Stable multispecific antibodies

Abstract

The present invention relates to multispecific antibody constructs comprising Fab fragments having a specific set of mutations at the interface of the CH1 and CL domains, said mutations preventing heavy / light chain mismatch.

Description

[0001] This application is a divisional application of Chinese invention patent application No. 201880029825.2, filed on March 27, 2018, with a priority date of March 27, 2017, entitled "Stable Multispecific Antibody". Technical Field

[0002] This invention relates to the production of multispecific antibody molecules useful in the medical field. Background Technology

[0003] Bispecific antibodies (bsAbs) combine the specificity of two antibodies and target different antigens or epitopes simultaneously. BsAbs with "two-target" functionality can interfere with multiple surface receptors or ligands, for example, those associated with cancer, proliferation, or inflammatory processes. BsAbs can also bring targets close together to support the formation of protein complexes on a single cell or trigger cell-to-cell contact. Examples of "forced binding" functionality include bsAbs that support protein complexation in coagulation cascades or immune cell recruiters and / or activators targeting tumors. After years of research and development, the first bsAb was approved in 2009.

[0004] Initially, bispecific antibodies were prepared via chemical conjugation or by fusing two hybridoma cell lines that produce two different mAbs into quadromas. However, chemical conjugation can sometimes alter the antigen-binding site, thereby impairing the antibody's biological properties. The quadromama method has the drawback that the random pairing of the heavy and light chains from two different antibodies theoretically produces ten equally possible combinations, resulting in a mixture of immunoglobulin molecules, only one of which is the desired bispecific product, and this must be separated from the mispaired products.

[0005] Genetic modification has become the preferred method for producing bispecific antibodies and has led to the development of a variety of different recombinant bispecific antibody forms. Some of these bispecific antibodies are very simple, derived from single-chain Fv (scFv) fragments of two (or more) different antibodies linked by appropriate peptide linkers. These antibodies are relatively easy to produce, and because they are formed from a single polypeptide chain and contain only the Fv region of the parent antibody, there are no mismatch problems between the chains. However, they are smaller than full-length immunoglobulins and lack constant regions, particularly the Fc region. While this may be advantageous in some applications, such as when it is desirable to avoid Fc-mediated effects, it is disadvantageous when Fc-mediated effector functions are required. Moreover, due to their small size and lack of an Fc region, they have a very short in vivo half-life.

[0006] Therefore, other bispecific recombinant antibody forms have been designed that more closely mimic naturally occurring immunoglobulin molecules, particularly those with intact Fc regions. These can be classified into two main forms.

[0007] In the first antibody (IgG scFv), an scFv fragment from antibody A is fused to the terminal (usually C-terminus) of the heavy chain of antibody B. The resulting antibody has only one type of heavy chain (containing the VH, CH1, CH2, and CH3 domains of antibody B and the VH and VL domains of antibody A) and one type of light chain (containing the VL and CL domains of antibody B), without inter-chain mismatch. This form is described, for example, by Qu et al. (Blood, 111, 2211-2219, 2008).

[0008] In the second antibody, the heavy and light chains of antibody A pair with the heavy and light chains of antibody B. This form regenerates the bispecific antibody produced by tetradoma, thus also generating a similar mispairing problem. To address the heavy chain mispairing problem, it has been proposed to mutate the CH3 domain of the antibody to promote its heterodimerization (i.e., heavy chain A pairs with heavy chain B) and prevent its homodimerization. This is accomplished through the so-called "knob into holes" method (Ridgway et al., Protein Eng, 9,617-21, 1996; US Patent 7,695,936). A "knob" mutation is introduced at the CH3 dimer interface of the heavy chain of antibody A, which consists of replacing a smaller amino acid with a larger amino acid, creating steric hindrance that prevents homodimerization. Simultaneously, to promote heterodimerization, a complementary "hole" mutation consisting of replacing a larger amino acid with a smaller amino acid is introduced into the CH3 domain of antibody B. To address the heavy / light chain mismatch problem, it has been suggested to use antibodies with different specificities but sharing a common light chain, which have previously been identified from scFv phage libraries (Merchant et al., Nat Biotechnol, 16, 677-81, 1998; US Patent 7,183,076). A drawback of this approach is the difficulty in identifying antibodies with a common light chain. International patent application WO2013 / 005194 proposes that heavy / light chain mismatch can be prevented by mutating certain key residues at the interface between the CH1 and CL domains, thus ensuring the desired chain pairing. Summary of the Invention

[0009] The inventors have now discovered that a new set of mutations can further improve chain matching in multispecific antibody molecules.

[0010] This invention relates to multispecific, such as bispecific antibody constructs, comprising different Fab fragments having specific group mutations at the interface of the CH1 and CL domains, said mutations promoting homologous pairing of heavy / light chains and preventing their mismatch.

[0011] The sequence position numbers used for the CH1 and CL domains in this article refer to the Kabat numbers (Kabat, EA et al., Sequences of proteins of immunological interest. 5th ed. - US Department of Health and Human Services, NIH Publication, No. 91-3242, pp. 662, 680, 689, 1991).

[0012] This article provides Fab fragments selected from the following mutations:

[0013] a) The Fab fragment, which consists of the following:

[0014] - The VH and VL domains of the target antibody;

[0015] -The CH1 domain derived from the CH1 domain of immunoglobulins is obtained by replacing the threonine residue at position 192 of the CH1 domain with aspartic acid.

[0016] - The CL domain derived from the immunoglobulin CL domain by replacing the asparagine residue at position 137 of the CL domain with a lysine residue and replacing the serine residue at position 114 of the CL domain with an alanine residue; and

[0017] b) The Fab fragment, which consists of the following:

[0018] - The VH and VL domains of the target antibody;

[0019] - The CH1 domain derived from the CH1 domain of immunoglobulins is obtained by replacing the leucine residue at position 124 of the CH1 domain with glutamine and replacing the serine residue at position 188 of the CH1 domain with a valine residue; and

[0020] - The CL domain derived from the CL domain of immunoglobulins is formed by replacing the valine residue at position 133 of the CL domain with a threonine residue and replacing the serine residue at position 176 of the CL domain with a valine residue.

[0021] Any construct containing this mutated Fab fragment, preferably any protein construct, is part of this invention.

[0022] Specifically, a multispecific antigen-binding fragment is provided, comprising at least two Fab fragments with different CH1 and CL domains, wherein each Fab fragment recognizes a different target epitope, and the Fab fragments are arranged in tandem in any order, wherein the C-terminus of the CH1 domain of the first Fab fragment is linked to the N-terminus of the VH domain of the subsequent Fab fragment via a peptide linker.

[0023] At least one Fab fragment, more preferably two fragments, are selected from:

[0024] a) The Fab fragment, which consists of the following:

[0025] The CH1 domain, derived from the CH1 domain of immunoglobulins, is obtained by replacing the threonine residue at position 192 of the CH1 domain with aspartic acid.

[0026] - The CL domain derived from the immunoglobulin CL domain by replacing the asparagine residue at position 137 of the CL domain with a lysine residue and replacing the serine residue at position 114 of the CL domain with an alanine residue; and

[0027] b) The Fab fragment, which consists of the following:

[0028] - The VH and VL domains of the target antibody;

[0029] - The CH1 domain derived from the CH1 domain of immunoglobulins is obtained by replacing the leucine residue at position 124 of the CH1 domain with glutamine and replacing the serine residue at position 188 of the CH1 domain with a valine residue; and

[0030] - The CL domain derived from the CL domain of immunoglobulins is formed by replacing the valine residue at position 133 of the CL domain with a threonine residue and replacing the serine residue at position 176 of the CL domain with a valine residue.

[0031] According to a preferred embodiment, the CH1 domain is derived from IgG immunoglobulin, advantageously being the IgG1 subtype. The CL domain is preferably κ-type. Preferably, when used for human treatment, the immunoglobulin from which the mutated CH1 and mutated CL domains are derived is a human immunoglobulin.

[0032] The VH and VL domains can be derived from any natural or genetically modified antibody that recognizes the epitopes that people wish to target.

[0033] The mutated Fab fragment of the present invention can be used in any multispecific antibody construct in which it is necessary to promote homologous pairing of heavy / light chains and prevent their mismatch.

[0034] Advantageously, mutated Fab fragments can be used in multispecific antibody molecules containing antigen-binding fragments, each of which consists substantially of tandemly arranged Fab fragments separated by linkers.

[0035] Therefore, another objective of the present invention is a multispecific antigen-binding fragment comprising at least two, and up to six, different Fab fragments selected from the following:

[0036] - Fab fragments containing the wild-type CH1 and CL domains of immunoglobulins (also defined in this paper as "wild-type Fab fragments")

[0037] - Fab fragments with mutations as defined above (a);

[0038] - Fab fragments with mutations as defined above (b);

[0039] Each Fab fragment identifies a different target epitope, and the Fab fragments are arranged in tandem in any order, with the C-terminus of the CH1 domain of the first Fab fragment connected to the N-terminus of the VH domain of the subsequent Fab fragment via a peptide linker.

[0040] One or two Fab fragments are arranged in tandem in any order and are connected to the C-terminus of the CH3 domain or the C-terminus of the hinge sequence via a polypeptide linker, while another Fab or two tandemly arranged Fab fragments are attached to the N-terminus of the hinge sequence in any order.

[0041] The aforementioned hinge sequence is typically derived from the natural hinge sequence of IgG1, IgG2, IgG3, IgG4, IgA1, IgA2, or IgD, and is most preferably derived from IgG1. Alternatively, it may be a sequence modified by point mutation and / or addition of amino groups, such that the modified hinge sequence consists of fewer than 80 amino acids, preferably fewer than 60 amino acids, and still more preferably fewer than 40 amino acids.

[0042] An "antigen-binding fragment" is defined in this paper as a molecule having two or more antigen-binding regions, each region recognizing a different epitope. Different epitopes can be carried by the same antigen molecule or by different antigen molecules.

[0043] In a preferred embodiment, a multispecific antibody having two identical antigen-binding arms is further provided, each arm consisting of a multispecific antigen-binding fragment, i.e., an antigen-binding fragment capable of binding multiple (i.e., more than one) antigens, as described above.

[0044] In a specific implementation, the multispecific antibody has an immunoglobulin-like structure and contains...

[0045] - Two identical antigen-binding arms, each consisting of such multispecific antigen-binding fragments as defined above;

[0046] -The CH2 and CH3 domains of immunoglobulin dimerization;

[0047] The hinge region of -IgA, IgG or IgD connects the C-terminus of the CH1 domain of the antigen-binding arm to the N-terminus of the CH2 domain.

[0048] More particularly, one subject of the invention is a multispecific, preferably bispecific antibody comprising two heavy chains and four light chains, wherein each heavy chain comprises

[0049] - The Fc region of an immunoglobulin containing a hinge-CH2-CH3 domain.

[0050] - The Fc region is connected to the Fab heavy chain CH1-VH of antibody 1 (Ab1) via the hinge domain.

[0051] It then links to the CH1-VH domain of the Fab heavy chain of antibody 2 (Ab2) via a peptide linker sequence, and this peptide linker sequence connects the N-terminus of the VH domain of the Fab heavy chain of Ab1 to the C-terminus of the CH1 domain of Ab2.

[0052] Furthermore, the four light chains contain the Fab light chain CL-VL of Ab1 and the Fab light chain CL-VL of Ab2, which are related to their cognate heavy chain structural domains.

[0053] Ab1 and Ab2 identify different epitopes.

[0054] Furthermore, the Fab CH1 domain of either Ab1 or Ab2 is a mutated domain, derived from the immunoglobulin CH1 domain by replacing the threonine residue at position 192 with aspartic acid, and the homologous CL domain is a mutated domain, derived from the immunoglobulin CL domain by replacing the asparagine residue at position 137 with lysine and the serine residue at position 114 with alanine, and / or

[0055] The Fab CH1 domain of either Ab1 or Ab2 is a mutant domain, derived from the immunoglobulin CH1 domain by replacing the leucine residue at position 124 of the CH1 domain with glutamine and replacing the serine residue at position 188 of the CH1 domain with a valine residue. The homologous CL domain is a mutant domain, derived from the immunoglobulin CL domain by replacing the valine residue at position 133 of the CL domain with a threonine residue and replacing the serine residue at position 176 of the CL domain with a valine residue.

