ID NO:56; or selected from CDRH1 of SEQ ID NO: 42, CDRH2 of SEQ ID NO: 43, CDRH3 of SEQ ID NO: 44, CDRL1 of SEQ ID NO: 57, CDRL2 of SEQ ID NO: 58, and CDRL3 of SEQ ID NO: 59; 3. The ULBP6 binding protein of claim 1 or 2, selected from CDRH1 of SEQ ID NO: 45, CDRH2 of SEQ ID NO: 46, CDRH3 of SEQ ID NO: 47, CDRL1 of SEQ ID NO: 60, CDRL2 of SEQ ID NO: 61, and CDRL3 of SEQ ID NO:

62.

4. or selected from CDRH1 of SEQ ID NO: 36, CDRH2 of SEQ ID NO: 37, CDRH3 of SEQ ID NO: 38, CDRL1 of SEQ ID NO: 51, CDRL2 of SEQ ID NO: 52, and CDRL3 of SEQ ID NO: 53; or selected from CDRH1 of SEQ ID NO:39, CDRH2 of SEQ ID NO:40, CDRH3 of SEQ ID NO:41, CDRL1 of SEQ ID NO:54, CDRL2 of SEQ ID NO:55, and CDRL3 of SEQ ID NO:56; or selected from CDRH1 of SEQ ID NO: 42, CDRH2 of SEQ ID NO: 43, CDRH3 of SEQ ID NO: 44, CDRL1 of SEQ ID NO: 57, CDRL2 of SEQ ID NO: 58, and CDRL3 of SEQ ID NO: 59; A ULBP6 binding protein comprising six CDRs selected from CDRH1 of SEQ ID NO:45, CDRH2 of SEQ ID NO:46, CDRH3 of SEQ ID NO:47, CDRL1 of SEQ ID NO:60, CDRL2 of SEQ ID NO:61, and CDRL3 of SEQ ID NO:

62.

5. The binding protein is a VH region that is at least 80% identical to SEQ ID NO:21 and a VL region that is at least 80% identical to SEQ ID NO:31; a VH region that is at least 80% identical to SEQ ID NO:22 and a VL region that is at least 80% identical to SEQ ID NO:32; a VH region that is at least 80% identical to SEQ ID NO:23 and a VL region that is at least 80% identical to SEQ ID NO:33; or 5. The ULBP6 binding protein of claim 4, comprising a VH region that is at least 80% identical to SEQ ID NO:24 and a VL region that is at least 80% identical to SEQ ID NO:

34.

6. The binding protein is a) a heavy chain (HC) sequence that is at least 80% identical to any one of SEQ ID NOs: 75-78, 85-88, or 95-98; and / or b) A ULBP6 binding protein according to any one of the preceding claims, comprising a light chain (LC) sequence which is at least 80% identical to any one of SEQ ID NOs: 105 to 108.

7. A ULBP6-binding protein that binds to ULBP6 and ULBP2 with a KD of 1 nM or less, and binds to ULBP5 with a KD of 1 nM or more.

8. 4. The ULBP6 binding protein of any one of the preceding claims, wherein the binding protein inhibits the interaction of human ULBP6 with human NKG2D by 80% or more.

9. Binds to human ULBP6, and for binding to human ULBP6, comprising the VH domain sequence of SEQ ID NO:21 and the VL domain sequence of SEQ ID NO:31; comprising the VH domain sequence of SEQ ID NO: 22 and the VL domain sequence of SEQ ID NO: 32; comprising the VH domain sequence of SEQ ID NO: 16 and the VL domain sequence of SEQ ID NO: 26; or A ULBP6 binding protein that competes with a reference ULBP6 binding protein comprising the VH domain sequence of SEQ ID NO: 17 and the VL domain sequence of SEQ ID NO:

27.

10. ULBP6 binding protein according to any one of the preceding claims, wherein the binding protein is an antibody or an antigen-binding fragment thereof.

11. The ULBP6 binding protein of claim 10, wherein the antibody or antigen-binding fragment thereof inhibits binding of soluble ULBP6 to NKG2D.

12. The ULBP6 binding protein of claim 11, wherein the antibody or antigen-binding fragment thereof inhibits binding of soluble ULBP6 to NKG2D by 80% or more.

13. The ULBP6 binding protein of any one of claims 10 to 12, wherein the antibody or antigen-binding fragment thereof comprises a modified Fc region.

14. 14. The ULBP6 binding protein of claim 13, wherein the modified Fc region comprises the amino acid substitutions S239D and I332E (numbered according to the EU index).

15. The ULBP6 binding protein of any one of claims 10 to 13, wherein the antibody or antigen-binding fragment thereof is defucosylated.

