scFvs and antibodies with reduced multimerization

Abstract

Methods are disclosed for generating variants of scFv that have a reduced tendency to form multimers.

Description

[Technical field]

[0001] The present invention relates to variants of scFvs and antibodies that have a reduced tendency to form multimers. In particular, the present invention relates to bispecific antibodies comprising said scFvs.

[0002] The present invention further relates to pharmaceutical compositions comprising one or more antibodies of the invention and their use for the treatment of cancer. [Background technology]

[0003] scFv fragments and antibodies, such as bispecific antibodies, have found wide use in antibody-assisted diagnostics and therapy.

[0004] Pretargeted radioimmunotherapy (PRIT) is an example of a pharmaceutical method that exploits the efficient binding of antibodies to specific targets, allowing treatment to be concentrated at the targeted site.

[0005] WO 2018 / 204873 discloses SADA technology, which benefits from the SADA domain, which has the ability to assemble or disassemble in a concentration-dependent manner. This property is particularly beneficial in the context of PRIT. Bispecific antibodies capable of binding to cytotoxic agents and targeting sites and connected to SADA domains can be administered in a multimeric form, particularly a tetrameric form, and can bind to the antibody targeting site, while unbound molecules are degraded and removed from the plasma stream via the kidney before the cytotoxic agent is administered. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] International Publication No. 2018 / 204873 Summary of the Invention [Problem to be solved by the invention]

[0007] It is important for SADA-PRIT therapy that the molecular size of the multimeric forms, especially the tetrameric form, is above the renal clearance limit, thereby providing a long plasma half-life, and that the monomeric form is below the renal clearance limit, providing a short half-life.

[0008] In pharmaceutical compositions comprising one or more antibodies, it is also important that the one or more antibodies remain stable during the shelf-life of the pharmaceutical composition, avoiding degradation of the antibodies, as well as avoiding unintended agglomeration or cross-reaction between the antibodies and / or between the antibodies and other components of the composition. [Means for solving the problem]

[0009] Summary of the Invention In a first aspect, the present invention provides a method for generating a variant of an scFv domain comprising a light chain variable domain (VL), a heavy chain variable domain (VH) and one or more disulfide bonds between the VL and VH, comprising: a. identifying cysteine ​​residues that form the one or more disulfide bonds between VL and VH; and b. replacing one or more disulfide bond forming cysteine ​​residues identified in step a with an amino acid other than cysteine. The present invention relates to a method comprising the steps of:

[0010] The mutated scFvs of the present invention are themselves scFvs and have a reduced ability and / or tendency to form multimers compared to scFvs having a disulfide bond between the VH and VL domains.

[0011] In another embodiment, the present invention relates to an scFv domain prepared by the method of the present invention.

[0012] In a further embodiment, the present invention relates to a bispecific antibody comprising a first scFv domain capable of binding to a DOTA or DOTAM metal chelate, a second scFv domain capable of binding to a tumor antigen and a SADA domain, wherein the first scFv domain and / or the second scFv domain does not contain a disulfide bond between the VH and VL domain.

[0013] In a further aspect, the present invention relates to compositions, in particular pharmaceutical compositions, comprising an scFv or a bispecific antibody of the invention and to the use of such compositions for diagnosing or treating cancer.

[0014] In a further aspect, the present invention relates to a kit comprising an scFv or a bispecific antibody of the invention.

[0015] The present invention also relates to polynucleotides, expression vectors or constructs comprising such polynucleotides, host cells comprising such polynucleotides or expression vectors or constructs, and the use of such host cells for the preparation of scFvs or bispecific antibodies of the invention.

[0016] Further aspects are provided in the claims.

[0017] [Definition] DOTA: DOTA (dodecanetetraacetic acid) is also known as 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid. 16 H 28 It has the formula (CH2CH2NCH2CO2H)4, also known as N4O8·xH2O.

[0018] DOTA metal chelate: refers to DOTA having a complex bound to a metal ion.

[0019] Derivatives of DOTA: is intended to mean compounds that contain a DOTA ring system and are capable of chelating metal ions. Examples of such compounds include benzyl DOTA and the bispecific chelators disclosed in WO 2019 / 010299. Additional DOTA derivatives are disclosed in WO 2010 / 099536.

[0020] DOTAM: A chelating agent that contains a ring system capable of binding to metal ions. It has the systematic name 1,4,7,10-tetraazacyclododecane-1,7-bis(acetate)-4,10-bis(acetamide) and the formula C 16 H 30 It has N6O6·2H2O.

[0021] Amino acid substitution: intended to mean the replacement of one amino acid with a different amino acid. As used herein, the term amino acid substitution with reference to a reference sequence is intended to mean that the amino acid sequence can be generated by starting from a reference sequence and introducing said amino acid substitution, even if the amino acid sequence was generated by another process that does not involve the reference sequence.

[0022] Sequence identity: The term sequence identity is intended to mean a measure of the relatedness of two nucleic acid or amino acid sequences. Sequence identity is determined by aligning two sequences, finding the longest overlap, counting the number of matches in the overlap, and dividing the number of matches by the number of nucleotide or amino acid residues in the overlap to calculate the sequence identity. Sequence identity is typically expressed as a percentage (%).

[0023] A variety of computational algorithms are available to those skilled in the art for generating sequence alignments and calculating sequence identity. As used herein, sequence alignment refers to pairwise alignment. Several algorithms perform this, including the sequence alignment program Clustal Omega [doi:10.1038 / msb.2011.75].

[0024] As used herein, sequence alignment is performed using the following algorithm: Algorithm: Clustal Omega (1.2.4), (http: / / www.clustal.org / omega / ).

[0025] Antibody or antibody fragment: An antibody fragment is a portion of an antibody, such as F(ab')2, F(ab)2, Fab', Fab, Fv, scFv, etc. Regardless of structure, an antibody fragment binds to the same antigen recognized by the intact antibody. For example, the 3F8 monoclonal antibody fragment binds to the epitope recognized by 3F8. The term "antibody fragment" also includes any synthetic or genetically engineered protein that acts like an antibody by binding to a specific antigen to form a complex. For example, antibody fragments include isolated fragments consisting of variable regions, such as "Fv" fragments consisting of heavy and light chain variable regions, recombinant single chain polypeptide molecules in which light and heavy variable regions are connected by a peptide linker ("scFv protein"), and minimal recognition units consisting of amino acid residues that mimic the hypervariable region.

[0026] scFv domain: A single chain polypeptide consisting of the variable regions of the light (VL) and heavy (VH) antibody chains, usually separated by a linker sequence. The order of the VL and VH regions can vary, and it is not uncommon that an scFv with a VL-VH order can be changed to an scFv with a VH-VL order without significant change in specificity. The linker sequence separating the VL and VH typically comprises or consists of small hydrophobic amino acid residues, e.g., G, S and T, and can have a size of 5 to 50 amino acids. One preferred linker consists of 4 glycine residues and 1 serine residue or a repeat of such a sequence, e.g., 2, 3, 4 or 5 repeats of this sequence.

[0027] Substitution: is understood in the usual way to replace one amino acid residue in a polypeptide with a different amino acid residue. Substitutions are described in the present specification and claims in the format [original amino acid][position][new amino acid], and one-letter code is used. When more than one amino acid is a possible substitution at a given position, the possible substitutions are separated by a comma. For example, the substitution of a glycine residue at position 9 of a polypeptide with an alanine residue is indicated as G9A, and the substitution of a glycine residue at position 9 with either alanine or valine is indicated as G9A,V.

[0028] Domain: is intended to mean a polypeptide or a portion of a polypeptide that has its own structure and function defined by its amino acid sequence. A polypeptide can be a single domain polypeptide, where a single domain forms the entire polypeptide, or a polypeptide can be a multi-domain polypeptide, where the domains are arranged as separate parts of the sequence. In multi-domain polypeptides, linker sequences are often provided between the domains to ensure distance between the domains so that each domain can fold and perform its function without steric hindrance from the other parts (domains) of the polypeptide.

[0029] CDR: Complementarity determining regions (CDRs) are parts of the variable region of an antibody and are of major importance for the binding specificity of the antibody. A typical antibody, consisting of two heavy and two light chains, has six CDR sequences, three in the light chain variable domain (VL) and three in the heavy chain variable domain (VH).

[0030] Pharmaceutical composition: As used herein, the term "pharmaceutical composition" is intended to mean a composition for administering a drug or medicine to a patient in need thereof. Pharmaceutical compositions are prepared from pharmaceutical grade ingredients using methods and techniques known in the pharmaceutical or pharmacological arts, for example as described in the European Pharmacopoeia, 10th Edition. [Brief description of the drawings]

[0031] [Figure 1] Figure 2 shows the SE-HPLC chromatogram of the GD2-SADA construct. For further details, see Example 1. [Diagram 2] Figure 2 shows a SE-HPLC chromatogram of a truncated GD2-SADA construct lacking the SADA domain. For further details, see Example 1. [Diagram 3] Figure 1 shows a Coomassie stained SDS-PAGE gel under non-reducing conditions of the GD2-SADA construct and a truncated version lacking the SADA domain. The latter was also analyzed under reducing conditions. For further details, see Example 2. [Figure 4] 1 shows a Coomassie stained SDS-PAGE gel of variants of the CD20-SADA construct. For further details, see Example 3. [Diagram 5] 1 shows a Coomassie stained SDS-PAGE gel of variants of the CD38-SADA construct. For further details, see Example 4. [Figure 6] Figure 1 shows a Coomassie stained SDS-PAGE gel of variant YMS9d under non-reducing and reducing conditions. For further details see Example 4. [Figure 7] The dilution regimen and the SE-HPLC chromatograms of the diluted samples are shown. For further details, see Example 6. [Figure 8] 1 shows a Coomassie stained SDS-PAGE gel of RSV-SADA constructs with and without disulfide bonds between the VH and VL sequences. For further details, see Example 10. [Figure 9] 1 shows a Coomassie stained SDS-PAGE gel of B7H3-SADA constructs with and without a disulfide bond between the VH and VL sequences. For further details, see Example 11. [Figure 10A] 1 shows a Coomassie stained SDS-PAGE gel of trastuzumab-based HER2-SADA constructs with and without a disulfide bond between the VH and VL sequences. For further details, see Example 12. [Figure 10B] Figure 1 shows a Coomassie stained SDS-PAGE gel of Pertuzumab-based HER2-SADA constructs with and without a disulfide bond between the VH and VL sequences. For further details, see Example 12. [Figure 11] 1 shows a Coomassie stained SDS-PAGE gel of the CD20-DOTAM-SADA construct. For further details, see Example 13. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0032] Some embodiments of the invention are set forth in the claims.

