Antibodies binding to death receptor 4 and death receptor 5

Abstract

The present invention relates to monoclonal antibodies that bind to death receptor 4 and / or death receptor 5, pharmaceutical compositions comprising the antibodies, and methods of treatment by administering the pharmaceutical compositions to patients.

Description

[0001] Divisional Application Instructions This application is a divisional application of the invention patent application filed on October 28, 2016, with application number 201680062988.1 and invention title "Antibody that binds to death receptor 4 and death receptor 5".

[0002] Cross-references to related applications

[0003] This application claims the benefit of US 62 / 248,782, filed October 30, 2015, which is incorporated herein by reference for all purposes. Technical Field

[0004] The present invention relates generally to a combination of monoclonal antibodies (mAbs) and recombinant DNA technologies for developing novel biological agents, and more specifically, to the production of monoclonal antibodies, for example, that bind to and activate death receptors 4 and 5. Background Technology

[0005] Tumor necrosis factor-associated apoptosis-inducing ligand (TRAIL, also known as Apo2 ligand, or Apo2L or Apo2L / TRAIL) is a member of the TNF ligand superfamily (reviewed in IAM van Roosmalen et al., Biochem Pharmacol 91:447-456, 2014). TRAIL is expressed in a stimulation-dependent manner on many cells of the immune system and modulates immune responses, for example, as a key effector molecule in NK cell-mediated cytotoxicity (C Falschlehner et al., Immunol 127:145-154, 2009). TRAIL induces the killing of susceptible cells by binding to cell membrane receptors death receptor 4 (DR4, also known as TRAIL-R1) and / or death receptor 5 (DR5, also known as TRAIL-R2 or Apo2) to activate the extrinsic apoptosis pathway. More specifically, this active, soluble form of TRAIL is a self-assembled, non-covalent homotrimer that binds to the extracellular domains of three receptor molecules with high affinity. This binding induces oligomerization of the intracellular death domain and the formation of homo- or hetero-polymeric complexes (FC Kischkel et al., Immunity 12: 611-620, 2000), subsequently recruiting Fas-associated proteins with the death domain (FADD) and forming the death-inducing signaling complex (DISC). This DISC leads to the activation of caspase, followed by apoptosis, and programmed cell death (see van Roosmalen et al., as cited above).

[0006] Many cancer cells express DR4 and / or DR5, and TRAIL can selectively induce apoptosis in cancer cells (reviewed in J Lemke et al., Cell Death Differ 21: 1350-1364, 2014) and in a cross-linked form in tumor endothelial cells (NSWilson et al., Cancer Cell 22: 80-90, 2012). TRAIL itself and various TRAIL receptor agonists have been shown to have strong antitumor activity against cancer cell lines and preclinical models (e.g., A Ashkenazi et al., J Clin Invest 104: 155-162, 1999). These include recombinant human TRAIL (rhTRAIL;dulanermin) composed of the extracellular region of human TRAIL (R Pitti et al., J. Biol Chem 271: 12687-12690, 1996) and many agonist monoclonal antibodies (mAbs) against DR4 or DR5, including murine, chimeric, humanized and human monoclonal antibodies (mAbs) (see van Roosmalen et al., as cited above). Such mAbs include 4H6.17.8 mAb (referred to as 4H6 in this paper; US Patent No. 7,252,994) and mapatumumab (HGS-ETR1; Pukac et al., Br J Cancer 92: 1430-1441, 2005) targeting DR4; and 3H3.14.5 mAb (referred to as 3H3 in this paper; US Patent No. 6,252,050), conatumumab (AMG 655; P Kaplan-Lefko et al., Cancer Biol Ther 9: 618-631, 2010), drozitumab (Apomab; C Adams et al., Cell Death Differ 15:751-761, 2008), and lexatumumab (HGS-ETR2; G Georgakis et al., Br J Haematol) targeting DR5. 130:501-510, 2005) and TRA-8 and its humanized form tigatuzumab (CS-1008, A Yada et al., Ann Oncol 19:1060-1067, 2008); and other mAbs listed by van Roosmalen (as previously cited, see Table 1 on page 450).

[0007] However, while all these drugs are highly effective at killing tumor cells in vitro and in typical animal models (see the references cited above), none of them have shown strong activity in clinical trials, either alone or in combination with other agents, and none have entered phase III trials (reviewed in PM Holland, Cytokine Growth Factor Rev 25:185-193, 2014 and Lemke et al., as cited above. See in particular Holland's Table 1 and Lemke et al.'s Tables 2 and 3, and the references cited therein). For mAbs, one reason may be the requirement to cross-link mAbs to oligomerize death receptors to trigger the apoptosis pathway. This requirement can be met in vivo by binding the Fc domain of mAbs to Fcγ receptors (FcγR) on immune cells, but the number of immune cells infiltrating tumors may be too small, or this binding may not have sufficiently high affinity. Cancer cells may also express inhibitors of apoptosis proteins (see Lemke et al., as cited above). To improve efficacy, modified forms of TRAIL are being developed, including fusion with other protein domains to enhance stability, oligomerization, and / or targeting (Lemke et al., as cited above, and Holland, as cited above). Similarly, constructs containing multiple antibody domains that bind DR4 and / or DR5 have been developed (K Miller et al., J Immunol 170:4854-4861, 2003; W Wang et al., Immunol Cell Biol 91:360-367, 2013; JE Allen et al., Mol Cancer Ther 11:2087-95, 2012; H Huet et al., Cancer Res 72abstract 3853, 2012; WO 2014 / 022592; WO2015 / 017822), but these constructs have not yet entered clinical trials or have not been successful in clinical trials (KP Papadopoulos et al., Cancer Chemother Pharmacol; Feb 27, 2015, PMID:25721064); their very unnatural structure may limit their clinical utility. In another approach, the combination or combination of TRAIL and anti-DR5 mAb AMG655 is more effective than either of them alone in vitro and in vivo (JD Graves et al., Cancer Cell 26: 177-189, 2014). Summary of the Invention

[0008] This invention provides monoclonal antibodies (mAbs) that bind to DR4 and DR5, such as D114 and G4.2, and their humanized forms. In one embodiment, the invention provides a bispecific monoclonal antibody comprising two pairs of heavy and light chains, wherein each heavy / light chain pair comprises a DR4-binding domain and a DR5-binding domain. In other embodiments, the invention provides a multimeric mAb having four DR4-binding domains or four DR5-binding domains. In a preferred embodiment, the antibody is in the form of a Bs(scFv)4-IgG antibody, the term as defined below. In any bispecific or multimeric mAb, each binding domain is preferably derived from a humanized or human mAb, such as the humanized forms of D114 and G4.2. Furthermore, any mAb of the present invention may contain a constant region comprising mutations that enhance binding to the cellular Fcγ receptor (FcγR), such as FcγRIIb, or, for example, the S267E and / or L328F mutations (according to the Kabat number based on the EU index) in the γ-1 constant region, preferably two or more such mutations. Advantageously, the mAb inhibits the growth of human tumor xenografts in mice. On the other hand, the present invention provides pharmaceutical compositions comprising any of these mAbs. A third aspect is the administration of such pharmaceutical compositions to patients to treat cancer or other diseases. Attached Figure Description

[0009] Figure 1A B. A simplified structural diagram of the Bs(scFv)4-IgG antibody form, showing individual variable and constant regions (A) or domains formed by folding each light chain region together with its respective heavy chain region (B). H 1 (or V) L 1) = The variable region of the heavy chain (or light chain) of the first antibody; for the V of the second antibody H 2 and V L The same applies to 2. C H 1, C H 2, C H 3 (or C) L ) = Heavy (light) constant region; H = Hinge region. V1 (or V2) = All variable regions of the first (or second) antibody.

[0010] Figure 2A -D. Figure 2A The combination of 4H6 and D114 mAbs with DR4 is shown; Figure 2B The results showed that 4H6 and D114 mAbs inhibited the binding of Apo2L to DR4; Figure 2C The combination of 3H3 and G4.2 mAbs with DR5 is shown; Figure 2DThe results showed that 3H3 and G4.2 mAbs inhibited the binding of Apo2L to DR5.

[0011] mlgG is a mouse negative control mAb.

[0012] Figure 3A - D. Cell viability of H60 lung tumor cells (A) and SW480 colon tumor cells (B) after treatment with 4H6 and D114 mAbs in the presence of goat anti-mouse IgG-Fc antibody (anti-mIgG-Fc); Cell viability of H60 lung tumor cells (C) and COLO 205 colon tumor cells (D) after treatment with 3H3 and G4.2 mAbs in the presence of anti-mIgG-Fc.

[0013] Figure 4A - D. The single-letter amino acid sequences of the mature heavy chain (A) and light chain (B) variable regions of 4H6mAb and the mature heavy chain (C) and light chain (D) variable regions of 3H3mAb (SEQ ID NOS: 1-4, respectively).

[0014] Figure 5 A-D. Alignment of the amino acid sequences of the mature variable regions of the HuD114-L1 light chain (A) and the HuD114-H1 and HuD114-H2 heavy chains (B) with the mouse D114 and human receptor V regions; alignment of the amino acid sequences of the mature variable regions of the HuG4.2-L1 and HuG4.2-L2 light chains (C) and the HuG4.2-H1 and HuG4.2-H2 heavy chains (D) with the mouse G4.2 and human receptor V regions. The sequences are (A) SEQ ID NOS.5-7, (B) SEQ ID NOS.8-11, (C) SEQ ID NOS.12-15, and (D) SEQ ID NOS.16-19, respectively. CDRs are underlined in the D114 (or G4.2) sequence. The CDRs -L1, -L2, -L3, -H1, -H2, and -H3 of D114 are represented as SEQ ID NOS.20-25, while those of G4.2 are represented as SEQ ID NOS.26-31. Amino acids in the HuD114 (or HuG4.2) sequence that are replaced by mouse D114 (or G4.2) amino acids are marked with double underscores. All light and heavy chains in the figures here use single-letter amino acid codes and the Kabat numbering system.

[0015] Figure 6AB. Single-letter amino acid sequences of the complete heavy chain (A) and light chain (B) of the bispecific antibody B-3H3 / 4H6-hFc** (SEQ ID NOs. 32 and 33), including a cleaved signal peptide that is therefore not present in the mature protein, which is SEQ ID NO: 34 and 35. Arrows under the sequences separate the regions, which are respectively the signal peptide (marked with strikethrough), 4H6 V H Zone (A) or 3H3 V H Zone (B), Connector Sequence (underlined), 4H6 V L Zone (A) or 3H3 V L Region (B) and heavy chain (A) or light chain (B) constant regions. Mutated amino acids 267E and 328F are marked with bold underlines.

[0016] Figure 7A B. Cell viability of SW480 colon tumor cells treated with the reagents was determined by WST-8 assay in the presence (B) or absence (A) of goat anti-hIgG-Fc. In the figure, hIgG is the control human mAb.

