Anti-HLA-G antibodies

Abstract

The present invention relates to antibodies against HLA-G, and formulations comprising same. The present invention further relates to the use of HLA-G antibodies and formulations in therapy, in particular in the treatment of solid tumors.

Description

[Technical field]

[0001] The present invention relates to antibodies against HLA-G, and formulations comprising same. The present invention further relates to the use of HLA-G antibodies and formulations in therapy, in particular in the treatment of solid cancers. [Background technology]

[0002] Class I human leukocyte antigens (HLA-I) include the classical antigens HLA-A, HLA-B and HLA-C, and the non-classical antigens HLA-E, HLA-F and HLA-G. Human leukocyte antigen G (HLA-G) is a non-classical HLA class I molecule expressed in humans and encoded by the HLA-G gene. HLA-G is composed of three globular domains (α,β-β) associated with a light chain, i.e., beta 2 microglobulin (B2m). 1 , α 2 and α 3 ), a heavy chain having a

[0003] Seven isoforms of HLA-G have been identified, four membrane-bound (HLA-G1, HLA-G2, HLA-G3, HLA-G4) and three soluble (HLA-G5, HLA-G6, and HLA-G7), which are the result of alternative splicing of the HLA-G primary transcript.

[0004] HLA-G is normally expressed on the placental cytotrophoblast. It has been reported that HLA-G expression is associated with pathological conditions such as inflammatory diseases and cancer. Notably, HLA-G is reported to be a tolerogenic molecule that is specifically upregulated in solid cancers and associated with poor prognosis.

[0005] HLA-G is known to exert immunomodulatory activity by binding to at least three receptors expressed on a variety of myeloid and lymphoid cells: The inhibitory receptor LILRB1 (for leukocyte immunoglobulin-like receptor B1), also called ILT2 or CD85j, expressed on lymphoid cells (B cells, some T cells, and NK cells) and myeloid cells (monocytes, macrophages, and dendritic cells) The inhibitory receptor LILRB2 (for leukocyte immunoglobulin-like receptor B2), also called ILT4 or CD85d, expressed on myeloid cells (monocytes, macrophages and dendritic cells); and · The regulatory receptors KIR2DL4 or CD158d expressed on NK cells.

[0006] HLA-G inhibits immune cell functions by direct binding to its inhibitory receptors. HLA-G is mainly involved in the α 3 It has been reported that the IL-1 domain has a tolerogenic function mediated by its interaction with ILT2 and ILT4. ILT2 recognizes only B2m-associated HLA-G molecules, whereas ILT4 recognizes both B2m-associated and B2m-free HLA-G molecules.

[0007] Due to this immune-inhibitory function, HLA-G expression by tumors can induce an immunosuppressive environment, enable tumor immune evasion, and ultimately reduce patient survival. Therefore, antibody-mediated blockade of HLA-G may provide an effective strategy to alleviate tumor local immune suppression, promote the development of anticancer immunity, and provide sustained treatment.

[0008] Theoretically, the antitumor efficacy of HLA-G blocking antibodies could be increased by including an active Fc component that could bind FcγR to immune effector cells to enable direct killing of HLA-G+ tumor cells; however, the reported expression of HLA-G mRNA and protein in several normal tissues, including the pancreas and pituitary, suggests that the use of active Fc may result in unacceptable toxicity, thus precluding its use in therapy.

[0009] HLA-G shares high similarity with other HLA-I molecules (i.e., HLA-A, HLA-B, HLA-C, HLA-E and HLA-F). Of the 338 amino acid positions in the HLA-G protein, only 20 have residues unique to HLA-G, and these 20 do not occur at that position in any of the other approximately 5000 human HLA-I molecules. This makes it extremely difficult to produce antibodies with high specificity to HLA-G without cross-reactivity to other HLA-I molecules.

[0010] Today, commercially available HLA-G antibodies are available. However, some are reported to lack specificity for HLA-G (e.g., cross-reactive with other HLA-I molecules) and all have epitopes in the α1 and α2 domains of HLA-G, which are distant from the HLA-G ILT2 / 4 binding site in the α3 domain, and therefore are not predicted to block the interaction between HLA-G and ILT2 and / or ILT4. Antibody 87G, which binds to an epitope in the HLA-G α1 domain, has been reported to alleviate HLA-G inhibition of immune cell function. However, it has not been reported that 87G and other commercially available HLA-G antibodies were able to block HLA-G ILT2 / 4 interactions. Due to lack of specificity and / or blocking activity, commercially available antibodies are not suitable for the development of HLA-G therapeutic antibodies.

[0011] Other antibodies that bind to HLA-G have been reported in WO20019202040 and WO2020069133, and such antibodies appear to modulate one or more HLA-G activities, although their binding sites (i.e., epitopes) on HLA-G have not been characterized.

[0012] To date, the efficacy of antibodies against HLA-G has not been demonstrated in patients, particularly for the treatment of solid tumors.

[0013] Thus, there remains a need to provide antibodies that bind to HLA-G and have therapeutically useful biological properties, such as improved pharmacokinetic properties and / or improved biological function in humans (e.g., specificity, binding affinity, neutralization, and / or cell cytotoxicity and phagocytosis) and / or reduced toxicity. Summary of the Invention

[0014] The present invention addresses the above needs by providing new antibodies against HLA-G with the structural and functional properties described herein, in particular high specificity for HLA-G, ability to block the interaction of HLA-G with its receptors ILT2 and ILT4, useful in therapy, in particular in the treatment of solid cancers. The inventors provide the first evidence supporting the use of an antibody format containing an active Fc, whereby direct tumor cell killing is possible in patients, as the epitopes recognized by the antibodies of the present invention are absent in normal tissues, including the pituitary and pancreas.

[0015] In particular, the present invention relates to an antibody that specifically binds to human HLA-G, a. CDR-L1 comprising SEQ ID NO:1; CDR-L2 comprising SEQ ID NO:2, and CDR-L3 comprising SEQ ID NO:3 and a light chain variable region comprising b. CDR-H1 comprising SEQ ID NO:4; CDR-H2 comprising SEQ ID NO:5, and CDR-H3 comprising SEQ ID NO:6 and a heavy chain variable region comprising the [Brief description of the drawings]

[0016] [Figure 1] HLA-G isoforms (Figure 1B in Carosella et al., Blood, Vol. 111, n° 10, 2008) [Diagram 2]Humanization of the rabbit variable light chain sequence of antibody 12389. Grafts 12389gL1, gL2 and gL3 are humanized grafts of the rabbit variable light chain of antibody 12389 using the IGKV1D-13 human germline as the acceptor framework. The CDRs are shown in bold / underlined. Donor residues are shown in bold / italics and shaded grey: V3 and Q70. [Diagram 3] Humanization of rabbit variable heavy chain sequence of antibody 12389. Grafts 12389 gH1, gH4, gH5, gH6, gH8, gH9, gH11, gH12, gH13, gH14, gH15 and gH16 are humanized grafts of rabbit variable heavy chain of antibody 12389 using IGHV3-66 human germline as acceptor framework. CDRs are shown in bold / underlined. Donor residues are shown in bold / italic and shaded grey: V24, I48, G49, K71, S73, V78 and G96. [Figure 4] Specificity of HLA-G02 in PBMC assays from 50 different donors. Results are expressed as MFI of each CD4+ cell population for each donor and antibody. [Diagram 5] Specificity of HLA-G02; binding to HLA-G1, HLA-G2, HLA-G3 and HLA-G4 expressed on cells (Figure 5A). Binding to HLA-G2 (Figure 5B). [Figure 6] Percentage of depleted Epcam+GFP+ HCT116 target cells after treatment with various anti-HLA-G antibodies or an IgG1 isotype control antibody. Each antibody was tested at two different concentrations, either 1 μg / ml (white bars) or 0.01 μg / ml (striped bars). The E:T ratio was 3.5:1. Each bar represents the mean (and range) of three data points, and each dot / square is an individual replicate. Data are from one representative donor. HLA-G01-HLA-G08 are represented by their respective ID numbers (01-08). [Figure 7]Percentage of Epcam+GFP+ HCT116 target cells depleted after treatment with anti-HLA-G antibodies HLA-G01 and HLA-G02 or IgG1 isotype control antibody from three separate experiments (three different donors). Antibodies were tested at 1 µg / ml (Figure 7A) or 0.01 µg / ml (Figure 7B). E:T ratios ranged from 2.5 to 3:1. Each bar represents the mean (and range) of data from an individual experiment, each dot, square or triangle is an individual replicate. [Figure 8] Percent depletion of Epcam+GFP+ HCT116 cells (vertical axis) after treatment with titrations of anti-HLA-G antibodies HLA-G01 and HLA-G02 (horizontal axis: antibody concentration in μg / ml) compared to isotype control IgG1. The E:T ratio was 4:1. Each point represents the mean (and range) of three replicates. Data shown are from a single representative donor. [Figure 9A] Percent depletion of JEG3 cells (vertical axis) after treatment with conventional HLA-G02 IgG1 (solid line) or defucosylated HLA-G02 IgG1 ("aF HLA-G02", dotted line). Horizontal axis: antibody concentration (μg / ml). E:T ratio was 10:1. Each point represents the mean (and range) of two replicates. Data from a single representative donor are shown. [Figure 9B] Percent depletion (vertical axis) of Epcam+GFP+ HCT116 cells after treatment with conventional HLA-G02 IgG1 (solid line) or defucosylated HLA-G02 IgG1 ("aF HLA-G02", dotted line) compared to isotype control IgG1. Horizontal axis: antibody concentration (μg / ml). E:T ratio was 5:1. Each point represents the mean (and range) of three replicates. Data from a single representative donor are shown. [Figure 10] Titration of HLA-G specific phagocytic activity of HLA-G02 against mock transfected K562 target cells (FIG. 10A) and HLA-G / B2m transfected K562 target cells (FIG. 10B) in comparison with anti-CD47 antibody ("aCD47") and isotype control IgG1. Vertical axis: percentage of CTY+CD11b+ double positive cells; horizontal axis: antibody concentration (μg / ml). [Figure 11] Titration of HLA-G specific phagocytic activity of conventional and defucosylated (aF) formats of HLA-G02 against mock-transfected K562 target cells (FIG. 11A) and HLA-G expressing K562 target cells (FIG. 11B) in comparison with anti-CD47 antibody and isotype control IgG1. Vertical axis: percentage of CTY+CD11b+ double positive cells; horizontal axis: antibody concentration (μg / ml). [Figure 12] VR12389 (black surface representation) blocks the interaction of HLA-G with ILT-2 and ILT4. Figure 12A: Crystal structure of ILT2 (cartoon, white) complexed with HLA-G (cartoon, grey) and ss2M (mesh, grey) (PDB ID 6AEE). Figure 12B: Superposition with the crystal structure of VR12389 complexed with HLA-G and ss2M shows that VR12389 (surface, black) blocks the interaction of ILT2 with HLA-G. Figure 12C: Crystal structure of ILT4 (cartoon, white) complexed with HLA-G (cartoon, grey) and ss2M (mesh, grey) (PDB ID 2DYP). FIG. 12D: Superposition of the crystal structure of VR12389 in complex with HLA-G and ss2M shows that VR12389 (surface, black) blocks the interaction of ILT4 with HLA-G. [Figure 13] Tumor cell killing assay. Figure 13A: Data obtained with anti-PDL1 from RCC. Figure 13B: Data obtained with anti-PDL1 from CRC. Figure 13C: Data obtained with HLA-G02 from RCC. Figure 13D: Data obtained with HLA-G02 from CRC. Data are expressed as % dead cells for isotype control in light grey, anti-PDL1 or HLA-G02 in dark grey, and cultures treated with anti-PDL1 or HLA-G02; in black, a 1.5-fold or greater increase in cell death was observed. [Figure 14]Depletion of HLA-G transfected HCT116 cells at various effector:target ratios 2.5-3 hours after treatment with conventional HLA-G02 or defucosylated HLA-G02. The number of live HLA-G GFP+ HCT116 cells was determined by flow cytometry and percent depletion (vertical axis) was calculated compared to a control without antibody. Horizontal axis: E:T ratio (effector:target ratio). Data shown are from a single donor (1). Each data point represents the mean of three replicates. Error bars represent 95% confidence intervals. [Figure 15] Specificity of defucosylated HLA-G02 ("aF HLA-G02") in PBMC assays from 10 different donors. Results are expressed as MFI of each CD4+ cell population for each donor and antibody. [Figure 16] CDC mediated by defucosylated HLA-G02. Figure 16A: Effect of serum activity on HLAG-β2m-Reh cell lysis (concentration response curves showing aF HLA-G02-mediated CDC of HLAG-β2m-Reh in the presence of activated or heat-inactivated serum. Vertical axis: lysis (%) normalized to the minimum and maximum controls (N=1, 4-PL fit); Horizontal axis: antibody concentration [M]). Figure 16B: Dependence of aF HLA-G for HLA-G02-mediated lysis (concentration response curves showing aF HLA-G02-mediated CDC of HLAG-β2m-Reh cells or Reh cells. Vertical axis: lysis (%) normalized to the minimum and maximum controls (N=1, 4-PL fit); Horizontal axis: antibody concentration [M]). FIG. 16C: aF HLA-G02-mediated CDC of HLAG-β2m-Reh cells (Concentration-response curve showing aF HLA-G02-mediated CDC of HLAG-β2m-Reh. Vertical axis: lysis (%) normalized to minimum and maximum controls (N=3, mean ± SEM, 4-PL fit); Horizontal axis: antibody concentration [M]). [Figure 17]Concentration-dependent effect of HLA-G02 and defucosylated HLA-G02 on phagocytosis of HLA-G expressing target cells or mock transfected cells compared to αCD47 antibody. Figures 17A and B: Representative data (mean ± SD) (ordinate: percentage of CTY+CD11b+ double positive cells; abscissa: antibody concentration (μg / ml)). Figures 17C and D: Combined data generated using monocytes from three separate donors in duplicate, across two independent experiments (ordinate: percentage of phagocytosis adjusted for isotype control = (antibody mean - isotype mean) ± SD; abscissa: antibody concentration (μg / ml)). [Figure 18] Target cell killing mediated by defucosylated HLA-G02. Percentage of depletion of HLA-G expressing target cells (vertical axis) by ADCP with different concentrations of defucosylated HLA-G02 or αCD47 (horizontal axis (μg / ml)) after overnight incubation (from four separate donors, in duplicate, across two independent experiments). Mean and standard deviation compared to isotype control. [Figure 19] Representative data of the concentration-dependent effect of defucosylated HLA-G02 and HLA-G02 IgG4P FALA on phagocytosis of HLA-G expressing cells alone (Figure 19A) or in combination with anti-CD47 Ab (1 μg / mL, Figure 19B) compared to the respective isotype controls. Vertical axis: percentage of CTY+CD11b+ double positive cells; horizontal axis: antibody concentration (μg / ml). [Figure 20] Phagocytosis of HLA-G expressing cells treated with defucosylated HLA-G02 alone or in combination with anti-CD47 antibody (1 μg / mL). Figure 20A represents data for defucosylated HLA-G02 (3 donors, duplicates, 2 independent experiments). Figure 20B represents data with HLA-G02 IgG4P FALA (combined data from 4 donors, each analyzed in duplicate, 3 independent experiments). Phagocytosis values ​​(%, ordinate) were adjusted to the respective isotype control (Ab mean - isotype mean) and are shown as mean ± SD. Abscissa: antibody concentration (μg / ml). [Figure 21]Levels of cytokines (vertical axis, pg / ml) produced in PBMC cultures in the presence and absence of JEG3 cells over a range of concentrations of defucosylated HLA-G02 (horizontal axis, μg / ml) (Figure 21A: IFN gamma; Figure 21B: TNF alpha; Figure 21C: IL-2; Figure 21D: IL-6; Figure 21E: IL-8; Figure 21F: IL-10). Values ​​are shown as mean + / - standard deviation for all 16 PBMC donors. Dotted lines indicate the mean levels of each cytokine produced in the presence of 50 μg / ml of defucosylated isotype control antibody, 50 μg / ml of anti-CD3, or 100 ng / ml of LPS. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0017] The present disclosure will now be described with respect to certain non-limiting aspects and embodiments thereof, and with reference to certain figures and examples.

[0018] Technical terms are used according to their common sense unless otherwise indicated. When a specific meaning is conveyed to a particular term, the definition of the term is given in the context in which the term is used.

[0019] Where the term "comprising" is used in the present description and claims, it does not exclude other elements. For the purposes of this disclosure, the term "consisting of" is considered to be a preferred embodiment of the term "comprising of."

[0020] Where an indefinite or definite article is used when referring to a singular noun such as "a", "an" or "the", this includes a plural of that noun, unless otherwise stated.

[0021] The present disclosure provides an antibody against HLA-G. In a first aspect, the present invention provides an antibody that specifically binds to HLA-G, comprising: a. CDR-L1 comprising SEQ ID NO:1; CDR-L2 comprising SEQ ID NO:2, and CDR-L3 comprising SEQ ID NO:3 and a light chain variable region comprising b. CDR-H1 comprising SEQ ID NO:4; CDR-H2 comprising SEQ ID NO:5, and CDR-H3 comprising SEQ ID NO:6 and a heavy chain variable region comprising the

[0022] HLA-G The term "HLA-G" or "human HLA-G" refers to the classical HLA-I molecule, human leukocyte antigen G, also known as the human major histocompatibility complex I molecule (MHC). Typically, HLA-G forms an MHC-I complex with B2m.

[0023] Unless otherwise specified, the term "HLA-G" refers to any alternative splice variant or naturally occurring variant or alternative splice isoform or naturally occurring isoform of human HLA-G that is naturally expressed by a cell. Exemplary sequences of full human HLA-G include the sequence shown in SEQ ID NO: 107.

[0024] In some embodiments, the antibodies of the invention selectively bind to the extracellular domain (ECD) of HLA-G (or "HLA-G ECD"). Exemplary sequences of HLA-G ECD include those set forth in SEQ ID NO:108.

[0025] The amino acid and nucleic acid sequences of HLA-G and its isoforms are also well known in the art.

[0026] Seven isoforms of HLA-G have been identified, four membrane-bound (HLA-G1, HLA-G2, HLA-G3, HLA-G4) and three soluble (HLA-G5, HLA-G6, and HLA-G7), which are the result of alternative splicing of the HLA-G primary transcript (Figure 1). Additional isoforms, some of which lack the α1 domain, have been suggested by mRNA analysis of renal tumors, but these have not been confirmed at the protein level (Tronik-Le Roux et al., Molecular Oncology 11(2017)1561-1578).

[0027] HLA-G1 and HLA-G5 are composed of three globular domains (α 1 , α 2 and α 3 HLA-G1 and HLA-G5 have a structure similar to that of classical HLA-I molecules, i.e., heterodimeric molecules, containing heavy chains exhibiting the B2m α-receptor domain. HLA-G1 and HLA-G5 can also exist as free alpha chains, i.e., not complexed with B2m.

[0028] HLA-G2 and HLA-G6 contain only the α1 and α3 domains, HLA-G4 contains only the α1 and α2 domains, and HLA-G3 and HLA-G7 contain only the α1 domain.

[0029] In one embodiment, the antibodies of the invention bind to at least one of HLA-G1, HLA-G2, HLA-G3, HLA-G4, HLA-G5, HLA-G6 and HLA-G7. In one embodiment, the antibodies of the invention bind to all HLA-G isoforms containing the alpha3 domain, i.e., HLA-G1, HLA-G2, HLA-G5 and HLA-G6, including HLA-G1 and HLA-G5 when complexed with B2m and when expressed as B2m free molecules.

[0030] HLA-G1 and HLA-G5 can form homomultimers, such as disulfide-linked dimers and disulfide-linked trimers. In one embodiment, the antibody of the present invention binds to monomeric HLA-G. In one embodiment, the antibody of the present invention binds to dimeric HLA-G. In one embodiment, the antibody of the present invention binds to trimeric HLA-G. In one embodiment, the antibody of the present invention binds to monomeric HLA-G, dimeric HLA-G and trimeric HLA-G.

[0031] Antibodies that bind to HLA-G Antibodies for use in connection with this disclosure include whole antibodies and functionally active fragments thereof (i.e., molecules that contain an antigen-binding domain that specifically binds to an antigen (also referred to as antigen-binding fragments)). Features described herein with respect to antibodies also apply to antibody fragments, unless the context dictates otherwise. Antibodies may be (or may be derived from) monoclonal, multivalent, multispecific, bispecific, fully human, humanized or chimeric.

[0032] Whole antibodies, also known as "immunoglobulins (Ig)", generally refer to intact or full-length antibodies, i.e., antibodies that contain two heavy and two light chain elements interconnected by disulfide bonds that assemble to define a characteristic Y-shaped three-dimensional structure. Classical native whole antibodies are monospecific, in that they bind to one antigen type, and bivalent, in that they have two independent antigen-binding domains. The terms "intact antibody", "full-length antibody" and "whole antibody" are used interchangeably to refer to monospecific bivalent antibodies that have a structure similar to the native antibody structure, including the Fc region as defined herein.

[0033] In whole antibodies, each light chain is composed of a light chain variable region (abbreviated herein as VL) and a light chain constant region (CL). Each heavy chain is composed of a heavy chain variable region (abbreviated herein as VH) and a heavy chain constant region (CH) composed of three constant domains CH1, CH2 and CH3, or four constant domains CH1, CH2, CH3 and CH4, depending on the Ig class. The "class" of an Ig or antibody refers to the type of constant region and includes IgA, IgD, IgE, IgG and IgM, some of which may be further divided into subclasses, e.g., IgG1, IgG2, IgG3, IgG4. The constant region of an antibody may mediate the binding of the immunoglobulin to host tissues or factors, including various cells of the immune system (e.g., effector cells) and the first component (Clq) of the classical complement system.

[0034] The VH and VL regions of the antibodies according to the invention can be further subdivided into regions of hypervariability (or "hypervariable regions" or HVRs) that determine the recognition of antigens, called complementarity determining regions (CDRs), interspersed with more structurally conserved regions called framework regions (FRs). Each VH and VL is composed of three CDRs and four FRs arranged from the amino-terminus to the carboxy-terminus in the following order: FR1, CDR1, FR2, CDR2, FR3, CDR3, FR4. The CDRs and FRs together form the variable region. By convention, the CDRs in the heavy chain variable region of an antibody or antigen-binding fragment thereof are referred to as CDR-H1, CDR-H2 and CDR-H3, and the CDRs in the light chain variable region are referred to as CDR-L1, CDR-L2 and CDR-L3. They are numbered consecutively from the N-terminus to the C-terminus of each chain.

[0035] CDRs are conventionally numbered according to the system devised by Kabat et al., which is described in Kabat et al., 1991, in Sequences of Proteins of Immunological Interest, US Department of Health and Human Services, NIH, USA (hereinafter "Kabat et al., supra"). This numbering system is used herein unless otherwise indicated.

[0036] The Kabat residue designations do not necessarily correspond directly to the linear numbering of the amino acid residues. The actual linear amino acid sequence may contain fewer or additional amino acids than the strict Kabat numbering, which correspond to shortening of, or insertion into, structural elements of the basic variable domain structure, whether in framework regions or complementarity determining regions (CDRs). The correct Kabat numbering of residues can be determined for a given antibody by alignment of the homologous residues in the antibody sequence with the "standard" Kabat numbering sequence.

[0037] The CDRs of the heavy chain variable domain are located at residues 31-35 (CDR-H1), residues 50-65 (CDR-H2) and residues 95-102 (CDR-H3) according to the Kabat numbering system, except that according to Chothia (Chothia, C. and Lesk, AMJ Mol. Biol., 196, 901-917 (1987)), the loop equivalent to CDR-H1 extends from residue 26 to residue 32. Thus, unless otherwise indicated, "CDR-H1" as used herein is intended to refer to residues 26-35 as described by a combination of the Kabat numbering system and the topological loop definition of Chothia.

[0038] The CDRs of the light chain variable domain are located at residues 24-34 (CDR-L1), residues 50-56 (CDR-L2) and residues 89-97 (CDR-L3) according to the Kabat numbering system.

[0039] In addition to the CDR loops, a fourth loop exists between CDR-2 (CDR-L2 or CDR-H2) and CDR-3 (CDR-L3 or CDR-H3) formed by framework 3 (FR3). The Kabat numbering system defines framework 3 as positions 66-94 of the heavy chain and positions 57-88 of the light chain.

[0040] Numbering schemes have been proposed based on alignment of the sequences of various members of the immunoglobulin family and are described, for example, in Kabat et al., 1991, and Dondelinger et al., 2018, Frontiers in Immunology, Vol 9, article 2278.

[0041] The antibody of the present invention comprises a light chain variable region comprising CDR-L1 comprising SEQ ID NO:1, CDR-L2 comprising SEQ ID NO:2, and CDR-L3 comprising SEQ ID NO:3, and a heavy chain variable region comprising CDR-H1 comprising SEQ ID NO:4, CDR-H2 comprising SEQ ID NO:5, and CDR-H3 comprising SEQ ID NO:6.

[0042] In one embodiment, an antibody of the invention comprises a light chain variable region comprising the CDRs of the light chain variable region of SEQ ID NO:19 and a heavy chain variable region comprising the CDRs of the heavy chain variable region of SEQ ID NO:93.

[0043] Antibodies containing such CDR sequences are particularly inventive, since they provide antibodies with high affinity for HLA-G, high specificity for HLA-G (especially without cross-reactivity with other HLA-I molecules despite their very high homology), high inhibition of HLA-G biological functions, and high stability, essential for manufacturability, and, when combined with an active Fc, provide antibodies with the ability to directly kill HLA-G+ tumor cells.

[0044] As used herein, the terms "constant domain", "constant region" are used interchangeably to refer to the domain of an antibody outside the variable region. The constant domain is identical in all antibodies of the same isotype, but differs from one isotype to another. Typically, the constant region of the heavy chain is formed by CH1-hinge-CH2-CH3-optionally CH4 from the N-terminus to the C-terminus, which includes three or four constant domains.

[0045] The constant region domain of the antibody molecule of the present invention, if any, can be selected taking into account the proposed function of the antibody molecule, in particular the effector function that may be required. For example, the constant region domain can be a human IgA, IgD, IgE, IgG or IgM domain. In particular, when the antibody molecule is intended for therapeutic use and antibody effector function is required, human IgG constant region domains, in particular IgG1 and IgG3 isotypes, can be used. Alternatively, when the antibody molecule is intended for therapeutic use and antibody effector function is not required, IgG2 and IgG4 isotypes can be used. It will be understood that sequence variants of these constant region domains can also be used. For example, an IgG4 molecule in which the serine at position 241 (numbered according to the Kabat numbering system) has been changed to proline, as described by Angal et al. (Angal et al., 1993. A single amino acid substitution abolishes the heterogeneity of chimeric mouse / human (IgG4) antibody as observed during SDS-PAGE analysis Mol Immunol 30, 105-108) and referred to herein as IgG4P, may be used.

[0046] "Fc", "Fc fragment" and "Fc region" are used interchangeably to refer to the C-terminal region of an antibody that contains the constant region of the antibody, excluding the first constant region immunoglobulin domain. Thus, Fc refers to the C-terminal region of an antibody that contains the last two constant domains of IgA, IgD and IgG, the C-terminal region of the antibody. H2 and C H3, or the last three constant domains of IgE and IgM, and the flexible hinge N-terminal to these domains. The human IgG1 heavy chain Fc region is defined herein to include residue C226 to its carboxyl terminus, with numbering according to the EU index as in Kabat. In the context of human IgG1, the lower hinge refers to positions 226-236, the CH2 domain refers to positions 237-340, and the CH3 domain refers to positions 341-447, according to the EU index as in Kabat. Corresponding Fc regions of other immunoglobulins can be identified by sequence alignment.

[0047] In the context of the present disclosure, the constant region or Fc region, if present, may be native as defined above or may be modified in various ways, so long as it contains a functional FcR binding domain, preferably a functional FcRn binding domain. Preferably, the modified constant region or Fc region improves functionality and / or pharmacokinetics. The modification may include deletion of certain parts of the Fc fragment. The modification may further include various amino acid substitutions that can affect the biological properties of the antibody. There may also be mutations to increase FcRn binding and thus in vivo half-life. The modification may further include modification of the glycosylation profile of the antibody. The native Fc fragment is glycosylated in the CH2 domain, where an N-glycan attached to the asparagine residue at position 297 (Asn297) is present in each of the two heavy chains. In the context of the present disclosure, antibodies can be glyco-modified, i.e., engineered, to have a particular glycosylation profile that confers improved properties, e.g., improved effector function and / or improved serum half-life.

[0048] The antibodies described herein are isolated. An "isolated" antibody is one that has been separated (e.g., by purification means) from a component of its natural environment.

[0049] The term "antibody" encompasses monovalent, i.e., antibodies that contain only one antigen-binding domain (e.g., a one-arm antibody comprising a full-length heavy chain and an interconnected full-length light chain, also called a "half antibody"), and multivalent, i.e., antibodies that contain multiple antigen-binding domains.

[0050] The term "antibody" according to the present invention also encompasses antigen-binding fragments of antibodies.

[0051] Antigen-binding fragments of antibodies include single chain antibodies (e.g., scFv and dsscfv), Fab, Fab', F(ab') 2 , Fv, single domain antibodies or nanobodies (e.g. VH or VL, or VHH or VNAR). Other antibody fragments for use in the present invention include the Fab and Fab' fragments described in International Patent Applications WO 2011 / 117648, WO 2005 / 003169, WO 2005 / 003170 and WO 2005 / 003171.

[0052] Methods for generating and producing these antibody fragments are well known in the art (see, for example, Verma et al., 1998, Journal of Immunological Methods, 216, 165-181).

[0053] As used herein, the term "Fab fragment" refers to a light chain fragment containing the VL (variable light chain) domain and the constant domain of the light chain (CL), and an antibody fragment containing the VH (variable heavy chain) domain and the first constant domain of the heavy chain (CH1).

[0054] A typical "Fab' fragment" comprises a pair of heavy and light chains, the heavy chain comprising a variable domain VH, a constant domain CH1 and a native or modified hinge region, and the light chain comprising a variable domain VL and a constant domain CL. A Fab' dimer according to the present disclosure is represented as F(ab') 2 where, for example, dimerization may be via the hinge.

[0055] As used herein, the term "single domain antibody" refers to an antibody fragment consisting of a single monomeric variable antibody domain. Examples of single domain antibodies include VH or VL or VHH or VNAR.

[0056] "Fv" refers to two variable domains, eg, a cognate pair or affinity matured variable domains, ie, cooperating variable domains such as a VH and VL pair.

[0057] As used herein, a "single-chain variable fragment" or "scFv" refers to a single-chain variable fragment stabilized by a peptide linker between the VH and VL variable domains.

[0058] As used herein, a "disulfide-stabilized single-chain variable fragment" or "dsscFv" refers to a single-chain variable fragment stabilized by a peptide linker between the VH and VL variable domains and also containing an interdomain disulfide bond between VH and VL (see, e.g., Weatherill et al., Protein Engineering, Design & Selection, 25(321-329), 2012; WO 2007109254).

[0059] In one embodiment, the disulfide bond between the variable domains VH and VL is between two of the residues listed below (Kabat numbering is used in the following list unless the context indicates otherwise). When Kabat numbering is referred to, the relevant reference is Kabat et al., 1991 (5 th edition, Bethesda, Md.), in Sequences of Proteins of Immunological Interest, US Department of Health and Human Services, NIH, USA.

[0060] In one embodiment, the disulfide bond is at a position selected from the group including: VH37+VL95C, see for example Protein Science 6,781-788 Zhu et al (1997); VH44+VL100, see, for example, Weatherill et al., Protein Engineering, Design & Selection, 25(321-329), 2012; VH44+VL105, see, for example, J Biochem. 118, 825-831 Luo et al (1995); VH45+VL87, see for example Protein Science 6,781-788 Zhu et al (1997); · VH55+VL101, see e.g. FEBS Letters 377 135-139 Young et al (1995); VH100+VL50, see for example Biochemistry 29 1362-1367 Glockshuber et al (1990); VH100b+VL49; see for example Biochemistry 29 1362-1367 Glockshuber et al (1990); VH98+VL46, see for example Protein Science 6,781-788 Zhu et al (1997); VH101+VL46; see for example Protein Science 6, 781-788 Zhu et al (1997); VH105+VL43, see, for example, Proc. Natl. Acad. Sci. USA Vol. 90 pp. 7538-7542 Brinkmann et al (1993); or Proteins 19, 35-47 Jung et al (1994); VH106+VL57, see e.g. FEBS Letters 377 135-139 Young et al (1995) and the corresponding position or positions in the variable region pair located within the molecule.

[0061] In one embodiment, the disulfide bond is formed between positions VH44 and VL100.

[0062] multispecific antibodies The antibody of the present invention can be a multispecific antibody. As used herein, "multispecific antibody or multi-specific antibody" refers to an antibody described herein that has at least two binding domains, i.e., two or more binding domains, for example, two or three binding domains, and at least two binding domains independently bind to two different antigens, or two different epitopes on the same antigen. A multispecific antibody is generally monovalent for each specificity (antigen). The multispecific antibody described herein encompasses monovalent and multivalent, for example, bivalent, trivalent, tetravalent, multispecific antibodies.

[0063] In one embodiment, the construct is a bispecific antibody. As used herein, "bispecific antibody or bi-specific antibody" refers to an antibody with two antigen-binding specificities. In one embodiment, the antibody comprises two antigen-binding domains, one binding domain binds to ANTIGEN 1 and the other binding domain binds to ANTIGEN 2, i.e., each binding domain is monovalent for each antigen. In one embodiment, the antibody is a tetravalent bispecific antibody, i.e., the antibody comprises four antigen-binding domains, for example, two binding domains bind to ANTIGEN 1 and the other two binding domains bind to ANTIGEN 2. In one embodiment, the antibody is a trivalent bispecific antibody.

[0064] In one embodiment, the antibody construct is a trispecific antibody. As used herein, "trispecific antibody or tri-specific antibody" refers to an antibody that has three antigen-binding specificities. For example, the antibody is an antibody that has three antigen-binding domains (trivalent) that independently bind to three different antigens or three different epitopes on the same antigen, i.e., each binding domain is monovalent for each antigen.

[0065] A paratope is a region of an antibody that recognizes and binds to an antigen. The antibody of the present invention can be a multiparatopic antibody. As used herein, "multiparatopic antibody" refers to an antibody described herein that contains two or more different paratopes that interact with different epitopes from either the same antigen or two different antigens. The multiparatopic antibody described herein can be biparatopic, triparatopic, or tetraparatopic.

[0066] As used herein, "antigen binding domain" refers to a portion of an antibody that comprises part or all of one or more variable domains that specifically interact with a target antigen, such as part or all of a pair of variable domains VH and VL. A binding domain may comprise a single domain antibody. In one embodiment, each binding domain is monovalent. Preferably, each binding domain comprises no more than one VH and one VL.