[0033] In one embodiment, the present invention provides a method for generating a variant of an scFv domain comprising a light chain variable domain (VL), a heavy chain variable domain (VH) and one or more disulfide bonds between the VL and VH, comprising: a. identifying cysteine ​​residues that form the one or more disulfide bonds between VL and VH; and b. replacing one or more disulfide bond forming cysteine ​​residues identified in step a with an amino acid other than cysteine. The present invention relates to a method comprising the steps of:

[0034] According to the invention, one or all of the disulfide bonds identified in step 1 can be eliminated by substituting the cysteines forming said bonds with other amino acids. It is preferred to substitute both cysteines forming the disulfide bond with other amino acids to avoid any free cysteines.

[0035] The present invention is based on the observation that many scFv domains and constructs containing scFv domains can form multimers, e.g. dimers or trimers, or multiple forms of monomeric forms. This is undesirable for compounds intended for pharmaceutical use, where high compound homogeneity and purity are generally desirable. Furthermore, the heterogeneity of scFv domains and constructs containing scFv domains complicates recovery and purification compared to analogous compounds with high homogeneity.

[0036] The inventors understand that the disulfide bond between the VH and VL of an scFv is responsible for multimerization and the formation of alternative disulfide bonds that result in the observed formation of multimers and multimeric forms of scFvs, and that it can be demonstrated using the methods of the invention that scFvs that do not have a disulfide bond between the VH and VL domains are less prone to forming multimers or multimeric forms.

[0037] The resulting scFv variants have similar properties to the scFv derived by the methods of the invention.Those skilled in the art will further appreciate that variants derived from scFvs by the invention are in fact also scFvs themselves.

[0038] In natural antibodies, the VL and VH sequences are part of light and heavy immunoglobulin chains, and in nature, the light and heavy chains are connected by one or more disulfide bonds found in the constant regions adjacent to the VL and VH sequences. However, since scFvs only consist of VL and VH sequences connected by a linker, the disulfide bonds found in the constant regions adjacent to the VL and VH sequences and connecting the chains containing the VL and VH chains in natural antibodies are not present in scFvs, and therefore it is common practice to introduce disulfide bonds between the VL and VH sequences of scFvs in order to improve the stability of scFvs. The present invention is based on the understanding of the inventors that such stabilizing disulfide bonds introduced between the VH and VL domains of scFvs may result in heterogeneity, which may be disadvantageous for at least some uses of scFvs.

[0039] It is known in the art that VH and VL domains may contain additional disulfide bonds between two cysteine ​​residues within the same domain (intracene disulfide bonds), and the inventors further understand that these intra-domain disulfide bonds, in contrast to inter-domain disulfide bonds (between the VL and VH domains), are not important, or at least less important, for the observed heterogeneity.

[0040] In one embodiment, the invention relates to a method for generating variants of an scFv domain, said variants resulting in a less multimeric form compared to the original scFv domain.

[0041] In another embodiment, the invention relates to the use of an scFv that does not have a disulfide bond between the VH and VL domains in a polypeptide construct comprising the scFv and an additional domain, wherein the polypeptide construct has a lower propensity for multimerization or at least a lower propensity to form multimers compared to a similar polypeptide that has a disulfide bond between the VH and VL domains in the scFv that has the same sequence except for the additional disulfide bond.

[0042] Multimeric forms can be detected using techniques known in the art for determining molecular weight, such as chromatographic methods.

[0043] In one embodiment of the invention, the multimeric form is determined by SDS-PAGE gel electrophoresis.

[0044] In one embodiment of the invention, the scFv domain is part of a polypeptide that comprises additional antibody fragments, for example, the scFv can be part of a polypeptide that, in addition to the scFv domain, comprises one or more of the following: additional scFv domains, immunoglobulin heavy and / or light chains, an Fc domain, a hinge region, etc.

[0045] In one preferred embodiment, the scFv domain is part of a bi- or multispecific antibody. One form of such a bi- or multispecific antibody is a bispecific antibody comprising two chains of a fusion polypeptide comprising two heavy immunoglobulin chains and an immunoglobulin light chain fused C-terminally to an scFv domain, where a first binding specificity is provided by the variable regions of the heavy and light immunoglobulin chains and a second binding specificity is provided by the scFv domain.

[0046] Another form of such a bi- or multispecific antibody is a polypeptide comprising two or more scFv domains, each providing a binding specificity.

[0047] In one embodiment, the present invention relates to a bi- or multispecific antibody further comprising a SADA domain, also known as a tetramerization domain.

[0048] SADA (self assembly and disassembly) domains are short amino acid domains that can spontaneously assemble and disassemble in solution depending on the concentration. Complexes containing SADA domains typically exist in at least two unique forms, a tetrameric form at high concentrations and a monomeric form at low concentrations. Self assembly and disassembly (SADA) technology is disclosed in WO 2018 / 204873, which is incorporated by reference in its entirety.

[0049] SADA conjugates may be designed such that the tetrameric form has a molecular weight well above the renal clearance limit and the monomeric form has a molecular weight below the renal clearance limit, meaning that the tetrameric form has a high plasma half-life since it is not excreted in the urine and the monomeric form has a low plasma half-life since it is excreted in the urine.

[0050] Preferred SADA domains for use in the present invention include domains comprising the sequence of SEQ ID NO:5, SEQ ID NO:6, SEQ ID NO:7, SEQ ID NO:8, SEQ ID NO:9, SEQ ID NO:10, SEQ ID NO:11 or SEQ ID NO:12, or a sequence having at least 80%, at least 85%, at least 90%, at least 95%, at least 97% sequence identity to one of these sequences.

[0051] Preferred SADA domains for use in the present invention include: a. sequence SEQ ID NO: 5 or a sequence differing from this sequence by 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10 substitutions; b. SEQ ID NO: 6 or a sequence differing from this sequence by 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10 substitutions; c. SEQ ID NO: 7 or a sequence differing from this sequence by 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10 substitutions; d. SEQ ID NO: 8 or a sequence differing from this sequence by 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10 substitutions; e. SEQ ID NO: 9 or a sequence differing from this sequence by 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10 substitutions; f. SEQ ID NO: 10 or a sequence differing from this sequence by 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10 substitutions; g. SEQ ID NO: 11 or a sequence differing from this sequence by 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10 substitutions; or h. Sequence SEQ ID NO: 12 or a sequence differing from this sequence by 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10 substitutions. Examples of such domains include:

[0052] A preferred SADA domain of the present invention is a domain comprising a sequence having at least 80% sequence identity to amino acids 6 to 36 of SEQ ID NO:5 and differing from the sequence of SEQ ID NO:5 by one or more substitutions, which domain retains the ability to dimerize or tetramerize.

[0053] Those skilled in the art can easily determine whether such domains with a given substitution retain the ability to dimerize or tetramerize by simple routine experimentation, or can find such information in the literature, for example, J. Gencel-Augusto and G- Lozano; Genes & Development 34:1128-1146 (incorporated by reference).

[0054] A preferred SADA domain of the invention has the sequence of amino acids 6 to 36 of SEQ ID NO:5 and, using the numbering of SEQ ID NO:5, the following: E6V, Q, K, G, D or A; Y7S, N, H, F, D or C; F8Y, V, S, L, I or C; T9S, P, N or A; L10V, I or F; Q11R, L, K, H or E; I12V, T, M, L or F; R13S, P, L, H, G or C; G14W, R or A; R15S, P, L, H, G or C; E16V, Q, K, G, D or A; F18Y, V, S, L, I or C; E19V, Q, K, G, D or A; M20V, T, R, L, K or I; F21L or I; R22L or G; E23V, Q, K, G, D or A; L24M; N25S, I or D; E26V, Q, K, G, D or A; A27V, T, S, G or D; L28W, V, M or F; E29Q, G or D; L30V, R, I, H or F; K31T, R, Q, N, M or E; D32Y, V, N, H, G or A; A33V, T, S, P, G or D; Q34R, L, K, H or E; The domains have amino acid sequences which differ by one, two, three, four or five substitutions selected from the following substitutions:

[0055] The L24P substitution should not be applied as it completely abolishes tetramerization.

[0056] The p53 tetramerization domain comprising the sequence of amino acids 6 to 36 of SEQ ID NO:5 is a preferred SADA domain.

[0057] The present invention is particularly useful in the context of SADA technology using constructs comprising one or more scFvs and SADA domains, constructed such that, for example, a tetrameric complex comprising four monomers has a size above the renal clearance limit, while the monomers of the complex have a size below the renal clearance limit. Typically, such constructs comprising one or more scFvs and SADA domains have a molecular weight of about 50-60 kD in monomeric form. After the complex is administered in tetrameric form, it binds to the target and the unbound complex degrades in plasma and is rapidly excreted because its size is below the renal clearance limit. However, if part of the polypeptide forms a multimer, mediated by disulfide bonds between the VH and VL of the scFv, theoretically and depending on the actual construct, clearance of the degraded complex may be less efficient due to the multimers formed.

[0058] Therefore, using the present invention in conjunction with SADA technology to reduce multimer formation is particularly beneficial.

[0059] In one embodiment, the invention relates to an scFv domain comprising a VL and a VH, capable of binding to an antigen and obtainable by a method according to any of the preceding claims. Preferably, the VH and VL are not connected by any disulfide bond.

[0060] In one embodiment, the scFv domain of the invention further comprises a linker between VH and VL. Linkers, sometimes also known as spacers, are short amino acid sequences created to separate multiple domains in a single protein. Linkers are known in the art and the invention is not limited to any particular sequence of a linker. In general, the purpose of a linker is to connect and / or separate different elements of a complex and is typically composed primarily of small hydrophilic amino acids, such as glycine, serine and threonine.