[0017] Figure 8A B. Cell viability of H460 lung tumor cells treated with the indicated reagents in the absence of goat anti-hIgG-Fc (A) and in the presence of human PBMCs (B).

[0018] Figure 9A (A) Cell viability of COLO 205 colon tumor cells after incubation with PBMCs from 4 donors and 0.5 μg / ml of the reagent shown. (B) Cell viability of the cells after incubation with PBMCs and 1.0 μg / ml of the indicator. The error bars in the figure represent the standard deviation of the mean (SEM).

[0019] Figure 10A (A) Cell viability of COLO 205 cells treated with the indicated reagent in the presence of PBMCs (short dashed line, 4H6-hFc; long dashed line, 4H6-hFc *). (B) Inhibition of growth of COLO 205 colon tumor xenografts by the indicated reagent.

[0020] Figure 11A -C. The reagent shown inhibits the growth of COLO 205 colon cancer xenografts (A, C) and SW480 colon cancer xenografts (B).

[0021] Figure 12AThe reagents shown in Figure B inhibited the growth of COLO 205 colon tumor xenografts (A) and MIA PaCa-2 pancreatic tumor xenografts (B).

[0022] Figure 13A Figure B shows the binding of anti-DR4 mAbs to DR4 (A) and anti-DR5 mAbs to DR5 (B). In all figures in this paper, the suffixes -Fc and -Fc** are synonymous with -hFc and -hFc**, respectively (e.g., HuD114-Fc#1 is synonymous with HuD114-hFc#1). In this figure and the figures below, as shown, solid lines generally represent the hFc** form of mAbs, dashed lines represent the hFc or hFc* form, and dotted lines represent mouse mAbs.

[0023] Figure 14A B. Cell viability of COLO205 colon tumor cells treated with anti-DR4 Ab (A) and anti-DR5 mAb (B) in the presence of goat anti-hIgG-Fc. Different concentration units are used in (A) and (B).

[0024] Figure 15A -D. Cell viability of COLO205 colon tumor cells (A), SW480 colon tumor cells (B), COLO 205 cells (C), and MIA PaCa-2 pancreatic tumor cells (D) after treatment with the indicated reagents in the presence of anti-mlgG-Fc (corresponding reagent is mouse TRA-8 mAb) or in the presence of goat anti-hIgG-Fc (corresponding reagent is humanized or human mAbs).

[0025] Figure 16A -B. Cell viability of H60 lung tumor cells (A) and SW480 colon tumor cells (B) after treatment with the indicated reagents in the presence of goat anti-hIgG-Fc.

[0026] Figure 17A -F. Cell viability of COLO 205 colon tumor cells (A, C, and E) and H60 lung tumor cells (B, D, and F) after treatment with the indicated reagents in the presence of human PBMCs.

[0027] Figure 18A -D. The reagents shown inhibited the growth of COLO 205 colon tumor xenografts (A), H60 lung tumor xenografts (B), COLO 205 xenografts (C), and Ramos lymphoma xenografts (D).

[0028] Figure 19AThe reagents shown in Figure B inhibited the growth of COLO 205 tumor xenografts (A) and MIA PaCa-2 pancreatic tumor xenografts (B).

[0029] Figure 20A - D. ELISA results of the various reagents shown in (A, C) binding simultaneously with DR4 and DR5. In (A), the curves of 4H6-hFc** and 3H3-hFc** overlap and cannot be distinguished. (B, D) Cell viability of COLO 205 cells treated with the shown reagents in the absence of goat anti-mlgG-Fc. Detailed Implementation

[0030] 1. Antibody As used herein, “antibody” refers to a protein containing one or more domains capable of binding antigens, wherein such domains are derived from or homologous to variable domains of natural antibodies. A monoclonal antibody (“mAb”) is a distinct antibody, as opposed to a mixture of different antibodies. Antibodies described herein are generally monoclonal unless otherwise stated. The “antigen” of an antibody is the compound that the antibody specifically binds to, typically a polypeptide, but can also be a small peptide or a small molecule hapten, a carbohydrate, or other fraction. Examples of antibodies include natural, full-length tetrameric antibodies; antibody fragments such as Fv, Fab, Fab', and (Fab')2; single-chain antibodies (scFv) (Huston et al., Proc. Natl. Acad. Sci. USA 85:5879, 1988; Bird et al., Science 242:423, 1988); single-arm antibodies (Nguyen et al., Cancer Gene Ther 10:840, 2003); and bispecific antibodies, chimeric antibodies, and humanized antibodies, terms which will be further explained below. Antibodies can be derived from any vertebrate species, including chickens, rodents (e.g., mice, rats, and hamsters), rabbits, camels, primates, and humans. Antibodies containing constant domains can be any known isotype, IgG, IgA, IgM, and IgE and their isotypes, such as human IgG1, IgG2, IgG3, IgG4, and mouse IgG1, IgG2a, IgG2b, and IgG3, and their allotypes and isoallotypes, including the arrangement of polymorphic site residues in allotypes and isoallotypes. Antibodies can also be chimeric isotypes, meaning that one or more of their constant regions (C) can contain regions of different isotypes, such as a γ-1C. H Region 1 and hinge region, with C from γ-2, γ-3 and / or γ-4 genes H 2 and / or C HThe three domains are grouped together. Antibodies can also contain substitutions in the constant region to reduce or increase effector functions, such as complement-mediated cytotoxicity or ADCC (see, for example, Winter et al., US Patent No. 5,624,821; Tso et al., US Patent No. 5,834,597; and Lazar et al., Proc Natl Acad Sci USA 103:4005, 2006), or to prolong the half-life in the human body (see, for example, Hinton et al., J Biol Chem 279:6213, 2004).

[0031] Natural antibody molecules are typically tetramers composed of two identical heterodimers, each containing a paired heavy chain and a light chain. Each light and heavy chain is separated by a variable region (V0). L Refers to a light chain, V H (Referring to the heavy chain, or V referring to both) and the subsequent constant region (C) L Or C H Or C) constitutes it. C H The area itself contains C H 1. Hinge (H), C H 2 and C H Zone 3. In three-dimensional (3D) space, V L and V H The regions fold together to form a V-domain, which is also called a binding domain because it binds to the antigen. L District and C H Zone 1 is folded together, making the V of the light chain... L -C L and heavy chain V H -C H Region 1 together forms part of the antibody, known as the Fab: therefore, a natural "Y-type" antibody contains two Fabs, one of which in each heterodimer forms the Y arm. A heterodimer's C... H Region 2 and another heterodimer C H The two regions are located relative to each other, and their respective C H The three regions fold together to form the antibody’s individual Fc domain (the basis of Y), which interacts with other components of the immune system.

[0032] Within each variable region of the light or heavy chain, there are three short segments (averaging about 10 amino acids in length) called complementarity-determining regions (“CDRs”). The six CDRs within the antibody variable region (three from the light chain and three from the heavy chain) fold together in 3D space to form the actual antibody-binding site, which locks onto the target antigen. The position and length of the CDRs have been defined by E. Kabat et al. (Sequences of Proteins of Immunological Interest, USDapartment of Health and Human Services, 1983, 1987, 1991). The variable region portion outside the CDRs is called the framework, which constitutes the environment of the CDRs. Chothia et al., J. Mol. Biol. 196:901, 1987, have defined the related concept of structure-determined hypervariable regions or loops.

[0033] The “genetically engineered” mAbs described herein are based on recombinant DNA technology, constructing genes or placing them in non-natural environments (e.g., human genes placed in mice or bacteriophages), and therefore include chimeric antibodies and humanized antibodies as described below, but exclude mouse or other rodent mAbs prepared using conventional hybridoma techniques. Chimeric antibodies (or chimeric antibody light or heavy chains, respectively) are antibodies (or antibody light or heavy chains, respectively) in which the variable region of a mouse (or other non-human species) antibody (or antibody light or heavy chain, respectively) is combined with the constant region of a human antibody; their genetic engineering construction methods are well known. Such antibodies retain the binding specificity of mouse antibodies while maintaining approximately two-thirds that of human antibodies.

[0034] Humanized antibodies are genetically engineered antibodies in which a CDR from a non-human "donor" (e.g., chicken, mouse, rat, rabbit, or hamster) is grafted into a human "recipient" antibody sequence, allowing the humanized antibody to retain the binding specificity of the donor antibody (see, for example, Queen, U.S. Patent Nos. 5,530,101 and 5,585,089; Winter, U.S. Patent No. 5,225,539; Carter, U.S. Patent No. 6,407,213; Adair, U.S. Patent Nos. 5,859,205 and 6,881,557; Foote, U.S. Patent No. 6,881,557). The recipient antibody sequence can be, for example, a mature human antibody sequence, a shared sequence of human antibody sequences, a germline human antibody sequence, or a combination of two or more such sequences. Therefore, a humanized antibody is a antibody having partially or entirely of the CDRs derived entirely or substantially from the donor antibody, as well as a variable region framework sequence and a constant region (if present) derived entirely or substantially from the human antibody sequence. Similarly, a humanized light chain (or heavy chain) has at least one, two, and usually all three CDRs derived entirely or substantially from the donor antibody light chain (or heavy chain) and the light chain (or heavy chain) variable region framework, and a light chain (or heavy chain) constant region (if present) derived substantially from the human receptor light chain (heavy chain). Humanized antibodies typically comprise a humanized heavy chain and a humanized light chain. The CDRs in a humanized antibody are substantially derived from the corresponding CDRs of a non-human antibody when at least 85%, 90%, 95%, or 100% of the corresponding amino acids (as defined by Kabat) are identical between the two. The variable or constant regions of an antibody chain are primarily derived from the corresponding human variable or constant regions when at least 85%, 90%, 95%, or 100% of the corresponding amino acids (as defined by Kabat) are identical.

[0035] Here, and elsewhere in this application, percentage sequence identity follows the Kabat numbering convention (C H The Eu index of a region is determined by the maximum match of the antibody sequence. After alignment, if the test antibody region (e.g., a complete mature variable region of the heavy or light chain) is compared with the same region of the reference antibody, the percentage sequence identity of the test antibody and reference antibody regions is the number of positions occupied by the same amino acid in the test antibody and reference antibody regions divided by the total number of aligned positions in the two regions, excluding vacancies, multiplied by 100 to convert to a percentage.

[0036] To retain the high binding affinity of humanized antibodies, at least one of two additional structural elements can be used. See U.S. Patent Nos. 5,530,101 and 5,585,089, which are incorporated herein by reference, for a detailed description of the construction of humanized antibodies. In the first structural element, the framework of the heavy chain variable region of the receptor or humanized antibody is selected to have a high sequence identity (between 65% and 95%) with the framework of the heavy chain variable region of the donor antibody, by adaptively selecting the receptor antibody heavy chain from a number of known human antibodies. In the second structural element, during the construction of the humanized antibody, selected amino acids within the human receptor antibody framework (outside the CDR) are replaced with corresponding amino acids from the donor antibody, according to prescribed rules. Specifically, the amino acids to be replaced in the framework are selected based on their ability to interact with the CDR. For example, the replaced amino acids may be adjacent to the CDR in the donor antibody sequence or within 4–6 Å of the CDR in the humanized antibody, a distance measured in three-dimensional space.

