HK1174663A

HK1174663A - Polypeptide modification method for purifying polypeptide multimers

Info

Publication number: HK1174663A
Application number: HK13101681.7A
Authority: HK
Inventors: 井川智之; 三瓶全次郎; 若林哲也; 伊藤绘里子
Original assignee: 中外制药株式会社
Priority date: 2009-12-25
Filing date: 2010-12-24
Publication date: 2013-06-14

Description

Modification method of polypeptide for purifying polypeptide multimer

Technical Field

The present invention relates to a method for producing and purifying a polypeptide multimer, a polypeptide multimer in which the binding force to protein A is altered, and the like.

Background

As a method for producing an IgG-type bispecific antibody having a human constant region (an IgG-type antibody having a human constant region, one arm of which has binding specificity to antigen a and the other arm of which has binding specificity to antigen B), several methods have been reported so far. Generally, an IgG-type bispecific antibody is composed of two H chains (i.e., an H chain against antigen a and an H chain against antigen B) and two L chains (i.e., an L chain against antigen a and an L chain against antigen B). In order to express two kinds of H chains and two kinds of L chains when such an IgG-type bispecific antibody is expressed, 10 kinds of combinations are considered as combinations of H2L 2. One of them is a combination with the target specificity (one arm with the binding specificity to antigen a and the other arm with the binding specificity to antigen B). Therefore, in order to obtain a bispecific antibody of interest, it is necessary to purify one antibody of interest from 10 antibodies, which is extremely inefficient and difficult.

As a method for solving this problem, a method of forming a common L chain having the same amino acid sequence using an anti-antigen a L chain and an anti-antigen B L chain has been reported (patent documents 1 and 2). In order to express two kinds of H chains and one kind of common L chain when an IgG-type bispecific antibody having such common L chains is expressed, three kinds of combinations are considered as combinations of H2L 2. One of them is a bispecific antibody of interest. The three combinations are: monospecific antibodies directed against antigen a (homologous antibodies directed against the H chain of antigen a), bispecific antibodies directed against antigen a and antigen B (heterologous antibodies directed against the H chain of antigen a and against the H chain of antigen B) and monospecific antibodies directed against antigen B (homologous antibodies directed against the H chain of antigen B). Since their ratio is typically 1:2:1, the expression efficiency of the bispecific antibody of interest is about 50%. A method of heteroassociating two H chains to further improve the efficiency has been reported (patent document 3). Therefore, the expression efficiency of the target bispecific antibody can be improved to about 90-95%. In addition, as a method for efficiently removing impurities of these two kinds of homologous antibodies, the following methods have been reported: a method in which two kinds of homologous antibodies and a target bispecific antibody (a heterologous antibody) can be purified by ion exchange chromatography by introducing amino acid substitutions into the variable regions of two kinds of H chains to give a difference in isoelectric points (patent document 4). By combining these methods, bispecific antibodies (heterologous antibodies) of the IgG class, which are antibodies with human constant regions, can be efficiently produced.

On the other hand, in the commercial production of IgG type antibodies, a purification step using protein a chromatography must be employed, but ion exchange chromatography is not necessarily used in the purification step. Therefore, the use of ion exchange chromatography for the preparation of bispecific antibodies in high purity leads to an increase in the preparation cost. In addition, in a purification method using only ion exchange chromatography, there is a possibility that the recalcitrance of the purification method as a pharmaceutical product cannot be secured, and it is desired to remove impurities by a plurality of chromatography steps.

In any case, it is preferable that the bispecific antibody can be highly purified even in a chromatography step having a separation mode different from that of ion exchange chromatography, and as the separation mode, it is desired that the bispecific antibody can be highly purified by protein a chromatography which is essentially used in the commercial production of an IgG-type antibody.

As a method for purifying a bispecific antibody (a heterologous antibody) using protein a, there have been reported so far: methods using bispecific antibodies composed of the H chain of mouse IgG2a that binds to protein a and the H chain of rat IgG2b that does not bind to protein a. It has been reported that the target bispecific antibody can be purified to 95% purity by this method using only a protein a purification step (non-patent documents 1 and 5). However, in this method, ion exchange chromatography is used in order to further improve the purity of the bispecific antibody. That is, purification of the bispecific antibody with high purity cannot be achieved only by the purification step using protein a chromatography. Furthermore, the half-life of the bispecific antibody made of the H chain of mouse IgG2a and the H chain of rat IgG2b, namely, cetuximab (castumaxomab), in humans is about 2.1 days, and is extremely short compared with the half-life of normal human IgG1, which is 2 to 3 weeks (non-patent document 2). In addition to the short half-life, immunogenicity is extremely high due to the use of mouse and rat constant regions (non-patent document 3). Therefore, the bispecific antibody obtained by this method is considered to be unsuitable as a pharmaceutical product.

On the other hand, from the viewpoint of immunogenicity, it is suggested that the possibility of using the human IgG3 constant region as a constant region not binding to protein a may be used (non-patent document 1). However, it is known that almost no association between H chains occurs in human IgG1 and human IgG3 (non-patent document 1), and therefore, it is impossible to produce a target bispecific antibody using an H chain of human IgG1 and an H chain of human IgG3 by the same method as in a bispecific antibody composed of an H chain of mouse IgG2a and an H chain of rat IgG2 b. Further, there are reports that: human IgG3 generally has a shorter half-life in humans than human IgG1, human IgG2, and human IgG4 (non-patent documents 4 and 5). Therefore, it is considered that the bispecific antibody using human IgG3 may have a short half-life in human as the bispecific antibody using mouse IgG2a and rat IgG2 b. The reason why the association between H chains hardly occurs in human IgG1 and human IgG3 is suggested to be due to the hinge sequence of human IgG3 (non-patent document 1), but the reason why the half-life of the constant region of human IgG3 is short is not sufficiently clarified. Thus, there has been no report to date on bispecific antibodies using the human IgG3 constant region as a constant region not binding to protein A. Moreover, there is no report on a method for purifying or preparing a bispecific antibody exhibiting a half-life as long as that of human IgG1 and having a human constant region with high purity and efficiency only by a protein a purification step.

Documents of the prior art

Patent document

Patent document 1: WO 98050431;

patent document 2: WO 2006109592;

patent document 3: WO 2006106905;

patent document 4: WO 2007114325;

patent document 5: WO 95033844;

non-patent document

Non-patent document 1: the Journal of Immunology, 1995, 155: 219- > 225;

non-patent document 2: j Clin Oncol 26: 2008(May 20 supl; abstr 14006);

non-patent document 3: clin Cancer Res 200713: 3899-3905;

non-patent document 4: nat Biotechnol. 2007 Dec; 25(12): 1369-72;

non-patent document 5: J. clin Invest 1970; 49: 673-80.

Disclosure of Invention

Problems to be solved by the invention

In general, ordinary IgG-type antibodies can be efficiently produced as high-purity IgG through a protein a purification step. However, in order to produce a bispecific antibody with high purity, a purification step by ion exchange chromatography must be added. The additional purification step by ion exchange chromatography increases the complexity of manufacture and the cost of manufacture. Therefore, it is preferred to prepare bispecific antibodies of high purity by only the protein a purification step. The invention aims to: provided is a method for purifying or producing a bispecific antibody of an IgG type antibody having a human antibody heavy chain constant region with high purity and efficiency only by a protein A purification step.

In addition, since the protein a binding site in the Fc region is the same as the FcRn binding site in the Fc region, it is expected that it is difficult to adjust the binding activity to protein a while maintaining the binding force to human FcRn. Maintaining binding to human FcRn is extremely important for the characteristics of IgG-type antibodies, i.e. long plasma retention in humans (long half-life). The present invention provides: a method for purifying or producing a bispecific antibody which maintains plasma retentivity equivalent to or higher than that of human IgG1, with high purity and efficiency, using only a protein A purification step.

Means for solving the problems

The inventors and others found that: a method for purifying or producing a polypeptide multimer, particularly an IgG-type multispecific antibody having a human constant region, which can bind to two or more antigens with high purity and efficiency only by a protein A purification step by changing the binding force to protein A.

Furthermore, by combining the above-mentioned method with a method of modifying amino acids constituting an interface formed when the 1 st polypeptide having an antigen binding activity and the 2 nd polypeptide having an antigen binding activity are associated and adjusting the association between the above-mentioned polypeptides, a polypeptide multimer of interest can be purified or produced with high purity and efficiency.

The inventors and others have also found that: by modifying the amino acid residue at position 435 in the EU numbering system in the heavy chain constant region, the binding force to protein A can be adjusted while maintaining plasma retention equal to or higher than that of human IgG 1. According to this finding, a bispecific antibody having plasma retentivity equivalent to or higher than that of human IgG1 can be purified or produced with high purity and efficiency.

The present invention provides the following [1] to [55] based on the above findings.

[1] A method for producing a polypeptide multimer comprising a1 st polypeptide having antigen-binding activity and a2 nd polypeptide having antigen-binding activity or not, the method comprising:

(a) a step of expressing a DNA encoding the 1 st polypeptide having an antigen-binding activity and a DNA encoding the 2 nd polypeptide having an antigen-binding activity or not; and

(b) a step of recovering the expression product of step (a),

wherein one or more amino acid residues in either or both of the 1 st polypeptide having antigen binding activity and the 2 nd polypeptide having antigen binding activity or having no antigen binding activity are modified so as to differentiate the binding force between the 1 st polypeptide having antigen binding activity and the 2 nd polypeptide having antigen binding activity or having no antigen binding activity and protein A.

[2] [1] the method according to (b), wherein the expression product is recovered by protein A affinity chromatography.

[3] [1] the method according to [1] or [2], wherein one or more amino acid residues in both or either of the 1 st polypeptide having antigen binding activity and the 2 nd polypeptide having antigen binding activity or having no antigen binding activity are modified so that a difference exists between the pH of a solvent for eluting the 1 st polypeptide having antigen binding activity from the protein A and the pH of a solvent for eluting the 2 nd polypeptide having antigen binding activity or having no antigen binding activity from the protein A.

[4] The method according to any one of [1] to [3], wherein one or more amino acid residues in the 1 st polypeptide having antigen binding activity or the 2 nd polypeptide having antigen binding activity or not have antigen binding activity are modified to increase or decrease the binding force between the 1 st polypeptide having antigen binding activity or the 2 nd polypeptide having antigen binding activity or not having antigen binding activity and protein A.

[5] The method according to any one of [1] to [4], wherein one or more amino acid residues in the 1 st polypeptide having antigen binding activity and the 2 nd polypeptide having antigen binding activity or having no antigen binding activity are modified to increase the binding force between the protein A and the polypeptide of either the 1 st polypeptide having antigen binding activity or the 2 nd polypeptide having antigen binding activity or having no antigen binding activity, and to decrease the binding force between the protein A and the polypeptide of the other.

[6] The method according to any one of [1] to [5], wherein the purity of the recovered polypeptide multimer is 95% or more.

[7] The method according to any one of [1] to [6], wherein the 1 st polypeptide having an antigen-binding activity and the 2 nd polypeptide having an antigen-binding activity or having no antigen-binding activity comprise an amino acid sequence of an antibody Fc region or an antibody heavy chain constant region.

[8] [7] the method according to claim, wherein at least one amino acid residue selected from the group consisting of positions 250 to 255, positions 308 to 317 and positions 430 to 436 in the EU numbering system in the amino acid sequence of the Fc region or the heavy chain constant region of the antibody is modified.

[9] The method according to any one of [1] to [8], wherein the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity comprise an amino acid sequence of an antibody heavy chain variable region.

[10] [9] the method according to any one of the above aspects, wherein at least one amino acid residue in the amino acid sequences of FR1, CDR2 and FR3 in the variable region of the antibody heavy chain is modified.

[11] The method according to any one of [1] to [10], wherein the polypeptide multimer comprises one or two 3 rd polypeptides having antigen-binding activity, and step (a) comprises expressing DNA encoding the 3 rd polypeptides having antigen-binding activity.

[12] The method of [11], wherein the 3 rd polypeptide having antigen-binding activity comprises an amino acid sequence of an antibody light chain.

[13] The method of [11] or [12], wherein the polypeptide multimer further comprises a4 th polypeptide having antigen-binding activity, and step (a) comprises expressing a DNA encoding the 4 th polypeptide having antigen-binding activity.

[14] [13] the method according to any one of the above methods, wherein at least one of the 3 rd polypeptide and the 4 th polypeptide having an antigen-binding activity comprises an amino acid sequence of an antibody light chain.

[15] [13] the method, wherein, with antigen binding activity of the 1 st polypeptide contains antibody light chain variable region and antibody heavy chain constant region of the amino acid sequence, with antigen binding activity of the 2 nd polypeptide contains antibody heavy chain amino acid sequence, with antigen binding activity of the 3 rd polypeptide contains antibody heavy chain variable region and antibody light chain constant region of the amino acid sequence, with antigen binding activity of the 4 th polypeptide contains antibody light chain amino acid sequence.

[16] The method according to any one of [1] to [15], wherein the polypeptide multimer is a multispecific antibody.

[17] The method of [16], wherein the multispecific antibody is a bispecific antibody.

[18] The method according to any one of [1] to [8], which comprises a1 st polypeptide having an antigen-binding activity and a2 nd polypeptide having no antigen-binding activity, wherein the 1 st polypeptide having an antigen-binding activity comprises an antigen-binding domain of a receptor and an amino acid sequence of an antibody Fc region, and the 2 nd polypeptide having no antigen-binding activity comprises an amino acid sequence of an antibody Fc region.

[19] The method according to any one of [7] to [18], wherein the antibody Fc region or the antibody heavy chain constant region is derived from human IgG.

[20] A polypeptide multimer produced by the method according to any one of [1] to [19 ].

[21] A method for purifying a polypeptide multimer comprising a1 st polypeptide having antigen-binding activity and a2 nd polypeptide having antigen-binding activity or not, the method comprising:

(b) a step of recovering the expression product of step (a) by protein a affinity chromatography;

[22] [21] the method according to any one of the above methods, wherein one or more amino acid residues in the 1 st polypeptide having an antigen-binding activity or the 2 nd polypeptide having an antigen-binding activity or having no antigen-binding activity are modified so that the binding force between the 1 st polypeptide having an antigen-binding activity or the 2 nd polypeptide having an antigen-binding activity or having no antigen-binding activity and protein A is increased or decreased.

[23] The method of [20] or [21], wherein one or more amino acid residues in the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity or having no antigen-binding activity are modified to increase the binding force between the protein A and the polypeptide of either the 1 st polypeptide having antigen-binding activity or the 2 nd polypeptide having antigen-binding activity or having no antigen-binding activity, and to decrease the binding force between the protein A and the polypeptide of the other.

[24] The method according to any one of [21] to [23], wherein the purity of the recovered polypeptide multimer is 95% or more.

[25] The method according to any one of [21] to [24], wherein the 1 st polypeptide having an antigen-binding activity and the 2 nd polypeptide having an antigen-binding activity or having no antigen-binding activity comprise an amino acid sequence of an antibody Fc region or an antibody heavy chain constant region.

[26] [25] the method according to which at least one amino acid residue selected from the group consisting of positions 250 to 255, positions 308 to 317 and positions 430 to 436 in the EU numbering system in the amino acid sequence of an antibody Fc region or an antibody heavy chain constant region is modified.

[27] The method according to any one of [21] to [26], wherein the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity comprise an amino acid sequence of an antibody heavy chain variable region.

[28] [27] the method according to any one of the above aspects, wherein at least one amino acid residue in the amino acid sequences of FR1, CDR2 and FR3 in the variable region of the antibody heavy chain is modified.

[29] The method according to any one of [21] to [28], wherein the polypeptide multimer comprises one or two 3 rd polypeptides having antigen-binding activity, and step (a) comprises expressing DNA encoding the 3 rd polypeptides having antigen-binding activity.

[30] The method of [29], wherein the 3 rd polypeptide having antigen-binding activity comprises an amino acid sequence of an antibody light chain.

[31] The method of [29] or [30], wherein the polypeptide multimer further comprises a4 th polypeptide having antigen-binding activity, and step (a) comprises expressing a DNA encoding the 4 th polypeptide having antigen-binding activity.

[32] [31] the method according to any one of [3] and [4] wherein at least one of the polypeptides having antigen-binding activity comprises an amino acid sequence of an antibody light chain.

[33] [31] the method according to any one of the above methods, wherein the 1 st polypeptide having antigen-binding activity comprises the amino acid sequences of an antibody light chain variable region and an antibody heavy chain constant region, the 2 nd polypeptide having antigen-binding activity comprises the amino acid sequence of an antibody heavy chain, the 3 rd polypeptide having antigen-binding activity comprises the amino acid sequences of an antibody heavy chain variable region and an antibody light chain constant region, and the 4 th polypeptide having antigen-binding activity comprises the amino acid sequence of an antibody light chain.

[34] The method according to any one of [21] to [33], wherein the polypeptide multimer is a multispecific antibody.

[35] The method of [34], wherein the multispecific antibody is a bispecific antibody.

[36] The method according to any one of [25] to [35], wherein the antibody Fc region or the antibody heavy chain constant region is derived from human IgG.

[37] A polypeptide multimer comprising a1 st polypeptide having antigen-binding activity and a2 nd polypeptide having antigen-binding activity or not,

wherein the 1 st polypeptide having an antigen-binding activity and the 2 nd polypeptide having an antigen-binding activity or having no antigen-binding activity have different binding abilities from protein A.

[38] [37] the polypeptide multimer, wherein the pH of the solvent that elutes the 1 st polypeptide having antigen binding activity from protein A is different from the pH of the solvent that elutes the 2 nd polypeptide having antigen binding activity or not from protein A.

[39] The polypeptide multimer according to [37] or [38], wherein the 1 st polypeptide having antigen-binding activity or the 2 nd polypeptide having antigen-binding activity or having no antigen-binding activity comprises an amino acid sequence of an antibody Fc region or an antibody heavy chain constant region,

at least one amino acid residue selected from the group consisting of positions 250 to 255, positions 308 to 317 and positions 430 to 436 in the EU numbering system in the amino acid sequence of the antibody Fc region or the antibody heavy chain constant region is modified.

[40] [37] the polypeptide multimer according to any one of [39] to [37], wherein the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity or having no antigen-binding activity comprise an amino acid sequence of an antibody Fc region or an antibody heavy chain constant region,

in the polypeptide of any one of the 1 st polypeptide having an antigen-binding activity and the 2 nd polypeptide having an antigen-binding activity or having no antigen-binding activity, the amino acid residue at position 435 in the EU numbering system in the amino acid sequence of the antibody Fc region or the antibody heavy chain constant region is histidine or arginine,

in the other polypeptide, the amino acid residue at position 435 in the EU numbering in the amino acid sequence of the antibody Fc region or the antibody heavy chain constant region is a different amino acid residue from that of the one polypeptide.

[41] The polypeptide multimer according to any one of [37] to [40], wherein the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity or having no antigen-binding activity comprise an amino acid sequence of an antibody Fc region or an antibody heavy chain constant region,

in the polypeptide of any one of the 1 st polypeptide having an antigen-binding activity and the 2 nd polypeptide having an antigen-binding activity or having no antigen-binding activity, the amino acid residue at position 435 in the EU numbering system in the amino acid sequence of the antibody Fc region or the antibody heavy chain constant region is histidine,

in another polypeptide, the amino acid residue at position 435 in the EU numbering of the amino acid sequence of the antibody Fc region or the antibody heavy chain constant region is arginine.

[42] [37] the polypeptide multimer according to any one of [37] to [41], wherein the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity comprise an amino acid sequence of an antibody heavy chain variable region,

at least one amino acid residue in the amino acid sequences of FR1, CDR2 and FR3 of the heavy chain variable region is modified.

[43] The polypeptide multimer according to any one of [37] to [42], wherein the polypeptide multimer further comprises one or two 3 rd polypeptides having antigen-binding activity.

[44] [43] the polypeptide multimer, wherein the 3 rd polypeptide having antigen-binding activity comprises the amino acid sequence of an antibody light chain.

[45] The polypeptide multimer of [43] or [44], wherein the polypeptide multimer further comprises a4 th polypeptide having antigen-binding activity.

[46] [45] the polypeptide multimer, wherein at least one of the 3 rd and 4 th polypeptides having antigen-binding activity comprises an amino acid sequence of an antibody light chain.

[47] [45] the polypeptide multimer, wherein the 1 st polypeptide having antigen-binding activity comprises the amino acid sequences of the antibody light chain variable region and the antibody heavy chain constant region, the 2 nd polypeptide having antigen-binding activity comprises the amino acid sequence of the antibody heavy chain, the 3 rd polypeptide having antigen-binding activity comprises the amino acid sequences of the antibody heavy chain variable region and the antibody light chain constant region, and the 4 th polypeptide having antigen-binding activity comprises the amino acid sequence of the antibody light chain.