[0061] In one embodiment, the present invention is capable of binding to a DOTA metal chelate, 6 CDR sequences consisting of sequences of SEQ ID NOs: 44 to 49 or sequences which differ from the sequences of SEQ ID NOs: 44 to 49 by one or two substitutions; a VL sequence comprising the sequence of SEQ ID NO:1 or a sequence having at least 90%, 95%, 96%, 97%, 98% or 99% sequence identity to SEQ ID NO:1; and A VH sequence comprising the sequence of SEQ ID NO:2 or a sequence having at least 90%, 95%, 96%, 97%, 98% or 99% sequence identity to SEQ ID NO:2. An antibody or antibody fragment comprising: The amino acid corresponding to position 111 of SEQ ID NO:1 and position 45 of SEQ ID NO:2 is not a cysteine ​​for the antibody or antibody fragment.

[0062] Antibodies and antibody fragments capable of binding to DOTA can be derived from the affinity matured murine antibody 2D12.5 to increase the affinity for DOTA (WO 2010 / 099536). When scFvs were generated based on 2D12.5, stabilizing disulfide bonds were generated by inserting cysteines at positions corresponding to positions 111 and 179 of SEQ ID NO: 3. This disulfide stabilized scFv was affinity matured several times to generate several modified antibodies with improved affinity for DOTA-metal, including the scFv called C825 (SEQ ID NO: 3). The inserted disulfide bonds appear to have been considered essential, since WO 2010 / 099526 explains that all variants that emerged in the last affinity maturation were discarded because they had lost the stabilizing disulfide bonds.

[0063] C825 is subsequently frequently used and humanized to obtain an antibody with superior DOTA-metal binding properties, likely to produce fewer adverse reactions when administered to humans, and which maintains the disulfide bond located between positions 111 and 179 of SEQ ID NO:3 of the corresponding mC825 scFv unchanged.

[0064] The inventors have surprisingly discovered that the disulfide bond between positions 111 and 179 is not essential to obtain a functional antibody, and in fact it is beneficial to remove this disulfide bond, including the cysteines corresponding to positions 111 and 179 of mC825 scFv (SEQ ID NO: 3).

[0065] The fact that the scFv can bind to the DOTA metal chelate indicates that the scFv is approximately 10 -4 M or less, e.g. 10 -4 M~10 -12 Range of M, e.g. 10 -5 M~10 -10 Range of M, e.g. 10 -6 M~10 -9 DOTA metal chelates with binding constants Kd in the range of M 175 It is intended to mean that it is capable of specifically binding to DOTA that is bound to Lu.

[0066] In one embodiment, the scFv of the invention comprises the VL and VH domains of SEQ ID NOs:1 and 2.

[0067] In one embodiment, the scFv comprises or consists of the sequence of SEQ ID NO:4, or comprises or consists of a sequence having at least 90% sequence identity to SEQ ID NO:4, such as at least 95% sequence identity, for example at least 96% sequence identity, such as at least 97% sequence identity, for example at least 98% sequence identity or at least 99% sequence identity.

[0068] In one embodiment, the scFv domain of the invention is capable of binding to GD2, 6 CDR sequences consisting of sequences of SEQ ID NOs: 13 to 18 or sequences differing from SEQ ID NOs: 13 to 18 by one or two substitutions; a VL sequence comprising the sequence of SEQ ID NO:19 or a sequence having at least 90%, 95%, 96%, 97%, 98% or 99% sequence identity to SEQ ID NO:19; and VH sequence comprising the sequence of SEQ ID NO:20 or a sequence having at least 90%, 95%, 96%, 97%, 98% or 99% sequence identity to SEQ ID NO:20 Including, The amino acid corresponding to position 97 of SEQ ID NO:19 and position 44 of SEQ ID NO:20 is not a cysteine.

[0069] The fact that scFv can bind to GD2 indicates that the scFv is approximately -4 M or less, e.g. 10 -4 M~10 -12 Range of M, e.g. 10 -5 M~10 -10 Range of M, e.g. 10 -6 M~10 -9 It is intended to mean that it is capable of specifically binding to GD2 with a binding constant Kd in the range of M.

[0070] In one embodiment, the scFv domain of the invention comprises or consists of the sequence of SEQ ID NO:21, or comprises or consists of a sequence having at least 90% sequence identity to SEQ ID NO:21, such as at least 95% sequence identity, such as at least 96% sequence identity, for example at least 97% sequence identity, such as at least 98% sequence identity or at least 99% sequence identity.

[0071] In one embodiment, the scFv domain of the invention is capable of binding to CD38, 6 CDR sequences consisting of sequences of SEQ ID NOs: 22 to 27 or sequences differing from SEQ ID NOs: 22 to 27 by one or two substitutions; a VL sequence comprising the sequence of SEQ ID NO:28 or a sequence having at least 90%, 95%, 96%, 97%, 98% or 99% sequence identity to SEQ ID NO:28; and VH sequence comprising the sequence of SEQ ID NO:29 or a sequence having at least 90%, 95%, 96%, 97%, 98% or 99% sequence identity to SEQ ID NO:29 Including, The amino acid corresponding to position 100 of SEQ ID NO:28 and position 44 of SEQ ID NO:29 is not cysteine.

[0072] The fact that scFv can bind to CD38 indicates that the scFv is approximately -4 M or less, e.g. 10 -4 M~10 -12 Range of M, e.g. 10 -5 M~10 -10 Range of M, e.g. 10 -6 M~10 -9 It is intended to mean that the antibody is capable of specifically binding to CD38 with a binding constant Kd in the range of M.

[0073] In one embodiment, the CDR sequences consist of SEQ ID NOs:22-27.

[0074] In one embodiment, the scFv of the invention comprises or consists of the sequence of SEQ ID NO: 30, or comprises or consists of a sequence having at least 90% sequence identity to SEQ ID NO: 30, such as at least 95% sequence identity, such as at least 96% sequence identity, for example at least 97% sequence identity, such as at least 98% sequence identity or at least 99% sequence identity.

[0075] In one embodiment, the scFv domain of the invention is capable of binding to CD20, 6 CDR sequences consisting of sequences of SEQ ID NOs: 31 to 36 or sequences which differ from the sequences of SEQ ID NOs: 31 to 36 by one or two substitutions; a VL sequence comprising a sequence of SEQ ID NO: 37 or a sequence having at least 90%, 95%, 96%, 97%, 98% or 99% sequence identity to SEQ ID NO: 37; and VH sequence comprising the sequence of SEQ ID NO: 38 or a sequence having at least 90%, 95%, 96%, 97%, 98% or 99% sequence identity to SEQ ID NO: 38. Including, The amino acid corresponding to position 99 of SEQ ID NO:37 and position 44 of SEQ ID NO:38 is not cysteine.

[0076] The fact that scFv can bind to CD20 indicates that the scFv is approximately-4 M or less, e.g. 10 -4 M~10 -12 Range of M, e.g. 10 -5 M~10 -10 Range of M, e.g. 10 -6 M~10 -9 It is intended to mean that the antibody is capable of specifically binding to CD20 with a binding constant Kd in the range of M.

[0077] Preferably, the CDR sequences consist of SEQ ID NOs:31-36.

[0078] In one embodiment, the scFv of the invention comprises or consists of the sequence of SEQ ID NO: 39, or comprises or consists of a sequence having at least 90% sequence identity to SEQ ID NO: 39, such as at least 95% sequence identity, such as at least 96% sequence identity, for example at least 97% sequence identity, such as at least 98% sequence identity or at least 99% sequence identity.

[0079] In one embodiment, the scFv domain of the invention is capable of binding to GPA33, six CDR sequences each consisting of the sequences of SEQ ID NOs: 50 to 55 or of sequences which differ from the sequences of SEQ ID NOs: 50 to 55 by one or two substitutions; a VL sequence having the sequence SEQ ID NO: 56 or a sequence having at least 90% sequence identity to SEQ ID NO: 56, such as at least 95% sequence identity, such as at least 96% sequence identity, such as at least 97% sequence identity, such as at least 98% sequence identity or at least 99% sequence identity, wherein the amino acid at position 44 is not a cysteine; and VH sequence having the sequence SEQ ID NO: 57 or a sequence having at least 90% sequence identity to SEQ ID NO: 57, such as at least 95% sequence identity, such as at least 96% sequence identity, such as at least 97% sequence identity, such as at least 98% sequence identity or at least 99% sequence identity, wherein the amino acid at position 100 is not a cysteine. Includes.

[0080] The scFv of this embodiment may comprise or consist of the sequence of SEQ ID NO: 61, or may comprise or consist of a sequence having at least 90% sequence identity to SEQ ID NO: 61, such as at least 95% sequence identity, such as at least 96% sequence identity, such as at least 97% sequence identity, for example at least 98% sequence identity or at least 99% sequence identity.

[0081] In one embodiment, the scFv of the invention is capable of binding to RSV, six CDR sequences consisting of amino acids 26-35, 53-59, 98-109, 177-181, 199-201, and 238-246 of SEQ ID NO: 62, or sequences which differ from these sequences by one or two substitutions; a VL sequence having the sequence of amino acids 1 to 120 of SEQ ID NO: 62 or a sequence having at least 90% sequence identity to amino acids 1 to 120 of SEQ ID NO: 62, such as at least 95% sequence identity, for example at least 96% sequence identity, such as at least 97% sequence identity, for example at least 98% sequence identity or at least 99% sequence identity; and a VH sequence having the sequence of amino acids 151 to 256 of SEQ ID NO: 62 or a sequence having at least 90% sequence identity, such as at least 95% sequence identity, for example at least 96% sequence identity, such as at least 97% sequence identity, for example at least 98% sequence identity or at least 99% sequence identity to amino acids 151 to 256 of SEQ ID NO: 62, The VL and VH are not connected by a disulfide bond.

[0082] In one embodiment, the scFv of the invention is capable of binding to B7H3, six CDR sequences consisting of amino acids 26-33, 51-58, 97-107, 175-180, 198-200, and 237-245 of SEQ ID NO: 63, or sequences which differ from these sequences by one or two substitutions; a VL sequence having at least 90% sequence identity, such as at least 95% sequence identity, for example at least 96% sequence identity, such as at least 97% sequence identity, for example at least 98% sequence identity or at least 99% sequence identity, to amino acids 149 to 255 of SEQ ID NO: 63 or to amino acids 149 to 256 of SEQ ID NO: 63; and a VH sequence having the sequence of amino acids 1 to 118 of SEQ ID NO: 63 or a sequence having at least 90% sequence identity to amino acids 1 to 118 of SEQ ID NO: 63, such as at least 95% sequence identity, for example at least 96% sequence identity, such as at least 97% sequence identity, for example at least 98% sequence identity or at least 99% sequence identity, The VL and VH are not connected by a disulfide bond.