[0037] Other methods for designing humanized antibodies can also yield the same results as those described in U.S. Patent Nos. 5,811,429 and 5,585,089, such as "hyperhumanization" (see Tan et al., J Immunol 169:1119,2002, and U.S. Patent No. 6,881,557) or Studnicak et al., Protein Eng. 7:805, 1994. Furthermore, other methods for preparing genetically engineered, immunogenic-reduced mAbs include "remodeling," "high chimerism," and faceted / surface reconstruction, such as Vaswami et al., Annals of Allergy, Asthma and Immunology 81:105, 1998; Roguska et al., Protein Eng. 9:895, 1996; and U.S. Patent Nos. 6,072,035 and 5,639,641. More human-like faceted antibodies are prepared by replacing specific amino acids in the variable region framework of non-human donor antibodies that may form B or T cell epitopes, such as exposed residues (Padlan, Mol. Immunol. 28: 489, 1991). Other types of genetically engineered antibodies include human antibodies prepared using phage display methods (Dower et al., WO91 / 17271; Mcafferty et al., WO92 / 001047; Winter, WO92 / 20791; and Winter, FEBS Lett. 23: 92, 1998, each of which is cited herein and incorporated herein) or using transgenic animals (Lonberg et al., WO93 / 12227; Kucherlapati WO91 / 10741, each of which is cited herein and incorporated herein).

[0038] The term "antibody" or "mAb" also includes bispecific antibodies. A bispecific antibody is an antibody containing a first domain that binds to a first antigen and a second (different) domain that binds to a second antigen, wherein the first and second domains are derived from or homologous to the variable domain of a natural antibody. The first and second antigens can be the same antigen, in which case the first and second domains can bind to different epitopes on the antigen. The term bispecific antibody also includes multispecific antibodies that, in addition to the first and second domains, contain one or more other domains that bind to other antigens and are derived from or homologous to the variable domain of a natural antibody. The term bispecific antibody also includes antibodies containing a variable domain derived from or homologous to a natural antibody and a second binding domain from another protein, such as the extracellular domain of a receptor (bispecific antibody-immunoadhesin). Furthermore, the term bispecific antibody also includes "two-in-one" antibodies with bispecific binding domains (G Schaefer et al., Cancer Cell 20:472-486, 2011).

[0039] Bispecific antibodies can be prepared in various forms (see, for example, Kontermann, MAbs 4:182-197, 2012 and the references cited therein), such as single-chain variable fragments (scFv), Fab-scFv and scFv-scFv fusion proteins (Coloma et al., Nat Biotechnol 15:125-126, 1997; Lu et al., J Immunol Methods 267:213-226, 2002; Mallender, J Biol Chem 269:199-206, 1994), Bs(scFv)4-IgG (Z Zuo et al., ProteinEng 13:361-367, 2000), dual variable domain antibodies (C Wu et al., Nat Biotechnol 25:1290-1297, 2007), and bispecific antibodies (Holliger et al., Proc Natl Acad Sci USA). 90:6444-6448, 1993). Bispecific F(ab')2 antibody fragments were prepared by chemical conjugation (Brennan et al., Science 229:81, 1985) or by using leucine zippers (Kostelny et al., J Immunol 148:1547-1553, 1992). More natural-shaped bispecific antibodies, where each heavy-light chain pair has a different V region, can be prepared, for example, by chemically cross-linking two heavy-light chain pairs separately (Karpovsky et al., J Exp Med 160:1686-1701, 1984). Natural-shaped bispecific antibodies can also be prepared by expressing the desired heavy and light chains in a single cell, or by fusing two hybridoma cell lines (“quadruplex”; Milstein et al., Nature 305:537-540), or by transfection. The binding of the correct light and heavy chains expressed in cells can be promoted through the “knobs-into-holes” technique (Ridgway et al., Protein Eng 9: 617-621, 1996; Atwell et al., J Mol Biol 270: 26-35, 1997; and U.S. Patent No. 7,695,936) to form targeted bispecific antibodies; optionally, the heavy and light chain domains within the antigen-binding fragment (Fab) of a light-heavy chain pair are exchanged or “crossed”, thereby producing bispecific antibodies called “CrossMabs” (Schaefer et al., Proc Natl Acad Sci USA 108: 11187-11192, 2011; WO 2009 / 080251; WO 2009 / 080252; WO 2009 / 080253).

[0040] Related to the concept of bispecific antibodies is the "multimeric" antibody, which is a mAb containing more than two binding domains, each binding the same antigen. Many forms of bispecific monoclonal antibodies, such as Bs(scFv)4-IgG and bivariate domains, can be modified into multimeric mAbs by using the same variable domain as each binding domain. Thus, for example, a multimeric Bs(scFv)4-IgG antibody contains four copies of the same binding domain.

[0041] An antibody is considered to specifically bind to an antigen if its binding to that antigen is significantly more pronounced than the binding to other unrelated antibodies, and therefore it typically possesses at least 10% specificity against that antigen. 6 However, 10 is preferred. 7 10 8 10 9 Or 10 10 M -1The binding affinity (Ka) of an antibody is the binding affinity of the antibody to an antigen. Generally, when we say an antibody binds to an antigen, we mean specific binding. If we say an antibody does not bind to an antigen, it means that within the experimental error range, any signal indicating binding is indistinguishable from the signal of an unrelated control antibody. An epitope of a mAb is the region on its antigen to which the mAb binds. If two antibodies competitively inhibit (block) the binding of another antibody to an antigen, it is determined that the two antibodies bind to the same or overlapping epitopes. Competitive inhibition of binding means that a 1-fold or 5-fold excess of one antibody inhibits the binding of another antibody by at least 50%, but preferably 75%, or a 10-fold, 20-fold, or 100-fold excess of one antibody inhibits the binding of another antibody by at least 75%, but preferably 90%, or even 95% or 99%, by a competitive binding assay (see Junghans et al., Cancer Res. 50:1495, 1990). If, in a competitive binding assay, the second mAb inhibits the binding of the first mAb by an IC50 that is equivalent to, within 2 or 3 times, the IC50 that the first antibody inhibits its own binding, then one mAb (the second mAb) is said to "completely" compete with the other mAb for binding to the antigen. If the IC50 of the second mAb inhibiting (the first mAb's) binding is actually greater than, for example, 3, 5, or 10 times, the IC50 of the first antibody inhibiting its own binding, then the second mAb is said to "partially" compete with the first mAb for binding to the antigen. Generally, if two mAbs completely compete with each other for binding to an antigen, then the two mAbs have the same epitope on the antigen, and if one mAb at least partially competes with another mAb for binding, then the two mAbs have overlapping epitopes. Alternatively, if almost all amino acid mutations in the antigen that reduce or eliminate the binding of that antigen to one antibody also reduce or eliminate its binding to another antibody, then the two antibodies have the same epitope; and if some amino acid mutations in the antigen reduce or eliminate the binding of that antigen to another antibody, then the two antibodies have overlapping epitopes.

[0042] 2. Anti-DR4 and anti-DR5 antibodies A monoclonal antibody that binds to DR4 (i.e., anti-DR4 mAb) or DR5 (i.e., anti-DR5 mAb) and induces apoptosis by transmitting apoptotic signals to at least certain cell types upon binding to the cell membrane receptors DR4 or DR5 is called an agonist. Such an antibody can be blocking or non-blocking, i.e., inhibiting or not inhibiting the binding of Apo2L / TRAIL to DR4 or DR5, respectively. The agonist mAb of the present invention can be a bispecific antibody that binds to one of DR4 and DR5 and a secondary antigen, or simultaneously binds to DR4 and DR5, inhibiting cell viability or inducing apoptosis by approximately 25%, 50%, 75%, 90%, 95%, 99%, or more at concentrations such as 0.1, 1, 10, 100, 1000, or 10,000 ng / ml, as determined, for example, by inhibiting cellular metabolism, such as using the WST-8 assay. The cell lines used here can be, for example, COLO 205 or SW480 colon cancer lines, H460 lung cancer lines, or BxPC-3 pancreatic cancer cell lines. This inhibition of cell viability or induction of apoptosis can be achieved by using the mAb alone, or in the presence of human peripheral blood mononuclear cells or other human cells, or in the presence of an antibody that cross-links the mAb, such as goat anti-IgG-Fc (at, for example, 10 µg / ml). The effect is typically measured after adding the mAb to the cells and any other drug, incubating overnight at 37°C, and then measuring the results.

[0043] The MAbs used in this invention are preferably specific to DR4 or DR5, meaning they do not (specifically) bind, or bind only to a much lower degree (e.g., less than tenfold) to proteins associated with DR4 or DR5, such as tumor necrosis factor (TNF) receptors TNFR1 and TNFR2 and other members of the TNFR superfamily, other death receptors such as Apo-1 (Fas), and decoy receptors DcR1 and DcR2. The MAbs used in this invention are generally specific to DR4 or DR5 and have at least 10 7 M -1 But preferred 10 8 M -1 Or higher binding affinity (Ka), with 10 being the most preferred. 9 M -1 10 10 M -1 Or 10 11 M -1Or higher. Such mAbs bind to human DR4 or DR5, but also preferentially bind to DR4 or DR5 from other species (e.g., mice or non-human primates, such as cynomolgus monkeys), ideally having a binding affinity similar to (e.g., within 10-fold) that of human DR4 or DR5. The sequences of human DR4 and DR5 are provided by G Pan et al., Science. 276: 111-113, 1997 (GenBank: AAC51226.1) and H Walczak et al., EMBO J 16: 5386-5397, 1997 (GenBank: CAG46696.1), respectively.

[0044] Exemplary antibodies of the present invention are D114 and G4.2 and their chimeric and humanized forms, such as HuD114 and HuG4.2. These, as well as other anti-DR4 mAbs such as 4H6 (produced by hybridoma ATCC HB-12455; see U.S. Patent No. 7,252,994) and anti-DR5 mAbs such as 3H3 (produced by hybridoma ATCC HB-12534; see U.S. Patent No. 6,252,050), or other antibodies containing 4H6 or 3H3 of all CDRs, can be used in the preparation of bispecific and multimeric mAbs of the present invention, wherein chimeric, humanized, or human agonist mAb forms are preferred. Other embodiments of the invention include, preferably, agonist anti-DR4 or anti-DR5 mAbs—ordinary IgG or bispecific or multimeric antibodies—containing a mutation in a constant region that increases binding to an Fc receptor, such as an IgG antibody receptor (FcγR), such as the FcγRIIb receptor, such as HuD114-hFc** and HuG4.2-hFc** mAbs described below. Exemplary mutations are individual G236D, L328F, S239D, and S267E mutations, and preferably double mutations such as G236D / S267E, S239D / S267E, and most preferably S267E / L328F, which provide greater binding affinity than any of the aforementioned mutations (SY Chu Mol Immunol 45: 3926-3933, 2008), and also provide greater binding affinity than other amino acid position mutations, all of which are defined herein using the Kabat numbering system. The S267E single mutation has been shown to increase the potency against DR5 mAb in vitro and in vivo (F Li et al., ProcNatl Acad Sci USA 109:10966-10971, 2012). Most preferably, the mAb of the present invention inhibits the growth of human tumor xenografts in mice, as determined by any of the methods described in the examples or other methods known in the art.