[48] The polypeptide multimer according to any one of [37] to [47], which is a multispecific antibody.

[49] [48] the polypeptide multimer, wherein the multispecific antibody is a bispecific antibody.

[50] The polypeptide multimer according to any one of [37] to [41], which comprises a1 st polypeptide having antigen-binding activity and a2 nd polypeptide having no antigen-binding activity, wherein the 1 st polypeptide having antigen-binding activity comprises an antigen-binding domain of a receptor and an amino acid sequence of an antibody Fc region, and the 2 nd polypeptide having no antigen-binding activity comprises an amino acid sequence of an antibody Fc region.

[51] The polypeptide multimer according to any one of [39] to [50], wherein the antibody Fc region or the antibody heavy chain constant region is derived from human IgG.

[52] A nucleic acid encoding a polypeptide constituting the polypeptide multimer according to any one of [20] and [37] to [51 ].

[53] A vector into which the nucleic acid according to [52] has been inserted.

[54] A cell comprising the nucleic acid according to [52] or the vector according to [53 ].

[55] A pharmaceutical composition comprising the polypeptide multimer according to any one of [20] and [37] to [51] as an active ingredient.

Effects of the invention

In the present invention, there are provided: a method for purifying or producing a polypeptide multimer (multispecific antibody) having binding activity to two or more antigens with high purity and efficiency by only a protein A purification step by changing the binding force to protein A. By using the method of the present invention, a target polypeptide multimer can be purified or produced with high purity and efficiency without impairing the effect of other target amino acid modifications. In particular, by combining with a method of modulating the association between two protein domains, a polypeptide multimer of interest can be purified or produced with higher purity and efficiency.

The invention relates to a method for purifying and preparing a multispecific antibody, which is characterized by comprising the following steps: modifying amino acid residues in the antibody heavy chain constant region and/or the antibody heavy chain variable region. By introducing the amino acid residue modification of the present invention into these regions, the binding force to protein a is changed. Furthermore, other effects of modification of target amino acids, for example, retention in plasma equal to or higher than that of human IgG1, can be obtained. The method of the present invention enables highly pure and efficient production of a multispecific antibody having such an amino acid modification effect.

In general, in order to prepare an IgG-type multispecific antibody in high purity, a purification step by ion exchange chromatography must be performed. However, adding such a purification step increases the complexity of the preparation and the cost of the preparation. On the other hand, purification by ion exchange chromatography alone may lack robustness as a method for purifying a pharmaceutical product. Therefore, it is an object to develop a method for producing an IgG-type bispecific antibody only by a protein a purification step or a highly robust production method using both a protein a purification step and an ion exchange chromatography step. The present invention solves the above problems.

Drawings

FIG. 1 is a graph showing the assessment of plasma retention of MRA-IgG1 and MRA-z106/z107k in human FcRn transgenic mice;

figure 2 is a diagram showing the manner in which the same site of the Fc region of an antibody binds to protein a and FcRn;

FIG. 3 is a graph showing the change in plasma concentration after administration of Q499-z118/J339-z119/L377-k and Q499-z121/J339-z119/L377-k to human FcRn transgenic mice;

FIG. 4 is a schematic diagram of a GC33-IgG1-CD3-scFv molecule that binds divalent to a cancer specific antigen, i.e., glypican-3(GPC3), and monovalent to a T cell antigen, i.e., CD 3;

FIG. 5 is a graph showing the results of size exclusion chromatography analysis of NTA1L/NTA1R/GC33-k0 and NTA2L/NTA2R/GC33-k0 purified using protein A;

FIG. 6 is a schematic representation of a monovalent anti-GPC 3 IgG antibody molecule that binds glypican-3;

FIG. 7 is a graph showing the results of size exclusion chromatography analysis of NTA4L-cont/NTA4R-cont/GC33-k0, NTA4L-G3/NTA4R-cont/GC33-k0, and NTA4L/NTA4R/GC33-k0 purified using protein A;

FIG. 8 is a diagram showing chromatograms of protein A column chromatography purification under pH gradient elution conditions for NTA4L-cont/NTA4R-cont/GC33-k0, NTA4L-G3/NTA4R-cont/GC33-k0, and NTA4L/NTA4R/GC33-k 0;

FIG. 9 is a schematic diagram of a monovalent Fcalpha receptor Fc fusion protein molecule that binds IgA;

FIG. 10 is a graph showing the results of size exclusion chromatography analysis of IAL-cont/IAR-cont and IAL/IAR purified using protein A;

FIG. 11 is a schematic diagram of a natural anti-IL-6 receptor-anti-GPC 3 bispecific antibody no 1;

FIG. 12 is a schematic diagram of no2 in which the VH domain and the VL domain of anti-GPC 3 antibody were exchanged in no 1;

FIG. 13 is a schematic diagram of no3 in which a modification for changing the isoelectric point of each chain is introduced into no 2;

FIG. 14 is a schematic diagram of no5 in which a modification for promoting H chain heteroassociation and a modification for purifying an antibody heteroassociation with protein A are introduced into no 3;

FIG. 15 is a schematic diagram of no6 in which a modification that promotes the association of a target H chain and a target L chain is introduced into no 5;

fig. 16 is a diagram showing a chromatogram for evaluating the expression patterns of no1, no2, no3, no5, no6 as an anti-IL-6 receptor-anti-GPC 3 bispecific antibody by cation exchange chromatography;

FIG. 17 is a diagram showing a chromatogram of CM no6 eluted with a pH gradient using a HiTrap protein A HP column (GE Healthcare);

fig. 18 is a diagram showing a chromatogram for evaluating the main peak component obtained by purifying the protein a purification component of no6 using SP Sepharose HP column (GE Healthcare) by cation exchange chromatography analysis.

Detailed Description

The present invention provides a method for producing a polypeptide multimer comprising a1 st polypeptide having antigen-binding activity and a2 nd polypeptide having antigen-binding activity or not. The method for producing a polypeptide multimer of the present invention is characterized by comprising:

(b) a step of recovering the expression product of step (a),

The preparation method of the polypeptide polymer can be expressed as follows: a method for preparing polypeptide polymer with changed binding force with protein A.

In addition, in the present invention, "a polypeptide having the 1 st antigen-binding activity" may be expressed as "the 1 st polypeptide having an antigen-binding activity". The "polypeptide having the 2 nd antigen-binding activity or having no antigen-binding activity" may be expressed as "the 2 nd polypeptide having an antigen-binding activity or having no antigen-binding activity". The expression "a polypeptide having an antigen-binding activity of the 3 rd" or "a polypeptide having an antigen-binding activity of the 4 th" described later can be similarly used.

In the present invention, "comprising" means "comprising" or "consisting of".

The present invention also provides a method for purifying a polypeptide multimer comprising the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity or having no antigen-binding activity. The method for purifying a polypeptide multimer of the present invention is characterized by comprising:

(b) a step of recovering the expression product of step (a) by protein A affinity chromatography,

A polypeptide having antigen-binding activity in which one or more amino acid residues are modified can be obtained as follows: a DNA encoding a polypeptide having an antigen-binding activity or not, a modification of one or more nucleotides of the DNA, introduction into cells known to those skilled in the art, culturing the cells to express the DNA, and recovering the expression product.

Therefore, the preparation method of the polypeptide multimer of the present invention can also be expressed as: a method comprising the following steps (a) to (d).

(a) A step of providing a DNA encoding the 1 st polypeptide having an antigen-binding activity and a DNA encoding the 2 nd polypeptide having an antigen-binding activity or not;

(b) modifying one or more nucleotides in both or either of the DNA encoding the 1 st polypeptide and the DNA encoding the 2 nd polypeptide of step (a) to differentiate the binding force between the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity or not having antigen-binding activity and protein A;

(c) introducing the DNA of step (b) into a host cell, and culturing the host cell to express the DNA; and

(d) a step of recovering the expression product of step (c) from the host cell culture.

In addition, the purification method of the polypeptide multimer of the present invention can be expressed as: a method comprising the following steps (a) to (d).

(d) a step of recovering the expression product of step (c) from the host cell culture by protein a affinity chromatography.

In the present invention, the polypeptide multimer refers to a heteromultimer comprising the 1 st polypeptide and the 2 nd polypeptide. Preferably, the 1 st polypeptide and the 2 nd polypeptide exhibit binding activity to different antigens from each other. The 1 st polypeptide and the 2 nd polypeptide which show different antigen binding activities from each other are not limited at all as long as one polypeptide has an antigen binding site (amino acid sequence) different from that of the other polypeptide. For example, as shown in FIG. 4 described later, one polypeptide may be fused to an antigen-binding active site different from the other polypeptide. Alternatively, as shown in FIG. 4, FIG. 6, or FIG. 9 described later, one of the polypeptides may be a polypeptide that binds to an antigen in one valence without having an antigen-binding active site of the other polypeptide. The polypeptide multimer of the present invention also includes polypeptide multimers comprising such 1 st and 2 nd polypeptides.

Examples of the multimer include, but are not limited to, dimers, trimers, tetramers, and the like.

In addition, the 1 st polypeptide and/or the 2 nd polypeptide of the present invention can form multimers with one or two 3 rd polypeptides.

Accordingly, the present invention provides a method for preparing a polypeptide multimer comprising a1 st polypeptide having antigen-binding activity, a2 nd polypeptide having antigen-binding activity or not, and one or two 3 rd polypeptides having antigen-binding activity, the method comprising the steps of:

(a) a step of expressing a DNA encoding the 1 st polypeptide having an antigen-binding activity, a DNA encoding the 2 nd polypeptide having an antigen-binding activity, and DNAs encoding the two 3 rd polypeptides having an antigen-binding activity; and

(b) a step of recovering the expression product of step (a),

or

(a) A step of expressing a DNA encoding a1 st polypeptide having an antigen-binding activity, a DNA encoding a2 nd polypeptide having no antigen-binding activity and a DNA encoding a 3 rd polypeptide having an antigen-binding activity; and

(b) a step of recovering the expression product of step (a),

The above method may also be expressed as: a method comprising the following steps (a) to (d).

(a) A step of providing a DNA encoding the 1 st polypeptide having an antigen-binding activity, a DNA encoding the 2 nd polypeptide having an antigen-binding activity, and DNAs encoding the two 3 rd polypeptides having an antigen-binding activity;

(b) modifying one or more nucleotides in both or either of the DNAs encoding the 1 st polypeptide and the 2 nd polypeptide of step (a) to differentiate the binding force between the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity and protein A;

(c) introducing DNA encoding the 1 st polypeptide, the 2 nd polypeptide and the two 3 rd polypeptides into a host cell, and culturing the host cell to express the DNA; and

Or

(a) A step of providing a DNA encoding a1 st polypeptide having an antigen-binding activity, a DNA encoding a2 nd polypeptide having no antigen-binding activity and a DNA encoding a 3 rd polypeptide having an antigen-binding activity;

(b) modifying one or more nucleotides in both or either of the DNAs encoding the 1 st polypeptide and the 2 nd polypeptide of step (a) to differentiate the binding force between the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having no antigen-binding activity and protein A;

(c) introducing DNA encoding the 1 st polypeptide, the 2 nd polypeptide and the 3 rd polypeptide into a host cell, and culturing the host cell to express the DNA; and

In addition, the 1 st and 2 nd polypeptides of the present invention can form multimers with the 3 rd and 4 th polypeptides.

Accordingly, the present invention provides a method for preparing a polypeptide multimer comprising the 1 st polypeptide having antigen-binding activity, the 2 nd polypeptide having antigen-binding activity, the 3 rd polypeptide having antigen-binding activity and the 4 th polypeptide having antigen-binding activity, the method comprising the steps of:

(a) a step of expressing a DNA encoding the 1 st polypeptide having an antigen-binding activity, a DNA encoding the 2 nd polypeptide having an antigen-binding activity, a DNA encoding the 3 rd polypeptide having an antigen-binding activity and the 4 th polypeptide having an antigen-binding activity; and

(b) a step of recovering the expression product of step (a),

wherein one or more amino acid residues in both or the other of the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity are modified so as to differentiate the binding force between the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity and protein A.

(a) A step of providing a DNA encoding the 1 st polypeptide having an antigen-binding activity, a DNA encoding the 2 nd polypeptide having an antigen-binding activity, a DNA encoding the 3 rd polypeptide having an antigen-binding activity and the 4 th polypeptide having an antigen-binding activity;

(c) introducing DNA encoding the 1 st polypeptide, the 2 nd polypeptide, the 3 rd polypeptide and the 4 th polypeptide into a host cell, and culturing the host cell to express the DNA; and

The present invention provides a method for purifying a polypeptide multimer comprising a1 st polypeptide having antigen-binding activity, a2 nd polypeptide having antigen-binding activity or not and one or two 3 rd polypeptides having antigen-binding activity, the method comprising the steps of:

(b) a step of recovering the expression product of step (a),

or

(a) Allowing the DNA encoding the 1 st polypeptide having antigen-binding activity, the DNA encoding the 2 nd polypeptide having no antigen-binding activity and the DNA encoding one 3 rd polypeptide having antigen-binding activity to express; and

(b) a step of recovering the expression product of step (a),

wherein one or more amino acid residues in either or both of the 1 st polypeptide having antigen binding activity and the 2 nd polypeptide having antigen binding activity or having no antigen binding activity are modified to differentiate the binding force between the 1 st polypeptide having antigen binding activity and the 2 nd polypeptide having antigen binding activity or having no antigen binding activity and protein A.

(a) A step of providing a DNA encoding the 1 st polypeptide having an antigen-binding activity, a DNA encoding the 2 nd polypeptide having an antigen-binding activity or not and a DNA encoding the two 3 rd polypeptides having an antigen-binding activity;

Or

The present invention also provides a method for purifying a polypeptide multimer comprising a1 st polypeptide having antigen-binding activity, a2 nd polypeptide having antigen-binding activity, a 3 rd polypeptide having antigen-binding activity and a4 th polypeptide having antigen-binding activity, the method comprising the steps of:

(a) a step of expressing a DNA encoding the 1 st polypeptide having an antigen-binding activity, a DNA encoding the 2 nd polypeptide having an antigen-binding activity, a DNA encoding the 3 rd polypeptide having an antigen-binding activity and a DNA encoding the 4 th polypeptide having an antigen-binding activity; and

wherein one or more amino acid residues in either or both of the 1 st polypeptide having antigen binding activity and the 2 nd polypeptide having antigen binding activity are modified so as to differentiate the binding force between the 1 st polypeptide having antigen binding activity and the 2 nd polypeptide having antigen binding activity and protein A.

(a) A step of providing a DNA encoding the 1 st polypeptide having an antigen-binding activity, a DNA encoding the 2 nd polypeptide having an antigen-binding activity, a DNA encoding the 3 rd polypeptide having an antigen-binding activity and a DNA encoding the 4 th polypeptide having an antigen-binding activity;

(b) modifying one or more nucleotides in both or either of the DNA encoding the 1 st polypeptide and the DNA encoding the 2 nd polypeptide of step (a) to differentiate the binding force between the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity and protein A;

In the polypeptide multimer of the present invention comprising the 1 st polypeptide, the 2 nd polypeptide, and one or two 3 rd polypeptides, the 1 st polypeptide and the 2 nd polypeptide, respectively, can form a multimer (dimer) with the 3 rd polypeptide. In addition, the dimers formed may form multimers among each other. Two 3 rd polypeptides may have identical amino acid sequences (may have binding activity to the same antigen). Alternatively, the antigen may have two or more activities (for example, may have binding activities to 2 or more different antigens) although they have the same amino acid sequence. In addition, when there is one polypeptide 3, the polypeptide 3 may form a dimer with either the polypeptide 1 or the polypeptide 2, thereby forming a polypeptide multimer.

In the polypeptide multimer of the present invention, it is preferable that the 1 st polypeptide and the 2 nd polypeptide have binding activity to different antigens. The 3 rd polypeptide may be a polypeptide having a binding activity to the same antigen as that of either or both of the 1 st polypeptide and the 2 nd polypeptide. Alternatively, the 3 rd polypeptide may be a polypeptide having an antigen-binding activity different from that of the 1 st polypeptide or the 2 nd polypeptide.

Alternatively, the polypeptide multimer of the present invention may also be a polypeptide multimer comprising the 1 st, 2 nd, 3 rd and 4 th polypeptides. In such polypeptide multimers, the 1 st and 2 nd polypeptides can form multimers (dimers) with the 3 rd and 4 th polypeptides, respectively. For example, dimers may be formed by forming disulfide bonds between the 1 st and 3 rd polypeptides, and the 2 nd and 4 th polypeptides.

In the polypeptide multimer of the present invention, it is preferable that the 1 st polypeptide and the 2 nd polypeptide have binding activity to different antigens. The 3 rd polypeptide may be a polypeptide having a binding activity to the same antigen as that of either or both of the 1 st polypeptide and the 2 nd polypeptide. Alternatively, the 3 rd polypeptide may be a polypeptide having a binding activity to a different antigen than the 1 st polypeptide or the 2 nd polypeptide. The 4 th polypeptide may be a polypeptide having a binding activity to the same antigen as either the 1 st polypeptide or the 2 nd polypeptide. Alternatively, the 4 th polypeptide may be a polypeptide having a binding activity to a different antigen than the 1 st polypeptide or the 2 nd polypeptide.

Specifically, for example, when the 1 st polypeptide and the 2 nd polypeptide are a polypeptide comprising the amino acid sequence of an antibody heavy chain against antigen a and a polypeptide comprising the amino acid sequence of an antibody heavy chain against antigen B, respectively, the 3 rd polypeptide may be made a polypeptide comprising the amino acid sequence of an antibody light chain against antigen a, and the 4 th polypeptide may be made a polypeptide comprising the amino acid sequence of an antibody light chain against antigen B. When the polypeptide multimer of the present invention has the 3 rd polypeptide and the 4 th polypeptide comprising different amino acid sequences of two antibody light chains, respectively, the pI values of the 3 rd polypeptide and the 4 th polypeptide can be brought to different values according to the method described later, in addition to the difference in the binding force between the 1 st polypeptide and the 2 nd polypeptide and protein A; alternatively, by differentiating the binding force to protein L, the target polypeptide multimer can be purified or prepared with higher purity and efficiency.

In addition, for example, when the 1 st polypeptide is a polypeptide comprising an amino acid sequence of an antibody heavy chain against antigen a, the 2 nd polypeptide is a polypeptide comprising an amino acid sequence of an antibody light chain variable region against antigen B and an amino acid sequence of an antibody heavy chain constant region, the 3 rd polypeptide is a polypeptide comprising an amino acid sequence of an antibody light chain against antigen a, and the 4 th polypeptide is a polypeptide comprising an amino acid sequence of an antibody heavy chain variable region against antigen B and an amino acid sequence of an antibody light chain constant region, the polypeptide multimers having the 1 st to 4 th polypeptides of interest can also be purified or produced with higher purity and efficiency by using the present invention. In this case, as described in example 12 described later, by introducing an amino acid mutation that changes the pI value of the polypeptide or an amino acid mutation that promotes the association of a target polypeptide into the polypeptide of the present invention that has a difference in binding force to protein a (WO2006/106905), it is possible to purify or produce the target polypeptide multimers having the 1 st to 4 th polypeptides with higher purity and efficiency. As amino acid mutations introduced to promote polypeptide association, the following methods may also be employed: protein Eng. 1996 Jul, 9(7): 617-21., Protein Eng Des Sel. 2010 Apr, 23(4): 195-202., J Biol chem. 2010 Jun 18, 285(25): 19637-46., WO2009080254, etc. by modifying the CH3 domain of the heavy chain constant region, a method for heterologous association of two polypeptides comprising a heavy chain constant region; and methods for promoting association of a specific combination of a heavy chain and a light chain as described in WO2009080251, WO2009080252, WO2009080253, and the like.

In the present invention, the term "polypeptide having an antigen binding activity" refers to a peptide or protein having a domain (region) capable of binding to a protein or peptide such as an antigen or a ligand, such as a variable region of an antibody heavy chain or light chain, a receptor, a fusion peptide of a receptor and an Fc region, and a Scaffold, and a fragment thereof, and having a length of 5 amino acids or more. That is, the polypeptide having antigen-binding activity may comprise an antibody variable region, a receptor, a fusion peptide of a receptor and an Fc region, a Scaffold, or an amino acid sequence of these fragments.