[0083] In one embodiment, the scFv of the invention is capable of binding to HER2, six CDR sequences consisting of amino acids 27-32, 50-52, 89-97, 164-171, 189-196, and 235-247 of SEQ ID NO:64, or sequences which differ from these sequences by one or two substitutions; a VL sequence having the sequence of amino acids 1 to 108 of SEQ ID NO: 64 or a sequence having at least 90% sequence identity to amino acids 1 to 108 of SEQ ID NO: 64, such as at least 95% sequence identity, for example at least 96% sequence identity, such as at least 97% sequence identity, for example at least 98% sequence identity or at least 99% sequence identity; and a VH sequence having at least 90% sequence identity, such as at least 95% sequence identity, for example at least 96% sequence identity, such as at least 97% sequence identity, for example at least 98% sequence identity or at least 99% sequence identity, to amino acids 138 to 258 of SEQ ID NO: 64 or to amino acids 138 to 256 of SEQ ID NO: 64, The VL and VH are not connected by a disulfide bond.

[0084] In one embodiment, the scFv of the invention is capable of binding to HER2, six CDR sequences consisting of amino acids 26-33, 51-58, 97-108, 176-181, 199-201, and 238-246 of SEQ ID NO:66, or sequences which differ from these sequences by one or two substitutions; a VL sequence having the sequence of amino acids 150 to 256 of SEQ ID NO: 66 or a sequence having at least 90% sequence identity to amino acids 150 to 256 of SEQ ID NO: 66, such as at least 95% sequence identity, for example at least 96% sequence identity, such as at least 97% sequence identity, for example at least 98% sequence identity or at least 99% sequence identity; and a VH sequence having the sequence of amino acids 1 to 119 of SEQ ID NO: 66 or a sequence having at least 90% sequence identity to amino acids 1 to 119 of SEQ ID NO: 66, such as at least 95% sequence identity, for example at least 96% sequence identity, such as at least 97% sequence identity, for example at least 98% sequence identity or at least 99% sequence identity, The VL and VH are not connected by a disulfide bond.

[0085] In one embodiment, the scFv of the invention is capable of binding to DOTAM, six CDR sequences consisting of amino acids 302-310, 327-333, 372-387, 455-462, 480-482, and 519-530 of SEQ ID NO:68, or sequences which differ from these sequences by one or two substitutions; a VL sequence having the sequence of amino acids 429 to 540 of SEQ ID NO: 68 or a sequence having at least 90% sequence identity to amino acids 429 to 540 of SEQ ID NO: 68, such as at least 95% sequence identity, for example at least 96% sequence identity, such as at least 97% sequence identity, for example at least 98% sequence identity or at least 99% sequence identity; and a VH sequence having the sequence of amino acids 278 to 398 of SEQ ID NO: 68 or a sequence having at least 90% sequence identity, such as at least 95% sequence identity, for example at least 96% sequence identity, such as at least 97% sequence identity, for example at least 98% sequence identity or at least 99% sequence identity to amino acids 278 to 398 of SEQ ID NO: 68, The VL and VH are not connected by a disulfide bond.

[0086] In one embodiment, the invention relates to a bispecific antibody comprising a first and a second binding site, wherein at least one of the first and second binding sites is an scFv that does not contain a disulfide bond between the VH and VL domains.

[0087] Several forms of bispecific antibodies are known in the art, and the present invention is not limited to any of such forms. One example of a bispecific antibody is an antibody comprising two antibody heavy chains and two fusion polypeptides, the fusion polypeptides comprising an antibody light chain in which an scFv sequence is fused to the C-terminus of the light chain. Another example of a bispecific antibody is a molecule comprising two or more scFv sequences contiguously linked to each other.

[0088] In one embodiment, a bispecific antibody of the invention comprises a first scFv domain that does not contain a disulfide bond connecting the VH and VL and / or a second scFv domain that does not contain a disulfide bond connecting the VH and VL.

[0089] In one embodiment, the bispecific antibody of the invention further comprises one or more linker sequences.

[0090] In one preferred embodiment, the present invention relates to a bispecific antibody comprising a first scFv domain capable of binding to a chelator, a second scFv domain capable of binding to a tumor antigen and a SADA domain, wherein the first scFv domain and / or the second scFv domain comprises / does not comprise a disulfide bond between the VH and VL domain. Preferably, the first and / or the second scFv is an scFv of the present invention. The tumor antigen may be any antigen known to be predominantly present on the surface of tumor cells, in particular on the surface of solid tumors.

[0091] Examples of such tumor antigens include HER2, B7-H3, CA6, CD138, CD20, CD19, CD22, CD27L, CD30, CD33, CD37, CD38, CD47, CD56, CD66e, CD70, CD74, CD79b, EGFR, EGFRvIII, FRα, GCC, GPNMB, mesothelin, MUC16, NaPi2b, nectin4, PSMA, STEAP1, Trop-2, 5T4, AGS-16, alpha v beta 6, CA19.9, CAIX, CD138, CD174, CD180, CD227, CD326, CD79a, CEACAM5, CRIPTO, DLL3, DS6, endothelin B receptor, FAP, GD2, mesothelin, PMEL 17, SLC44A4, TENB2, TIM-1, CD98, Endosialin / CD248 / TEM1, Fibronectin extra domain B, LIV-1, Mucin 1, p-cadherin, peritosin, Fyn, SLTRK6, tenascin c, VEGFR2 and PRLR.

[0092] Preferred examples of tumor antigens include GD2, CD38, B7-H3, CD33, and GPA33.

[0093] GD2 is a disialoganglioside expressed in tumors of neuroectodermal origin, such as neuroblastoma and melanoma, and its expression in normal tissues is highly restricted.

[0094] CD38, also known as cyclic ADP ribose hydrolase, is a cytoplasmic enzyme that binds to CD4 + , CD8 + It is found on the surface of many immune cells, including B lymphocytes and natural killer cells. Its expression is highly elevated in melanoma cells.

[0095] B7-H3, also known as B7 homolog 3 and CD276, is a type I transmembrane protein that exists in two isoforms. It has restricted expression in normal tissues and is frequently expressed in many different types of cancer, including neuroblastoma.

[0096] CD33, also known as sialic acid-binding Ig-like lectin 3, is a cell surface antigen. It is expressed on cells of the myeloid lineage. It can be aberrantly expressed in some cases of plasma cell myeloma.

[0097] Glycoprotein 33, GPA33, is a cell surface antigen expressed on more than 95% of human colon cancers.

[0098] The binding site that can be bound to a chelator or a chelator that binds to a metal ion can be any such binding site known in the art.Preferred examples of chelators include DOTA, DOTAM and variants thereof.Examples of suitable binding sites that can be bound to a chelator or a chelator that binds to a metal ion can be found in WO 2010 / 099539, which discloses a binding site that can be bound to DOTA or a derivative of DOTA and a binding site based on antibody 2D12.5, and in WO 2019 / 201959, which discloses a rabbit antibody that can be bound to DOTAM, which are incorporated by reference.

[0099] In one embodiment, the bispecific antibody of the invention further comprises a SADA domain.

[0100] Preferably, the bispecific antibody of the invention comprises: a. a first scFv binding site capable of binding to a chelator; b. a second scFv binding site capable of binding to a tumor antigen; and c. SADA domain Includes.

[0101] In one example, the bispecific antibody of the invention comprises: a. a first scFv binding site capable of binding to a DOTA metal chelate and comprising a VL sequence of SEQ ID NO: 1 and a VH sequence of SEQ ID NO: 2; b. a second scFv binding site capable of binding to a tumor antigen; and c. SADA domain Includes.

[0102] In another example, the bispecific antibody of the invention comprises: a. a first scFv binding site capable of binding to a DOTAM metal chelate and comprising a VL sequence of amino acids 429 to 540 of SEQ ID NO:68 and a VH sequence of amino acids 278 to 398 of SEQ ID NO:68; b. a second scFv binding site capable of binding to a tumor antigen; and c. SADA domain Includes.

[0103] Preferred examples of bispecific antibodies of the invention include bispecific antibodies comprising or consisting of one of the sequences SEQ ID NO:40, SEQ ID NO:41, SEQ ID NO:42, SEQ ID NO:58, SEQ ID NO:59, SEQ ID NO:62, SEQ ID NO:63, SEQ ID NO:64, SEQ ID NO:65, SEQ ID NO:66, SEQ ID NO:67, SEQ ID NO:68 or SEQ ID NO:69.

[0104] In one embodiment, the invention relates to a composition comprising an scFv or bispecific antibody of the invention. Preferably, the composition is a pharmaceutical composition.

[0105] In one embodiment, the invention relates to the use of a bispecific antibody of the invention for diagnosing or treating cancer.

[0106] The cancer is preferably a solid cancer or tumor.

[0107] In one embodiment, the present invention provides a. administering the bispecific antibody to a patient in need thereof; and b. administering DOTA, DOTAM or a derivative thereof after a holding period. A method comprising: DOTA, DOTAM or their derivatives bind to radionuclides The present invention also relates to the use of a bispecific antibody of the present invention in a method.

[0108] In one embodiment, the present invention provides a. administering the bispecific antibody to a patient in need thereof; and b. administering DOTA, DOTAM or a derivative thereof after a holding period. A method comprising: DOTA, DOTAM or their derivatives bind to radionuclides The method relates to the use of a bispecific antibody of the invention comprising a first scFv capable of binding to a DOTA-metal or a DOTAM-metal, a second scFv capable of binding to a tumor antigen and a SADA domain.

[0109] In this embodiment, the bispecific antibody comprising a first scFv capable of binding to a DOTA-metal, a DOTAM-metal or a derivative thereof, a second scFv capable of binding to a tumor antigen and a SADA domain is preferably administered in a tetrameric form.