[0056] Further descriptions are provided for any protein chains selected from the following:

[0057] - The light chain of the mutated Fab fragment of the present invention;

[0058] -The heavy chain of the mutated Fab fragment of the present invention;

[0059] -The heavy chain of the antigen-binding fragment of the present invention;

[0060] -The heavy chain of the immunoglobulin-like multispecific antibody of the present invention.

[0061] This disclosure further provides polynucleotides comprising sequences encoding the protein chains of the present invention. The polynucleotides may also comprise additional sequences: in particular, they may advantageously comprise sequences encoding leader sequences or signal peptides that allow the secretion of the protein chains. Attached Figure Description

[0062] Figure 1 The SDS-polyacrylamide gel electrophoresis of BiXAb-6567 under reducing and non-reducing conditions is shown. Lane 1: Migration of BiXAb-6567 under reducing conditions; Lane 2: Molecular weight markers indicating the weight of each band; Lane 3: Migration of BiXAb-6567 under non-reducing conditions.

[0063] Figure 2 Size exclusion chromatography analysis of BiXAb-6567 is shown.

[0064] Figure 3 The melting curves of two parental antibodies (anti-CD38 and anti-PD-L1) and BiXAb-6567, as determined by differential scanning calorimetry, are shown.

[0065] Figure 4A The binding curves of two parental antibodies (anti-CD38 and anti-PD-L1) and BiXAb-6567 in the direct CD38 antigen binding ELISA are shown. Figure 4B The binding curves of two parental antibodies (anti-CD38 and anti-PD-L1) and BiXAb-6567 in the direct PD-L1 antigen binding ELISA are shown. Figure 4C The binding curves of BiXAb-6567 in ELISA are shown for the dual antigens (PD-L1 and CD38).

[0066] Figures 5A to 5C The fluorescence-activated cell sorting curves of two parental mAbs (anti-CD38 and anti-PD-L1) and BiXAb-6567 are shown in three different cell lines. Figure 5A 5B: Multiple myeloma RPMI-8226; 5C: CHO cells stably transfected with full-length CD38; and 5D: Ovarian cancer cell line SKOV-3.

[0067] Figure 6 The titration binding curves of two parental antibodies (anti-CD38 and anti-PD-L1), BiXAb-6567, and the negative control anti-CD20 antibody on the CHO-CD38 cell line are shown.

[0068] Figure 7 The cytotoxic activity curves of two parental antibodies (anti-CD38 and anti-PD-L1), BiXAb-6567, and two negative control antibodies (anti-CD20 and anti-HER2) are shown in ADCC assays using the multiple myeloma cell line RPMI-8226 as the target cell and ungraded, non-preactivated monocytes as effector cells.

[0069] Figure 8 The cytotoxic activity curves of two parental antibodies (anti-CD38 and anti-PD-L1), BiXAb-6567, and two negative control antibodies (anti-CD20 and anti-HER2) are shown in ADCC assays using CHO-CD38 cell lines as target cells and ungraded, non-preactivated monocytes as effector cells.

[0070] Figure 9 The cytotoxic activity curves of two parental antibodies (anti-CD38 and anti-PD-L1), BiXAb-6567, and two negative control antibodies (anti-CD20 and anti-HER2) are shown in an ADCC assay using SKOV-3 cell lines as target cells and IL-12-enriched pre-activated NK cells as effector cells.

[0071] Figure 10 The cytotoxic activity curves of two parental antibodies (anti-CD38 and anti-PD-L1), BiXAb-6567, and two negative control antibodies (anti-CD20 and anti-HER2) are shown in an ADCC assay using SKOV-3 cell lines as target cells and IL-15-enriched pre-activated NK cells as effector cells.

[0072] Figure 11This is a schematic diagram of the bispecific antibody of the present invention, which comprises two heavy chains and four light chains, and shows different combinations of mutations. Detailed Implementation

[0073] definition:

[0074] The basic structure of naturally occurring antibody molecules is a Y-type tetramer quaternary structure composed of two identical heavy chains and two identical light chains, which are linked together by non-covalent interactions and interchain disulfide bonds.

[0075] In mammalian species, there are five types of heavy chains: α, δ, ε, γ, and μ, which determine the class (isotype) of immunoglobulins: IgA, IgD, IgE, IgG, and IgM, respectively. Following the N-terminal variable domain (VH) of the heavy chain is the constant region. The γ, α, and δ heavy chains contain three domains (numbered CH1, CH2, and CH3 from the N-terminus to the C-terminus), while the constant regions of the µ and ε heavy chains consist of four domains (numbered CH1, CH2, CH3, and CH4 from the N-terminus to the C-terminus). The CH1 and CH2 domains of IgA, IgG, and IgD are separated by flexible hinges whose lengths vary between different types. For IgA and IgG, the lengths differ between different subtypes: IgG1, IgG2, IgG3, and IgG4 have hinges of 15, 12, 62 (or 77), and 12 amino acids, respectively, while IgA1 and IgA2 have hinges of 20 and 7 amino acids, respectively.

[0076] There are two types of light chains: λ and κ, which can bind to any heavy chain isotype, but in a given antibody molecule they are all of the same type. The two light chains appear to be functionally identical. Their N-terminal variable domain (VL) is followed by a constant region, which consists of a single domain called CL.

[0077] Heavy and light chains pair via protein-protein interactions between the CH1 and CL domains and via VH / VL interactions, and the two heavy chains are bound together via protein-protein interactions between their CH3 domains. The structure of immunoglobulin molecules is generally stabilized by interchain disulfide bonds between the CH1 and CL domains and between the hinges.

[0078] The antigen-binding region corresponds to the arms of the Y-shaped structure, which consists of each complete light chain paired with the VH and CH1 domains of the heavy chain, and is called a Fab fragment (used for fragment antigen binding). Fab fragments are first produced by the digestion of native immunoglobulin molecules by papain, which cleaves the antibody molecule at the hinge region on the amino-terminal side of the interchain disulfide bond, thereby releasing two identical antigen-binding arms. Other proteases (such as pepsin) can also cleave the antibody molecule at the hinge region on the carboxyl-terminal side of the interchain disulfide bond, releasing a fragment consisting of two identical Fab fragments held together by a disulfide bond; reduction of the disulfide bond in the F(ab′)2 fragment produces the Fab′ fragment.

[0079] The antigen-binding region corresponding to the VH and VL domains is called the Fv fragment (variable fragment); it contains the CDR (complementarity-determining region), which forms the antigen-binding site (also called the complementarity site).

[0080] The effector region of an antibody responsible for binding to effector molecules or cells corresponds to the stem of a Y-shaped structure and contains paired CH2 and CH3 domains (or CH2, CH3, and CH4 domains, depending on the antibody class) of the heavy chain, and is called the Fc (crystallizable fragment) region.

[0081] Due to the identity of the two heavy chains and the two light chains, naturally occurring antibody molecules have two identical antigen-binding sites and therefore bind to two identical epitopes simultaneously.

[0082] In the context of this invention, a "multispecific antigen-binding fragment" is defined herein as a molecule having two or more antigen-binding regions, each region recognizing a different epitope. The different epitopes may be carried by the same antigen molecule or different antigen molecules. The terms "recognizing" or "recognizes" refer to the fragment specifically binding to the target antigen.

[0083] A "multispecific antibody" contains at least two multispecific antigen-binding fragments / arms and is capable of binding to two, three or more different antigens.

[0084] An antibody is said to "specifically bind" to a target antigen if it binds to it with higher affinity, faster binding speed, and / or longer duration of binding compared to binding to other substances. "Specific binding" or "preferential binding" does not necessarily require (although it may include) exclusive binding. Generally, but not necessarily, the mention of binding refers to preferential binding.

[0085] The terms "mutated derivative," "mutant," or "functional variant" refer to a sequence that differs from the parental sequence it refers to by deleting, substituting, or inserting one or more amino acids. Preferably, the mutated derivative exhibits at least 80%, more preferably at least 85%, and still more preferably at least 90% homology with the native sequence. In certain embodiments, the mutation substantially does not affect the function of the antibody.

[0086] Mutant derivatives or functional variants may comprise a VH chain (containing an amino acid sequence having at least 85% (e.g., 90%, 92%, 94%, 95%, 96%, 97%, 98%, or 99%) identity with any reference sequence described herein), a VL chain (containing an amino acid sequence having at least 85% (e.g., 90%, 92%, 94%, 95%, 96%, 97%, 98%, or 99%) identity with any reference sequence described herein), or both. These variants are capable of binding target antigens. In some instances, the variants have similar antigen-binding affinity relative to the aforementioned reference antibodies (e.g., having less than 1 x 10⁻⁶). -7 M, 10 -8 M, preferably less than 1 x 10 -9 or 1 x 10 -10 M's KD).

[0087] The affinity of a binding reaction is defined by the terms ka (binding rate constant), kd (dissociation rate constant), or KD (equilibrium dissociation). Typically, when used against an antibody, specific binding refers to a rate of less than 10-1. -7 M, preferably less than 10 -8 M, for example, less than 10 -9 M or 10 -10 The affinity (KD) value of M indicates the specific binding ("recognition") of the antibody to its target. A lower KD value indicates a higher binding affinity (i.e., a stronger binding force), therefore 10 -9 The KD value represents a value higher than 10. -8 It has a high KD value and high binding affinity.

[0088] The "percentage of identity" between two amino acid sequences was determined using an algorithm modified as described in Karlin and Altschul Proc. Natl. Acad. Sci. USA 87:2264-68, 1990, and in Karlin and Altschul Proc. Natl. Acad. Sci. USA 90:5873-77, 1993. This algorithm was incorporated into the NBLAST and XBLAST programs (version 2.0) of Altschul et al. J. Mol. Biol. 215:403-10, 1990. BLAST protein searches can be performed using the XNBLAST program with a score of 50 and a word length of 3 to obtain amino acid sequences homologous to the target protein molecule. In the case of vacancies between two sequences, vacancy BLAST can be used as described in Altschul et al., Nucleic Acids Res. 25(17):3389-3402, 1997. When using BLAST and vacancy BLAST procedures, the default parameters of each procedure (such as XBLAST and NBLAST) can be used.

[0089] In other embodiments, the functional variants described herein may contain one or more mutations (e.g., conserved substitutions) that preferably do not occur on residues predicted to interact with one or more CDRs.

[0090] This article describes mutant derivatives or functional variants that are essentially the same as the reference antibody.

[0091] The terms "substantially identical" or "not substantially different" mean that the relevant amino acid sequence of the variant (e.g., in the framework region (FR), CDR, VH, or VL domains) is not substantially different from that of the reference antibody (e.g., including conserved amino acid substitutions), thus the variant has substantially similar binding activity (e.g., affinity, specificity, or both) and biological activity relative to the reference antibody. Such variants may include minor amino acid changes, such as one or two substitutions in a specified region of five amino acids. Typically, more substitutions may be made in the FR region compared to the CDR region, as long as they do not adversely affect the binding function of the antibody (e.g., reducing binding affinity by more than 50% compared to the original antibody). In some embodiments, the sequence identity between the original antibody and the modified antibody may be about 85%, 90%, 95%, 96%, 97%, 98%, 99%, or higher. In some embodiments, the modified antibody has the same binding specificity and at least 50% of the affinity of the original antibody.

[0092] Conservative substitutions will produce molecules with similar functions and chemical properties to those from which such modifications are made. For example, a “conservative amino acid substitution” can involve replacing a native amino acid residue with another residue that has little or no effect on the polarity or charge of the amino acid residue at that position. The desired amino acid substitution (whether conserved or non-conservative) can be determined by those skilled in the art. For example, amino acid substitutions can be used to identify important residues in a molecular sequence or to increase or decrease the affinity of the molecules described herein. Variants containing one or more conserved amino acid substitutions can be prepared according to methods known to those skilled in the art for altering polypeptide sequences, methods which can be found, for example, in references compiling such methods, such as *Molecular Cloning: A Laboratory Manual*, J. Sambrook et al., eds., 2nd edition, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, 1989, or *Current Protocols in Molecular Biology*, FMAusubel et al., eds., John Wiley & Sons, Inc., New York. Conservative substitutions of amino acids include substitutions between amino acids in the following groups: (a) M, I, L, V; (b) F, Y, W; (c) K, R, H; (d) A, G; (e) S, T; (f) Q, N; and (g) E, D.