Any polypeptide can be used as the scaffold as long as it is a polypeptide having a stable steric structure that can bind to at least one antigen. Examples of such polypeptides include, but are not limited to, those described in Nygren et al (Current Opinion in Structural Biology, 7: 463-469(1997), Journal of immunological Methods, 290: 3-28(2004)), Binz et al (Nature Biotech 23: 1257-1266(2005)), and Hosse et al (Protein Science 15: 14-27(2006)) in addition to antibody variable region fragments, fibronectin, Protein A domain, LDL receptor A domain, lipocalin, and the like.

Methods for obtaining the variable regions of antibodies, receptors, fusion peptides of receptors and Fc regions, Scaffold, and fragments thereof are well known to those skilled in the art.

The polypeptide having an antigen-binding activity may be a polypeptide derived from a living organism or an artificially designed polypeptide. It may be any polypeptide derived from a natural protein, a synthetic protein, a recombinant protein, or the like. Furthermore, the peptide or the fragment of the protein may have a domain (region) capable of binding to a protein or a peptide such as an antigen or a ligand and have a length of 10 amino acids or more, and may include a plurality of regions capable of binding to an antigen (including a ligand), as long as the peptide or the fragment has a binding ability to an antigen.

The polypeptide having antigen binding activity may also be expressed as: a polypeptide having an antigen binding protein domain.

In the present invention, the "polypeptide having no antigen binding activity" refers to a peptide or protein having a length of 5 amino acids or more, such as a fragment of an antibody having no antigen binding activity, an Fc region, a Scaffold, and a fragment thereof. That is, the polypeptide having no antigen binding activity may comprise an amino acid sequence of an antibody constant region, an Fc region, a Scaffold, or a fragment thereof, but is not limited thereto. By combining a polypeptide having no antigen-binding activity and a polypeptide having antigen-binding activity, a polypeptide multimer that binds to an antigen in one valence can also be prepared.

In addition, the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity or having no antigen-binding activity of the present invention may comprise an amino acid sequence of an antibody heavy chain constant region or an amino acid sequence of an antibody Fc region. As the amino acid sequence of the antibody Fc region or the antibody heavy chain constant region, there may be mentioned, but not limited to, the amino acid sequence of a human IgG type constant region or Fc region. The constant region or Fc region of IgG type may be any of the homologous types of natural IgG1, IgG2, IgG3, and IgG 4. Alternatively, they may be variants thereof.

In addition, the 3 rd polypeptide having antigen binding activity and the 4 th polypeptide having antigen binding activity of the present invention may comprise an amino acid sequence of an antibody light chain constant region. The amino acid sequence of the antibody light chain constant region may be, but is not limited to, the amino acid sequence of human kappa-type or human lambda-type constant regions. Alternatively, they may be variants thereof.

In addition, the polypeptide having an antigen binding activity of the present invention may comprise an amino acid sequence of an antibody variable region (e.g., amino acid sequences of CDR1, CDR2, CDR3, FR1, FR2, FR3, FR 4).

In addition, the polypeptide having an antigen-binding activity of the present invention may comprise an amino acid sequence of an antibody heavy chain or an amino acid sequence of an antibody light chain. More specifically, the 1 st polypeptide having an antigen-binding activity, and the 2 nd polypeptide having an antigen-binding activity or having no antigen-binding activity may comprise an amino acid sequence of an antibody heavy chain, and the 3 rd polypeptide having an antigen-binding activity and the 4 th polypeptide having an antigen-binding activity may comprise an amino acid sequence of an antibody light chain.

In addition, when the target polypeptide multimer is a tetramer in which the 1 st polypeptide and the 3 rd polypeptide form a dimer, the 2 nd polypeptide and the 4 th polypeptide form a dimer, and these dimers form a multimer with each other, for example, a polypeptide having an antigen-binding activity in which the 1 st polypeptide and the 2 nd polypeptide comprise an amino acid sequence of an antibody heavy chain, a polypeptide having an antigen-binding activity in which the 3 rd polypeptide and the 4 th polypeptide comprise an amino acid sequence of an antibody light chain, a polypeptide having an antigen-binding activity in which the 1 st polypeptide comprises an amino acid sequence of an antibody heavy chain, a polypeptide having an antigen-binding activity in which the 2 nd polypeptide comprises an amino acid sequence of an antibody light chain variable region and an amino acid sequence of an antibody heavy chain constant region, a polypeptide having an antigen-binding activity in which the 3 rd polypeptide comprises an amino acid sequence of an antibody light chain, a tetramer, a, The 4 th polypeptide having antigen-binding activity comprises a polypeptide comprising the amino acid sequence of an antibody heavy chain variable region and the amino acid sequence of an antibody light chain constant region.

That is, the polypeptide multimer of the present invention may be a multispecific antibody.

In the present invention, a "multispecific antibody" refers to an antibody capable of specifically binding to at least two different antigens.

In the present invention, the term "different antigens" includes not only the case where the antigenic molecules themselves are different but also the case where the antigenic molecules are the same and the antigenic determinants are different. Thus, a "different antigen" of the invention, for example, comprises different antigenic determinants within a single molecule. In the present invention, an antibody recognizing each different epitope within such a single molecule is regarded as an "antibody capable of specifically binding to a different antigen".

Examples of the multispecific antibody of the present invention include, but are not limited to, bispecific antibodies capable of specifically binding to two antigens. Preferred bispecific antibodies of the present invention include IgG antibodies of H2L2 type (composed of two H chains and two L chains) having human IgG constant regions. More specifically, examples include, but are not limited to, chimeric antibodies of the IgG type, humanized antibodies, and human antibodies.

The polypeptide having an antigen-binding activity may be, for example, a molecule having a structure in which at least two of a heavy chain variable region, a light chain variable region, a heavy chain constant region, and a light chain constant region are connected in the form of a single chain. Alternatively, the antibody may have a structure in which at least two of a heavy chain variable region, a light chain variable region, an Fc region (constant region lacking CH1 domain), and a light chain constant region are linked in a single chain.

In the present invention, the phrase "the binding force between a polypeptide having antigen-binding activity and protein a is different" means that the binding force between two or more polypeptides and protein a is different (different) by modifying the surface amino acids of the polypeptide having antigen-binding activity. More specifically, it means that, for example, the binding force of the 1 st polypeptide having antigen-binding activity to protein A is different from the binding force of the 2 nd polypeptide having antigen-binding activity to protein A. The difference in binding force to protein A can be confirmed, for example, by using protein A affinity chromatography.

The strength of the binding force of the polypeptide having antigen binding activity to protein a is related to the pH of the solvent used for elution, and the stronger the binding force of the polypeptide to protein a, the lower the pH of the solvent used for elution. Therefore, "the binding force between the polypeptide having antigen binding activity and protein a is different" may be expressed as "when two or more polypeptides having antigen binding activity are eluted by protein a affinity chromatography, the pH of the elution solvent differs between the polypeptides". The difference in pH of the elution solvent is not less than 0.1, preferably not less than 0.5, and more preferably not less than 1.0, but is not limited thereto.

In the present invention, it is preferable to change the binding force with protein a without reducing other activities (for example, retention in plasma) of the polypeptide having an antigen binding activity.

A polypeptide multimer of a target comprising a1 st polypeptide having antigen-binding activity and a2 nd polypeptide having antigen-binding activity or having no antigen-binding activity can be prepared or purified by protein A affinity chromatography using the difference in binding force between the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity or having no antigen-binding activity and protein A. Specifically, for example, when the polypeptide multimer of the present invention is a bispecific antibody having L chain commonalization (3 rd polypeptide and 4 th polypeptide having the same amino acid sequence), the polypeptide multimer can be prepared or purified by the following method. First, a nucleic acid encoding a1 st polypeptide having an antigen-binding activity (more specifically, a1 st antibody heavy chain) in which the amino acid at position 435 in the EU numbering system in the amino acid sequence of the antibody heavy chain constant region is arginine (R), a nucleic acid encoding a2 nd polypeptide having an antigen-binding activity (more specifically, a2 nd antibody heavy chain) in which the amino acid at position 435 is histidine (H), and a nucleic acid encoding a 3 rd polypeptide having an antigen-binding activity (common L chain) are introduced into a host cell, and the cell is cultured to express DNA. Next, the obtained expression product was loaded on a protein a column, and after washing, elution was performed in the order from an eluent with a high pH to an eluent with a low pH. In a cognate antibody consisting of two antibody 1 heavy chains and two common L chains, there is no binding site for protein a in the heavy chain constant region. In bispecific antibodies consisting of a1 st antibody heavy chain, a2 nd antibody heavy chain and two common L chains, there is one protein a binding site in the heavy chain constant region. In a cognate antibody consisting of two antibody 2 heavy chains and two common L chains, there are two protein a binding sites in the heavy chain constant region. As described above, the binding force between the polypeptide and protein a is related to the pH of the solvent for eluting the polypeptide in protein a affinity chromatography, and the stronger the binding force between the polypeptide and protein a, the lower the pH of the elution solvent. Thus, when elution is performed in the order from an eluent with a high pH to an eluent with a low pH, the antibody is eluted in the following order:

seeding of homologous antibodies consisting of two heavy chains of the 1 st antibody and two common L chains;

bispecific antibodies composed of a1 st antibody heavy chain, a2 nd antibody heavy chain and two common L chains;

seeding homologous antibody consisting of two heavy chains of the 2 nd antibody and two common L chains.

Thus, a target polypeptide multimer (bispecific antibody) can be prepared or purified.

The polypeptide multimer obtained according to the preparation method or the purification method of the present invention has a purity of at least 95% or more (e.g., 96%, 97%, 98%, 99% or more).

Examples of the modification of the amino acid residue that can differentiate the binding force between protein A and the 1 st polypeptide having antigen-binding activity and the binding force between protein A and the 2 nd polypeptide having antigen-binding activity or not include the following modifications (1) to (3), but are not limited thereto.

(1) Modifying one or more amino acid residues in the amino acid sequence of the polypeptide of either one of the 1 st polypeptide having antigen binding activity and the 2 nd polypeptide having antigen binding activity or not having antigen binding activity so as to increase the binding force to protein a;

(2) modifying one or more amino acid residues in the amino acid sequence of the polypeptide of either one of the 1 st polypeptide having antigen binding activity and the 2 nd polypeptide having antigen binding activity or not having antigen binding activity so that the binding force to protein a is reduced;

(3) one or more amino acid residues in the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity or having no antigen-binding activity are modified so that the binding force between the polypeptide having antigen-binding activity and the protein A is increased and the binding force between the polypeptide having antigen-binding activity and the protein A is decreased.

In the present invention, it is preferable to modify amino acids located on the surface of a polypeptide having an antigen-binding activity or having no antigen-binding activity. In addition, it is preferred to consider reducing the effect of the modification on other activities of the polypeptide.

Thus, in the present invention, for example, the following amino acid residues are preferably modified: TLMISR at positions 250-255, VLHQDWLNGK at positions 308-317 and EALHNHY at positions 430-436 in the EU numbering in the Fc region or heavy chain constant region of the antibody;

preferably TLMIS at positions 250 and 254, LHQD at positions 309 and 312, LN at positions 314 and 315, E at position 430 and LHNHY at position 432 and 436;

further preferably LMIS at positions 251 and 254, LHQ at positions 309 and 311, L at position 314, LHNH at positions 432 and 435;

in particular the amino acid residues of MIS at position 252 and 254, L at position 309, Q at position 311 and NHY at position 434 and 436.

Preferred modification positions for the amino acid modification in the antibody heavy chain variable region include FR1, CDR2, and FR 3. More preferred modification positions include, for example, EU numbers H15-H23, H56-H59, H63-H72 and H79-H83.

As the modification of the amino acid, a modification that does not reduce binding to FcRn and does not deteriorate plasma retention in a human FcRn transgenic mouse is more preferable.

More specifically, as modifications for enhancing the binding force of the polypeptide to protein a, there can be mentioned: the amino acid residue at EU-numbering position 435 in the amino acid sequence of the antibody Fc region or the antibody heavy chain constant region is substituted with histidine (His), but is not limited thereto.

In addition, as modifications for reducing the binding force of the polypeptide to protein a, there can be mentioned: the substitution of arginine for the amino acid residue at position 435 in the EU numbering in the amino acid sequence of the antibody Fc region or the antibody heavy chain constant region is not limited to these.

In addition, in the heavy chain variable region of an antibody, since the heavy chain variable region of VH3 subclass (subclass) has binding ability to protein a, it is preferable that the amino acid sequence of the modification site is identical to the heavy chain variable region sequence of VH3 subclass in order to improve the binding ability to protein a; in order to reduce the binding force to protein A, it is preferably identical to the heavy chain variable region sequences of other subclasses.

As described later, the modification of the amino acid residue can be carried out by modifying one or more nucleotides in the DNA encoding the polypeptide and allowing the DNA to be expressed in a host cell. As a person skilled in the art, the number, position or kind of nucleotides to be modified can be easily determined depending on the kind of amino acid residues after modification.

In the present invention, the modification means any one of substitution, deletion, addition, and insertion, or a combination thereof.

In addition, the polypeptide having antigen-binding activity may further comprise additional modifications in addition to the above-described modifications of the amino acid sequence. The additional modification may be any of amino acid substitution, deletion, and modification, or selected from a combination thereof, for example. Specifically, polypeptides comprising the following modifications in the amino acid sequence are encompassed by the present invention:

seeding amino acid modifications to enhance the heteroassociation rate of the two H chains of the bispecific antibody;

modification of amino acids for stabilizing disulfide bonds between the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity or having no antigen-binding activity;

modification of amino acids for improving retention in plasma of antibodies;

modification for increasing stability under acidic conditions;

modification for reducing heterogeneity;

modification for suppressing deamidation reaction is used as a seed;

modification for introducing isoelectric point difference between two polypeptides is used for seed;

modification for altering binding to Fc γ receptor.

These amino acid modifications are described below.

Two for enhancing bispecific antibodies H Amino acid modification of strand heteroassociation rates

The amino acid modification of the present invention may be combined with the amino acid modification described in WO 2006106905. The modification site is not limited as long as it is an amino acid that forms an interface between two polypeptides having antigen binding activity. Specifically, for example, when the heavy chain constant region is modified, there can be enumerated: modifying at least one combination of amino acids having the same charge among combinations at positions 356 and 439, 357 and 370, and 399 and 409 in the EU numbering system in the amino acid sequence of the heavy chain constant region of the polypeptide 1 having an antigen binding activity; in a combination of 356 nd and 439 nd, 357 nd and 370 th, and 399 nd and 409 th positions in the EU numbering of the heavy chain constant region of the 2 nd polypeptide having antigen binding activity or not, at least one combination is modified to an amino acid having an opposite charge to that of the 1 st polypeptide having antigen binding activity. More specifically, for example, in the amino acid sequences of the heavy chain constant regions of the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity, a mutation substituting Glu at position 356 in the EU numbering into Lys is introduced into either of the polypeptides, and a mutation substituting Lys at position 439 in the EU numbering into Glu is introduced into the other polypeptide. By combining such modifications with the modifications of the present invention, the target polypeptide can be obtained with higher purity only by purification using protein a.

Further, the Kabat numbering position 38 and/or the EU numbering position 123 of the light chain variable region of the polypeptide 3 having antigen binding activity is modified to have amino acids of opposite charges at the Kabat numbering position 39 and/or the EU numbering position 213 of the heavy chain constant region of the heavy chain variable region of the polypeptide 1 having antigen binding activity and the Kabat numbering position 39 and/or the EU numbering position 213 of the heavy chain variable region of the polypeptide 2 having antigen binding activity or not having antigen binding activity, respectively, and the light chain variable region of the 4 th polypeptide having antigen-binding activity, wherein the amino acid residues at position 38 by Kabat numbering and/or position 123 by EU numbering are modified to have oppositely charged amino acids, respectively, can be purified or prepared with higher purity and efficiency to obtain a multimer of the target polypeptides comprising the 1 st to 4 th polypeptides having antigen-binding activity.

For imparting antigen-binding activity 1 Polypeptide, and the second antibody having antigen-binding activity or not 2 Disulfide bond stabilized amino acid modifications between polypeptides

As described in publicly known documents (mol. Immunol. 1993, 30, 105-108 and mol. Immunol. 2001, 38, 1-8), by substituting Ser at position 228 in the EU numbering system in the amino acid sequence of the IgG4 heavy chain constant region with Pro, heterogeneity of IgG4 was eliminated and a stable structure could be maintained.

Amino acid modifications for improving retention in plasma of antibodies

In order to modulate retention in plasma, amino acid modifications of the invention can be combined with amino acid modifications used to alter the pI value of an antibody. As for the modification of the constant region, for example, there can be mentioned: modification of amino acids at positions 250 or 428 in the EU numbering system described in the publicly known documents (J. Immunol. 2006, 176(1): 346-356 or nat. Biotechnol. 199715 (7): 637-640, etc.). Examples of the modification of the variable region include amino acid modifications described in WO2007/114319 or WO 2009/041643. The amino acid to be modified is preferably an amino acid exposed on the surface of the polypeptide having an antigen-binding activity. For example, the substitution of the amino acid at position 196 in the EU numbering system in the amino acid sequence of the heavy chain constant region can be mentioned. When the heavy chain constant region is IgG4, for example, substitution of lysine at position 196 with glutamine can lower the pI value and improve the plasma retention.

In addition, by altering the binding to FcRn, plasma retention can also be modulated. Examples of amino acid modifications for altering The binding to FcRn include known amino acid modifications (The Journal of Biological Chemistry, Vol. 276, Nos. 96591-6604, 2001 or Molecular Cell, Vol. 7, 867-877, 2001 orCurr Opin Biotechnol. 2009, 20(6): 685-91.) Amino acid substitutions of the antibody heavy chain constant region described in (1). Examples thereof include: amino acid substitution at positions 233, 238, 253, 254, 255, 256, 258, 265, 272, 276, 280, 285, 288, 290, 292, 293, 295, 296, 297, 298, 301, 303, 305, 307, 309, 311, 312, 315, 317, 329, 331, 338, 360, 362, 376, 378, 380, 382, 415, 424, 433, 434, 435, 436 and the like in the EU numbering system.

Modifications for improving stability under acidic conditions

When IgG4 is used as the heavy chain constant region, it is preferable to suppress hemimolarity of IgG4 under acidic conditions and maintain a stable 4-chain structure (H2L2 structure). Therefore, it is preferable to replace arginine (arginine 2002, 105, 9-19) which is an amino acid at position 409 in the EU numbering system and plays an important role in maintaining the 4-chain structure with lysine of the IgG1 type which maintains a stable 4-chain structure even under acidic conditions. Further, methionine, which is an amino acid at position 397 in the EU numbering system, may be substituted with valine in order to improve the acid stability of IgG 2. The above modifications may be used in combination with the amino acid modifications of the present invention.

Modifications for reducing heterogeneity

The amino acid modification of the invention may be combined with the method described in WO 2009041613. Specifically, for example, a modification in which two amino acids at the C-terminal end of the constant region of the IgG1 heavy chain, i.e., glycine at position 446 and lysine at position 447 in the EU numbering system, are deleted may be combined with the amino acid modifications according to the examples of the present invention.

Modification for inhibition of deamidation reactions

The amino acid modification of the present invention may be combined with an amino acid modification for inhibiting deamidation reaction. It has been reported that deamidation reactions are likely to occur particularly at sites (seeds, etc.) where asparagine (N) is adjacent to glycine (G) (Geiger et al, J. Bio. chem. 1987; 262: 785. cna 794). When a site adjacent to asparagine and glycine is present in the polypeptide multimer (multispecific antibody) of the present invention, deamidation can be inhibited by modifying the amino acid sequence. Specifically, for example, an amino acid of one or both of asparagine and glycine is substituted with another amino acid. More specifically, for example, there may be mentioned: asparagine was substituted for aspartic acid.

Modifications for introducing differences in isoelectric point between two polypeptides

The amino acid modification of the present invention may be combined with an amino acid modification for introducing a difference in isoelectric point. A specific method is described, for example, in WO 2007/114325. In addition to the modification of the present invention, the amino acid sequences of the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity or having no antigen-binding activity are modified to make differences in isoelectric point values of these polypeptides, so that the target polypeptide can be purified or produced with higher purity and efficiency. In addition, by making the 3 rd polypeptide having antigen binding activity and the 4 th polypeptide having antigen binding activity have a difference in isoelectric point, it is also possible to purify or prepare a target polypeptide multimer comprising the 1 st to 4 th polypeptides with higher purity and efficiency. As specific modification sites, when the 1 st polypeptide and the 2 nd polypeptide comprise an amino acid sequence of an antibody heavy chain, there can be exemplified: kabat numbering 1, 3, 5, 8, 10, 12, 13, 15, 16, 19, 23, 25, 26, 39, 42, 43, 44, 46, 68, 71, 72, 73, 75, 76, 81, 82b, 83, 85, 86, 105, 108, 110, 112. When the 3 rd polypeptide and the 4 th polypeptide comprise an amino acid sequence of an antibody light chain, there can be enumerated, for example: kabat numbering 1, 3, 7, 8, 9, 11, 12, 16, 17, 18, 20, 22, 38, 39, 41, 42, 43, 45, 46, 49, 57, 60, 63, 65, 66, 68, 69, 70, 74, 76, 77, 79, 80, 81, 85, 100, 103, 105, 106, 107, 108. The difference in isoelectric point can be produced by making at least one amino acid residue at the above-mentioned position in one polypeptide an amino acid having a charge, and making at least one amino acid residue at the above-mentioned position in the other polypeptide an amino acid residue having a charge opposite to the charge or an uncharged amino acid residue.