[0045] CDRs (e.g., three CDRs in the light chain and three CDRs in the heavy chain) that, individually or collectively, are at least 90%, 95%, or 98% or completely equivalent to D114 (or G4.2) in their amino acid sequence (as defined by Kabat) and retain their functional properties, such as humanized D114 (or G4.2) mAbs, or D114 (or G4.2) with a small number of functionally irrelevant amino acid substitutions (e.g., conserved substitutions as defined below), deletions, or insertions, are embodiments of the present invention, in which case they may or may not have the amino acid mutations described above.

[0046] Once a single, prototype anti-human DR4 (or anti-human DR5) mAb with the desired properties described herein, such as D114 (or G4.2), can be isolated, other mAbs with similar properties, including those that competitively bind to DR4 (or DR5) and / or have the same epitopes, can be directly produced using methods known in the art. For example, mice can be immunized with the extracellular domain of DR4 (or DR5) to generate hybridomas, and mAbs can be screened based on their ability to competitively bind to DR4 (or DR5) with D114. Mice can also be immunized with a smaller fragment of DR4 (or DR5) containing an epitope that binds to D114 (or G4.2). Epitopes can be located, for example, by binding to a series of overlapping peptides spanning DR4 (or DR5). Mouse mAbs produced in these ways can then be humanized. Alternatively, the method used by Jespers et al., Biotechnology 12:899, 1994, can be used to guide the selection of mAbs with the same epitopes and thus similar properties to D114 (or G4.2), which is cited herein and incorporated herein by reference. Using phage display, the heavy chain of D114 (or G4.2) is first paired with a set of (preferably human) light chains to select mAbs that bind DR4 (or DR5), and then a new light chain is paired with a set of (preferably human) heavy chains to select mAbs that bind DR4 (or DR5) with the same epitopes as D114 (or G4.2). Alternatively, variants of D114 (or G4.2) can be obtained by mutagenesis of the DNA encoding both the heavy and light chains of D114 (or G4.2).

[0047] Genetically engineered mAbs can be expressed using a variety of methods known in the art, such as chimeric, humanized, or bispecific mAbs. For example, genes encoding their light and heavy chain V regions can be synthesized by overlapping oligonucleotides and inserted into an expression vector (e.g., available from Invitrogen) along with a usable C region, the vector providing the necessary regulatory regions, such as promoters, enhancers, polyA sites, etc. A CMV promoter-enhancer is preferred. The expression vector can then be transfected into various mammalian cell lines, such as CHO or nonproductive myeloma including Sp2 / O and NSO, using a variety of well-known methods, such as lipid transfection or electroporation, and selected for antibody expression using appropriate antibiotics. See, for example, U.S. Patent No. 5,530,101. Larger quantities of antibodies can be produced by culturing cells in commercially available bioreactors. Therefore, the present invention provides cell lines expressing any of the mAbs described herein.

[0048] As described above, the mAbs of the present invention, including bispecific mAbs, can be produced according to standard procedures in the art, such as microfiltration, ultrafiltration, protein A or G affinity chromatography, size exclusion chromatography, anion exchange chromatography, cation exchange chromatography, and / or other forms of affinity chromatography based on organic dyes, etc. For pharmaceutical use, substantially pure antibodies having at least about 90 or 95% homology are preferred, and most preferably 98% or 99% or more homology. It should also be understood that when mAbs are manufactured using conventional procedures, one or more amino acids at the amino or carboxyl terminus of the light and / or heavy chains, such as the C-terminal lysine of the heavy chain, may be partially or completely deleted or derived, and such compositions are still considered to be the same mAb.

[0049] 3. Bispecific and multimeric antibodies In one embodiment, the present invention provides a bispecific monoclonal antibody comprising at least one DR4-binding domain and at least one DR5-binding domain. Such an antibody is referred herein as a bispecific DR4 / DR5 antibody or mAb. In a preferred embodiment, each binding domain is derived from a humanized or human mAb. In a preferred embodiment, the bispecific antibody comprises two or more DR4-binding domains and two or more DR5-binding domains; typically, the two DR4-binding domains are identical, and the two DR5-binding domains are identical. In any case, the bispecific mAb as described above is preferably used as an agonist, i.e., to induce apoptosis signaling in susceptible cells such as cancer cells via DR4 and / or DR5. Exemplary bispecific mAbs include the binding domains of the 4H6 anti-DR4 mAb or D114 mAb disclosed herein and their humanized forms, and the binding domains of the 3H3 anti-DR5 mAb or G4.2 mAb disclosed herein and their humanized forms.

[0050] The bispecific antibody of the present invention can be in any form, such as any of the forms listed by Kontermann above. In a preferred embodiment, the bispecific antibody is in the form of Bs(scFv)4-IgG described by Zuo et al. above, and in... Figure 1A As shown in B. In this form, a binding structural field of the single-chain (scFv) form is connected to C. L The region thus becomes the N-terminal structural domain of the light chain, while another binding structural domain of the scFv form connects to C. H One domain, thus becoming the N-terminal domain of the heavy chain, and the two light chains and two heavy chains form a homodimer similar to ordinary IgG antibodies, but with two of each binding domain. Therefore, the advantage of the Bs(scFv)4-IgG form is that it is a homodimer, with each monomer containing the same heavy and light chains, thus eliminating the need for any precautions to ensure proper heterodimerization. Typically, (G4S)3GS or ASGS(G4S)3 is chosen as the V-terminal domain within each scFv. L and V H The joint of the area. Each scFv bonding structural domain can be V L -Connector-V H form or V H -Connector-V L Forms (such as) Figure 1AAs shown), or the binding domain may be part of the light chain and part of the heavy chain, so a total of 2×2×2 = 8 variants of Bs(scFv)4-IgG antibody can be prepared from two given binding domains (e.g., binding domains of HuD114 and HuG4.2), which may have different properties.

[0051] In another embodiment of the invention, the bispecific antibody is, for example, in the form of dual variable domains described by Wu et al. cited above (see...). Figure 1A (and its annotation). Similar to Bs(scFv)4-IgG mAbs, this bispecific mAb is a homodimer, with each monomer containing one of each binding domain, linked in a V1-V2-constant domain sequence. Various peptide linkers can be used to link the first and second variable regions, such as ASTKGPSVFPLAP in the heavy chain and RTVAAPSVIFIPP in the light chain, or ASGS(G4S)3 in both chains. For example, in such a bispecific mAb, the variable domain of HuG4.2 can serve as the first domain, and the variable domain of HuD114 can serve as the second domain; the linker can be one of those mentioned above, and the constant region can be the human IgG1,κ isotype.

[0052] In other preferred embodiments of the invention, a monomer containing a light chain and a heavy chain pair of anti-DR4 mAb is paired with a monomer containing a light chain and a heavy chain of anti-DR5 mAb to form a heterodimer having a normal IgG molecular conformation. If all four chains are expressed in cells, this can be achieved by inserting bulges and holes into the C-terminal of the corresponding heavy chain. H Region 3 promotes the formation of target heterodimer bispecific antibodies rather than homodimers (Ridgway et al., Protein Eng 9:617-21, 1996; Atwell et al., JMolBiol270:26-35, 1997; and U.S. Patent No. 7,695,936), while promoting the correct pairing of light and heavy chains through "cross-linking" in the heavy and light chain domains of one of the monomers to form each anti-DR4 and anti-DR5 monomer (Schaefer et al., Proc Natl Acad Sci USA 108:11187-92, 2011; WO2009 / 080251; WO2009 / 080252; WO2009 / 080253).

[0053] In another embodiment, the present invention provides a multimeric antibody containing three or more DR4-binding domains or three or more DR5-binding domains, for example, four binding domains. In a preferred embodiment, the anti-DR4 or anti-DR5 multimeric mAb is present as Bs(scFv)4-IgG or a dual variable domain form, such that it contains exactly four binding domains targeting DR4 or DR5, respectively.

[0054] In a particularly preferred embodiment of the invention, any form of bispecific DR4 / DR5 or multimeric anti-DR4 or anti-DR5 antibody, including mutations in the constant region that increase binding to the Fc receptor, such as the FcγRIIb receptor, includes those forms specifically described above, such as Bs(scFv)4-IgG and bivariate domains. Exemplary mutations are individual G236D, L328F, S239D, and S267E mutations, and preferably double mutations of G236D / S267E, S239D / S267E, most preferably S267E / L328F (as described by SY Chu et al. cited above), as well as other mutations at these amino acid sites. Any bispecific antibody having a V1 domain that binds DR5, a V2 domain that binds DR4, and a natural human constant region, such as Bs(scFv)4-IgG or bivariate domain (see [link to original text]). Figure 1B The term B-DR5 / DR4-hFc is used to describe a antibody that has both V1 and V2 binding to DR4 (or DR5). Further, if it contains the S267E / L328F double mutation, it is called B-DR5 / DR4-hFc**. Similarly, antibodies with V1 binding to DR4 and V2 binding to DR5 are called B-DR4 / DR5-hFc and B-DR4 / DR5-hFc**. A multimeric Bs(scFv)4-IgG or a dual variable-domain antibody whose V1 and V2 both bind to DR4 (or DR5) is called B-DR4 / DR4-hFc (or B-DR5 / DR5-hFc), and if it further contains the S267E / L328F mutation, it is called B-DR4 / DR4-hFc** (or B-DR5 / DR5-hFc**).

[0055] This invention also provides variant antibodies comprising bispecific antibodies, the light and heavy chains of which differ from those specifically described herein in that they have a small number (e.g., typically no more than 1, 2, 3, 5, or 10) of substitutions, deletions, or insertions, typically within the C or V region framework, but also possibly within CDRs. With respect to the substituted amino acids, substitutions typically made in the variant sequence are conserved. Amino acids can be grouped as follows to determine conserved substitutions, i.e., substitutions within a group: Group I (hydrophobic side chains): met, ala, val, leu, ile; Group II (neutral hydrophilic side chains): cys, ser, thr; Group III (acidic side chains): asp, glu; Group IV (basic side chains): asn, gln, his, lys, arg; Group V (residues affecting chain orientation): gly, pro; Group VI (aromatic side chains): trp, tyr, phe. Preferably, the substitutions in the antibody have no substantial effect on the binding affinity or potency of the antibody, i.e., its ability to transmit apoptotic signals via DR4 and / or DR5. Preferably, the variant sequence is at least 90%, more preferably at least 95%, and most preferably at least 98% identical to the original sequence. Alternatively, other allotypes or homotypes of the constant region can be used.