[0093] Amino acid sequence variants of antibodies can be prepared by introducing appropriate nucleotide changes into antibody nucleic acids or by peptide synthesis. Such modifications include, for example, deletion and / or insertion and / or substitution of residues within the antibody amino acid sequence. Any combination of deletions, insertions, and substitutions can be performed to obtain the final construct, provided that the final construct possesses the desired characteristics. Nucleic acid molecules encoding amino acid sequence variants of antibodies are prepared using a variety of methods known in the art. These methods include, but are not limited to, oligonucleotide-mediated (or site-directed) mutagenesis, PCR mutagenesis, and cassette mutagenesis of earlier prepared variants or non-variant (natural) versions of antibodies. In one embodiment, the equilibrium dissociation constant (KD) value of the antibody of the present invention is less than 10. -7 M, especially less than 10 -8 M, 10 -9 M or 10 -10 M. Binding affinity can be determined using techniques known in the art, such as ELISA or biospecific interaction analysis (e.g., using surface plasmon resonance), or other techniques known in the art.

[0094] Any molecule described herein can be examined using conventional methods to determine its properties, such as antigen-binding activity, antigen-binding specificity, and biological function.

[0095] The terms “subject,” “individual,” and “patient” are used interchangeably in this document and refer to a mammal being evaluated for treatment and / or being treated. Subjects may be humans, but also include other mammals, particularly those that can be used as laboratory models of human diseases, such as mice, rats, rabbits, dogs, etc.

[0096] The term "treatment" or "treating" refers to the operation, application, or method of treatment in which a subject, including humans, receives medical assistance for the purpose of directly or indirectly improving the subject's condition. Specifically, in some embodiments, the term refers to reducing morbidity or alleviating symptoms, eliminating recurrence, preventing recurrence, preventing morbidity, improving symptoms, improving prognosis, or a combination thereof. Those skilled in the art will understand that treatment does not necessarily result in the complete absence or elimination of symptoms. For example, in the context of cancer, "treatment" or "treatment" can refer to slowing the growth, proliferation, or metastasis of tumors or malignant cells, preventing or delaying the development of tumors or malignant cells' growth, proliferation, or metastasis, or some combination thereof.

[0097] Preferred design of multispecific antibodies:

[0098] This document provides multispecific antigen-binding fragments and multispecific antibody constructs comprising the aforementioned fragments, wherein each multispecific antigen-binding fragment consists substantially of Fab fragments arranged in tandem and separated by the linkers of this invention.

[0099] Such fragments and constructs preferably contain chains derived from human immunoglobulins, preferably IgG, and still preferably IgG1.

[0100] In the case of a multispecific antigen-binding fragment comprising two or more different Fab fragments, the polypeptide linkers separating the Fab fragments may be the same or different. According to a preferred embodiment of the multispecific antibody of the invention, it has two identical antigen-binding arms, each arm consisting of an antigen-binding fragment capable of binding multiple (more than one) antigens, as defined above. Depending on the intended use of the antibody, the antigen-binding arms may be linked together in various ways.

[0101] If you want to obtain an antibody without Fc-mediated effects, the antibody will not contain an Fc region. In this case, the two antigen-binding arms can be linked together, for example:

[0102] - Homodimerization via an antigen-binding arm through an interchain disulfide bond provided by a polypeptide linker of the isolated Fab fragment, if the linker contains a cysteine ​​residue; and / or

[0103] - By adding a polypeptide extension containing a cysteine ​​residue to the C-terminus of each antigen-binding arm, interchain disulfide bonds are allowed to form, and homodimerization of the polypeptide extension produces a hinge-like structure; by way of non-limiting example, the polypeptide extension may be a hinge sequence of, for example, IgG1, IgG2, IgG3, IgA or IgD.

[0104] - A semi-rigid connector links the C-terminus of the heavy chain of two antigen-binding arms to form a single polypeptide chain and keep the antigen-binding arms sufficiently far apart from each other.

[0105] Alternatively, if effector functions are required, such as antibody-dependent cell-mediated cytotoxicity (ADCC), antibody-dependent phagocytosis (ADP), and complement-dependent cytotoxicity (CDC), the multispecific antibodies of the present invention may further include an Fc domain that provides these effector functions. The choice of the Fc domain will depend on the type of effector function desired.

[0106] In this case, the multispecific antibody of the present invention has an immunoglobulin-like structure and comprises:

[0107] - Two identical multispecific antigen binding arms as defined above;

[0108] -The CH2 and CH3 domains of immunoglobulin dimerization;

[0109] - The C-terminus of the CH1 domain of the antigen-binding arm is connected to the hinge region of the N-terminus of the CH2 domain for IgA, IgG, or IgD, or when the CH4 domain following the CH3 domain is from IgM or IgE, in which case the C-terminus of the CH1 domain of the antigen-binding arm can be directly connected to the N-terminus of the CH2 domain.

[0110] Preferably, the CH2 and CH3 domains, hinge region and / or CH4 domain are derived from immunoglobulins that are identical to or have the same type and subclass as the CH1 domain of the antigen-binding arm.

[0111] The CH2, CH3, and optionally CH4 domains, as well as the hinge region, derived from innate immunoglobulins can be used. They may also be mutated if desired, for example, to modulate the effector function of the antibody. In some cases, all or part of the CH2 or CH3 domain may be omitted.

[0112] More specifically, the present invention provides multispecific, preferably bispecific, tetravalent antibodies that include two binding sites for each of their targets, as well as a functional Fc domain that allows activation of effector functions such as antibody-dependent cell-mediated cytotoxicity (ADCC), phagocytosis, and complement-dependent cytotoxicity (CDC).

[0113] The preferred antibodies of this invention are full-length antibodies. They preferably comprise heavy and light chains derived from human immunoglobulins, preferably IgG, and even more preferably IgG1.

[0114] The light chain can be a λ or κ light chain, with the κ light chain being preferred.

[0115] In a preferred embodiment, the polypeptide linker used in the present invention is a tetra-Fab multispecific, preferably bispecific antibody that links two pairs of IgG Fab domains, the amino acid sequence of which comprises heavy chain sequences of at least two Fabs linked by the polypeptide linker, followed by a native hinge sequence, and then an IgG Fc sequence, which is co-expressed with a suitable IgG light chain sequence.

[0116] Figure 11 Examples of preferred bispecific antibodies of the present invention having an IgG-like structure are shown in the figure.

[0117] In a preferred embodiment, the bispecific antibody of the present invention typically comprises

[0118] - A continuous heavy chain composed of Fc (hinge-CH2-CH3).

[0119] - This is followed by the Fab heavy chain of antibody 1 (CH1-VH) and the continuous Fab heavy chain of antibody 2 (CH1-VH), the latter linked by a hinge-derived peptide linker sequence.

[0120] Furthermore, during protein expression, the resulting heavy chains assemble into dimers, while the co-expressed antibody 1 and antibody 2 light chains (VL-CL) bind to their homologous heavy chains to form the final tandem F(ab)'2-Fc molecule.

[0121] Antibody 1 (Ab1) and antibody 2 (Ab2) are different.

[0122] In a preferred embodiment, a bispecific antibody is described, which comprises

[0123] - Two Fab fragments with different mutant CH1 and mutant CL domains, which consist of the following

[0124] a) A Fab fragment containing mutant CH1 and mutant C-κ domains derived from human IgG1 / κ, and VH and VL domains of Ab1.

[0125] b) A Fab fragment containing mutant CH1 and mutant C-κ domains derived from human IgG1 / κ, and VH and VL domains of Ab2.

[0126] c) Mutated light chain constant domain derived from the human κ constant domain,

[0127] The Fab fragments are arranged in series in the following order.

[0128] The C-terminus of the mutated CH1 domain of the Ab1 Fab fragment is linked to the N-terminus of the VH domain of the Ab2 Fab fragment via a peptide linker.

[0129] - The hinge region of human IgG1 connects the C-terminus of the mutated CH1 domain of the Ab2 fragment to the N-terminus of the CH2 domain.

[0130] - Dimerization of the CH2 and CH3 domains of human IgG1.

[0131] Ab1 and Ab2 can be any target antibody, especially any therapeutic target antibody.

[0132] In a particular implementation, the different Ab1 and Ab2 are independently selected from anti-CD38 antibodies (e.g., daratumumab) and anti-PD-L1 antibodies (e.g., atezolizumab).

[0133] In another specific embodiment, the different Ab1 and Ab2 are independently selected from anti-EGFR antibody and anti-HER2 / neu receptor. In a preferred embodiment, the different Ab1 and Ab2 are independently selected on the one hand from cetuximab or a mutant derivative thereof, and on the other hand from trastuzumab or a mutant derivative thereof.

[0134] These antibodies can be used as drugs, and more specifically for the treatment of cancer.

[0135] In a specific instance, the bispecific molecule is a bispecific anti-CD38, anti-PD-L1 antibody comprising, preferably, the following: a) two heavy chains, each comprising SEQ ID NO: 7, preferably comprising, and b) four light chains, two comprising SEQ ID NO: 15, preferably comprising, and the other two comprising SEQ ID NO: 18, preferably comprising, and b) four light chains, two comprising SEQ ID NO: 15, preferably comprising, and the other two comprising SEQ ID NO: 18, preferably comprising, and b) such a bispecific antibody is called BiXAb-6567.

[0136] SEQ ID NO: 7 (heavy chain) is:

[0137] EVQLLESGGGLVQPGGSLRLSCAVSGFTFNSFAMSWVRQAPGKGLEWVSAISGSGGGTYYADSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYFCAKDKILWFGEPVFDYWGQGTLVTVSSASTKGPSVFPQAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSLVSVVTVPSSSLGTQTYICNVNHKPSNTKVDKRVEPKSCDKTHTSPPAPAPELLGGPAAPPAPAPAGGEVQLVESGGGLVQPGGSLRLSCAASGFTFSDSWIHWVRQAPGKGLEWVAWISPYGGSTYYADSVKGRFTISADTSKNTAYLQMNSLRAEDTAVYYCARRHWPGGFDYWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSLSSVVDVPSSSLGTQTYICNVNHKPSNTKVDKRVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSREEMTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPGK

[0138] This heavy chain sequence contains

[0139] - the VH of daratumumab (SEQ ID NO: 8)

[0140] EVQLLESGGGLVQPGGSLRLSCAVSGFTFNSFAMSWVRQAPGKGLEWVSAISGSGGGTYYADSVKGRFTISRDNSKNTLYLQMNSLRAEDTAVYFCAKDKILWFGEPVFDYWGQGTLVTVSS

[0141] -daratumumab Fab's CH1 domain (G1m(3) allotype of human IgG1 with mutations L124Q and S188V) (SEQ ID NO: 9)

[0142] ASTKGPSVFPQAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSLVSVVTVPSSSLGTQTYICNVNHKPSNTKVDKRV

[0143] -AP connector (SEQ ID NO: 3)

[0144] EPKSCDKTHTSPPAPAPELLGGPAAPPAPAPAGG

[0145] -VH of atezolizumab (SEQ ID NO: 10)

[0146] EVQLVESGGGLVQPGGSLRLSCAASGFTFSDSWIHWVRQAPGKGLEWVAWISPYGGSTYYADSVKGRFTISADTSKNTAYLQMNSLRAEDTAVYYCARRHWPGGFDYWGQGTLVTVSS

[0147] -CH1 domain (Fab atezolizumab G1m(3) allotype of human IgG1 with mutation T192D (SEQ ID NO: 11)

[0148] ASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSLSSVVDVPSSSLGTQTYICNVNHKPSNTKVDKRV

[0149] - The hinge of human IgG1 (SEQ ID NO: 12)