[0067] A variety of multispecific antibody formats have been produced. Although various classifications have been proposed, multispecific IgG antibody formats generally include bispecific IgG, adduct IgG, multispecific (e.g., bispecific) antibody fragments, multispecific (e.g., bispecific) fusion proteins and multispecific (e.g., bispecific) antibody conjugates, as described, for example, in Spiess et al., Alternative molecular formats and therapeutic applications for bispecific antibodies. Mol Immunol. 67 (2015): 95-106.

[0068] Techniques for making bispecific antibodies include, but are not limited to, CrossMab technology (Klein et al. Engineering therapeutic bispecific antibodies using CrossMab technology, Methods 154 (2019) 21-31), Knobs-in-holes engineering (e.g., WO 1996027011, WO 1998050431), DuoBody technology (e.g., WO 2011131746), Azymetric technology (e.g., WO 2012058768). Further techniques for making bispecific antibodies are described, for example, in Godar et al., 2018, Therapeutic bispecific antibody formats: a patent applications review (1994-2017), Expert Opinion on Therapeutic Patents, 28: 3, 251-276. Bispecific antibodies include, among others, CrossMab antibodies, DAF (two-in-one), DAF (four-in-one), DutaMab, DT-lgG, knob-in-hole common LC, knob-in-hole assembly, charge pair, Fab arm exchange, SEEDbody, Triomab, LUZ-Y, Fcab, κλ-body and orthogonal Fab.

[0069] A loaded IgG classically comprises a full-length IgG engineered by adding additional antigen-binding domains or antigen-binding fragments to the N-terminus and / or C-terminus of the IgG heavy and / or light chains. Examples of such additional antigen-binding fragments include sdAb antibodies (e.g., VH or VL), Fv, scFv, dsscFv, Fab, scFav. Additional IgG antibody formats include, among others, DVD-IgG, lgG(H)-scFv, scFv-(H)lgG, lgG(L)-scFv, scFv-(L)lgG, lgG(L,H)-Fv, lgG(H)-V, V(H)-lgG, lgC(L)-V, V(L)-lgG, KIH IgG-scFab, 2scFv-lgG, lgG-2scFv, scFv4-lg, Zybody, and DVI-IgG (four-in-one), as described, for example, in Spiess et al., Alternative molecular formats and therapeutic applications for bispecific antibodies. Mol Immunol. 67(2015):95-106.

[0070] Multispecific antibody fragments include nanobodies, nanobody-HSA, BiTE, diabody, DART, TandAb, scDiabody, sc-Diabody-CH3, Diabody-CH3, Triple Body, Miniantibody; Minibody, Tri Bi minibody, scFv-CH3 KIH, Fab-scFv, scFv-CH-CL-scFv, F(ab')2, F(ab')2-scFv2, scFv-KIH, Fab-scFv-Fc, tetravalent HCAb, scDiabody-Fc, Diabody-Fc, Tandem scFv-Fc; and intrabody, as described, for example, in Spiess et al., Alternative molecular formats and therapeutic applications for bispecific antibodies. Mol Immunol. 67 (2015): 95-106.

[0071] Multispecific fusion proteins include Dock and Lock, ImmTAC, HSAbody, scDiabody-HAS, and Tandem scFv-Toxin.

[0072] Multispecific antibody conjugates include IgG-lgG; Cov-X-Body; and scFv1-PEG-scFv2.

[0073] Additional multispecific antibody formats are described, for example, in Brinkmann and Kontermann, The making of bispecific antibodies, mAbs, 9:2, 182-212 (2017), in particular FIG. 2, e.g., tandem scFv, triplex, Fab-VHH, taFv-Fc, scFv 4 -Ig, scFv 2 -Fcab, scFv 4 -IgG. Bibodies, tribodies and methods for their production are disclosed, for example, in WO 99 / 37791.

[0074] Examples of antibodies for use in the present invention include added IgGs and added Fabs where complete IgG or Fab fragments have been engineered by adding at least one additional antigen binding domain (e.g. two, three or four additional antigen binding domains), such as single domain antibodies (such as VH or VL, or VHH), scFv, dsscFv, dsFv to the N-terminus and / or C-terminus of the heavy and / or light chain of said IgG or Fab, respectively, as described, for example, in WO 2009 / 040562, WO 2010 / 035012, WO 2011 / 030107, WO 2011 / 061492, WO 2011 / 061246 and WO 2011 / 086091. In particular, the Fab-Fv format is described in WO 2009 / 040562 and its disulfide-stabilized version, the Fab-dsFv, is described in WO 2010 / 035012. Single linker Fab-dsFv, in which the dsFv is connected to the Fab via a single linker between the VL or VH domain of the Fv and the C-terminus of the LC or HC of the Fab, is described in WO 2014 / 096390. Added IgGs, including full-length IgG1s engineered by adding dsFv to the C-terminus of the IgG heavy or light chain, are described in WO 2015 / 197789.

[0075] Another exemplary antibody for use in the present invention comprises a Fab linked to two scFvs or dsscFvs, each scFv or dsscFv binding to the same or different targets (e.g., one scFv or dsscFv that binds to a therapeutic target and one scFv or dsscFv that increases half-life, e.g., by binding to albumin). Such antibodies are described in WO 2015 / 197772. Another exemplary antibody for use in the fragments of the present invention comprises a Fab linked to only one scFv or dsscFv, e.g., as described in WO 2013 / 068571 and Dave et al., Mabs, 8(7)1319-1335 (2016).

[0076] Other known formats of multispecific antibodies include:

[0077] As used herein, a diabody refers to two Fv pairs, a first VH / VL pair and a further VH / VL pair, with two inter-Fv linkers such that the VH of the first Fv is linked to the VL of the second Fv and the VL of the first Fv is linked to the VH of the second Fv.

[0078] Triabodies as used herein refer to a format similar to diabodies, which contain three Fvs and three inter-Fv linkers.

[0079] Tetrabodies as used herein refer to a format similar to diabodies, which contain four Fvs and four inter-Fv linkers.

[0080] Tandem scFv, as used herein, refers to at least two scFvs linked via a single linker such that there is a single inter-Fv linker.

[0081] Tandem scFv-Fc, as used herein, refers to at least two tandem scFvs, each of which is attached, for example via a hinge, to the N-terminus of the CH2 domain of the constant region fragment -CH2CH3.

[0082] As used herein, Fab-Fv refers to an Fv fragment in which the variable region is added to the C-terminus of each of the CH1 of the heavy chain and the CL of the light chain. The format may be provided as a PEGylated version thereof.

[0083] As used herein, Fab'-Fv is similar to FabFv, where the Fab portion is replaced by Fab'. The format may be provided as a PEGylated version thereof.

[0084] As used herein, Fab-dsFv refers to a FabFv in which an intra-Fv disulfide bond stabilizes the additional C-terminal variable region. The format may be provided as a PEGylated version thereof.

[0085] As used herein, Fab-scFv refers to a Fab molecule in which scFv is added to the C-terminus of the light chain or heavy chain.

[0086] As used herein, Fab'-scFv refers to a Fab' molecule in which scFv is added to the C-terminus of the light or heavy chain.

[0087] As used herein, DiFab refers to two Fab molecules linked via the C-terminus of the heavy chains.

[0088] As used herein, DiFab' refers to two Fab' molecules linked via one or more disulfide bonds in their hinge regions.

[0089] As used herein, an sc diabody is a diabody that contains an intra-Fv linker such that the molecule contains three linkers to form a normal scFv in which the VH and VL termini are each linked to one of the variable regions of an additional Fv pair.

[0090] As used herein, an Sc diabody-Fc is two sc diabodies, each attached, for example via a hinge, to the N-terminus of the CH2 domain of the constant region fragment -CH2CH3.

[0091] As used herein, ScFv-Fc-scFv refers to four scFvs, one of each, appended to the N- and C-termini of both the heavy and light chains of a -CH2CH3 fragment.

[0092] As used herein, Sc diabody-CH3, for example, refers to two sc diabody molecules each linked via a hinge to a CH3 domain.

[0093] As used herein, an IgG-scFv is a full-length antibody having an scFv at the C-terminus of each heavy or light chain.

[0094] As used herein, scFv-IgG is a full-length antibody having an scFv at the N-terminus of each heavy or light chain.

[0095] As used herein, a V-IgG is a full-length antibody that has a variable domain at the N-terminus of each heavy or light chain.

[0096] As used herein, an IgG-V is a full-length antibody that has a variable domain at the C-terminus of each heavy or light chain.

[0097] DVD-Ig (also known as double-V domain IgG) is a full-length antibody that has four additional variable domains, one at the N-terminus of each heavy chain and each light chain.

[0098] The present disclosure provides a multispecific antibody comprising one binding domain that specifically binds to HLA-G, said binding domain comprising: a. CDR-L1 comprising SEQ ID NO:1; CDR-L2 comprising SEQ ID NO:2, and CDR-L3 comprising SEQ ID NO:3 and a light chain variable region comprising b. CDR-H1 comprising SEQ ID NO:4; CDR-H2 comprising SEQ ID NO:5, and CDR-H3 comprising SEQ ID NO:6 and a heavy chain variable region comprising:

[0099] The antibody of the present invention specifically (or selectively) binds to HLA-G. An antibody "specifically binds" to a protein if it preferentially or with high affinity binds to the protein of interest (e.g., HLA-G) but does not substantially bind to other proteins. In other words, the antibody binds to the protein of interest without significant cross-reactivity to any other molecule. The specificity of the antibody can be further tested by determining whether the antibody binds to or distinguishes between other related proteins as described above.

[0100] In particular, an antibody that "specifically binds" to HLA-G is not cross-reactive with other human proteins, particularly other HLA-I molecules. In one embodiment, the antibody of the present invention does not substantially bind to any of HLA-A, HLA-B, HLA-C, HLA-E and HLA-F. In one embodiment, the antibody of the present invention does not bind to any of HLA-A, HLA-B, HLA-C, HLA-E and HLA-F. Exemplary sequences of HLA-A, HLA-B, HLA-C, HLA-E and HLA-F are SEQ ID NOs: 132, 134, 136, 138 and 140, respectively. In one embodiment, the antibody of the present invention does not bind to B2m.

[0101] The antibody of the present invention can specifically bind to the alpha 3 domain of HLA-G. By "specifically bind to the alpha 3 domain of HLA-G", it is understood that the antibody binds to the alpha 3 domain of HLA-G without cross-reactivity with other human proteins and without cross-reactivity with other domains of HLA-G (e.g., the alpha 1 domain or the alpha 2 domain).

[0102] Cross-reactivity can be evaluated, for example, by any suitable method described herein.Antibody cross-reactivity can be considered significant if the antibody binds to another molecule at least about 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90% or 100% as strongly as it binds to the protein of interest.A specific (or selective) antibody can bind to another molecule less than about 90%, 85%, 80%, 75%, 70%, 65%, 60%, 55%, 50%, 45%, 40%, 35%, 30%, 25% or 20% as strongly as it binds to the protein of interest. An antibody may bind to other molecules with less than about 20%, less than about 15%, less than about 10%, or less than about 5%, less than about 2%, or less than about 1% of the strength with which it binds to a protein of interest.

[0103] In one embodiment, according to the invention, the binding of an antibody to HLA-G has a dissociation constant (K D ) is its characteristic.

[0104] As used herein, the term "K D " is K d Against K a (i.e., K d / K a ) and expressed as a molar concentration (M). d and K. a and K refer to the dissociation and association rates, respectively, of a particular antigen-antibody (or antigen-binding fragment thereof) interaction. D Values ​​can be determined using methods well established in the art. D A method for determining K is by using recombinant HLA-G, or a suitable fusion protein / polypeptide thereof, using surface plasmon resonance (SPR), such as, for example, the Biacore® system described in the examples herein. Typically, K DThe value is determined by SPR at a temperature of 25°C. In one example, affinity is measured using recombinant HLA-G extracellular domain (ECD) expressed in complex with B2m, as described in the examples herein. For surface plasmon resonance, the target molecule is immobilized on a solid phase and exposed to the ligand in a mobile phase that moves along a flow cell. Binding of the ligand to the immobilized target changes the local refractive index, resulting in a change in the SPR angle, which can be monitored in real time by detecting changes in the intensity of the reflected light. The rate of change of the SPR signal can be analyzed to obtain the apparent rate constants for the association and dissociation phases of the binding reaction. The ratio of these values ​​gives the apparent equilibrium constant (affinity) (see, e.g., Wolff et al, Cancer Res. 53:2560-65 (1993)).

[0105] The term "affinity" refers to the strength of the interaction between an antibody and HLA-G.

[0106] Binding affinity to HLA-G may be measured for HLA-G or HLA-G ECD associated or not with B2m. For example, binding to HLA-G may be assessed by measuring binding affinity to a soluble HLA-G ECD comprising the sequence of SEQ ID NO: 108 or 110. In one example, binding to HLA-G is assessed by measuring binding affinity to a soluble HLA-G ECD comprising the sequence of SEQ ID NO: 108 or 110, for example, associated with B2m. In one embodiment, the antibody of the invention has a dissociation constant (K D ) to the ECD of HLA-G. In one embodiment, the dissociation constant is determined by SPR between an antibody of the invention expressed as a full-length antibody and a monomeric form of HLA-G (e.g., HLA-G ECD associated with B2m) at a temperature of 25° C. In one embodiment, the dissociation constant is determined by SPR as described in Example 7.1.

[0107] In another example, binding to HLA-G can be assessed by measuring binding affinity to cell membrane expressed HLA-G comprising the sequence of SEQ ID NO: 107. Binding to cell membrane expressed HLA-G can be analyzed by FACS. Typically, HLA-G expressed on the surface of cells is expressed in a dimeric form. In one embodiment, the antibody of the present invention has a dissociation constant (K D ) on cells. In one embodiment, the antibodies of the invention bind to HLA-G with a dissociation constant (K) of less than 1 nM as determined by FACS. D ) to bind to JEG3 cells.

[0108] In one embodiment, the invention relates to an antibody that specifically binds to HLA-G, comprising a light chain variable region and a heavy chain variable region, wherein the light chain variable region comprises CDR-L1 comprising SEQ ID NO: 1, CDR-L2 comprising SEQ ID NO: 2, and CDR-L3 comprising SEQ ID NO: 3; and the heavy chain variable region comprises CDR-H1 comprising SEQ ID NO: 4, CDR-H2 comprising SEQ ID NO: 5, and CDR-H3 comprising SEQ ID NO: 6, and wherein the antibody has a dissociation constant (K D In one embodiment, the dissociation constant is determined by SPR at a temperature of 25° C. between an antibody of the invention expressed as a full-length antibody and a monomeric form of HLA-G.

[0109] In one embodiment, the invention provides an antibody that specifically binds to HLA-G, comprising a light chain variable region and a heavy chain variable region, wherein the light chain variable region comprises CDR-L1 comprising SEQ ID NO:1, CDR-L2 comprising SEQ ID NO:2, and CDR-L3 comprising SEQ ID NO:3, and the heavy chain variable region comprises CDR-H1 comprising SEQ ID NO:4, CDR-H2 comprising SEQ ID NO:5, and CDR-H3 comprising SEQ ID NO:6, and wherein the antibody expressed as a full-length antibody has a dissociation constant (K D ) to provide an antibody that binds to JEG3 cells.

[0110] In one embodiment, the antibody of the invention is a blocking antibody. The term "blocking" (or "blocks") in relation to an antibody refers to an antibody that is capable of inhibiting or attenuating the binding of its target (HLA-G) to its receptor. In one embodiment, the antibody of the invention blocks the interaction between HLA-G and ILT2. In one embodiment, the antibody of the invention blocks the interaction between HLA-G and ILT4. In one embodiment, the antibody of the invention blocks the interaction between HLA-G and ILT2 and between HLA-G and ILT4. In one embodiment, the antibody of the invention blocks the interaction between HLA-G and ILT2 and / or between HLA-G and ILT4 when HLA-G is expressed as a monomer and / or dimer and / or trimer.

[0111] Blocking of HLA-G binding to ILT2 and / or ILT4 can be assessed by measuring the blocking of the interaction between the extracellular domain (ECD) of HLA-G, with or without B2m association, expressed on the surface of cells, and ILT2 and / or ILT4 expressed as a fusion protein, such as an Fc fusion protein (ITT2-Fc.ILT4-Fc). For example, an ILT2-rabbit Fc fusion protein comprising SEQ ID NO: 142 can be used. For example, an ILT4-rabbit Fc fusion protein comprising SEQ ID NO: 144 can be used. Blocking of HLA-G binding to ILT2 and / or ILT4 can be assessed as described in Example 8.

[0112] In some embodiments, the antibodies of the invention do not block the association of HLA-G with B2m, hi some embodiments, the antibodies of the invention do not block the association of HLA-G with its cognate peptide that is naturally expressed in a complex with HLA-G.

[0113] In some embodiments, the antibodies of the invention inhibit the multimerization of HLA-G. In some embodiments, the antibodies of the invention inhibit the dimerization of HLA-G. In some embodiments, the antibodies of the invention inhibit the trimerization of HLA-G.

[0114] In one embodiment, the antibody according to the invention has an IC of less than 50 pM for blocking binding of ILT2 to HLA-G. 50 and preferably the antibody according to the invention has an IC of less than 40 pM, or less than 30 pM, or less than 20 pM for blocking the binding of ILT2 to HLA-G naturally expressed on the surface of JEG3 cells, as determined using an in vitro assay using a high reaction volume, e.g., as described in Example 8. 50 In a preferred embodiment, the antibody according to the invention has an IC of less than 20 pM for blocking the binding of ILT2 to HLA-G naturally expressed on the surface of JEG3 cells. 50 In one embodiment ILT2 is expressed as an ILT2-rabbit Fc fusion protein, for example comprising SEQ ID NO: 142. In one embodiment the antibody according to the invention has an IC of less than 1800 pM for blocking binding of ILT4 to HLA-G. 50 and preferably the antibody according to the invention has an IC of less than 1500 pM or less than 1400 pM for blocking binding of ILT4 to HLA-G in an in vitro assay as described herein. 50 In one embodiment, ILT4 is expressed as an ILT4-rabbit Fc fusion protein, for example comprising SEQ ID NO: 144. Blockade of HLA-G binding to ILT4 can be assessed as described in Example 8.

[0115] In one embodiment, the invention provides an antibody that specifically binds to HLA-G, comprising a light chain variable region and a heavy chain variable region, wherein the light chain variable region comprises CDR-L1 comprising SEQ ID NO:1, CDR-L2 comprising SEQ ID NO:2, and CDR-L3 comprising SEQ ID NO:3, the heavy chain variable region comprises CDR-H1 comprising SEQ ID NO:4, CDR-H2 comprising SEQ ID NO:5, and CDR-H3 comprising SEQ ID NO:6, and the antibody is a. An IC of less than 20 pM to block binding of ILT2 to HLA-G naturally expressed on the surface of JEG3 cells, as determined using an in vitro assay using a high reaction volume, e.g., as described in Example 8. 50 and / or b. An IC of less than 1400 pM for blocking ILT4 binding to HLA-G, as determined, for example, as described in Example 8. 50 The present invention provides an antibody having the formula:

[0116] As used herein, IC 50 The term IC refers to the half-maximal inhibitory concentration, which is a measure of the effectiveness of a substance, such as an antibody, in inhibiting a particular biological or biochemical function, which in this invention is the binding activity of ILT2 or ILT4 to HLA-G. 50 is a quantitative measure of how much of a particular substance is needed to inhibit half of a given biological process or function or activity.

[0117] In some embodiments, the antibodies of the invention inhibit HLA-G mediated immunosuppressive function. In some embodiments, the antibodies of the invention inhibit HLA-G mediated immunosuppressive function by blocking the interaction between HLA-G and ILT2 and / or the interaction between HLA-G and ILT4. In some embodiments, the antibodies of the invention inhibit HLA-G mediated suppressive function of NK cells. In some embodiments, the antibodies of the invention inhibit HLA-G mediated suppressive function of cytotoxic T lymphocytes, such as CD8+ T lymphocytes and / or CD4+ T lymphocytes. In some embodiments, the antibodies of the invention inhibit HLA-G mediated suppressive function of regulatory T lymphocytes. In some embodiments, the antibodies of the invention inhibit HLA-G mediated suppressive function of B cells. In some embodiments, the antibodies of the invention inhibit HLA-G mediated suppressive function of monocytes. In some embodiments, the antibodies of the invention inhibit HLA-G mediated suppressive function of macrophages. In some embodiments, the antibodies of the invention inhibit HLA-G mediated suppressive function of dendritic cells. In some embodiments, the antibodies of the invention inhibit HLA-G mediated suppression of neutrophils. In some embodiments, the antibodies of the invention inhibit HLA-G mediated inhibition of phagocytosis.

[0118] In some embodiments, the antibodies of the invention induce the production of inflammatory cytokines, e.g., TNF alpha, IL-1, IL-1 beta, IL-2, IL-4, IL-6, IL-8, IL-10, IL-12 (also known as IL-12p70), IL-13, IL-15, IL-18, IFN gamma, GM-CSF, CCL2, CCL3, CCL4, CCL5, TNF alpha. In some embodiments, the antibodies of the invention promote the recruitment of immune cells (monocytes, macrophages, dendritic cells, B cells, T cells, or NK cells) to the tumor microenvironment. In some embodiments, the antibodies of the invention inhibit HLA-G function against tumor cells expressing HLA-G. In some embodiments, the antibodies of the invention induce activation of myeloid cells. In some embodiments, the antibodies of the invention induce tumor cell killing, e.g., by ADCC or CDC. In some embodiments, the antibodies of the invention induce tumor cell phagocytosis, e.g., ADCP. In some embodiments, the antibodies of the invention inhibit angiogenesis. In some embodiments, the antibodies of the invention inhibit tumor cell metastasis, hi some embodiments, the antibodies of the invention inhibit tumor cell proliferation.

[0119] In one embodiment, the present invention provides an antibody that specifically binds to HLA-G, comprising: a. CDR-L1 comprising SEQ ID NO:1; CDR-L2 comprising SEQ ID NO:2, and CDR-L3 comprising SEQ ID NO:3 and a light chain variable region comprising b. CDR-H1 comprising SEQ ID NO:4; CDR-H2 comprising SEQ ID NO:5, and CDR-H3 comprising SEQ ID NO:6 and a heavy chain variable region comprising The antibody is provided, which blocks HLA-G binding to ILT2 and / or ILT4, preferably ILT2 and ILT4, and which has a dissociation constant (KD) for HLA-G of less than 10 nM. In one embodiment, the KD value is determined by SPR at a temperature of 25° C. using a full-length antibody of the invention. In one embodiment, the dissociation constant is determined for the monomeric form of HLA-G.

[0120] In one embodiment, the present invention provides an antibody that specifically binds to HLA-G, comprising: a. CDR-L1 comprising SEQ ID NO:1; CDR-L2 comprising SEQ ID NO:2, and CDR-L3 comprising SEQ ID NO:3 and a light chain variable region comprising b. CDR-H1 comprising SEQ ID NO:4; CDR-H2 comprising SEQ ID NO:5, and CDR-H3 comprising SEQ ID NO:6 and a heavy chain variable region comprising:

[0121] Antibodies for use in the present invention may be chimeric, humanized or fully human.

[0122] In one embodiment, the antibody is chimeric. The term "chimeric" antibody refers to an antibody in which the variable domains of the heavy and / or light chains (or at least a portion thereof) are derived from a particular source or species, e.g., mouse, rat, rabbit, etc., and the remaining portions of the heavy and / or light chains (i.e., the constant regions) are derived from another species, such as human. (Morrison; PNAS 81,6851 (1984)). Chimeric antibodies are composed of elements from two different species, such that the elements retain the properties of the species from which they are derived. A subcategory of "chimeric antibodies" is "humanized antibodies."

[0123] Chimeric antibodies are typically produced using recombinant DNA techniques. The DNA can be modified by substituting coding sequences for human L and H chain constant regions in place of the corresponding non-human (e.g., mouse or rabbit) H and L constant regions.

[0124] Humanized antibodies (including CDR-grafted antibodies) are antibody molecules having one or more complementarity determining regions (CDRs) from a non-human species and a framework region from a human immunoglobulin molecule (see, e.g., U.S. Pat. No. 5,585,089; WO 91 / 09967). It is understood that only the specificity determining residues of the CDRs may be transferred, rather than the entire CDR (see, e.g., Kashmiri et al., 2005, Methods, 36, 25-34). Humanized antibodies may optionally further comprise one or more framework residues from the non-human species from which the CDRs are derived.

[0125] A fully human antibody is an antibody in which the variable and constant regions (if any) of both heavy and light chains are all of human origin or are substantially identical to sequences of human origin, but are not necessarily derived from the same antibody.Examples of fully human antibodies can include, for example, the antibody produced by the above-mentioned phage display method, as well as the antibody produced by mouse, in which mouse immunoglobulin variable genes and optionally constant region genes are replaced by their human counterparts, as described in general terms in, for example, EP 0546073, U.S. Patent No. 5,545,806, U.S. Patent No. 5,569,825, U.S. Patent No. 5,625,126, U.S. Patent No. 5,633,425, U.S. Patent No. 5,661,016, U.S. Patent No. 5,770,429, EP 0438474 and EP 0463151.

[0126] In one embodiment, the antibody is human. A human antibody comprises a heavy or light chain variable region or a full-length heavy or light chain that is "the product of" or "derived from" a particular germline sequence when the variable region or full-length chain of the antibody is obtained from a system that uses human germline immunoglobulin genes. Such systems include immunizing a transgenic mouse carrying human immunoglobulin genes with the antigen of interest, or screening a human immunoglobulin gene library displayed on a phage with the antigen of interest. A human antibody or fragment thereof that is "the product of" or "derived from" a human germline immunoglobulin sequence can be identified as such by comparing the amino acid sequence of the human antibody with that of human germline immunoglobulins and selecting the human germline immunoglobulin sequence that is closest in sequence to the sequence of the human antibody (i.e., the highest % identity). A human antibody that is "the product of" or "derived from" a particular human germline immunoglobulin sequence may contain amino acid differences compared to the germline sequence, for example, due to naturally occurring somatic mutations or deliberate introduction of site-specific mutations. However, the selected human antibody will typically be at least 90% identical in amino acid sequence to the amino acid sequence encoded by a human germline immunoglobulin gene when compared to the germline immunoglobulin amino acid sequence of another species (e.g., a mouse germline sequence) and will contain amino acid residues that identify the human antibody as human. In certain cases, a human antibody may be at least 60%, 70%, 80%, 90%, or at least 95%, or even at least 96%, 97%, 98%, or 99% identical in amino acid sequence to the amino acid sequence encoded by the germline immunoglobulin gene. Typically, a human antibody derived from a particular human germline sequence will exhibit no more than 10 amino acid differences from the amino acid sequence encoded by the human germline immunoglobulin gene. In certain cases, a human antibody may display no more than 5, or even no more than 4, 3, 2, or 1 amino acid difference from the amino acid sequence encoded by the germline immunoglobulin gene.

[0127] Human antibodies can be produced by many methods known to those skilled in the art. Human antibodies can be produced by hybridoma methods using human myeloma cell lines or mouse-human heteromyeloma cell lines (Kozbor, J Immunol; (1984) 133:3001; Brodeur, Monoclonal Isolated Antibody Production Techniques and Applications, pp51-63, Marcel Dekker Inc, 1987). Alternative methods include the use of phage libraries or transgenic mice, both of which utilize human variable region repertoires (Winter G; (1994) Annu Rev Immunol 12:433-455, Green LL, (1999) J Immunol Methods 231:1 1-23).

[0128] The antibody according to the present invention can be obtained using any suitable method known in the art. HLA-G, including its fusion protein, cells expressing HLA-G (recombinantly or naturally) can be used to produce antibodies that specifically recognize HLA-G. Various forms of HLA-G as described herein may be used.

[0129] In another embodiment, the antigen used is HLA-G, preferably expressed on the surface of rabbit fibroblasts, preferably produced as described in the examples below. In one embodiment, the antigen used is HLA-G complexed with B2m, preferably expressed on the surface of rabbit fibroblasts, preferably produced as described in the examples below.

[0130] HLA-G or a fragment thereof for use in immunizing a host can be prepared by processes well known in the art from genetically engineered host cells containing an expression system. HLA-G or a fragment thereof can in some instances be part of a larger protein, such as a fusion protein, e.g., fused to an affinity tag or the like.

[0131] Antibodies generated against HLA-G can be obtained by administering HLA-G or a part thereof to an animal, preferably a non-human animal, using well-known routine protocols, if immunization of the animal is required. See, for example, Handbook of Experimental Immunology, DM Weir (ed.), Vol 4, Blackwell Scientific Publishers, Oxford, England, 1986). Many animals, such as rabbits, mice, rats, sheep, cattle, camels or pigs, may be immunized. However, mice, rabbits, pigs and rats are commonly used. In one embodiment, the antibodies of the present invention are obtained by administering rat fibroblasts expressing HLA-G on their surface.

[0132] Monoclonal antibodies can be produced by a variety of techniques, including, but not limited to, hybridoma methods, recombinant DNA methods, phage display methods, and methods utilizing transgenic animals containing all or part of the human immunoglobulin loci. Some exemplary methods for producing monoclonal antibodies are described herein.

[0133] For example, monoclonal antibodies can be prepared using hybridoma technology (Kohler & Milstein, 1975, Nature, 256:495-497), trioma technology, human B cell hybridoma technology (Kozbor et al., 1983, Immunology Today, 4:72) and EBV-hybridoma technology (Cole et al., Monoclonal Antibodies and Cancer Therapy, pp77-96, Alan R Liss, Inc., 1985).

[0134] Antibodies for use in the present invention can also be produced using single lymphocyte antibody techniques, for example, by cloning and expressing immunoglobulin variable region cDNA generated from a single lymphocyte selected for production of a particular antibody by the methods described in Babcook, J. et al., 1996, Proc. Natl. Acad. Sci. USA 93(15):7843-7848l; WO 92 / 02551; WO 2004 / 051268 and International Patent Application No. WO 2004 / 106377.

[0135] Monoclonal antibodies may also be generated using various phage display methods known in the art, including those disclosed by Brinkman et al. (in J. Immunol. Methods, 1995, 182:41-50), Ames et al. (J. Immunol. Methods, 1995, 184:177-186), Kettleborough et al. (Eur. J. Immunol. 1994, 24:952-958), Persic et al. (Gene, 1997 187 9-18), Burton et al. (Advances in Immunology, 1994, 57:191-280). In a particular phage display method, repertoires of VH and VL genes can be cloned separately by polymerase chain reaction (PCR), randomly recombined in a phage library, and then screened for antigen-binding phages, as described in Winter et al., Ann. Rev. Immunol, 12:433-455 (1994). Phages typically display antibody fragments as either single-chain Fv (scFv) fragments or Fab fragments. Libraries from immunized sources provide high affinity antibodies to immunogens without the need to construct hybridomas. Alternatively, naive repertoires can be cloned (e.g., from humans) without immunization to provide a single source of antibodies against a wide range of non-self and self antigens, as described in Griffiths et al., EMBO J 12:725-734 (1993). Finally, naive libraries can also be generated synthetically by cloning unrearranged V gene segments from stem cells and using PCR primers containing random sequences to encode the hypervariable CDR3 regions and achieve rearrangement in vitro, as described by Hoogenboom and Winter, J. Mol. Biol, 227:381-388 (1992).Patent publications describing human antibody phage libraries include, for example, U.S. Pat. No. 5,750,373, as well as U.S. Pat. No. 2005 / 0079574, U.S. Pat. No. 2005 / 0119455, U.S. Pat. No. 2005 / 0266000, U.S. Pat. No. 2007 / 0117126, U.S. Pat. No. 2007 / 0160598, U.S. Pat. No. 2007 / 0237764, U.S. Pat. No. 2007 / 0292936 and U.S. Pat. No. 2009 / 0002360.

[0136] Antibody screening can be performed using an assay that measures binding to HLA-G and / or measures the ability to block the binding of HLA-G to one or more of its receptors.An example of a binding assay is, for example, an ELISA that uses a fusion protein of a target polypeptide immobilized on a plate and uses a conjugated secondary antibody to detect the antibody that binds to the target.An example of a blocking assay is a flow cytometry-based assay that measures the blocking of ligand protein that binds to a target polypeptide.A fluorescently labeled secondary antibody is used to detect the amount of such ligand protein that binds to a target polypeptide.

[0137] Antibodies can be isolated by screening combinatorial libraries for antibodies with the desired activity or activities. For example, various methods are known in the art for screening such libraries for antibodies with the desired binding characteristics.

[0138] The antibody according to the invention may comprise the framework region of the animal in which the antibody was raised. For example, the antibody according to the invention comprises the CDRs defined above and the framework region of a rabbit antibody, such as an antibody comprising a light chain variable region according to SEQ ID NO:7 (the nucleotide sequence of which is shown in SEQ ID NO:8) and a heavy chain variable region according to SEQ ID NO:11 (the nucleotide sequence of which is shown in SEQ ID NO:12).

[0139] In a preferred embodiment, the antibody according to the invention is humanized.

[0140] In a preferred embodiment, the antibody that binds to HLA-G, which is a humanized antibody, comprises a variable light chain and a variable heavy chain, a. the variable light chain comprises CDR-L1 comprising SEQ ID NO:1, CDR-L2 comprising SEQ ID NO:2, and CDR-L3 comprising SEQ ID NO:3; b. The variable heavy chain comprises CDR-H1 comprising SEQ ID NO:4, CDR-H2 comprising SEQ ID NO:5, and CDR-H3 comprising SEQ ID NO:6.

[0141] As used herein, the term "humanized" antibody refers to an antibody whose heavy and / or light chains contain one or more CDRs (including one or more modified CDRs, if desired) from a donor antibody (e.g., a non-human antibody such as a mouse or rabbit monoclonal antibody) grafted onto the heavy chain variable region framework and / or light chain variable region framework of an acceptor antibody (e.g., a human antibody). For a review, see Vaughan et al, Nature Biotechnology, 16, 535-539, 1998. In one embodiment, instead of transferring the entire CDR, only one or more of the specificity determining residues from any one of the CDRs described hereinabove are transferred to the human antibody framework (see, e.g., Kashmiri et al., 2005, Methods, 36, 25-34). In one embodiment, only the specificity determining residues from one or more of the CDRs described hereinabove are transferred to the human antibody framework. In another embodiment, only the specificity determining residues from each of the CDRs described hereinabove are transferred to the human antibody framework.

[0142] When CDRs are grafted, any suitable acceptor variable region framework sequence may be used, taking into account the class / type of the donor antibody from which the CDRs are derived, including murine, primate and human framework regions.

[0143] Preferably, the humanized antibody according to the invention has a variable domain comprising human acceptor framework regions and one or more of the CDRs specifically provided herein.Thus, in one embodiment, a humanized antibody that binds to HLA-G is provided, the variable domain comprising human acceptor framework regions and non-human donor CDRs.

[0144] Examples of human frameworks that can be used in the present invention include KOL, NEWM, REI, EU, TUR, TEI, LAY and POM (Kabat et al., supra). For example, KOL and NEWM can be used for the heavy chain, REI can be used for the light chain, and EU, LAY and POM can be used for both the heavy and light chains. Alternatively, human germline sequences can be used, which are available at http: / / www.imgt.org / .