[0056] 4. Treatment methods In a preferred embodiment, the present invention provides a pharmaceutical formulation comprising any antibody described herein, such as B-DR5 / DR4-hFc, B-DR5 / DR4-hFc**, B-DR4 / DR5-hFc or B-DR4 / DR5-hFc** bispecific mAb, or B-DR4 / DR4-hFc B-DR4 / DR4-hFc**, B-DR5 / DR5-hFc or B-DR5 / DR5-hFc** multimeric mAb, and humanized forms of D114 and G4.2, such as HuD114-hFc** and HuG4.2-hFc**. The pharmaceutical formulation contains an mAb placed in a physiologically acceptable carrier and optionally contains excipients or stabilizers, and is in lyophilized or aqueous solution form. Acceptable carriers, excipients, or stabilizers are non-toxic to the recipient at the dosage and concentration used, and include buffers such as phosphates, citrates, or acetates, with a pH typically between 5.0 and 8.0, most commonly between 6.0 and 7.0; salts such as sodium chloride, potassium chloride, etc., prepared as isotonic solutions; antioxidants, preservatives, low molecular weight peptides, proteins, hydrophilic polymers such as polysorbate 80, amino acids, carbohydrates, chelating agents, sugars, and other standard ingredients known to those skilled in the art (Remington's Pharmaceutical Science 16). th(ed., Osol, A.Ed. 1980). mAb is usually present at concentrations of 0.1-1 mg / kg or 1-100 mg / ml, but most commonly 10-50 mg / ml, for example 10, 20, 30, 40 or 50 mg / ml.

[0057] In another preferred embodiment, the present invention provides a method of treating patients by administering any antibody of the present invention in pharmaceutical formulation form, such as B-DR5 / DR4-hFc, B-DR5 / DR4-hFc**, B-DR4 / DR5-hFc or B-DR4 / DR5-hFc** bispecific mAb or B-DR4 / DR4-hFc, B-DR4 / DR4-hFc**, B-DR5 / DR5-hFc or B-DR5 / DR5-hFc** polyab, and humanized forms of D114 and G4.2 (e.g., HuD114-hFc** and HuG4.2-hFc**), which are generally designed to destroy harmful cells such as cancer cells expressing DR4 and / or DR5. The mAbs prepared in pharmaceutical formulations can be administered to patients via any suitable route, particularly by intravenous infusion or bolus, intramuscular or subcutaneous administration. Intravenous infusion can be completed within 15 minutes, but more commonly 30 minutes, or 1, 2, or even 3 hours. mAbs can also be injected directly into the lesion (e.g., a tumor) or encapsulated in a carrier reagent such as liposomes. The administered dose is sufficient to relieve the treated condition (“therapeutic effective dose”) and may be 0.1–5 mg / kg body weight, such as 1, 2, 3, 4, or 5 mg / kg, but can be as high as 10 mg / kg or even 15, 20, or 30 mg / kg, for example, in the range of 0.1–1 mg / kg, 1–10 mg / kg, or 1–20 mg / kg. Fixed unit doses can also be given, such as 50, 100, 200, 500, or 1000 mg, or the dose can be based on the patient’s surface area, such as 1000 mg / m². 2 Typically, doses of 1 to 8 (e.g., 1, 2, 3, 4, 5, 6, 7, or 8) are administered to treat cancer, but doses of 10, 20, or more may be given. For example, once daily, mAbs are administered twice weekly, once weekly, every other week, once monthly, or at other intervals, depending on factors such as the mAb's half-life, lasting 1 week, 2 weeks, 4 weeks, 6 weeks, 8 weeks, 3–6 months, or longer. Repeated courses of treatment, such as chronic dosing, are also possible.

[0058] Diseases particularly susceptible to treatment with the mAbs of the present invention include those associated with cells (e.g., cancer cells) that express elevated levels of DR4 and / or DR5 compared to normal cells of the same tissue type, such as ovarian cancer, breast cancer, lung cancer (small cell or non-small cell), colon cancer, prostate cancer, pancreatic cancer, gastric cancer, liver cancer (hepatocellular carcinoma), kidney cancer (renal cell carcinoma), head and neck tumors, melanoma, sarcoma, and brain tumors (e.g., glioblastoma). Hematologic malignancies such as leukemia and lymphoma may also be susceptible to the mAbs of the present invention. Optionally, the expression of DR4 and / or DR5 in the cancer cells of the treated subject can be assessed before initiating or continuing treatment. Expression can be assessed at the mRNA level or preferably at the protein level, for example by immunoassay of biopsy specimens, such as immunohistochemistry (IHC). The expression of DR4 and / or DR5 is preferably at least higher than the background assessed by immunoassay with irrelevant control antibodies, and more preferably higher than the expression levels of non-cancer cells of the same tissue type. Other diseases that are sensitive to treatment with the mAb of the present invention include autoimmune diseases such as rheumatoid arthritis, multiple sclerosis, and inflammatory bowel disease.

[0059] In a preferred embodiment, the mAb of the present invention is administered in combination with other treatment regimens (i.e., together with, before, during, or after). For example, for the treatment of cancer, the mAb can be administered in combination with any one or more known chemotherapeutic agents, such as alkylating agents like carmustine, chlorambucil, cisplatin, carboplatin, oxaliplatin, procarbazine, and cyclophosphamide; antimetabolites such as fluorouracil, fluorouridine, fludarabine, gemcitabine, methotrexate, and hydroxyurea; natural medicines, including plant alkaloids and antibiotics such as bleomycin, doxorubicin, daunorubicin, idarubicin, etoposide, mitomycin, mitoxantrone, vincristine, vinblastine, and Taxol (paclitaxel) or related compounds such as Taxotere®; topoisomerase 1 inhibitors such as irinotecan; tyrosine kinase inhibitors such as... Gleevec® (imatinib), Sutent® (sunitinib), Nexavar® (sorafenib), Tarceva® (erlotinib), Tykerb® (lapatinib), Iressa® (gefitinib), and Xalkori® (crizotinib); rapamycin (sirolimus) and other mTOR inhibitors; and angiogenesis inhibitors; and all approved and investigational anticancer agents listed in WO 2005 / 017107 A2 (which is incorporated herein by reference). mAbs can be used in combination with one, two, three, or more of these other agents, preferably in a standard chemotherapy regimen. Typically, the other agents are those that are known or have been identified as effective against the specific type of cancer being treated.

[0060] Other agents that can be used in combination with the mAbs of the present invention for the treatment of cancer include biologics, such as monoclonal antibodies, including Herceptin® or Perjeta® (petuzumab) against the HER2 antigen, Avastin® against VEGF, antibodies against the epidermal growth factor (EGF) receptor, such as Erbitux® (cetuximab) and Vectibix® (panitumumab), immune system activators, such as Yervoy® (ipilimumab), anti-PD-1 mAbs such as Opdivo® (nivolurnab) and Keytruda® (pembroluzimab), and antibody-drug conjugates such as Kadcyla™ (ado-trastuzumab emtansine). Furthermore, the mAbs can be used in combination with any form of surgery and / or radiation therapy.

[0061] Treatment regimens combining the mAbs of the present invention (e.g., standard chemotherapy) can, compared to the same treatment regimens without mAbs (e.g., chemotherapy), increase median progression-free survival or overall survival in patients with the specific types of cancer listed above by at least 20%, 30%, or 40%, but preferably 50%, 60% to 70%, or even 100% or longer; or (at least) by 2, 3, 4, 6, or 12 months. Additionally or alternatively, treatment regimens combining the mAbs (e.g., standard chemotherapy) can, compared to the same treatment regimens without mAbs (e.g., chemotherapy), increase the complete response rate, partial response rate, or objective response rate (complete + partial) in patients (particularly relapsed or refractory) by at least 30%, but preferably 50%, 60% to 70%, or even 100%.

[0062] Typically, in clinical trials (e.g., phase II, II / III, or III trials), the increase in median progression-free survival and / or response rate (mentioned above) in patients treated with chemotherapy combined with a monoclonal antibody mAb, compared to a control group receiving chemotherapy alone (or with placebo), is statistically significant, for example at p = 0.05, 0.01, or even 0.001. It should also be understood that response rates are determined using objective criteria commonly used in cancer clinical trials, such as those accepted by the National Cancer Institute and / or the Food and Drug Administration, such as the RECIST criteria (Responsiveness Assessment Criteria for Solid Tumors).

[0063] 5. Other methods The mAbs of this invention can also be used for diagnosis, prognosis, and experimentation. They can be used to measure the levels of DR4 and / or DR5 in tumors or the circulatory system of cancer patients, thereby guiding and directing cancer treatment. For example, tumors associated with high levels of DR4 and / or DR5 are particularly susceptible to mAb treatment. In specific embodiments, such as in tumor biopsy specimens or serum, the mAbs can be used for ELISA or radioimmunoassay to determine the levels of DR4 and / or DR5. The simultaneous use of anti-DR4 mAb and anti-DR5 mAb is particularly useful for detecting cells that co-express DR4 and DR5. For various assay methods, the mAbs can be labeled with fluorescent molecules, spin-labeled molecules, enzymes, or radioisotopes, and can be provided in the form of kits containing all the necessary reagents for the assay. In other uses, monoclonal antibodies are used to purify DR4 or DR5, for example, by affinity chromatography.

[0064] 6. Example Example 1: Preparation of anti-DR4 and anti-DR5 mAbs To prepare and measure mAbs binding to human DR4, several fusion proteins were constructed using standard molecular biology methods. For DR4-hFc, cDNA encoding the extracellular domain of human DR4 (amino acids 109-239) was prepared and inserted into an Invitrogen-derived pCI vector, linked at the C-terminus to the human Igγ-1 Fc region (hinge-CH2-CH3), transfected, and expressed in 293F mammalian cells. Using a similar approach, the monkey DR4 extracellular domain was cloned from cynomolgus monkey liver cDNA (Biochain) to construct the cDR4-hFc fusion protein to obtain the monkey DR4 domain. These Fc-fusion proteins were purified from the 293F culture supernatant using a protein A column. Similarly, the DR4-KF fusion protein was prepared by linking the human DR4 extracellular domain to the C-terminal human Igκ constant region containing the FLAG peptide (DYKDDDDK); a similar protein, cDR4-KF, was prepared using cynomolgus monkey DR4. These FLAG-tagged fusion proteins were purified using an anti-FLAG column. To generate and determine mAbs binding to human DR5, an entirely similar set of proteins, DR5-hFc and DR5-KF, as well as the extracellular domains (amino acids 54-183) of cDR5-hFc and cDR5-KF, were prepared using human and cynomolgus monkey DR5 in the same manner as DR4. Blocking assays were performed using human Apo2L / TRAIL protein (amino acids 95-281) (R&Dsystems). Commercially available human DR4-Fc and DR5-Fc fusion proteins (R&Dsystems) were also used.