[0150] EPKSCDKTHTCPPCP

[0151] -CH2 domain of human IgG1 (SEQ ID NO: 13)

[0152] APELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAK

[0153] -G1m(3) CH3 domain of allogeneic human IgG1 (SEQ ID NO: 14)

[0154] GQPREPQVYTLPPSREEMTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPGK

[0155] The light chain SEQ ID NO: 15 (daratumumab) is

[0156] EIVLTQSPATLSLSPGERATLSCRASQSVSSYLAWYQQKPGQAPRLLIYDASNRATGIPARFSGSGSGTDFTLTISSLEPEDFAVYYCQQRSNWPPTFGQGTKVEIK RTVAAPSVFIFPPSDEQLKSGTASVTCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLVSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC

[0157] The light chain sequence contains

[0158] -VL of daratumumab (SEQ ID NO: 16)

[0159] EIVLTQSPATLSLSPGERATLSCRASQSVSSYLAWYQQKPGQAPRLLIYDASNRATGIPARFSGSGSGTDFTLTISSLEPEDFAVYYCQQRSNWPPTFGQGTKVEIK

[0160] - The CKappa domain of daratumumab with mutant V133T and S176V (SEQ ID NO: 17)

[0161] RTVAAPSVFIFPPSDEQLKSGTASVTCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLVSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC

[0162] The light chain SEQ ID NO: 18 (atezolizumab) is

[0163] DIQMTQSPSSSLSASVGDRVTITCRASQDVSTAVAWYQQKPGKAPKLLIYSASFLYSGVPSRFSGSGSGTDFTLTISSLQPEDFATYYCQQYLYHPATFGQGTKVEIK RTVAAPAVFIFPPSDEQLKSGTASVVCLLKNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC

[0164] The light chain sequence contains

[0165] - VL of atezolizumab (SEQ ID NO: 19)

[0166] DIQMTQSPSSSLSASVGDRVTITCRASQDVSTAVAWYQQKPGKAPKLLIYSASFLYSGVPSRFSGSGSGTDFTLTISSLQPEDFATYYCQQYLYHPATFGQGTKVEIK

[0167] - The CKappa domain of atezolizumab with mutants S114A and N137K (SEQ ID NO: 20)

[0168] RTVAAPAVFIFPPSDEQLKSGTASVVCLLKNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC

[0169] The third pair of Fab domains (for antigen targets different from the first and second pairs of Fab domains) can also be linked to the same or different peptide linkers: i) via the C-terminus of the heavy chain of the third pair of Fab domains to the N-terminus of the heavy chain of the outer (i.e., Fc distal) Fab domain, ii) or via the C-terminus or N-terminus of the heavy chain of the third pair of Fab domains to the C-terminus of the two heavy chains of the Fab domain.

[0170] Any molecule described herein may be modified to include additional non-protein moieties known in the art and readily available, for example, through PEGylation or high glycosylation. Modifications may be included to improve serum half-life or stability against proteolytic degradation.

[0171] The antibodies of the present invention may or may not be glycosylated, or may exhibit a variety of glycosylation profiles. In a preferred embodiment, the antibody is not glycosylated at the variable region of the heavy chain, but is glycosylated at the Fc region.

[0172] Humanized forms of reference non-human antibodies can be used. In the humanization approach, the complementarity-determining region (CDR) and certain other amino acids from the donor variable region are transplanted into the human variable receptor region, which is then linked to the human constant region. See, for example, Riechmann et al., Nature 332:323-327 (1988); U.S. Patent No. 5,225,539.

[0173] Another example of the construct of the present invention is a bispecific antigen-binding fragment Fab-Fab that does not contain an Fc domain.

[0174] Such Fab-Fab constructs typically contain two distinct Fab domains. These antibodies have only one Fab domain, each binding to antigen 1 and antigen 2 respectively. They have the same light chain as the corresponding BiXAb antibodies; however, the heavy chain of the Fab-Fab is shortened in such a way that their C-terminal residue is cysteine ​​220 (according to EU designation).

[0175] Another example of the construct of the present invention is a dimerized Fab-Fab construct (e.g., through disulfide bonds in the joint or through natural disulfide bonds at the hinge of the C-terminus of the C-terminal proximal Fab structural domain).

[0176] Connector design

[0177] A polypeptide linker, also named a "hinge-derived polypeptide linker sequence" or "pseudo-hinge linker," comprises all or part of the hinge region sequence of one or more immunoglobulins selected from preferably human-origin IgA, IgG, and IgD. The polypeptide linker may comprise all or part of the hinge region sequence of only one immunoglobulin. In this case, the immunoglobulin may belong to the same isotype and subclass as the immunoglobulin derived from the adjacent CH1 domain, or to different isotypes or subclasses. Alternatively, the polypeptide linker may comprise all or part of the hinge region sequences of at least two immunoglobulins of different isotypes or subclasses. In this case, the N-terminal portion of the polypeptide linker directly following the CH1 domain preferably consists of all or part of the hinge region of an immunoglobulin that belongs to the same isotype and subclass as the immunoglobulin derived from the CH1 domain.

[0178] Optionally, the polypeptide linker may further comprise a sequence of 2 to 15, preferably 5 to 10, N-terminal amino acids of the CH2 domain of an immunoglobulin.

[0179] The polypeptide linker sequence typically consists of fewer than 80 amino acids, preferably fewer than 60 amino acids, and still more preferably fewer than 40 amino acids.

[0180] In some cases, sequences from the natural hinge region can be used; in others, these sequences can be point-mutated, particularly by replacing one or more cysteine ​​residues in the natural IgG1, IgG2, or IgG3 hinge sequences with alanine or serine, to avoid unwanted intra- or inter-chain disulfide bonds.

[0181] In a particular embodiment, the peptide linker sequence comprises or consists of the following amino acid sequences: EPKX1CDKX2HX3X4PPX5PAPELLGGPX6X7PPX8PX9PX 10 GG (SEQ ID NO:1), wherein X1, X2, X3, X4, X5, X6, X7, X8, X9, and X10 are the same or different, and are any amino acids. In particular, the polypeptide linker sequence may comprise or consist of sequences selected from or composed of the following:

[0182] EPKSCDKTHTSPPAPAPELLGGPGGPPGPGPGGG (SEQ ID NO: 2);

[0183] EPKSCDKTHTSPPAPAPELLGGPAAPPAPAPAGG (SEQ ID NO: 3);

[0184] EPKSCDKTHTSPPAPAPELLGGPAAPPGPAPGGG (SEQ ID NO:4);

[0185] EPKSCDKTHTCPPCPAPELLGGPSTPPTPSPSGG (SEQ ID NO:5) and EPKSCDKTHTSPPSPAPELLGGPSTPPTPSPSGG (SEQ ID NO:6).

[0186] In a specific implementation, X1, X2, and X3 may be the same or different, being either threonine (T) or serine (S).

[0187] In another specific implementation, X1, X2 and X3 may be the same or different, selected from Ala (A), Gly (G), Val (V), Asn (N), Asp (D) and Ile (I), with X1 still preferred. X2 and X3 may be the same or different, and may be Ala (A) or Gly (G).

[0188] Alternatively, X1, X2, and X3 may be the same or different, and may be Leu (L), Glu (E), Gln (Q), Met (M), Lys (K), Arg (R), Phe (F), Tyr (T), His (H), or Trp (W), with Leu (L), Glu (E), or Gln (Q) being preferred.

[0189] In a particular implementation, X4 and X5 may be the same or different, and are any amino acid selected from serine (S), cysteine ​​(C), alanine (A), and glycine (G).

[0190] In a preferred embodiment, X4 is serine (S) or cysteine ​​(C).

[0191] In a preferred embodiment, X5 is alanine (A) or cysteine ​​(C).

[0192] In specific implementation schemes, X6, X7, X8, X9, X 10 The same or different amino acids are any amino acids other than threonine (T) or serine (S). Preferably, X6, X7, X8, X9, and X10 are the same or different amino acids selected from Ala (A), Gly (G), Val (V), Asn (N), Asp (D), and Ile (I).

[0193] Or, X6, X7, X8, X9, X 10 The same or different can be Leu (L), Glu (E), Gln (Q), Met (M), Lys (K), Arg (R), Phe (F), Tyr (T), His (H), Trp (W), with Leu (L), Glu (E) or Gln (Q) preferred.

[0194] In the preferred embodiment, X6, X7, X8, X9, X 10 Same or different, selected from Ala (A) and Gly (G).

[0195] In yet another preferred embodiment, X6 and X7 are the same, and are preferably selected from Ala (A) and Gly (G).

[0196] In a preferred embodiment, the polypeptide adapter sequence comprises or consists of the sequence SEQ ID NO: 1, wherein...

[0197] X1, X2 and X3 may be the same or different, and are threonine (T) and serine (S).

[0198] X4 is either serine (S) or cysteine ​​(C).

[0199] X5 is either alanine (A) or cysteine ​​(C);

[0200] X6, X7, X8, X9, X 10 Same or different, selected from Ala (A) and Gly (G).

[0201] In another preferred embodiment, the polypeptide adapter sequence comprises or consists of the sequence SEQ ID NO: 1, wherein

[0202] If X1, X2, and X3 are the same or different, it is either Ala (A) or Gly (G).

[0203] X4 is either serine (S) or cysteine ​​(C).

[0204] X5 is either alanine (A) or cysteine ​​(C);

[0205] X6, X7, X8, X9, X 10 Same or different, selected from Ala (A) and Gly (G).

[0206] In the case of multispecific antigen-binding fragments containing two or more different Fab fragments, the peptide linkers separating the Fab fragments can be the same or different.

[0207] The generation of multispecific antibodies:

[0208] Those skilled in the art may refer to international patent application WO2013 / 005194, the general technique for expressing multispecific antibodies of which is incorporated herein by reference.

[0209] Additionally, this document describes polynucleotides comprising sequences of protein chains encoding molecules or antibodies of the present invention. These polynucleotides may also contain additional sequences: specifically, they may advantageously contain sequences encoding leader sequences or signal peptides that allow the secretion of said protein chains. Host cells transformed with said polynucleotides are also disclosed.

[0210] Typically, DNA sequences are engineered using the amino acid sequences of different monoclonal antibodies, optionally after codon optimization for mammalian expression. For the heavy chain, synthetic DNA is used, encoding a signal peptide, the variable region of Fab1 and the constant CH1 domain, followed by a hinge head and the variable region and constant CH1 domain of Fab2, along with flanking sequences for restriction enzyme digestion. For the light chain, synthetic DNA is used, encoding a signal peptide and both variable and constant κ regions.

[0211] The nucleic acids encoding the heavy and light chains of the antibody of the present invention are inserted into an expression vector. The light and heavy chains can be cloned in the same or different expression vectors. The DNA segment encoding the immunoglobulin chain is effectively linked to a control sequence in the expression vector (which ensures the expression of the immunoglobulin polypeptide). Such control sequences include signal sequences, promoters, enhancers, and transcription termination sequences. The expression vector typically replicates in the host organism either as an episome or as an integrated portion of the host chromosomal DNA. Typically, the expression vector will contain a selection marker (e.g., tetracycline or neomycin) to allow detection of those cells transformed with the desired DNA sequence.

[0212] In one instance, the heavy and light chain coding sequences (e.g., sequences encoding VH and VL, VH-CH1 and VL-CL, or full-length heavy and light chains) are both contained in a single expression vector. In another instance, each heavy and light chain of the antibody is cloned into a separate vector. In the latter case, the expression vectors encoding the heavy and light chains can be co-transfected into a single host cell to express both chains, which can then be assembled in vivo or in vitro to form a complete antibody. Alternatively, the expression vectors encoding the heavy chain and / or those encoding the light chain can be introduced into different host cells to express each of the heavy and light chains, which can then be purified and assembled in vitro to form a complete antibody.

[0213] In a particular implementation, host cells are co-transfected with three independent expression vectors (e.g., plasmids), resulting in the co-production of all three chains (referred to as the heavy chain HC and the two light chains LC1 and LC2, respectively) and the secretion of multispecific antibodies, such as bispecific antibodies.