For changing and Fc modification of binding to gamma receptors

Amino acid modifications of the invention may be combined with amino acid modifications for altering (increasing or decreasing) binding to Fc γ receptors. Examples of the modification for changing the binding property to the Fc γ receptor include, but are not limited to, the modifications described in Curr Opin Biotechnol. 2009, 20(6): 685-91. Specific methods include, for example: binding to Fc γ receptors can be altered by combining the modifications of the invention with modifications that replace the leucine at positions 234 and 235 and asparagine at position 272 of the EU numbering of the constant region of the heavy chain of IgG1 with other amino acids. The substituted amino acid is, for example, alanine, but is not limited thereto.

The following describes the preparation of a DNA encoding a polypeptide having an antigen-binding activity, the modification of one or more nucleotides, the expression of the DNA, and the recovery of an expression product.

Encoding a polypeptide having antigen-binding activity DNA Preparation of

In the present invention, the DNA encoding a polypeptide having antigen-binding activity or the DNA encoding a polypeptide having no antigen-binding activity may use the whole length or a part of a known sequence (naturally occurring sequence, non-existing sequence), or a combination thereof. Such DNA can be obtained by a method known to those skilled in the art. For example, the antibody can be obtained from an antibody library or a gene encoding an antibody can be cloned from a hybridoma producing a monoclonal antibody.

For the antibody library, a plurality of antibody libraries are already known, and a method for preparing an antibody library is also known, and a person skilled in the art can obtain an appropriate antibody library. For example, with regard to phage antibody libraries, reference may be made to Clackson et al, Nature 1991, 352: 624-8; marks et al, j. mol. biol. 1991, 222: 581-97; water houses et al, Nucleic Acids Res. 1993, 21: 2265-6; griffiths et al, EMBO J. 1994, 13: 3245-60; vaughan et al, Nature Biotechnology 1996, 14: 309-14, and Japanese patent application laid-open No. Hei 20-504970. Further, known methods such as a method using eukaryotic cells as a library (WO 95/15393) and a ribosome-indicating method can be used. Furthermore, a technique for obtaining a human antibody by panning using a human antibody library is also known. For example, a phage capable of binding to an antigen can be selected by expressing the variable region of a human antibody as a single-chain antibody (scFv) on the surface of the phage by phage display. Analysis of the genes of the selected phage allows the determination of the DNA sequence encoding the variable region of the human antibody that binds to the antigen. Once the DNA sequence of scFv that binds to the antigen is clarified, an appropriate expression vector can be prepared based on the sequence to obtain a human antibody. The above-mentioned methods are well known and reference may be made to WO92/01047, WO92/20791, WO93/06213, WO93/11236, WO93/19172, WO95/01438, WO 95/15388.

The method for obtaining a gene encoding an antibody from a hybridoma can basically employ a known technique. Specifically, a desired antigen or a cell expressing a desired antigen is used as a sensitizing antigen, and the immunization is performed according to a usual immunization method. The obtained immune cells are fused with known parent cells by a usual cell fusion method, and cells (hybridomas) producing monoclonal antibodies are selected by a usual screening method. By transcribing the mRNA of the resulting hybridoma using a reverse transcriptase, cDNA for the variable region (V region) of the antibody can be obtained. The gene encoding the antibody can be obtained by ligating the cDNA to DNA encoding the constant region (C region) of the desired antibody.

More specifically, the following method can be exemplified, but is not limited thereto.

The sensitizing antigen used for obtaining the antibody genes encoding the heavy chain and the light chain of the antibody includes both a complete antigen with immunogenicity and an incomplete antigen without immunogenicity, and the incomplete antigen includes a hapten and the like. For example, a full-length protein or a partial peptide of the target protein, or the like can be used. It is known that a substance composed of a polysaccharide, a nucleic acid, a lipid, or the like can be used as an antigen, and the antigen is not particularly limited in the present invention. The antigen can be prepared by a method known to those skilled in the art, for example, a method using baculovirus (for example, WO 98/46777) and the like. Hybridomas can be prepared, for example, by the method of Milstein et al (G. Kohler and C. Milstein, Methods Enzymol. 1981, 73: 3-46). When the immunogenicity of the antigen is low, the antigen can be immunized by binding it to a macromolecule having immunogenicity such as albumin. If necessary, the antigen may be bound to another molecule to become a soluble antigen. When a transmembrane molecule such as a receptor is used as an antigen, an extracellular region portion of the receptor may be used as a fragment, or a cell expressing the transmembrane molecule on the cell surface may be used as an immunogen.

Antibody-producing cells can be obtained by immunizing an animal with the above-mentioned appropriate sensitizing antigen. Alternatively, antibody-producing lymphocytes are immunized in vitro to become antibody-producing cells. Various mammals can be used as the animal to be immunized, but rodents, lagomorphs, and primates are generally used. There can be exemplified: rodents such as mice, rats, hamsters, and the like; rabbit-shaped mesh such as rabbit; primates such as monkeys including cynomolgus monkey, rhesus monkey, baboon, and chimpanzee. Furthermore, transgenic animals having all the components of human antibody genes (reporters) are also known, and human antibodies can be obtained by using these animals (see WO 96/34096; Mendez et al, nat. Genet. 1997, 15: 146-56). Instead of using the above transgenic animal, a desired human antibody having an antigen-binding activity can also be obtained by sensitizing human lymphocytes with a desired antigen or cells expressing a desired antigen in vitro and fusing the sensitized lymphocytes with human myeloma cells, for example, U266 (see Japanese patent publication No. Hei 1-59878). Alternatively, a desired human antibody can be obtained by immunizing a transgenic animal having all the components of human antibody genes with a desired antigen (see WO93/12227, WO92/03918, WO94/02602, WO96/34096, WO 96/33735).

Immunization of animals can be carried out as follows: the sensitizing antigen is diluted and suspended in a Phosphate Buffered Saline (PBS) or a physiological saline as appropriate, mixed with an adjuvant as necessary, emulsified, and then injected into the abdominal cavity or the subcutaneous space of an animal to immunize. After that, the sensitizing antigen mixed in Freund's incomplete adjuvant is preferably administered several times every 4 to 21 days. The production of antibodies can be confirmed by measuring the titer of the antibody of interest in the serum of an animal by a conventional method.

Hybridomas can be prepared as follows: antibody-producing cells obtained from animals immunized with the desired antigen or lymphocytes (Monoclonal Antibodies): Principles and Practice, Academic Press, 1986, 59-103) or lymphocytes are fused with myeloma cells using commonly used fusion agents (e.g., polyethylene glycol). The binding specificity of the antibody produced by the hybridoma is measured by culturing and proliferating the hybridoma as necessary, and then performing a known analysis method such as immunoprecipitation, Radioimmunoassay (RIA), and enzyme-linked immunosorbent assay (ELISA). Thereafter, the antibody-producing hybridomas are subcloned by a limiting dilution method or the like as necessary, and the target specificity, affinity, or activity of the antibodies are measured.

Next, the gene encoding the selected antibody can be cloned from a hybridoma or an antibody-producing cell (sensitized lymphocyte or the like) using a probe capable of specifically binding to the antibody (for example, an oligonucleotide complementary to a sequence encoding the antibody constant region or the like). Cloning from mRNA can also be performed by RT-PCR. Immunoglobulins are classified as: IgA, IgD, IgE, IgG and IgM five different classes. Furthermore, these classes are divided into several subclasses (homologous types) (e.g., IgG-1, IgG-2, IgG-3, and IgG-4; IgA-1, IgA-2, etc.). In the present invention, the heavy chain and the light chain used for the production of the antibody are not particularly limited, and may be derived from an antibody belonging to any of the above-mentioned classes and subclasses, and particularly preferably from IgG.

Here, genes encoding the heavy and light chains may also be modified by genetic engineering techniques. For example, an antibody such as a mouse antibody, a rat antibody, a rabbit antibody, a hamster antibody, a goat antibody, or a camel antibody may be appropriately modified to reduce its foreign antigenicity to humans, and a genetically recombinant antibody, such as a chimeric antibody or a humanized antibody, may be appropriately prepared. A chimeric antibody is an antibody composed of the variable regions of the heavy and light chains of a non-human mammal such as a mouse antibody and the constant regions of the heavy and light chains of a human antibody, and can be produced by ligating a DNA encoding the variable region of a mouse antibody and a DNA encoding the constant region of a human antibody, inserting the ligated DNA into an expression vector, and introducing the resulting vector into a host. Humanized antibodies, also known as reshaped (reshaped) human antibodies, can be synthesized by PCR from a plurality of oligonucleotides that are made to have overlapping portions at the ends of DNA sequences designed to link the Complementarity Determining Regions (CDRs) of non-human mammals, such as mouse antibodies. The resulting DNA is ligated with a DNA encoding a human antibody constant region, inserted into an expression vector, and introduced into a host to produce a humanized antibody (see EP 239400; WO 96/02576). FRs of human antibodies connected via CDRs are selected for which complementarity determining regions form good antigen binding sites. If desired, amino acids in the framework regions of the variable regions of the antibody may be substituted so that the complementarity determining regions of the reshaped human antibody form appropriate antigen binding sites (K. Sato et al, Cancer Res. 1993, 53: 851-. The monoclonal antibody of the present invention includes such a humanized antibody or a chimeric antibody.

When the antibody of the present invention is a chimeric antibody or a humanized antibody, a human antibody-derived constant region is preferably used as the constant region of the antibodyAnd (5) positioning. For example, C can be used in the heavy chainγ1、Cγ2、Cγ3、Cγ4; in the light chain, C may be usedκ、Cλ. In order to improve the stability of the antibody or its production, the human antibody constant region may be modified as necessary. The chimeric antibody of the present invention is preferably composed of a variable region derived from an antibody derived from a non-human mammal and a constant region derived from a human antibody. The humanized antibody is preferably composed of CDR derived from an antibody derived from a mammal other than human and FR and C regions derived from a human antibody. The constant region derived from a human antibody has an amino acid sequence inherent to each of IgG (IgG1, IgG2, IgG3, IgG4), IgM, IgA, IgD and IgE isologous types. The constant region used in the humanized antibody of the present invention may be a constant region of an antibody belonging to any homologous type. The constant region of human IgG is preferably used, but not limited thereto. The FR derived from a human antibody to be used for a humanized antibody is also not particularly limited, and may be the FR of an antibody belonging to any homologous type.

The variable and constant regions of the chimeric and humanized antibodies of the present invention may be modified by deletion, substitution, insertion and/or addition, as long as they show the binding specificity of the original antibody.

Chimeric antibodies and humanized antibodies using human-derived sequences are considered to be effective when administered to humans for therapeutic purposes and the like because of their reduced antigenicity in humans.

In the present invention, amino acids may be modified in order to alter the biological properties of the antibody.

Small molecule antibodies are useful as antibodies, both from the viewpoint of in vivo kinetic properties and from the viewpoint of being able to be produced at low cost using e.coli, plant cells, and the like.

The antibody fragment is a small molecule antibody. Small molecule antibodies also include antibodies having an antibody fragment as a portion thereof. The small molecule antibody of the present invention is not particularly limited in its structure, production method, and the like, as long as it has an antigen binding ability. Among small molecule antibodies, there are also antibodies with higher activity than full-length antibodies (Orita et al, Blood (2005) 105: 562-566). In the present specification, the "antibody fragment" is not particularly limited as long as it is a part of a full-length antibody (for example, full-length IgG), and preferably includes a heavy chain variable region (VH) or a light chain variable region (VL). Examples of preferred antibody fragments are: fab, F (ab ') 2, Fab', Fv, etc. The amino acid sequence of the heavy chain variable region (VH) or the light chain variable region (VL) in the antibody fragment may be modified by substitution, deletion, addition and/or insertion. Further, a part of the amino acid sequence of the variable region (VH) or the variable region (VL) may be deleted as long as the antigen-binding ability is maintained. For example, of the above antibody fragments, "Fv" is the smallest antibody fragment that includes the entire antigen recognition site and binding site. "Fv" is a dimer of a heavy chain variable region (VH) and a light chain variable region (VL) strongly bound by a non-covalent bond (VH-VL dimer). An antigen binding site is formed on the surface of the VH-VL dimer by 3 complementary Chain Determining Regions (CDRs) of each variable region. The 6 CDRs confer an antigen binding site on the antibody. However, even a single variable domain (or half of an Fv comprising only 3 antigen-specific CDRs), although having a low affinity compared to the full binding site, has the ability to recognize and bind antigen. Accordingly, molecules smaller than the above-described Fv are also included in the antibody fragment of the present invention. The variable regions of antibody fragments may also be chimeric and humanized.

The small molecule antibody preferably comprises both a heavy chain variable region (VH) and a light chain variable region (VL). Examples of small molecule antibodies are: antibody fragments of Fab, Fab ', F (ab') 2 and Fv, as well as scFv (single chain Fv) made using antibody fragments (Huston et al, Proc. Natl. Acad. Sci. USA (1988) 85: 5879-83; Pluckthun "The Pharmacology of Monoclonal Antibodies Pharmacology)" volume 113, Reaenburg and Moore, Springer Verlag, New York, pp. 269-315, (1994)), diabody (diabody) (Holliger et al, Proc. Natl. Acad. Sci. USA (1993) 90: 6444-8; EP 404097; WO 93/11161; Johnson et al, Method in Enzymology (1991) 203: 88-98; Holliger et al, Protein Engineering (1996) 9: 299, History et al (1996) 2: 597; Hippon et al, Protein Engineering et al (1996) 2: 5926; Pihnson et al, It. At20: 1994), j immunol. Methods (1999) 231: 177-89 parts of a base; orita et al, Blood (2005) 105: 562-566), three-chain antibodies (triabody) (Journal of Immunological Mtthouses (1999) 231: 177-89) and tandem diabodies (Cancer Rrsearch (2000) 60: 4336-41), and the like.

Antibody fragments can be obtained by treating the antibody with an enzyme such as a protease such as papain, pepsin, etc. (cf. Morimoto et al, J. biochem. Biophys. Methods (1992) 24: 107-17; Brennan et al, Science (1985) 229: 81). The antibody fragment may be prepared by gene recombination based on the amino acid sequence of the antibody fragment.

Small molecule antibodies having a modified structure of an antibody fragment can be constructed using an antibody fragment obtained by enzyme treatment or gene recombination. Alternatively, genes encoding the entirety of the small molecule antibody can also be constructed and introduced into expression vectors, which are then expressed in appropriate host cells (see, e.g., Co et al, J. Immunol. (1994) 152: 2968-76; Better and Horwitz, Methods Enzymol. (1989) 178: 476-96; Pluckthun and Skerra, Methods Enzymol. (1989) 178: 497 515; Lamoyi, Methods Enzymol. (1986) 121: 652-63; Rousseaux et al, Methods Enzymol. (1986) 121: 663-9; Bird and Walker, Trends Biotechnol. (1991) 9: 132-7).

The "scFv" is a single-chain polypeptide in which two variable regions are linked via a linker or the like as necessary. The scFv comprises two variable regions, typically one VH and one VL, but may also comprise two VH or two VL. Typically, scFv polypeptides contain a linker between the VH and VL domains through which the paired portions of VH and VL necessary for antigen binding are formed. In general, in order to form a pair portion between VH and VL within the same molecule, the linker connecting VH and VL is generally made to be a peptide linker of 10 amino acids or more in length. However, the linker of the scFv in the present invention is not limited to the peptide linker as long as it does not inhibit the formation of scFv. For an overview of scFv, reference may be made to Pluckthun "The Pharmacology of Monoclonal antibodies" Vol 113 (eds. Rosenburg and Moore, Springer Verlag, NY, pp. 269-315 (1994)).

Further, "diabody (Db)" means a bivalent antibody fragment constructed by gene fusion (P. Holliger et al, Proc. Natl. Acad. Sci. USA 90: 6444-6448(1993), EP404,097, WO93/11161, etc.). Diabodies are dimers composed of two polypeptide chains, each of which is linked in the same chain by a linker that is short enough not to bind to each other, for example, around 5 residues. The VL and VH encoded on the same polypeptide chain form a dimer because the linker between them is short and cannot form a single-chain V-region fragment (fragment), and thus the diabody has two antigen-binding sites. In this case, if a combination of VLa-VHb and VLb-VHa linked by a linker of about 5 residues is expressed simultaneously with respect to VL and VH of two different epitopes (a, b), it is secreted as a bispecific Db.

Diabodies comprise two molecules of scFv and therefore four variable regions, and as a result, diabodies have two antigen binding sites. In order to form a diabody, a linker of about 5 amino acids is generally used when the linker connecting VH and VL in each scFv molecule is a peptide linker, unlike the case of an scFv that does not form a dimer. However, the linker of the scFv that forms the diabody is not limited to the peptide linker as long as it does not interfere with the expression of the scFv and the formation of the diabody.

It is preferred that the small molecule antibodies and antibody fragments of the present invention further have the amino acid sequence of the antibody heavy chain constant region and/or light chain constant region.

Alteration of one or more nucleotides

In the present invention, "nucleotide change" means that a polypeptide encoded by a DNA has a target amino acid residue by performing genetic manipulation or mutation treatment in which at least one nucleotide is inserted, deleted or substituted in the DNA. That is, it means that a codon encoding an original amino acid residue is converted into a codon encoding a target amino acid residue. The nucleotide changes described above can be carried out by site-specific mutagenesis (see, for example, Kunkel (1985) Proc. Natl. Acad. Sci. USA 82: 488), PCR mutagenesis, cassette mutagenesis, and the like. Generally, an antibody mutant having improved biological properties has homology and/or similarity of 70% or more, more preferably 80% or more, and still more preferably 90% or more (e.g., 95% or more, 97%, 98%, 99%, etc.) of the amino acid sequence of the variable region of the original antibody. In the present specification, homology and/or similarity of sequences is defined as: sequence alignment and gap introduction (gap introduction) is performed as required to maximize sequence homology, followed by the proportion of amino acid residues that are identical (same residues) or similar (classified as amino acid residues of the same group according to the nature of the general amino acid side chains) to the original amino acid residues. In general, natural amino acid residues are divided into the following groups according to the nature of their side chains: (1) hydrophobicity: alanine, isoleucine, valine, methionine and leucine; (2) neutral hydrophilicity: asparagine, glutamine, cysteine, threonine, and serine; (3) acidity: aspartic acid and glutamic acid; (4) alkalinity: arginine, histidine and lysine; (5) residues that influence chain orientation: glycine and proline; and (6) aromatic: tyrosine, tryptophan and phenylalanine. The number of modified amino acids may be, for example, 10, 9, 8, 7, 6, 5, 4, 3, two or one amino acid, but is not limited thereto.

Typically, a total of 6 complementarity determining regions (hypervariable regions; CDRs) present in the variable regions of the heavy and light chains interact to form the antigen binding site of an antibody. Even one of the variable regions is known to have the ability to recognize and bind to an antigen, although the affinity is low as compared with the variable region including all the binding sites. Accordingly, the polypeptide having an antigen-binding activity of the present invention may encode a fragment portion including each antigen-binding site of the heavy chain and the light chain of the antibody as long as it maintains the binding property to a desired antigen.

As described above, the method of the present invention can efficiently obtain, for example, a desired polypeptide multimer which retains the activity of the polypeptide.

In a preferred embodiment of the present invention, the amino acid residues to be used for the "modification" are appropriately selected from, for example, the amino acid sequences of the variable regions of the heavy and light chains of an antibody, and the amino acid sequences of the variable regions of the light and light chains of an antibody.