[0110] The holding period should be selected to allow sufficient time for the bispecific antibody to find and bind to the tumor antigen and for the unbound bispecific antibody in tetrameric form to degrade to monomeric form and thereby be rapidly cleared from the bloodstream.

[0111] The fixed holding period may be selected in the range of 48 to 96 hours.

[0112] In one embodiment, the invention further comprises administering a removal agent after step a and before step b.

[0113] In one embodiment of the present invention, DOTA or its derivatives include DOTA, benzyl DOTA and bis-chelate compounds:

[0114] [ka] [where, X 1 , X 2 , X 3 and X 4 are each independently a lone pair (i.e., donating an oxygen anion) or H; X 5 , X 6 and X 7 are each independently a lone pair (i.e., donating an oxygen anion) or H; Y 1 is O or S; n is 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, or 22; M1 is, 175 Lu 3+ , 45 Sc 3+ , 69 Ga 3+ , 71 Ga 3+ , 89 Y 3+ , 113 In 3+ , 115 In 3+ , 139 La 3+ , 136 Ce 3+ , 138 Ce 3+ , 140 Ce 3+ , 142 Ce 3+ , 151 EU 3+ , 153 EU 3+ ,159 Tb 3+ , 154 Gd 3+ , 155 Gd 3+ , 156 Gd 3+ , 157 Gd 3+ , 158 Gd 3+ or 160 Gd 3+ Selected from among; M2 is selected from among radionuclides.

[0115] In one embodiment, the radionuclide is 211 At, 51 Cr, 57 Co, 58 Co, 67 Cu, 152 EU, 67 Ga, 111 In, 59 Fe, 212 Pb, 177 Lu, 223 Ra, 224 Ra, 186 Re, 188 Re, 75 Se, 99m Tc, 227 Th, 89 Zr, 90 Y, 94m Tc, 64 Cu, 68 Ga, 66 Ga, 86 Y, 82 Rb, 110m In, 209 Bi, 211 Bi, 212 Bi, 213 Bi, 210 Po, 211 Po, 212 Po, 214 Po, 215 Po, 216 Po, 218 Po, 211 At, 215 At, 217 At, 218 At, 221 Fr,223 Ra, 224 Ra, 226 Ra, 225 Ac, 227 Ac, 227 Th, 228 Th, 229 Th, 230 Th, 232 Th, 231 Pa, 233 U, 234 U, 235 U, 236 U, 238 U, 237 Np, 238 Pu, 239 Pu, 240 Pu, 244 Pu, 241 Am, 244 Cm, 245 Cm, 248 Cm, 249 Cf and 252 Cf, preferably 177 Lu, 99 mTc, 64 Cu and 89 Zr is selected.

[0116] The chelating agent that binds to a radionuclide, such as DOTA or DOTAM or derivatives thereof bound to a radionuclide, may be administered two or more times. If the chelating agent that binds to a radionuclide is administered more than once, it is recommended that the individual administrations be separated by at least 24 hours.

[0117] The two or more administrations of a chelating agent that binds a radionuclide may use the same radionuclide or each administration may use a different radionuclide.

[0118] Such repeated administration of a chelator that binds a radionuclide, e.g., DOTA or a derivative thereof that binds a radionuclide, is disclosed in WO 2021 / 242848 (incorporated by reference) for GD2-SADA, however, the inventors understand that methods using repeated administration of a chelator that binds a radionuclide, e.g., DOTA, are not necessarily limited to GD2-SADA, but may also be applied to the bispecific antibodies of the present invention.

[0119] In one example, a bispecific antibody of the invention capable of binding to a tumor antigen is administered to a patient in need of such treatment or diagnosis, and 48 hours later, a chelator that binds an alpha emitter is administered to the patient, and 24 hours after administration of the chelator that binds the alpha emitter, a second administration of a chelator that binds a beta emitter is administered. By using such a method, the benefits of treatment with alpha emitters and the benefits of beta emitters are combined.

[0120] In another example, a bispecific antibody of the invention capable of binding to a tumor antigen is administered to a patient in need of such treatment or diagnosis, and 48 hours later, a chelating agent that binds a radionuclide suitable for PET or SPECT scanning is administered to the patient and a PET or SPECT scan is performed. Depending on the scan results, the treatment procedure can begin with the administration of a chelating agent that binds a radionuclide suitable for treating cancer 24 hours after the first administration of the chelating agent that binds the radionuclide, and treatment may be continued by administering a second or subsequent dose of a radionuclide suitable for treating cancer.

[0121] Thus, in one embodiment, the present invention provides a method for producing a composition comprising: a. administering the bispecific antibody to a patient in need thereof; b. after a holding period, administering a chelating agent that binds the radionuclide; c. after a further holding period, administering a chelating agent that binds the radionuclide; and d. Optionally, repeating step c one or more times. The present invention relates to the use of a bispecific antibody of the present invention in a method comprising the steps of:

[0122] In one embodiment, the method further comprises detecting the localization of the radionuclide. Detection can be performed using well-known methods and equipment for detecting radionuclides, such as a PET or SPECT scanner.

[0123] In one embodiment, the cancer is selected from osteosarcoma, liposarcoma, fibrosarcoma, malignant fibrous histiocytoma, leiomyosarcoma, spindle cell sarcoma, brain tumor, small cell lung cancer, retinoblastoma, and HTLV-1-infected T-cell leukemia.

[0124] The present invention relates to a kit comprising the bispecific antibody of the present invention. Preferably, the kit further comprises a chelating agent such as DOTA, DOTAM or a derivative of DOTA.

[0125] The kit may further include instructions for use or a link to such instructions.

[0126] The scFv domains and / or bispecific antibodies of the present invention may be prepared using methods known in the art.

[0127] One preferred method of providing scFv domains and / or bispecific antibodies of the invention is to provide a polynucleotide encoding the desired scFv or bispecific antibody, providing the polynucleotide with suitable regulatory sequences, e.g., promoter, terminator, enhancer, ribosome binding site, Kozak sequence, polyadenylation site, etc., inserting the construct into a suitable host cell, followed by growth under conditions which result in expression of the desired scFv or bispecific antibody.

[0128] Polynucleotide sequences can be assembled using techniques known in the art, for example using PCR techniques, or can be synthesized, for example using such sequences commercially available.

[0129] Thus, in one embodiment, the invention relates to a polynucleotide encoding an scFv domain or a bispecific antibody of the invention, an expression vector or construct comprising such a polynucleotide sequence, or a host cell comprising the polynucleotide, expression vector or construct.

[0130] In one embodiment, the present invention provides a method for producing an scFv or bispecific antibody of the invention, comprising the steps of: a. providing a host cell comprising a polynucleotide, expression vector or construct encoding an scFv or bispecific antibody of the invention; b. growing the host cell under conditions conducive to expression of the scFv or bispecific antibody; and c. Recovering the scFv or bispecific antibody from the growth broth The present invention relates to a method comprising the steps of:

[0131] [array] SEQ ID NO:1: VL amino acid sequence of DOTA-metal binding to the antibody of the invention; SEQ ID NO:2: VH amino acid sequence of DOTA-metal binding antibody of the invention; SEQ ID NO:3: Amino acid sequence of mC825; SEQ ID NO: 4: Amino acid sequence of huC825 scFv of the present invention (without disulfide bonds); SEQ ID NOs: 5 to 12: amino acid sequences of SADA domain; SEQ ID NOs: 13 to 18: CDR sequences of GD2 scFv.

[0132] SEQ ID NO: 19: VL sequence of GD2 scFv; SEQ ID NO: 20: VH sequence of GD2 scFv; SEQ ID NO:21: GD2 scFv without cysteine; SEQ ID NOs: 22 to 27: CDR sequences of CD38 scFv.

[0133] SEQ ID NO: 28: VL sequence of anti-CD38 scFv; SEQ ID NO: 29: VH sequence of anti-CD38 scFv; SEQ ID NO: 30: Shows the amino acid sequence of anti-CD38 without cysteine.

[0134] SEQ ID NOs: 31 to 36: CDR sequences of anti-CD20 scFv.

[0135] SEQ ID NO: 37: VL sequence of anti-CD20 scFv; SEQ ID NO: 38: VH sequence of anti-CD20 scFv; SEQ ID NO: 39: Shows the amino acid sequence of anti-CD20 scFv without cysteine.

[0136] SEQ ID NO: 40: Shows the amino acid sequence of the GD2-SADA construct without disulfide bonds.

[0137] SEQ ID NO: 41: Shows the amino acid sequence of the CD38-SADA construct without disulfide bonds.

[0138] SEQ ID NO: 42: Shows the amino acid sequence of the CD20-SADA construct without disulfide bonds.

[0139] SEQ ID NO: 43: Shows the amino acid sequence of the GD2-SADA construct with disulfide bonds.

[0140] SEQ ID NOs: 44 to 49 show the CDR sequences of C825.

[0141] SEQ ID NOs: 50-55: CDR sequences of anti-GPA33; SEQ ID NO:56: VL sequence of anti-GPA33; SEQ ID NO:57: VH sequence of anti-GPA33; SEQ ID NO:58: GPA33-SADA construct without cysteine; SEQ ID NO:59: GPA33-SADA construct with cysteine ​​in DOTA scFv; SEQ ID NO: 60: GPA33-SADA construct with cysteines in DOTA scFv and GPA33 scFv; SEQ ID NO:61: GPA33 scFv without cysteine; SEQ ID NO: 62 shows the amino acid sequence of the RSV-SADA construct without a disulfide bond between the VL and VH sequences. Amino acids 1-120 are the VH sequence of the anti-RSV scFv, and amino acids 151-256 are the VL sequence of the anti-RSV scFv. The light chain CDR sequence of the anti-RSV scFv is amino acids 177-181, 199-201, and 238-246 of SEQ ID NO: 62, and the heavy chain CDR sequence of the anti-RSV scFv is amino acids 26-35, 53-59, and 98-109 of SEQ ID NO: 62.

[0142] SEQ ID NO: 63 shows the amino acid sequence of the B7H3-SADA construct that does not have a disulfide bond between the VL and VH sequences. Amino acids 1 to 118 are the VH sequence of anti-B7H3, and amino acids 149 to 255 are the VL sequence of anti-B7H3 scFv. The light chain CDR sequence of anti-B7H3 scFv is amino acids 175 to 180, 198 to 200, and 237 to 245 of SEQ ID NO: 63, and the heavy chain CDR sequence of anti-B7H3 scFv is amino acids 26 to 33, 51 to 58, and 97 to 107 of SEQ ID NO: 63.