[0214] More specifically, these three carriers can be used advantageously in a molecular ratio of 2:1:1 (HC:LC1:LC2).

[0215] The host cell can be further transformed with at least two polynucleotides encoding two different light chains by an expression vector containing a polynucleotide encoding a heavy chain as defined herein: a first light chain that specifically pairs with the first VH / CH1 region of the heavy chain; and a second light chain that specifically pairs with the second VH / CH1 region of the heavy chain.

[0216] In other embodiments, the host cell may be additionally transformed with a polynucleotide encoding a third light chain, which is different from the first and second light chains and specifically pairs with the third VH / CH1 region of the heavy chain.

[0217] Recombinant vectors used to express the antibodies described herein typically contain nucleic acids encoding an antibody amino acid sequence that is efficiently linked to a constitutive or inducible promoter. The vector may be adapted for replication and integration in prokaryotes, eukaryotes, or both. Typical vectors contain transcription and translation terminators, initiation sequences, and promoters that can be used to regulate the expression of the nucleic acid encoding the antibody. Optionally, the vector contains a universal expression cassette containing at least one independent terminator sequence, a sequence allowing the cassette to replicate in both eukaryotes and prokaryotes (i.e., a shuttle vector), and selection markers for prokaryotic and eukaryotic systems.

[0218] Multispecific, such as bispecific, antibodies described herein can be produced in prokaryotic or eukaryotic expression systems, such as bacteria, yeast, filamentous fungi, plant insects (e.g., using baculovirus vectors), and mammalian cells. The recombinant antibodies of the present invention need not be glycosylated or expressed in eukaryotic cells; however, expression in mammalian cells is generally preferred. Examples of useful mammalian host cell lines are human embryonic kidney cells (293 cells), hamster kidney cells (BHK cells), Chinese hamster ovary cells / − or + DHFR (CHO, CHO-S, CHO-DG44, Flp-in CHO cells), African green monkey kidney cells (VERO cells), and human hepatocytes (Hep G2 cells).

[0219] Mammalian tissue cell cultures are preferred for expressing and producing peptides because many suitable host cell lines capable of secreting intact immunoglobulins have been developed in the art, including CHO cell lines, various Cos cell lines, HeLa cells, preferably myeloma cell lines (e.g., NSO) or transformed B cells or hybridomas.

[0220] In the most preferred embodiment, the multispecific, such as bispecific, antibody of the present invention is produced by using a CHO cell line, most advantageously a CHO-S or CHO-DG-44 cell line or a derivative thereof.

[0221] The expression vectors of these cells may include expression control sequences, such as origin of replication, promoters, and enhancers, as well as essential processing information sites, such as ribosome binding sites, RNA splicing sites, polyadenylation sites, and transcription terminator sequences. Preferred expression control sequences are promoters derived from immunoglobulin genes, SV40, adenovirus, bovine papillomavirus, cytomegalovirus, etc.

[0222] Vectors containing target polynucleotide sequences (e.g., heavy and light chain coding sequences and expression control sequences) can be transferred into host cells using well-known methods, which vary depending on the type of cell host. For example, calcium phosphate treatment or electroporation can be used for other cell hosts. (See Sambrook et al., Molecular Cloning: A Laboratory Manual (Cold Spring Harbor Press, 2nd ed., 1989)). When cloning heavy and light chains on separate expression vectors, co-transfection of the vectors is performed to obtain complete expression and assembly of the immunoglobulin.

[0223] Transform or transfect host cells with a vector (e.g., by chemical transfection or electroporation) and culture them in a standard nutrient medium (or a modified medium where appropriate) to induce promoters, select transformants, or amplify genes encoding desired sequences.

[0224] Antibody expression can be transient or stable.

[0225] Preferably, multispecific antibodies, such as bispecific antibodies, are generated through a stable expression method, in which cell lines stably transfected with DNA encoding all polypeptide chains of a multispecific antibody, such as BiXAb-6567, are capable of sustained expression, which allows for the preparation of therapeutic agents. Stable expression in CHO cell lines is particularly advantageous, for example.

[0226] Once expressed, the complete antibody of the present invention, its dimer, single light and heavy chains, or other immunoglobulin forms can be further isolated or purified to obtain a substantially homogeneous formulation for further assays and applications. Standard protein purification methods known in the art can be used. For example, suitable purification procedures may include fractionation on an immunoaffinity column or ion exchange column, ethanol precipitation, high-performance liquid chromatography (HPLC), sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), ammonium sulfate precipitation, and gel filtration (see Scopes, Protein Purification (Springer-Verlag, NY, 1982)). For pharmaceutical use, a substantially pure composition with at least about 90-95% homogeneity is preferred, and most preferably 98-99% or higher homogeneity.

[0227] In vitro production allows for scale-up to produce large quantities of the desired multispecific, such as bispecific, antibodies of the present invention. Such methods can employ homogeneous suspension culture in, for example, an airlift reactor or a continuous stirred reactor, or immobilized or entrapped cell culture on agarose beads or ceramic tubes, for example, in hollow fibers or microcapsules.

[0228] application

[0229] The protein constructs or multispecific antibodies of the present invention can be used in all applications of multispecific antibodies. In particular, they can be used to obtain pharmaceuticals that are suitable for a wide range of therapeutic and diagnostic medical applications, either in patients (by combining with radioactive, fluorescent, chemiluminescent, or any other labeling) or in vitro (e.g., using immunohistochemistry, immunoblotting, and immunofluorescence techniques for cell and tissue staining). These pharmaceuticals or diagnostic products are also part of the objectives of this invention.

[0230] Other aspects of the invention are pharmaceutical compositions comprising a protein construct or antibody according to the invention. Another aspect of the invention is the use of a protein construct or antibody according to the invention in the preparation of a pharmaceutical composition. Other aspects of the invention are methods for manufacturing pharmaceutical compositions comprising a protein construct or antibody according to the invention.

[0231] In another aspect, the present invention provides compositions comprising, for example, pharmaceutical compositions, protein constructs or antibodies as defined herein formulated together with a drug carrier.

[0232] The compositions of the present invention can be applied by a variety of methods known in the art.

[0233] This application further relates to the following implementation schemes:

[0234] 1. A protein construct comprising a Fab fragment selected from the following mutants:

[0235] a) The Fab fragment, which consists of the following:

[0236] - The VH and VL domains of the target antibody;

[0237] The -CH1 domain is derived from the CH1 domain by replacing the threonine residue at position 192 of the CH1 domain of the immunoglobulin with aspartic acid.

[0238] -CL domain, which is derived from the CL domain by replacing the asparagine residue at position 137 of the immunoglobulin's CL domain with a lysine residue and replacing the serine residue at position 114 of the CL domain with an alanine residue; and

[0239] b) The Fab fragment, which consists of the following:

[0240] - The VH and VL domains of the target antibody;

[0241] -CH1 domain, which is derived from the CH1 domain by replacing the leucine residue at position 124 of the CH1 domain of the immunoglobulin with glutamine and replacing the serine residue at position 188 of the CH1 domain with a valine residue; and

[0242] -CL domain, which is derived from the CL domain by replacing the valine residue at position 133 of the immunoglobulin with a threonine residue and replacing the serine residue at position 176 of the CL domain with a valine residue.

[0243] 2. The protein construct of embodiment 1, comprising or composed of a multispecific antigen-binding fragment, wherein the multispecific antigen-binding fragment comprises at least two Fab fragments having different CH1 and CL domains, wherein each Fab fragment recognizes a different target epitope, and the Fab fragments are arranged in tandem in any order, wherein the C-terminus of the CH1 domain of the first Fab fragment is connected to the N-terminus of the VH domain of a subsequent Fab fragment via a peptide linker.

[0244] At least one of the Fab fragments consists of a mutated Fab fragment as defined in Embodiment 1, and preferably, two of the Fab fragments are mutated Fab fragments as defined in Embodiment 1.

[0245] 3. The protein construct of embodiment 2, which is a multispecific antibody having two identical antigen-binding arms, each arm consisting of a multispecific antigen-binding fragment as defined in embodiment 2.

[0246] 4. The protein construct of implementation scheme 3, having an immunoglobulin-like structure, including:

[0247] - Two identical antigen-binding arms, each arm consisting of a multispecific antigen-binding fragment as defined in Embodiment 1;

[0248] -The CH2 and CH3 domains of immunoglobulin dimerization;

[0249] - A hinge region of IgA, IgG, or IgD that connects the C-terminus of the CH1 domain of the antigen-binding arm to the N-terminus of the CH2 domain.

[0250] 5. The protein construct of embodiment 4, preferably a bispecific antibody, comprising at least two heavy chains and four light chains.

[0251] Each heavy chain contains

[0252] a. The Fc region of an immunoglobulin containing a hinge-CH2-CH3 domain.

[0253] b. The Fc region is connected to the Fab heavy chain CH1-VH of antibody 1 (Ab1) via the hinge domain.

[0254] c. It then links to the CH1-VH domain of the Fab heavy chain of antibody 2 (Ab2) via a peptide linker sequence, and the peptide linker sequence connects the N-terminus of the VH domain of the Fab heavy chain of Ab1 to the C-terminus of the CH1 domain of Ab2.

[0255] Furthermore, the four light chains contain the Fab light chain CL-VL of Ab1 and the Fab light chain CL-VL of Ab2, which are related to their cognate heavy chain structural domains.

[0256] Ab1 and Ab2 identify different epitopes.

[0257] Furthermore, the Fab CH1 domain of either Ab1 or Ab2 is a mutated domain, derived from the CH1 domain by replacing the threonine residue at position 192 of the immunoglobulin with an aspartic acid residue, and the homologous CL domain is a mutated domain, derived from the CL domain by replacing the asparagine residue at position 137 of the immunoglobulin with a lysine residue and replacing the serine residue at position 114 of the CL domain with an alanine residue, and / or

[0258] The Fab CH1 domain of either Ab1 or Ab2 is a mutated domain derived from the CH1 domain by replacing the leucine residue at position 124 with glutamine and the serine residue at position 188 with valine. The homologous CL domain is a mutated domain derived from the CL domain by replacing the valine residue at position 133 with threonine and the serine residue at position 176 with valine.

[0259] 6. The protein construct of embodiment 5, wherein the Fab CH1 domain of Ab1 is a mutated domain derived from the CH1 domain by replacing the threonine residue at position 192 of the immunoglobulin CH1 domain with an aspartic acid residue, and the homologous CL domain is a mutated domain derived from the CL domain by replacing the asparagine residue at position 137 of the immunoglobulin CL domain with a lysine residue and replacing the serine residue at position 114 of the CL domain with an alanine residue, and wherein the Fab of Ab2... The CH1 domain is a mutated domain derived from the CH1 domain by replacing the leucine residue at position 124 with glutamine and the serine residue at position 188 with valine. The homologous CL domain is a mutated domain derived from the CL domain by replacing the valine residue at position 133 with threonine and the serine residue at position 176 with valine.

[0260] 7. The protein construct of embodiment 5, wherein the Fab CH1 domain of Ab2 is a mutated domain derived from the CH1 domain by replacing the threonine residue at position 192 of the immunoglobulin CH1 domain with an aspartic acid residue, and the homologous CL domain is a mutated domain derived from the CL domain by replacing the asparagine residue at position 137 of the CL domain with a lysine residue and replacing the serine residue at position 114 of the CL domain with an alanine residue, and wherein the Fab of Ab1... The CH1 domain is a mutated domain derived from the CH1 domain by replacing the leucine residue at position 124 with glutamine and the serine residue at position 188 with valine. Similarly, the homologous CL domain is a mutated domain derived from the CL domain by replacing the valine residue at position 133 with threonine and the serine residue at position 176 with valine.

[0261] 8. A protein construct according to any one of embodiments 2 to 7, wherein the polypeptide linker sequence comprises or consists of the following amino acid sequence: EPKX1CDKX2HX3X4PPX5PAPELLGGPX6X7PPX8PX9PX 10 GG (SEQ ID NO:1), where X1, X2, X 3、 X4, X5, X6, X7, X8, X9, X 10 It can be any amino acid, whether it is the same or different.