DNA Expression of

The DNA encoding the polypeptide thus modified is cloned (inserted) into an appropriate vector and introduced into a host cell. The vector is not particularly limited as long as the inserted nucleic acid is stably retained, and for example, when Escherichia coli is used as a host, a cloning vector is exemplified. As the cloning vector, pBluescript vector (Stratagene) and the like are preferable, but various commercially available vectors can be used. When a vector is used for the production of the polypeptide multimer or polypeptide of the present invention, the expression vector is particularly useful. The expression vector is not particularly limited, and may be any vector capable of expressing the polypeptide in vitro, in Escherichia coli, in cultured cells, or in an individual organism. For example, pBEST vector (manufactured by Promega corporation) is preferable for expressing a polypeptide in vitro; when the polypeptide is expressed in E.coli, pET vector (manufactured by Invitrogen) is preferable; when the polypeptide is expressed in cells, pME18S-FL3 vector (GenBank Accession number AB009864) is preferred; when the polypeptide is expressed in an individual organism, the pME18S vector (Mol Cell biol. 8: 466-472(1988)) and the like are preferred. The insertion of DNA into a vector can be carried out by a conventional method, for example, by ligase reaction using restriction enzyme sites (Current protocols in Molecular Biology edge. Ausubel et al (1987) published John Wiley & sons. Section 11.4-11.11).

The host cell is not particularly limited, and various host cells can be used according to the purpose. Examples of the cell for expressing the polypeptide include: bacterial cells (e.g., Streptococcus: (A), (B), (C), (Streptococcus) Staphylococcus (1)Staphylococcus) Escherichia coli (E.coli)E. coli) Streptomyces (I), (II)Streptomyces) Bacillus subtilis (A) and (B)Bacillus subtilis) ); fungal cells (e.g., yeast: (A))Yeast) Aspergillus (A), (B), (C)Aspergillus) ); insect cells (e.g., Drosophila S2: (A)DrosophilaS2), Spodoptera frugiperda SF9(SpodopteraSF 9)); animal cells (e.g., CHO, COS, HeLa, C127, 3T3, BHK, HEK293, Bowes melanoma cells) and plant cells. The vector can be introduced into the host cell by, for example, calcium phosphate precipitation or electroporation (Current protocols in Molecular Biology, Ausubel et al (1987) published by John Wiley&Sons. Section 9.1-9.9), lipofection, microinjection, and the like.

Appropriate secretion signals can be inserted into the polypeptide of interest to cause secretion of the polypeptide expressed in the host cell into the lumen of the endoplasmic reticulum, into the periplasmic space or into the extracellular environment. The signal may be an endogenous signal or a heterologous signal with respect to the target polypeptide.

In constructing the expression vectors for the 1 st to 4 th polypeptides, the DNA encoding the 1 st to 4 th polypeptides may be introduced into a separate vector to form an expression vector. Alternatively, a plurality of DNAs (for example, a DNA encoding the 1 st polypeptide and a DNA encoding the 2 nd polypeptide) among DNAs encoding the 1 st to 4 th polypeptides may be introduced into one vector to form an expression vector. When a plurality of DNAs are introduced into one vector to form an expression vector, the combination of the introduced DNAs encoding polypeptides is not limited.

Recovery of expression products

With respect to recovery of the expression product, the medium is recovered when the polypeptide is secreted into the medium. When the polypeptide is produced intracellularly, the cell is first lysed, after which the polypeptide is recovered.

When the polypeptide is recovered and purified from a recombinant cell culture, the following known methods can be used: ammonium sulfate or ethanol precipitation, acid extraction, anion or cation exchange chromatography, phosphocellulose chromatography, hydrophobic interaction chromatography, affinity chromatography, hydroxylapatite chromatography and lectin chromatography.

In the present invention, protein a affinity chromatography is preferred.

Examples of the column using protein A include, but are not limited to, Hyper D (manufactured by PALL), POROS (manufactured by Applied Biosystems), Sepharose F. (manufactured by GE), and ProSep (manufactured by Millipore). In addition, in protein a affinity chromatography, a resin to which a ligand that mimics the IgG binding ability of (mimic) protein a is bound can also be used. When the mimic protein A is used, the amino acid modification of the present invention can be used to differentiate the binding force of the mimic protein A, thereby separating and purifying the target polypeptide multimer. The mimetic protein A is, for example, mabSelect SuRE (product of GE Healthcare), but is not limited thereto.

The present invention also provides: a polypeptide multimer obtained by the production method or purification method of the present invention.

Furthermore, the present invention provides a polypeptide multimer comprising a1 st polypeptide having antigen-binding activity and a2 nd polypeptide having antigen-binding activity or not, wherein the 1 st polypeptide and the 2 nd polypeptide have different binding force to protein A.

The polypeptide multimer described above can be obtained by the method described in the present specification. In addition, the polypeptide polymer specific structure or properties as described above. The outline thereof is shown below.

The polypeptide polymer of the present invention has a binding force to protein A that is changed compared to that before the amino acid modification. More specifically, the binding force between protein A and either or both of the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity or having no antigen-binding activity is changed. In the polypeptide multimer of the present invention, the binding force between the 1 st polypeptide having antigen-binding activity and protein A is different from the binding force between the 2 nd polypeptide having antigen-binding activity or not having antigen-binding activity and protein A. Thus, in affinity chromatography, the pH of the solvent that elutes the 1 st polypeptide from protein A is different from the pH of the solvent that elutes the 2 nd polypeptide from protein A.

In addition, the 1 st polypeptide and/or the 2 nd polypeptide and one or two of the 3 rd polypeptides can form a polymer.

Accordingly, the present invention relates to a polypeptide multimer comprising a1 st polypeptide having antigen-binding activity, a2 nd polypeptide having antigen-binding activity or not, and one or two 3 rd polypeptides having antigen-binding activity, wherein the 1 st polypeptide and the 2 nd polypeptide have different binding force to protein a. The polypeptide multimer can also be obtained by the method described in the present specification.

Furthermore, the polypeptide multimer may contain the 4 th polypeptide. Either the 1 st polypeptide or the 2 nd polypeptide may form a multimer with the 3 rd polypeptide, while the other may form a multimer with the 4 th polypeptide.

Accordingly, the present invention relates to a polypeptide multimer comprising a1 st polypeptide having antigen-binding activity, a2 nd polypeptide having antigen-binding activity or not, a 3 rd polypeptide having antigen-binding activity and a4 th polypeptide having antigen-binding activity, wherein the 1 st polypeptide and the 2 nd polypeptide differ in binding force to protein a. The polypeptide multimer can also be obtained by the method described in the present specification.

The 1 st polypeptide having an antigen-binding activity and the 2 nd polypeptide having an antigen-binding activity or having no antigen-binding activity described above may comprise an amino acid sequence of an antibody heavy chain constant region or an amino acid sequence of an antibody Fc region. The amino acid sequence of the antibody heavy chain constant region or the antibody Fc region includes, but is not limited to, the amino acid sequence of a constant region derived from human IgG.

The 3 rd polypeptide having an antigen binding activity and the 4 th polypeptide having an antigen binding activity described above may comprise an amino acid sequence of an antibody light chain constant region.

The polypeptide having antigen-binding activity may comprise an amino acid sequence of an antibody variable region (e.g., amino acid sequences of CDR1, CDR2, CDR3, FR1, FR2, FR3, FR 4).

In addition, the 1 st polypeptide having an antigen binding activity and the 2 nd polypeptide having an antigen binding activity or not may include: an antibody heavy chain or an amino acid sequence consisting of an antibody light chain variable region and an antibody heavy chain constant region. The 3 rd polypeptide having antigen binding activity and the 4 th polypeptide having antigen binding activity may comprise: an antibody light chain or an amino acid sequence consisting of an antibody heavy chain variable region and an antibody light chain constant region.

The polypeptide multimer of the invention can be a multispecific antibody. Examples of the multispecific antibody of the present invention include: a bispecific antibody capable of specifically binding to two antigens, but is not limited thereto.

In the polypeptide multimer of the present invention, one or more amino acid residues are modified so that the binding force of the 1 st polypeptide having antigen-binding activity to protein A and the binding force of the 2 nd polypeptide having antigen-binding activity or not having antigen-binding activity to protein A are different (different). Examples of the modification site include, as described above: TLMISR at positions 250-255, VLHQDWLNGK at positions 308-317 and EALHNHY at positions 430-436 in the EU numbering in the Fc region or heavy chain constant region of the antibody; preferably TLMIS at positions 250 and 254, LHQD at positions 309 and 312, LN at positions 314 and 315, E at position 430 and LHNHY at position 432 and 436; further preferably LMIS at positions 251 and 254, LHQ at positions 309 and 311, L at position 314, LHNH at positions 432 and 435; in particular, but not limited to, those amino acid residues of MIS at position 252 and 254, L at position 309, Q at position 311, and NHY at position 434 and 436. Preferred modification positions for the amino acid modification in the antibody heavy chain variable region include FR1, CDR2, and FR 3.

More specifically, as the polypeptide multimer of the present invention, the following polypeptide multimers can be exemplified, but not limited thereto: wherein, in the polypeptide of any one of the 1 st polypeptide having an antigen-binding activity and the 2 nd polypeptide having an antigen-binding activity or having no antigen-binding activity, the amino acid residue at position 435 in the EU numbering system in the amino acid sequence of the antibody Fc region or the antibody heavy chain constant region is histidine or arginine;

in the other polypeptide, the amino acid residue at position 435 in the EU numbering in the amino acid sequence of the antibody Fc region or the antibody heavy chain constant region is different from that of the one polypeptide.

The polypeptide multimer of the present invention includes, but is not limited to, the following polypeptide multimers: wherein, in the polypeptide of any one of the 1 st polypeptide having an antigen-binding activity and the 2 nd polypeptide having an antigen-binding activity or having no antigen-binding activity, the amino acid residue at position 435 in the EU numbering system in the amino acid sequence of the antibody heavy chain constant region is histidine;

in another polypeptide, the amino acid residue at position 435 in the EU numbering system in the amino acid sequence of the heavy chain constant region of the antibody is arginine.

Furthermore, as the invention of the 1 st and 2 nd polypeptide polymer, can exemplify the following polypeptide polymer, but not limited to these.

(1) A polypeptide multimer comprising a1 st polypeptide or a2 nd polypeptide comprising the following amino acid sequence: amino acid residues at positions 435 and 436 in the EU numbering in the amino acid sequence of the antibody heavy chain constant region from human IgG are modified into amino acid sequences of histidine (His) and tyrosine (Tyr), respectively.

Examples of such polypeptide multimers include: polypeptide multimers comprising the 1 st or 2 nd polypeptide comprising the amino acid sequence set forth in SEQ ID NO 9, SEQ ID NO 11, SEQ ID NO 13, or SEQ ID NO 15, but are not limited thereto.

(2) A polypeptide multimer comprising a1 st polypeptide or a2 nd polypeptide comprising the following amino acid sequence: amino acid residues at positions 435 and 436 in the EU numbering in the amino acid sequence of the antibody heavy chain constant region derived from human IgG are modified to amino acid sequences of arginine (Arg) and phenylalanine (Phe), respectively.

Examples of such polypeptide multimers include: polypeptide multimers comprising the 1 st or 2 nd polypeptide comprising the amino acid sequence set forth in SEQ ID NO 10 or SEQ ID NO 12, but are not limited thereto.

(3) A polypeptide multimer comprising a1 st polypeptide or a2 nd polypeptide comprising the following amino acid sequence: amino acid residues at positions 435 and 436 in the EU numbering in the amino acid sequence of the antibody heavy chain constant region derived from human IgG are modified to amino acid sequences of arginine (Arg) and tyrosine (Tyr), respectively.

Examples of such polypeptide multimers include: polypeptide multimers comprising the 1 st or 2 nd polypeptide comprising the amino acid sequence set forth in SEQ ID NO. 14, but are not limited thereto.

(4) A polypeptide multimer comprising the following polypeptides: a polypeptide comprising an amino acid sequence in which amino acid residues at positions 435 and 436 in the EU numbering in the amino acid sequence of an antibody heavy chain constant region derived from human IgG are modified into histidine (His) and tyrosine (Tyr), respectively, in the polypeptide of any one of the 1 st polypeptide or the 2 nd polypeptide; and a polypeptide in which amino acid residues at positions 435 and 436 in the EU numbering system included in the amino acid sequence of the heavy chain constant region of the antibody are modified to amino acid sequences of arginine (Arg) and phenylalanine (Phe), respectively.

Examples of such polypeptide multimers include: polypeptide multimers comprising a1 st polypeptide comprising the amino acid sequence of SEQ ID NO 9, SEQ ID NO 11, SEQ ID NO 13 or SEQ ID NO 15 and a2 nd polypeptide comprising the amino acid sequence of SEQ ID NO 10 or SEQ ID NO 12, but are not limited thereto.

(5) A polypeptide multimer comprising the following polypeptides: a polypeptide comprising an amino acid sequence in which amino acid residues at positions 435 and 436 in the EU numbering in the amino acid sequence of an antibody heavy chain constant region derived from human IgG are modified into histidine (His) and tyrosine (Tyr), respectively, in the polypeptide of any one of the 1 st polypeptide or the 2 nd polypeptide; and a polypeptide comprising an amino acid sequence in which amino acid residues at positions 435 and 436 in the EU numbering system in the amino acid sequence of the heavy chain constant region of the antibody are modified to arginine (Arg) and tyrosine (Tyr), respectively, in the other polypeptide.

Examples of such polypeptide multimers include: polypeptide multimers comprising a1 st polypeptide comprising the amino acid sequence of SEQ ID NO 9, SEQ ID NO 11, SEQ ID NO 13 or SEQ ID NO 15 and a2 nd polypeptide comprising the amino acid sequence of SEQ ID NO 14, but are not limited thereto.

(6) A polypeptide multimer comprising the following polypeptides: a polypeptide comprising an amino acid sequence in which amino acid residues at positions 435 and 436 in the amino acid sequence of an antibody heavy chain constant region derived from human IgG are modified to arginine (Arg) and phenylalanine (Phe), respectively, in any one of the polypeptides 1 and 2; and a polypeptide comprising an amino acid sequence in which amino acid residues at positions 435 and 436 in the EU numbering system in the amino acid sequence of the heavy chain constant region of the antibody are modified to arginine (Arg) and tyrosine (Tyr), respectively, in the other polypeptide.

Examples of such polypeptide multimers include: polypeptide multimers comprising, but not limited to, a1 st polypeptide comprising the amino acid sequence of SEQ ID NO. 10 or SEQ ID NO. 12 and a2 nd polypeptide comprising the amino acid sequence of SEQ ID NO. 14.

The 1 st and 2 nd polypeptides may further comprise an antibody heavy chain variable region. The polypeptide multimers of (1) to (6) above may have the 3 rd polypeptide and/or the 4 th polypeptide.

The present invention also provides a polypeptide mutant comprising a polypeptide having a mutation at any one of amino acid residues 435 to 436 in the EU numbering system. Examples of such polypeptide mutants include: polypeptide mutants including the polypeptides listed in the examples are not limited thereto.

The present invention also provides a nucleic acid encoding a polypeptide (a polypeptide having antigen-binding activity) constituting the polypeptide multimer of the present invention. The invention also provides a vector carrying the nucleic acid.

The invention also provides a host cell comprising the nucleic acid or vector described above. The host cell is not particularly limited, and examples thereof include Escherichia coli, various plant cells, and animal cells. The host cell can be used, for example, as a production system for preparing or expressing a polypeptide multimer or polypeptide of the invention. A production system for producing a polypeptide multimer or polypeptide comprising: in vitro (in vitro) and in vivo (in vivo) production systems. Examples of in vitro production systems are: production systems using eukaryotic cells and production systems using prokaryotic cells.

Examples of eukaryotic cells that can be used as host cells are: animal cells, plant cells, fungal cells. The animal cells include: mammalian cells, such as CHO (J. exp. Med. (1995) 108: 945), COS, HEK293, 3T3, myeloma cells, BHK (baby hamster kidney), HeLa, Vero, etc.; amphibian cells, e.g. Xenopus oocytes (Xenopus laevis oocytes) (Valle et al, Nature (1981) 291: 338-340); and insect cells such as Sf9, Sf21, Tn 5. CHO-DG44, CHO-DX11B, COS7 cells, HEK293 cells, BHK cells are suitably used in the expression of the polypeptide multimers or polypeptides of the present invention. Among animal cells, CHO cells are particularly preferable for the purpose of large-scale expression. The introduction of the vector into the host cell can be carried out, for example, by the following method: the calcium phosphate method, the DEAE-dextran method, the method using cationic liposome DOTAP (manufactured by Boehringer Mannheim), the electroporation method, the lipofection method, and the like.

Plant cells such as tobacco-derived cells and duckweed (Lemna minor) are known as protein production systems, and the polypeptide multimer or polypeptide of the present invention can be produced by culturing the cells using callusA peptide. Protein expression systems using fungal cells which can be used as hosts for producing polypeptide multimers or polypeptides of the invention are well known, such as: yeasts, e.g. yeasts: (Saccharomyces) Genus cells (Saccharomyces cerevisiae: (Saccharomyces cerevisiae) Schizosaccharomyces pombe (Schizosaccharomyces pombe)Saccharmyces pombe) Etc.); and filamentous fungi, e.g. Aspergillus (A)Aspergillus) Cells of (A. niger)Aspergillus niger) Etc.).

When prokaryotic cells are used, there are production systems using bacterial cells. As bacterial cells, it is known that Escherichia coli (E.coli) is usedE.coli) In addition, there are production systems using Bacillus subtilis, and these bacterial cells can be used to produce the polypeptide multimers or polypeptides of the present invention.

When the host cell of the present invention is used to produce a polypeptide multimer or polypeptide, the host cell transformed with an expression vector comprising a polynucleotide encoding the polypeptide multimer or polypeptide of the present invention may be cultured to express the polynucleotide. The culture can be carried out according to a known method. For example, when animal cells are used as hosts, DMEM, MEM, RPMI1640, or IMDM can be used as a culture medium. In this case, a serum supplement such as FBS or Fetal Calf Serum (FCS) may be used in combination, or the cells may be cultured in serum-free culture. The pH during the culture is preferably about 6 to 8. Usually, the culture is carried out at about 30 to 40 ℃ for about 15 to 200 hours, and the medium may be exchanged or aerated or stirred as necessary.

On the other hand, examples of systems for producing a polypeptide in vivo include: production systems using animals or production systems using plants. The target polynucleotide is introduced into these animals or plants, and the polypeptide is produced in vivo in the animals or plants and recovered. The "host" in the present invention includes the above-mentioned animals and plants.

When animals are used, there are production systems using mammals and insects. As the mammal, goat, pig, sheep, mouse, cow, etc. can be used (Vicki Glaser, SPECTRUM Biotechnology Applications (1993)). In addition, when mammals are used, transgenic animals may be used.

For example, a polynucleotide encoding a polypeptide multimer or polypeptide of the present invention is prepared as a fusion gene with a gene encoding a polypeptide inherently produced in milk such as goat β -casein. Then, a polynucleotide fragment comprising the fusion gene is injected into a goat embryo, and the embryo is transferred into a female goat. The goat which received the embryo gives rise to a transgenic goat from which the antibody of interest can be obtained or from the milk produced in its offspring. In order to increase the amount of antibody-containing milk produced by the transgenic goat, the above transgenic goat may be administered with an appropriate hormone (Ebert et al, Bio/Technology (1994) 12: 699-).

As an insect producing the polypeptide multimer or polypeptide of the present invention, for example, silkworm can be used. When silkworms are used, a polypeptide multimer or polypeptide of interest can be obtained from the body fluid of the silkworms by infecting the silkworms with a baculovirus into which a polynucleotide encoding the polypeptide multimer or polypeptide of interest is inserted (Susumu et al, Nature (1985) 315: 592-4).

Furthermore, when a plant is used for producing the polypeptide multimer or polypeptide of the present invention, for example, tobacco (tabaco) can be used. When tobacco is used, a polynucleotide encoding a polypeptide multimer or polypeptide of interest is inserted into a plant expression vector such as pMON 530, and the vector is introduced into Agrobacterium tumefaciens (II)Agrobacterium tumefaciens) And the like. Infecting tobacco such as tobacco with the bacterium (Nicotiana tabacum) The desired polypeptide multimer or polypeptide can be obtained from the leaves of this tobacco (Ma et al, eur. j. Immunol. (1994) 24: 131-8). Furthermore, the same bacteria were infected with duckweed (Lemna minor) and the desired polypeptide multimers or polypeptides could be obtained from the cells of duckweed after cloning (Cox KM et al. nat. Biotechnol. 2006 Dec; 24(12): 1591-.

The polypeptide multimer or polypeptide thus obtained can be isolated from the inside or outside of the host cell (medium, milk, etc.) and purified into a substantially pure and homogeneous polypeptide multimer or polypeptide. The polypeptide multimer or the polypeptide can be isolated and purified by the isolation and purification method used for the purification of ordinary polypeptides, but is not limited thereto. For example, the antibody can be separated and purified by appropriately selecting and combining a chromatography column, a filter, ultrafiltration, salting out, solvent precipitation, solvent extraction, distillation, immunoprecipitation, SDS-polyacrylamide gel electrophoresis, isoelectric point electrophoresis, dialysis, recrystallization, and the like.