[0143] SEQ ID NO: 64 shows the amino acid sequence of the trastuzumab-based HER2-SADA construct TR-4 (anti-HER2 (VL-VH) x anti-DOTA (VH-VL)) without a disulfide bond between the VL and VH sequences. Amino acids 1-108 are the VL sequence of the HER2 scFv, and amino acids 138-258 are the VH sequence of the anti-HER2 scFv. The light chain CDR sequence of the anti-HER2 scFv is amino acids 27-32, 50-52, and 89-97 of SEQ ID NO: 64, and the heavy chain CDR sequence of the anti-HER2 scFv is amino acids 164-171, 189-196, and 235-247 of SEQ ID NO: 64.

[0144] SEQ ID NO: 65: Amino acid sequence of the Trastuzumab-based HER2-SADA construct TR-7 (anti-HER2(VH-VL) x anti-DOTA(VH-VL)) without a disulfide bond between the VL and VH sequences. Amino acids 1-120 are the VH sequence of the anti-HER2 scFv, and amino acids 151-258 are the VL sequence of the anti-HER2 scFv.

[0145] SEQ ID NO: 66 shows the amino acid sequence of the pertuzumab-based HER2-SADA construct PE-1 (anti-HER2(VH-VL)×anti-DOTA(VH-VL)) without a disulfide bond between the VL and VH sequences. Amino acids 1-119 are the VH sequence of anti-HER2 scFv and amino acids 150-256 are the VL sequence of anti-HER2 scFv. The light chain CDR sequence of anti-HER2 scFv is amino acids 176-181, 199-201 and 238-246 of SEQ ID NO: 66, and the heavy chain CDR sequence of anti-HER2 scFv is amino acids 26-33, 51-58 and 97-108 of SEQ ID NO: 66.

[0146] SEQ ID NO: 67: Amino acid sequence of Pertuzumab-based HER2-SADA construct PE-3 (anti-HER2(VL-VH)×anti-DOTA(VH-VL)) without disulfide bond between VL and VH sequences. Amino acids 1-107 are the VL sequence of anti-HER2 scFv, and amino acids 138-256 are the VH sequence of anti-HER2 scFv.

[0147] SEQ ID NO: 68 shows the amino acid sequence of CD20-DOTAM-SADA construct Ri-12 (anti-CD20 (VL-VH) x anti-DOTAM (VH-VL)), which does not have a disulfide bond between the VL and VH sequences. Amino acids 278 to 398 are the VH sequence of anti-DOTAM scFv, and amino acids 429 to 540 are the VL sequence of anti-DOTAM scFv. The light chain CDR sequence of anti-DOTAM scFv is amino acids 455 to 462, 480 to 482, and 519 to 530 of SEQ ID NO: 68, and the heavy chain CDR sequence of anti-DOTAM scFv is amino acids 302 to 310, 327 to 333, and 372 to 387 of SEQ ID NO: 68.

[0148] SEQ ID NO: 69 shows the amino acid sequence of CD20-DOTAM-SADA construct Ri-13 (anti-CD20(VL-VH) x anti-DOTAM(VL-VH)), which does not have a disulfide bond between the VL and VH sequences. Amino acids 278 to 389 are the VL sequence of the anti-DOTAM scFv, and amino acids 420 to 540 are the VH sequence of the anti-DOTAM scFv.

[0149] All cited references are incorporated by reference.

[0150] The accompanying drawings and examples are provided to illustrate, not to limit, the present invention. It is apparent to one skilled in the art that the aspects, embodiments, claims and any items of the present invention may be combined.

[0151] All percentages are on a weight / weight basis unless otherwise stated. All measurements are made under standard conditions (ambient temperature and pressure) unless otherwise stated. Test conditions comply with European Pharmacopoeia 8.0 unless otherwise stated. EXAMPLES

[0152] Materials and Methods Protein Production: The nucleotide sequence for the intended protein, including regulatory sequences to direct expression, was synthesized and inserted into an expression vector.

[0153] The expression vector was transfected into CHO cells and transformants were grown in standard medium for protein expression, after which the protein was harvested from the broth.

[0154] SDS-PAGE gel electrophoresis: Precast gels, Thermo Fisher Bolt Bis-Tris 4-12% gels, were provided by Thermo Fisher Scientific, MA USA and used according to the manufacturer's instructions. Electrophoresis gels were stained with Coomassie using the manufacturer's instructions.

[0155] [Example 1] Disulfide cross-linking in GD2-SADA A GD2-SADA construct having the amino acid sequence of SEQ ID NO:43 was prepared and purified.

[0156] The GD2-SADA construct comprises the GD2 scFv (amino acids 1-252), DOTA bound to the scFv (amino acids 275-533), and a SADA domain (amino acids 545-583). The construct contains one disulfide bond formed by cysteines C97 and C179 between the VH and VL of the GD2 scFv, and one disulfide bond formed by cysteines C369 and C513 between the VH and VL of the DOTA bound to the scFv.

[0157] The purified construct was analyzed using SE-HPLC (see FIG. 1) and found to be predominantly in tetrameric form, although the peak appeared broad and a high molecular weight shoulder was observed, suggesting that some heterogeneity may be present in the peak.

[0158] To resolve the heterogeneity, a truncated form of GD2-SADA was prepared, called GD2-SADA minus P53 domain, in which the molecule was truncated after amino acid G533, meaning that the SADA domain was missing.

[0159] The truncated forms were also analyzed by SE-HPLC, see Figure 2. As expected, the truncated forms lacked the ability to tetramerize due to the lack of the SADA domain, and thus were found predominantly as monomers, although some dimers, trimers and tetramers could also be seen in the chromatograms (see Figure 2).

[0160] This experiment demonstrated that the GD2-SADA construct formed multimers, primarily dimers, and that multimerization was not driven by the SADA domain alone.

[0161] [Example 2] SDS-PAGE analysis of multimers The GD2-SADA constructs and truncated forms prepared in Example 1 were further analyzed by SDS-PAGE chromatography. See FIG.

[0162] GD2-SADA and truncated GD2-SADA were separated under non-reducing conditions and consistently showed the presence of multimers, particularly dimers and trimers. When the truncated forms were analyzed under reducing conditions, all forms collapsed into monomeric forms, confirming that the observed multimerization was caused by disulfide bonds.

[0163] [Example 3] CD20-SADA In this example, bispecific antibody variants capable of binding to CD20 and DOTA were generated. The anti-CD20 sites varied in the order of the VH and VL regions and with or without a disulfide bond connecting the VH and VL. The DOTA binding site was an scFv as disclosed by SEQ ID NO:4 and the SADA domain was the domain as disclosed by SEQ ID NO:5.

[0164] The amino acid sequence of the VL sequence of the anti-CD20 scFv is disclosed by SEQ ID NO: 37, in which the cysteine ​​at position 99 was replaced with glycine (G) to form an anti-CD20 scFv that has no disulfide bonds. The amino acid sequence of the VH sequence of the anti-CD20 scFv is disclosed by SEQ ID NO: 38, in which the cysteine ​​at position 44 was replaced with glycine (G) to form an anti-CD20 scFv that has no disulfide bonds.

[0165] The sequence of construct Ri-3A is disclosed in SEQ ID NO:39.

[0166] The following constructs were generated:

[0167] [Table 1]

[0168] The four constructs were separated by SDS-PAGE under reducing and non-reducing conditions (see FIG. 4).

[0169] The figure shows that under non-reducing conditions, constructs containing a disulfide bond between VH and VH (Ri-2A and Ri-4A) form high molecular weight multimers and that the multimer content is strongly reduced or even absent in constructs without a disulfide bond between VH and VH (Ri-1A and Ri-3A).

[0170] Under reducing conditions, all four constructs collapsed into monomeric forms.

[0171] [Example 4] CD38-SADA In this example, bispecific antibody variants capable of binding to CD38 and DOTA were generated. The DOTA binding site is based on the scFv disclosed by SEQ ID NO: 3 and is the scFv disclosed by SEQ ID NO: 4. The DOTA binding site, which includes one disulfide bond between VH and VL, contains cysteines at positions 111 and 194. The SADA domain was the domain disclosed by SEQ ID NO: 5.

[0172] The amino acid sequence of the VL sequence of the anti-CD38 scFv is disclosed by SEQ ID NO: 28, in which the cysteine ​​at position 100 was replaced with glutamine (Q) to form an anti-CD38 scFv that does not have disulfide bonds. The amino acid sequence of the VH sequence is disclosed by SEQ ID NO: 29, in which the cysteine ​​at position 44 was replaced with glycine (G) to form an anti-CD38 scFv that does not have disulfide bonds.

[0173] The following constructs were generated:

[0174] [Table 2]

[0175] The constructs were analyzed by non-reducing SDS-PAGE, see Figure 5.

[0176] The results show that YMS9a and YMS9c contained significant amounts of multimers, whereas the amount of multimers was significantly reduced or absent in YMS9d.

[0177] The results also showed that YMS9a and YMS9c produced some heterogeneity in the monomer band, which disappeared under reducing conditions.

[0178] The YMS9d product was further analyzed by loading various amounts of 3.2 μg, 1.6 μg, 1.1 μg, and 0.5 μg on an SDS-PAGE gel under non-reducing and reducing conditions. The results, shown in Figure 6, showed that the protein eluted as a single band under both reducing and non-reducing conditions, and that only the high protein loading lane had a few additional faint bands visible under non-reducing conditions.

[0179] [Example 5] In vitro titer of CD38-SADA by SPR In this example, the binding properties of the SADA constructs of the invention were examined by SPR analysis.

[0180] YMS9a (having a disulfide bond between the VL and VH of the DOTA-binding site and a disulfide bond between the VL and VH of the CD38-binding site) and YMS9d (having no disulfide bond between VL and VH) prepared in Example 4 were analyzed for both binding to DOTA and binding to CD38 by SPR analysis.

[0181] The results showed no significant difference in in vitro binding efficacy between YMS9a and YMS9d.