[0145] In a humanized antibody according to the invention, the acceptor heavy and light chains do not necessarily have to be derived from the same antibody and can, if desired, comprise composite chains having framework regions derived from different chains.

[0146] A suitable framework region for the light chain of the humanized antibody according to the invention has SEQ ID NO:103 and is derived from the human germline IGKV1D-13 IGKJ4, the nucleotide sequence of which is shown in SEQ ID NO:104.

[0147] A suitable framework region for the heavy chain of the humanized antibody according to the invention has the sequence shown in SEQ ID NO:105 and is derived from the human germline IGHV3-66 IGHJ4, the nucleotide sequence of which is shown in SEQ ID NO:106.

[0148] Thus, in one embodiment, there is provided a humanized antibody that binds to HLA-G, comprising a light chain variable region and a heavy chain variable region: a. The light chain variable region is i. a CDR-L1 comprising SEQ ID NO:1, and ii. A CDR-L2 comprising SEQ ID NO:2; and iii. comprises a CDR-L3 comprising SEQ ID NO:3; b. The heavy chain variable region is i. a CDR-H1 comprising SEQ ID NO:4, and ii. CDR-H2 comprising SEQ ID NO:5; and iii. comprises a CDR-H3 comprising SEQ ID NO:6; A humanized antibody is provided in which the light chain framework regions are derived from human germline IGKV1D-13 IGKJ4 comprising SEQ ID NO:103 and the heavy chain framework regions are derived from human germline IGHV3-66 IGHJ4 comprising SEQ ID NO:105.

[0149] In one embodiment, the antibody of the invention comprises: a. a light chain variable region comprising SEQ ID NO: 19, or 15, or 23, and / or b. Contains a heavy chain variable region comprising SEQ ID NO: 93, 27, 33, 57, 69, 75, 81, or 87.

[0150] In one embodiment, an antibody of the invention comprises a light chain variable region comprising SEQ ID NO:19 and a heavy chain variable region comprising SEQ ID NO:93.

[0151] In one embodiment, an antibody of the invention comprises a light chain variable region comprising SEQ ID NO: 15 and a heavy chain variable region comprising SEQ ID NO: 27. In one embodiment, an antibody of the invention comprises a light chain variable region comprising SEQ ID NO: 19 and a heavy chain variable region comprising SEQ ID NO: 27. In one embodiment, an antibody of the invention comprises a light chain variable region comprising SEQ ID NO: 23 and a heavy chain variable region comprising SEQ ID NO: 27. In one embodiment, an antibody of the invention comprises a light chain variable region comprising SEQ ID NO: 19 and a heavy chain variable region comprising SEQ ID NO: 33. In one embodiment, an antibody of the invention comprises a light chain variable region comprising SEQ ID NO: 19 and a heavy chain variable region comprising SEQ ID NO: 57. In one embodiment, an antibody of the invention comprises a light chain variable region comprising SEQ ID NO: 19 and a heavy chain variable region comprising SEQ ID NO: 69. In one embodiment, an antibody of the invention comprises a light chain variable region comprising SEQ ID NO: 19 and a heavy chain variable region comprising SEQ ID NO: 75. In one embodiment, an antibody of the invention comprises a light chain variable region comprising SEQ ID NO: 19 and a heavy chain variable region comprising SEQ ID NO: 81. In one embodiment, an antibody of the invention comprises a light chain variable region comprising SEQ ID NO: 19 and a heavy chain variable region comprising SEQ ID NO: 87. In one embodiment, an antibody of the invention comprises a light chain variable region comprising SEQ ID NO: 23 and a heavy chain variable region comprising SEQ ID NO: 93.

[0152] In one embodiment, the antibody of the invention is an IgG1. In one embodiment, the antibody of the invention is an IgG1, a. a light chain comprising SEQ ID NO: 21 or 17, or 25, and / or b. Contains a heavy chain comprising SEQ ID NO: 95, 29, 35, 59, 71, 77, 83, or 89.

[0153] In one embodiment, an antibody of the invention comprises a light chain comprising SEQ ID NO:21 and a heavy chain comprising SEQ ID NO:95.

[0154] In one embodiment, an antibody of the invention comprises a light chain comprising SEQ ID NO: 17 and a heavy chain comprising SEQ ID NO: 29. In one embodiment, an antibody of the invention comprises a light chain comprising SEQ ID NO: 21 and a heavy chain comprising SEQ ID NO: 29. In one embodiment, an antibody of the invention comprises a light chain comprising SEQ ID NO: 25 and a heavy chain comprising SEQ ID NO: 29. In one embodiment, an antibody of the invention comprises a light chain comprising SEQ ID NO: 21 and a heavy chain comprising SEQ ID NO: 35. In one embodiment, an antibody of the invention comprises a light chain comprising SEQ ID NO: 21 and a heavy chain comprising SEQ ID NO: 59. In one embodiment, an antibody of the invention comprises a light chain comprising SEQ ID NO: 21 and a heavy chain comprising SEQ ID NO: 71. In one embodiment, an antibody of the invention comprises a light chain comprising SEQ ID NO: 21 and a heavy chain comprising SEQ ID NO: 77. In one embodiment, an antibody of the invention comprises a light chain comprising SEQ ID NO: 21 and a heavy chain comprising SEQ ID NO: 83. In one embodiment, an antibody of the invention comprises a light chain comprising SEQ ID NO: 21 and a heavy chain comprising SEQ ID NO: 89. In one embodiment, an antibody of the invention comprises a light chain comprising SEQ ID NO: 25 and a heavy chain comprising SEQ ID NO: 95.

[0155] Advantageously, IgG1 comprises an active Fc fragment, i.e. has Fc-mediated effector functions. Thus, in one embodiment, the antibody of the invention comprises Fc-mediated effector functions.

[0156] The term "effector function" refers to the biological activity attributable to the Fc region of an antibody, which varies depending on the antibody isotype. Examples of antibody effector functions include Clq binding and complement-dependent cytotoxicity (CDC), Fc receptor binding, antibody-dependent cell-mediated cytotoxicity (ADCC), and antibody-dependent cell-mediated phagocytosis (ADCP).

[0157] The term "complement dependent cytotoxicity" or "CDC" refers to a cell death-inducing mechanism in which the Fc effector domain of a target-binding antibody binds and activates complement component C1q, which then activates the complement cascade, resulting in target cell death.

[0158] The term "antibody-dependent cellular cytotoxicity" or "ADCC" is a cell death-inducing mechanism that depends on the interaction of antibody-coated target cells with effector cells with lytic activity, such as natural killer cells, monocytes, macrophages, and neutrophils, via Fc gamma receptors (FcγR) expressed on the effector cells.

[0159] The term "antibody-dependent cellular phagocytosis" or "ADCP" is a mechanism for inducing phagocytosis that depends on the interaction of antibody-coated target cells or antibody-coated soluble targets with effector cells that have phagocytic activity, such as macrophages and neutrophils, via Fc gamma receptors (FcγR) expressed on the effector cells.

[0160] As described in the Examples, the expression pattern of HLA-G in normal non-tumor tissues was investigated and it was advantageously found that the form of HLA-G containing the epitope bound by the antibody of the present invention is not expressed in healthy tissues, in particular in pancreatic and pituitary tissues.

[0161] This is in contrast to what has been reported previously in the literature, where expression of HLA-G protein in pancreatic islets was reported by Cirulli et al. (Cirulli et al, DIABETES, Vol. 55, May 2006), who observed a marked upregulation of HLA-G in pancreatic islet cells cultured on an extracellular matrix that supports cell replication. Gene expression of HLA-G in the pituitary gland, as well as in pancreatic islets and testes, for example, has also been reported by Boegel et al. (Boegel et al, BMC Medical Genomics (2018) 11:36).

[0162] Thus, the results described in the examples herein are surprising and show that, contrary to what was expected from the teachings of the prior art, for example, antibodies against HLA-G, which can kill cells expressing HLA-G via Fc-mediated effector functions, are potential candidates for the treatment of solid tumors, and no toxicity to patients is expected via binding to normal tissues. Moreover, antibodies containing active Fc can promote the recruitment and activation of effector cells within the tumor microenvironment (TME), and further, direct killing of HLA-G+ tumor cells can induce the release of tumor antigens into the local environment, which further stimulates the immune response.

[0163] The dual mechanism of such antibodies described herein, capable of blocking the interaction between HLA-G and its inhibitory receptor and capable of killing cells, represents a considerable advantage for the treatment of patients with, for example, upregulation of HLA-G in solid tumors.

[0164] Furthermore, tumor heterogeneity suggests that the importance of each mechanism (HLA-G blockade to promote immune cell activation, and direct tumor cell killing via active Fc-dependent mechanisms) may vary between patients. Thus, multiple mechanisms of tumor cell killing may benefit an even broader range of patients by virtue of their ability to engage different mechanisms in tumors with diverse characteristics, e.g., diverse HLA-G expression patterns.

[0165] Methods for selecting HLA-G antibodies of the present invention Due to the particular challenges associated with the production of antibodies against HLA-G (e.g., high homology with other HLA-I, identification of antibodies able to block the interaction between HLA-G and its inhibitory receptors), and in order to identify antibodies that would be useful therapeutically, special discovery, screening and testing strategies had to be developed that involved measurement of binding to HLA-G, evaluation of the affinity and specificity of binding (no cross-reactivity to other HLA-I), and evaluation of the functional properties of the test antibodies, as well as high-throughput measurements of the structural aspects of binding (target epitope residues).

[0166] Therefore, there is provided a method for identifying an antibody according to the invention, comprising the steps of: a) immunizing a non-human mammal with an HLA-G immunogenic composition; b) recovering B cells from the non-human mammal; c) selecting an antibody produced by said B cells having the following properties: i. Dissociation constant K less than 20 nM D binds to HLA-G with an affinity represented by ii. does not bind to HLA-I other than HLA-G; and iii. Blocking the binding between HLA-G and ILT2 and / or between HLA-G and ILT4

[0167] Process a) "Immunogenic composition" refers to a composition capable of generating an immune response in a non-human mammal to which the composition is administered. An immunogenic composition typically allows for the expression of an immunogenic antigen of interest in the mammal to which the composition is administered, against which antibodies can be generated as part of the immune response. "HLA-G immunogenic composition" refers to a composition capable of generating an immune response against HLA-G in a mammal to which the composition is administered.

[0168] "Protein immunization" refers to the technique of administration of an immunogenic protein, which comprises an antigen of interest or an immunogenic portion of said protein that comprises said antigen of interest or an immunogenic portion thereof.

[0169] In one embodiment, the immunogenic composition comprises a full-length protein. In another embodiment, the immunogenic composition comprises an immunogenic portion of a protein. For example, in one embodiment, the immunogenic composition comprises full-length HLA-G complexed with B2m. In another embodiment, the immunogenic composition comprises full-length HLA-G in the absence of B2m. In another embodiment, the immunogenic composition comprises an immunogenic portion of HLA-G associated or not with B2m. In another embodiment, the immunogenic composition comprises the extracellular domain of HLA-G complexed with B2m. In another embodiment, the immunogenic composition comprises the extracellular domain of HLA-G in the absence of B2m.

[0170] "DNA immunization" refers to a technique in which engineered nucleic acid molecules (also referred to herein as nucleic acid or DNA vaccines) encoding a full-length protein or immunogenic portion thereof comprising an antigen of interest are administered directly to mammalian cells to generate an immunological response in said cells against said antigen of interest. DNA immunization uses the host cell machinery to express peptides corresponding to the administered nucleic acid molecule and / or to achieve the expected effect, particularly antigen expression at the cellular level, as well as immunotherapeutic effects at the cellular level or within the host organism.

[0171] "Cellular immunization" refers to the technique of administering cells that naturally express or are transfected with an immunogenic protein, the immunogenic protein comprising an antigen of interest or an immunogenic portion of said protein, including said antigen of interest or an immunogenic portion thereof. In one embodiment, the immunization in step a) is performed using cellular immunization with fibroblasts transfected with an immunogenic protein, the immunogenic protein comprising an antigen of interest or an immunogenic portion of said protein, including said antigen of interest or an immunogenic portion thereof.

[0172] By "immunogenic portion" is meant a portion of a protein or antigen of interest that retains the ability to induce an immune response in a non-human mammal administered said portion of the protein or antigen of interest, or the DNA encoding it, to enable the production of an antibody of the invention.

[0173] HLA-G, including its fusion proteins, and cells expressing HLA-G (recombinantly or naturally) can be used to produce antibodies that specifically recognize HLA-G. Various forms of HLA-G as described herein may be used.

[0174] HLA-G or a fragment thereof for use in immunizing a host can be prepared by processes well known in the art from genetically engineered host cells containing expression systems, or can be recovered from natural biological sources. HLA-G or a fragment thereof can in some instances be part of a larger protein, such as a fusion protein, e.g., fused to an affinity tag, etc.

[0175] In one embodiment, the immunization step may be carried out using protein immunization, DNA immunization or cell immunization or any combination thereof.

[0176] In one embodiment, the non-human mammal is a mouse. In one embodiment, the non-human mammal is a rat. In one embodiment, the non-human mammal is a rabbit. In one embodiment, the rabbit is a New Zealand White rabbit. In one embodiment, the mammal is immunized by subcutaneous injection of the immunogenic composition. In one embodiment, the immunogenic composition comprises rabbit fibroblasts that transiently express HLA-G on the cell surface. In one example, the rabbit fibroblasts are transfected with a DNA sequence encoding HLA-G comprising SEQ ID NO: 111. In another embodiment, the immunogenic composition comprises rabbit fibroblasts that transiently co-express HLA-G and B2m on the cell surface. In one embodiment, the rabbit fibroblasts are Rab9 rabbit fibroblasts.

[0177] The immunization step may be performed using a prime-boost immunization protocol, which refers to an initial administration of the immunogenic composition (prime immunization or prime administration), followed by at least one further administration (boost immunization or boost administration) separated in time from the initial administration during the course of the immunization protocol. A boost immunization encompasses one, two, three or more administrations. In one embodiment, the boost immunization comprises two administrations of the immunogenic composition, spaced 14 days apart. In one embodiment, the immunization step comprises an initial administration of rabbit fibroblasts that transiently express HLA-G on their cell surface, followed by two boost immunizations, spaced 14 days apart. In one embodiment, the rabbit fibroblasts are Rab9 rabbit fibroblasts.

[0178] In one embodiment, the prime immunization includes administration of an adjuvant. In one embodiment, the adjuvant is administered at a site different from the site of injection of the immunogenic composition. In one embodiment, the adjuvant is Freund's adjuvant. In one embodiment, both the prime immunization and the boost immunization include administration of an adjuvant, such as Freund's adjuvant.

[0179] In one embodiment, the immunization step comprises a prime immunization in the presence of a first adjuvant, followed by at least one boost immunization in the presence of a second adjuvant.

[0180] In one embodiment, the immunogenic composition is administered by subcutaneous injection, for example, in the shoulder.

[0181] "Adjuvant" refers to an immunostimulating agent. Adjuvants are substances well known in the art. Conventional adjuvants that act as immunostimulating agents or antigen delivery systems, or both, include, for example, alum, polysaccharides, liposomes, nanoparticles based on biodegradable polymers, lipopolysaccharides. For example, the adjuvant can be Freund's adjuvant, Montanide adjuvant, or Fama adjuvant.

[0182] Step b) Methods for isolating B cells are well known and generally involve isolating B cells from PBMC (peripheral blood mononuclear cells), bone marrow or from secondary lymphoid organs, i.e. lymph nodes or spleen. In one embodiment, the isolation of antigen-specific memory B cells is carried out 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29 or 30 days after the immunization step a). In one embodiment, the isolation of B cells is carried out 14 days after the immunization step a). In one embodiment, step b) involves sorting of antigen-specific B cells by flow cytometry.

[0183] Process c) The screening step in step c) may be carried out, for example, according to methods for measuring binding and blocking activity as described in the present disclosure.

[0184] i. Dissociation constant (K D ) binds to HLA-G with an affinity represented by Binding may be determined for soluble and / or membrane-bound HLA-G. Binding to HLA-G may be determined, for example, by using surface plasmon resonance, such as the Biacore® system, as described in the Examples. In one example, affinity for recombinant HLA-G ECD complexed with B2m was measured by Biacore, as described in the Examples herein, and found to have a K of less than 20 nM. D Antibodies that bind at the dissociation constant K can be selected for further analysis. D is determined by SPR at a temperature of 25° C. between an antibody expressed as a full-length antibody and HLA-G in monomeric form.

[0185] Alternatively, or in addition, binding to HLA-G can be determined by FACS on cells expressing HLA-G, such as HEK293 transfected with HLA-G and B2m, or JEG3 cells that naturally express HLA-G. Methods for measuring the affinity of antibodies to HLA-G by FACS are provided in the examples described herein.

[0186] Advantageously, antibodies may be screened using binding assays against both soluble and cell-expressed HLA-G.

[0187] ii. Does not bind to HLA-I other than HLA-G A special screening strategy has been developed to assess the specificity of binding, i.e., lack of cross-reactivity to other HLA-I, involved in the generation of the HLA-G construct variant, in which amino acids specific for HLA-G are replaced by consensus amino acids found in other HLA-I ("HLA-G null construct"). HLA-G null constructs as described in Example 1 may be particularly useful for screening for antibodies specific for HLA-G.

[0188] "HLA-G Null 1,2,3" corresponds to HLA-G variants in which amino acids specifically expressed on HLA-G α1, α2 and α3 are replaced by consensus amino acids (20 mutated amino acids) expressed on other HLA-I.

[0189] In one embodiment, the method for identifying an antibody according to the invention comprises screening the antibodies recovered after step b) against "HLA-G Null 1,2,3" and selecting antibodies for which no binding is detected. Screening can be performed against cell surface expressed "HLA-G Null 1,2,3" (e.g. comprising SEQ ID NO: 115) or against soluble "HLA-G Null 1,2,3" (ECD) (e.g. comprising SEQ ID NO: 113).

[0190] Binding to other HLA-I can be further evaluated according to the method described in the example.HEK293 cells can be transfected with DNA sequences encoding any of HLA-A, B, C, E or F (e.g., DNA sequences comprising SEQ ID NO: 131, 133, 135, 137 and 139, respectively) and B2m (e.g., DNA sequences comprising SEQ ID NO: 130).Binding to cell-expressed HLA-I can be evaluated by FACS.

[0191] iii. Blocking the binding between HLA-G and ILT2 and / or between HLA-G and ILT4 Blockade of HLA-G binding to ILT2 and / or ILT4 can be assessed by measuring blockade of the interaction of HLA-G associated with B2m expressed on the surface of a cell (e.g., naturally expressed on the surface of JEG3 cells or transiently expressed on the surface of HCT116 cells as described in the Examples) with ILT2 and / or ILT4 expressed as a fusion protein, such as, for example, an Fc fusion protein (ITT2-Fc.ILT4-Fc).

[0192] Antibodies having the required properties and selected after step c) can be further characterized and identified based on additional assays including, for example, specificity assays, ADCC, ADCP, CDC, biophysical assays and stability assays, as well as the ability to modulate the immune environment and induce tumor cell killing in ex vivo cultured human primary tumors, for example the methods described in the examples herein.

[0193] Epitope Within the present invention, the term "epitope" is used interchangeably for both conformational and linear epitopes: conformational epitopes are composed of discontinuous portions of the primary amino acid sequence of an antigen, whereas linear epitopes are formed by sequences formed by contiguous amino acids.

[0194] In one embodiment, the antibody of the present invention specifically binds to the HLA-G alpha 3 domain. In one embodiment, the antibody of the present invention binds to an epitope of HLA-G comprising residues F195 and Y197 with reference to SEQ ID NO: 107.

[0195] In some embodiments, an antibody of the invention binds to an epitope on HLA-G, said epitope comprising residues V194, F195, Y197, E198, Q224, Q226, D227, V248, V249, P250 and Y257 (numbering according to SEQ ID NO: 107).

[0196] In one embodiment, the antibody of the present invention does not bind to B2m. Thus, advantageously, the antibody of the present invention binds to HLA-G complexed with B2m or in the absence of B2m. In one embodiment, the antibody of the present invention does not bind to the same binding site on HLA-G as the HLA-G cognate peptide that is naturally expressed in a complex with HLA-G. Thus, in one embodiment, the antibody of the present invention does not block the association of HLA-G with its cognate peptide that is naturally expressed in a complex with HLA-G.

[0197] In one embodiment, the invention provides an anti-HLA-G antibody that binds to an epitope on HLA-G, wherein the epitope includes at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, or all of the residues selected from the list consisting of V194, F195, Y197, E198, Q224, Q226, D227, V248, V249, P250 and Y257 of HLA-G (SEQ ID NO: 107). In one embodiment, the invention provides a humanized IgG1 antibody that binds to an epitope of HLA-G, wherein the epitope comprises residues V194, F195, Y197, E198, Q224, Q226, D227, V248, V249, P250 and Y257 of human HLA-G (SEQ ID NO: 107). In one embodiment, the antibody is a defucosylated IgG1.

[0198] In one embodiment, the invention provides an anti-HLA-G antibody that binds to an epitope on HLA-G, said epitope comprising at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, or all of the residues selected from the list consisting of V194, F195, Y197, E198, Q224, Q226, D227, V248, V249, P250 and Y257 of HLA-G (SEQ ID NO: 107), as determined by a contact distance of less than 4 A. In one embodiment, the antibody is a defucosylated IgG1.

[0199] In one embodiment, the invention provides a humanized IgG1 antibody that binds to an epitope of HLA-G, said epitope comprising at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, or all of the residues selected from the list consisting of V194, F195, Y197, E198, Q224, Q226, D227, V248, V249, P250 and Y257 of HLA-G (SEQ ID NO: 107), when determined with a contact distance of less than 4 A. In one embodiment, the antibody is a defucosylated IgG1.

[0200] In one embodiment, the invention provides a humanized IgG1 antibody that binds to an epitope on HLA-G, said epitope comprising at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, or all of the residues selected from the list consisting of V194, F195, Y197, E198, R219, Q224, Q226, D227, V248, V249, P250, E253 and Y257 of HLA-G (SEQ ID NO: 107), as determined by a contact distance of less than 5 A. In one embodiment, the antibody is a defucosylated IgG1.

[0201] The epitope can be identified by any suitable epitope mapping method known in the art in combination with any one of the antibodies provided by the present invention. Examples of such methods include screening peptides of various lengths derived from full-length HLA-G for binding to the antibody of the present invention or fragments thereof, and identifying the smallest fragment that can specifically bind to the antibody containing the sequence of the epitope recognized by the antibody. HLA-G peptides can be produced synthetically or by proteolytic digestion of HLA-G. The peptides that bind to the antibody can be identified, for example, by mass spectrometry. Methods such as X-ray crystallography, nuclear magnetic resonance (NMR) spectroscopy or hydrogen deuterium exchange mass spectrometry (HDX-MS) can be used to identify the epitope bound by the antibody. Typically, when epitope determination is performed by X-ray crystallography, the amino acid residues of the antigen within 4 Å of the CDR are considered to be the amino acid residues of the epitope. After identification, the epitope can be useful for preparing fragments that bind to the antibody of the present invention, and if necessary, can be used as an immunogen to obtain additional antibodies that bind to the same epitope.

[0202] The epitopes shown in the aspects and embodiments of the present invention are preferably epitopes characterized by X-ray crystallography.In one embodiment, the present invention provides an anti-HLA-G antibody that binds to an epitope on HLA-G, said epitope comprising at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10 or all of the residues selected from the list consisting of V194, F195, Y197, E198, Q224, Q226, D227, V248, V249, P250 and Y257 (SEQ ID NO: 107) of HLA-G when determined by a contact distance of less than 4 Å, and said epitope is characterized by X-ray crystallography.

[0203] In one embodiment, the invention provides a humanized IgG1 antibody that binds to an epitope of HLA-G, said epitope comprising at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, or all of the residues selected from the list consisting of V194, F195, Y197, E198, Q224, Q226, D227, V248, V249, P250 and Y257 of HLA-G (SEQ ID NO: 107), as determined by a contact distance of less than 4 Å, and wherein the epitope is characterized by X-ray crystallography. In one embodiment, the antibody is a defucosylated IgG1.

[0204] In one embodiment, the invention provides a humanized IgG1 antibody that binds to an epitope of HLA-G, said epitope comprising at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, or all of the residues selected from the list consisting of V194, F195, Y197, E198, R219, Q224, Q226, D227, V248, V249, P250, E253 and Y257 of HLA-G (SEQ ID NO: 107), as determined by X-ray crystallography with a contact distance of less than 5 A. In one embodiment, the antibody is a defucosylated IgG1.

[0205] Additionally, HDX-MS and NMR can be used to analyze interactions in solution and may indicate allosteric or conformational changes not necessarily evident by crystallography. For example, from HDX-MS with 30 s deuterium incubation, potential binding regions identified were 178-MLQRADPPKTHVTHHPVFD-196, and 214-ILTWQRDGEDQTQDVEL-230.

[0206] In one embodiment, the epitope determined by NMR, where stringency is defined as increasing as the average of all calculated shifts is exceeded (>0.0764), includes residues T200, L201, L215, W217, R219, D220, E229, A245, A246, V247, V249, S251, E253, Q255, T258, H260, V261 and W274.

[0207] In one embodiment, the epitope determined by NMR, where stringency is defined as increasing as the calculated total shift exceeds the mean plus one standard deviation (>0.1597), includes residues H191, Y197, E198, R202, L230, V248, G252, C259 and K275.

[0208] The antibody may compete for binding to HLA-G with, or bind to the same epitope as, an antibody defined above with respect to its light chain, heavy chain, light chain variable region, heavy chain variable region or CDR sequences.

[0209] In particular, the present invention provides an antibody that competes for binding to HLA-G or binds to the same epitope as an antibody that comprises the combination of CDR-L1 / CDR-L2 / CDR-L3 / CDR-H1 / CDR-H2 / CDR-H3 sequences of SEQ ID NO: 1 / 2 / 3 / 4 / 5 / 6. The antibody may compete for binding to HLA-G or bind to the same epitope as an antibody that comprises the VL and VH sequence pairs of SEQ ID NO: 19 and 93, respectively. The antibody may compete for binding to HLA-G or bind to the same epitope as an IgG1 that comprises the combination of CDR-L1 / CDR-L2 / CDR-L3 / CDR-H1 / CDR-H2 / CDR-H3 sequences of SEQ ID NO: 1 / 2 / 3 / 4 / 5 / 6. The antibody may compete for binding to HLA-G or bind to the same epitope as an IgG1 that comprises the VL and VH sequence pairs of SEQ ID NO: 19 and 93, respectively.

[0210] In one embodiment, the invention provides an antibody that cross-competes for binding to HLA-G with an antibody comprising the CDR-L1 / CDR-L2 / CDR-L3 / CDR-H1 / CDR-H2 / CDR-H3 sequence combination of SEQ ID NOs: 1 / 2 / 3 / 4 / 5 / 6.

[0211] In the context of the present invention, the antibodies provided herein that compete for binding to HLA-G with a reference antibody according to the present invention or that bind to the same epitope retain the advantageous properties of the reference antibody described in the above section, e.g. specificity for HLA-G, high affinity, ILT2 and / or ILT4 blocking activity. In one example, an antibody that competes for binding to HLA-G with a reference antibody according to the present invention or that binds to the same epitope has: a dissociation constant (K) between said antibody expressed as a full-length antibody and a monomeric form of HLA-G of less than 20 nM, particularly less than 15 nM, particularly less than 10 nM, particularly less than 9 nM, particularly less than 8 nM, particularly less than 7 nM, particularly less than 6 nM or particularly less than 5 nM, as determined, for example, by SPR at a temperature of 25° C. D ); and / or b. An IC of less than 20 pM for blocking the binding of ILT2 to HLA-G naturally expressed on the surface of JEG3 cells, as determined using an in vitro assay using a high reaction volume, e.g., as described in Example 8. 50 and / or c. An IC of less than 1400 pM for blocking ILT4 binding to HLA-G, as determined, for example, as described in Example 8. 50 .

[0212] To determine whether an antibody competes with a reference antibody for binding, the above-mentioned binding method is carried out in two different experimental settings. In the first setting, the reference antibody is bound to the antigen under saturating conditions, and then the binding of the test antibody to the antigen is evaluated. In the second setting, the test antibody is bound to the antigen under saturating conditions, and then the binding of the reference antibody to the protein / peptide is evaluated. In both experimental settings, if only the first (saturating) antibody can bind to the protein / peptide, it is concluded that the test antibody and the reference antibody compete for binding to the antigen. As will be understood by those skilled in the art, an antibody that competes with a reference antibody for binding does not necessarily bind to the same epitope as the reference antibody, but may sterically block the binding of the reference antibody by binding to an overlapping or adjacent epitope, or may cause a conformational change that results in the lack of binding.

[0213] Two antibodies bind to the same or overlapping epitopes if each competitively inhibits (blocks) the binding of the other to the antigen. Alternatively, two antibodies have the same epitope if essentially every amino acid mutation in the antigen that reduces or eliminates binding of one antibody reduces or eliminates binding of the other. Two antibodies have overlapping epitopes if some amino acid mutations that reduce or eliminate binding of one antibody reduce or eliminate binding of the other.

[0214] Additional routine experiments (e.g., peptide mutations and binding analysis) can then be performed to confirm whether the observed lack of binding of the test antibody is indeed due to binding to the same antigenic moiety as the reference antibody, or whether steric blocking (or another phenomenon) is responsible for the observed lack of binding. This type of experiment can be performed using ELISA, RIA, surface plasmon resonance, flow cytometry, or any other quantitative or qualitative antibody binding assay available in the art.

[0215] Antibody variants It will also be understood by those skilled in the art that antibodies may undergo various post-translational modifications. The type and extent of these modifications often depend on the host cell line used to express the antibody and the culture conditions. Such modifications may include variations in glycosylation, methionine oxidation, diketopiperazine formation, aspartic acid isomerization and asparagine deamidation. A common modification is the loss of carboxy-terminal basic residues (such as lysine or arginine) by the action of carboxypeptidase (as described in Harris, RJ. Journal of Chromatography 705:129-134, 1995). Thus, the C-terminal lysine of antibody heavy chains may not be present.

[0216] In one embodiment, the C-terminal amino acid from the antibody is cleaved during post-translational modification.

[0217] In one embodiment, the N-terminal amino acid from the antibody is cleaved during post-translational modification.

[0218] In certain embodiments, antibody variants are provided that have one or more amino acid substitutions, insertions and / or deletions.The target sites for substitution mutagenesis include CDR and FR.Amino acid substitutions can be introduced into the antibody of interest, and the product can be screened for desired activity, such as retained / improved antigen binding, reduced immunogenicity, or improved ADCC, CDC and / or ADCP.

[0219] In certain embodiments, amino acid sequence variants of the antibodies described herein are contemplated. For example, it may be desirable to improve the binding affinity and / or other biological properties of the antibody. Amino acid sequence variants of anti-HLA-G antibodies can be prepared by introducing appropriate modifications into the nucleotide sequence encoding the protein or by peptide synthesis. Such modifications include, for example, deletion from, and / or insertion into, and / or substitution of residues within the amino acid sequence of anti-HLA-G antibodies (e.g., within one or more CDR and / or framework sequences, or within VH and / or VL domains). Any combination of deletion, insertion and substitution can be made to arrive at the final construct, as long as the final construct has the desired properties.

[0220] In certain embodiments of the variant VH and variant VL sequences provided herein, each HVR is unchanged or contains no more than one, two or three amino acid substitutions.

[0221] It will be understood that one or more amino acid substitutions, additions and / or deletions may be made to the CDRs provided by the present invention without significantly altering the ability of the antibody to bind to HLA-G and neutralize HLA-G activity. The effect of any amino acid substitution, addition and / or deletion can be easily tested by one skilled in the art, for example, by determining inhibition of HLA-G binding and HLA-G interaction with its natural ligand, using the methods described herein, particularly those illustrated in the examples.

[0222] As a result, in certain embodiments of the variant VH and variant VL sequences, each CDR contains no more than one, two or three amino acid substitutions, where such amino acid substitutions are conservative and the antibody retains its binding properties to HLA-G.

[0223] Thus, the present invention provides an anti-HLA-G antibody comprising one or more CDRs selected from CDR-L1 (comprising SEQ ID NO: 1), CDR-L2 (comprising SEQ ID NO: 2), CDR-L3 (comprising SEQ ID NO: 3), CDR-H1 (comprising SEQ ID NO: 4), CDR-H2 (comprising SEQ ID NO: 5) and CDR-H3 (comprising SEQ ID NO: 6), wherein one or more amino acids in one or more of the CDRs have been replaced by another amino acid, e.g. a similar amino acid as defined herein below.

[0224] In one embodiment, the invention provides an anti-HLA-G antibody comprising CDR-L1 (comprising SEQ ID NO:1), CDR-L2 (comprising SEQ ID NO:2), CDR-L3 (comprising SEQ ID NO:3), CDR-H1 (comprising SEQ ID NO:4), CDR-H2 (comprising SEQ ID NO:5) and CDR-H3 (comprising SEQ ID NO:6), e.g., one or more amino acids in one or more of the CDRs have been replaced by another amino acid, e.g., a similar amino acid as defined herein below.

[0225] In one embodiment, the anti-HLA-G antibody of the invention comprises a light chain variable region comprising three CDRs, in which the sequence of CDR-L1 comprises a sequence having at least 70%, 80%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity or similarity to the sequence shown in SEQ ID NO:1, and / or the sequence of CDR-L2 comprises a sequence having at least 70%, 80%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity or similarity to the sequence shown in SEQ ID NO:2, and / or the sequence of CDR-L3 comprises a sequence having at least 70%, 80%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity or similarity to the sequence shown in SEQ ID NO:3.

[0226] In one embodiment, the anti-HLA-G antibody of the invention comprises a heavy chain variable region comprising three CDRs, wherein the sequence of CDR-H1 comprises a sequence having at least 70%, 80%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity or similarity to the sequence shown in SEQ ID NO: 4, and / or the sequence of CDR-H2 comprises a sequence having at least 70%, 80%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity or similarity to the sequence shown in SEQ ID NO: 5, and / or the sequence of CDR-H3 comprises a sequence having at least 70%, 80%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity or similarity to the sequence shown in SEQ ID NO: 6.

[0227] In one embodiment, an anti-HLA-G antibody of the invention comprises a light chain variable region comprising a sequence having at least 70%, 80%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity or similarity to the sequence set forth in SEQ ID NO:19 or SEQ ID NO:15 or SEQ ID NO:23.

[0228] In one embodiment, an anti-HLA-G antibody of the invention comprises a heavy chain variable region comprising a sequence having at least 70%, 80%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity or similarity to the sequence set forth in SEQ ID NO: 93, 27, 33, 57, 69, 75, 81 or 87.

[0229] In one embodiment, an anti-HLA-G antibody of the invention comprises a light chain variable region and a heavy chain variable region, wherein the light chain variable region comprises a sequence having at least 70%, 80%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity or similarity to the sequence set forth in SEQ ID NO: 19, and / or the heavy chain variable region comprises a sequence having at least 70%, 80%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity or similarity to the sequence set forth in SEQ ID NO: 93.