[0262] 9. The protein construct of embodiment 8, wherein the polypeptide linker sequence comprises or consists of sequences selected from the following:

[0263] EPKSCDKTHTSPPAPAPELLGGPGGPPGPGPGGG (SEQ ID NO: 2);

[0264] EPKSCDKTHTSPPAPAPELLGGPAAPPAPAPAGG (SEQ ID NO: 3);

[0265] EPKSCDKTHTSPPAPAPELLGGPAAPPGPAPGGG (SEQ ID NO: 4);

[0266] EPKSCDKTHTCPPCPAPELLGGPSTPPTPSPSGG (SEQ ID NO:5) and EPKSCDKTHTSPPSPAPELLGGPSTPPTPSPSGG (SEQ ID NO:6).

[0267] 10. A protein construct according to any one of implementation schemes 1 to 9, which recognizes both EGFR and HER2 / neu.

[0268] 11. The protein construct of embodiment 10, which is a multispecific antibody, preferably a bispecific antibody, wherein Ab1 and Ab2 are different, and are independently selected from cetuximab or a mutant derivative thereof on the one hand, or from trastuzumab or a mutant derivative thereof on the other hand.

[0269] 12. A protein construct according to any one of implementation schemes 1 to 9, which recognizes both CD38 and PD-L1.

[0270] 13. The protein construct of embodiment 12, which is a multispecific antibody, preferably a bispecific antibody, wherein Ab1 and Ab2 are different, and are independently selected from daratumumab or a mutant derivative thereof on the one hand, and atezolizumab or a mutant derivative thereof on the other hand.

[0271] 14. The protein construct of embodiment 12, which is a bispecific antibody, comprises, preferably, the following: a) two heavy chains, each heavy chain comprising SEQ ID NO: 7, preferably consisting of SEQ ID NO: 7; b) four light chains, two of which comprise SEQ ID NO: 15, preferably consisting of SEQ ID NO: 15, and the other two comprise SEQ ID NO: 18, preferably consisting of SEQ ID NO: 18.

[0272] 15. A polypeptide comprising a heavy chain of an antigen-binding fragment or a heavy chain of a multispecific antibody as defined in any one of embodiments 2 to 14, preferably consisting of a heavy chain of an antigen-binding fragment or a heavy chain of a multispecific antibody as defined in any one of embodiments 2 to 14.

[0273] 16. A polynucleotide containing a sequence encoding the polypeptide of embodiment 15.

[0274] 17. Host cells transfected with an expression vector containing the polynucleotides of implementation scheme 16.

[0275] 18. The host cell of embodiment 17 is further transformed with at least two polynucleotides encoding two different light chains: the first light chain specifically pairs with the first VH / CH1 region of the heavy chain; and the second light chain specifically pairs with the second VH / CH1 region of the heavy chain.

[0276] 19. A method for generating an antigen-binding fragment or multispecific antibody as defined in any one of embodiments 2 to 14, the method comprising the steps of: a) culturing host cells expressing an antibody heavy chain as defined in any one of embodiments 2 to 14 and an antibody light chain as defined in any one of embodiments 2 to 14 in a suitable culture medium and culture conditions; and b) recovering the generated antibody from the culture medium or the cultured cells.

[0277] 20. A multispecific antigen-binding fragment or multispecific antibody as defined in any of embodiments 2 to 14, which is used as a medicine.

[0278] The following embodiments are provided by way of example and are not intended to limit the invention.

[0279] Example

[0280] Example 1: Preparation of the bispecific antibody BiXAb-6567 of the present invention

[0281] Gene synthesis

[0282] After codon optimization for mammalian expression, DNA sequences were designed using the amino acid sequences of anti-CD38 (daratumumab) and anti-PDL1 (atezolizumab) using the GeneScript program. These antibodies were referred to as “parental” anti-CD38 and “parental” anti-PD-L1 mAb.

[0283] The heavy chain DNA construct is designed as follows: a signal peptide, followed by the sequence SEQ ID NO: 7 [consisting of: a variable region, followed by a constant CH1 domain of Fab1 (anti-CD38) with mutations Leu to Gln and Ser to Val introduced at Kabat positions 124 and 188, respectively, followed by an adapter, then the variable region, and then the constant CH1 domain of Fab2 (anti-PD-L1) with the mutation Thr to Asp introduced at Kabat position 192]; flanking sequences for restriction enzyme digestion are introduced at both ends of the heavy chain DNA construct. The light chain DNA construct is designed as follows: a signal peptide (SEQ ID NO: 21), followed by a variable region, followed by a constant κ region. For the anti-CD38 light chain (SEQ ID NO: 15), mutations are introduced at Kabat positions 133 (Val to Thr) and 176 (Ser to Val) of the constant κ domain. For the anti-PDL1 light chain (SEQ ID NO: 18), mutations at positions 114 (Ser to Ala) and 137 (Asn to Lys) of the Kabat domain were introduced into the constant κ domain. All DNA constructs were synthesized by GeneArt.

[0284] PCR was performed using PfuTurbo Hot Start to amplify the insert fragment. The heavy and light chains were then digested with NotI and ApaI, and NotI and HindIII, respectively. The double-digested heavy chain fragment was ligated into a NotI and ApaI-treated pcDNA3.1 expression vector (Invitrogen), which already contained the human IgG1 hinge and CH2-CH3 domain. The double-digested light chain fragment was also ligated into a NotI and HindIII-treated pcDNA3.1 expression vector (Invitrogen). The plasmid DNA was validated by double-stranded DNA sequencing.

[0285] Expression and purification

[0286] Bispecific antibody BiXAb-6567 was generated using transient gene expression in serum-free medium adapted to suspension (CHO SFM-II medium, Life Technologies™) by co-transfecting three genes encoding separate vectors at a molecular ratio of 2:1:1 = HC:LC1:LC2 (1 continuous heavy chain (HC) and 2 light chains (LC)). Typically, for 50 mL scale expression, a total of 50 μg plasmid DNA (25 μg heavy chain, 12.5 μg anti-CD38 light chain, and 12.5 μg anti-PD-L1 light chain) was mixed in a 1.5 mL Eppendorf tube, followed by the addition of 1 mL CHO SFM medium containing 25 μL of 3 mg / mL PEI transfection reagent at pH 7.0 (Polyplus), and the reaction was incubated at room temperature for 20 min. The DNA-PEI mixture was then added to 49 mL of Life Technologies Invitrogen FreeStyle™ CHO-S cells (1–2 x 10⁻⁶ cells) in a 125 mL shake flask. 6 / mL). Cells were agitated for 6 days. The supernatant was collected by centrifugation at 3,000 rpm for 15 minutes. The expression titer of BiXAb-6567 in the supernatant was determined using a ForteBio protein A biosensor (Octet® Systems). BiXAb-6567 was then purified on protein A affinity resin (MabSelect SuRe, GE Healthcare Life Sciences). The antibody was eluted from protein A with 0.1 M glycine at pH 3.5, and the elution buffer was neutralized with 1 M TRIS. The purified antibody in Dulbecco PBS (Lonza) was sterile filtered (0.2 μM sterile filter, Techno Plastic Products AG), and the final concentration was determined by reading the optical density (OD) at 280 nm (Eppendorf BioSpectrometer®).

[0287] BiXAab-6567 typically exhibits good expression titers (>180 mg / L) in transient CHO expression. This expression level is comparable to that seen with conventional monoclonal antibodies.

[0288] SDS-Pyramid Gel Electrophoresis

[0289] To assess the quality of purified BiXAb-6567, we performed SDS-PAGE (Experion™ automated electrophoresis system, BioRad). In the presence of sodium dodecyl sulfate (SDS) in the run buffer, the rate of antibody migration in the gel depends primarily on its size, thus allowing for molecular weight determination. The assay was performed under both non-reducing and reducing conditions; the latter allows for the disruption of disulfide bonds and therefore visualization of individual polypeptide chains (light and heavy chains).

[0290] exist Figure 1 SDS-PAGE data are presented. Under non-reducing conditions, the quaternary structure of the antibody is maintained, and the observed molecular weight should represent the sum of the molecular weights of the different heavy and light chains. The bispecific antibody of this invention (BiXAb-6567) consists of six chains: two heavy chains and four light chains. The theoretical molecular weight of BiXab-6567 is 244.40 kDa, not considering post-translational modifications (PTM), such as N-glycosylation in the Fc region of asparagine at position 297. The gel was calibrated using a mixture of standards with known molecular weights. The non-reducing data show a main band close to the 250 kDa molecular weight standard, consistent with the calculated molecular weight and the expected glycosylation of the two asparagine molecules at position 297 of the Fc domain. Under reducing conditions, dithiothreitol (DTT) further denatures BiXAb-6567 by reducing disulfide bonds and disrupting the quaternary structure, thus the six polypeptide chains should migrate separately in the gel according to their molecular weights. The two identical heavy chains of BiXAb-6567 co-migrated as one band, while the two pairs of light chains co-migrated as a second band due to their nearly identical molecular weights. Therefore, based on migration rates calibrated by molecular weight, the data showed two main bands at approximately 75 kDa and 25 kDa. Each heavy chain shares an N-glycosylation site at asparagine 297, which explains the width of the higher molecular weight band, and the observed molecular weight is slightly higher than the calculated molecular weight (75.44 kDa); this broadening is typical for glycosylated proteins. The calculated molecular weights of the anti-CD38 (23.40 kDa) and anti-PD-L1 (23.36 kDa) light chains are very similar, thus contributing to their co-migrating.

[0291] In summary, when the presence of N-glycosylation sites in the heavy chain is taken into account, the SDS-PAGE of BiXAb-6567 exhibits the expected spectra under both reducing and non-reducing conditions, and is consistent with the calculated theoretical molecular weight.

[0292] Size exclusion chromatography analysis

[0293] Protein aggregation is frequently observed in modified protein molecules. We performed analytical size exclusion chromatography (SEC) to determine the high molecular weight species content of the one-step affinity-purified BiXAb-6567 formulation (see Expression and Purification of Variants). We used an SEC-s3000 (300 x 7.8 mm) column (BioSep) and an Aktapurifier 10 system (GE Healthcare); the assay was performed using PBS buffer at pH 7.4 at a flow rate of 1 mL / min.

[0294] Figure 2 The SEC chromatogram presented confirms that the main peak corresponds to the expected size of the monomer BiXAb-6567; this peak accounts for 98.2% of the total sample. Furthermore, a small peak was observed, corresponding to a higher molecular weight species (possibly a dimer); this peak accounts for 1.8% of the total sample. Therefore, we conclude that the percentage of higher molecular weight species is very small and similar to that of conventional monoclonal antibodies produced in CHO expression systems. The narrow and symmetrical shape of the monomer peaks suggests that BiXAb-6567 has been correctly assembled and is represented by a single species.

[0295] Example 2: Characterization of BiXAb-6567 by Differential Scanning Calorimetry (DSC)

[0296] Differential scanning calorimetry (DSC) was used to compare the thermal stability of BiXAb-6567, parental anti-CD38 mAb, and parental anti-PD-L1 mAb. A Microcal™ VP-Capillary DSC system (Malvern Instruments) was used to perform the differential scanning calorimetry experiments.

[0297] All samples were centrifuged (20,000 x g, 5 min, 4°C) and their protein content was quantified before DSC analysis using a Nanodrop ND-1000 spectrophotometer (Thermo Scientific) with an IgG analysis program. For assay, all samples were diluted in PBS to a final concentration of 1 mg / mL.

[0298] A pre-equilibration time of 3 minutes was used, and the resulting thermal chromatogram was obtained at a scan rate of 60°C / h between 20 and 110°C, with a filtration time of 25 seconds and moderate feedback. Five buffer / buffer scans were performed to stabilize the instrument before sample analysis, with buffer / buffer scans performed between each protein / buffer scan. Data conformed to a non-2-state unfolding model, and pre- and post-conversion values ​​were adjusted by subtracting the baseline.