[0182] [Example 6] HMW form of CD38-SADA without interchain DS bonds A solution of 10 mg / ml of compound YMS9d prepared in Example 4 was prepared. After the solution was prepared, it was equilibrated at room temperature for 3 hours. The solution was analyzed by SE-HPLC.

[0183] Samples of the stock solution were concentrated to 20 mg / ml. After the solutions were prepared, they were equilibrated at room temperature for 3 hours. The solutions were analyzed by SE-HPLC.

[0184] The 10 mg / ml and 20 mg / ml solutions were each diluted to 1 mg / ml. After the solutions were prepared, they were equilibrated at room temperature for 3 hours. The solutions were analyzed by SE-HPLC.

[0185] The results are shown in Figure 7 and indicate that the compound formed higher molecular weight forms at higher concentrations (peaks circled in Figure 7) and that virtually all of these higher molecular weight forms degraded to tetramers upon dilution.

[0186] [Example 7] CD38 and Lu-DOTA binding by SPR The binding properties of samples diluted from the 10 mg / ml and 20 mg / ml solutions described in Example 6 were analyzed by SPR. The results are shown in the table below:

[0187] [Table 3]

[0188] [Table 4]

[0189] The results showed that the formation and subsequent degradation of the HMW form did not significantly alter the binding properties.

[0190] [Example 8] CD38-SADA binds to Daudi cells in vitro Compounds YMS9c, which contains one disulfide bond between the VH and VL chains of CD38 scFv, and YMS9d, which does not have any disulfide bond between VH and VL, were used in this example. The compounds were prepared as described in Example 4.

[0191] Compound 125The cells were labeled with I and incubated with Daudi cells that contain the surface-exposed CD38 antigen. After incubation, the cells were rinsed and cell-bound radioactivity was counted.

[0192] [Table 5]

[0193] The results showed that YMS9d, which does not have a VH-VL disulfide bond, had higher binding to the anti-CD38 site compared to YMS9c, which has a disulfide bond.

[0194] [Example 9] In vivo distribution in Daudi-bearing mice Daudi tumor-bearing mice were given a 10 mg / kg injection of YMS9c, which contains one disulfide bond between the VH and VL chains of CD38 scFv, and YMS9d, which does not have any disulfide bond between VH and VL, as prepared in Example 4.

[0195] Forty-eight hours after administration of the CD38-SADA compound, 5MBq 177 Lu-DOTA / 177 Lu-Bn-DOTA was administered to mice.

[0196] Two and 24 hours after administration of radioactivity, some mice (n=4) were euthanized, dissected, and biodistribution was determined by counting the amount of radioactivity found in selected tissues: blood, tumor, and kidney.

[0197] Tumor:blood ratios were calculated:

[0198] [Table 6]

[0199] This example demonstrated higher tumor:blood uptake for CD38-SADA conjugates that do not have a disulfide bond between the VL and VH compared to conjugates that have a disulfide bond between the VH and VL of CD38 scFv.

[0200] [Example 10] RSV-SADA In this example, a bispecific antibody variant capable of binding to RSV and DOTA was generated, in which the DOTA binding site was an scFv as disclosed by SEQ ID NO:4 and the SADA domain was the domain as disclosed by SEQ ID NO:5.

[0201] Two types of RSV-SADA conjugates were generated: one type was PalDOT-SAD, which has a disulfide bond between the VL and VH of the RSV-binding scFv, and the other type was PA-3A, which does not have a disulfide bond between the VL and VH of the RSV-binding scFv.

[0202] The sequence of construct PA-3A is disclosed in SEQ ID NO:62.

[0203] The two constructs were expressed and run on a non-reducing SDS-PAGE gel as shown in Figure 8, lane 2 is PA-3A and lane 4 is PalDOT-SAD. The results showed that the construct without a disulfide bond between VH and VL ran as a single distinct band, while the version with one disulfide bond between VH and VL of the anti-RSV scFv showed the presence of several bands.

[0204] [Example 11] B7H3-SADA In this example, bispecific antibody variants capable of binding to B7H3 and DOTA were generated, where the DOTA binding site was the scFv disclosed by SEQ ID NO:4 and the SADA domain was the domain disclosed by SEQ ID NO:5.

[0205] Two types of B7H3-SADA conjugates were generated, one type, 3BH-4, which has a disulfide bond between the VL and VH of the B7H3-binding scFv, and the other type, 3BH-5, which does not have a disulfide bond between the VL and VH of the B7H3-binding scFv.

[0206] The sequence of construct 3BH-5 is disclosed in SEQ ID NO:63.

[0207] The two constructs were expressed and run on a non-reducing SDS-PAGE gel as shown in Figure 9, lane 2 is 3BH-5 and lane 4 is 3BH-4. The results showed that the construct with no disulfide bond between VH and VL ran as a single clearly distinct band, while the version with one disulfide bond between VH and VL of the anti-B7H3 scFv showed the presence of at least one high molecular weight band.

[0208] [Example 12] HER2-SADA In this example, bispecific antibody variants capable of binding to HER2 and DOTA were generated. The SADA domain was the domain disclosed by SEQ ID NO:5.

[0209] Eight versions of HER2-SADA conjugates were generated, with one series (TR series) using an anti-HER2 scFv derived from the clinical antibody trastuzumab, with four constructs:

[0210] [Table 7]

[0211] One series (PE series) used an anti-HER2 scFv derived from the clinical antibody pertuzumab, with four constructs:

[0212] [Table 8]

[0213] The TR series constructs differ from the PE series constructs in that the VH and VL sequences of the anti-HER2 scFv moiety of the TR series differ from the VH and VL sequences of the anti-HER2 scFv moiety of the PE series.

[0214] The sequence of construct TR-4 is disclosed in SEQ ID NO:64.

[0215] The sequence of construct TR-7 is disclosed in SEQ ID NO:65.

[0216] The sequence of construct PE-1 is disclosed in SEQ ID NO:66.

[0217] The sequence of construct PE-3 is disclosed in SEQ ID NO:67.

[0218] The constructs were expressed and run on a non-reducing SDS-PAGE gel. Figure 10A shows the SDS-PAGE gel of the TR series, and Figure 10B shows the SDS-PAGE gel of the PE series. The results showed that the constructs without a disulfide bond between VH and VL ran as a single distinct band, while the versions with one disulfide bond between VH and VL of HER2 scFv showed the presence of at least one high molecular weight band.

[0219] [Example 13] Anti-CD20-anti-DOTAM-SADA In this example, bispecific antibody variants were generated that were capable of binding to CD20 and DOTAM. The SADA domain was the domain disclosed by SEQ ID NO:5.

[0220] Two types of anti-CD20-anti-DOTAM-SADA conjugates were generated.

[0221] [Table 9]

[0222] The sequence of construct Ri-12 is disclosed in SEQ ID NO:68.

[0223] The sequence of construct Ri-13 is disclosed in SEQ ID NO:69.

[0224] The two constructs were expressed and run on a non-reducing SDS-PAGE gel as shown in FIG. 11, lane 2 is Ri-12 and lane 3 is Ri-13.

[0225] [array]

[0226] [ka] TIFF2024546810000013.tif234165TIFF2024546810000014.tif246170TIFF2024546810000015.tif243168TIFF20245468100 00016.tif242168TIFF2024546810000017.tif239168TIFF2024546810000018.tif252168TIFF2024546810000019.tif125168

Claims

1. A bispecific antibody comprising a first scFv domain capable of binding to a chelator or a chelator that binds to a metal ion, a second scFv domain capable of binding to a tumor antigen, and a SADA domain, wherein the VH and VL domains of the first scFv domain and / or the second scFv domain are not connected by a disulfide bond.

2. 2. The bispecific antibody of claim 1, wherein the first scFv domain does not contain a disulfide bond connecting the VH and VL, and the second scFv domain does not contain a disulfide bond connecting the VH and VL.

3. 2. The bispecific antibody of claim 1, wherein the first scFv domain capable of binding to a chelator or a chelator that binds to a metal ion is selected from an scFv capable of binding to DOTA, a derivative of DOTA, DOTAM, or any of these that bind to a metal ion.

4. a. The first scFv is capable of binding to DOTA-metal and comprises six CDR sequences each consisting of a sequence of SEQ ID NOs: 44 to 49, or b. The first scFv is capable of binding to DOTAM and comprises six CDR sequences consisting of amino acids 302-310, 327-333, 372-387, 455-462, 480-482 and 519-530 of SEQ ID NO: 68; The bispecific antibody of claim 3.

5. i) a. a VL sequence having the sequence of SEQ ID NO: 1 or a sequence having at least 90% sequence identity, at least 95% sequence identity, at least 96% sequence identity, at least 97% sequence identity, at least 98% sequence identity, or at least 99% sequence identity to SEQ ID NO: 1, wherein the amino acid at position 111 is not a cysteine; and b. a VH sequence having the sequence of SEQ ID NO: 2 or a sequence having at least 90% sequence identity, at least 95% sequence identity, at least 96% sequence identity, at least 97% sequence identity, at least 98% sequence identity, or at least 99% sequence identity to SEQ ID NO: 2, wherein the amino acid at position 45 is not a cysteine; or ii) a. a VL sequence having the sequence of amino acids 429-540 of SEQ ID NO:68, or a sequence with at least 90% sequence identity, at least 95% sequence identity, at least 96% sequence identity, at least 97% sequence identity, at least 98% sequence identity, or at least 99% sequence identity to amino acids 429-540 of SEQ ID NO:68; and b. A VH sequence having the sequence of amino acids 278-398 of SEQ ID NO:68 or a sequence having at least 90% sequence identity, at least 95% sequence identity, at least 96% sequence identity, at least 97% sequence identity, at least 98% sequence identity, or at least 99% sequence identity to amino acids 278-398 of SEQ ID NO:

68.