[0065] To prepare anti-DR4 mAbs, Balb / c mice were immunized twice weekly for 11 weeks via individual hind paw pads (using 1 μg DR4-hFc (TRAILR1-Fc from R&D Systems)), followed by 5 immunizations with 5 μg DR4-hFc resuspended in MPL / TDM adjuvant (Sigma-Aldrich). DR4-hFc preparation was described above. Three days after the final booster immunization, popliteal lymph node cells were fused with mouse myeloma cells P3X63AgU.1 (ATCC CRL 1597) using 35% polyethylene glycol. Hybridomas were screened in HAT medium according to Chuntharapai and Kim, J Immunol 163:766, 1997. Ten days after fusion, hybridoma culture supernatants were screened by DR4 binding ELISA as described below. Selected hybridomas were cloned twice using limiting dilution. MAbs in ascites and culture media of the selected hybridomas were purified by protein A / G column purification. Of the several antibodies screened for further analysis, antibody D114 was selected for the next step because it exhibited superior agonist activity in the cell killing assay. The isotype of D114 was determined using an isotype kit and identified as IgG1,κ.

[0066] To prepare anti-DR5 mAbs, Balb / c mice were immunized twice a week for 12 weeks via individual hind paw pads (using 5 μg DR5-hFc followed by a single immunization with 1 μg DR5-hFc resuspended in MPL / TDM adjuvant (Sigma-Aldrich). Three days after the final booster, popliteal lymph node cells were fused with mouse myeloma cells, and hybridomas were screened using HAT medium as described above for anti-DR4 mAbs. Ten days after fusion, the hybridoma culture supernatant was screened using a DR5 binding ELISA as described below. The selected hybridoma clones were then cloned twice by limiting dilution as described above for anti-DR4 mAbs and purified. Further testing of the selected mAbs against COLO was conducted as described below. The effect of antibodies on 205 cell viability and their binding activity to monkey DR5 was investigated. Among several screened antibodies, antibody G4.2 was selected for the next step due to its superior agonist activity in the cell killing assay. G4.2 was isotyped using an isotype assay kit, and its isotype was determined to be IgG1,κ.

[0067] Example 2: Antibody Binding and Blocking Activity Each ELISA assay step described in this patent application is performed by incubation with suitable reagents at room temperature for 1 hour, except for the initial plate coating step which requires overnight incubation at 4°C, followed by a 1-hour blockade with 2% BSA. Between each step, the plates are washed three times with PBS containing 0.05% Tween 20. Data points are typically taken two or three times; there is usually little variation between replicate data points.

[0068] To determine the binding of mAbs to DR4 (or DR5), a capture assay is typically used: plates are first coated overnight with goat anti-mouse IgG-Fc (2 μg / ml), blocked with 2% BSA, incubated with hybridoma supernatant or incrementally concentrated purified antibody, and then incubated with 0.3 μg / ml DR4-KF (or DR5-KF). The captured DR4-KF (or DR5-KF) is detected by adding HRP-goat anti-human κ, followed by a TMB matrix. However, to measure the binding of the HuG4.2 variant to DR5, a direct binding assay is used as described below: plates are coated with Dr5-Fc (0.5 μg / ml), blocked, captured, and incubated with incrementally concentrated mAbs, while the bound mAbs are detected using HRP-goat anti-human κ.

[0069] To determine the blocking activity of the mAbs, plates were first coated overnight with goat anti-human IgG g-Fc (2 μg / ml), blocked with 2% BSA, and then incubated with 0.5 μg / ml DR4-hFc (or DR5-hFc). Then, 50 ng / ml Apo2L (TRAIL, R&D Systems) was mixed with various concentrations of mAbs and added to the wells. The binding of Apo2L was detected by adding 0.2 μg / ml biotinylated anti-TRAIL antibody (R&D Systems), followed by HRP-streptavidin, and then TMB substrate.

[0070] Using these ELISA assays, it was shown that D114 binds to DR4 in the same way as 4H6 ( Figure 2A And, like 4H6, it blocks the binding of Apo2L to DR4. Figure 2B Similarly, results from binding assays using cDR4-hFc protein showed that D114 binds to cynomolgus monkey DR4, as it does to human DR4. G4.2 also showed good binding to DR5. Figure 2C And effectively block the binding of Apo2L to DR5. Figure 2DFurthermore, according to the binding assay using cDR5-hFc protein described above, G4.2 also binds well to cynomolgus monkey DR5. However, 3H3 did not bind to cynomolgus monkey DR5 in this assay, which gives G4.2 an advantage over 3H3 and also indicates that G4.2 must have a different epitope than 3H3.

[0071] Example 3: Cytotoxic effects of D114 and G4.2 To conduct experiments to detect a decrease in cell viability (referred to as cell killing in this paper), human tumor cell lines (in DMEM / 10% FCS) were subjected to a concentration of 2 × 10⁻⁶ cells / mL. 4 Cells were seeded at 100 μL / well in 96-well plates and incubated overnight (under standard conditions of 37°C and 5% CO2). The medium was then removed and replaced with 100 μL of the same medium containing an ascending gradient concentration of mAbs, with or without 10 μg / ml goat anti-mouse IgG-Fc (anti-mIgG-Fc) against mice and goat anti-human IgG-Fc (anti-hIgG-Fc) against humans, humanized, or chimeric antibodies. In some assays, when tumor cells were seeded into the plates, the mAbs and anti-IgG-Fc were included in the medium. Each concentration of mAb was administered in duplicate. After overnight incubation under standard conditions, 10 μL / well of WST-8 was added, and after 1 hour, cell viability was determined by measuring absorbance at OD 450. In some assays, the purpose of using goat anti-mIgG-Fc or anti-hIgG-Fc is to crosslink (oligopolymerize) the tested anti-DR4 or anti-DR5 antibody, thus potentially delivering a stronger apoptotic signal to the cells. This cross-linking of anti-IgG-Fc antibodies is thought to mimic the cross-linking that occurs when anti-DR4 or anti-DR5 antibodies bind to DR4 or DR5, thereby forming a cross-link with leukocytes infiltrating tumors via FcγR on their cell surface.

[0072] Using this assay, the D114-inhibited H460 lung tumor cell line (ATCC HTB-177) was measured in the presence of anti-mIgG-Fc. Figure 3A ) and SW480 colon cancer cell line (ATCC CCL-228; Figure 3B The effects of G4.2 on cell viability in H460 cells are similar to those in 4H6, as are those on other cell lines tested. Similarly, G4.2 inhibits H460 cells (…). Figure 3C ) and COLO 205 colon tumor cells (ATCC CCL-222; Figure 3D Its effect on cell vitality is almost the same as that of 3H3.

[0073] Example 4: Antibody Construction and Identification Cloning of the heavy and light chain variable regions of mAbs, construction and expression of various mAbs, and introduction of mutations were all performed using standard molecular biology methods, such as those described in U.S. Patent No. 7,632,926 (incorporated herein by reference for all purposes). The amino acid sequences of the mature heavy and light chain variable regions of 4H6 are shown in [images / details]. Figure 4A and 4B In the diagram, the mature heavy and light chain variable regions of 3H3 are shown in... Figure 4C and 4D The mature heavy and light chain variable regions of D114 are shown in [the diagram]. Figure 5A and 5B In the middle, it is listed at the top line and marked as D114; the mature heavy chain and light chain variable regions of G4.2 are shown respectively in Figure 5C and 5D In the middle, the top line is marked G4.2.

[0074] In addition, to compare with the mAbs disclosed herein, several other anti-DR5 mAbs were constructed based on the published sequences: human mAbs drozitumab (Apomab; Figures 17 and 18 of US Patent No. 8,030,023) and conatumumab (AMG 655; Figure 19 of WO2012 / 106556) and mouse mAb TRA-8 (Figures 23 and 24 of US Patent No. 7,244,429).

[0075] As is the convention used in this article, the suffix "hFc **" applied to antibodies (also written as Fc ** in some figures) indicates that they contain human γ-1 constant regions with S267E and L328F mutations, the sequence of which is as follows: Figure 6A A portion of the image is shown (see) Figure 6A (See the description), and the suffix "hFc" indicates that it contains the human γ-1 constant region without the S267E and L328F mutations (and therefore S and L at those amino acid positions). We constructed chimeric antibodies of 3H3-hFc, 4H6-hFc, 0114-hFc, and G4.2-hFc by combining the variable region of the corresponding mouse antibody with the human κ constant region and the human γ-1 constant region. We also constructed corresponding chimeric antibodies of 3H3-hFc**, 4H6-hFc**, D114-hFc**, and G4.2-hFc** containing the human γ-1 constant region with the S267E and L328F mutations. For comparison in some experiments, we also constructed the antibody Apomab-hFc** by combining the human Apomab variable region with the human γ-1 constant region containing the S267E and L328F mutations.

[0076] Bispecific antibody B-4H6 / 3H3-hFc was constructed in the form of Bs(scFv)4-IgG. The heavy and light chain amino acid sequences of this mAb are shown below. Figure 6A and Figure 6B Each mature chain begins with the first amino acid following the signal peptide. Therefore, this mAb has Figure 1A The structure shown has V from 4H6 mAb H 1 and V L 1 and V from 3H3 mAb H 2 and V L 2.

[0077] Example 5: Cell-killing effect of antibodies In the aforementioned cell killing assays, the ability of the various antibodies to kill different cell lines (i.e., reduce cell viability) was tested. In the absence of anti-hIgG-Fc, neither 3H3-hFc** nor 4H6-hFc**, nor even combinations of these two mAbs (mixtures), showed any killing effect on SW480 colon tumor cells. Figure 7A ) or H460 lung tumor cells ( Figure 8A The effect of these cross-linking agents is that, without the cross-linking agents that oligomerize the antibodies, they cannot transmit death signals to the cells. However, in the presence of cross-linking agents against hIgG-Fc, 3H3-hFc* and 4H6-hFc** (to a greater extent) can kill SW480 cells. Figure 7B Adding 3H3-hFc to 4H6-hFc will not further increase its lethality.

[0078] To determine whether a mutated antibody has an advantage in killing tumor cells in the presence of cells expressing FcγRIIb, peripheral blood mononuclear cells (PBMCs), mainly composed of monocytes and lymphocytes, were isolated from human blood using a Ficoll-Paque PLUS (GE Healthcare) according to the manufacturer's instructions. Cell viability was measured as described above, typically using a medium containing 2 × 10⁻⁶ mAbs and 10⁻⁶ cells / mL. 5 Individual PBMCs. In the presence of PBMCs, 4H6-hFc is more effective than 4H6-hFc in killing H460 tumor cells. Figure 8B ).