Examples of chromatography include: affinity chromatography, ion exchange chromatography, hydrophobic chromatography, gel filtration, reverse phase chromatography, adsorption chromatography, etc. (stratgies for Protein Purification and Characterization, a Laboratory courtesy, manual Ed Daniel r. Marshark et al (1996) Cold Spring Harbor Laboratory Press). The chromatography can be performed by liquid chromatography, for example, HPLC, FPLC and the like. The columns used for affinity chromatography are: protein A column and protein G column. Examples of protein a columns are: hyper D, POROS, Sepharose F.F. (Pharmacia), etc., but are not limited thereto.

If necessary, an appropriate protein-modifying enzyme may be allowed to act on the polypeptide multimer or polypeptide to arbitrarily modify or partially remove the peptide, either before or after purification of the polypeptide multimer or polypeptide. As the protein-modifying enzyme, for example, used are: trypsin, chymotrypsin, lysyl endopeptidase, protein kinase, glycosidase, and the like.

Still another preferred embodiment of the present invention is a method for preparing the polypeptide multimer or polypeptide of the present invention, the method comprising the steps of: the host cell of the invention is cultured in the manner described above and the polypeptide is recovered from the cell culture.

The present invention also relates to a pharmaceutical composition (medicament) comprising: the polypeptide polymer or polypeptide of the invention and a pharmaceutically acceptable carrier. In the present invention, the pharmaceutical composition generally refers to a drug for the treatment or prevention of a disease or for examination or diagnosis.

The pharmaceutical composition of the present invention can be formulated according to methods known to those skilled in the art. For example, a sterile solution or suspension with water or a pharmaceutically acceptable liquid other than water can be prepared for parenteral administration in the form of an injection. For example, a pharmacologically acceptable carrier or vehicle, specifically, sterile water or physiological saline, vegetable oil, emulsifier, suspending agent, surfactant, stabilizer, flavoring agent, excipient, vehicle (vehicle), preservative, binder, and the like may be suitably combined and mixed in a unit dosage form required for generally accepted pharmaceutical practice to prepare a preparation. The amount of the active ingredient in the above preparation is set to an appropriate amount to obtain the indicated range.

The sterile composition for injection can be prepared according to a usual preparation procedure using a vehicle such as distilled water for injection.

Examples of the aqueous solution for injection include: physiological saline and isotonic solutions containing glucose or other adjuvants (e.g., D-sorbitol, D-mannose, D-mannitol, sodium chloride). Suitable cosolvents such as alcohols (ethanol, etc.), polyols (propylene glycol, polyethylene glycol, etc.), nonionic surfactants (Tween 80(TM), HCO-50, etc.) may be used in combination.

As the oily liquid, sesame oil and soybean oil are mentioned, and benzyl benzoate and/or benzyl alcohol may be used in combination as a cosolvent. It can also be mixed with buffers (such as phosphate buffer and sodium acetate buffer), demulcents (such as procaine hydrochloride), stabilizers (such as benzyl alcohol and phenol), and antioxidants. The prepared injection is usually filled in an appropriate ampoule.

The pharmaceutical compositions of the present invention are preferably administered parenterally. For example, the composition can be prepared into an injection, a nasal administration, a pulmonary administration, or a transdermal administration. For example, systemic or local administration can be carried out by intravenous injection, intramuscular injection, intraperitoneal injection, subcutaneous injection, and the like.

The method of administration may be appropriately selected depending on the age and symptoms of the patient. The pharmaceutical composition containing the polypeptide multimer or polypeptide or polynucleotide encoding the same can be administered in an amount, for example, set to: each time, the weight of the composition is 0.0001-1000 mg per kg body weight. Alternatively, the amount to be administered may be, for example, 0.001 to 100000 mg per patient, but the present invention is not necessarily limited to the above values. The dose and the administration method vary depending on the body weight, age, symptoms, and the like of the patient, and one skilled in the art can set an appropriate dose and administration method in consideration of the above conditions.

If necessary, the multispecific antibody of the present invention may be combined with other pharmaceutical ingredients to prepare a formulation.

All prior art documents cited in this specification are incorporated herein by reference.

Examples

The present invention will be described more specifically with reference to examples. However, the technical scope of the present invention is not limited to these examples.

[ Examples 1] Preparation of antibody Gene expression vector and expression of Each antibody

As the antibody H chain variable region, the following H chain variable region was used. Q153 (H chain variable region of anti-human F.IX antibody, SEQ ID NO: 1), Q407 (H chain variable region of anti-human F.IX antibody, SEQ ID NO: 2), J142 (H chain variable region of anti-human F.X antibody, SEQ ID NO: 3), J300 (H chain variable region of anti-human F.X antibody, SEQ ID NO: 4), MRA-VH (H chain variable region of anti-human interleukin-6 receptor antibody, SEQ ID NO: 5).

As the antibody L chain, the following L chain was used. L180-k (anti-human F.IX antibody/anti-human F.X antibody common L chain, SEQ ID NO: 6), L210-k (anti-human F.IX antibody/anti-human F.X antibody common L chain, SEQ ID NO: 7), MRA-k (anti-human interleukin-6 receptor antibody common L chain, SEQ ID NO: 8).

As the antibody H chain constant region, the following constant regions were used: g4d (SEQ ID NO: 9) constructed by introducing a mutation to IgG4 in which Ser at position 228 in the EU numbering system is substituted with Pro and removing Gly and Lys at the C-terminal; z72(SEQ ID NO: 10) constructed by introducing a mutation of Arg for His at position 435 in the EU numbering system, a mutation of Phe for Tyr at position 436 in the EU numbering system, and a mutation of Pro for Leu at position 445 in the EU numbering system into G4 d; z7(SEQ ID NO: 11) constructed by introducing a mutation of substituting Glu at position 356 in the EU numbering system with Lys into G4 d; z73(SEQ ID NO: 12) constructed by introducing a mutation to z72 wherein Glu is substituted by Lys at position 439 in the EU numbering system; z106(SEQ ID NO: 13) constructed by introducing a mutation wherein Lys at position 196 in the EU numbering system is substituted with Gln, a mutation wherein Phe at position 296 in the EU numbering system is substituted with Tyr, and a mutation wherein Arg at position 409 in the EU numbering system is substituted with Lys into z 7; z107(SEQ ID NO: 14) constructed by introducing a mutation wherein Lys at position 196 in the EU numbering system is substituted by Gln, a mutation wherein Phe at position 296 in the EU numbering system is substituted by Tyr, a mutation wherein Arg at position 409 in the EU numbering system is substituted by Lys, and a mutation wherein Phe at position 436 in the EU numbering system is substituted by Tyr into z 73; g1d (SEQ ID NO: 15) constructed by removing Gly and Lys from the C-terminal of IgG 1. The substitution OF Glu at position 356 in EU numbering to Lys and the substitution OF Lys at position 439 in EU numbering to Glu are due to the heterologous molecules used to efficiently form each H chain when generating heterologous antibodies ((WO 2006/106905) PROCESS FOR PRODUCTION OF POLYPEPTIDE BY REGIONATION OF ASSEMBLY).

Anti-human f.ix antibody H chain genes Q153-G4d or Q153-z7 were prepared by linking G4d or z7 downstream of Q153. The anti-human f.ix antibody H chain gene Q407-z106 was prepared by linking z106 to the downstream of Q407. Anti-human f. X antibody H chain genes J142-G4d, J142-z72, or J142-z73 were made by linking G4d, z72, or z73 downstream of J142. The anti-human F.X antibody H chain gene J300-z107 was prepared by ligating z107 to the downstream of J300. The anti-human interleukin-6 receptor antibody H chain gene MRA-G1d, MRA-z106 or MRA-z107 was prepared by linking G1d, z106 or z107 to the downstream of MRA-VH.

Each antibody gene (Q153-G4d, Q153-z7, Q407-z106, J142-G4d, J142-z72, J142-z73, J300-z106, MRA-G1d, MRA-z106, MRA-z107, L180-k, L210-k, MRA-k) was inserted into an animal cell expression vector.

The following antibodies were transfected into FreeStyle293 cells (invitrogen) using the prepared expression vector, and were overexpressed. The transfected multiple antibody genes are arranged and are represented as antibody names as follows:

MRA-G1d/MRA-k；

MRA-z106/MRA-z107/MRA-k；

Q153-G4d/J142-G4d/L180-k；

Q153-G4d/J142-z72/L180-k；

Q153-z7/J142-z73/L180-k；

Q407-z106/J300-z107/L210-k。

[ examples 2] Protein A Investigation of elution conditions for affinity chromatography

Elution conditions for protein A affinity chromatography were examined by using, as a sample, a FreeStyle293 cell culture solution (hereinafter abbreviated as CM) obtained by transiently expressing Q153-G4d/J142-G4d/L180-k and Q153-G4d/J142-z72/L180-k in a single manner. CM filtered through a filter of 0.22 μm was loaded on a rProtein A Sepharose Fast Flow column (GE Healthcare) equilibrated with D-PBS, and washing 1, washing 2, and elution 1 to 5 shown in Table 1 were performed in stages. The loading of CM was adjusted so that the amount of antibody loaded on the column reached 20mg/mL resin (resin). And separating and collecting the eluted components under each condition, and identifying the components contained in each eluted component by cation exchange chromatography analysis. The following samples were used for the control: each CM was loaded on rProtein G Sepharose Fast Flow resin (GE Healthcare), and a sample was purified by batch elution. Since protein G binds to the Fab portion of the antibody, all antibodies present in CM can be purified by using protein G regardless of the affinity of protein a (bispecific antibody (heterologous antibody) constructed by heteroassociation of two target H chains, and monospecific homologous antibody constructed by homologously association of one H chain of an impurity).

[ Table 1]

Cation exchange chromatography analysis was performed on each eluted fraction (elution 1-5) of CM-expressing Q153-G4d/J142-G4d/L180-k and Q153-G4d/J142-z72/L180-k protein A columns. The method is characterized in that: as the fraction changed from elution 1 to elution 5, that is, as the pH of the solvent used in elution decreased, the antibody components contained in the fractions of Q153-G4d/J142-G4d/L180-k changed from the homologous antibody J142-G4d/L180-k to the sequence of the heterologous antibody Q153-G4d/J142-G4d/L180-k and the homologous antibody Q153-G4 d/L180-k. The elution order is believed to depend on the strength of the binding to protein a. That is, the cognate antibody Q153-G4d/L180-k, which remains bound until reaching low pH, binds strongly to protein A as compared to the cognate J142-G4d/L180-k (cognate antibody against FX) which elutes at high pH. It was confirmed that the variable region J142 is a sequence not binding to protein A. That is, the cognate J142-G4d/L180-k (cognate antibody against FX) has two binding sites for protein A, the heterologous antibody Q153-G4d/J142-G4d/L180-k has 3 binding sites for protein A, and the homologous antibody Q153-G4d/L180-k (cognate antibody against FIX) has 4 binding sites for protein A. Therefore, it is found that: the greater the number of binding sites to protein a, the stronger the binding to protein a and the lower the pH necessary for elution.

On the other hand, it was found that: in Q153-G4d/J142-z72/L180-k, the antibody components contained in the respective fractions become the order of the hetero antibody Q153-G4d/J142-z72/L180-k followed by the homo antibody Q153-G4d/L180-k as changing from elution 1 fraction to elution 5 fraction. Since almost no cognate antibody J142-z72/L180-k (cognate antibody against FX) was detected in each eluted fraction, it was suggested that it lacked the ability to bind to protein A. It is considered that the protein was not bound to protein A only because of the mutation introduced into J142-z72 in which His at position 435 in the EU numbering system was substituted with Arg. The cognate antibody J142-z72/L180-k (cognate antibody against FX) has no binding site for protein A, the heterologous antibody Q153-G4d/J142-z72/L180-k has two binding sites for protein A, and the cognate antibody Q153-G4d/L180-k (cognate antibody against FIX) has 4 binding sites for protein A. Since cognate antibody J142-z72/L180-k (cognate antibody against FX) did not bind to protein A, no cognate antibody J142-z72/L180-k was detected in each eluted fraction. It also implies that: the possibility of separating both the heterologous and the homologous antibodies Q153-G4d/L180-k (homologous antibodies against FIX) at pH3.6 and below, Q153-G4d/J142-G4d/L180-k and Q153-G4d/J142-z72/L180-k, respectively.

[ Examples 3] Using proteins A Chromatographic separation and purification of heterologous antibody

CM of antibodies shown below was used as a sample:

・Q153-G4d/J142-G4d/L180-k；

・Q153-G4d/J142-z72/L180-k；

・Q153-z7/J142-z73/L180-k；

・Q407-z106/J300-z107/L210-k。

CM filtered through a filter of 0.22 μm was loaded on an rProtein A Sepharose Fast Flow column (GE Healthcare) equilibrated with D-PBS, and washing 1, washing 2, elution 1, and elution 2 shown in Table 2 were performed (only elution 1 was performed for Q407-z106/J300-z 107/L210-k). The elution conditions are referenced to the results of example 2. The loading of CM was adjusted to achieve a loading of 20mg/mL of resin. And separating and collecting the eluted components under various conditions, and identifying the components contained in the eluted components through cation exchange chromatography analysis. In the control, as in example 2, a sample was used in which each CM was loaded on rProtein G Sepharose Fast Flow resin (GE Healthcare) and was purified by batch elution.

[ Table 2]

The results of the cation exchange chromatography analysis of each eluted fraction are shown in Table 3 below. The area value of the elution peak is expressed as a percentage. In antibodies other than Q153-G4d/J142-G4d/L180-k, almost no homologous antibody against FX was detected in all eluted fractions. It was found that not only the homologous antibody J142-z72 (homologous antibody against FX) shown in example 2 but also the homologous antibodies J142-z73 and J300-z107 (homologous antibody against FX) became not bound to protein A. This is believed to be due to: the reason is that the binding to protein A in the anti-FX homologous antibody is lost by a mutation introduced into the H chain constant region of the anti-FX antibody, which replaces His at position 435 in the EU numbering system with Arg. The hetero-antibody as the bispecific antibody of interest was mostly detected in fraction eluted 1, and a trace amount of the homologous antibody against FIX was also detected in fraction eluted 1, but most of the homologous antibody against FIX was eluted by fraction eluted 2. The fraction of bispecific antibody of interest, i.e.heterologous antibody, of the eluted fraction at pH3.6 was greatly increased in Q153-z7/J142-z73/L180-k and Q407-z106/J300-z107/L210-k compared to Q153-G4d/J142-z 72/L180-k. Thus, it can be seen that: by introducing a mutation for replacing His at position 435 in the EU numbering system with Arg and a mutation for replacing Glu at position 356 in the EU numbering system with Lys and a mutation for replacing Lys at position 439 in the EU numbering system with Glu for efficiently forming a hetero molecule of each H chain, a hetero antibody as a target bispecific antibody can be purified with a purity of 98% or more only by a protein A purification step.

From the above results, it was found that: by using the difference in the number of protein A binding sites between the homologous antibody and the heterologous antibody, the heterologous antibody can be isolated and purified with high purity and efficiency by using only a protein A chromatography step.

[ Table 3]

[ Table 4]

[ Table 5]

[ Table 6]

。

[ Examples 4] Human being FcRn Pharmacokinetic evaluation in transgenic mice

The study of example 3 found that: by using z106(SEQ ID NO: 13) and z107(SEQ ID NO: 14) in each H chain constant region of the bispecific antibody, the heterologous antibody as the bispecific antibody of interest can be purified with a purity of 98% or more using only the protein A step. On the other hand, since protein A and human FcRn recognize the same site of IgG antibody (J Immunol. 2000164 (10): 5313-8.), the binding to human FcRn is likely to be lost by losing the binding to protein A. Indeed, although: as a method for purifying a bispecific antibody to 95% purity using protein a, a method using the H chain of rat IgG2b (which does not bind to protein a) has been reported, but the bispecific antibody purified by this method, namely, castorezumab, has a half-life in a human body of about 2.1 days and is extremely short in comparison with the half-life of ordinary human IgG1 of 2 to 3 weeks (non-patent document 2). Thus, the pharmacokinetics of the antibody having z106(SEQ ID NO: 13) and z107(SEQ ID NO: 14) as constant regions used in example 3 was evaluated.

As a pharmacokinetic experiment for predicting half-life in humans, a human FcRn transgenic mouse (b6.mfcrn-/-. hFcRn Tg line 276 +/+ mouse, Jackson Laboratories) was used for pharmacokinetic evaluation as follows. Mice were given a single intravenous administration of 1mg/kg of MRA-G1d/MRA-k (hereinafter referred to as MRA-IgG1) having IgG1 as a constant region and MRA-z106/MRA-z107/MRA-k (hereinafter referred to as MRA-z106/z107) having z106/z107 as a constant region, respectively, and blood was collected at appropriate times. The collected blood was immediately centrifuged at 4 ℃ at 15,000 rpm for 15 minutes to obtain plasma. The separated plasma was stored in a refrigerator set to-20 ℃ or lower until the measurement was performed. Plasma concentrations were determined by ELISA.

As shown in FIG. 1, the results of the evaluation of plasma retention of MRA-IgG1 and MRA-z106/z107k in human FcRn transgenic mice show that MRA-z106/z107 has plasma retention higher than or equal to that of MRA-IgG 1. It was thus found that: the z106/z107, which is a constant region for purifying or producing a heterologous antibody with high purity and efficiency, can be purified or produced only by the protein A purification step, and has plasma retention equal to or higher than that of human IgG 1.

[ Examples 5] Preparation of antibody Gene expression vector and expression of Each antibody

As the antibody H chain variable region, the following H chain variable regions were used:

seed Q499 (H chain variable region of anti-human F.IX antibody, SEQ ID NO: 16);

zizan J339 (H chain variable region of anti-human F.X antibody, SEQ ID NO: 17).

As the antibody L chain, the following L chains were used:

seed L377-k (antihuman F.IX antibody/antihuman F.X antibody common L chain, SEQ ID NO: 18).

As the antibody H chain constant region, the following H chain constant regions were used:

as a result of introducing a mutation wherein Leu at position 405 in the EU numbering system is substituted with Phe into z106(SEQ ID NO: 19) described in seed example 1;

seed introduction into z118 z121(SEQ ID NO: 20) constructed by introducing a mutation in which Arg is substituted by His at position 435 in the EU numbering system;

as a result of introduction of Z119(SEQ ID NO: 21) into Z118, a mutation wherein Lys at position 356 in the EU numbering system was replaced with Glu, and a mutation wherein Glu at position 439 in the EU numbering system was replaced with Lys were introduced.

Anti-human F.IX antibody H chain gene Q499-z118 or Q499-z121 was made by ligating z118 or z121 downstream of Q499. An anti-human F.X antibody H chain gene J339-z119 was prepared by ligating z119 downstream of J339.

Each antibody gene (Q499-z118, Q499-z121, J339-z119, L377-k) was inserted into an animal cell expression vector.

The following antibodies were transfected into FreeStyle293 cells (invitrogen) using the prepared expression vector to overexpress them. The transfected multiple antibody genes are arranged and are represented as antibody names as follows:

Q499-z118/J339-z119/L377-k；

Q499-z121/J339-z119/L377-k。

the two antibodies differ only in the amino acid at position 435 in the EU numbering of the H chain of the anti-human f.ix antibody. The 435-position in z118 is His, which has the ability to bind to protein A, but the 435-position in z121 is Arg, which is known from example 2 to have no binding to protein A. Since Q499 is predicted to bind to protein A by its sequence, there are two binding sites for the cognate J339-z119/L377-k (cognate antibody against FX) to protein A, 3 binding sites for the heterologous antibody Q499-z118/J339-z119/L377-k to protein A, and 4 binding sites for the cognate antibody Q499-z118/L377-k (cognate antibody against FIX) to protein A. On the other hand, there are two binding sites for the homologue J339-z119/L377-k introduced with the protein A non-binding modified Q499-z121/J339-z119/L377-k to protein A, two binding sites for the heterologous antibody Q499-z121/J339-z119/L377-k to protein A, and two binding sites for the homologous antibody Q499-z121/L377-k to protein A. That is, even when a protein A non-binding modification (for example, a modification in which Arg is substituted for the amino acid at position 435 in the EU numbering system) is introduced only into the H chain binding to protein A in the variable region, the effect of efficiently separating and purifying a heterologous antibody with high purity by only the protein A purification step cannot be obtained. However, by using a modified protein A which does not bind to Q499, i.e., MabSelct Sure (GE Healthcare), the effect of non-binding modification of protein A can be seen. The MabSelect SuRe is a chromatographic carrier for antibody purification developed to meet industrial requirements, and the ligand is recombinant protein a which maintains alkali resistance in genetic engineering. Since it has high pH stability, it can be washed efficiently with NaOH at a low cost. It also has the characteristic of not binding to the heavy chain variable region of the VH3 subclass of Q499 et al. The cognate of Q499-z118/J339-z119/L377-k has two binding sites for MabSelect Sure, the heterologous antibody Q499-z118/J339-z119/L377-k has two binding sites for MabSelect Sure, and the homologous antibody Q499-z118/L377-k has two binding sites for MabSelect Sure. On the other hand, the homology of Q499-z121/J339-z119/L377-k has two binding sites for the homology of J339-z119/L377-k to MabSelect Sure, the heterologous antibody Q499-z121/J339-z119/L377-k has one binding site for the heterologous antibody Q499-z121/L377-k to MabSelect Sure, and the homologous antibody Q499-z121/L377-k to MabSelect Sure has a binding site of 0. Namely, the following is considered: by combining a modified protein A such as MabSelect Sure that does not bind to the antibody variable region with a protein A-non-binding modification, a heterologous antibody can be isolated and purified with high purity and efficiency only by a protein A purification step, regardless of the protein A binding ability of the heavy chain variable region.