[0230] In one embodiment, the anti-HLA-G antibody of the invention comprises a CDR-L1 / CDR-L2 / CDR-L3 / CDR-H1 / CDR-H2 / CDR-H3 sequence comprising SEQ ID NO:1, 2, 3, 4, 5 and 6, respectively, and the remainder of the light chain variable region and heavy chain variable region have at least 70%, 80%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity or similarity to SEQ ID NO:19 and SEQ ID NO:93, respectively.

[0231] In one embodiment, the anti-HLA-G antibody of the invention comprises a light chain comprising a sequence having at least 70%, 80%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity or similarity to the sequence set forth in SEQ ID NO:21 or SEQ ID NO:17 or SEQ ID NO:25.

[0232] In one embodiment, an anti-HLA-G antibody of the invention comprises a heavy chain comprising a sequence having at least 70%, 80%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity or similarity to the sequence set forth in SEQ ID NO: 95, 29, 35, 59, 71, 77, 83 or 89.

[0233] In one embodiment, the anti-HLA-G antibody of the invention is an IgG1 comprising a light chain comprising a sequence having at least 70%, 80%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity or similarity to the sequence set forth in SEQ ID NO:21, and a heavy chain comprising a sequence having at least 70%, 80%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity or similarity to the sequence set forth in SEQ ID NO:95.

[0234] In one embodiment, the anti-HLA-G antibody of the invention is an IgG1 comprising the CDR-L1 / CDR-L2 / CDR-L3 / CDR-H1 / CDR-H2 / CDR-H3 sequences shown in SEQ ID NOs: 1, 2, 3, 4, 5 and 6, respectively, and the remainder of the light and heavy chains having at least 70%, 80%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity or similarity to SEQ ID NO: 21 and SEQ ID NO: 95, respectively.

[0235] The anti-HLA-G antibody variants provided herein by the invention retain the advantageous properties of the parent antibody (i.e. the unmodified antibody), i.e. the functional properties mentioned above, e.g. high specificity, high affinity, ILT2 and / or ILT4 blocking activity. In one example, the anti-HLA-G antibody variants provided by the invention have: a dissociation constant (K) between said antibody expressed as a full-length antibody and a monomeric form of HLA-G of less than 20 nM, particularly less than 15 nM, particularly less than 10 nM, particularly less than 9 nM, particularly less than 8 nM, particularly less than 7 nM, particularly less than 6 nM or particularly less than 5 nM, as determined, for example, by SPR at a temperature of 25° C. D ); and / or b. An IC of less than 20 pM for blocking the binding of ILT2 to HLA-G naturally expressed on the surface of JEG3 cells, as determined using an in vitro assay using a high reaction volume, e.g., as described in Example 8. 50 and / or c. An IC of less than 1400 pM for blocking ILT4 binding to HLA-G, as determined, for example, as described in Example 8. 50 .

[0236] Sequence identity and similarity The degree of identity and similarity between sequences can be easily calculated. "Sequence identity %" (or "sequence similarity %") is calculated by: (1) comparing two optimally aligned sequences over a comparison window (e.g., the length of the longer sequence, the length of the shorter sequence, a specified window, etc.), (2) determining the number of positions that contain identical (or similar) amino acids (e.g., identical amino acids are present in both sequences, similar amino acids are present in both sequences) to obtain the number of matched positions, (3) dividing the number of matched positions by the total number of positions within the comparison window (e.g., the length of the longer sequence, the length of the shorter sequence, a specified window), and (4) multiplying the result by 100 to obtain the sequence identity % or sequence similarity %.

[0237] The method of aligning sequences for comparison is well known in the art.The optimal alignment of sequences for comparison can be carried out, for example, by the local homology algorithm of Smith&Waterman, Adv.Appl.Math.2:482(1981), by the homology alignment algorithm of Needleman&Wunsch, J.Mol.Biol.48:443(1970), by the similarity search method of Pearson&Lipman, Proc.Nat'l.Acad.Sci.USA 85:2444(1988), by computer implementation of these algorithms (GAP, BESTFIT, FASTA and TFASTA in Wisconsin Genetics Software Package, Genetics Computer Group, 575 Science Dr., Madison, Wis.), or by manual alignment and visual inspection (see, for example, Current Protocols in Molecular Biology (Ausubel et al., eds.1995 supplement)).

[0238] Preferred examples of algorithms suitable for determining percent sequence identity and percent sequence similarity include the BLAST and BLAST 2.0 algorithms described in Altschul et al., Nuc.Acids Res.25:3389-3402 (1977) and Altschul et al., J.Mol.Biol.215:403-410 (1990).Polypeptide sequences can also be compared using FASTA using default or recommended parameters.FASTA (e.g., FASTA2 and FASTA3) provides alignment and percent sequence identity of the best overlapping regions between query and search sequences.

[0239] In certain embodiments, substitutions, insertions or deletions may be made in one or more CDRs, so long as such changes do not substantially reduce the ability of the antibody to bind to its target.

[0240] For example, conservative changes may be made to the CDRs that do not substantially reduce binding affinity. Such changes may be made outside the antigen contact residues within the CDRs.

[0241] Conservative substitutions are shown in Table 1, along with more specific "exemplary substitutions." [Table 1]

[0242] Substantial modifications to the biological properties of antibody variants can be achieved by selecting substitutions that differ significantly in their effect on the structure of the polypeptide backbone in the region of the substitution, on the charge or hydrophobicity of the molecule at the target site, or on maintaining the bulk of the side chain. Amino acids can be grouped according to the similarity of the properties of their side chains (ALLehninger, Biochemistry second ed., pp. 73-75, Worth Publishers, New York (1975)).

[0243] One type of substitution variant involves substituting one or more CDR region residues of a parent antibody (humanized or human antibody). Generally, the resulting variants selected for further testing have changes in certain biological properties (e.g., increased affinity, reduced immunogenicity) compared to the parent antibody and / or substantially retain certain biological properties of the parent antibody. An exemplary substitution variant is an affinity matured antibody, which can be conveniently generated, for example, using phage display-based affinity maturation techniques. In summary, one or more CDR residues are mutated, and the variant antibodies are displayed on phage and screened for a certain biological activity (e.g., binding affinity).

[0244] For example, changes (e.g., substitutions) may be made to CDRs to improve antibody affinity. Such changes may be made to HVR "hot spots," i.e., residues encoded by codons that undergo frequent mutation during the somatic maturation process (see, e.g., Chowdhury, Methods Mol. Biol. 207:179-196 (2008)), and / or to residues that contact the antigen, and the resulting mutant VH or VL are tested for binding affinity. Affinity maturation by construction of secondary libraries and reselection from the secondary libraries is described, for example, in Hoogenboom et al. Methods in Molecular Biology 178:1-37 (O'Brien et al., ed., Human Press, Totowa, NJ, (2001)). In some embodiments of affinity maturation, diversity is introduced into the variable genes selected for maturation by any of a variety of methods (e.g., error-prone PCR, chain shuffling, or oligonucleotide-directed mutagenesis). A secondary library is then created. The library is then screened to identify any antibody variants with the desired affinity.

[0245] One method that can be used to identify antibody residues or regions that can be targeted for mutagenesis is alanine scanning mutagenesis (Cunningham and Wells (1989) Science, 244:1081-1085). In this method, a residue, or a number of target residues, are identified and replaced with alanine to determine whether the interaction of the antibody with the antigen is affected. Alternatively, or in addition, an X-ray structure of an antigen-antibody complex can be used to identify contact points between the antibody and its antigen. Mutants may be screened to determine whether they contain the desired properties.

[0246] Constant region mutants In some embodiments, one or more amino acid modifications may be introduced into the Fc region of an antibody provided herein, thereby generating an Fc region variant. The Fc region variant may comprise a human Fc region sequence (e.g., a human IgG1, IgG2, IgG3, or IgG4 Fc region) that contains an amino acid modification (e.g., a substitution) at one or more amino acid positions.

[0247] Certain antibody variants with improved or diminished binding to FcRs have been described (see, e.g., U.S. Pat. No. 6,737,056; WO 2004 / 056312; and Shields et al., J. Biol. Chem. 9(2):6591-6604 (2001).

[0248] Antibodies with increased half-lives and improved binding to the neonatal Fc receptor (FcRn) are described, for example, in US 2005 / 0014934. These antibodies comprise an Fc region with one or more substitutions that improve binding of the Fc region to FcRn.

[0249] In certain embodiments, the antibody variant comprises an Fc region with one or more amino acid substitutions that improve ADCC, e.g., substitutions at positions 298, 333 and / or 334 (EU numbering of residues) of the Fc region.

[0250] Antibodies with reduced effector function include those with substitutions at one or more of Fc region residues 234, 235, 237, 238, 265, 269, 270, 297, 327 and 329 (see, e.g., U.S. Patent No. 6,737,056). Such Fc variants include Fc variants with substitutions at two or more of amino acid positions 265, 269, 270, 297 and 327, where amino acid residues are numbered according to the EU numbering system.

[0251] In order to confirm the reduction / depletion of CDC and / or ADCC activity, in vitro and / or in vivo cytotoxicity assays can be performed. For example, Fc receptor (FcR) binding assays can be performed to ensure that the antibody lacks FcγR binding (and therefore likely lacks ADCC activity) but retains FcRn binding ability. The primary cells for mediating ADCC, NK cells, express only FcγRIII, whereas monocytes express FcRI, FcγRII and FcγRIII. FcR expression on hematopoietic cells is summarized in Ravetch and Kinet, Annu. Rev. Immunol. 9:457-492 (1991). Non-limiting examples of in vitro assays for evaluating the ADCC activity of a molecule of interest are described in U.S. Patent No. 5,500,362, U.S. Patent No. 5,821,337. Alternatively, or in addition, the ADCC activity of the molecule of interest can be evaluated in vivo, for example, in animal models such as those disclosed in Clynes et al.Proc.Nat l Acad.Sci.USA 95:652-656 (1998).Clq binding assay can be carried out to confirm that the antibody cannot bind to Clq and therefore lacks CDC activity.See, for example, the Clq and C3c binding ELISA in WO2006 / 029879 and WO2005 / 100402. To assess complement activation, a CDC assay may be performed (see, e.g., Gazzano-Santoro et al, J. Immunol. Methods 202:163 (1996); Cragg, MS et al, Blood 101:1045-1052 (2003); and Cragg, MS and MI Glennie, Blood 103:2738-2743 (2004)). FcRn binding and in vivo clearance / half-life determinations can also be performed using methods known in the art (see, e.g., Petkova, SB et al, Int l. Immunol. 18(12):1759-1769 (2006)).

[0252] The constant region domain of the antibody molecule of the present invention, if any, can be selected taking into account the proposed function of the antibody molecule, in particular the effector function that may be required. For example, the constant region domain can be a human IgA, IgD, IgE, IgG or IgM domain. In particular, when the antibody molecule is intended for therapeutic use and antibody effector function is required, human IgG constant region domains, in particular IgG1 and IgG3 isotypes, can be used. Alternatively, when the antibody molecule is intended for therapeutic use and antibody effector function is not required, IgG2 and IgG4 isotypes can be used. It will be understood that sequence variants of these constant region domains can also be used.

[0253] In some embodiments, the antibodies of the invention are wild-type human IgG1 (referred to as IgG1).

[0254] In some embodiments, the antibody of the invention is IgG1 LALA, a mutant of the wild-type human IgG1 isoform in which the amino acid substitutions L234A / L235A (according to EU numbering) are introduced in the constant region of IgG1. In one embodiment, the antibody of the invention comprises a light chain comprising the sequence of SEQ ID NO:21 and a heavy chain comprising the sequence of SEQ ID NO:97.

[0255] In some embodiments, the antibody of the invention is IgG1 LALAGA, a variant of the wild-type human IgG1 isoform in which the amino acid substitutions L234A / L235A / G237A (according to EU numbering) are introduced in the IgG1 constant region.

[0256] In some embodiments, the antibody of the invention is IgG4P, a mutant of the wild-type human IgG4 isoform in which amino acid 228 (according to EU numbering) is replaced by proline, e.g., as described in Angal et al., Molecular Immunology, 1993, 30(1), 105-108. In one embodiment, the antibody of the invention comprises a light chain comprising the sequence of SEQ ID NO:21 and a heavy chain comprising the sequence of SEQ ID NO:99.

[0257] In some embodiments, the antibodies of the invention are IgG4 FALA, i.e. mutants of wild-type human IgG4 isoform in which the substitutions F234A / L235A (according to EU numbering) have been introduced into the constant region of IgG4.

[0258] In some embodiments, the antibody of the invention is IgG4P FALA, a mutant of the wild-type human IgG4 isoform in which amino acid 228 (according to EU numbering) is replaced by proline and the amino acid substitutions F234A / L235A (according to EU numbering) are introduced in the IgG4 constant region. In one embodiment, the antibody of the invention comprises a light chain comprising the sequence of SEQ ID NO:21 and a heavy chain comprising the sequence of SEQ ID NO:101.

[0259] Glycosylation variants In certain embodiments, the antibodies provided herein are altered to increase or decrease the extent to which the antibody is glycosylated. Addition or deletion of glycosylation sites to an antibody can be conveniently accomplished by altering the amino acid sequence such that one or more glycosylation sites are created or removed.

[0260] In one embodiment, the antibody of the present invention is glycosylated. In one embodiment, the antibody of the present invention has a low or no fucose content. By "fucose content" is meant the proportion of fucosylated forms in the N-glycan attached to the Asn297 residue of the Fc fragment of each heavy chain of each antibody. By "low fucose content" is meant a fucose content of 65 percent or less. In fact, it is now known that the low fucose content of an antibody composition plays a decisive role in the ability of said composition to induce a strong ADCC response via Fc gamma RIII. Advantageously, the fucose content is 65 percent or less, preferably 60 percent, 55 percent or 50 percent or less, or even 45 percent, 40 percent, 35 percent, 30 percent, 25 percent or 20 percent or less. However, the fucose content does not have to be null, and may be, for example, 5 percent, 10 percent, 15 percent or 20 percent or less.

[0261] In one embodiment, the antibody of the present invention is a defucosylated IgG1. Methods for producing defucosylated IgG1 are well known and include the production of cells genetically modified for the production of defucosylated antibodies. In one embodiment, the defucosylated IgG1 of the present invention is produced in CHO cells in which the gene encoding alpha 1,6 fucosyltransferase (FUT8) has been genetically knocked out by methods well known in the art. For example, KO FUT8 CHOSXE / DG44 cells may be used, as described in the examples provided herein.

[0262] In one embodiment, the antibodies of the invention have improved ADCC and / or ADCP and / or improved ability to deplete tumor cells expressing HLA-G. In one embodiment, the ADCC and / or ADCP function and / or ability to deplete tumor cells expressing HLA-G of the defucosylated antibodies according to the invention is improved compared to the corresponding conventional (i.e. fucosylated) antibody (i.e. an antibody comprising the same amino acid sequence but not modified and comprising fucose, for example, produced in CHO cells expressing FUT8). In the context of the present invention, "improved" activity (e.g. ADCC and / or ADCP and / or tumor cell depletion) means that the activity (e.g. ADCC and / or ADCP and / or tumor cell depletion) of the defucosylated antibody is at least 1%, 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50% higher than the same activity of the reference antibody (i.e. the corresponding conventional, i.e. fucosylated antibody).

[0263] In vitro and / or in vivo cytotoxicity assays well known in the art can be performed to confirm, for example, the increased ADCC and / or ADCP activity described herein.

[0264] In one embodiment, the antibody of the invention comprises: a. a light chain comprising SEQ ID NO: 21, or 17, or 25, and / or b. A defucosylated IgG1 comprising a heavy chain comprising SEQ ID NO: 95, 29, 35, 59, 71, 77, 83 or 89.

[0265] In one embodiment, the antibody of the invention comprises: a. a light chain comprising at least 90% identity or similarity to SEQ ID NO: 21, or 17, or 25; and / or b. A defucosylated IgG1 comprising a heavy chain comprising at least 90% identity or similarity to SEQ ID NO: 95, 29, 35, 59, 71, 77, 83 or 89.

[0266] In one embodiment, the anti-HLA-G antibody of the invention is a defucosylated IgG1 comprising a light chain comprising a sequence having at least 70%, 80%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity or similarity to the sequence set forth in SEQ ID NO:21, and a heavy chain comprising a sequence having at least 70%, 80%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity or similarity to the sequence set forth in SEQ ID NO:95.

[0267] In one embodiment, the anti-HLA-G antibody of the invention is a defucosylated IgG1 comprising the CDR-L1 / CDR-L2 / CDR-L3 / CDR-H1 / CDR-H2 / CDR-H3 sequences shown in SEQ ID NOs: 1, 2, 3, 4, 5 and 6, respectively, and the remainder of the light and heavy chains having at least 70%, 80%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity or similarity to SEQ ID NO: 21 and SEQ ID NO: 95, respectively.

[0268] In one embodiment, the anti-HLA-G antibody of the invention is a defucosylated IgG1 comprising a light chain variable region comprising SEQ ID NO: 19 and a heavy chain variable region comprising SEQ ID NO: 93, and the remainder of the light and heavy chains have at least 90% identity or similarity to SEQ ID NOs: 21 and 95, respectively.

[0269] In one embodiment, an antibody of the invention comprises a light chain comprising SEQ ID NO: 21 and a heavy chain comprising SEQ ID NO: 95 and has improved ADCC and / or ADCP function and / or improved ability to deplete tumor cells expressing HLA-G. In one embodiment, the ADCC and / or ADCP function and / or ability to deplete tumor cells expressing HLA-G is improved compared to a corresponding conventional antibody (i.e., a fucosylated antibody comprising a light chain comprising SEQ ID NO: 21 and a heavy chain comprising SEQ ID NO: 95).

[0270] In one embodiment, an antibody of the invention comprises a light chain comprising SEQ ID NO: 21 and a heavy chain comprising SEQ ID NO: 95 and has improved CDC function and / or improved ability to deplete tumor cells expressing HLA-G. In one embodiment, the CDC function and / or ability to deplete tumor cells expressing HLA-G is improved compared to a corresponding conventional antibody (i.e. a fucosylated antibody comprising a light chain comprising SEQ ID NO: 21 and a heavy chain comprising SEQ ID NO: 95).

[0271] In one embodiment, the anti-HLA-G antibody of the invention is a defucosylated IgG1 comprising a light chain variable region comprising SEQ ID NO: 19 and a heavy chain variable region comprising SEQ ID NO: 93, or a light chain comprising SEQ ID NO: 21 and a heavy chain comprising SEQ ID NO: 95, with improved ADCC and / or ADCP and / or CDC function and with improved ability to deplete tumor cells expressing HLA-G.

[0272] Biological molecules such as antibodies contain acidic and / or basic functional groups, which give the molecule a net positive or negative charge. The amount of overall "observed" charge depends on the absolute amino acid sequence of the entity, the local environment of the charged groups within the 3D structure, and the environmental conditions of the molecule. The isoelectric point (pI) is the pH at which a particular molecule, or its solvent-accessible surface, carries no net charge. In one example, an antibody that binds to HLA-G can be engineered to have a suitable isoelectric point. This can result in an antibody with more robust properties, particularly favorable solubility and / or stability profiles, and / or improved purification properties.

[0273] Antibodies can be engineered by replacing amino acid residues, for example by replacing an acidic amino acid residue with one or more basic amino acid residues. Alternatively, basic amino acid residues can be introduced or acidic amino acid residues can be removed. Alternatively, if the molecule has an unacceptably high pI value, acidic residues can be introduced to lower the pI, if necessary. It is important to note that when engineering the pI, care must be taken to retain the desired activity of the antibody or fragment. Thus, in one embodiment, the engineered antibody has the same or substantially the same activity as the "unmodified" antibody or fragment.

[0274] **Programs such as ExPASY http: / / www.expasy.ch / tools / pi_tool.html, and http: / / www.iut-arles.up.univ-mrs.fr / w3bb / d_abim / compo-p.html may be used to predict the isoelectric point of antibodies.

[0275] Effector molecules If desired, the antibodies for use in the present invention may be conjugated to one or more effector molecules.

[0276] It will be understood that effector molecule can comprise a single effector molecule or two or more such molecules linked to form a single moiety that can be bound to the antibody of the present invention.If it is desired to obtain an antibody fragment linked to an effector molecule, this can be prepared by standard chemical or recombinant DNA procedures, in which the antibody fragment is linked to the effector molecule directly or via a coupling agent.Technologies for conjugating such effector molecules to antibodies are well known in the art (see Hellstrom et al., Controlled Drug Delivery, 2nd Ed., Robinson et al., eds., 1987, pp.623-53; Thorpe et al., 1982, Immunol.Rev., 62:119-58 and Dubowchik et al., 1999, Pharmacology and Therapeutics, 83, 67-123). Particular chemical procedures include, for example, those described in WO 93 / 06231, WO 92 / 22583, WO 89 / 00195, WO 89 / 01476 and WO 03031581. Alternatively, where the effector molecule is a protein or polypeptide, linkage may be achieved using recombinant DNA procedures, for example as described in WO 86 / 01533 and EP 0392745.

[0277] The term effector molecule as used herein includes, for example, anti-neoplastic agents, drugs, toxins, biologically active proteins such as enzymes, other antibodies or antibody fragments, synthetic or naturally occurring polymers, nucleic acids and fragments thereof such as DNA, RNA and fragments thereof, radionuclides, particularly radioactive iodides, radioisotopes, chelated metals, nanoparticles and reporter groups such as fluorescent compounds or compounds that can be detected by NMR or ESR spectroscopy.

[0278] Examples of effector molecules may include cytotoxins or cytotoxic agents, including any agent that is detrimental to (e.g., kills) cells. Examples include combrestatins, dolastatins, epothilones, staurosporines, maytansinoids, spongiostatins, rhizoxins, halichondrins, roridin, hemiasterlin, taxol, cytochalasin B, gramicidin D, ethidium bromide, emetine, mitomycin, etoposide, tenoposide, vincristine, vinblastine, colchicine, doxorubicin, daunorubicin, dihydroxyanthracin dione, mitoxantrone, mithramycin, actinomycin D, 1-dehydrotestosterone, glucocorticoids, procaine, tetracaine, lidocaine, propranolol, and puromycin, as well as analogs or homologs thereof.

[0279] Effector molecules also include, but are not limited to, antimetabolites (e.g., methotrexate, 6-mercaptopurine, 6-thioguanine, cytarabine, 5-fluorouracil decarbazine), alkylating agents (e.g., mechlorethamine, thioepaclorambucil, melphalan, carmustine (BSNU) and lomustine (CCNU), cyclophosphamide, busulfan, dibromomannitol, streptozotocin, mitomycin C and cis-dichlorodiamineplatinum(II) (DDP) cisplatin), anthracyclines (e.g., daunorubicin (formerly daunomycin) and doxorubicin), antibiotics (e.g., dactinomycin (formerly actinomycin), bleomycin, mithramycin, anthramycin (AMC), calicheamicin or duocarmycin), and antimitotic agents (e.g., vincristine and vinblastine).

[0280] Other effector molecules include 111 In and 90 Y.,Lu 177 , Bismuth 213 , Californium 252 ,iridium 192 and tungsten 188 / rhenium 188or drugs such as, but not limited to, alkylphosphocholines, topoisomerase I inhibitors, taxoids and suramin.

[0281] Other effector molecules include proteins, peptides and enzymes. Enzymes of interest include, but are not limited to, proteolytic enzymes, hydrolases, lyases, isomerases, transferases. Proteins, polypeptides and peptides of interest include, but are not limited to, immunoglobulins, toxins such as abrin, ricin A, pseudomonas exotoxin or diphtheria toxin, proteins such as insulin, tumor necrosis factor, alpha-interferon, beta-interferon, nerve growth factor, platelet-derived growth factor or tissue plasminogen activator, thrombotic or antiangiogenic agents such as angiostatin or endostatin, or biological response modifiers such as lymphokines, interleukin-1 (IL-1), interleukin-2 (IL-2), granulocyte macrophage colony stimulating factor (GM-CSF), granulocyte colony stimulating factor (G-CSF), nerve growth factor (NGF) or other growth factors and immunoglobulins.

[0282] Other effector molecules can include, for example, detectable substances useful for diagnosis.Examples of detectable substances include various enzymes, prosthetic groups, fluorescent materials, luminescent materials, bioluminescent materials, radionuclides, positron-emitting metals (for use in positron emission tomography) and non-radioactive paramagnetic metal ions.For the metal ions that can be conjugated to antibodies for use as diagnostic agents, see generally U.S. Patent No. 4,741,900. Suitable enzymes include horseradish peroxidase, alkaline phosphatase, beta-galactosidase or acetylcholinesterase; suitable prosthetic groups include streptavidin, avidin and biotin; suitable fluorescent materials include umbelliferone, fluorescein, fluorescein isothiocyanate, rhodamine, dichlorotriazinylamine fluorescein, dansyl chloride and phycoerythrin; suitable luminescent materials include luminol; suitable bioluminescent materials include luciferase, luciferin and aequorin; suitable radionuclides include, 125 I, 131 I, 111 In and 99 Examples include Tc.

[0283] In another example, the effector molecule may increase the half-life of the antibody in vivo and / or reduce the immunogenicity of the antibody and / or enhance delivery of the antibody across an epithelial barrier to the immune system. Examples of suitable effector molecules of this type include polymers, albumin, albumin binding proteins, or albumin binding compounds such as those described in WO 05 / 117984.

[0284] When the effector molecule is a polymer, it may generally be a synthetic or naturally occurring polymer, such as an optionally substituted linear or branched polyalkylene, polyalkenylene or polyoxyalkylene polymer or a branched or unbranched polysaccharide, such as a homo- or heteropolysaccharide.

[0285] Particular optional substituents which may be present on the above synthetic polymers include one or more hydroxy, methyl or methoxy groups.

[0286] Specific examples of synthetic polymers include optionally substituted linear or branched poly(ethylene glycol), poly(propylene glycol) poly(vinyl alcohol) or derivatives thereof, in particular optionally substituted poly(ethylene glycol), such as methoxypoly(ethylene glycol) or derivatives thereof.

[0287] Exemplary naturally occurring polymers include lactose, amylose, dextran, glycogen or derivatives thereof.

[0288] "Derivatives" as used herein is intended to include reactive derivatives, e.g., thiol-selective reactive groups such as maleimides and the like. The reactive group may be linked to the polymer directly or via a linker segment. It will be understood that the residue of such a group will, in some instances, form part of the product as the linking group between the antibody fragment and the polymer.

[0289] The size of the polymer may vary if desired, but generally ranges from 500 Da to 50,000 Da, e.g., 5,000 to 40,000 Da, e.g., 20,000 to 40,000 Da, in average molecular weight. The polymer size may be selected based on, among other things, the intended use of the product, e.g., its ability to localize to a particular tissue, such as a tumor, or its ability to extend circulatory half-life (for a review, see Chapman, 2002, Advanced Drug Delivery Reviews, 54, 531-545). Thus, for example, if the product is intended to leave the circulation and penetrate tissues, it may be advantageous to use a low molecular weight polymer, e.g., having a molecular weight of about 5,000 Da. For applications in which the product will remain in the circulation, it may be advantageous to use a high molecular weight polymer, e.g., having a molecular weight in the range of 20,000 Da to 40,000 Da.

[0290] Suitable polymers include polyalkylene polymers, such as poly(ethylene glycol) or especially methoxypoly(ethylene glycol) or derivatives thereof, especially those having a molecular weight in the range of about 15,000 Da to about 40,000 Da.

[0291] In one example, the antibody for use in the present invention is conjugated to a poly(ethylene glycol) (PEG) moiety. In one particular example, the antibody is an antibody fragment, and the PEG molecule can be attached via any available amino acid side chain or terminal amino acid functional group located in the antibody fragment, such as any free amino, imino, thiol, hydroxyl or carboxyl group. Such amino acids can be naturally present in the antibody fragment or can be engineered into the fragment using recombinant DNA methods (see, for example, U.S. Pat. No. 5,219,996; U.S. Pat. No. 5,667,425; WO 98 / 25971). In one example, the antibody molecule of the present invention is a modified Fab fragment, where the modification is the addition of one or more amino acids to the C-terminus of its heavy chain to allow for the attachment of an effector molecule. Suitably, the additional amino acids form a modified hinge region containing one or more cysteine ​​residues to which the effector molecule can be attached. Multiple sites can be used to attach two or more PEG molecules.

[0292] Suitably, the PEG molecule may be covalently attached via a thiol group of at least one cysteine ​​residue located in the antibody fragment. Each polymer molecule attached to the modified antibody fragment may be covalently attached to the sulfur atom of a cysteine ​​residue located within the fragment. The covalent bond is generally a disulfide bond, or in particular a sulfur-carbon bond. When a thiol group is used as the attachment point, an appropriately activated effector molecule may be used, for example, a thiol-selective derivative such as maleimide and cysteine ​​derivatives. An activated polymer may be used as a starting material in preparing the above polymer-modified antibody fragment. An activated polymer may be any polymer containing a thiol-reactive group, such as an α-halo carboxylic acid or ester, for example, iodoacetamide, an imide, for example, maleimide, vinyl sulfone or disulfide. Such starting materials may be obtained commercially (e.g., from Nektar, formerly Shearwater Polymers Inc., Huntsville, AL, USA) or may be prepared from commercially available starting materials using conventional chemical procedures. Particular PEG molecules include 20K methoxy-PEG-amine (available from Nektar, formerly Shearwater; Rapp Polymere; and SunBio) and M-PEG-SPA (available from Nektar, formerly Shearwater).

[0293] In one embodiment, the antibody is PEGylated, i.e., is a modified Fab fragment or diFab to which PEG (poly(ethylene glycol)) has been covalently attached, e.g., according to the methods disclosed in EP 0948544 or EP 1090037 [see also “Poly(ethyleneglycol) Chemistry, Biotechnical and Biomedical Applications”, 1992, J. Milton Harris (ed), Plenum Press, New York, “Poly(ethyleneglycol) Chemistry and Biological Applications”, 1997, J. Milton Harris and S. Zalipsky (eds), American Chemical Society, Washington DC and “Bioconjugation Protein Coupling Techniques for the Biomedical Sciences”, 1998, M. Aslam and A. Dent, Grove Publishers, New York; Chapman, A. 2002, Advanced Drug Delivery Reviews 2002, 54:531-545]. In one example, PEG is attached to a cysteine ​​in the hinge region. In one example, the PEG-modified Fab fragment has a maleimide group covalently attached to a single thiol group in the modified hinge region. A lysine residue may be covalently attached to the maleimide group, and a methoxypoly(ethylene glycol) polymer having a molecular weight of about 20,000 Da may be attached to each amine group on the lysine residue. Thus, the total molecular weight of the PEG attached to the Fab fragment may be about 40,000 Da.

[0294] In one embodiment, the antibody is a modified Fab' fragment having at the C-terminus of its heavy chain a modified hinge region containing at least one cysteine ​​residue to which an effector molecule is attached. Suitably, the effector molecule is PEG and is attached using the methods described in WO 98 / 25971 and WO 2004072116 or WO 2007 / 003898. Effector molecules can be attached to the antibody fragment using the methods described in International Patent Applications WO 2005 / 003169, WO 2005 / 003170 and WO 2005 / 003171.

[0295] In one embodiment, the antibody is not conjugated to an effector molecule.

[0296] Polynucleotides and Vectors The present invention also provides isolated polynucleotides encoding antibodies or portions thereof according to the invention, such as those amino acid sequence numbers listed in Table 2. The term "isolated" is used throughout this specification to mean that the polynucleotide is present in a physical environment different from that in which it may occur in nature.

[0297] An isolated polynucleotide according to the present invention can include synthetic DNA produced, for example, by chemical processing, cDNA, genomic DNA, or any combination thereof. [Table 2]

[0298] Examples of suitable sequences are provided herein. Thus, in one embodiment, the invention provides an isolated polynucleotide encoding an antibody comprising the sequence set forth in SEQ ID NO: 16, 18, 20, 22, 24, 26, 28, 30, 32, 34, 36, 38, 40, 42, 44, 46, 48, 50, 52, 54, 56, 58, 60, 62, 64, 66, 68, 70, 72, 74, 76, 78, 80, 82, 84, 86, 88, 90, 92, 94, 96, 98, 100, 102.

[0299] In one embodiment, the invention provides an isolated polynucleotide encoding an antibody of the invention, comprising: i. at least 90% identical to SEQ ID NO: 20, 16, or 24; or ii. An isolated polynucleotide comprising or consisting of SEQ ID NO:20, 16 or 24 is provided.

[0300] In one embodiment, the invention provides an isolated polynucleotide encoding an antibody of the invention, comprising: i. at least 90% identical to SEQ ID NO: 94, 28, 34, 58, 70, 76, 82, or 88; or ii. An isolated polynucleotide comprising or consisting of SEQ ID NO: 94, 28, 34, 58, 70, 76, 82 or 88 is provided.

[0301] In one embodiment, the invention provides an isolated polynucleotide encoding an antibody of the invention, comprising: i. at least 90% identical to SEQ ID NO: 22, 18, or 26; or ii. An isolated polynucleotide comprising or consisting of SEQ ID NO: 22, 18 or 26 is provided.

[0302] In one embodiment, the present invention provides an isolated polynucleotide encoding an antibody of the present invention, comprising: i. at least 90% identical to SEQ ID NO: 96, 30, 36, 60, 72, 78, 84, or 90; or ii. An isolated polynucleotide comprising or consisting of SEQ ID NO: 96, 30, 36, 60, 72, 78, 84 or 90 is provided.

[0303] In one embodiment, the invention provides an isolated polynucleotide encoding the heavy chain of an IgG1 antibody of the invention comprising the sequence shown in SEQ ID NO: 96, 30, 36, 60, 72, 78, 84 or 90.

[0304] An isolated polynucleotide encoding the light chain of an IgG1 antibody of the invention comprising the sequence shown in SEQ ID NO: 22, 18 or 26 is also provided.

[0305] In a preferred embodiment, the invention provides isolated polynucleotides encoding the heavy and light chains of an IgG1 antibody of the invention, wherein the polynucleotide encoding the heavy chain comprises the sequence set forth in SEQ ID NO: 96 and the polynucleotide encoding the light chain comprises the sequence set forth in SEQ ID NO: 22.

[0306] The present invention also provides a cloning or expression vector comprising one or more of the polynucleotides described herein. In one example, a cloning or expression vector according to the invention comprises one or more of the isolated polynucleotides described above.

[0307] Standard techniques of molecular biology may be used to prepare DNA sequences encoding the antibodies of the present invention. The desired DNA sequence may be fully or partially synthesized using oligonucleotide synthesis techniques. Site-directed mutagenesis and polymerase chain reaction (PCR) techniques may be used as needed.

[0308] General methods for constructing vectors, transfection methods and culture methods are well known to those skilled in the art. In this regard, see "Current Protocols in Molecular Biology", 1999, FMAusubel (ed), Wiley Interscience, New York and the Maniatis Manual produced by Cold Spring Harbor Publishing.