[0299] Figure 3The DSC curves presented (covering the range of 50 to 100°C) demonstrate that different Fv regions can lead to different ways of unfolding the Fab spectrum; the experiment also demonstrates that the Fv region determines the apparent stability of the Fab. The DSC spectrum of the CD38 mAb resistance shows two transitions: with 170 kcal / mole / o C's Cp max and 70.9 o The large peak of C at Tm1 corresponds to the unfolding of both the CH2 and Fab domains, and has a value of 20 kcal / mole / o C's Cp max and 81.5 o The small peak at Tm2 of C corresponds to the unfolding of the CH3 domain. The DSC spectrum of the anti-PD-L1 mAb shows two transitions: one with 20 kcal / mole / o C's Cp max and 69.9 o The small peak of Tm1 in C corresponds to the unfolding of the CH2 domain and has a value of 160 kcal / mole / o C's Cp max and 83.4 o The large peak of Tm2 in C corresponds to the unfolding of the CH3 and Fab domains.

[0300] The DSC spectrum of BiXAb-6567 also showed two transitions with two large peaks. The first peak has a value of 130 kcal / mole / o C's Cp max and 71.5 o The Tm1 of C corresponds to the unfolding of the CH2 and Fab domains, which are resistant to CD38 mAb. The second peak has 170 kcal / mole / o C's Cp max and 81.5 o The Tm2 of C corresponds to the unfolding of the CH3 and Fab domains of the anti-PD-L1 mAb. Therefore, the DSC spectrum of BiXAb-6567 is similar to the superposition of the two DSC spectra of the two parent mAbs, illustrating the excellent assembly and stability of BiXAb-6567. The Tonset (63.3) of BiXAb-6567... o C) Similar to the parent mAb in Tonset (anti-CD38 Tonset = 63.5) o C, while anti-PD-L1 Tonset = 63.2 o(C) indicates that BiXAb-6567 has stability properties similar to those of the parent antibodies. The calculated ΔH of BiXAb-6567 is 1560 kcal / mol, reflecting the larger size of the bispecific molecule relative to the two parent antibodies (anti-CD38 ΔH = 963 kcal / mol, while anti-PD-L1 ΔH = 820 kcal / mol).

[0301] definition:

[0302] Tm, or denaturation / melting temperature, is the point at which the concentrations of unfolded and folded substances are equal, and it represents the midpoint of the unfolding transition. As a parameter, it describes a protein's sensitivity to thermal denaturation and is therefore related to its stability. The higher the Tm, the more stable the protein.

[0303] Tonset is the temperature at which the unfolding transition begins. This parameter is typically 5 to 10°C lower than Tm. It is also a parameter describing protein stability, but is related to resistance to heat denaturation.

[0304] ΔH is the enthalpy of pyrolysis and reflects the disruption of intramolecular interactions in a protein (i.e., the disruption of intradomain and interdomain interactions). Pyrolysis is endothermic and therefore produces a positive enthalpy. Enthalpy (ΔH) is the area below the pyrolysis transition peak.

[0305] Example 3: Cell-free binding properties of BiXAb-6567

[0306] Direct CD38 antigen binding plate ELISA assay

[0307] 100 μl of parental mAbs, each at a concentration of 3 μg / mL, prepared by dilution with PBS pH 7.4, anti-CD38, or anti-PDL1 were used to coat Maxisorp plates overnight at 4°C. Additionally, BiXAb-6567 at a concentration of 5 µg / mL, prepared by dilution with PBS pH 7.4, was used to coat Maxisorp plates overnight at 4°C. Plates were washed five times with 1x PBS (PBST) containing 0.05% Tween-20, and then blocked for 200 h at room temperature with 200 μL / well of 1% BSA in 1x PBS. The plates were then washed five times with 1x PBST. A 7-point 3-fold dilution series of recombinant CD38 His / Flag-labeled (Creative Biomart) starting at 1 µg / mL in 1x PBS was prepared; 100 µL of each dilution step was added to each well. Plates were incubated at room temperature for 1 h and washed five times with 1x PBST. Add 100 µL / well of anti-Flag-tag antibody-conjugated HRP (Abcam) diluted 10,000 times in 1x PBS and incubate the plates at room temperature for 1 hour. After washing 5 times with 1x PBST, add 100 x L / well of TMB substrate in 1x PBS for colorimetric reading and incubate the plates at room temperature for 15 minutes for color development. Collect the assay data at 650 nm using a Victor2 microplate reader (Perkin Elmer).

[0308] BiXAb-6567 exhibited a dose-dependent binding curve very similar to that of the parental anti-CD38 antibody. Figure 4A The EC50 values ​​for CD38 binding to the two antibodies were as follows: EC50 [BiXAb-6567] = 171 ng / mL, while EC50 [anti-CD38] = 199 ng / mL. This result indicates that BiXAb-6567 possesses a correctly assembled anti-CD38 Fab domain, as it exhibits similar binding to the parental anti-CD38 mAb. As expected, the parental anti-PDL1 mAb, used as a negative control, did not show any binding.

[0309] Direct PDL1 antigen binding plate ELISA analysis.

[0310] 100 µL of biotinylated human PD-L1 protein (AcroBiosystems) at a concentration of 1 µg / mL, prepared by dilution with 1x PBS at pH 7.4, was used to coat Maxisorp plates overnight at 4°C. The plates were washed five times with PBST and then blocked for 2 hours at room temperature with 200 μL / well of 1% BSA in 1x PBS. The plates were then washed five times with 1x PBST. A seven-point 3-fold dilution series of anti-CD38 mAb (starting at 0.3 mg / mL), anti-PD-L1 mAb (starting at 0.3 mg / mL), or BiXAb-6567 (starting at 0.5 mg / mL) in 1x PBS solution was prepared; 100 µL of each dilution step was added to each well. The plates were incubated at room temperature for 1 hour and washed five times with 1x PBST. Add 100 µL / well of anti-human antibody (IgG H&L) conjugated HRP (Abliance) diluted 5,000 times in 1x PBS and incubate the plates at room temperature for 1 hour. After washing 5 times with 1x PBST, add 100 x L / well of TMB substrate in 1x PBS for colorimetric reading and incubate the plates at room temperature for 15 minutes for color development. Collect the assay data at 650 nm using a Victor2 microplate reader (Perkin Elmer).

[0311] BiXAb-6567 exhibited a dose-dependent binding curve very similar to that of the parental anti-PD-L1 antibody. Figure 4B The EC50 values ​​for PD-L1 binding to the two antibodies were as follows: EC50 [BiXAb-6567] = 93 ng / mL, while EC50 [anti-PD-L1] = 72 ng / mL. This result indicates that BiXAb-6567 possesses a correctly assembled anti-PD-L1 Fab domain, as it exhibits similar binding to the parental anti-PD-L1 mAb. As expected, the parental anti-CD38 mAb, used as a negative control, did not show any binding.

[0312] Double antigen binding ELISA assay

[0313] 100 µL of 2 µg / mL recombinant human Fc-labeled CD38 (Creative BioMart), prepared by dilution with 1x PBS at pH 7.4, was used to coat Maxisorp plates overnight at 4 °C. The plates were washed five times with 1x PBST and then blocked for 2 hours at room temperature with 1% BSA in 200 µL / well of 1x PBS. The plates were washed five times with 1x PBST. A seven-point 3-fold dilution series of BiXAb-6567 in 1x PBS (starting at 1 µg / mL) was prepared, and 100 µL of each dilution step was added to each well. The plates were incubated at room temperature for 1 hour, followed by five washes with 1x PBST. 100 µL / well of 1 µg / mL biotinylated human PD-L1 (AcroBiosystems) in 1x PBS was added, and the plates were incubated at room temperature for 1 hour. After washing five times with 1x PBST, add 100 µL / well of 0.1 µg / mL streptavidin-conjugated HRP (Biotechne) prepared by dilution with 1x PBS. Incubate the plates at room temperature for 1 hour. After washing five times with 1x PBST, add 100 xL / well of TMB substrate in 1x PBS for colorimetric reading, and incubate the plates at room temperature for 15 minutes for color development. Data were collected at 650 nm using a Victor2 microplate reader (Perkin Elmer).

[0314] BiXAb-6567 exhibited a dose-dependent binding curve in a dual ELISA format, indicating that it possesses correctly assembled anti-CD38 and anti-PD-L1 Fab domains. Figure 4C This indicates that BiXAb-6567 is a bispecific antibody capable of simultaneously binding to CD38 and PD-L1 at an EC50 of 144 ng / mL. As expected, neither of the two parental mAbs (anti-CD38 or anti-PD-L1) showed any binding in this dual ELISA format.

[0315] Example 4: Determination of relative binding activity by fluorescence-activated cell sorting (FACS)

[0316] CHO-CD38 cells (CHO cells stably transfected with full-length human CD38) were cultured in DMEM-Glutamax-I medium supplemented with 100 µg / ml penicillin, 100 µg / ml streptomycin, 10% fetal bovine serum, and 500 µg / ml genimycin. SKOV-3 cells and RPMI-8226 cells were cultured in RPMI 1640-Glutamax-I medium supplemented with 100 µg / ml penicillin, 100 µg / ml streptomycin, and 10% fetal bovine serum.

[0317] Three x 10⁵ cells (CHO-CD38, SKOV-3, or RPMI-8226) were used for each sample. Cells were washed once with PBA solution (PBS supplemented with 1% BSA and 0.05% sodium azide). To determine the FACS profile, cells were stained with each antibody at a concentration of 50 μg / ml in 30 μl volumes. To titrate BiXAb-6567 and parental anti-CD38 antibodies, and subsequently determine binding parameters, CHO-CD38 cells were stained with each antibody at indicated concentrations in 30 µl volumes. Cells were incubated on ice for 30 min and then washed twice with 1 ml of PBA solution. Cells were incubated on ice in the dark for 30 min with fluorescently labeled anti-human κ or anti-human IgG Fcγ-specific secondary antibodies and then washed twice with 1 ml of PBA solution; finally, the cells were resuspended in a final volume of 500 µl of PBA solution. Samples were analyzed using an Episics-XL or Navios flow cytometer (Beckman Coulter). 10,000 events are obtained in each experiment.

[0318] exist Figure 5A -C shows the binding profiles of BiXAb-6567 with parental anti-CD38 and anti-PD-L1 antibodies. We chose to test the multiple myeloma cell line RPMI-8226, which expresses high levels of CD38 and negligible levels of PD-L1 (…). Figure 5A CHO-CD38 cell lines that express very high levels of CD38 due to stable transfection with full-length CD38 (); Figure 5B ); and the ovarian cancer cell line SKOV-3, which is known to express PD-L1 ( Figure 5C These spectra show a single peak for BiXAb-6567, which is very similar to the spectra of the two parental antibodies on the three cell lines. This indicates that BiXAb-6567 folds correctly and has similar binding properties to the parental antibodies. As expected, CHO-CD38 expresses only CD38 and not PD-L1, while SKOV-3 expresses only PD-L1 and not CD38.

[0319] To quantitatively confirm that the binding properties of BiXAb-6567 are similar to those of the parental anti-CD38 antibody, titration of BiXAb-6567 and the parental anti-CD38 antibody was performed using CHO-CD38 cells. Figure 6The EC50 of BiXAb-6567 was determined to be 17.1 nM, while that of the parental anti-CD38 was 8.5 nM, confirming the similar binding properties of the anti-CD38 Fab domain in both BiXAb-6567 and the parental anti-CD38 antibody. As expected, the negative controls in this experiment (anti-PD-L1 and anti-CD20 antibodies) did not demonstrate binding to CHO-CD38 cells.

[0320] Example 5: Antibody-dependent cell-mediated cytotoxicity (ADCC) using ungraded, non-pre-activated monocytes (MNCs)

[0321] CHO-CD38, SKOV-3, and RPMI-8226 cells were cultured as described in Example 4 above.