[ Examples 6] By modifying proteins A Affinity chromatography separation and purification of heterologous antibody

Modified protein A affinity chromatography was performed for CM expressing Q499-z118/J339-z119/L377-k and Q499-z121/J339-z 119/L377-k. CM filtered through a 0.22 μm filter was loaded on a Mab Select SuRe column (GE Healthcare) equilibrated with D-PBS, and washed 1, 2 and eluted as shown in Table 7. The Mab Select SuRe is composed of a structure in which the B domain is genetically engineered within 5 domains (a to E) of recombinant protein a having IgG binding ability so that the modified B domain forms a tetramer. Mab Select SuRe lacks the ability to bind to the variable region of the antibody, and thus has the characteristic of eluting the antibody under milder conditions than the common recombinant protein a. Moreover, the alkali resistance is enhanced, and the resin can be fixedly cleaned by 0.1-0.5M NaOH, so that the resin is more suitable for production. In this example, 50mM acetic acid (pH 3.0 or so as measured without pH adjustment) was used as shown in Table 7, instead of stepwise elution at pH3.6 and pH2.7 as in example 3. Separating and collecting the eluted components, and identifying the components contained in each eluted component by cation exchange chromatography analysis. In the control, as in example 2, a sample was used in which each CM was loaded on rProtein G Sepharose Fast Flow resin (GE Healthcare) and was purified by batch elution.

The protein a eluate fraction was then subjected to ion exchange chromatography. Protein A eluate fractions diluted 3-fold with an equilibration buffer after neutralization with 1.5M Tris-HCl (pH7.4) were loaded onto an SP Sepharose High Performance column (GE Healthcare) equilibrated with an equilibration buffer (20mM NaPhosphate buffer, pH 6.0). The antibody bound to the column was eluted by a gradient of NaCl concentration from 50mM to 350mM in 25 Column Volumes (CV). The fractions eluted from the hetero-antibody were subjected to gel filtration chromatography using superdex200 to purify, and the monomer fractions were separated and collected. This was used for pharmacokinetic evaluation in human FcRn transgenic mice as described in example 7.

[ Table 7]

The results of cation exchange chromatography analysis of each eluted fraction are shown in tables 8 and 9. In Q499-z118/J339-z119/L377-k, as shown in Table 8, the ratio of each component contained in the eluted fraction was not much changed from that of the control. This is believed to be due to: all 3 modified protein A binding sites were two for J339-z119/L377-k (homologous antibody against F.X), Q499-z118/L377-k (homologous antibody against F.IX), Q499-z118/J339-z119/L377-k (heterologous antibody), so no difference was generated in binding or dissociation in the protein A purification step.

On the other hand, as shown in Table 9, the ratio of Q499-z121/L377-k (homologous antibody against F.IX) contained in the eluted fraction of Q499-z121/J339-z119/L377-k was significantly reduced as compared with the control. In contrast, with the decrease in Q499-z121/L377-k, the ratios of the elution components of J339-z119/L377-k (homologous antibody against F.X) and Q499-z121/J339-z119/L377-k (heterologous antibody) increased relatively compared to the control. This is believed to be due to: j339-z119/L377-k (homologous antibody against F.X) binds to modified protein A at two sites, Q499-z121/J339-z119/L377-k (heterologous antibody) binds to modified protein A at one site, Q499-z121/L377-k (homologous antibody against F.IX) binds to modified protein A at 0 site, and almost all Q499-z121/L377-k have passed without binding to modified protein A.

From the above results, it was found that: even in antibodies having a variable region having protein A binding ability, by combining a protein A non-binding modification with a modified protein A, it is possible to significantly reduce one homologous antibody and improve the purity of a heterologous antibody only by a protein A purification step.

[ Table 8]

[ Table 9]

。

Examples 7] Human being FcRn Pharmacokinetic evaluation in transgenic mice

Pharmacokinetic evaluations were performed on Q499-z118/J339-z119/L377-k and Q499-z121/J339-z119/L377-k prepared in example 6.

As shown in fig. 2, since protein a and human FcRn recognize the same site of IgG antibody (J immunol. 2000164 (10): 5313-8.), it is expected that it is difficult to modulate the binding activity to protein a while maintaining the binding to human FcRn. Maintaining binding to human FcRn is extremely important for the characteristics of IgG-type antibodies, i.e., long plasma retention (long half-life) in humans. Thus, the pharmacokinetics of Q499-z118/J339-z119/L377-k and Q499-z121/J339-z119/L377-k prepared in example 6 were compared.

As a pharmacokinetic experiment for predicting half-life in humans, a human FcRn transgenic mouse (b6.mfcrn-/-. hFcRn Tg line 276 +/+ mouse, Jackson Laboratories) was used for pharmacokinetic evaluation as follows. Q499-z118/J339-z119/L377-k and Q499-z121/J339-z119/L377-k are respectively dosed to mice at 5mg/kg in a single intravenous administration mode, and blood is collected timely. The collected blood was immediately centrifuged at 4 ℃ at 15,000 rpm for 15 minutes to obtain plasma. The separated plasma was stored in a refrigerator set to-20 ℃ or lower until the measurement was performed. Plasma concentrations were determined by ELISA.

As a result, Q499-z118/J339-z119/L377-k and Q499-z121/J339-z119/L377-k showed the same degree of retention in plasma as shown in FIG. 3. It was thus found that: z121/z119, which is a constant region in which a protein A non-binding modification is introduced into one H chain, has a plasma retention property equivalent to that of z118/z119 in which a protein A non-binding modification is not introduced. Therefore, regardless of the variable region, the heterologous antibody can be isolated and purified with high purity and efficiency only by the protein a purification step, and a protein a non-binding modification (for example, a modification in which the amino acid at position 435 in the EU numbering is substituted with Arg) which does not affect pharmacokinetics is found.

[ Examples 8] To the direction of GC33-IgG1-CD3-scFv Is/are as follows CH3 Introduction of mutations in the structural domains exclusively through the protein A Purification step to prepare the target molecule

Introduced through a protein A By carrying out GC33-IgG1-CD3-scFv Purified mutation of molecules

Of the two H chains of the anti-GPC 3 IgG antibody, the production of a molecule that imparts an anti-CD 3 scFv antibody to only one H chain was examined (FIG. 4). The molecule is considered to bind to glypican-3(GPC3) as a cancer-specific antigen in a bivalent manner and to bind to CD3 as a T-cell antigen in a monovalent manner, thereby supplying T cells to cancer cells and killing the cancer cells. In order to bind CD3 at a single valence, it is necessary to add an anti-CD 3 scFv antibody to only one of the two H chains, and it is therefore necessary to purify the molecule formed by heteroassociation of these two H chains.

Therefore, whether or not the target molecule can be purified BY protein a chromatography alone was verified BY introducing a mutation in which His at position 435 in the EU number is substituted with Arg into one H chain and combining the mutation with a mutation described in WO2006/106905 (PROCESS FOR PRODUCTION OF POLYPEPTIDE BY regeneration OF association OF assessment OF association) as a modification FOR promoting heteroassociation OF two H chains (substitution OF Asp at position 356 in the EU number OF one H chain with Lys and substitution OF Lys at position 439 in the EU number OF the other H chain with Glu) in the same manner as in example 3.

Preparation of antibody Gene expression vector and expression of Each antibody

As the antibody H chain variable region, a gene encoding GPC3 (anti-human Glypican-3 antibody H chain variable region, SEQ ID NO: 22) was prepared by a method known to those skilled in the art. As the antibody L chain, a gene encoding GC33-k0 (anti-human Glypican-3 antibody L chain, SEQ ID NO: 23) was prepared by a method known to those skilled in the art. As the antibody H chain constant region, the following genes were prepared by a method known to those skilled in the art:

as a seed for IgG1, LALA-G1d (SEQ ID NO: 24) was constructed by introducing a mutation in which Leu at positions 234 and 235 in the EU numbering system was replaced with Ala and Asn at position 297 was replaced with Ala, and removing Gly and Lys at the C-terminus;

as for scFv of heavy seed CD3 (scFv in which H chain variable region of anti-human CD3 antibody and L chain variable region of anti-human CD3 antibody are bound via polypeptide linker) at C-terminus to LALA-G1d to construct LALA-G1d-CD3(SEQ ID NO: 25);

as a result of introduction of a mutation substituting His at position 435 in the EU numbering system for Arg and a mutation substituting Lys at position 439 in the EU numbering system for Glu into LALA-G3S3E-G1d (SEQ ID NO: 26) and LALA-S3K-G1d-CD3(SEQ ID NO: 27) which was constructed by introducing a mutation substituting Asp at position 356 in the EU numbering system for Lys into LALA-G1d-CD3 were seeded.

Anti-human GPC3 antibody H chain gene NTA1L or NTA1R was prepared by linking LALA-G1d-CD3 as an anti-CD 3 scFv antibody imparting an H chain constant region or LALA-G1d as an H chain constant region to the downstream of H chain variable region GPC3 of an anti-human Glypican-3 antibody. Further, LALA-S3K-G1d-CD3 constructed by introducing an anti-CD 3 scFv antibody to the H chain constant region and a mutation substituting Asp at position 356 in the EU numbering system for Lys, or LALA-G3S3E-G1d constructed by introducing a mutation substituting His at position 435 in the EU numbering system for Arg and a mutation substituting Lys at position 439 in the EU numbering system for Glu were ligated downstream of GPC3 to prepare an anti-human GPC3 antibody H chain gene NTA2L or NTA 2R. The genes were prepared as follows:

h chain

・NTA1L：GPC3-LALA-G1d-CD3；

・NTA1R：GPC3-LALA-G1d；

・NTA2L：GPC3-LALA-S3K-G1d-CD3；

・NTA2R：GPC3-LALA-G3S3E-G1d；

L chain

・GC33-k0。

Each antibody gene (H chain; NTA1L, NTA1R, NTA2L, NTA2R, L chain; GC33-k0) was inserted into an animal cell expression vector. Using the expression vector thus prepared, the following antibodies were transfected into FreeStyle293 cells (invitrogen) by a method known to those skilled in the art, and were overexpressed. The transfected multiple antibody genes were arranged and expressed as antibody names as follows (first H chain/second H chain/L chain):

・NTA1L/NTA1R/GC33-k0；

・NTA2L/NTA2R/GC33-k0。

protein purification of expression samples and evaluation of heterodimer yield

Culture supernatants (CM) of FreeStyle293 cells using the antibodies shown below were used as samples:

・NTA1L/NTA1R/GC33-k0；

・NTA2L/NTA2R/GC33-k0。

CM filtered through a 0.22 μm filter was loaded on a rProtein A Sepharose Fast Flow column (GE Healthcare) equilibrated with D-PBS, and washing 1, washing 2, and elution 1 shown in Table 10 were performed. The loading of CM was adjusted to achieve a loading of 20mg/mL of resin. Separating and collecting the eluted components, and identifying the components contained in the eluted components by size exclusion chromatography analysis.

[ Table 10]

The results of size exclusion chromatography analysis of the eluted fractions are shown in figure 5 and table 11 below. The area values of the elution peaks are expressed as percentages. Few homologous antibodies with anti-CD 3 scFv antibodies in the duplex (NTA1L homologous antibody and NTA2L homologous antibody) were detected in NTA1L/NTA1R/GC33-k0 and NTA2L/NTA2R/GC33-k 0. The reason is considered to be that: since the expression level of scFv molecules is generally low, the expression level of H chain of an anti-CD 3 scFv antibody is extremely low. About 76% of the homologous antibodies to NTA1R were detected in NTA1L/NTA1R/GC33-k0, whereas about 2% of the homologous antibodies to NTA2R were detected in NTA2L/NTA2R/GC33-k0, with respect to the homologous antibodies without anti-CD 3 scFv antibody in the duplex. That is, it was found that a foreign antibody of the target molecular form can be efficiently purified with a purity of 98% or more only by a protein a purification step by introducing a mutation for replacing His at position 435 in the EU numbering with Arg, a mutation for replacing Glu at position 356 in the EU numbering with Lys, and a mutation for replacing Lys at position 439 in the EU numbering with Glu, which are used for efficiently forming a foreign molecule of each H chain.

[ Table 11]

。

[ Examples 9] To monovalent antibodies CH3 Introduction of mutations in the structural domains exclusively through the protein A Purification step to prepare the target molecule

Introduced through a protein A Mutation of purification of monovalent antibody molecules

A common anti-GPC 3 IgG antibody binds to glypican-3(GPC3) as a cancer-specific antigen in two valences via two H chains. In this example, the production of an anti-GPC 3 IgG antibody molecule (FIG. 6) that binds to glypican-3 at a monovalent level was examined. It is believed that the binding of the molecule to glypican-3(GPC3) as a cancer-specific antigen in one valency is not based on avidity but on affinity, and furthermore, the antigen can be bound without cross-linking, as compared to ordinary bivalent antibodies. In order to bind to Glypican-3(GPC3) at one valence, one H chain is required to be a normal H chain, and the other H chain is required to be an H chain of a hinge Fc domain lacking the variable region and CH1 domain.

Then, it was verified whether the target molecule could be purified BY protein a chromatography alone BY introducing a modification in which His at position 435 in the EU number was substituted with Arg into one H chain and further combining the modification with a mutation described in WO2006/106905 (PROCESS FOR PRODUCTION OF POLYPEPTIDE BY regeneration OF association OF assessment OF association) which is a modification promoting heteroassociation OF two H chains (substitution OF Asp at position 356 in the EU number OF one H chain with Lys and substitution OF Lys at position 439 in the EU number OF the other H chain with Glu) in the same manner as in example 3.

Preparation of antibody Gene expression vector and expression of Each antibody

GPC3 (anti-human Glypican-3 antibody H chain variable region, SEQ ID NO: 22).

As the antibody L chain, the following L chains were used:

GC33-k0 (L chain of anti-human Glypican-3 antibody, SEQ ID NO: 23).

introduction of a mutation for substituting His at position 435 in the EU numbering system with Arg into LALA-G1d for seed-LALA-G3-G1 d (SEQ ID NO: 28);

as a seed, LALA-G3S3E-G1d (SEQ ID NO: 26) constructed by introducing a mutation in which Lys at position 439 of the EU numbering system is substituted with Glu into LALA-G3-G1 d;

as a result of the introduction of a mutation into G1Fc, LALA-G1Fc (SEQ ID NO: 29) which had been constructed by deleting positions 1 to 215 in the EU numbering system of LALA-G1d, and LALA-G1Fc-S3K (SEQ ID NO: 30) which had been constructed by substituting Lys for Asp at position 356 in the EU numbering system.

An anti-human GPC3 antibody H chain gene NTA4L-cont, NTL4L-G3, or NTA4L was prepared by linking LALA-G1d as an H chain constant region, LALA-G3-G1d constructed by introducing a mutation substituting His at position 435 in the EU numbering system for Arg, or LALA-G3S3E-G1d constructed by introducing a mutation substituting His at position 435 in the EU numbering system for Arg and a mutation substituting Lys at position 439 for Glu to the downstream of H GPC3 of an anti-human Glyphan-3 antibody. Furthermore, as LALA-G1Fc which is an anti-human hinge Fc domain or LALA-G1Fc-S3K which is a hinge Fc domain constructed by introducing a mutation substituting Lys for Asp at position 356 in the EU numbering system, Fc gene NTA4R-cont or NTA4R was prepared. The genes were prepared as follows.

H chain

・NTA4L-cont：GPC3-LALA-G1d

・NTA4L-G3：GPC3-LALA-G3-G1d

・NTA4L：GPC3-LALA-G3S3E-G1d

・NTA4R-cont: LALA-G1Fc

・NTA4R: LALA-G1Fc-S3K

L chain

・GC33-k0

Each antibody gene (NTA4L, NTA4L-cont, NTA4L-G3, NTA4R, NTA4R-cont, GC33-k0) was inserted into an animal cell expression vector.

・NTA4L-cont/NTA4R-cont/GC33-k0；

・NTA4L-G3/NTA4R-cont/GC33-k0；

・NTA4L/NTA4R/GC33-k0。

protein purification of expression samples and evaluation of heterodimer yield

CM of antibodies shown below was used as a sample:

・NTA4L-cont/NTA4R-cont/GC33-k0；

・NTA4L-G3/NTA4R-cont/GC33-k0；

・NTA4L/NTA4R/GC33-k0。

CM filtered through a 0.22 μm filter was loaded on a rProtein A Sepharose Fast Flow column (GE Healthcare) equilibrated with D-PBS, and washing 1, washing 2, and elution 1 shown in Table 12 were performed. The loading of CM was adjusted to achieve a loading of 20mg/mL of resin. Separating and collecting the eluted components, and identifying the components contained in the eluted components by size exclusion chromatography analysis.

[ Table 12]

。

The results of size exclusion chromatography analysis of the eluted fractions are shown in FIG. 7 and Table 13 below. The area values of the elution peaks are expressed as percentages.

In NTA4L-cont/NTA4R-cont/GC33-k0, the homologous antibody binding to GPC3 in a bivalent form (NTA4L-cont homologous antibody) and the homologous molecule not having a GPC3 binding site (NTA4R-cont homologous antibody) were eluted, and the target NTA4L-cont/NTA4R-cont heterologous antibody was only 46.5%.

In NTA4L-G3/NTA4R-cont/GC33-k0, although almost no homologous antibody binding to GPC3 in a bivalent state (NTA4L-G3 homologous antibody) was detected, a large amount of homologous molecule not having a binding site for GPC3 (NTA4R-cont homologous antibody) was contained, and the amount of the target NTA4L-G3/NTA4R-cont heterologous antibody was 66.7%. In NTA4L/NTA4R/GC33-k0, although almost no homologous antibody binding to GPC3 in divalent form (NTA4L homologous antibody) was detected, the proportion of homologous molecules not having a GPC3 binding site (NTA4R) was further greatly reduced, and the proportion of the target NTA4L/NTA4R heterologous antibody was greatly increased to 93.0%. Namely, it is clear that: by introducing a mutation for replacing His at position 435 in the EU numbering system with Arg and introducing a mutation for replacing Asp at position 356 in the EU numbering system with Lys and a mutation for replacing Lys at position 439 in the EU numbering system with Glu for efficiently forming a hetero-molecule of each H chain, a hetero-antibody in the form of a target molecule can be efficiently purified with a purity of 93% or more only by a protein A purification step.

[ Table 13]

。

[ Examples 10] By means of passing pH Gradient elution of proteins A Column chromatography purification step for preparing heterologous antibody

As shown in example 9, it was found that the heterologous antibody can be efficiently purified only BY the protein A purification step BY combining a mutation in which His at position 435 in the EU numbering system is substituted with Arg with a mutation described in WO2006/106905 (PROCESSS FOR PRODUCTION OF POLYPEPTIDE BY REGULATION OF ASSEMBLY) and a combination OF the mutations (Asp at position 356 in the EU numbering system OF one H chain or Fc is substituted with Lys and Lys at position 439 in the EU numbering system OF the other H chain is substituted with Glu). However, when elution is carried out only with eluent 1(2 mM HCl, pH2.7), the purity of the heterologous antibody is not said to be sufficiently high and further purification steps are necessary.

Thus, in the present embodiment, it is verified that: whether or not the heterologous antibody can be isolated and purified to a higher purity can be achieved by purification by protein A column chromatography using a pH gradient elution in which the number of binding sites to protein A is increased, the binding to protein A is increased, and the pH required for elution is decreased. If the purity of the heterologous antibody can be increased to about 100% by pH gradient elution, cost reduction and efficiency of the purification step can be achieved.