[0309] Host cells for producing antibodies Also provided is a host cell comprising one or more cloning or expression vectors comprising one or more isolated polynucleotide sequences according to the invention or one or more isolated polynucleotide sequences encoding the antibody of the invention. Any suitable host cell / vector system may be used for the expression of the polynucleotide sequence encoding the antibody of the invention. Bacterial, such as E. coli, and other microbial systems may be used, or eukaryotic, such as mammalian, host cell expression systems may be used. Suitable mammalian host cells include CHO cells, myeloma cells, or hybridoma cells.

[0310] In further embodiments, a host cell comprising such a nucleic acid or vector is provided. In one such embodiment, the host cell comprises (e.g., is transformed by) (1) a vector comprising a nucleic acid encoding an amino acid sequence comprising the VL of an anti-HLA-G antibody and an amino acid sequence comprising the VH of an anti-HLA-G antibody, or (2) a first vector comprising a nucleic acid encoding an amino acid sequence comprising the VL of an anti-HLA-G antibody and a second vector comprising a nucleic acid encoding an amino acid sequence comprising the VH of an anti-HLA-G antibody. In one embodiment, the host cell is a eukaryotic cell, such as a Chinese Hamster Ovary (CHO) cell or a lymphoid cell (e.g., Y0, NS0, Sp20 cell). In one embodiment, the host cell is a prokaryotic cell, such as an E. coli cell. In one embodiment, a method for producing an anti-HLA-G antibody is provided, comprising culturing a host cell comprising a nucleic acid encoding the antibody under conditions suitable for the expression of the antibody, and optionally recovering the antibody from the host cell (or host cell culture medium).

[0311] Suitable host cells for cloning or expressing antibody-encoding vectors include prokaryotic or eukaryotic cells as described herein. For example, antibodies can be produced in bacteria, particularly when glycosylation and Fc effector functions are not required. For the expression of antibody fragments and polypeptides in bacteria, see, for example, U.S. Patent No. 5,648,237, U.S. Patent No. 5,789,199 and U.S. Patent No. 5,840,523. (See also Charlton, Methods in Molecular Biology, Vol. 248 (BKC Lo, ed., Humana Press, Totowa, NJ, 2003), pp. 245-254, which describes the expression of antibody fragments in E. coli). After expression, antibodies can be isolated from bacterial cell paste in a soluble fraction and further purified.

[0312] In addition to prokaryotes, eukaryotic microbes such as filamentous fungi or yeast, including "humanized" fungal and yeast strains whose glycosylation pathways result in the production of antibodies with partial or fully human glycosylation patterns, are suitable cloning or expression hosts for antibody-encoding vectors. See Gerngross, Nat. Biotech. 22:1409-1414 (2004), and Li et al., Nat. Biotech. 24:210-215 (2006).

[0313] Suitable types of Chinese Hamster Ovary (CHO cells) for use in the present invention may include CHO cells and CHO-K1 cells, including dhfr-CHO cells, such as CHO-DG44 cells and CHO-DXB11 cells, which may be used with a DHFR selectable marker, or CHOK1-SV cells, which may be used with a glutamine synthetase selectable marker. Other cell types used for expression of antibodies include lymphocytic cell lines, such as NS0 myeloma cells and SP2 cells, COS cells. Host cells may be stably transformed or transfected with the isolated polynucleotide sequences or expression vectors according to the present invention.

[0314] In one embodiment, the antibody of the present invention is produced in a host cell that is genetically modified to reduce or disable the function of alpha 1,6 fucosyltransferase (FUT8). In one embodiment, the genetically modified cell is a CHO cell, and the FUT8 gene is knocked out (KO FUT8). In one embodiment, the host cell for producing the antibody of the present invention is a CHO-DG44. In one embodiment, the host cell for producing the antibody of the present invention is a KO FUT8 CHOSXE / DG44 cell, which can be produced according to the method described in the examples provided herein.

[0315] The antibody production process The present invention also provides a process for the production of an antibody according to the invention, comprising culturing a host cell according to the invention under conditions suitable for producing an antibody according to the invention, and isolating the antibody.

[0316] The invention also provides a process for the production of a pharmaceutical composition comprising an antibody according to the invention, comprising culturing a host cell according to the invention under conditions suitable for producing an antibody according to the invention, isolating the antibody, and formulating the antibody into a pharmaceutical composition.

[0317] The antibody may contain only heavy or light chain polypeptides, in which case it is necessary to use only heavy or light chain polypeptide coding sequences to transfect the host cell. For the production of an antibody containing both heavy and light chains, two vectors may be transfected into the cell line, the first vector encoding the light chain polypeptide, and the second vector encoding the heavy chain polypeptide. Alternatively, a single vector may be used that contains sequences encoding light and heavy chain polypeptides.

[0318] Thus, a process is provided for culturing a host cell, expressing the antibody, isolating the antibody, and optionally purifying the antibody to provide an isolated antibody, in one embodiment, the process further comprises conjugating an effector molecule to the isolated antibody.

[0319] The invention also provides a process for the production of an antibody according to the invention, which comprises culturing a host cell containing a vector of the invention under conditions suitable to result in expression of protein from DNA encoding the antibody molecule of the invention, and isolating the antibody molecule.

[0320] The antibodies according to the invention are expressed at good levels from the host cells and therefore the properties of the antibodies are likely to be optimized for commercial processing.

[0321] In one embodiment, antibodies, e.g. humanized antibodies, particularly antibodies according to the invention, are provided in a substantially purified form, particularly purified antibodies free or substantially free of endotoxins and / or host cell proteins or DNA.

[0322] Substantially free of endotoxin is generally intended to refer to an endotoxin content of less than or equal to 1 EU per mg of antibody product, for example 0.5 or 0.1 EU per mg of product.

[0323] Substantially free of host cell proteins or DNA is generally intended to refer to a host cell protein and / or DNA content of less than 400 μg per mg of antibody product, for example less than 100 μg / mg, in particular less than 20 μg / mg, as appropriate.

[0324] Pharmaceutical Compositions, Dosages and Dosage Regimens The antibody of the present invention may be provided in a pharmaceutical or diagnostic composition. Thus, the present invention also provides a pharmaceutical or diagnostic composition comprising an antibody according to the present invention in combination with one or more pharma- ceutically acceptable carriers, excipients or diluents.

[0325] Preferably, the pharmaceutical or diagnostic composition comprises an antibody that specifically binds to HLA-G, the antibody comprising ai CDR-L1 comprising SEQ ID NO:1, ii. A CDR-L2 comprising SEQ ID NO:2; and iii. CDR-L3 comprising SEQ ID NO:3 and a light chain variable region comprising bi CDR-H1 comprising SEQ ID NO:4, ii. CDR-H2 comprising SEQ ID NO:5; and iii. CDR-H3 comprising SEQ ID NO:6 and a heavy chain variable region comprising:

[0326] In one embodiment, the antibody according to the invention is the only active ingredient. In another embodiment, the antibody according to the invention is combined with one or more additional active ingredients. In one embodiment, the antibody according to the invention is combined with an antibody against CD47. Thus, in one embodiment, an antibody that specifically binds to HLA-G, ai CDR-L1 comprising SEQ ID NO:1, ii. A CDR-L2 comprising SEQ ID NO:2; and iii. CDR-L3 comprising SEQ ID NO:3 and a light chain variable region comprising bi CDR-H1 comprising SEQ ID NO:4, ii. CDR-H2 comprising SEQ ID NO:5; and iii. CDR-H3 comprising SEQ ID NO:6 and a heavy chain variable region comprising Antibodies are provided that are combined with a second antibody that binds CD47.

[0327] Alternatively, the pharmaceutical compositions may contain an antibody according to the invention, which is the only active ingredient, and be administered separately to a patient in combination (e.g. simultaneously, sequentially or separately) with other therapeutic, diagnostic or palliative agents.

[0328] The pharmaceutical composition according to the present invention can be suitably administered to a patient to identify the therapeutically effective amount required. The term "therapeutically effective amount" as used herein refers to the amount of therapeutic agent required to treat, improve or prevent the targeted disease or condition, or to show a detectable therapeutic or prophylactic effect. For any antibody, the therapeutically effective amount can be first estimated either in cell culture assays or animal models, usually rodents, rabbits, dogs, pigs or primates. Animal models can be used to determine the appropriate concentration range and route of administration. Such information can then be used to determine the useful dose and route of administration in humans.

[0329] The exact therapeutically effective amount for a human subject depends on the severity of the disease state, the general health of the subject, the age, weight and sex of the subject, diet, time and frequency of administration, drug combinations, reaction sensitivities, and tolerance / response to treatment. In general, the therapeutically effective amount is 0.01 mg / kg to 500 mg / kg, e.g., 0.1 mg / kg to 200 mg / kg, e.g., 100 mg / kg. Pharmaceutical compositions can be conveniently provided in unit dose forms containing a predetermined amount of the active agent of the present invention per dose.

[0330] Pharmaceutically acceptable carriers in therapeutic compositions may further contain liquids, such as water, saline, glycerol, and ethanol. Additionally, auxiliary substances, such as wetting or emulsifying agents or pH buffering substances, may be present in such compositions.

[0331] Forms suitable for administration include those suitable for parenteral administration, for example, by injection or infusion, for example, by bolus injection or continuous infusion, in intravenous, inhalable or subcutaneous form. If the product is for injection or infusion, it may take the form of a suspension, solution or emulsion in an oily or aqueous vehicle, and may contain formulatory agents such as suspending agents, preservatives, stabilizers and / or dispersing agents. Alternatively, the antibody according to the invention may be in dry form, for reconstitution with an appropriate sterile liquid before use. Solid forms suitable for dissolution or suspension in liquid vehicles before injection may also be prepared.

[0332] Once formulated, the compositions of the invention can be administered directly to a subject.Thus, there is provided herein the use of an antibody according to the invention for the manufacture of a medicament.

[0333] Preferably, the pharmaceutical compositions according to the invention are adapted for administration to human subjects.

[0334] Thus, in another aspect, the present invention provides an antibody that specifically binds to HLA-G, or a pharmaceutical composition comprising said antibody, for use in therapy, said antibody comprising: ai CDR-L1 comprising SEQ ID NO:1, ii. A CDR-L2 comprising SEQ ID NO:2; and iii. CDR-L3 comprising SEQ ID NO:3 and a light chain variable region comprising bi CDR-H1 comprising SEQ ID NO:4, ii. CDR-H2 comprising SEQ ID NO:5; and iii. CDR-H3 comprising SEQ ID NO:6 and a heavy chain variable region comprising:

[0335] Treatment indications As used herein, the terms "treatment", "treating" and the like refer to obtaining a desired pharmacological and / or physiological effect. The effect may be preventative in that it completely or partially prevents the disease or its symptoms, and / or it may be therapeutic in that it partially or completely cures the disease and / or the adverse effects caused by the disease. Thus, treatment encompasses any treatment of a mammalian, particularly a human, disease, including (a) preventing the disease from occurring in a subject who may have a predisposition to the disease but has not yet been diagnosed as having it, (b) inhibiting the disease, i.e., arresting its development, and (c) relieving the disease, i.e., causing regression of the disease.

[0336] "Therapeutically effective amount" refers to an amount of HLA-G antibody that, when administered to a mammal or other subject for treating a disease, is sufficient to effect such treatment of the disease. The therapeutically effective amount varies depending on the anti-HLA-G antibody, the disease and its severity, and the age, weight, etc., of the subject to be treated.

[0337] The antibody of the present invention, its preparation or pharmaceutical composition can be administered for preventive and / or therapeutic treatment.In preventive application, the antibody, preparation or composition is administered to a subject at risk of a disorder or condition as described herein in an amount sufficient to prevent or reduce the subsequent effects of the condition, or one or more of its symptoms.In therapeutic application, the antibody is administered to a subject already suffering from a disorder or condition as described herein in an amount sufficient to cure, alleviate or partially stop the condition, or one or more of its symptoms.Such therapeutic treatment can result in a reduction in the severity of disease symptoms, or an increase in the frequency or duration of symptom-free periods.

[0338] The present invention provides a method of treating a disorder or condition described herein in a subject in need thereof, comprising administering to the subject an antibody or pharmaceutical composition according to the invention, wherein such antibody is administered in a therapeutically effective amount.

[0339] The present invention also provides an antibody or a pharmaceutical composition of the invention for use in therapy, in particular for use in the treatment of a disorder or condition described herein.

[0340] The present invention also provides the use of an antibody or a pharmaceutical composition of the invention for the manufacture of a medicament, in particular for use in the treatment of a disorder or condition as described herein.

[0341] The antibodies of the invention may be used to treat, prevent or ameliorate any condition associated with HLA-G activity, for example any condition that results in whole or in part from signaling through the HLA-G receptor.

[0342] HLA-G associated diseases or disorders include, inter alia, cancer (or tumors), infections and autoimmune disorders.

[0343] In one embodiment, the invention provides an antibody or a pharmaceutical composition of the invention for use in the treatment of a disease characterized by overexpression of HLA-G.

[0344] The antibody of the present invention may be particularly useful for treating or preventing cancer, including cancer characterized by overexpression of HLA-G. Thus, in one embodiment, the present invention provides an antibody or pharmaceutical composition of the present invention for use in treating cancer. In one embodiment, the present invention provides an antibody or pharmaceutical composition of the present invention for use in treating cancer characterized by overexpression of HLA-G.

[0345] In the context of the present invention, cancer includes, for example, renal clear cell carcinoma (RCC), colorectal cancer (CRC), pancreatic cancer, ovarian cancer, breast cancer, head and neck cancer, gastric cancer, hepatocellular carcinoma, lung cancer, neuroblastoma and blood cancer. The antibodies of the present invention may be useful for treating or preventing liquid cancers, such as blood cancers.

[0346] The antibody of the present invention may be particularly useful for treating or preventing solid tumors. Thus, in one embodiment, the present invention provides an antibody or pharmaceutical composition of the present invention for use in treating a solid tumor. In one embodiment, the present invention provides an antibody or pharmaceutical composition of the present invention for use in treating a solid tumor characterized by overexpression of HLA-G. In one embodiment, the present invention provides an antibody or pharmaceutical composition of the present invention for use in treating renal clear cell carcinoma (RCC), colorectal cancer (CRC), pancreatic cancer, ovarian cancer, head and neck cancer, gastric cancer or hepatocellular carcinoma. In one particular embodiment, the solid tumor is renal clear cell carcinoma (RCC). In another particular embodiment, the solid tumor is colorectal carcinoma (CRC).

[0347] In one embodiment, the present invention provides the use of an antibody or pharmaceutical composition of the present invention for the manufacture of a medicament for use in the treatment of a solid tumor. In one embodiment, the present invention provides the use of an antibody or pharmaceutical composition of the present invention for the manufacture of a medicament for use in the treatment of a solid tumor characterized by overexpression of HLA-G. In one embodiment, the present invention provides the use of an antibody or pharmaceutical composition of the present invention for the manufacture of a medicament for use in the treatment of renal clear cell carcinoma (RCC), colorectal cancer (CRC), pancreatic cancer, ovarian cancer, head and neck cancer, gastric cancer or hepatocellular carcinoma. In one particular embodiment, the solid tumor is renal clear cell carcinoma (RCC). In another particular embodiment, the solid tumor is colorectal carcinoma (CRC).

[0348] In one embodiment, the invention provides a method of treating a solid tumor in a patient, comprising administering to said patient a therapeutically effective amount of an antibody or pharmaceutical composition of the invention.In one embodiment, the invention provides a method of treating a solid tumor in a patient, characterized by overexpression of HLA-G, comprising administering to said patient a therapeutically effective amount of an antibody or pharmaceutical composition of the invention.

[0349] In one embodiment, the solid tumor is selected from renal cell carcinoma (RCC), colorectal carcinoma (CRC), pancreatic cancer, ovarian cancer, head and neck cancer, gastric cancer and hepatocellular carcinoma.In one particular embodiment, the solid tumor is renal cell carcinoma (RCC).In another particular embodiment, the solid tumor is colorectal carcinoma (CRC).

[0350] The invention also provides the use of the antibodies of the invention as diagnostically active agents or in diagnostic assays, for example to diagnose a disease or its severity.

[0351] In one embodiment, the present invention provides a method for diagnosing a solid tumor by using an antibody or a pharmaceutical composition according to the present invention.In one embodiment, the present invention provides a method for diagnosing a solid tumor characterized by overexpression of HLA-G by using an antibody or a pharmaceutical composition according to the present invention.

[0352] In one embodiment, the present invention provides a method for diagnosing renal clear cell carcinoma (RCC), colorectal carcinoma (CRC), pancreatic cancer, ovarian cancer, head and neck cancer, gastric cancer or hepatocellular carcinoma by using the antibody or pharmaceutical composition according to the present invention. In a particular embodiment, the solid tumor is renal clear cell carcinoma (RCC). In another particular embodiment, the solid tumor is colorectal carcinoma (CRC).

[0353] Diagnosis may preferably be performed on biological samples. "Biological samples" encompass a variety of sample types obtained from an individual and may be used in diagnostic or monitoring assays. The definition encompasses cerebrospinal fluid, blood, such as plasma and serum, as well as other liquid samples of biological origin, such as urine and saliva, solid tissue samples, such as biopsy specimens or tissue cultures, or cells derived therefrom, and their progeny. This definition also includes samples that have been manipulated in some way after their procurement, for example, by treatment with reagents, solubilization, or enrichment of certain components, such as polynucleotides.

[0354] Diagnostic tests can be preferably performed on biological samples that are not in contact with the human or animal body. Such diagnostic tests are also called in vitro tests. In vitro diagnostic tests can rely on in vitro methods to detect HLA-G in biological samples obtained from subjects.

[0355] In one embodiment, the present invention provides a method for diagnosing a solid tumor expressing HLA-G in a biological sample by using an antibody or pharmaceutical composition according to the present invention.In one embodiment, the present invention provides a method for diagnosing renal clear cell carcinoma (RCC), colorectal cancer (CRC), pancreatic cancer, ovarian cancer, head and neck cancer, gastric cancer or hepatocellular carcinoma in a biological sample by using an antibody or pharmaceutical composition according to the present invention.

[0356] example Example 1: Generation of HLA-G, HLA-I and ILT2, ILT4 proteins for use in screening assays 1.1. HLA-G Protein: Sequence Sequences of the most abundant isoforms of HLA-G, HLA-G1 (membrane-bound) or HLA-G5 (soluble), containing the alpha 1, 2 and 3 domains of HLA-G, were used to generate HLA-G constructs for screening antibodies against HLA-G.

[0357] Based on the crystal structure analysis, the sequence of the HLA-G extracellular domain or ECD was defined (alpha 1 domain (italics)-alpha 2 domain-alpha 3 domain (underlined)-final residues KQ as determined by crystallography). The 20 residues specific for HLA-G are in bold:

number

[0358] ·Soluble HLA-G (HLA-G ECD): Proteins were expressed with an AVItev10HisTag (signal peptide (bold), avidin affinity tag or AVI tag for biotinylation (underlined), Tev protease site (underlined and italicized), 10His Tag (italicized)):

number

[0359] The final purified protein sequence used in the screening assay contains the following sequence (including the AVItev10His Tag):

number

[0360] HLA-G DNA sequence for cell membrane-bound expression:

number

[0361] The corresponding membrane-associated protein contains the following sequence (signal peptide cleaved after expression):

number

number

[0362] "HLA-G Null 1,2,3" "HLA-G Null 1,2,3" corresponds to HLA-G in which amino acids specifically expressed on HLA-G α1, α2 and α3 have been replaced by consensus amino acids expressed on other HLA-I (20 mutated amino acids (bold in the sequence below)).

[0363] HLA-I consensus amino acids were obtained from sequence information obtained from the Immuno Polymorphism Database at EBI. HLA-I full-length proteins were analyzed and residue profile plots were generated for each domain (alpha 1-3) across the entire HLA-I set (20 in total) which allowed for the identification of HLA-G specific residues.

[0364] The sequence information was used to generate allelic consensus sequences for each HLA protein in which positions in the canonical sequence were replaced by the most common residue found across all alleles.

[0365] To obtain the HLA-G Null 1,2,3 sequence, the HLA-G specific residues were changed to consensus residues found in other HLA molecules at 20 specific positions.

[0366] Soluble protein (HLA-G Null 1,2,3 ECD) was expressed with an AVItev10HisTag (signal peptide (bold), AVI tag (underlined), Tev protease site (underlined and italicized), 10His Tag (italicized)):

number

[0367] The peptide signal was cleaved after expression and the final purified protein sequence used in the screening assay contained the following sequence:

number

[0368] "HLA-G Null 1,2,3" DNA sequence for cell membrane-bound expression:

number

[0369] The corresponding membrane-associated protein contains the following sequence (signal peptide cleaved after expression):

number

number

[0370] "HLA-G Null 1,3" "HLA-G Null 1,3" corresponds to HLA-G in which amino acids specifically expressed on HLA-G α1 and α3 have been replaced by consensus amino acids expressed on other HLA-I (mutated amino acids (bold in sequence SEQ ID NO: 117 below)).

[0371] The protein was expressed on the surface of the cell. The DNA sequence for cell membrane-bound expression includes:

number

[0372] The corresponding membrane-associated protein contains the following sequence (signal peptide cleaved after expression):

number

number

[0373] "HLA-G Null 3" "HLA-G Null3" corresponds to HLA-G in which the amino acids specifically expressed on HLA-G α3 are replaced by consensus amino acids expressed on other HLA-I (the five mutated amino acids (bold in the sequence of SEQ ID NO: 119 below)).

[0374] The protein was expressed on the surface of the cell. The DNA sequence for cell membrane-bound expression includes:

number

[0375] The corresponding membrane-associated protein contains the following sequence (signal peptide cleaved after expression):

number

number

[0376] "HLA-G Null3 2AA" "HLA-G Null3 2AA" corresponds to HLA-G in which only two amino acids (F195, Y197 of HLA-G) that are specifically expressed on HLA-G α3 and reported to interact with ILT2 / ILT4 have been replaced by consensus amino acids (mutated amino acids (bold)) expressed on other HLA-I.

[0377] The protein was expressed on the surface of the cell. The DNA sequence for cell membrane-bound expression includes:

number

[0378] The corresponding membrane-associated protein contains the following sequence (signal peptide cleaved after expression):

number

number

[0379] ·Isolated HLA-G alpha 3 domain (wild type) Soluble protein was expressed as a Tev-human Fc fusion protein (signal peptide (bold), Tev (underlined), Fc fragment (italic)) and not co-expressed with B2m ("B2m free isolated HLA-G alpha 3 domain"):

number

[0380] The peptide signal was cleaved after expression and the final purified protein sequence used in the screening assay contained the following sequence (the Fc tag was cleaved):

number

[0381] DNA sequence for cell membrane-bound expression:

number

[0382] The corresponding membrane-associated protein contains the following sequence (signal peptide, transmembrane and cytoplasmic domains cleaved after expression (italics)):

number

[0383] Isolated HLA-G alpha 3 domain, null (5 amino acids mutated (bold)) The protein was expressed as a 10HistevAVI tagged protein (tag and GS linker (italics), signal peptide (bold)).

number

[0384] The peptide signal was cleaved after expression and the 10His Tag was removed during purification. The final purified protein sequence used in the screening assay contains the following sequence (N-terminal AVI Tag and GS linker (italics)):

number

[0385] B2m All HLA-G constructs described above used in the screening assay (except B2m free isolated HLA-G alpha 3 domain, wild type) were co-expressed with B2m. For expression of soluble HLA-G constructs, B2m was co-expressed with the following sequence (signal peptide (bold)):

number

[0386] The signal peptide was cleaved after expression and the final purified protein sequence used in the screening assay contained the following sequence:

number

[0387] The entire membrane-bound HLA-G construct (containing the cell-expressed isolated alpha 3 domain) was co-expressed with B2m. The DNA sequence used for transfection is as follows:

number

[0388] The corresponding B2m complexed with the membrane-bound HLA-G construct contains the sequence of SEQ ID NO: 129 (signal peptide MSRSVALAVLALLSLSGLEA, cleaved after expression).

[0389] Methods for the generation and purification of HLA-G constructs expressed as soluble proteins Protein expression and purification of isolated HLA-G α3 domain, wild type (TevHumanFc or TevHFc) HLA-G α3 TevHFc was co-expressed with β2m using the Expi293™ Expression System (Life technologies™) according to the manufacturer's protocol. Cells were harvested 5 days after transfection and the supernatant was used immediately for purification.

[0390] The supernatant containing HLA-G α3 TevHFc+β2m protein was applied to a Hitrap Protein A column. Unbound proteins and contaminants were washed with PBS, HLA-G α3 TevHFc+β2m protein was eluted with 0.1M citrate buffer (pH 2), and the peak fraction was neutralized with 0.5ml of 2M Tris PH 8. Fractions containing purified HLA-G α3 TevHFc+β2m protein were pooled and the HFc tag was removed by incubating the protein with tev protease at a ratio of 1:100 for 2 hours at room temperature and 2 hours at 4°C. The protein was concentrated and further purified by size exclusion chromatography on a S75 26 / 60 column equilibrated with PBS buffer. Fractions containing purified HLA-G α3 protein were pooled, concentrated, and aliquots were stored at -80C until required.

[0391] Protein expression and purification of isolated HLA-G alpha 3 domain, null 10histevAVI HLA-G α3 null was co-expressed with β2m using the Expi293™ Expression System (Life technologies™) according to the manufacturer's protocol. Cells were harvested 5 days after transfection and the supernatant was used immediately for purification. The supernatant containing 10histevAVI HLAG α3 null+B2m protein was applied to a HisTrap Excel column (GE Healthcare) using an Akta Purifier (GE Healthcare). Unbound protein and contaminants were washed with Cytiva HyClon™ Phosphate Buffered Saline (PBS), 500 mM NaCl, pH 7.5. 10 mM Imidazole and protein was eluted with Cytiva HyClon™ Phosphate Buffered Saline (PBS), 500 mM NaCl, pH 7.5, 500 mM Imidazole. Fractions containing purified 10histevAVI HLAG α3 null+B2m protein were pooled and the 10his tag was removed by incubating the protein with tev protease at a ratio of 1:100 for 2 hours at room temperature and 2 hours at 4° C. The protein was concentrated and further purified by size exclusion chromatography on a S75 26 / 60 equilibrated with Cytiva HyClon™ Phosphate Buffered Saline (PBS). Fractions containing purified AVI HLAG α3 null+B2m protein were pooled, concentrated, and aliquots were stored at −80° C. until needed.

[0392] Protein expression and purification of HLA-G ECD (wild type or null mutant) co-expressed with B2m HLA-G ECD (WT or null mutant) was co-expressed with β2m using the CHO-SXE expression system according to the manufacturer's protocol. Briefly, cells were incubated at 8% CO in serum-free CD CHO medium (Gibco) supplemented with Gibco® GlutaMAX™ (1:1000). 2 Culture 6 x 10 CHO-SXE cells in a 37 °C shaking incubator containing6 The cells were grown to a cell density of 1000μg / mL. The cells were then centrifuged at 1500 rpm and resuspended in fresh ExpiCHO™ Expression Medium (Gibco). The cells were transfected with 1 mg / L of DNA at a 1:1 ratio of HLA-G ECD and β2m. Transfection was performed using the ExpiFectamine™ CHO Transfection Kit and OptiPRO™ SFM. Conditioned medium containing secreted proteins was collected 96 hours after transfection. The filtered cell culture supernatant was loaded onto a 5 ml HisTrap Excel column (GE Healthcare) using an Akta Purifier (GE Healthcare). The column was washed with Cytiva HyClone™ Phosphate Buffered Saline (PBS), 500 mM NaCl (pH 7.5), and the protein was eluted with the same buffer containing 500 mM imidazole. Fractions containing protein were analyzed by SDS-PAGE using NuPAGE 4-20% Tris-Glycine (Thermo) stained with Quick Coomassie Stain (VWR) and NuPAGE MES SDS Running Buffer (Thermo). Pure fractions were pooled and then concentrated using an Amicon® Ultra-15 Centrifugal Filter Unit (Millipore). Protein was then further purified using a Superdex 200 16 / 600 column (GE Healthcare) using Cytiva HyClone™ Phosphate Buffered Saline (PBS) as the running buffer.

[0393] Protein purity was assessed by analytical size-exclusion HPLC and sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). Proteins were greater than 97% pure (generally at least 99%). Proteins were also analyzed by liquid chromatography-mass spectrometry (LC-MS) to confirm that the sequence molecular weight (MW) was as predicted.

[0394] Methods for producing cells expressing HLA-G constructs on their cell surface Transient expression of HLA-G with B2m (including null construct) on the surface of ExpiHEK293 Co-transfection of Expi293™ suspension cells with HLA-G and B2m expression vectors at a 1:1 ratio was achieved using ExpiFectamine™ 293 transfection reagent (ThermoFisher Scientific), resulting in expression of cell surface proteins from 24 hours onwards.

[0395] Transient expression of HLA-G with B2m (including null construct) on the surface of CHO (PHAGE panning) Proprietary CHO-SXE cells were co-transfected with the B2m expression vector and HLA-G or HLA-G Null 1,2,3 in a 1:1 ratio using the ExpiFectamine™ CHO Transfection Kit (Gibco) according to the manufacturer's recommendations. Cells expressing HLA-G on their surface were harvested 48 hours later.

[0396] HLA-G expressing HCT116 cells The day before transfection, HCT116 cells (ATCC CCL-247) were cultured at 37 °C for 2 h in 20 ml of complete RPMI growth medium at 75 °C. 2 4 x 10 per flask 6 Seed the cells and incubate for 24 hours at 37°C, 5% CO 2The cells were incubated at 4°C for 30 min at 37°C. On the day of transfection, the growth medium was removed and replaced with 16 ml of complete growth medium. For each flask of cells to be transfected, 20 μg of HLA-G and β2m plasmids (1:1 ratio) were diluted in 4 ml of Opti-MEM® I Reduced Serum Media without serum. 80 μl of Lipofectamine LTX® Reagent was added to the above diluted Opti-MEM® DNA solution and incubated at room temperature for 30 min. After incubation, the DNA-Lipofectamine LTX® Reagent complex was added directly to each flask containing cells and the flasks were incubated at 37°C for 30 min ... 2 Placed in incubator for 22±2 hours.

[0397] 1.2.HLA-I Constructs HLA-I constructs were co-expressed with β2m using the Expi293™ Expression System (Life technologies) according to the manufacturer's protocol.

[0398] The HLA-I consensus sequences were derived from the publicly reported amino acid sequences of HLA-I alleles. The relevant sequence information carried forward for analysis was obtained from the Immuno Polymorphism Database at EBI. This information was used to generate allelic consensus sequences for each HLA protein in which positions in the canonical sequence were replaced by the most common residue found across all alleles.

[0399] The DNA sequences encoding the consensus amino acid sequences and used for transfection / cell membrane expression are listed below. The DNA sequence encoding B2m was as described above. The signal peptide in the secreted HLA-I protein was removed after expression.

[0400] HLA-A DNA sequences used for transfection / cell membrane expression:

number

[0401] The corresponding membrane-associated protein contains the following sequence:

number

[0402] HLA-B DNA sequences used for transfection / cell membrane expression:

number

[0403] The corresponding membrane-associated protein contains the following sequence:

number

[0404] HLA-C DNA sequences used for transfection / cell membrane expression:

number

[0405] The corresponding membrane-associated protein contains the following sequence:

number

[0406] HLA-E DNA sequences used for transfection / cell membrane expression:

number

[0407] The corresponding membrane-associated protein contains the following sequence:

number

[0408] HLA-F DNA sequences used for transfection / cell membrane expression:

number

[0409] The corresponding membrane-associated protein contains the following sequence:

number

[0410] 1.3.ILT2 / ILT4 construct (Fc fusion): ILT2 was expressed as a soluble ILT2 ECD-rabbit Fc fusion protein (signal peptide (bold), rabbit Fc (underlined)):

number

[0411] The final purified protein sequence used in the screening assay comprises the following sequence:

number

[0412] ILT4 was expressed as a soluble ILT4 ECD-rabbit Fc fusion protein (signal peptide (bold), rabbit Fc (underlined)):

number

[0413] The final purified protein sequence used in the screening assay comprises the following sequence:

number

[0414] Expression and purification of ILT2rbFc and ILT4rbFc Proteins were expressed by transient transfection using the Expi293™ HEK Expression System (Life technologies™) according to the manufacturer's protocol. Cells were harvested 5 days after transfection and the supernatant was used immediately for purification.

[0415] The supernatant containing ILT2rbFc or ILT4rbFc protein was applied to a Hitrap Protein A column. Unbound proteins and contaminants were washed away with PBS, and ILT2rbFc or ILT4rbFc protein was eluted with endo free 0.1M citrate buffer, pH 2, and the peak fraction was neutralized with 0.5ml of 2M Tris PH8. Fractions containing purified protein were pooled, concentrated, and further purified by size exclusion chromatography on a S200 26 / 60 using Cytiva HyClone™ Phosphate Buffered Saline (PBS) as running buffer. Fractions containing purified ILT2rbFc or ILT4rbFc protein were pooled, concentrated, aliquoted, and stored at -80°C.

[0416] Example 2: Generation of antibodies by immunization with HLA-G Due to the particular challenges associated with the generation of antibodies against HLA-G (e.g., high homology with other HLA-I molecules, identification of antibodies capable of blocking the interaction between HLA-G and its inhibitory receptors), and to identify antibodies that would be useful therapeutically, special discovery strategies, including special screening and testing strategies, had to be developed, as described below.

[0417] Immunization and screening strategies Syngeneic cells expressing different forms of HLA-G, with or without B2m co-expression, were used to immunize several animals across different species (including mice and rabbits). After 3-5 injections, animals were sacrificed and PBMCs, spleens, bone marrow and lymph nodes were harvested. Serum was monitored for binding to the immunogens used.

[0418] Memory B cell cultures were prepared and supernatants were first screened for the ability to bind HLA-G above irrelevant controls in a multiplexed no-wash assay using either the TTP Labtech Mirrorball system (plate reader) or Intellicyt iQue (flow cytometry). Cultures were screened against an irrelevant control protein, an in-house generated HLA-G protein, and / or using EXPI 293 HEKs transiently expressing the construct of interest on the cell surface (HLA-G, HLA-G Null 1,2,3, HLA-G Null3). Protein reagents were biotinylated to allow streptavidin capture to beads. Fluorescently labeled species-specific anti-Fc secondary antibodies were used in the assay to detect the test antibodies.

[0419] Approximately 3800 HLA-G specific positive hits were identified in the primary screen from a total of 18 B cell culture experiments, each containing 100-300 plates. Positive supernatants from the primary screen were then progressed for further characterization in binding assays (soluble isolated alpha 3 domain null, and cell-expressed isolated alpha 3 domain, HLA-G Null 1,3, HLA-G Null3 2AA).

[0420] Using the fluorescent foci method and binding to HLA-G ECD protein, wells with the desired profile were progressed for variable V-region retrieval.

[0421] In parallel, plasma cells from bone marrow and lymph nodes were also directly screened for their ability to bind human HLA-G or specifically bind to the α3 domain using fluorescent focusing. Here, B cells secreting HLA-G specific antibodies were sorted on biotinylated human HLA-G ECD or isolated wild-type HLA-G α3 domain immobilized on streptavidin beads. A goat anti-species Fc-FITC conjugate reveal reagent was used. Approximately 1700 direct foci were selected.