[0322] To prepare MNCs, the following procedure was used. Freshly drawn peripheral blood was anticoagulated with citrate. Subsequently, 5 ml of Ficoll-Paque PLUS solution was separated from 6 ml of anticoagulated whole blood. The sample was centrifuged at 2500 rpm for 20 minutes at room temperature without centrifuge rupture. MNCs were collected from the plasma / Ficoll interface. The MNC cell suspension was diluted 1:10 in PBS and centrifuged at 1800 rpm for 5 minutes at room temperature. The supernatant was removed, and red blood cells were lysed by adding 45 ml of ice-cold distilled water to the cell suspension for 30 seconds, followed by the addition of 5 ml of 10x PBS. The cells were centrifuged at 1800 rpm for 5 minutes at room temperature and washed three times with 1x PBS to remove platelets. Finally, the cells were resuspended in 5 ml of cell culture medium. The cell count was adjusted to achieve a 40:1 effector cell:tumor cell ratio in the ADCC assay.

[0323] For ADCC 51 Chromium release assay, using 1x10 6 Target cells (RPMI 8226, SKOV-3, or CHO-CD38) were incubated with 100 μCi 51Cr in 200 μl PBS at 37°C and 5% CO2 for 2 hours. After 2 hours of incubation, the cells were washed three times with 7 ml of culture medium and finally resuspended at a concentration of 0.1 × 10⁶ cells / ml. Target cells (5,000 cells / well) and MNCs were incubated in 96-well microtiter plates (200 µl assay volume) at 37°C and 5% CO2 for 3 hours in the presence of antibody. Triton X-100 was added to determine the maximum target cell lysis (= maximum cpm). To determine the baseline... 51Chromium release (=basal cpm) was performed without further treatment of the target cells. After 4 hours of incubation, the microtiter plate was centrifuged at 2000 rpm for 5 minutes, and 25 µl of the supernatant was mixed with 125 µl of Optiphase Supermix (Perkin Elmer) and incubated in a shaking incubator for 1 minute. The samples were measured using a MicroBeta TriLux (Perkin Elmer) β-counter instrument. Target cell lysis was calculated using the following formula:

[0324] Fracturing percentage = (experimental CPM – base CPM) / (maximum CPM – base CPM) x 100.

[0325] All measurements were performed in triplicate.

[0326] ADCC assay of CD38+ cells (RPMI-8226 and CHO-CD38) was performed using non-pre-activated MNCs as effector cells. Figure 7 and 8 The assays showed that BiXAb-6567 and the anti-CD38 antibody exhibited potent cytotoxicity against RPMI-8226 cells with EC50 values ​​of 0.8 nM and 0.3 nM, respectively; on CHO-CD38 cells, BiXAb-6567 and the anti-CD38 antibody showed cytotoxicity with EC50 values ​​of 0.2 nM and 0.07 nM, respectively. Anti-PD-L1 activity was minimal in both cell lines; as expected, the two negative control mAbs (anti-CD20 and anti-HER2) did not promote any lysis. These results demonstrate the potent ADCC activity of BMX-6567 against CD38+ cells, similar to the activity of the parental anti-CD38 antibody.

[0327] Example 6: ADCC using enriched pre-activated NK cells

[0328] SKOV3 cells, RPMI 8226, and CHO-CD38 cells were cultured as described in Example 4. MNCs were prepared as described in Example 5. NK cells were isolated from the MNCs using the “NK Cell Isolation Kit, Human” (Miltenyi) according to the manufacturer’s instructions, through negative selection. Cells were cultured in RPMI medium supplemented with 10% fetal bovine serum at a concentration of 2 x 10⁻⁶ cells / mL. 6 NK cells were cultured overnight at a seeding density of 10 cells / ml. IL-12 or IL-15 was added to a final concentration of 10 ng / ml. ADCC assays were performed as outlined in Example 5, except that the effector cell:tumor cell ratio was maintained at 10:1 and the reaction duration was reduced to 3 hours.

[0329] The ADCC properties of the anti-PD-L1 moiety of BiXAb-6567 were determined using enriched NK cells pre-activated with IL-12 or IL-15 on the PD-L1+ cell line SKOV-3. Results are presented in... Figure 9 and 10 In this experiment, the ADCC properties of BiXAb-6567 were compared with those of the parental anti-PD-L1 antibody. Since SKOV-3 cells are PD-L1+ / HER2+ / CD20- / CD38-, an anti-HER2 antibody was used as a positive control, and an anti-CD20 antibody and a parental anti-CD38 antibody were used as negative controls. Figure 9 and 10 The potent ADCC activity of BiXAb-6567 and the parental anti-PD-L1 antibody was demonstrated, independent of the use of IL-12 or IL-15 in cultured NK cells. When IL-12 was used, the EC50 values ​​for BiXAb-6567 and the parental anti-PD-L1 antibody were 0.007 nM and 0.03 nM, respectively. The situation was even more similar when IL-15 was used; however, the curve fitting did not converge, thus hindering the calculation of EC50 values. These results demonstrate the potent ADCC activity of BMX-6567 against PD-L1+ cells, which is similar to the ADCC activity of the parental anti-PD-L1 antibody.

Claims

1. A protein construct comprising a mutated Fab fragment, said mutated Fab fragment being a Fab fragment composed of: - The VH and VL domains of the target antibody; -CH1 domain, which is derived from the CH1 domain by replacing the leucine residue at position 124 of the CH1 domain of the immunoglobulin with glutamine and replacing the serine residue at position 188 of the CH1 domain with a valine residue; and -CL domain, which is derived from the CL domain by replacing the valine residue at position 133 of the CL domain of the immunoglobulin with a threonine residue and replacing the serine residue at position 176 of the CL domain with a valine residue. The sequence position number used here for the CH1 and CL structural domains refers to the Kabat number.

2. The protein construct of claim 1, comprising or composed of a multispecific antigen-binding fragment, said multispecific antigen-binding fragment comprising at least two Fab fragments having different CH1 and CL domains, wherein each Fab fragment recognizes a different target epitope, and said Fab fragments are arranged in tandem in any order, wherein the C-terminus of the CH1 domain of the first Fab fragment is connected to the N-terminus of the VH domain of a subsequent Fab fragment via a peptide linker. At least one of the Fab fragments consists of a mutated Fab fragment as defined in claim 1.

3. The protein construct of claim 2, which is a multispecific antibody having two identical antigen-binding arms, each arm consisting of a multispecific antigen-binding fragment as defined in claim 2.

4. The protein construct of claim 3, having an immunoglobulin-like structure, comprising: - Two identical antigen-binding arms, each arm consisting of a multispecific antigen-binding fragment as defined in claim 1; -The CH2 and CH3 domains of immunoglobulin dimerization; - A hinge region of IgA, IgG, or IgD that connects the C-terminus of the CH1 domain of the antigen-binding arm to the N-terminus of the CH2 domain.

5. The protein construct of claim 4, preferably a bispecific antibody, and comprising at least two heavy chains and four light chains. Each heavy chain contains a. The Fc region of an immunoglobulin containing a hinge-CH2-CH3 domain. b. The Fc region is connected to the Fab heavy chain CH1-VH of antibody 1 (Ab1) via the hinge domain. c. It then links to the CH1-VH domain of the Fab heavy chain of antibody 2 (Ab2) via a peptide linker sequence, and the peptide linker sequence connects the N-terminus of the VH domain of the Fab heavy chain of Ab1 to the C-terminus of the CH1 domain of Ab2. Furthermore, the four light chains contain the Fab light chain CL-VL of Ab1 and the Fab light chain CL-VL of Ab2, which are related to their cognate heavy chain structural domains. Ab1 and Ab2 identify different epitopes. The Fab CH1 domain of either Ab1 or Ab2 is a mutated domain derived from the CH1 domain by replacing the leucine residue at position 124 with glutamine and the serine residue at position 188 with valine. The homologous CL domain is a mutated domain derived from the CL domain by replacing the valine residue at position 133 with threonine and the serine residue at position 176 with valine. Furthermore, the sequence position number used for the CH1 and CL structural domains here refers to the Kabat number.

6. The protein construct of claim 5, wherein the Fab CH1 domain of Ab2 is a mutated domain derived from the CH1 domain by replacing a leucine residue at position 124 of the immunoglobulin with glutamine and replacing a serine residue at position 188 of the CH1 domain with a valine residue, and the homologous CL domain is a mutated domain derived from the CL domain by replacing a valine residue at position 133 of the immunoglobulin with a threonine residue and replacing a serine residue at position 176 of the CL domain with a valine residue.

7. The protein construct of claim 5, wherein the Fab CH1 domain of Ab1 is a mutated domain derived from the CH1 domain by replacing a leucine residue at position 124 of the immunoglobulin with glutamine and replacing a serine residue at position 188 of the CH1 domain with a valine residue, and the homologous CL domain is a mutated domain derived from the CL domain by replacing a valine residue at position 133 of the immunoglobulin with a threonine residue and replacing a serine residue at position 176 of the CL domain with a valine residue.

8. The protein construct of any one of claims 2 to 7, wherein the polypeptide linker sequence comprises or is composed of the following amino acid sequence: EPKX1CDKX2HX3X4PPX5PAPELLGGPX6X7PPX8PX9PX 10 GG (SEQ IDNO:1), where X1, X2, X 3、 X4, X5, X6, X7, X8, X9, X 10 It can be any amino acid, whether it is the same or different.

9. The protein construct of claim 8, wherein the polypeptide linker sequence comprises or consists of sequences selected from the following: EPKSCDKTHTSPPAPAPELLGGPGGPPGPGPGGG (SEQ ID NO: 2); EPKSCDKTHTSPPAPAPELLGGPAAPPAPAPAGG (SEQ ID NO: 3); EPKSCDKTHTSPPAPAPELLGGPAAPPGPAPGGG (SEQ ID NO: 4); EPKSCDKTHTCPPCPAPELLGGPSTPPTPSPSGG (SEQ ID NO:5) and EPKSCDKTHTSPPSPAPELLGGPSTPPTPSPSGG (SEQ ID NO:6).

10. The protein construct of any one of claims 1 to 9, which recognizes both EGFR and HER2 / neu.

11. The protein construct of claim 10, which is a multispecific antibody, preferably a bispecific antibody, wherein Ab1 and Ab2 are different, each being independently selected on one hand from cetuximab or a mutant derivative thereof, or on the other hand from trastuzumab or a mutant derivative thereof.

12. The protein construct of any one of claims 1 to 9, which recognizes both CD38 and PD-L1.

13. The protein construct of claim 12, which is a multispecific antibody, preferably a bispecific antibody, wherein Ab1 and Ab2 are different, each independently selected on one hand from daratumumab or a mutant derivative thereof, and on the other hand from atezolizumab or a mutant derivative thereof.

14. The protein construct of claim 12, which is a bispecific antibody, comprising, preferably, the following: a) two heavy chains, each containing SEQ ID NO:7, preferably consisting of SEQ ID NO:7; and b) four light chains, two of which contain SEQ ID NO:15, preferably consisting of SEQ ID NO:15, and the other two of which contain SEQ ID NO:18, preferably consisting of SEQ ID NO:

18.

15. A polypeptide comprising a heavy chain of an antigen-binding fragment or a heavy chain of a multispecific antibody as defined in any one of claims 2 to 14, preferably consisting of a heavy chain of an antigen-binding fragment or a heavy chain of a multispecific antibody as defined in any one of claims 2 to 14.

16. A polynucleotide comprising a sequence encoding the polypeptide of claim 15.

17. Host cells transfected with an expression vector comprising the polynucleotide of claim 16.

18. The host cell of claim 17, further comprising conversion with at least two polynucleotides encoding two different light chains: a first light chain specifically paired with a first VH / CH1 region of the heavy chain; and a second light chain specifically paired with a second VH / CH1 region of the heavy chain.

19. A method for generating an antigen-binding fragment or a multispecific antibody as defined in any one of claims 2 to 14, the method comprising the steps of: a) Culture host cells expressing the antibody heavy chain as defined in any one of claims 2 to 14 and the antibody light chain as defined in any one of claims 2 to 14 in suitable culture medium and culture conditions; and b) Recover the generated antibody from the culture medium or the cultured cells.

20. The multispecific antigen-binding fragment or multispecific antibody as defined in any one of claims 2 to 14, used as a medicament.

Images