2. The bispecific antibody of claim 1 , comprising:

6. Tumor antigens include HER2, B7-H3, CA6, CD138, CD20, CD19, CD22, CD27L, CD30, CD33, CD37, CD38, CD47, CD56, CD66e, CD70, CD74, CD79b, EGFR, EGFRvIII, FRα, GCC, GPNMB, mesothelin, MUC16, NaPi2b, nectin-4, PSMA, STEAP1, Trop-2, 5T4, AGS-16, alpha v beta 6, CA19.9, CAIX, CD138, CD174, CD180, CD227, CD326, CD79a, CEACAM5, CRIPTO, DLL3, DS6, endothelin B receptor, FAP, GD2, mesothelin, and PMEL. 17, SLC44A4, TENB2, TIM-1, CD98, endosialin / CD248 / TEM1, fibronectin extra domain B, LIV-1, mucin 1, p-cadherin, peritosin, Fyn, SLTRK6, tenascin c, VEGFR2 and PRLR.

7. 7. The bispecific antibody of claim 6, wherein the tumor antigen is selected from the group consisting of GD2, CD38, CD20, B7-H3, GPA33, RSV, and HER2.

8. i) the second scFv is capable of binding to GD2, and a. a VL sequence having the sequence of SEQ ID NO: 19 or a sequence having at least 90% sequence identity, at least 95% sequence identity, at least 96% sequence identity, at least 97% sequence identity, at least 98% sequence identity, or at least 99% sequence identity to SEQ ID NO: 19, wherein the amino acid at position 97 is not a cysteine; and b. A VH sequence having the sequence of SEQ ID NO: 20 or a sequence having at least 90% sequence identity, at least 95% sequence identity, at least 96% sequence identity, at least 97% sequence identity, at least 98% sequence identity, or at least 99% sequence identity to SEQ ID NO: 20, wherein the amino acid at position 44 is not a cysteine. contains, or ii) the second scFv is capable of binding to CD38, and a. a VL sequence having the sequence of SEQ ID NO:28 or a sequence having at least 90% sequence identity, at least 95% sequence identity, at least 96% sequence identity, at least 97% sequence identity, at least 98% sequence identity, or at least 99% sequence identity to SEQ ID NO:28, wherein the amino acid at position 100 is not a cysteine; and b. A VH sequence having the sequence of SEQ ID NO:29 or a sequence having at least 90% sequence identity, at least 95% sequence identity, at least 96% sequence identity, at least 97% sequence identity, at least 98% sequence identity, or at least 99% sequence identity to SEQ ID NO:29, wherein the amino acid at position 44 is not a cysteine. contains, or iii) the second scFv is capable of binding to CD20; and a. a VL sequence having the sequence of SEQ ID NO: 37 or a sequence having at least 90% sequence identity, at least 95% sequence identity, at least 96% sequence identity, at least 97% sequence identity, at least 98% sequence identity, or at least 99% sequence identity to SEQ ID NO: 37, wherein the amino acid at position 99 is not a cysteine; and b. A VH sequence having the sequence of SEQ ID NO: 38 or a sequence having at least 90% sequence identity, at least 95% sequence identity, at least 96% sequence identity, at least 97% sequence identity, at least 98% sequence identity, or at least 99% sequence identity to SEQ ID NO: 38, wherein the amino acid at position 44 is not a cysteine. contains, or iv) the second scFv is capable of binding to GPA33, and a. a VL sequence having the sequence of SEQ ID NO: 56 or a sequence having at least 90% sequence identity, at least 95% sequence identity, at least 96% sequence identity, at least 97% sequence identity, at least 98% sequence identity, or at least 99% sequence identity to SEQ ID NO: 56, wherein the amino acid at position 44 is not a cysteine; and b. A VH sequence having the sequence of SEQ ID NO: 57 or a sequence having at least 90% sequence identity, at least 95% sequence identity, at least 96% sequence identity, at least 97% sequence identity, at least 98% sequence identity, or at least 99% sequence identity to SEQ ID NO: 57, wherein the amino acid at position 100 is not a cysteine. contains, or v) the second scFv is capable of binding to RSV; and a. a VL sequence having the sequence of amino acids 151-256 of SEQ ID NO: 62, or a sequence with at least 90% sequence identity, at least 95% sequence identity, at least 96% sequence identity, at least 97% sequence identity, at least 98% sequence identity, or at least 99% sequence identity to amino acids 151-256 of SEQ ID NO: 62; and b. A VH sequence having the sequence of amino acids 1 to 120 of SEQ ID NO:62 or a sequence having at least 90% sequence identity, at least 95% sequence identity, at least 96% sequence identity, at least 97% sequence identity, at least 98% sequence identity, or at least 99% sequence identity to amino acids 1 to 120 of SEQ ID NO:

62. contains, or vi) the second scFv is capable of binding to B7H3; and a. a VL sequence having the sequence of amino acids 149-255 of SEQ ID NO: 63, or a sequence with at least 90% sequence identity, at least 95% sequence identity, at least 96% sequence identity, at least 97% sequence identity, at least 98% sequence identity, or at least 99% sequence identity to amino acids 149-255 of SEQ ID NO: 63; and b. A VH sequence having the sequence of amino acids 1 to 118 of SEQ ID NO:63 or a sequence having at least 90% sequence identity, at least 95% sequence identity, at least 96% sequence identity, at least 97% sequence identity, at least 98% sequence identity, or at least 99% sequence identity to amino acids 1 to 118 of SEQ ID NO:

63. contains, or vii) the scFv is capable of binding to HER2; and a. a VL sequence having the sequence of amino acids 1-108 of SEQ ID NO:64, or a sequence with at least 90% sequence identity, at least 95% sequence identity, at least 96% sequence identity, at least 97% sequence identity, at least 98% sequence identity, or at least 99% sequence identity to amino acids 1-108 of SEQ ID NO:64; and b. A VH sequence having the sequence of amino acids 138-258 of SEQ ID NO:64 or a sequence having at least 90% sequence identity, at least 95% sequence identity, at least 96% sequence identity, at least 97% sequence identity, at least 98% sequence identity, or at least 99% sequence identity to amino acids 138-258 of SEQ ID NO:

64. contains, or viii) the scFv is capable of binding to HER2; and a. a VL sequence having the sequence of amino acids 150-256 of SEQ ID NO:66, or a sequence with at least 90% sequence identity, at least 95% sequence identity, at least 96% sequence identity, at least 97% sequence identity, at least 98% sequence identity, or at least 99% sequence identity to amino acids 150-256 of SEQ ID NO:66; and b. A VH sequence having the sequence of amino acids 1 to 119 of SEQ ID NO:66 or a sequence having at least 90% sequence identity, at least 95% sequence identity, at least 96% sequence identity, at least 97% sequence identity, at least 98% sequence identity, or at least 99% sequence identity to amino acids 1 to 119 of SEQ ID NO:

66.

8. The bispecific antibody of claim 7, comprising:

9. 5. The SADA domain has the amino acid sequence of amino acids 6 to 36 of SEQ ID NO:5, or amino acids 6 to 36 of SEQ ID NO:5 and, using the numbering of SEQ ID NO:5, the following: E6V, Q, K, G, D or A; Y7S, N, H, F, D or C; F8Y, V, S, L, I or C; T9S, P, N or A; L10V, I or F; Q11R, L, K, H or E; I12V, T, M, L or F; R13S, P, L, H, G or C; G14W, R or A; R15S, P, L, H, G or C; E16V, Q, K, G, D or A; F18Y, V, S, L, I or C; E19V, Q, K, G, D or A; M20V, T, R, L, K or I; F21L or I; R22L or G; E23V, Q, K, G, D or A; L24M; N25S, I or D; E26V, Q, K, G, D or A; A27V, T, S, G or D; L28W, V, M or F; E29Q, G or D; L30V, R, I, H or F; K31T, R, Q, N, M or E; D32Y, V, N, H, G or A; A33V, T, S, P, G or D; Q34R, L, K, H or E; 2. The bispecific antibody of claim 1 , comprising a sequence that differs by one or more substitutions selected from:

10. 2. The bispecific antibody of claim 1, comprising one of the sequences of SEQ ID NOs: 40, 41, 42, 58, 62, 63, 64, 65, 66, 67, 68 and 69.

11. 1. A method for producing a variant of an scFv domain comprising a light chain variable domain (VL), a heavy chain variable domain (VH) and one or more disulfide bonds between the VL and VH, comprising: a. Identifying cysteine ​​residues that form one or more disulfide bonds between VL and VH; and b. Substituting one or more disulfide bond-forming cysteine ​​residues identified in step a with an amino acid other than cysteine. A method comprising:

12. 12. An scFv domain comprising a VL and a VH and capable of binding to an antigen, obtainable by the method of claim 11.

13. The scFv domain of claim 12, wherein the scFv is capable of binding to CD38 and comprises six CDR sequences consisting of the sequences of SEQ ID NOs: 22 to 27, respectively.

14. The scFv a. a VL sequence having the sequence of SEQ ID NO:28 or a sequence having at least 90% sequence identity, at least 95% sequence identity, at least 96% sequence identity, at least 97% sequence identity, at least 98% sequence identity, or at least 99% sequence identity to SEQ ID NO:28, wherein the amino acid at position 100 is not a cysteine; and b. A VH sequence having the sequence of SEQ ID NO:29 or a sequence having at least 90% sequence identity, at least 95% sequence identity, at least 96% sequence identity, at least 97% sequence identity, at least 98% sequence identity, or at least 99% sequence identity to SEQ ID NO:29, wherein the amino acid at position 44 is not a cysteine. The scFv domain of claim 13, comprising:

15. A composition comprising the bispecific antibody of claim 1 or the scFv of claim 12.

16. 16. The composition of claim 15 for use in the diagnosis or treatment of cancer.

17. A polynucleotide encoding the bispecific antibody of claim 1 or the scFv of claim 12.

18. 18. An expression vector or construct comprising the polynucleotide of claim 17.

19. A host cell comprising the polynucleotide of claim 17.

20. A host cell comprising the expression vector or construct described in claim 18.

21. 13. A method for producing the bispecific antibody of claim 1 or the scFv of claim 12, comprising: a. providing a host cell according to claim 19; b. growing the host cell under conditions that induce expression of the polynucleotide; and c. Recovering the scFv or bispecific antibody from the growth broth A method comprising:

22. 13. A method for producing the bispecific antibody of claim 1 or the scFv of claim 12, comprising: a) providing a host cell according to claim 20; b. growing the host cell under conditions that induce expression of the polynucleotide; and c. Recovering the scFv or bispecific antibody from the growth broth A method comprising:

Images