[0079] We then investigated whether two mutations in the human γ-1 constant region provided higher cytotoxic activity than a single mutation. Therefore, we constructed a chimeric 4H6-Fc* mAb (indicated by Fc* with a single asterisk) containing only the S267E mutation, unlike the S267E / L328F double mutation in 4H6-hFc**. Considering the potential variability of PBMCs, the above cytotoxicity assay used a fixed concentration of 0.5 μg / ml mAb, 4 x 10⁻⁶ wells. 4 COLO 205 tumor cells and 4x10 cells per well from four different human donors. 5 PBMCs. For each donor, 4H6-hFc moderately reduced tumor cell viability (compared to control hlgG mAb set at 100% survival for each donor), with 4H6-hFc* having a single mutation resulting in a slightly to significantly greater reduction in cell viability than 4H6-hFc, but 4H6-hFc** resulting in a significantly greater reduction in viability than 4H6-hFc*. Figure 9A This indicates that the two mutations enhance the binding to FcγRIIb, which is superior to such a single mutation. To determine whether the same is true for other combinations of at least two mutations, we introduced other mutation pairs into 4H6-hFc according to the table below.

[0080] Table 1: Mutants

[0081] In the presence of human PBMCs, a fixed antibody concentration of 1 μg / ml was used to assay COLO205 cells, MDA-MB-231 cells (ATCC HTB-26), and H460 cells ( ) under the specified conditions. Figure 9B All four variants showed roughly the same reduction in cell viability, with the greatest effect observed in COLO 205 cells and only a moderate effect observed in H60 cells. Therefore, various combinations of two or more mutations, each capable of increasing binding to FcγRIIb, are also well-suited for enhancing tumor cell killing.

[0082] To further explore the advantages of double mutations over single mutations, the ability of 4H6-hFc, 4H6-hFc*, and 4H6-hFc** to kill COLO 205 cells was tested at a certain range of antibody concentrations in the presence of PBMCs. Figure 10A Experiments showed that no killing effect of 4H6-hFc on COLO 205 cells was observed, while 4H6-hFc* had a weak killing effect on COLO 205 cells and 4H6-hFc** had a strong killing effect. This again demonstrates the advantage of double mutation over single mutation.

[0083] As a final experiment in this regard, the inhibition of COLO 205 tumor xenograft growth in mice by 4H6-hFc, 4H6-hFc*, and 4H6-hFc** was compared according to the methods described below. Each mAb was administered twice weekly at 2 mg / kg. The ability of the double-mutant 4H6-hFc** to inhibit xenograft growth was stronger than that of the unmutated 4H6-Fc or the single-mutant 4H6-Fc*. Figure 10B This is consistent with the results of cell killing.

[0084] Example 6: The ability of antibodies to inhibit the growth of tumor xenografts Xenograft experiments were generally performed using previously known techniques (Kim et al., Nature 362: 841, 1993). Human tumor cells were typically cultured in complete DMEM medium and harvested in HBSS. Female immunodeficient mice (4-6 weeks old) were subcutaneously injected into the dorsal region with 2-10 × 10⁻⁶ cells suspended in 0.1-0.2 ml of HBSS. 6 Cells, and in some experiments, also contain Matrigel (Corning). When the tumor size reaches approximately 100 mm... 3 Mice were randomly assigned to groups and typically received intraperitoneal injections of 0.5–5 mg / kg mAbs in 0.1 ml volumes, either once or twice weekly. Tumor size was determined twice weekly by measuring two dimensions [length (a) and width (b)]. The tumor size was determined according to V = ab 2 / 2. Tumor volume was calculated and expressed as mean tumor volume ± SEM. Typically, there were 5–7 mice in each treatment group. Statistical analysis could be performed, for example, using Student's t-test at the final data points.

[0085] We first conducted xenograft experiments to determine whether, due to the enhanced FcγR binding provided by the mutation in the hFc** constant region, consistent with the enhanced cell killing in the presence of the aforementioned PBMCs, the hFc** form is more effective than the hFc form in vivo in combating DR4 and DR5 mAbs. In fact, at a single dose of 2 mg / kg, the anti-DR4 mAb D114-hFc only weakly inhibited the growth of COLO 205 xenografts, while under the same conditions, D114-hFc** almost completely inhibited the xenografts (…). Figure 11A (As described below, the results obtained by inhibiting COLO 205 xenografts with humanized forms of HuD114-hFc and HuD114-hFc** are very similar;) Figure 18AFurthermore, when administered in this manner, D114-hFc did not significantly inhibit the growth of SW480 colon tumor xenografts, but D114-hFc** weakly inhibited xenograft growth under these conditions. Figure 11B Finally, when administered twice weekly at 2 mg / kg, another anti-DR4 mAb, 4H6-hFc, weakly inhibited the growth of COLO205 xenografts, but 4H6-hFc **completely inhibited xenograft growth under these conditions ( Figure 11C ).

[0086] Regarding anti-DR5 mAbs, 3H3-Fc did not significantly inhibit the growth of COLO 205 xenografts when administered once at a dose of 0.5 mg / kg, but even at very low dose levels, 3H3-hFc** completely inhibited the growth of these xenografts. Figure 12A Similarly, when administered once at 1 mg / kg, 3H3-Fc only weakly inhibited the growth of MIA PaCa-2 (ATCCCRL-1420) pancreatic tumor xenografts, but even at such very low dose levels, 3H3-hFc** completely inhibited the growth of these xenografts. Figure 12B Similarly, as described below, when administered once at a dose of 1 mg / kg, humanized mAb HuG4.2-hFc** inhibited the growth of COLO 205 xenografts and MIA PaCa-2 xenografts better than HuG4.2-hFc. Figure 19A Therefore, for anti-DR4 and anti-DR5 mAbs, the mutated hFc form is more effective in vivo than the non-mutated hFc form.

[0087] Example 7: Construction and Identification of Humanized Antibodies The design, construction, expression, and purification of humanized D114 and humanized G4.2 mAbs were performed using standard molecular biology methods, such as the method for L2G7 mAb described in U.S. Patent No. 7,632,926, which is incorporated herein by reference for all purposes. More specifically, the methods for designing humanized D114 (or G4.2) mAbs are generally followed in U.S. Patent Nos. 5,530,101 and 5,585,089 by Queen et al. Figure 5A and 5BAs shown in (or 5C and 5D), the VK sequence AIT38746 and VH sequence AAC18293 (or human VK sequence AAQ02698 and VH sequence AAC50998) listed in the bottom row are used as receptor sequences for the D114 (or G4.2) VL and VH sequences, respectively, due to their particularly high framework homology (i.e., sequence identity). A computer-generated molecular model of the D114 (or G4.2) variable domain is used to locate amino acids in the framework sufficiently close to the CDRs that potentially interact with them. To design humanized D114 (or G4.2) light and heavy chain variable regions, a CDsR conceptually derived from a mouse D114 (or G4.2) mAb is first transplanted into the receptor framework region. Framework sites where the computer model shows significant contact with the CDRs (potentially necessary to maintain the CDR conformation) are replaced with amino acids from the mouse antibody, replacing the human framework amino acids.

[0088] For D114, these substitutions are at residues 48 and 71 of the light chain (HuD114-L1) and residues 48 and 71 of the heavy chain (HuD114-H1), or by adding additional heavy chain residues 67 and 69 to these residues (HuD114-H2). The light and heavy chain V-region sequences of humanized D114 are shown in [images / descriptions]. Figure 5A and 5B The middle row is labeled HuD114, where they are aligned with their respective D114 donor and human receptor V regions. mAbs having a HuD114-L1 light chain and a HuD114-H1 or HuD114-H2 heavy chain are named HuD114#1 and HuD114#2, respectively. Each of these is prepared in two forms: HuD114-hFc, where the human γ-1 constant region does not have the S267E and L328F mutations (and therefore S and L at positions 267 and 328), and HuD114-hFc**, where the constant region has these mutations. Using the binding assay described above, HuD114-hFc**#1 and HuD114-hFc**#2 both bind to DR4, just like the chimeric antibody D114-hFc described above. Figure 13A This indicates that there was no loss of affinity during the humanization process. In the presence of goat anti-hIgG-Fc, HuD114-HFC**#1 and HuD114-hFc**#2 had the same efficacy in killing COLO 205 cells. Figure 14ATo enhance cytotoxic activity, new versions of the HuD114 light and heavy chains with additional mouse substitutions were prepared: HuD114-L2 containing mouse substitutions at residues 48 and 71 plus 43 and 44, and HuD114-H3 containing mouse substitutions at residues 48, 67, 69, and 71 plus 91. Six antibodies were prepared from all combinations of HuD114-L1 and HuD114-L2 with HuD114-H1, HuD114-H2, and HuD114-H3, all exhibiting comparable binding and cytotoxic activity. Therefore, HuD114-hFc#2 and HuD114-hFc**#2 were used for further research and are subsequently labeled HuD114-hFc and HuD114-hFc** in the figures.

[0089] The invention also includes variants of these preferred antibodies, such as humanized antibodies comprising a light chain V region and a heavy chain V region, wherein the sequence of the light chain V region is consistent with... Figure 5A The HuD114-L1 in the heavy chain is at least 90%, 95%, 98%, or 99% identical, and the heavy chain V region has the same characteristics as... Figure 5B The HuD114-H1 or Hu0114-H2 in the antibody has at least 90, 95, 98, or 99% sequence identity. Preferably, all CDR residues in the antibody are those of the donor. Preferably, light chain positions 48 and 71 are occupied by V and Y, respectively, and heavy chain positions 48 and 71 are occupied by L and A, respectively.

[0090] Similarly, for G4.2 mice, substitutions occur at residues 4 and 68 of the light chain (HuG4.2-L1) or at these residues plus residue 1 (HuG4.2-L2), and at residues 27, 28, 30, and 93 of the heavy chain (HuG4.2-H1) or at these residues plus residue 47 (HuG4.2-H2). The V-region sequences of the light and heavy chains of HuG4.2 are shown below. Figure 5C and 5D In the table below, those labeled HuG4.2 are aligned with their respective G4.2 donor and human receptor V regions. Four different humanized G4.21 antibodies, named HuG4.2#1, #2, #3, and #4, were prepared by combining each humanized light chain with each humanized heavy chain, as shown in the table below, where the number of substitutions in each chain is indicated in parentheses. For HuD114, each of these mAbs was prepared in two forms: hFc without any γ-1 heavy chain constant region mutations, and hFc** with mutations S267E and L328F.

[0091] Table 2: HuG4.2 variants

[0092] All four versions of HuG4.2 integrate well with DR5, with HuG4.2-hFc **#1 and HuG4.2-hFc **#2 being slightly better than HuG4.2-hFc **#3 and HuG4.2-hFc **#4. Figure 13B In fact, compared to the chimeric antibody G4.2-hFc**, the binding ability of HuG4.2-hFc**#1 and #2 to DR5 is comparable, indicating little or no loss of affinity during humanization. In the presence of goat anti-hIgG-Fc ( Figure 14B In the case of [missing information], all four versions of HuG4.2 also killed COLO 205 cells well, with HuG4.2-hFc**#2 slightly outperforming the other antibodies and essentially on par with the chimeric antibody G4.2-hFc**. Therefore, HuG4.2-Fc#2 and HuG4.2-hFc**#2 were used for further research and are referred to as HuG4.2-Fc and HuG4.2-Fc** in the following figures and figures.