[0422] Following reverse transcription (RT) and PCR of selected cells, "transcriptionally active PCR" (TAP) products encoding the V-regions of the antibodies were generated and used to transiently transfect EXPI293 HEK cells. The resulting TAP supernatants containing the recombinant antibodies were tested in the following assays: Cell binding to HLA-G and null variants as above to help establish domain binding (multiplexed iQue) ILT2 Blockade Assay Affinity measurement by Biacore

[0423] The heavy and light chain variable region gene pairs from the TAP products of interest were then cloned as species-matched full-length IgG antibodies and re-expressed in a transient expression system. The recombinant cloned antibodies were then re-tested in the assays described above.

[0424] A total of 109 V (variable) regions were cloned and scored, of which only 30 were specific for HLA-G and did not bind other HLA-I, and of which only 9 showed blocking of HLA-G ILT2 interactions. All of these antibodies bound to the alpha 3 domain of HLA-G, and of the 9 specific and blocking antibodies, 5 had diverse sequences and these 5 were taken forward for further testing (listed in Table 3 below). [Table 3]

[0425] Several antibodies were subjected to humanization operations based on their properties and further evaluated in characterization assays, including HLA-G02 and HLA-G01, which appeared to be the best antibodies. Humanization of VR12389 is described in Example 4. HLA-G03 was also humanized, but nevertheless, the high affinity of the humanized antibody did not translate into improved functional activity, as described in further examples.

[0426] Additional assays performed on the purified antibodies included a cell-based specificity assay, an ILT4 blocking assay, ADCC. Data generated for the purified IgG1 antibodies is described in further examples below.

[0427] In this disclosure, antibody ID HLA-G02 refers to the humanized VR12389gL2gH16 IgG1 antibody.

[0428] Immunization strategy that led to the discovery of antibody 12389 Rab9 cells transiently expressing HLA-G Rab9 fibroblasts were cultured in RPMI medium + 10% FBS and 1% glutamine in 5-stack cell culture flasks. When the cells were 90-100% confluent, the medium was removed, the cells were washed with 100 ml of PBS, and the cells were removed from the cell culture flask using 100 ml of Accutase and incubated at room temperature for 10-15 min. The harvested cells were centrifuged and diluted with 5 × 10 7 The cells were resuspended at 3x10 cells / ml. HLA-G DNA was added to the cells at 250μg DNA per ml of cells. The Rab9+HLA-G DNA mixture was then added at 3x10 7Cells / cuvette were transferred into an electroporation cuvette. The cuvette was then pulsed with 150-170V of electricity (20ms 5.5 amps) using an in-house electroporator device (Zapper) and a Gene Pulser Xcell ShockPod Cuvette Chamber (BIORAD). After the electrical pulse, cells were quickly transferred into warmed Rab9 medium and transferred back into a fresh 5-stack cell culture flask.

[0429] After electroporating all the cells and transferring them to a new culture flask, they were incubated at 37°C and 5% CO2 for 24 h before harvesting the cells using Accutase (as described above), counting, and diluting them to 2 × 10 7 Cells / cryovial were transferred to cryovials and frozen in a -80 freezer. After 24 hours, frozen cells were transferred to a liquid nitrogen dewar for longer term storage.

[0430] Prior to freezing, 5×10 cells were counted for HLA-G expression by staining with Sigma APC-conjugated anti-HLA-G antibody (clone MEM / G9) for 1 h at 40°C and running samples on a FACS Calibre. 5 Transfected cells were tested.

[0431] On the day of immunization, for each shot, one vial of transfected cells was quickly thawed at 37° C., washed twice in 50 ml PBS, and then resuspended in 500 μl for administration to rabbits.

[0432] A DNA sequence encoding full-length HLA-G (SEQ ID NO: 111) was used for electroporation (the nucleic acid sequence of HLA-G optimized for expression in mammalian cells).

[0433] immunization 2 x 10 cells transiently expressing HLA-G on the cell surface, prepared as described above. 7Rab9 rabbit fibroblasts were used to subcutaneously immunize one female New Zealand White rabbit. An equal volume of complete Freund's adjuvant was injected subcutaneously into the rabbit at a separate site but at the same time as the cell immunization.

[0434] Rabbits were given two booster injections, 14 days apart, using Rab9 rabbit fibroblasts that transiently express HLA-G on the cell surface. Prior to each immunization, a heparinized bleed (200 μl) was taken from an ear vein. Serum was collected from the blood draw after centrifugation at 10,000 rpm for 5 min in a benchtop centrifuge and frozen at -20°C. Single cell suspensions of spleen, lymph node, bone marrow and peripheral blood mononuclear cells were prepared and frozen in 10% DMSO / FCS at -80°C until required for B cell discovery. At termination, blood was also drawn and serum was prepared as described above.

[0435] B cell harvesting and screening VR12389 was discovered in splenic memory B cell cultures. Splenocytes were cultured for 5 days at 37°C in 96-well plates with feeder cell lines and supplements. The cultures were then screened by a no-wash multiplexed flow cytometry assay (Intellicyt iQue). Culture supernatants containing secreted antibodies were mixed with the screening reagents described above (in-house generated HLA-G protein, and cell-expressed HLA-G constructs: HLA-G, HLA-G Null 1,2,3, HLA-G Null3). Screening cells used for screening were differentially stained to allow gating of the various populations, and antibody binding was identified using Dylight 405-labeled goat anti-rabbit antibody as the secondary antibody.

[0436] Hits were defined as HLA-G binders that were specific, i.e., did not bind to HLA-G Null 1,2,3 mutants or irrelevant control transfections. B cells involved in hits were returned to the original culture wells and then harvested using fluorescence focusing as previously described.

[0437] Selected cells for both leads were subjected to the same workflow after the focusing step. Reverse transcription (RT) and PCR of selected cells resulted in a "transcriptionally active PCR" (TAP) product encoding the V-region of the antibody that was used to transiently transfect EXPI293 HEK cells. The resulting TAP supernatant containing the recombinant antibody was characterized for cell binding, ILT2 blocking and affinity, after which the antibody was cloned and further progressed to be expressed on a large scale.

[0438] As mentioned above, several animals across different species (including rats, mice and rabbits) were immunized with syngeneic cells expressing different forms of HLA-G with or without B2m co-expression, but not all immunization strategies were successful in producing antibodies against HLA-G, or if production of antibodies against HLA-G was confirmed, the antibodies were not specific for HLA-G and / or did not block or could not bind to cell surface expressed HLA-G protein. Notably, immunization of rabbits with Rab9 cells expressing isolated alpha3 domain or expressing rabbit-human HLA-G chimera did not result in antibodies against HLA-G. Thus, the present disclosure provides an immunization method that is particularly useful for discovering therapeutically useful anti-HLA-G antibodies, the method comprising immunizing rabbits with Rab9 rabbit fibroblasts transiently expressing the full-length sequence of HLA-G on the cell surface.

[0439] Example 3: Generation of antibodies by phage display In order to identify antibodies that may be therapeutically useful, in parallel with the immunization campaign, a second approach was developed to attempt to identify antibodies from phage display libraries that specifically bind to HLA-G that may be therapeutically useful, and special screening and testing strategies were developed due to the particular challenges associated with generating therapeutically useful HLA-G antibodies.

[0440] Phage display library Three human naive combinatorial scFv phage libraries were utilized to obtain antibodies that bind to HLA-G using various constructs, with the aim of obtaining selective binders to HLA-G with no or minimal binding to other HLA-I. The libraries were biopanned using 3 or 4 rounds of selection using recombinant HLA-G expressed only on the cell surface, or followed by recombinant HLA-G extracellular domain protein in the final round. An optional subtraction step against HLA-G Null 1,2,3 (soluble protein or cell expression) was included in the final round to enrich for HLA-G specific binders.

[0441] In summary, biopanning on cells consisted of co-transfecting DNA constructs encoding human HLA-G and β2m in ExpiCHO for the first round, or Expi293 HEK for subsequent rounds, and incubating these cells with blocked phage virions pre-depleted on non-transfected or HLA-G Null 1,2,3 cells. Biopanning on proteins was performed by incubating blocked phage particles in solution with plate-coated HLA-G, or biotinylated HLA-G, followed by capture on streptavidin or neutravidin magnetic beads. After several washes with PBS Tween, target-bound phages were eluted and reamplified by infecting E. coli TG1s.

[0442] Phage screening After the final round of selection, 1692 monoclonal rescued phages were screened by ELISA using biotinylated HLA-G ECD captured on streptavidin-coated plates. Binding was detected by HRP-conjugated anti-M13 pVIII coat protein antibody. The specificity of these monoclonal phage clones was assessed using biotinylated HLA-G Null 1, 2, 3. 359 binders of interest were sequenced and analyzed for diversity based on their variable heavy chain CDR3 sequence motifs. 81 unique clones were then reformatted as scFv-rabbit IgG Fc fusions in mammalian expression vectors for further characterization.

[0443] Further characterization of HLA-G selective binders scFv-Fc were expressed in Expi293 HEK. Their binding and specificity were tested by flow cytometry using IntelliCyt iQue Screener Plus. Diluted antibody-containing supernatants were added to ExpiHEK co-transfected with human HLA-G or HLA-G and β2 microglobulin. Binding was detected using Fc fragment-specific fluorescent antibodies.

[0444] Twenty-one HLA-G cell binders that did not bind to HLA-G Null 1,2,3+B2m were further characterized by binding to HLA-A, -B, -C, -E, -F expressed on HEKs after reformatting as scFv-Fc and / or into full-length human IgG1 by SPR, ILT2 blocking assays, and by binding to JEG3 cells by flow cytometry.

[0445] Fourteen antibodies were identified as being highly specific for HLA-G with no or minimal binding to other HLA-I molecules, and of these, only six antibodies were found to block the interaction between HLA-G and ILT2.

[0446] Of the six specific blocking antibodies, three had diverse sequences and these three were taken forward for further testing (HLA-G06, HLA-G07, HLA-G08).

[0447] Additional assays performed on the purified antibodies included a cell-based specificity assay, an ILT4 blocking assay, and ADCC. Data generated on the purified antibodies is described in further examples below.

[0448] Example 4: Humanization of antibody 12389 Antibody 12389 was humanized by grafting CDRs from rabbit V regions onto human germline antibody V region frameworks. To restore antibody activity, some framework residues from the rabbit V regions were also retained in the humanized sequence. These residues were selected using the protocol outlined by Adair et al. (1991) (WO 91 / 09967). Alignments of rabbit antibody (donor) V region sequences with human germline (acceptor) V region sequences, along with the designed humanized sequences, are shown in Figures 2 and 3. The CDRs grafted from the donor to the acceptor sequences are as defined by Kabat (Kabat et al., 1987), except for CDR-H1 (see Adair et al., WO 91 / 09967), where the combined Chothia / Kabat definition is used.

[0449] For antibody 12389, the human V region IGKV1D-13+IGKJ4 J region (IMGT, http: / / www.imgt.org / ) was selected as the acceptor for the light chain CDRs. All light chain framework residues in the humanized grafted variants are from the human germline gene, but with none, one or two residues from the group including residues 3 and 70, with donor residues valine (V3) and glutamine (Q70), respectively, retained, with reference to SEQ ID NO: 7 (rabbit VL).

[0450] The human V region IGHV3-66+IGHJ4 J region (IMGT, http: / / www.imgt.org / ) was selected as the acceptor for the heavy chain CDRs of antibody 12389. Like many rabbit antibodies, the VH gene of antibody 12389 is shorter than the selected human acceptor. When aligned with the human acceptor sequence, framework 1 of the VH region of antibody 12389 lacks the N-terminal residues that are retained in the humanized antibody (Figure 3). Framework 3 of the 12389 rabbit VH region also lacks two residues (75 and 76, rabbit VH, with reference to SEQ ID NO: 11) in the loop between beta sheet strands D and E. In the humanized graft variants, the gaps are filled by the corresponding residues (lysine 75, K75; asparagine 76, N76) from the selected human acceptor sequence. The heavy chain framework residues in the humanized graft variants are all derived from the human germline gene, except for one or more residues from the group including residues 24, 48, 49, 71, 73, 78 and 96, in which donor residues valine (V24), isoleucine (I48), glycine (G49), lysine (K71), serine (S73), valine (V78) and glycine (G96), respectively, were retained, with reference to SEQ ID NO:11.

[0451] The mutant humanized antibody chains and combinations thereof were expressed and assessed for their binding affinity to human HLA-G compared to the parental antibody.

[0452] Expression in Expi293 cells Genes encoding the mutant heavy and light chain V-region sequences were designed and constructed by automated synthesis procedures by ATUM (Newark, Calif.) For transient expression in mammalian cells, the humanized light chain V-region genes were cloned into the light chain expression vector pMhCK, which contains DNA encoding the human kappa chain constant region (Km3 allotype). The humanized heavy chain V region genes were cloned into the human gamma 4 heavy chain expression vector pMhg4PFL, which contains DNA encoding the human gamma 4 heavy chain constant region with the hinge stabilizing mutation S228P (Angal S., King DJ, Bodmer MW, Turner A., ​​Lawson ADG, Roberts G., Pedley B. and Adair JRA single amino acid substitution abolishes the heterogeneity of chimeric mouse / human (IgG4) antibody. Mol. Immunol. 1993,30(1):105-8), or into the gamma 1 heavy chain expression vector pMhg1FL, which contains DNA encoding the human gamma 1 heavy chain constant region (G1m17, 1 allotype). Co-transfection of the resulting heavy and light chain vectors into Expi293™ suspension cells was achieved using ExpiFectamine™ 293 transfection reagent (A14525, ThermoFisher Scientific), resulting in expression of humanized recombinant IgG4P and IgG1 antibodies.

[0453] Affinity measurement by SPR As described below, the assay format was capture of anti-HLA-G IgG by immobilized anti-human IgG Fc-specific antibodies, followed by titration of HLA-G on the captured surface.

[0454] The affinity of anti-HLA-G IgG bound to HLA-G was determined by surface plasmon resonance using a Biacore T200 (GE Healthcare Biosciences AB). The assay was performed at 25°C. Affinipure F(ab') was coupled to the antibody via amine coupling chemistry.2 Fragmented goat anti-human IgG, Fc specific (Jackson ImmunoResearch) was immobilized on a Series S CM5 Sensor Chip (GE Healthcare Bio-Sciences AB) to a level of approximately 6000 response units (RU). + Buffer (10 mM HEPES pH 7.4, 0.15 M NaCl, 3 mM EDTA, 0.05% Surfactant P20, GE Healthcare Bio-Sciences AB) was used as the running buffer with a flow rate of 10 μL / min. Reference surfaces were prepared by activating and deactivating the appropriate flow cells.

[0455] For capture with immobilized anti-human IgG, Fc, a 10 μL injection of anti-HLA-G IgG at concentrations between 0.15 and 0.7 μg / mL was used. Human HLA-G ECD+B2m was titrated against the captured anti-HLA-G IgG at 50 nM at a flow rate of 30 μL / min for 60 s, followed by dissociation for 150 s. The surface was regenerated at a flow rate of 10 μL / min by injection of 10 μL of 40 mM HCl followed by 5 μL of 5 mM NaOH.

[0456] Background subtracted binding curves were analyzed using Biacore T200 Evaluation Software (version 3.0) using 1:1 binding fitted with local Rmax.

[0457] Antibodies were analyzed at the beginning and end of the assay, demonstrating good precision. High quality data was generated for all samples, as summarized in Tables 4 and 5. [Table 4] [Table 5]

[0458] As shown in Table 4, all grafts except 12389gL2gH5, 12389gL2gH6, 12389gL2gH8, 12389gL2gH11 had a KD of less than 10 nM and less than the KD measured for the parental 12389 (chimeric rabbit V region / human Fc).

[0459] Graft 12389gL2gH16, which retains donor residues I48, G49, K71 and V78 of the VH framework and donor residue V3 of the VL framework, had the highest affinity binding to human HLA-G as measured by surface plasmon resonance and retained functionality when expressed in different formats such as IgG4P and IgG1 (Tables 4-5), and graft 12389gL2gH16 was selected for further characterization.

[0460] Example 5. Expression and purification of HLA-G01-HLA-G08 The antibodies were transiently expressed as IgG1 in CHO cells transfected with DNA vectors encoding the LC and HC of the HLA-G01-HLA-G08 antibody (1:1 ratio of LC:HC) and purified using Protein A affinity chromatography according to well-known methods for further studies.

[0461] Protein concentration was determined by reading absorbance at 280 nm using a nanodrop, and purity was determined by analytical size-exclusion HPLC. Monomer content was determined using analytical size-exclusion chromatography and SDS Page electrophoresis. Endotoxin levels were determined using Charles River Endosafe® LAL Reagent Cartridge Technology and an Endosafe® nexgen-PTS reader, with levels below 1 EU / mg being acceptable quality.

[0462] The final purified sample was highly pure and contained greater than 98% monomer content. The final purified sample was analyzed by intact mass spectrometry to confirm heavy and light chain masses, predicted modifications and identities.

[0463] Example 6: Production of defucosylated HLA-G02 antibodies Methods for generating KO FUT8 CHOSXE / DG44 cells RNA guides (gRNAs) were designed to knock out two exons containing sequences encoding the alpha 1,6 fucosyltransferase (FUT8) active site. Two to eight gRNAs (seven total) were designed using Benchling software to create multiple deletions (the largest possible deletion is 4 kb) in both the forward and reverse strands, using the guides as a pool to maximize possible knockouts. The sequences of the gRNAs are shown in Table 6 below. [Table 6]

[0464] gRNA pools were prepared at final pmol concentrations as shown in Table 7. Cas09 (ThermoFisher) was prepared to the required concentrations as shown in Table 7.

[0465] 3x10 for nucleofection 6 CHO SXE / DG44 cells were prepared and centrifuged at 100xg for 8 min, after which the cells were washed in PBS, centrifuged again at 100xg for 8 min, and resuspended in 100 μl of nucleofection solution to give 3x10 4 A total of 1000 cells / µl were obtained. Pre-transfection mixtures were prepared as shown in Table 7 below and left without cells for 10-60 min at room temperature to allow the Cas09 / gRNA complexes to form. [Table 7]

[0466] Following the manufacturer's instructions, 16.67 μl of cells (3 × 10 4After adding 100 μl of cells / μl of the complex, nucleofection (Nucleofector 4D, Lonza) was proceeded. The cells were harvested by adding 900 μl of pre-warmed CD CHO medium and placed in a small T25 flask in an upright position. The incubation was continued at 37°C, 5% CO 2 After 24 h at RT, the cells were harvested and split, after which the medium was replaced with fresh pre-warmed CD CHO (with the antibiotics penicillin, streptomycin and amphotericin B added to reduce the risk of contamination) and transferred to 125 ml shake flasks (after 96–120 h).

[0467] Ten days after nucleofection, cells were ready to be sorted by FACS. Cells were stained with LCA stain (Lens Culinaris Agglutinin conjugated to Fluorescein). Cells were prepared for FACS by centrifugation at x100g for 8 min, cells were washed with PBS, centrifuged again at x100g for 8 min, resuspended in pre-warmed CD CHO medium, 20 μg / ml LCA stain was added, left for 45 min, cells were washed x2 with PBS (centrifugation at x100g for 8 min, resuspended in PBS) to remove unbound stain. LCA-bound fucose on the cell surface of FUT8-positive cells, FUT8-negative (knockout cells) was not stained and collected in pre-warmed CD CHO medium. Cells were placed at 37°C, 5% CO2 to recover and passaged until cells reached the required cell density.

[0468] Production of defucosylated HLA-G antibodies HLA-G02 antibody constructs were expressed in an engineered CHO-SXE cell line (Cain et al 2012) that was further modified to knock down the α-1,6-fucosyltransferase enzyme (FUT8) as described above. Defucosylated antibodies were transiently produced from these cells using the ExpiCHO transfection system (Thermo Fisher Scientific). Cells were seeded at a cell density of 6×10^6 cells / ml in ExpiCHO expression medium following the high yield protocol. For a 200 ml culture, 200 μg of DNA was diluted in 8 ml Opti-PRO serum-free medium (SFM) and mixed with 7.4 ml Opti-PRO SFM containing 640 μl ExpiFectamine transfection reagent before adding to the cells. Incubate at 8% CO on a shaking platform set at 190 RPM. 2 The cells were transferred to an incubator set at 37° C. and on day 1 post-transfection, 48 ml of feed and 1200 μl of enhancer were added to the cells and returned to the incubator and the temperature was reduced to 32° C. Cultures were harvested on day 10 by centrifugation at 4000 RPM for 1 hour. Cell culture supernatants were clarified by application to a 0.22 μM Stericup Filter.

[0469] The titer of the product was determined by loading 100 μl of the supernatant onto a Protein G column attached to an Infinity High Performance Liquid Chromatography (HPLC) system. The product was eluted from the column with 150 mM sodium chloride, pH 2.1, and A 280The signal was compared to purified Fab standards. The clarified supernatant was loaded onto a MabSelect Sure column (GE Healthcare) at 5 ml / min and washed with 3 column volumes (CV) of PBS pH 7.4. The captured protein was eluted from the column with a low pH buffer, 0.1 M sodium citrate buffer pH 3.6. The eluate was neutralized with 2 M Tris-HCl pH 8.5. The affinity purified protein was loaded onto a Superdex 16 / 60 gel filtration chromatography column equilibrated in 10 mM phosphate buffered saline (pH 7.4) at 1 ml / min to remove high molecular weight species, and the eluted fractions were analyzed by analytical SE-UPLC, after which the appropriate fractions were pooled. The pooled fractions were subjected to SDS-PAGE and SE-UPLC to determine protein quality and purity.

[0470] In this disclosure, unless otherwise specified, "HLA-G02" refers to unmodified conventional (i.e., fucosylated) VR12389gL2gH16 IgG1. "Defucosylated HLA-02" refers to the corresponding defucosylated IgG1 antibody produced according to the methods described herein.

[0471] Example 7: HLA-G antibody binding, affinity and specificity 7.1. Affinity for HLA-G WT as measured by SPR The binding affinity of anti-HLA-G antibodies (hIgG1 format) to HLA-G was determined by surface plasmon resonance using a Biacore T200 (GE Healthcare Biosciences AB). The assay was performed at 25°C. Affinipure F(ab') was bound to HLA-G via amine coupling chemistry. 2 Fragmented goat anti-human IgG, Fc specific (Jackson ImmunoResearch) was immobilized on a Series S CM5 Sensor Chip (GE Healthcare Bio-Sciences AB) to a level of approximately 6000 response units (RU). +Buffer (10 mM HEPES pH 7.4, 0.15 M NaCl, 3 mM EDTA, 0.05% Surfactant P20, GE Healthcare Bio-Sciences AB) was used as the running buffer with a flow rate of 10 μL / min. Reference surfaces were prepared by activating and deactivating the appropriate flow cells.

[0472] For capture with immobilized anti-human IgG, Fc, 10 μL injections of anti-HLA-G antibodies at concentrations between 0.6 and 0.9 μg / mL were used. Human HLA-G ECD+B2m was titrated against captured anti-HLA-G hIgG1 from various top concentrations of 4000 nM, 400 nM, 100 nM and 50 nM at a flow rate of 30 μL / min for 60, 90 or 120 s, followed by dissociation for 120, 180 or 240 s (Table 8). [Table 8]

[0473] The results are shown in Table 9. [Table 9] HLA-G06, HLA-G07 and HLA-G08 had the lowest affinity (high KD values) for HLA-G as determined by SPR. HLA-G01, HLA-G02 and HLA-G03 had the highest affinity for HLA-G.

[0474] In a separate assay, the affinity of defucosylated HLA-G02 was also evaluated and found to be similar to its conventional (fucosylated) HLA-G02 counterpart. The results are shown in Table 10 below. [Table 10]

[0475] 7.2. Affinity for HLA-G Null 1, 2, 3 Measured by SPR to Assess Specificity The binding affinity of anti-HLA-G IgG1 antibodies to HLA-G Null 1,2,3 was determined by surface plasmon resonance using a Biacore T200 (GE Healthcare Biosciences AB). The assay was performed at 25°C. Affinipure F(ab') was coupled to the antibody via amine coupling chemistry. 2 Fragmented goat anti-human IgG, Fc specific (Jackson ImmunoResearch) was immobilized on a Series S CM5 Sensor Chip (GE Healthcare Bio-Sciences AB) to a level of approximately 6000 response units (RU). + Buffer (10 mM HEPES pH 7.4, 0.15 M NaCl, 3 mM EDTA, 0.05% Surfactant P20, GE Healthcare Bio-Sciences AB) was used as the running buffer with a flow rate of 10 μL / min. Reference surfaces were prepared by activating and deactivating the appropriate flow cells.

[0476] For capture with immobilized anti-human IgG, Fc, a 10 μL injection of anti-HLA-G antibody at a concentration of 0.6-0.9 μg / mL was used. Human "HLA-G Null 1,2,3" mutant AVI tev 10HisTag+B2m was titrated from 20 μM against the captured anti-HLA-G IgG at a flow rate of 30 μL / min for 60 s, followed by dissociation for 150 s. The surface was regenerated at a flow rate of 10 μL / min by injection of 10 μL of 40 mM HCl followed by 5 μL of 5 mM NaOH.

[0477] Steady-state analysis was used to analyze background subtraction binding curves using Biacore T200 Evaluation Software (version 3.0). The results are shown in Table 11. [Table 11] No binding was detected for HLA-G01 to HLA-G07. HLA-G08 showed some binding to HLA-G Null 1, 2, and 3 and was therefore not very specific for HLA-G.

[0478] 7.3. Binding to HEK-expressed HLA-G wild type versus HLA-G Null 1, 2, 3 to assess antibody specificity Binding to HLA-G wild type expressed on human embryonic kidney (HEK293) cells was measured and compared to "HLA-G Null 1,2,3" to determine binding specificity. Advantageously, cell-based assays can be used to assess binding to dimeric HLA-G, but the SPR assay described above measures binding only to monomeric HLA-G.

[0479] HEK293 cells transfected with either HLA-G / β2m or HLA-G Null Null 1,2,3 / β2m were incubated with anti-HLA-G IgG1 for 2 h at 4°C in 384-well V-bottom plates (Greiner). IgG concentrations ranging from 100 nM to 0.05 nM were diluted in PBS, 1% FBS, 0.1% sodium azide. After the incubation period, cells were washed three times in assay buffer and then incubated with 20 μl of staining solution (R-Phycoerythrin AffiniPure F(ab')) for 20 min at 4°C. 2 Fragment Goat Anti-Human IgG(H+L) (Jackson ImmunoResearch)-3.75 μg / ml and Viability Dye e780 (Life Technologies). After a washing step, cells were incubated for 10 min at room temperature in 10% neutral buffered formalin solution (Sigma-Aldrich) protected from light. Cells were then washed and resuspended in 40 μl of PBS. Samples were run on a FACS Canto II instrument in HTS mode to determine the percentage of PE positive cells. EC50 and Emax were calculated from the median fluorescence intensity using FlowJo analysis software. The results are shown in Table 12. [Table 12]

[0480] HLA-G03 had the lowest affinity for HLA-G expressed on HEK293 compared to HLA-G01 and HLA-G02. Thus, HLA-G01 and HLA-G02 were preferred over HLA-G03. The affinity of defucosylated HLA-G02 for HEK-expressed HLA-G was similar to that of its conventional (fucosylated) counterpart. Binding of HLA-G08 to HLA-G Null 1, 2, and 3 was detected, confirming that this antibody is not very specific.

[0481] 7.4. Binding affinity to JEG3 cells as determined by FACS assay The binding affinity of anti-HLA-G IgG was measured in a flow cytometry cell-based assay using human choriocarcinoma trophoblastic (JEG3), which naturally expresses HLA-G. This assay can be advantageously used to measure binding to cells that naturally express HLA-G, including binding to dimeric HLA-G on the cells, whereas the SPR assay described above measures only binding to monomeric HLA-G.

[0482] JEG3 cells were incubated for 2 h at 4°C in a microcentrifuge tube (Eppendorf) with 1.5 ml of anti-HLA-G IgG1 solution. IgG concentrations ranging from 10 nM to 0.0005 nM were diluted in PBS, 1% FBS, 0.1% sodium azide. Cells were transferred to a 384-well V-bottom plate (Greiner), washed three times in assay buffer, and incubated for 20 min at 4°C with 20 μl of staining solution (R-Phycoerythrin AffiniPure F(ab') 2Fragment Goat Anti-Human IgG(H+L) (Jackson ImmunoResearch)-7.5 μg / ml and Viability Dye e780 (Life Technologies). After a washing step, cells were incubated for 10 min at room temperature in 10% neutral buffered formalin solution (Sigma-Aldrich) protected from light. Cells were then washed and resuspended in 20 μl of PBS. Samples were run on a FACS Canto II instrument in HTS mode to determine the percentage of PE positive cells. KD was calculated from the median fluorescence intensity using FlowJo analysis software. The results are shown in Table 13. [Table 13] HLA-G01 and HLA-G02 showed high affinity in FACS assays. Results confirmed that HLA-G03 has a lower affinity for HLA-G compared to HLA-G01 and HLA-G02, especially when expressed as a dimer on the surface of cells that naturally express HLA-G.

[0483] 7.5. Binding to HLA-I and JEG3 wild type (WT) / HLA-G knockdown The specificity of anti-HLA-G IgG binding was further investigated by measuring binding to the HLA-A / B / C / E / F consensus molecule expressed on HEK293 cells. In the same assay, binding to HLA-G expressed on JEG3 cells was also measured and compared to JEG3 HLA-G knockdown (KD).

[0484] HEK293 cells transfected with either HLA-A, B, C, E or F consensus sequence / β2m and JEG3 WT, as well as JEG3 KD, were incubated with anti-HLA-G IgG1 for 2 h at 4°C in 384-well V-bottom plates (Greiner). IgG concentrations ranging from 100 nM to 0.05 nM were diluted in PBS, 1% FBS, 0.1% sodium azide. After the incubation period, cells were washed three times in assay buffer and then incubated with 20 μl of staining solution (R-Phycoerythrin AffiniPure F(ab')) for 20 min at 4°C. 2 Fragment Goat Anti-Human IgG(H+L) (Jackson ImmunoResearch)-3.75 μg / ml and Viability Dye e780 (Life Technologies). After a washing step, cells were incubated for 10 min at room temperature in 10% neutral buffered formalin solution (Sigma-Aldrich) protected from light. Cells were then washed and resuspended in 40 μl of PBS. Samples were run on a FACS Canto II instrument in HTS mode to determine the percentage of PE positive cells. EC50 and Emax were calculated from the median fluorescence intensity using FlowJo analysis software. The results are shown in Table 14 and Table 15. [Table 14] No binding to the HLA-I consensus was detected ("ND"). The results confirmed that antibodies HLA-G01 and HLA-G02 are highly specific for HLA-G. [Table 15]

[0485] Furthermore, no binding to JEG3 KD was detected. HLA-G03 had a similar EC50 compared to HLA-G01 and HLA-G02 (or defucosylated HLA-G02), but a much lower Emax, and thus was not as good as HLA-G01 or HLA-G02.

[0486] 7.6. Determination of the binding domain of anti-HLA G antibodies on HLA-G The binding domain of anti-HLA-G IgG was characterized by measuring binding to HLA-G Null3, HLA-G Null1,3, and HLA-G Null3 2AA expressed on human embryonic kidney (HEK293) cells.

[0487] HEK293 cells transfected with either HLA-G Null3 / β2m, HLA-G Null1,3 / β2M, or HLA-G Null3 2AA / β2M were incubated with anti-HLA-G IgG1 for 2 h at 4°C in 384-well V-bottom plates (Greiner). IgG concentrations ranging from 100 nM to 0.05 nM were diluted in PBS, 1% FBS, 0.1% sodium azide. After the incubation period, cells were washed three times in assay buffer and then incubated with 20 μl of staining solution (R-Phycoerythrin AffiniPure F(ab')) for 20 min at 4°C. 2 Fragment Goat Anti-Human IgG(H+L) (Jackson ImmunoResearch)-3.75 μg / ml and Viability Dye e780 (Life Technologies). After a washing step, cells were incubated for 10 min at room temperature in 10% neutral buffered formalin solution (Sigma-Aldrich) protected from light. Cells were then washed and resuspended in 40 μl of PBS. Samples were run on a FACS Canto II instrument in HTS mode to determine the percentage of PE positive cells. EC50 and Emax were calculated from the median fluorescence intensity using FlowJo analysis software. The results are shown in Table 16. [Table 16] The data showed that antibodies including HLA-G02 were specific for the HLA-G alpha 3 domain. The data confirmed the lowest specificity of HLA-G08.

[0488] 7.7. Specificity of HLA-G02 assessed in PBMC experiments The specificity of HLA-G02 was further confirmed using PBMCs from 50 different donors representing a range of HLA-I alleles. The aim was to ensure that the anti-HLA-G antibodies were specific and did not cross-react with other HLA-I molecules expressed, particularly on PBMCs and CD4+ T lymphocytes.

[0489] PBMCs were purified from peripheral venous blood and stored in liquid nitrogen in freezing medium (90% FBS + 10% DMSO). PBMCs from 50 different donors were thawed and resuspended in 1 ml complete RPMI medium (RPMI 1640 medium + 10% fetal bovine serum, 2 mM Glutamax and 1% penicillin / streptomycin). Cells were centrifuged at 300 rpm for 10 min and washed twice with PBS. Cell pellets were resuspended in 1 ml Facs buffer (PBS, 0.5% BSA and 2 mM EDTA) and cells were seeded into 96-well plates with 50 μl cell suspension / well.

[0490] Cells were stained with anti-human CD4-APC (Biolegend, 2.5 μl / well) and with anti-HLA-G antibodies (HLA-G02 or control "pan-HLA", IgG1 that binds to HLA-I and is not specific for HLA-G) or isotype control (50 μl of solution, 20 μg / ml / well). Cells were incubated for 20 min at room temperature in the dark, then washed twice with Facs buffer and resuspended in 50 μl of Facs buffer containing the secondary antibody Goat Anti-Human IgG-FITC (Jackson ImmunoResearch, diluted 1 / 100). Cells were incubated for another 20 min at room temperature in the dark, then washed twice with Facs buffer. Cells were resuspended in 100 μl / well of Facs buffer and samples were acquired using a Canto II (HTS 1) and 10,000 events were collected per sample. Analysis was performed by measuring the mean fluorescence intensity (MFI) of each CD4+ cell population from each donor using FlowJo software v10.6.0. Graphs were generated using Graphpad Prism software.

[0491] The results are shown in Figure 4, which shows the lack of binding of the specific anti-HLA-G antibody HLA-G02 to CD4 T cells across 50 different donors compared to pan-HLA antibodies. Data are presented as mean fluorescence intensity (MFI) for each donor.