16. 14. The ULBP6 binding protein of claim 13, wherein the modified Fc region comprises amino acid substitutions selected from S239D / A330L / I332E, G236A / S239D / A330L / I332E, L235V / F243L / R292P / Y300L / P396L, or F243L / R292P / Y300L / V305I / P396L (numbered according to the EU index).

17. A nucleic acid sequence encoding one or both of the HC and LC of a ULBP6 binding protein according to any one of the preceding claims.

18. An expression vector comprising the nucleic acid of claim 17.

19. 19. A recombinant host cell comprising the nucleic acid sequence(s) of claim 17 or the expression vector(s) of claim 18.

20. 20. A method for producing a ULBP6 binding protein, comprising culturing a recombinant host cell according to claim 19 under conditions suitable for expression of the nucleic acid sequence(s) or vector(s), thereby producing a polypeptide comprising the ULBP6 binding protein.

21. 21. A ULBP6 binding protein produced by the method of claim 20.

22. A cell line engineered to express the ULBP6 binding protein of any one of claims 1 to 16 and 21.

23. A pharmaceutical composition comprising the ULBP6 binding protein according to any one of claims 1 to 16 and 21, and a pharma- ceutically acceptable excipient.

24. 23. A method for treating cancer in a subject in need of such treatment, comprising administering to the subject a therapeutically effective amount of a ULBP6 binding protein described in any one of claims 1 to 16 and 21, or a therapeutically effective amount of the pharmaceutical composition described in claim 23.

25. 25. The method of claim 24, wherein the cancer is head and neck cancer, cervical cancer, bladder cancer, lung cancer, colorectal cancer, pancreatic cancer, or glioblastoma.

26. 26. The method of claim 24 or 25, wherein the subject is a human.

27. 27. The method of claim 24, 25, or 26, further comprising determining whether the subject expresses soluble ULBP6.

28. A ULBP6 binding protein according to any one of claims 1 to 16 and 21 or a pharmaceutical composition according to claim 23 for therapeutic use.

29. The ULBP6 binding protein according to any one of claims 1 to 16 and 21 or the pharmaceutical composition according to claim 23, for use in the treatment of cancer.

30. 30. The ULBP6 binding protein of claim 29, wherein the cancer is head and neck cancer, cervical cancer, bladder cancer, lung cancer, colorectal cancer, pancreatic cancer, or glioblastoma.

31. A ULBP6 binding protein for the treatment of basal cell carcinoma.

32. ULBP6 antibody or fragment thereof, a) binds to one or more residues of SEQ ID NO: 125 and / or SEQ ID NO: 126; b) binds to a conformational epitope of SEQ ID NO: 125, and / or c) which protects one or more residues of SEQ ID NO: 125 or SEQ ID NO: 126 from deuterium exchange in HDX-MS analysis of SEQ ID NO: 124; d) binds to human ULBP6, resulting in peptides derived from SEQ ID NO: 125 and / or SEQ ID NO: 126 of human ULBP6 that are less prone to deuterium incorporation than the corresponding peptides derived from uncomplexed human ULBP6; e) binds to the NKG2D binding interface of human ULBP6; and / or f) The ULBP6 antibody or a fragment thereof, which interferes with the interaction of human ULBP6 with NKG2D by binding to one or more residues selected from E96, E154, Y184, and D189 of SEQ ID NO:

124.

33. ULBP6 antibody or fragment thereof, a) binds to one or more residues of SEQ ID NO: 128 and / or SEQ ID NO: 129; b) binds to a conformational epitope of SEQ ID NO: 128 and / or SEQ ID NO: 129; c) protects one or more residues of SEQ ID NO: 128 and / or SEQ ID NO: 129 from deuterium exchange in HDX-MS analysis; and / or d) binds to human ULBP6, resulting in peptides derived from SEQ ID NO: 128 and / or SEQ ID NO: 129 of human ULBP6 that are less prone to deuterium incorporation than the corresponding peptides derived from uncomplexed human ULBP6; e) binds to the NKG2D binding interface of human ULBP6; and / or f) The ULBP6 antibody or a fragment thereof, which interferes with the interaction of human ULBP6 with NKG2D by binding to one or more residues selected from E96, E154, Y184, and D189 of SEQ ID NO:

124.

34. The binding protein is a) binds to one or more residues of SEQ ID NO: 127; b) binds to a conformational epitope of SEQ ID NO: 127; c) protects one or more residues of SEQ ID NO: 127 from deuterium exchange in HDX-MS analysis; and / or d) A ULBP6 binding protein as described in claim 32 or claim 33, which binds to human ULBP6, resulting in a peptide derived from SEQ ID NO: 127 of human ULBP6 that is less susceptible to deuterium incorporation than the corresponding peptide derived from uncomplexed human ULBP6.

Images