[0093] The invention also includes variants of HuG4.2, which include the same as... Figure 5C The light chain V region of HuG4.2-L1 or HuG4.2-L2 in the sample has at least 90, 95, 98, or 99% sequence identity, and has a similarity to... Figure 5D The heavy chain V region of HuG4.2-H1 or HuG4.2-H2 exhibits at least 90, 95, 98, or 99% sequence identity. Preferably, all CDR residues in these antibodies are derived from the donor antibody. Preferably, light chain positions 4 and 68 are occupied by L and R, respectively, and heavy chain positions 27, 28, 30, and 93 are occupied by L, P, N, and T, respectively. Optionally, heavy chain position 47 is occupied by L.

[0094] The cell-killing ability of various humanized mAbs was tested in the presence of goat anti-IgG-Fc (10 μg / ml) as a cross-linking agent. Both HuD114-hFc and HuD114-hFc** killed COLO 205 colon tumor cells significantly better than the anti-DR4 mAb mapatumumab tested in clinical trials. Figure 15A ) and SW480 colon tumor cells ( Figure 15B However, there was no significant difference between HuD114-hFc and HuD114-hFc**, as expected, because the increased binding to FcγRIIb provided by the mutation should not have any effect in the absence of cells expressing FcγRIIb. Similarly, the ability of HuG4.2-hFc and / or HuG4.2-hFc** to kill COLO 205 cells against anti-DR5 mAbs Apomab, AMG655, and lexatumumab was compared ( Figure 15C And compared with Apomab and TRA-8 in their ability to kill MIA PaCa-2 pancreatic tumor cells ( Figure 15D Again, in the tested condition, there was no significant difference between HuG4.2-hFc and HuG4.2-hFc** due to the lack of cells expressing FcγRIIb. However, HuG4.2-hFc and HuG4.2-hFc** showed significantly better killing ability against both cell lines than other tested mAbs that have undergone clinical trials. Killing of H460 cells by various mAbs was also measured in the presence of anti-hIgG-Fc. Figure 16A ) and SW480 cells ( Figure 16B The ability of HuD114-hFc** and HuG4.2-hFc** to kill H460 cells was significantly better than that of anti-DR5 mAb Apomab and AMG 655, while the ability of HuD114-hFc** to kill SW480 cells was significantly better than that of HuG4.2-hFc**, and slightly better than that of Apomab and AMG655, which is consistent with the results mentioned above. Figure 15A -D).

[0095] Next, we compared the cell-killing abilities of various mAbs in the presence of human PBMCs. The hFc** form may be superior to the hFc form due to the enhanced binding to FcγRIIb provided by the S267E and L328F mutations. Indeed, in killing COLO 205 cells (… Figure 17A ) or H60 cells ( Figure 17B In terms of ability, HuG4.2-Fc** is much better than HuG4.2-Fc, and similarly, Apomab-hFc**, a form of Apomab with the same mutations, is better at killing cells than Apomab itself. Figure 17A However, HuG4.2-hFc** still killed COLO 205 or H60 cells substantially better than Apomab-hFc** (and superior to AMG 655), which again demonstrates the superior ability of HuG4.2-hRFc** against DR5mAb.

[0096] In the context of HuD114 and HuG4.2, to compare the effects of double and single mutations, we tested the cell-killing ability of unmutated mAbs HuD114-hFc (or HuG4.2-hFc), mAbs with only the S267E mutation HuD114-hFc* (or HuG4.2-hFc*), and mAbs with the S267E / L328F double mutation HuD114-hFc** (or HuG4.2-hFc**) in the presence of PBMCs. HuD114-hFc* killed COLO 205 cells ( Figure 17C ) and H460 cells ( Figure 17D The ability of HuG4.2 to kill cells is slightly better than that of HuD114-hFc, while HuD114-hFc has a much better ability to kill cells. Similarly, HuG4.2-hFc kills COLO 205 cells ( Figure 17E ) and H460 cells ( Figure 17F The ability of the double mutation to kill cells is slightly better than that of HuG4.2-hFc, while HuG4.2-hFc has a much better ability to kill cells. Therefore, the presence of a double mutation in the constant region is much more advantageous than a single mutation.

[0097] Finally, the ability of humanized mAbs HuD114 and HuG4.2 to inhibit xenograft growth was investigated. When administered once at 2 mg / kg, HuD114-hFc moderately inhibited COLO 205 xenograft growth, while under these conditions, HuD114-hFc **completely inhibited xenograft growth (…). Figure 18A A single administration of HuD114-hFc at 2 mg / kg also strongly inhibited the growth capacity of H460 lung tumor xenografts. Figure 18B In another experiment, under low-dose testing conditions, HuD114-h Fc** completely inhibited COLO 205 xenografts, while the antibody mapatumumab, which has already been tested in clinical trials, had no effect. Figure 18C When administered once at a dose of 2 mg / kg, HuD114-hFc** was also slightly more effective than mapatumumab in inhibiting the growth of Ramos (ATCC CRL-1596) lymphoma xenografts. Figure 18D A single dose of 1 mg / kg of HuG4.2-hFc or HuG4.2-hFc **inhibits COLO 205 xenografts ( Figure 19A ) and MIA PaCa-2 pancreatic tumor xenograft ( Figure 19BThe growth of these mAbs was significantly enhanced, with HuG4.2-hFc** being significantly more effective than HuG4.2-hFc. (HuD114-hFc** was also effective in this model; Figure 19B). In the xenograft model, the hFc** form of these mAbs was more potent than the hFc form, a result consistent with the greater ability of the hFc** form to kill tumor cells in vitro in the presence of the aforementioned PBMCs.

[0098] Example 8: Identification of Bispecific Antibodies To demonstrate the above-mentioned bispecific B-4H6 / 3H3-hFc** mAb ( Figure 6A (B) This method can bind both DR4 and DR5 simultaneously. First, the plate is coated overnight with an anti-DR4 mAb (2 μg / ml), which does not compete with DR4H6 for binding. It is then blocked with 2% BSA and incubated with DR4-hFc (0.5 μg / ml). Then, incremental concentrations of B-4H6 / 3H3-hFc mAb are added to the wells, followed by 0.5 μg / ml DR5-KF. The binding of DR5-KF is detected using HRP-anti-Flag M2 (Sigma) and the substrate. Since only mAbs capable of simultaneously binding DR4 (in the form of DR4-hFc) and DR5 (in the form of DR5-KF) will give a positive result in this ELISA analysis, the results ( Figure 20A The results show that B-4H6 / 3H3-hFc** has this capability, while mAbs 4H6-hFc** and 3H3-hFc** obviously do not.

[0099] The ability of B-4H6 / 3H3-hFc to kill COLO 205 cells in the absence of cross-linking agent-anti-mIgG-Fc was determined. 4H6-hFc failed to kill cells under these conditions, while 3H3-hFc showed some degree of cytotoxicity. Figure 20B This is because COLO 205 cells are highly sensitive to death receptor agonists, far more so than other cell lines such as H60 and SW480 used in this study. However, although the 4H6 component itself has no cytotoxic ability, the B-4H6 / 3H3-hFc bispecific mAb significantly outperformed the 3H3-hFc in killing cells. Figure 20B Conversely, simply mixing individual mAbs 3H3-hFc** and 4H6-hFc** does not increase the cytotoxicity of either mAb. Figure 7A B).

[0100] Another bispecific antibody, B-HuD114 / HuG4.2-hFc**, was constructed in the same manner as B-4H6 / 3H3-hFc** in the form of Bs(scFv)4-IgG, and also contains a double mutation in the constant region. Using the ELISA assay described in this example, it was found that B-HuD114 / HuG4.2-hFc** binds to both DR4 and DR5, while HuG4.2-hFc** does not (…). Figure 20C In the absence of anti-hIgG-Fc, B-HuD114 / HuG4.2-hFc mAb also killed COLO205 cells, even more effectively than the natural ligand TRAIL. Figure 20D ).

[0101] Although the invention has been described with reference to the present preferred embodiments, it should be understood that various modifications can be made without departing from the invention. Unless otherwise expressly stated herein, any step, element, embodiment, feature, or aspect of the invention may be used in conjunction with any other aspect. All cited publications, patents, and patent applications (including registration numbers, etc.) are incorporated herein by reference in their entirety, to the extent that each individual publication, patent, and patent application is expressly and individually indicated by reference for all purposes. The word “in this” means anywhere in this patent application, and not just within the portion where the word “in this” appears. If multiple sequences are associated with registration numbers at different times, the sequence associated with the registration number on the effective filing date of this application is contemplated, meaning the actual filing date or an earlier priority application disclosing the registration number.

Claims

1. A monoclonal antibody (mAb) that binds to death receptor 5, comprising a light chain V region and a heavy chain V region, wherein the light chain V region has three CDRs, namely CDR1 as shown in SEQ ID NO:26, CDR2 as shown in SEQ ID NO:27, and CDR3 as shown in SEQ ID NO:28, and the heavy chain V region has three CDRs, namely CDR1 as shown in SEQ ID NO:29, CDR2 as shown in SEQ ID NO:30, and CDR3 as shown in SEQ ID NO:

31.

2. The mAb according to claim 1 is a humanized antibody.

3. The mAb of claim 2, comprising a light chain V region having at least 90% sequence identity with HuG4.2-L1 as shown in SEQ ID NO: 13 or HuG4.2-L2 as shown in SEQ ID NO: 14, and a heavy chain V region having at least 90% sequence identity with HuG4.2-H1 as shown in SEQ ID NO: 17 or HuG4.2-H2 as shown in SEQ ID NO: 18, wherein light chain positions 4 and 68 of the Kabat number are occupied by L and R, respectively, and heavy chain positions 27, 28, 30 and 93 of the Kabat number are occupied by L, P, N and T, respectively.

4. The mAb according to claim 2, comprising a light chain V region of HuG4.2-L1 as shown in SEQ ID NO: 13 or HuG4.2-L2 as shown in SEQ ID NO: 14 and a heavy chain V region of HuG4.2-H1 as shown in SEQ ID NO: 17 or HuG4.2-H2 as shown in SEQ ID NO:

18.

5. The mAb according to claim 1, which inhibits the growth of human tumor xenografts in mice.

6. The mAb according to claim 1, wherein it is a bispecific antibody.

7. The mAb of claim 1, comprising a human constant region having one or more mutations that increase binding to the human Fcγ receptor, wherein the one or more mutations that increase binding to the human Fcγ receptor are one or both of the S267E and L328F mutations.

8. A pharmaceutical composition comprising the monoclonal antibody of claim 1 and a pharmaceutically acceptable carrier.

9. Use of the mAb of any one of claims 1-7 or the pharmaceutical composition of claim 8 in the preparation of a medicament for treating cancer, wherein the cancer is lung cancer or colon cancer.

Images