[0492] 7.8. Cell membrane-associated HLA-G isoforms: Binding of HLA-G02 to HLA-G1, HLA-G2, HLA-G3, or HLA-G4 Expi293F human cells were transfected with plasmids encoding either HLA-G1, HLA-G2, HLA-G3 or HLA-G4 using the xpiFectamine™ 293 Transfection Kit according to the manufacturer's protocol (ThermoFisher Scientific, ref. A14524). 24 hours after transfection, cells were harvested, washed in PBS (300 rpm, 10 min) and resuspended in PBS. Cells were seeded in 96-well / plates, stained with anti-HLA-G antibody HLA-G02 (human IgG1) or commercial antibody 4H84 (Abcam, mouse IgG1) at a final concentration of 10 μg / ml and incubated for 20 min in the dark. The cells were then washed twice with Facs buffer (PBS, 0.5% BSA and 2 mM EDTA) and resuspended in 50 μl of Facs buffer containing the secondary antibodies goat anti-human IgG-PE (Jackson ImmunoResearch, diluted 1 / 100) or goat anti-mouse IgG-PE (Jackson ImmunoResearch, diluted 1 / 100). The cells were incubated for a further 20 minutes at room temperature in the dark and then washed twice with Facs buffer. The cells were resuspended in 100 μl of Facs buffer and samples were acquired using a Canto II (HTS 1).

[0493] The results are shown in Figure 5. Figure 5A shows that the HLA-G02 antibody specific for the HLA-G alpha 3 domain recognizes both HLA-G1 and HLA-G2 (with 68 and 17% positive cells, respectively), but not HLA-G3 or HLA-G4, compared to the negative control (neg CTRL, irrelevant Ab). The commercially available mouse anti-HLA-G antibody 4H84 specific for the alpha 1 domain was used as a positive control, which recognizes all isoforms HLA-G1, HLA-G2, HLA-G3 or HLA-G4 (staining 81, 34, 67 and 16% of the cells, respectively).

[0494] Figure 5B more specifically shows that the HLA-G02 antibody specific for the HLA-G alpha3 domain recognizes HLA-G2 (17% positive cells) compared to a negative control (irrelevant Ab). The commercially available mouse anti-HLA-G antibody 4H84 specific for the alpha1 domain was used as a positive control for HLA-G2 (34% positive cells).

[0495] Example 8: Evaluation of the ability of HLA-G antibodies to block the interaction between HLA-G and ILT2 or ILT4 Blocking the interaction between JEG3 and ILT2 Anti-HLA-G IgGs were incubated with human choriocarcinoma trophoblasts (JEG3), which express HLA-G, and ILT2 rabbit Fc to determine their efficacy in blocking ILT2 binding to HLA-G.

[0496] JEG3 cells were incubated with anti-HLA-G IgG1 in 384-well V-bottom plates (Greiner) for 1 h at 4°C. IgG concentrations ranging from 100 nM to 0.05 nM were diluted in PBS, 1% FBS, 0.1% sodium azide. After the incubation period, cells were washed in assay buffer and then incubated with 20 μl of a 3 μg / ml ILT2rbFc solution for 1 h at 4°C. After incubation, 5 μl of staining solution (Fluorescein (FITC) AffiniPure F(ab') 2 Fragment Goat Anti-Rabbit IgG, Fc fragment specific (Jackson ImmunoResearch)-7.5 μg / ml, and Viability Dye e780 (Life Technologies) were added to each well for 20 min at 4°C. After a washing step, cells were incubated for 10 min at room temperature in 10% neutral buffered formalin solution (Sigma-Aldrich) protected from light. Cells were then washed and resuspended in 40 μl of PBS. Samples were run on a FACS Canto II instrument in HTS mode to determine the percentage of FITC positive cells. IC50 and percent inhibition were calculated from the median fluorescence intensity using FACSDiva analysis software. The results are shown in Table 17.

[0497] Blocking the interaction between JEG3 and ILT2: high reaction volumes In previous ILT2 blocking assays, some antibodies showed extremely low IC50. At these low concentrations, ligand depletion most likely occurs due to small reaction volumes that may result in overestimation of IC50 values. To overcome ligand depletion, high volumes of anti-HLA-G IgG solutions (of the best blocking antibodies identified in previous assays) were incubated with JEG3 expressing HLA-G, and ILT2 rabbit Fc to improve the measurement of their potency in blocking ILT2 binding to HLA-G.

[0498] JEG3 cells were incubated with 1.5 ml of anti-HLA-G IgG1 solution for 2 hours at 4°C in a microcentrifuge tube (Eppendorf). IgG concentrations ranging from 10 nM to 0.005 nM were diluted in PBS, 1% FBS, 0.1% sodium azide. Cells were transferred to a 384-well V-bottom plate (Greiner), washed three times in assay buffer, and then incubated with 20 μl of a 3 μg / ml ILT2rbFc solution for 1 hour at 4°C. After the incubation period, cells were washed and stained with 20 μl of staining solution (Fluorescein (FITC) AffiniPure F(ab') 2 The cells were incubated with Fragment Goat Anti-Rabbit IgG, Fc fragment specific (Jackson ImmunoResearch)-7.5 μg / ml, and Viability Dye e780 (Life Technologies) for 20 min at 4°C. After a washing step, the cells were incubated for 10 min at room temperature in 10% neutral buffered formalin solution (Sigma-Aldrich) protected from light. The cells were then washed and resuspended in 20 μl of PBS. The samples were run on a FACS Canto II instrument in HTS mode to determine the percentage of FITC positive cells. The IC50 and percent inhibition were calculated from the median fluorescence intensity using FlowJo analysis software. The results are shown in Table 17.

[0499] 8.2. Blocking the interaction between HLA-G and ILT2 expressed on HEK Anti-HLA-G IgGs were incubated with HLA-G and ILT2 rabbit Fc transfected HEK293 cells to measure their potency in blocking ILT2 binding to HLA-G.

[0500] HLA-G / β2M transfected HEK293 cells were incubated with anti-HLA-G IgG1 for 1 h at 4°C in 384-well V-bottom plates (Greiner). IgG concentrations ranging from 100 nM to 0.05 nM were diluted in PBS, 1% FBS, 0.1% sodium azide. After the incubation period, cells were washed in assay buffer and then incubated with 20 μl of 1 μg / ml ILT2rbFc solution for 1 h at 4°C. After incubation, 5 μl of staining solution (Fluorescein (FITC) AffiniPure F(ab') 2 Fragment Goat Anti-Rabbit IgG, Fc fragment specific (Jackson ImmunoResearch)-3 μg / ml, and Viability Dye e780 (Life Technologies) were added to each well for 20 min at 4°C. After a washing step, cells were incubated for 10 min at room temperature in 10% neutral buffered formalin solution (Sigma-Aldrich) protected from light. Cells were then washed and resuspended in 40 μl of PBS. Samples were run on a FACS Canto II instrument in HTS mode to determine the percentage of FITC positive cells. IC50 and percent inhibition were calculated from the median fluorescence intensity using FACSDiva analysis software. The results are shown in Table 17.

[0501] Blocking the interaction between HLA-G and ILT4 expressed on HCT116 Anti-HLA-G IgGs were incubated with HLA-G and ILT4 rabbit Fc transfected human colon carcinoma (HCT116) cells to determine their efficacy in blocking ILT4 binding to HLA-G.

[0502] HCT116 cells transfected with HLA-G / β2M were incubated with anti-HLA-G IgG for 1 h at 4°C in 384-well V-bottom plates (Greiner). IgG1 concentrations ranging from 100 nM to 0.05 nM were diluted in PBS, 1% FBS, 0.1% sodium azide. After the incubation period, cells were washed in assay buffer and then incubated with 20 μl of 0.4 μg / ml ILT4rbFc solution for 1 h at 4°C. After incubation, 5 μl of staining solution (Fluorescein (FITC) AffiniPure F(ab') 2 Fragment Goat Anti-Rabbit IgG, Fc fragment specific (Jackson ImmunoResearch)-1.5 μg / ml, and Viability Dye e780 (Life Technologies) were added to each well for 20 min at 4°C. After a washing step, cells were incubated for 10 min at room temperature in 10% neutral buffered formalin solution (Sigma-Aldrich) protected from light. Cells were then washed and resuspended in 40 μl of PBS. Samples were run on a FACS Canto II instrument in HTS mode to determine the percentage of FITC positive cells. IC50 and percent inhibition were calculated from the median fluorescence intensity using FACSDiva analysis software. The results are shown in Table 17. [Table 17] HLA-G01 and HLA-G02 were identified as the best blockers of HLA-G association with ILT2 and with ILT4. The blocking properties of defucosylated HLA-G02 were similar to those of its conventional (i.e., fucosylated) counterpart.

[0503] Example 9: Efficacy and potency of HLA-G01-HLA-G08 antibodies in ADCC of HLA-G expressing cells Antibody-dependent cellular cytotoxicity (ADCC) is an immune mechanism by which cells expressing Fc receptors, e.g., NK cells, can recognize and kill cells coated by antibodies. It is a critical process for anti-cancer responses and is a key mechanism underlying the efficacy of many anti-cancer therapies. Two different types of target cells expressing HLA-G were used to determine the ability of a panel of anti-HLA-G IgG1 antibodies to induce ADCC in vitro. These cells were either transfected with HLA-G and human beta-2-microglobulin (HCT116 colorectal cancer cells) or endogenously expressed the targets on their cell surface (JEG3 cells).

[0504] method: Transfection of HCT116 colorectal cancer cell line HCT116 cells were transfected with an HLA-G construct that also encoded a green fluorescent protein (GFP) tag. Thus, cells successfully transfected with HLA-G also expressed GFP and could be easily identified and accurately monitored by flow cytometry. A total of 40 μg of DNA (20 μg each of HLA-G ECD GFP and human β2M plasmid) and 80 μL of Lipofectamine LTX were used to transfect 4 × 10 cells in a T75 tissue culture flask. 6 HCT116 cells were transfected. After 24 hours, the cells were removed from the flask and used as target cells as described below.

[0505] In vitro ADCC assay HLA-G transfected HCT116 or JEG3 target cells were seeded (2 × 10 in a volume of 50 μL) in appropriate culture medium in polypropylene round-bottom plates. 4 cells / well). Anti-HLA-G or control antibodies were prepared as 4-fold concentrated stocks in the same medium and 50 μl / well was added to the appropriate wells. All antibodies were tested in either duplicates or triplicates depending on the number of donor NK cells available. Some target cells were left without antibody and used as untreated controls.

[0506] Primary human NK cells were isolated from whole blood by negative selection using a magnetic bead kit (Miltenyi Biotech). Purified NK cells were resuspended in RPMI + 10% FBS, 2 mM L-glutamine in the minimum volume required for the assay. 100 μl / well of NK cells were added on top of the target cells and antibodies to the appropriate wells of the assay plate. Effector:target ratios ranged from 10:1 to 3:1 depending on donor NK cell numbers.

[0507] Incubate the assay plate at 37°C in 5% CO 2 The cells were incubated at 300g for approximately 3 hours. After 3 hours, the number of live target cells was measured by flow cytometry. The assay plate was centrifuged at 300g for 3 minutes to pellet the cells, and each well was stained for the epithelial cell marker Epcam and the NK cell marker CD56. The staining antibodies (anti-Epcam PE and anti-CD56 BV421) were diluted 1 / 100 in cell staining buffer and 100 μL / well was added to each well. The plate was incubated at room temperature for 15 minutes. After staining, the cells were washed twice with 150 μl / well of PBS, and the plate was centrifuged at 300g for 3 minutes between each wash. At the end of staining, the cells in each well were resuspended in a final volume of 100 μL / well of PBS containing 50 nM TO-PRO™-3 cell viability dye.

[0508] After 10 minutes, BD FACS Canto II Instrument was used to take exactly 70μL samples from each well, and data was analyzed using FlowJoV10.60 software. The total number of live target cells was determined for each well. Live target cells were first identified as TO-PRO-™3 negative, then as CD56 negative and SSC high. Cells were then gated for Epcam (JEG3), or Epcam and GFP expression (HCT116). Percent depletion was calculated for each test sample compared to either untreated cells or isotype control, and data was transferred to GraphPad Prism 8.1.1 Software for analysis.

[0509] result: Figure 6 shows the percentage of Epcam+GFP+HCT116 target cells depleted after treatment with various anti-HLA-G antibodies or an IgG1 isotype control antibody. Each antibody was tested at two different concentrations, either 1 μg (white bars) or 0.01 μg / ml (striped bars). The E:T ratio was 3.5:1. Each bar represents the mean (and range) of three data points, and each dot / square is an individual replicate. Data are from one representative donor.

[0510] Table 18 lists the average depletion of Epcam+GFP+HCT116 cells by each antibody shown in FIG. 6 (ND: not detected). [Table 18]

[0511] Several antibodies showed similar mean % GFP+ cells depleted at the highest concentration of antibody (1 μg / mL). Antibodies HLA-G01 and HLA-G02 had the highest mean % GFP+ cells depleted observed at the lowest concentration (0.01 μg / mL). These antibodies were taken forward for further characterization in ADCC assays.

[0512] Figure 7 shows the percentage of Epcam+GFP+ HCT116 target cells depleted after treatment with anti-HLA-G antibodies HLA-G01 and HLA-G02 or an IgG1 isotype control antibody from three separate experiments (three different donors). Antibodies were tested at 1 µg / ml (Figure 7A) or 0.01 µg / ml (Figure 7B). E:T ratios ranged from 2.5 to 3:1. Each bar represents the mean (and range) of data from an individual experiment, and each dot, square or triangle is an individual replicate.

[0513] Table 19 lists the average percentage depletion of Epcam+GFP+HCT116 cells by 1 μg / mL of each antibody shown in FIG. 7A. [Table 19]

[0514] Table 20 lists the average depletion of Epcam+GFP+HCT116 cells by 0.01 μg / mL of each antibody shown in FIG. 7B. [Table 20] At both concentrations of antibody, antibody HLA-G02 had better cell depletion activity than HLA-G01, as determined by the ADCC assay.

[0515] FIG. 8 shows the percentage depletion of Epcam+GFP+HCT116 cells after treatment with titrations of anti-HLA-G antibodies HLA-G01 and HLA-G02 compared to isotype controls. The E:T ratio was 4:1. Each point represents the mean (and range) of three replicates. Data shown are from a single representative donor.

[0516] Based on these results, antibody HLA-G02 was selected for further characterization and development, in particular a defucosylated version of HLA-G02 was prepared as described above for comparison with conventional IgG1.

[0517] Defucosylation of antibodies has been shown to increase FcγRIII:Fc binding affinity and has been reported to increase the ADCC potential of IgG1 molecules. Therefore, we tested a defucosylated version of HLA-G02 for its ability to induce ADCC in transfected HCT116 or JEG3 cells. The defucosylated antibodies were compared to the same antibody V regions made in a conventional IgG1 format.

[0518] FIG. 9A shows the depletion rate of JEG3 cells after treatment with conventional HLA-G02 IgG1 (solid line) or defucosylated HLA-G02 IgG1 ("aF HLA-G02", dotted line). The E:T ratio was 10:1. Each point represents the mean (and range) of two replicates. Data from a single representative donor are shown.

[0519] FIG. 9B shows the depletion rate of Epcam+GFP+ HCT116 cells after treatment with conventional HLA-G02 IgG1 (solid line) or defucosylated HLA-G02 IgG1 ("aF HLA-G02", dotted line). The E:T ratio was 5:1. Each point represents the mean (and range) of three replicates. Data from a single representative donor are shown.

[0520] Antibody HLA-G02 was selected for further development as a candidate for therapeutic use because it showed good potency and efficacy in cell killing assays, and the defucosylated version of HLA-G02 had improved ADCC compared to its conventional (i.e., fucosylated) counterpart.

[0521] Example 10: Efficacy and potency of HLA-G antibody HLA-G02 in phagocytosis of HLA-G expressing cells HLA-negative K562 cells were used as targets and transfected to express HLA-G. Pre-labeled target cells (CTY+) were incubated with monocyte-derived macrophages (CD11b+) together with HLA-G02 antibody in hIgG1 format. Phagocytosis was analyzed by measuring the percentage of CTY+CD11b+ cells.

[0522] method: CD14+ monocytes were purified from peripheral venous blood using the Pan Monocyte Isolation Kit, Human (Miltenyi), an indirect magnetic labeling system for isolation of untouched monocytes. Cells were differentiated into macrophages with 50ng / ml recombinant MCSF in complete RPMI medium (RPMI 1640 medium + 10% fetal bovine serum, 2mM Glutamax and 1% penicillin / streptomycin) for 7 days at 37°C, 5% CO2.

[0523] HLA-negative erythroleukemia K562 cells were mock transfected or transfected with HLA-G and B2m using the 4D-Nucleofector System and the SF Cell Line 4D-Nucleofector™X Kit L (Lonza, ref. V4XC-2024) and cultured for 24 h at 37°C and 5% CO2 in complete RPMI. The next day, cells were harvested, washed, labeled with Cell Trace Yellow (Thermofisher), washed again and seeded at 25,000 cells / well in 100 μl complete RPMI in 96-well round-bottom Ultra low attachment plates (Corning Costar). Cells were subsequently incubated for 1 h at 37°C and 5% CO2 with either anti-CD47 or anti-HLA-G antibodies or isotype controls at 10 μg / ml. After washing, cells were combined with monocyte-derived macrophages (50,000 macrophages / well) at a macrophage:cell target ratio of 2:1. The mixed cells were incubated at 37°C, 5% CO2 for 2 hours. Cells were then washed and resuspended in PBS+10% purified human Fc gamma R binding inhibitor (Thermofisher) for 20 minutes at 4°C, and then stained with anti-CD11b-APC (Biolegend) for 20 minutes at 4°C. Cells were washed and resuspended in PBS+2mM EDTA+0.5% BSA in the presence of DAPI (500ng / ml) dead cell exclusion. Samples were acquired by flow cytometry using a BD FACSCanto. Analysis was performed by measuring the percentage of CTY+CD11b+ double positive cells using FlowJo software v10.6.0.

[0524] Anti-CD47 was used as a positive control. Studies have shown that blocking CD47 on target cells inhibits the interaction of CD47 with SIRPa, a receptor expressed on macrophages, leading to increased phagocytic activity. CD47 expression has been shown to be upregulated on tumor cells, and anti-CD47 antibodies are currently being tested in clinical trials. In phagocytosis assays, assessing the phagocytic activity of monocyte-derived macrophages is a good control, and it has been shown that both mock-transfected and HLA-G-transfected cells are able to induce phagocytosis.

[0525] result: Table 21 shows the HLA-G-specific phagocytic activity of anti-HLA-G antibodies against mock-transfected K562 compared to HLA-G-expressing K562 target cells (HLA-G / B2m K562). Data are expressed as percentage of CTY+CD11b+ double positive cells and are representative of three independent experiments in duplicate. [Table 21]

[0526] Table 22 shows statistically significant HLA-G specific phagocytic activity of HLA-G02 compared to IgG1 isotype control. Data represents pooled data from 6 donors (in duplicate). Data was exported to Excel and normalized to the average percentage of phagocytosis of mock transfected cells. [Table 22]

[0527] Figure 10 shows titration of HLA-G specific phagocytic activity of human anti-HLA-G IgG1 antibody HLA-G02 against mock transfected K562 target cells (Figure 10A) and HLA-G / B2m transfected K562 target cells (Figure 10B) compared to anti-CD47 antibody and isotype control. Data are expressed as percentage of CTY+CD11b+ double positive cells and are representative of one of two donors (Table 23). [Table 23]

[0528] Figure 11 shows titration of HLA-G specific phagocytic activity of conventional and defucosylated (aF) formats of HLA-G02 against mock-transfected K562 target cells (Figure 11A) and HLA-G expressing K562 target cells (Figure 11B) compared to anti-CD47 antibody and isotype control. Data are expressed as percentage of CTY+CD11b+ double positive cells and are representative of 1 of 3 donors in 2 independent experiments (Table 24). [Table 24] Defucosylated HLA-G02 exhibited improved ADCP compared to its conventional (ie, fucosylated) counterpart HLA-G02.

[0529] Example 11: Epitope mapping of the VR12389 antibody by X-ray crystallography Protein production of HLA-G fusion proteins Peptide B2mHLAG C42S mut tev10his-HLA-G(C42S) in which the cysteine ​​required for homodimerization was mutated to serine (signal peptide (bold) (cleaved after expression), peptide (underlined), GS linker (italic), B2m sequence, GS linker (italic), HLA-G sequence with C42S (shaded gray), tev cleavage site (bold and italic), GS linker (italic), 10 his tag),

number

[0530] The protein PeptideB2mHLAG C42S mut tev10his was expressed by transient transfection using the Expi293™ Expression System (Life technologies™) according to the manufacturer's protocol. Five days after transfection, cells were harvested and the supernatant was used immediately for purification. The supernatant containing the PeptideB2mHLAG C42S mut tev10his protein was applied to a Histrap NiExcel column. Unbound proteins and contaminants were washed with PBS, 500 mM NaCl, 20 mM imidazole, pH 7.4, and the PeptideB2mHLAG C42S mut tev10his protein was eluted with PBS, 500 mM NaCl, 500 mM imidazole, pH 7.4. Fractions containing purified PeptideB2mHLAG C42S mut tev10his protein were pooled and the his tag was removed by incubating the protein with tev protease at a ratio of 1:100 for 2 h at room temperature and 2 h at 4°C. The protein was concentrated and further purified by size exclusion chromatography on an S200 26 / 60 equilibrated with 20 mM Tris, 50 mM NaCl, pH 7.4 buffer. Fractions containing purified PeptideB2mHLAG C42S mut protein were pooled, concentrated to 2.94 mg / ml, and stored at -80°C in 1 mg aliquots.

[0531] The PeptideB2mHLAG C42S mut protein was characterized by SDS-PAGE and migrated to a position on the gel consistent with the predicted molecular weight (MW) of the glycosylated protein.

[0532] The amino acid sequence of the resulting protein, PeptideB2mHLAG C42S mut, used for complexation and crystal structure was as follows:

number

[0533] ·Fab purification VR12389 Rabbit Fab (VR12389) (light chain sequence represented by SEQ ID NO:9, and heavy chain sequence represented by SEQ ID NO:13) was expressed in ExpiCHO cells as a secreted protein. Light and heavy chain expression constructs were co-transformed at a molar ratio of 1:1. Secreted protein was purified by passing conditioned medium over protein G beads and eluted with 0.1 M glycine, pH 2.7. Fractions were neutralized by adding 2 M Tris-HCl, pH 8.5. Protein was dialyzed into PBS, pH 7.2, then concentrated to 5.62 mg / ml and stored at 4° C.

[0534] ·VR12389 PeptideB2mHLAG C42S mut protein with RbFab Fusion peptide_β2m_HLA-G was incubated with VR12389 RbFab at a molar ratio of 1:1.1 for 1 hour. The complex was then purified using a Superdex 200 16 / 600 column (GE Healthcare) using 10 mM Tris-HCl, 150 mM NaCl (pH 7.5) as running buffer. Fractions were analyzed by SDS-PAGE using NuPAGE 4-20% Tris-Glycine (Thermo), and the purest complex fraction was then concentrated to 10.4 mg / ml using an Amicon® Ultra-15 Centrifugal Filter Unit (Millipore).

[0535] VR12389 Crystallographic analysis of PeptideB2mHLAG C42S mut protein with RbFab Several commercial crystallization screens were used to identify crystallization conditions for the complex. These were performed in a sitting drop format using Swissci 96-well 2-drop MRC Crystallization plates (supplied by Molecular Dimensions, catalog number MD11-00-100). First, 75 μL of each crystallization condition within the screen was loaded into a reservoir using a Microlab STAR liquid handling system (Hamilton). Then, 300 nL of HLA-G / VR12389 complex and 300 nL of reservoir solution were dispensed into the wells of the crystallization plate using a Mosquito liquid handler (TTP LabTech). Crystals were identified in condition G4 of the ProPlex-HT96 screen (Molecular Dimensions), containing 2 M ammonium sulfate and 0.1 M Tris at pH 8.0. Crystals were frozen using a reservoir solution containing 25% glycerol as a cryoprotectant. Diffraction data were collected at Diamond Light Source. The structure was solved using molecular replacement in Phaser, with alternating cycles of automatic and manual refinement using Phenix.refine and Coot.

[0536] By superimposing the crystal structure of HLA-G complexed with VR12389 and the crystal structure of HLA-G complexed with ILT2 and ILT4 (as reported in the literature, see e.g. Q Wang et al., Cellular & Molecular Immunology, 2019 and Shiroishi, PNAS vol 103, No44, P 16412-16417, 2006), it was revealed that VR12389 prevents HLA-G from interacting with the ILT2 and ILT4 receptors by steric hindrance (Figure 12).

[0537] With contact distances <4 Å, the HLA-G epitope recognized by the VR12389 antibody includes HLA-G residues V194, F195, Y197, E198, Q224, Q226, D227, V248, V249, P250 and Y257.

[0538] With contact distances <5 Å, the HLA-G epitope recognized by the VR12389 antibody includes HLA-G residues V194, F195, Y197, E198, R219, Q224, Q226, D227, V248, V249, P250, E253 and Y257.

[0539] Example 12: Epitope mapping of the 12389 antibody by HDX-MS and NMR In contrast to crystallography, which is performed under static conditions, HDX-MS and NMR are techniques that make it possible to analyze interactions in solution and show allosteric or conformational changes that are not always evident by crystallography.

[0540] HDX-MS Materials and Methods Sample preparation and data acquisition For HDX-MS analysis, 12 μM human HLA-G ECD (SEQ ID NO: 110) was complexed with 36 μM 12389 antibody (expressed as either IgG1 or Fab) and incubated at 4° C. for 1 hour.

[0541] 4 μl of HLA-G or HLA-G complex was incubated at 25° C. 2 57 μL of 10 mM phosphate (pH 7.0) in O or D 2The deuterated samples were then incubated at 25° C. for 0.5, 2, 15, and 60 minutes. After reaction, all samples were quenched by mixing 1:1 with quench buffer (4 M guanidine hydrochloride, 250 mM Tris(2-carboxyethyl)phosphine hydrochloride (TCEP), 100 mM phosphate) at 1° C. The mixed solution had a final pH of 2.5. The mixture was immediately injected into a nanoAcquity HDX module (Waters Corp.) for digestion. Peptide digestion was then performed online using an Enzymatic online digestion column (Waters) in 0.2% aqueous formic acid at 20° C. and a flow rate of 100 μL / min. All deuterated time points, and non-deuterated controls, were performed in triplicate with blanks run between each data point.

[0542] Peptide fragments were then captured using an Acquity BEH C18 1.7 μM VANGUARD chilled precolumn for 3 min. Peptides were then eluted onto a chilled Acquity UPLC BEH C18 1.7 µM 1.0 × 100 using the following gradient: 0 min, 5% B; 6 min, 35% B; 7 min, 40% B; 8 min, 95% B; 11 min, 5% B; 12 min, 95% B; 13 min, 5% B; 14 min, 95% B; 15 min, 5% B (A: 0.2% HCOOH in HO, B: 0.2% HCOOH in acetonitrile). Peptide fragments were ionized by positive electrospray into a Synapt G2-Si mass spectrometer (Waters). Data acquisition was performed in ToF-only mode over the m / z range of 50–2000 Th using the MSe method (low collision energy, 4 V; high collision energy: gradient from 18 V to 40 V). The Glu-1-fibrinopeptide B peptide was used for internal lock mass correction.

[0543] HDX-MS data processing MS obtained from non-deuterated control samples of HLA-G were used for sequence identification using Waters Protein Lynx Global Server 2.5.1 (PLGS). EThe data were used to perform a peptide search against a database of HLA-G sequences only, with a precursor intensity threshold of 500 counts and three matched product ions required for assignment. The ion accounting files of the three control samples were combined into a peptide list that was imported into Dynamx v3.0 software.

[0544] Peptides were subjected to further filtering in DynamX. Filtering parameters used were minimum and maximum peptide sequence length of 4 and 25, respectively, minimum intensity of 1000, minimum MS / MS products of 2, minimum products per amino acid of 0.2, and maximum MH+ error threshold of 10 ppm. DynamX v3.0 was used to quantify the isotopic envelope resulting from deuterium incorporation for each peptide at each time point. Additionally, full spectra were inspected and visually confirmed to ensure accurate assignment of m / z peaks, and only peptides with high signal-to-noise ratios were used for HDX-MS analysis.

[0545] After manual filtering in Dynamx, statistical analysis and filtering was performed using Deuteros, which uses the statistical analysis published by Houde et al., 2011. Deuteros generates a Wood plot that displays peptide length, start and end residues, comprehensive coverage, and a y-axis metric that is absolute uptake (in Daltons), which is the difference in uptake in the presence of ligand (bound) and the apo form. The Wood plot first applies confidence filtering to all peptides at each time point. Peptides with different deuteration outside the selected confidence limits are not significant.

[0546] result: In the presence of VR12389, a total of 12 peptides were observed to show a statistically significant decrease in deuterium uptake upon antibody binding in HLA-G, of which 9 showed the majority of protection. The majority of protection encompassed residues 178-196 (MLQRADPPKTHVTHHPVFD) and 214-230 (ILTWQRDGEDQTQDVEL). Both regions are within the α3 domain (and the terminal 5 residues of α2). A region showing moderate protection upon antibody binding encompassed residues 234-249 (RPAGDGTFQKWAAVVV), likely due to conformational changes. Peptides showing similar exchange patterns in the presence and absence of antibody have insignificant deuterium uptake. [Table 25]

[0547] In conclusion, from HDX-MS with 30 s deuterium incubation, the potential binding domains of VR12389 are 178-MLQRADPPKTHVTHHPVFD-196 and 214-ILTWQRDGEDQTQDVEL-230.

[0548] Nuclear magnetic resonance (NMR) spectroscopy The epitope mapping of antibody VR12389 was determined by NMR spectroscopy using the Fab fragment of the antibody.

[0549] material HLA-G α3 domain 2 H / 13 C / 15 N-labeled expression 1 μg of HLA-G α3 short N-His ATUM#393044 (HLA-G residues: 207-300) was used to transform BL21(DE3) competent E. coli (New England BioLabs #C2527H) by standard heat shock. The transformed cells were plated on LB agar plates containing 100 μg / ml carbenicillin and incubated overnight at 37°C. The next day, a single colony was used to inoculate 10 ml LBroth (Merck #C1389) containing 100 μg / ml carbenicillin and grown for 5 hours at 37°C with shaking at 200 RMP (New Brunswick Excella E25). 1 ml of the starter culture was then used to inoculate 500 ml of 2 H / 13 C / 15 N-labeled minimal medium (described below) was inoculated and grown overnight in single-use 2 L Erlenmeyer flasks (VWR #734-1904) at 37°C with shaking at 200 RPM. The optical density (OD) of the overnight culture was measured the next day. 600 ) was recorded (Amersham Biosciences Ultrospec 3100 pro). A final OD of 0.1 was then obtained. 600 Expression cultures were inoculated with overnight cultures until 10 min.

[0550] O.D. of 0.9 600 in 500 ml batches in single-use 2 L Erlenmeyer flasks at 37 °C with shaking at 200 RMP until 2 H / 13 C / 15 N-labeled minimal medium expression cultures were grown. HLA-G α3 expression was then induced with 500 μM IPTG (Sigma # I6758). The induced cultures were then left for an additional 4 hours at 37° C. before being harvested by centrifugation at 7,000 g for 30 minutes (Beckman Coulter J6-MI). The harvested pellets were then frozen at −20° C. prior to cell lysis.

[0551] Unlabeled expression of human β2m BL21(DE3) competent E. coli was transformed with 1 μg of human β2m (residues: 21-119) ATUM#358573 and grown as above. The next day, a single colony was used to inoculate 10 ml LBroth (10 g / L tryptone, 5 g / L yeast extract, 5 g / L NaCl, 1 mM NaHO) containing 100 μg / ml carbenicillin and grown for 5 hours at 37°C with shaking at 200 RPM. 1 ml of the starter culture was then used to inoculate 500 ml LBroth containing 100 μg / ml carbenicillin and grown overnight in a single-use 2L Erlenmeyer flask at 37°C with shaking at 200 RPM. The following day, the OD of the overnight culture was measured. 600 A final OD of 0.1 was then recorded. 600 Expression cultures were inoculated with overnight cultures until 10 min.

[0552] OD 3.0 600 The 2x TY (tryptone 16g / L, yeast extract 10g / L, NaCl 5g / L) expression culture was grown again in 500 ml batches in single-use 2L Erlenmeyer flasks at 37°C with shaking at 200 RMP until a fermentation rate of 100% was reached. The incubator temperature was then reduced to 17°C. After 30 minutes, the culture was replenished with 20x feed solution (1M MOPS pH 7.2, 20 mM MgCl 2 , 20 mM MgSO 4 The cells were fed with 100 mM NaCl (20% glycerol) and expression was induced with 150 μM IPTG. The induced cultures were then left at 17° C. for 16 h before being harvested by centrifugation (7,000 g, 30 min). The harvested pellets were then frozen at −20° C. prior to cell lysis.

[0553] bacterial cell lysis The protocol for purification and refolding was adapted from Craig S. Clements et al. The production, purification and crystallization of a soluble heterodimeric form of a highly selected T-cell receptor in its unliganded and liganded state. Biological Crystallography, 2002.

[0554] Lyse the cell pellet in lysis buffer: 50 mM Tris pH 8.0, 1% (v / v) Triton X-100, 1% (w / v) sodium deoxycholate, 100 mM NaCl, 10 mM DTT, 1 mg DNAse I (Biomedicals), 5 mM MgCl 2, and dissolved in cOmplete protease inhibitor (Roche). After 10 min of continuous stirring at room temperature, 10 mM EDTA was added. The resuspended cell pellet was then passed through a CF Cell Disrupter (Constant systems) three times at 4 °C using a pressure of 20 psi. The lysate was then clarified by centrifugation at 48,000 g for 1 h at 4 °C (Beckman Coulter Avanti JXN-26). The insoluble pellet was then washed twice with wash buffer 1: 50 mM Tris pH 8.0, 0.5% (v / v) Triton X-100, 100 mM NaCl, 1 mM EDTA, 1 mM DTT, 0.2 mM cOmplete protease inhibitor. Between each wash, the resuspended inclusion bodies were centrifuged at 48,000 x g for 30 min. After the second wash, the inclusion bodies were washed a final time in Final Wash Buffer: 50 mM Tris pH 8.0, 1 mM EDTA, 1 mM DTT, cOmplete protease inhibitor. The purified inclusion bodies were then redissolved in 20 mM Tris pH 8.0, 8 M urea (Sigma #U5378), 0.5 mM EDTA, 1 mM DTT. The purified fractions were analyzed by SDS-PAGE using NuPAGE 4-12%, Bis-Tris (Thermo #NP0322) and NuPAGE MES SDS Running Buffer (Thermo #NP0002) stained with Quick Coomassie Stain.

[0555] 2 H / 13 C / 15 N HLA-G α3 and untagged human β2m refolding The redissolved inclusion bodies were then diluted with 1.5 M guanidine HCl, 5 mM sodium acetate, 5 mM EDTA to approximately 1 mg / ml before refolding.

[0556] Refolding buffer: Tris pH 8.5, 0.4M arginine, 0.5mM oxidized glutathione, 5mM reduced glutathione, 2mM EDTA. Human β2m was first added dropwise. Then, in a molar ratio of 1:1,2 H / 13 C / 15 N-labeled HLA-G α3 domain was added. The refolding reaction was left at room temperature for 16 hours with gentle agitation. The refolding reaction was then dialyzed in dialysis buffer: 5 mM Tris pH 8.5 with a 1:20 dilution factor using a 3,000 MWCO Spectra / Por dialysis membrane. The dialysis buffer was exchanged once during the 24 hours of dialysis at room temperature.

[0557] Purification of the folded complex The refolded HLA-G α3 / β2m complex was then purified using an AKTA Pure (GE Healthcare) system and a HiTrap Q column (Cytiva Life Sciences) using the following buffers and purification sequence: Buffer A: 10 mM Tris, 10 mM NaCl pH 8.5 Buffer B: 10 mM Tris, 500 mM NaCl pH 8.5. Purification sequence: run at 5 ml / min, equilibrate 5 CV buffer A, load dialyzed refolding reaction, wash 10 CV buffer A, elution: 0-40% B for 10 CV, hold at 40% for 10 CV, 40-100% B for 20 CV, hold at 100% B for 10 CV.

[0558] Fractions were analyzed by SDS-PAGE using NuPAGE 4-12%, Bis-Tris (Thermo) and NuPAGE MES SDS Running Buffer (Thermo) stained with Quick Coomassie Stain (VWR #SERA35081.01). Pure fractions were pooled and then concentrated using a 10,000 MWCO Amicon Ultra (Millipore) and run in 150 mM NaCl, 10 mM Tris pH 7.4, 0.02% NaN as running buffer. 3The purified sample was then loaded onto an S75 300 / 10 increasing gel filtration column (Cytiva Life Sciences) containing 100% DMSO. Fractions were again analyzed by SDS-PAGE and pure fractions were pooled. The final purified sample was again analyzed by SDS-PAGE and then concentrated to approximately 350 μM. Protein concentrations were determined using a Thermo Scientific Nanodrop 2000 spectrophotometer.

[0559] Purification of Fab Reagent: Fab reagents were purified using an AKTA Pure system (GE Healthcare) and packed Gammabind Plus Sepharose (Cytiva Life Sciences) columns. Supernatants were concentrated to >300mg / L using an AKTA Flux system (GE Healthcare) prior to capture. The following buffers and purification sequence were used: Buffer A: 10mM PBS pH7.4. Buffer B: 0.1M Glycine-HCl pH2.7. Equilibrate 5CV Buffer A. Load supernatant at a flow rate that ensures at least 20 minutes contact time. Wash 5CV Buffer A. Elute 100% Buffer B 5CV. Elution fractions were neutralized with 2M Tris pH8.5.

[0560] Fractions were analyzed by SDS-PAGE using NuPAGE 4–20% Tris-Glycine and NuPAGE MOPS SDS Running Buffer (stained with Quick Coomassie Stain). Pure fractions were pooled and then concentrated using a 10,000 MWCO Amicon Ultra (Millipore) and loaded onto an S200 26 / 60 filtration column (Cytiva Life Sciences) with 10 mM PBS pH 7.4 as running buffer. Fractions were analyzed by SDS-PAGE and Acquity UPLC H-Class System (Waters) again using a BEH200 SEC column (Waters). Pure fractions were pooled and then concentrated to >5 mg / ml. Protein concentration was determined using a Thermo Scientific Nanodrop 2000 spectrophotometer.

[0561] Backbone assignment of HLA-G α3 To obtain backbone assignments for the HLA-G α3 domain, we used 150 mM NaCl, 10 mM Tris pH 7.4, 0.02% NaN 3 500 μl of HLA-G α3 complexed with unlabeled β2M at a concentration of 320 μM 2 H / 13 C / 15N-labeled samples were prepared and transferred to 5 mm NMR tubes. All experiments were recorded at 35 °C using either a 600 MHz Bruker AVIII-HD or a 600 MHz Bruker AVANCE NEO spectrometer equipped with a cryogenically cooled probe. 3D TROSY-HNCACB(Wittekind and Mueller,1993 HNCACB,a High-Sensitivity 3D NMR Experiment to Correlate Amide-Proton and Nitrogen Resonances with the Alpha-and Beta-Carbon Resonances in Proteins.J.Magn.Reson.Ser.B 101,201-205.doi:10.1006 / jmrb.1993.1033;Salzmann et.al.,1999.TROSY-type Triple Resonance Experiments for Sequential NMR Assignment of Large Proteins.J.Am.Chem.Soc.121,844-848.doi:10.1021 / ja9834226) and 3D TROSY-HNCOCACB (Salzmann et.al.,1999,TROSY-type Triple Resonance Experiments for Sequential NMR Assignment of Large Proteins. J. Am. Chem. Soc. 121, 844-848; Eletsky et.al., 2001. TROSY NMR with partially deuterated proteins. J. Biomol. NMR 120, 177-180) were used to perform sequential connections between backbone NMR signals of residues in proteins. 13 C. 15 N and 1 TROSY-HNCACB was recorded in the H dimension with spectral widths of 75, 36 and 16 ppm and acquisition times of 9 (F1), 21 (F2) and 70 (F3) ms (16 scans per increment, relaxation delay 1.3 s, total acquisition time 5 days), respectively. 13 C.15 N and 1 TROSY-HNCOCACB was recorded in the H dimension with spectral widths of 75, 36 and 16 ppm and acquisition times of 9 (F1), 21 (F2) and 70 (F3) ms (8 scans per increment, relaxation delay 1.3 s, total acquisition time 2 days 17 h). NMR spectra were processed using NMRPipe (Delaglio et al., 1995 NMRPipe: a multidimensional spectral processing system based on UNIX pipes. J. Biomol. NMR 6, 277-93). Data analysis was performed using Sparky (Goddard and Kneller, DGSPARKY 3. In., University of California, San Francisco), resulting in the assignment of amide protons and nitrogen residues for 80 residues, which correspond to 93% of the residues in the native protein (excluding proline residues).

[0562] Mapping of Fab fragment binding sites 150 μM of unlabeled β2M complexed with 10% molar excess of unlabeled Fab 2 H / 13 C / 15 The N HLA-G α3 sample was used to map the binding site of the Fab fragment. 200 μl of the sample was prepared in the same buffer as above for backbone assignment of HLA-G α3 and transferred to a 3 mm NMR tube. TROSY (Pervushin et.al., 1998. Single Transition-to-single Transition Polarization Transfer (ST2-PT) in [ 15 N, 1 H]-TROSY. J. biomol. NMR 12, 345-348) spectra with equivalent control TROSY experiments of HLA-G α3 / β2M. 1 H and 15 The chemical shift changes of N were determined. 15 N and1 Control TROSY experiments for HLA-G α3 / β2M were recorded in the H dimension with spectral widths of 36 and 16 ppm and acquisition times of 60 (F1) and 80 (F2) ms (8 scans per increment, relaxation delay of 1.5 s, total acquisition time of 1 h), respectively. 15 N and 1 TROSY experiments of HLA-G α3 / β2M / Fab were recorded in the H dimension with spectral widths of 36 and 16 ppm and acquisition times of 60 (F1) and 80 (F2) ms (8 scans per increment, relaxation delay of 1.5 s, total acquisition time 1 h). Spectra were processed using NMRPipe (Delaglio et al., 1995 NMRPipe: a multidimensional spectral processing system based on UNIX pipes. J. Biomol. NMR 6, 277-93). Data analysis was performed using Sparky.

[0563] The chemical shift changes were analyzed using the minimum shift method (Williamson et al., 1997, Mapping the binding site for matrix metalloproteinase on the N-terminal domain of the tissue inhibitor of metalloproteinases-2 by NMR chemical shift perturbation. Biochemistry 36, 13882-9) using the following formula to calculate the combined chemical shift change (Δδ):

number

[0564] To identify the Fab binding site on HLA-G α3, a histogram of the combined minimum shift versus protein sequence was used to identify regions of HLA-G α3 that contained significantly perturbed signals. If the size of the combined minimum shift change for an amino acid exceeded a threshold value of the average of the combined chemical shift changes for all amino acids, these residues were selected for further evaluation as possible contact residues in the Fab binding site.

[0565] Two thresholds were applied to identify residues bound by the Fab: those whose minimum shift exceeds the mean of all calculated shifts, and those whose minimum shift exceeds the mean of all calculated minimum shifts plus one standard deviation. In these analyses, proline residues cannot be identified because they do not contain an amide proton.

[0566] result: When stringency is defined as increasing as the average of all calculated shifts is exceeded (>0.0764), the epitope determined by NMR includes residues T200, L201, L215, W217, R219, D220, E229, A245, A246, V247, V249, S251, E253, Q255, T258, H260, V261, and W274. When stringency is defined as increasing as the average of all calculated shifts is exceeded (>0.1597), the epitope determined by NMR includes residues H191, Y197, E198, R202, L230, V248, G252, C259, and K275.

[0567] Example 13: Characterization of antibody molecules by liquid chromatography-mass spectrometry (LC-MS). The molecular weights (MW) of HLA-G02 (VR12389gL2gH16) (unmodified (fucosylated) and defucosylated), VR12389 gL2gH15 and HLA-G01 were measured by separate heavy and light chains (reduced) by LC-MS using a Waters ACQUITY UPLC System equipped with a Xevo G2 Q-ToF mass spectrometer. Samples (approximately 5 μg) were reduced with 5 mM tris(2-carboxyethyl)phosphine (TCEP) in 150 mM ammonium acetate for 40 min at 37°C. The LC column was a Waters BioResolve™ RP mAb Polyphenyl, 450 Å, 2.7 μm, equilibrated with 95% solvent A (water / 0.02% trifluoroacetic acid (TFA) / 0.08% formic acid) and 5% solvent B (95% acetonitrile / 5% water / 0.02% TFA / 0.08% formic acid) at a flow rate of 0.6 mL / min and held at 80° C. Proteins were eluted with a gradient of 5% to 50% solvent B over 8.8 min, followed by washing and re-equilibration with 95% solvent B. UV data was acquired at 280 nm. MS conditions were as follows: ion mode: ESI positive ion, resolved mode, mass range: 400-5000 m / z, and external calibration with NaI.

[0568] Data was analyzed using Waters MassLynx™ and MaxEnt Software.

[0569] As shown in Table 26, the predicted MW from the sequence matched the MW observed for the heavy and light chains by LC-MS for the whole antibody, and for VR12389 gL2gH16, there was a mass difference of approximately 146 Da for the heavy chain of the corresponding defucosylated version, as expected. [Table 26]

[0570] Example 14: Thermal stability (Tm) measurements Thermal shift assays were used to determine the melting temperature (Tm), or the temperature of the midpoint of unfolding, to assess the conformational stability of the molecules and therefore their robustness to manufacturing, and long-term stability.

[0571] The fluorescent dye SYPRO® Orange was used to monitor the protein unfolding process by binding to hydrophobic regions that become exposed as the temperature increases. The reaction mixture contained 5 μL of 30× SYPRO® Orange Protein Gel Stain (Thermofisher scientific, S6651) diluted from a 5000-fold concentration with the test buffer. 45 μL of sample at 0.2 mg / mL in PBS pH 7.4 was added to the dye and mixed. 10 μL of this solution was dispensed in quadruplicate into a 384 PCR optical well plate and run using a QuantStudio 7 Real-Time PCR System (Thermofisher™). The PCR system heating device was set to 20° C. and ramped to 99° C. at a rate of 1.1° C. / min. The fluorescence change in the wells was monitored by a charge-coupled device. The increase in fluorescence intensity was plotted and the inflection point of the slope was used to generate the apparent midpoint temperature (Tm). The data are shown in Table 27. [Table 27]

[0572] The samples showed high and similar thermal stability, suggesting no substantial structural differences between the grafts. The thermal stability of the Fab domain can vary considerably (typical range is 70°C-84°C), with high thermal stability being preferred due to its high mechanical stability. As expected, no significant structural differences were observed between conventional (i.e., fucosylated) and defucosylated HLA-G02.

[0573] Example 15: Experimental isoelectric point (pI) determination The pI was experimentally determined using an iCE3™ whole-capillary imaging capillary isoelectric focusing (cIEF) system (ProteinSimple™).

[0574] The experimental pI of the main peak was found to be similar for HLA-G01 and HLA-G02. The pI was within the range predicted to be good for the manufacturing process and formulation buffer. The presence of small amounts of acidic and basic charged species was consistent with other IgG1 therapeutic molecules and could be attributed to common post-translational modifications.

[0575] Example 16: Solubility measurements using the polyethylene glycol (PEG) aggregation assay As a mimic of high concentration solubility, a PEG aggregation assay was used. PEG is a non-adsorbing, non-denaturing polymer, and due to its inert nature, it has been used to promote protein precipitation primarily through excluded volume effects. Samples were exposed to increasing concentrations of PEG 3350, and the amount of sample remaining in solution was determined by plotting the absorbance at A280 nm. Determination of the % PEG concentration at which half of the sample precipitated generated a PEG midpoint (PEGmdpnt) score. This score allows antibody molecules to be ranked based on their apparent native state aggregation propensity, with a lower PEGmdpnt score (e.g., ≦10) indicating a greater propensity for native state aggregation.

[0576] Stock 40% PEG 3350 solutions (w / v) were prepared in PBS pH 7.4 and pH 5.0 buffer (a common preformulation storage buffer). Serial titrations were performed by an ASSIST PLUS liquid handling robot (INTEGRA™ 4505) to yield PEG 3350 ranging from 40% to 15.4%. Sample preparation consisted of mixing antibody and PEG solutions in a 1:1 volumetric ratio to minimize non-equilibrium precipitation. 35 μL of PEG 3350 stock solution was added by the liquid handling robot to a 96-well v-bottom PCR plate (A1-H1). 35 μL of 2 mg / mL sample solution was added to the PEG stock solution to yield a test concentration of 1 mg / mL, and final PEG 3350 concentrations ranging from 20% to 7.7%. The solutions were mixed by automated slow repetitive pipetting and incubated at 37 °C for 0.5 h to redissolve non-equilibrium aggregates. Samples were then incubated at 20 °C for 24 h. The sample plate was then centrifuged at 4000×g for 1 hour at 20° C. 50 μL of the supernatant was dispensed into a UV-Star®, half-area, 96-well, μClear®, microplate (Greiner, 675801). Protein concentrations were determined by UV spectrophotometry at 280 nm using a FLUOstar® Omega multidetection microplate reader (BMG LABTECH™). The resulting values ​​were plotted using Graphpad Prism version 7.04 and the midpoint of a sigmoidal dose-response (variable slope) fit to the PEG midpoint (PEG mdpnt ) scores were derived.

[0577] Data showing that a higher PEG midpoint (%) corresponds to a higher concentration solubility is shown in Table 28. The NB* sample showed signs of aggregation at the lowest tested concentration of PEG 3350 (7.7%), therefore an accurate PEG midpoint could not be generated. [Table 28]

[0578] PEG aggregation assay data showed that all samples tested showed moderate aggregation tendencies in PBS. Notably, HLA-G01 showed a very high aggregation tendency at pH 5.0. In contrast, VR12389 grafts showed low aggregation tendencies in pH 5.0 buffer. No significant differences were observed between conventional (i.e., fucosylated) HLA-G02 and its defucosylated counterpart.

[0579] Example 17: kD interaction parameter measurements (colloid stability) Colloidal stability was assessed using the kD interaction parameter, with positive and negative values ​​relating to repulsive and attractive intermolecular forces, respectively.

[0580] Dynamic light scattering (DLS) was performed using a DynaPro III plate reader (Wyatt Technology Corp, Santa Barbara, CA, USA). Samples were diluted to 30 μL in PBS, pH 7.4 or pH 5 buffer in 1 mg / mL increments from 7 mg / mL to 1 mg / mL. The wells containing buffer were selected as the solvent offset and measurements were performed at 25 °C, the laser power was set to 20%, and auto attenuation was enabled. Each measurement was the average of 5, 5 s scans in triplicate (5 × 3). Diffusion coefficients were measured (Dm) and interaction parameters (kD) were calculated according to the following equation (where D 0 represents the diffusion coefficient at infinite dilution). Dm=D 0 (1+k D C) Formula: D given by the Debye plot at the Y-intercept 0 Slope = kD*D 0。

[0581] Diffusion coefficients were measured as a function of protein concentration and kD, which is used to assess colloidal stability, with positive and negative values ​​indicating repulsive and attractive intermolecular forces, respectively. For samples that exhibit attractive / self-association, the diffusion coefficient increases as a function of protein concentration, which is reflected by negative kD values. The data are shown in Table 29. [Table 29]

[0582] The kD interaction parameter was shown to be highly negative for HLA-G01 at pH 5.0 and less negative at pH 7.4, suggesting increased colloidal stability at physiological pH. The VR12389 graft had a slightly negative kD value, suggesting good colloidal stability at both pHs.

[0583] Example 18: Effect of mechanical stress on aggregation stability (aggregation assay) Proteins tend to unfold when exposed to an air-liquid interface, exposing their hydrophobic surfaces to a hydrophobic environment (air) and their hydrophilic surfaces to a hydrophilic environment (water). Agitation of a protein solution achieves a large air-liquid interface that can promote aggregation. This assay mimics the stresses molecules undergo during production (e.g., ultrafiltration) and serves to provide stringent conditions to try to distinguish between different antibody molecules.

[0584] Samples in PBS pH 7.4 or pH 5 buffer were stressed by vortexing using an Eppendorf Thermomixer Comfort™. Prior to vortexing, concentrations were adjusted to 1 mg / mL and absorbance at 595 nm was obtained using a Varian Cary 50-Bio Spectrophotometer® to establish a time 0 reading. Each sample was aliquoted into 1.5 mL conical Eppendorf® capped tubes (3×250 μL) and vortexed at 1400 rpm at 25° C. for up to 24 hours. Aggregation (turbidity) was monitored by measuring samples at 595 nm at 3 and 24 hours after vortexing using a Varian Cary® 50-Bio Spectrophotometer. Data are summarized in Table 30. [Table 30]

[0585] At 3 hours after vortexing, a low aggregation tendency (low absorbance at 595 nm) was observed for all antibody samples in both PBS pH 7.4 and pH 5 buffers. It was only possible to differentiate the samples and evaluate the buffer dependency at a longer time point (24 hours). At 24 hours, there was a slightly higher aggregation tendency in pH 5 buffer compared to PBS pH 7.4, as well as for HLA-G01 compared to the VR12389 grafted molecule. No important differences were observed between conventional (i.e. fucosylated) HLA-G02 and its defucosylated counterpart.

[0586] Example 19: Hydrophobic Interaction Chromatography (HIC) Hydrophobic interaction chromatography (HIC) separates molecules in order of increasing hydrophobicity. Molecules bind to a hydrophobic stationary phase in the presence of high concentrations of polar salts and desorb into the mobile phase as the concentration of salt decreases. The longer the retention time, the more hydrophobic the molecule.

[0587] Samples (2.0 mg / mL) were diluted 1:2 with 1.6 M ammonium sulfate and 30 μg (30 μL) of sample was injected into a Dionex ProPac™ HIC-10 column (100 mm×4.6 mm) in 100 mM phosphate pH 7.4 in series with an Agilent 1200 binary HPLC equipped with a fluorescence detector. Separation was monitored by autofluorescence (excitation and emission wavelengths, 280 nm and 340 nm, respectively). Samples were analyzed using gradient elution with buffer A (0.8 M ammonium sulfate 50 mM phosphate pH 7.4) and buffer B (50 mM phosphate pH 7.4) as follows: (i) hold at 0% B for 2 min, (ii) linear gradient (0.8 mL / min) from 0 to 100% B in 30 min, (iii) column was washed with 100% B for 2 min and re-equilibrated at 0% B for 10 min before the next sample injection. The column temperature was maintained at 20° C. Retention times (min) are shown in Table 31. [Table 31]

[0588] The molecule exhibited early elution times, suggesting low apparent hydrophobicity. Low hydrophobic potential has been reported to be a desirable property and may indicate increased developability due to reduced aggregation tendency (Jarasch A et al. 2015). No significant differences were observed between the conventional (i.e., fucosylated) HLA-G02 antibody and its defucosylated counterpart.

[0589] Example 20: HLA-G tissue cross-reactivity Using an antibody optimized for staining frozen tissues that recognizes a similar epitope as HLA-G02 ("HLA-G Ab"), we investigated the expression pattern of HLA-G in normal non-tumor tissues and surprisingly found that forms of HLA-G containing the epitope bound by HLA-G Ab (and therefore the epitope bound by HLA-G02) were not expressed in healthy tissues, particularly pancreatic and pituitary tissues.

[0590] This is in contrast to what has been reported previously in the literature, where expression of HLA-G protein in pancreatic islets was reported by Cirulli et al. (Cirulli et al, DIABETES, Vol. 55, May 2006), who observed a marked upregulation of HLA-G in pancreatic islet cells cultured on an extracellular matrix that supports cell replication. Gene expression of HLA-G in the pituitary gland, as well as in pancreatic islets and testes, for example, has also been reported by Boegel et al. (Boegel et al, BMC Medical Genomics (2018) 11:36).

[0591] method: Using the above HLA-G Abs that specifically bind to HLA-G, a tissue cross-reactivity study in human tissues was performed. The purpose of this tissue cross-reactivity (TCR) study was to evaluate the potential cross-reactivity of HLA-G antibodies using FITC-conjugated HLA-G antibodies in frozen human tissues and blood smears using immunohistochemistry (IHC) techniques.

[0592] A panel of 42 different frozen normal human tissues and blood smears (3 donors per tissue) were evaluated. Two concentrations of HLA-G Ab-FITC were used, set at 3 and 10 μg / mL, and a negative control IgG1-FITC was used at the highest concentration of 10 μg / mL.

[0593] result: HLA-G Ab-FITC produced membranous and variable cytoplasmic staining in the extravillous trophoblast of the placenta, and because HLA-G is a major histocompatibility gene expressed almost exclusively in the extravillous trophoblast at the fetal-maternal interface (the extravillous trophoblast invades the decidua and maternal spiral arteries), this pattern was thought to represent on-target binding of HLA-G Ab (Goldman Whol, 2000).

[0594] In contrast, no positive staining was observed in the following tissues: adrenal gland, blood cells, bone marrow, breast, cecum, cerebellum, cerebral cortex, colon, duodenum, endothelium (blood vessels), eye, esophagus, fallopian tube (fallopian tube), gallbladder, heart, ileum, jejunum, kidney, liver, lung, lymph node, muscle, nerve, ovary, pancreas, parotid gland, parathyroid gland, pituitary gland, prostate, rectum, skin, spinal cord, spleen, stomach, testis, thymus, thyroid, tonsils, ureter, bladder, and uterus (cervix and endometrium).

[0595] In conclusion, the form of HLA-G containing the epitope bound by HLA-G Ab was found to be expressed exclusively in extravillous trophoblast cells (as reported in the literature and used as a control in the present study) and not in other normal tissues tested.

[0596] Thus, the results are surprising and, contrary to what was expected from the teachings of the prior art, show that antibodies against HLA-G are potential candidates for the treatment of solid tumors, for example, capable of killing cells expressing HLA-G via Fc-mediated effector functions, and that no toxicity to patients via binding to normal tissues is predicted. Empirically, there are potential hypotheses that may explain the unexpected differences in the results obtained with this assay compared to those reported in the literature: i) mRNA expression as reported in the pituitary gland does not necessarily indicate membrane protein expression. For example, the mRNA may not be translated into protein, or the mRNA may code for a soluble HLA-G isoform that may not be detected in frozen tissue and would not present a problem in terms of toxicity since it is not membrane bound. The mRNA may also be expressed by infiltrating immune cells and not pituitary cells, ii) the commercially available antibody 4H84 was used to detect HLA-G protein in the pancreas, and this antibody is known to be non-specific. 4H84 also recognizes an epitope in the α1 domain of HLA-G, but the antibody of the present invention is highly specific for HLA-G and binds to the α3 domain. Thus, HLA-G isoforms expressed in the pancreas do not contain the α3 domain, but may still be detected by α1 binders (e.g., HLA-G3, HLA-G4, HLA-G7). The positive detection of HLA-G in placental trophoblasts and the lack of detection in normal tissues suggests that the antibody of the present invention may be able to bind to HLA-G protein expressed in tumors, which contains the ILT2 / 4 binding α3 domain required for the immunoregulatory function of HLA-G, but not to cells in normal tissues.

[0597] The dual mechanism of such antibodies described herein, capable of blocking the interaction between HLA-G and its inhibitory receptor and capable of killing cells, represents a considerable advantage for the treatment of patients with, for example, upregulation of HLA-G in solid tumors.

[0598] Example 21: Functional characterization of HLA-G antibodies in a 3D tumoroid model Nilogen Oncosystems' (Tampa, Florida, US) 3D tumoroid model technology was used to evaluate HLA-G antibody activity in a primary ex vivo human tumoroid platform. Nilogen's technology uses fresh patient-derived tumor tissue and results in the generation of 3D tumor organoids that retain an intact tumor microenvironment, including infiltrating immune cells, and capture complete tumor heterogeneity. This technology captures patient response rates similar to those seen in clinical trials, thus providing a useful model to evaluate the potential of immunotherapy candidates for the treatment of patients.

[0599] Ten colorectal adenocarcinoma tumors (CRC) and ten renal clear cell carcinoma tumors (RCC) were obtained and each was used to induce thousands of tumoroids containing all tumor cells (tumor cells, stromal cells, infiltrating immune cells) and matrix components using Nilogen's method.

[0600] Isolated tumoroids (100-400 per treatment well) were immediately exposed to the antibodies for 72 hours without prior culture. The level of tumor cell death was assessed at 24 and 72 hours using Nilogen's 3D-Explore™ imaging platform. Culture medium was collected at 24 hours for evaluation of cytokine levels. Tumoroids were disaggregated at 72 hours and analyzed by flow cytometry to evaluate the effect on immune cell activation profile. Additionally, prior to treatment, tumoroids were analyzed by flow cytometry to characterize their cellular composition and FFPE sections of each tumor were stained for HLA-G, ILT2 and ILT4.

[0601] The activity of HLA-G02 antibodies in an active IgG1 format was analyzed by comparing with an IgG1 isotype control.An anti-PD-L1 antibody with an active IgG1 format was used for comparison.

[0602] Tumor cell death Immediately after isolation, cultures containing approximately 100 tumoroids per well were cultured in the presence of 10 mg / ml of either isotype control IgG1, anti-PDL1 positive control or HLA-G02 IgG1. After 24 and 72 hours of culture, tumoroids were stained with a live / dead dye, imaged, and the percent dead cells were calculated using a proprietary algorithm.

[0603] The results are shown in Figure 13. Figure 13A: Data obtained with anti-PDL1 from RCC. Figure 13B: Data obtained with anti-PDL1 from CRC. Figure 13C: Data obtained with HLA-G02 from RCC. Figure 13D: Data obtained with HLA-G02 from CRC. Data are expressed as % dead cells for isotype control in light grey, anti-PDL1 or HLA-G02 in dark grey, and cultures treated with anti-PDL1 or HLA-G02; in black, a 1.5-fold or greater increase in cell death was observed.

[0604] Since each tumor may be different for each patient, the data fully represents the potential of the antibody of the present invention in the treatment of cancer, especially RCC and CRC.Overall, the data shows that the antibody of the present invention can kill tumor cells in the tumor environment under conditions where the response is entirely dependent on infiltrated immune cells.Advantageously, the data provides evidence that the antibody can have increased cell killing activity in certain tumors.

[0605] Of note, increased killing at 24 hours may no longer be detectable at 72 hours, as dead cells are lost from the culture over time. Furthermore, for the anti-PDL1 positive control, a total of five tumors showed increased cell killing at 24 hours (1xRCC and 4xCRC), and no increased killing was observed at 72 hours. For HLA-G02, six tumors showed increased killing at 24 hours (2xRCC and 4xCRC), and four tumors showed increased killing at 72 hours (1xRCC and 3xCRC). In total, only five of the twenty tumors showed increased killing with anti-PDL1, whereas nine of the twenty tumors showed increased killing with HLA-G02.

[0606] Thus, the data indicate that the antibodies of the invention may be advantageous in treating solid tumors compared to anti-immune checkpoint antibodies such as anti-PD-L1.

[0607] Example 22: Tissue cross-reactivity assay using defucosylated HLA-G02 Using the same method as described in Example 20, defucosylated HLA-G...

Claims

1. An antibody that specifically binds to HLA-G, comprising: a. i. CDR-L1 containing SEQ ID NO: 1, ii. CDR-L2 containing SEQ ID NO: 2, and iii. CDR-L3 containing SEQ ID NO: 3 in a light chain variable region, and b. i. CDR-H1 containing SEQ ID NO: 4, ii. CDR-H2 containing SEQ ID NO: 5, and iii. CDR-H3 containing SEQ ID NO: 6 in a heavy chain variable region, and an antibody comprising the same.

2. (a) An antibody that blocks HLA-G binding to ILT2 and ILT4, and / or inhibits HLA-G-mediated immunosuppressive function, and / or (b) has an equilibrium dissociation constant (KD) of less than 10 nM for HLA-G, and / or (c) specifically binds to the HLA-G alpha 3 domain, and / or (d) binds to an epitope of HLA-G containing residues F195 and Y197 with reference to SEQ ID NO: 107, and / or (e) binds to an epitope of HLA-G (SEQ ID NO: 107) containing V194, F195, Y197, E198, Q224, Q226, D227, V248, V249, P250, and Y257, optionally, the epitope being characterized by X-ray crystallography, and / or (f) a chimeric or humanized antibody according to claim 1.

3. (a) An antibody according to claim 1, comprising a light chain variable region containing SEQ ID NO: 19 or 15 or 23, and / or a heavy chain variable region containing SEQ ID NO: 93, 27, 33, 57, 69, 75, 81, or 87, or (b) an antibody according to claim 1, comprising a light chain variable region containing SEQ ID NO: 19 and a heavy chain variable region containing SEQ ID NO:

93.

4. A full-length antibody, optionally, the full-length antibody being IgG1, IgG1 LALA, IgG1LALAGA, IgG4, IgG4P, or IgG4P FALA, according to claim 1.

5. An antibody according to claim 1, which is IgG1, optionally, defucosylated IgG1.

6. (a) An antibody according to claim 1, comprising a light chain containing SEQ ID NO: 21 or 17, or 25, and / or a heavy chain containing SEQ ID NO: 95, 29, 35, 59, 71, 77, 83, or 89, or (b) a light chain comprising a sequence having at least 90% identity or similarity to SEQ ID NO: 21, 17, or 25, and / or a heavy chain comprising a sequence having at least 90% identity or similarity to SEQ ID NO: 95, 29, 35, 59, 71, 77, 83, or 89, or (c) comprising a light chain variable region comprising SEQ ID NO: 19 and a heavy chain variable region comprising SEQ ID NO: 93, wherein the remaining portions of the light chain and heavy chain each have at least 90% identity or similarity to SEQ ID NO: 21 and 95, respectively, or (d) comprising a light chain comprising SEQ ID NO: 21 and a heavy chain comprising SEQ ID NO: 95, The antibody according to claim 5.

7. The antibody according to claim 5, which inhibits HLA-G-mediated immunosuppressive function and inhibits tumor growth by depleting tumor cells expressing HLA-G.

8. An afucosylated IgG1, comprising a light chain comprising SEQ ID NO: 21 and a heavy chain comprising SEQ ID NO: 95, wherein the antibody has improved ADCC and / or ADCP and / or CDC functions and / or has an improved ability to deplete tumor cells expressing HLA-G. The antibody according to claim 5.

9. Fab, Fab', F(ab') ２ The antibody according to claim 1, which is dsFv, scFv or dsscFv.

10. An antibody that cross-competes with the antibody according to claim 1 for binding to HLA-G or binds to an epitope of HLA-G comprising V194, F195, Y197, E198, Q224, Q226, D227, V248, V249, P250, and Y257 of HLA-G (SEQ ID NO: 107).

11. An isolated polynucleotide encoding the antibody according to any one of claims 1 to 10, optionally, (A) a. The polynucleotide encodes a light chain variable region, and the polynucleotide is i. at least 90% identical to SEQ ID NO: 20, 16, or 24, or ii. comprises SEQ ID NO: 20, 16, or 24, or consists of SEQ ID NO: 20, 16, or 24, or b. The polynucleotide encodes a heavy chain variable region, and the polynucleotide is i. at least 90% identical to SEQ ID NO: 94, 28, 34, 58, 70, 76, 82, or 88, or ii. comprises SEQ ID NO: 94, 28, 34, 58, 70, 76, 82, or 88, or consists of SEQ ID NO: 94, 28, 34, 58, 70, 76, 82, or 88, or, (B) a. The polynucleotide encodes a light chain, and the polynucleotide is i. at least 90% identical to SEQ ID NO: 22, 18 or 26, or ii. includes SEQ ID NO: 22, 18 or 26, or consists of SEQ ID NO: 22, 18 or 26, or b. The polynucleotide encodes a heavy chain, and the polynucleotide is i. at least 90% identical to SEQ ID NO: 96, 30, 36, 60, 72, 78, 84 or 90, or ii. includes SEQ ID NO: 96, 30, 36, 60, 72, 78, 84 or 90, or consists of SEQ ID NO: 96, 30, 36, 60, 72, 78, 84 or 90, a polynucleotide.

12. A cloning vector or expression vector comprising one or more polynucleotides according to claim 11.

13. Comprising one or more polynucleotides according to claim 11, or one or more expression vectors comprising said polynucleotides, optionally A host cell that has been genetically modified to reduce or abolish the function of alpha1,6-fucosyltransferase.

14. A process for producing an antibody, comprising culturing the host cell according to claim 13 under conditions suitable for producing said antibody, and isolating said antibody, and optionally further comprising formulating said antibody into a pharmaceutical composition.

15. A pharmaceutical composition comprising the antibody according to any one of claims 1 to 10 and one or more pharmaceutically acceptable carriers, excipients or diluents.

16. The antibody according to any one of claims 1 to 10, or a pharmaceutical composition comprising said antibody, for use in therapy.

17. The antibody according to any one of claims 1 to 10, or a pharmaceutical composition comprising said antibody, for use in the treatment of a disease characterized by overexpression of HLA-G.

18. The antibody according to any one of claims 1 to 10, or a pharmaceutical composition comprising said antibody, for use in a method of treating a solid tumor in a patient, said method comprising administering a therapeutically effective amount of said antibody or said pharmaceutical composition to said patient, and optionally The solid tumor is selected from renal clear cell carcinoma (RCC), colorectal cancer (CRC), pancreatic cancer, ovarian cancer, head and neck cancer, gastric cancer and hepatocellular carcinoma, an antibody or pharmaceutical composition.

19. The antibody according to any one of claims 1 to 10, or a diagnostic composition comprising said antibody, for diagnosing renal clear cell carcinoma (RCC), colorectal cancer (CRC), pancreatic cancer, ovarian cancer, head and neck cancer, gastric cancer or hepatocellular carcinoma.

Images