CM of antibodies shown below was used as a sample:

・NTA4L-cont/NTA4R-cont/GC33-k0；

・NTA4L-G3/NTA4R-cont/GC33-k0；

・NTA4L/NTA4R/GC33-k0。

CM filtered through a filter of 0.22 μm in diameter was loaded on a HiTrap protein A HP column (GE Healthcare) equilibrated with D-PBS, and washing 1, washing 2, and elution with pH gradient of elution A and elution B shown in Table 14 were performed in this order. pH gradient elution to elute a: elution B (100:0) → (30:70) was carried out as a slope of a straight line for 35 minutes. The eluted fractions were separated and collected, and the components contained in each eluted fraction were identified by size exclusion chromatography analysis.

[ Table 14]

。

The chromatogram of the protein A column chromatography purification according to the pH gradient elution conditions of NTA4L-cont/NTA4R-cont/GC33-k0, NTA4L-G3/NTA4R-cont/GC33-k0 and NTA4L/NTA4R/GC33-k0 is shown in FIG. 8. A broad peak eluted in NTA4L-cont/NTA4R-cont/GC33-k 0. In NTA4L-G3/NTA4R-cont/GC33-k0, two elution peaks were confirmed by pH gradient elution, and the peak on the high pH side was defined as elution 1 and the peak on the low pH side was defined as elution 2. Although NTA4L/NTA4R/GC33-k0 and NTA4L-G3/NTA4R-cont/GC33-k0 gave approximately the same results, the peak area of elution 2 was small.

The results of size exclusion chromatography analysis of each eluted peak are shown in Table 15. In NTA4L-cont/NTA4R-cont/GC33-k0, 3 components of a homologous antibody binding to GPC3 in a bivalent range (NTA4L-cont homologous antibody), a heterologous antibody binding to GPC3 in a monovalent range (NTA4L-cont/NTA4R-conc heterologous antibody), and a homologous molecule not having a binding site for GPC3 (NTA4R-cont homologous antibody) were detected in the order of elution. Since each component has the same number of sites (two) to which protein a binds, it is considered that these 3 components cannot be separated by pH gradient elution. The method is characterized in that: in elution 1 of NTA4L-G3/NTA4R-cont/GC33-k0, the amount of the homologous antibody binding to GPC3 in two valences (NTA4L-G3 homologous antibody) and the homologous molecule not having a binding site for GPC3 (NTA4R-cont homologous antibody) were below the detection limit, and the amount of the heterologous antibody binding to GPC3 in one valency (NTA4L-G3/NTA4R-conc heterologous antibody) was 99.6%. In elution 2, 98.8% were found to be homologous molecules (NTA4R-cont homologous antibody) that did not have a GPC3 binding site. The NTA4L-G3 homologous antibody failed to bind to protein a due to a mutation replacing His at position 435 in EU numbering with Arg, and thus passed directly through (su tong り) protein a column. In addition, as for the number of binding sites to protein a, the number of binding sites of NTA4L-G3/NTA4R-conc heterologous antibody has one binding site, and the number of binding sites of NTA4R-cont homologous antibody has two binding sites, and the stronger the binding to protein a, the lower the pH necessary for elution, and therefore, it is considered that: compared to NTA4L-G3/NTA4R-conc alloantibody, NTA4R-cont alloantibody elutes at low pH. About the same results were obtained with NTA4L/NTA4R/GC33-k 0. Although almost the same composition ratio as NTA4L-G3/NTA4R-cont/GC33-k0 was obtained in the results of size exclusion chromatography analysis, there was a difference in the chromatogram of protein A, and in NTA4L/NTA4R/GC33-k0, the peak area of elution 2 was small relative to elution 1. This is believed to be due to: since a mutation for efficiently forming NTA4L-G3/NTA4R-cont heterologous antibody was introduced, the expression ratio of NTA4R-cont homologous antibody, which is the main component of elution 2, was decreased. By using the above amino acid mutation, the purification yield and the fastness of a heterologous antibody purified by protein A column chromatography using pH gradient elution can be improved.

From the above results, it was found that: the heterologous antibody can be isolated and purified with high purity and efficiency only by a protein a column chromatography purification step using pH gradient elution.

[ Table 15]

。

[ Examples 11] To one side Fcalpha Receptors Fc Of fusion proteins CH3 Introduction of mutations into the structural Domain solely by proteins A Purification step to prepare the target molecule

To the direction of CH3 Introduction of mutations in the structural domains and passage through proteins A Purification step preparation of monovalent Fcalpha Receptors Fc Fusion proteins

Common Fc receptor Fc fusion proteins, represented by Eternercept or Abatacept, are homodimers that can bind to ligands in a bivalent manner. In this example, the preparation of Fcalpha receptor Fc fusion (fig. 9) that binds IgA as a ligand with a monovalent valence was examined. In order to allow Fcalpha receptors to bind IgA monovalent, it is necessary that one side H chain of two Fc receptor Fc fusion H chains is a full-length H chain of a hinge Fc domain, and therefore, it is necessary to purify a molecule constructed by heterologously associating these two H chains.

Then, it was verified whether the target molecule could be purified only BY protein a chromatography BY introducing a mutation substituting His at position 435 in the EU number FOR Arg into one H-chain and combining the mutation with a mutation described in WO2006/106905 (PROCESS FOR product OF POLYPEPTIDEs BY weight) as a modification FOR promoting heteroassociation OF two H-chains (substitution OF Asp at position 356 in the EU number OF one H-chain FOR Lys and substitution OF Lys at position 439 in the EU number OF the other H-chain FOR Glu) in the same manner as in example 6.

Preparation of antibody Gene expression vector and expression of Each antibody

As the Fc receptor, Fcalphar (human IgA1 receptor, SEQ ID NO: 31) was used.

As the fusion H chain constant region, the following H chain constant regions were used:

seed IgG1, G1Fc (SEQ ID NO: 32) which is a human hinge Fc domain constructed by deletion of Gly and Lys at positions 1 to 223 in the EU numbering and at the C-terminal;

as for the seed G1Fc, G1Fc-G3S3K (SEQ ID NO: 33) constructed by introducing a mutation for substituting Lys for Asp at position 356 in the EU numbering system and a mutation for substituting Arg for His at position 435 was introduced;

as a result of introduction of a mutation for substituting Lys at position 439 of the EU numbering system with Glu into G1Fc, G1Fc-S3E (SEQ ID NO: 34) was constructed.

Fcalphar-Fc fusion proteins IAL-cont and IAL were prepared by connecting, via a polypeptide linker (SEQ ID NO: 35), G1Fc as an H chain constant region and G1Fc-G3S3K constructed by introducing mutations in which Asp at position 356 in the EU numbering system was substituted with Lys and His at position 435 was substituted with Arg to the downstream of FcalpharR.

Further, an Fc gene IAR-cont or IAR was prepared as G1Fc which is a human hinge Fc domain or G1Fc-S3E which is a hinge Fc domain constructed by introducing a mutation substituting Glu for Lys at position 439 in the EU numbering system. The genes were prepared as follows:

h chain

・IAL-cont：FcalphaR-G1Fc；

・IAL：FcalphaR-G1Fc-G3S3K；

・IAR-cont: G1Fc；

・IAR: G1Fc-S3E。

Each antibody gene (IAL-cont, IAL, IAR-cont, IAR) was inserted into an animal cell expression vector.

・IAL-cont/IAR-cont；

・IAL/IAR。

protein purification of expression samples and evaluation of heterodimer yield

CM of antibodies shown below was used as a sample:

・IAL-cont/IAR-cont；

・IAL/IAR。

CM filtered through a 0.22 μm filter was loaded on a rProtein A Sepharose Fast Flow column (GE Healthcare) equilibrated with D-PBS, and washing 1, washing 2, and elution 1 shown in Table 16 were performed. The loading of CM was adjusted to achieve a loading of 20mg/mL of resin. Separating and collecting the eluted components, and identifying the components contained in the eluted components by size exclusion chromatography analysis.

[ Table 16]

。

The results of size exclusion chromatographic analysis of the eluted fractions are shown in FIG. 10 and Table 17 below. The area values of the elution peaks are expressed as percentages. In IAL-cont/IAR-cont, the bivalent IgA-binding cognate antibody (IAL-cont cognate antibody) and the cognate molecule without IgA binding site (IAR-cont cognate antibody) are eluted, and the target IAL-cont/IAR-cont heterologous antibody is only 30%. In IAL/IAR, no homologous antibody that binds IgA in bivalent form (IAL homologous antibody) was detected, and the proportion of homologous molecules that do not have an IgA binding site (IAR homologous antibody) was greatly reduced, and the proportion of target IAL/IAR heterologous antibody was greatly increased to about 96%. Namely, it is clear that: by introducing a mutation for replacing His at position 435 in the EU numbering system with Arg and introducing a mutation for replacing Asp at position 356 in the EU numbering system with Lys and a mutation for replacing Lys at position 439 in the EU numbering system with Glu in order to efficiently form a heterologous molecule in each H chain, a heterologous antibody having a desired molecular form can be efficiently purified with a purity of 95% or more only by a protein A purification step.

[ Table 17]

。

[ Examples 12] 4 Chain IgG Preparation of bispecific antibody

Preparation of antibody Gene expression vector and expression of Each antibody

The bispecific antibodies against human f.ix and human F.X prepared in example 1 were composed of two H chains and a common L chain recognizing the respective antigens. It is not easy to obtain such bispecific antibodies with a common L chain. This is because it is difficult to recognize two kinds of antigens by an L chain having a common sequence. Thus, since it is extremely difficult to obtain a common L chain, it is considered that a bispecific antibody composed of two H chains and two L chains recognizing two antigens is preferable, but when two H chains and two L chains are expressed, they are randomly combined to express 10H 2L2 type IgG molecules. Of these 10H 2L 2-type IgG molecules, it was extremely difficult to purify the bispecific antibody of interest.

In this example, the preparation of bispecific antibodies composed of two H chains and two L chains of anti-human IL-6 receptor and human glypican-3(GPC3) was investigated. In order to efficiently produce a bispecific antibody composed of two H chains and two L chains, it is necessary to promote association of the H chains and L chains against the same antigen and promote hetero-association of the two H chains. Furthermore, it is necessary that bispecific antibodies with the correct combination can be purified from the resulting expression product.

To facilitate the association of H and L chains against the same antigen, the variable region (VH) of the H chain (GC33-VH-CH1-hinge-CH2-CH3) and the variable region (VL) of the L chain (GC33-VL-CL) of the anti-GPC 3 antibody, i.e. GC33, were exchanged with each other, making H chain (GC33-VL-CH1-hinge-CH2-CH3) and L chain (GC33-VH-CL) (VH domain and VL domain exchange). Although GC33-VL-CH1-hinge-CH2-CH3 was associated with GC33-VH-CL, the association of GC33-VL-CH1-hinge-CH2-CH3 with the L chain of an anti-IL-6 receptor antibody (MRA-VL-CL) was inhibited by the instability of the VL/VL interaction. Similarly, while the H chain of the anti-IL-6 receptor antibody (MRA-VH-CH1-hinge-CH2-CH3) is associated with MRA-VL-CL, the association of the H chain of the anti-IL-6 receptor antibody (MRA-VH-CH1-hinge-CH2-CH3) with the L chain of the anti-GPC 3 antibody (GC33-VH-CL) is inhibited by the instability of the VH/VH interaction. By doing so, it is possible to promote the association of H chains with L chains against the same antigen, and also to cause VH/VH and VL/VL interactions, but less stable than VH/VL interactions (reports for VH/VH: FEBS Lett. 2003 Nov 20; 554(3): 323-9., J Mol biol. 2003 Oct 17; 333(2): 355-65., reports for VL/VL: J Struct biol. 2002 Jun; 138(3): 171-86., Proc Natl Acad Sci U S A. 1985 Jul; 82(14): 4592-6.), and thus many undesirable associations between H chains and L chains occur. Thus, by exchanging only the VH domain and VL domain, about 10 combined products are expressed, although the ratio of the bispecific antibody of interest is increased.

Usually, it is extremely difficult to purify the target bispecific antibody from these 10 components, but the separation of these 10 components in ion exchange chromatography can be improved by introducing modifications such that the 10 components have different isoelectric points. Thus, a modification for lowering the isoelectric point was introduced into MRA-VH, which is the H chain variable region of the anti-IL-6 receptor antibody, to prepare H54-VH having a lowered isoelectric point. In the same manner, a modification for lowering the isoelectric point was introduced into the L chain variable region of the anti-IL-6 receptor antibody, namely MRA-VL, to produce L28-VL having a lowered isoelectric point. Furthermore, a modification for increasing the isoelectric point was introduced into the GC33-VH, which is the H chain variable region of the anti-GPC 3 antibody, to produce Hu22-VH having an increased isoelectric point.

By exchanging the VH and VL of the H chain and L chain of the anti-GPC 3 antibody, the combination of the target H chain and L chain was improved, but the interaction of H54-VH/Hu22-VH and L28-VL/GC33-VL was not completely inhibited, so that also many undesirable associations of H chain and L chain occurred. The 39 th position of the common antibody sequence is glutamine, which is believed to form hydrogen bonds with each other at the VH/VH interface when VH/VH interactions occur. Thereafter, to further attenuate the interaction of H54-VH/Hu22-VH, glutamine at position 39 of Kabat numbering was substituted with lysine. It is thus assumed that: the VH/VH interaction is greatly reduced due to electrostatic repulsion of the lysines from each other at the VH/VH interface. Then, H54-VH-Q39K and Hu22-VH-Q39K were prepared by substituting lysine for glutamine at position 39 of Kabat numbering in the sequences of H54-VH and Hu 22-VH. By the same token, in the VL/VL interaction, since the 38 th position of the normal antibody sequence is also glutamine, it is considered that glutamine forms a hydrogen bond with each other at the VL/VL interface. Then, to further attenuate the L28-VL/GC33-VL interaction, the glutamine at Kabat numbering position 38 was substituted with glutamic acid. It is thus believed that the VL/VL interaction is greatly reduced due to electrostatic repulsion of the glutamic acids from each other at the VL/VL interface. Then, L28-VL-Q38E and GC33-VL-Q38E were prepared by substituting glutamine at Kabat numbering position 39 in the sequences of L28-VL and GC33-VL with glutamic acid.

In order to express and purify a bispecific antibody OF interest more efficiently, a molecule constructed BY heteroassociation OF two H chains can be purified BY protein a chromatography alone, BY introducing a mutation substituting His at position 435 in the EU number FOR Arg into one H chain, and combining the mutation described in WO2006/106905 (PROCESS FOR product OF POLYPEPTIDE BY regeneration OF association) as a modification FOR promoting heteroassociation OF two H chains (substitution OF Asp at position 356 in the EU number OF one H chain FOR Lys, and substitution OF Lys at position 439 at the EU number OF the other H chain FOR Glu), in the same manner as in example 3.

That is, as the antibody H chain variable region, the following H chain variable regions were used:

seeds of MRA-VH (H chain variable region of anti-human interleukin-6 receptor antibody, SEQ ID NO: 36);

seed GC33-VH (H chain variable region of anti-GPC 3 antibody, SEQ ID NO: 37);

H54-VH (H chain variable region of anti-human interleukin-6 receptor antibody, SEQ ID NO: 38) which reduce the isoelectric point of MRA-VH;

hu22-VH (H chain variable region of anti-GPC 3 antibody, SEQ ID NO: 39) which enhances the isoelectric point of GC 33-VH;

H54-VH-Q39K (SEQ ID NO: 40) in which Gln at Kabat numbering 39 in the sequence of seed H54-VH is substituted with Lys;

as a seed, Hu22-VH-Q39K (SEQ ID NO: 41) in which Gln at position 39 of Kabat numbering was substituted with Lys in the sequence of Hu 22-VH.

In addition, as the antibody H chain constant region, the following H chain constant regions were used:

as for the sequences of H chain constant regions of IgG1, IgG1-LALA-N297A-CH (SEQ ID NO: 42) was constructed by substituting Leu at positions 234 and 235 in the EU numbering system with Ala, Asn at position 297 with Ala, and removing Gly and Lys at the C-terminus;

IgG1-LALA-N297A-CHr (SEQ ID NO: 43) constructed by adding two Ser residues to the N-terminus of the sequence of IgG 1-LALA-N297A-CH;

IgG1-LALA-N297A-s3-CH (SEQ ID NO: 44) in which Lys at position 439 in the EU numbering system in the sequence of IgG1-LALA-N297A-CH is substituted with Glu;

as a seed, IgG1-LALA-N297A-CHr (SEQ ID NO: 45) in which Asp at position 356 in the EU numbering system was replaced with Lys and His at position 435 was replaced with Arg was used as IgG1-LALA-N297A-G3s 3-CHr.

As the antibody L chain variable region, the following L chain variable regions were used:

as a seed, MRA-VL (L chain variable region of anti-human interleukin-6 receptor antibody, SEQ ID NO: 46)

As a seed GC33-VL (L chain variable region of anti-GPC 3 antibody, SEQ ID NO: 47)

As a seed, L28-VL (L chain variable region of anti-human interleukin-6 receptor antibody, SEQ ID NO: 48) which lowers the isoelectric point of MRA-VL

As a seed, L28-VL-Q38E (SEQ ID NO: 49) in which the Kabat-numbered 38 th Gln residue in the sequence of L28-VL is substituted with Glu

As a seed of GC33-VL, GC33-VL-Q38E (SEQ ID NO: 50) in which the Kabat-numbered Gln at position 38 is substituted with Glu in the sequence of Kabat-numbered amino acids

As the antibody L chain constant region, the following L chain constant regions were used:

as a seed IgG1-CL (L chain constant region of IgG1, SEQ ID NO: 51)

As a C-terminal replacement of Ala and Ser of the sequence of As a seed IgG1-CL, IgG1-CLr (SEQ ID NO: 52) in which Arg and Thr are substituted

The gene no1-Mh-H was prepared by ligating IgG1-LALA-N297A-CH to the downstream of MRA-VH. The gene no1-Mh-L was prepared by ligating IgG1-CL downstream of MRA-VL. The gene no1-Gh-H was prepared by ligating IgG1-LALA-N297A-CH to the downstream of GC 33-VH. The gene no1-Gh-L was prepared by ligating IgG1-CL downstream of GC 33-VL.

The gene no2-Gh-H was prepared by ligating IgG1-LALA-N297A-CHr downstream of GC 33-VL. The gene no2-Gh-L was prepared by ligating IgG1-CLr downstream of GC 33-VH.

The gene no3-Ml-H was prepared by ligating IgG1-LALA-N297A-CH downstream of H54-VH. The gene no3-Ml-L was prepared by ligating IgG1-CL downstream of L28-VL. The gene no3-Ghh-L was prepared by ligating IgG1-CLr downstream of Hu 22-VH.

The gene no5-Ml-H was generated by ligating IgG1-LALA-N297A-s3-CH downstream of H54-VH. The gene no5-Gh-H was generated by ligating IgG1-LALA-N297A-G3s3-CHr downstream of GC 33-VL.

The gene no6-Ml-H was prepared by ligating IgG1-LALA-N297A-s3-CH to the downstream of H54-VH-Q39K. The gene no6-Ml-L was created by ligating IgG1-CL downstream of L28-VL-Q38E. Gene no6-Gh-H was generated by ligating IgG1-LALA-N297A-G3s3-CHr downstream of GC 33-VL-Q38E. Gene no6-Ghh-L was prepared by ligating IgG1-CLr downstream of Hu 22-VH-Q39K.

Each gene (no1-Mh-H, no1-Mh-L, no1-Gh-H, no1-Gh-L, no2-Gh-H, no2-Gh-L, no3-Ml-H, no3-Ml-L, no3-Ghh-L, no5-Ml-H, no5-Gh-H, no6-Ml-H, no6-Ml-L, no6-Gh-H, no6-Ghh-L) is inserted into an animal cell expression vector.

The expression vectors of the combinations shown below were introduced into FreeStyle293-F cells to overexpress each target molecule.

A. Target molecule: no1 (FIG. 11)

Description of the drawings: natural anti-IL-6 receptor-anti-GPC 3 bispecific antibody

A polypeptide encoded by a polynucleotide inserted into an expression vector: NO1-Mh-H (SEQ ID NO: 53), NO1-Mh-L (SEQ ID NO: 54), NO1-Gh-H (SEQ ID NO: 55), NO1-Gh-L (SEQ ID NO: 56)

B. Target molecule: no2 (FIG. 12)

Description of the drawings: exchange of VH Domain and VL Domain of anti-GPC 3 antibody in no1

A polypeptide encoded by a polynucleotide inserted into an expression vector: NO1-Mh-H, NO1-Mh-L, NO2-Gh-H (SEQ ID NO: 57), NO2-Gh-L (SEQ ID NO: 58)

C. Target molecule: no3 (FIG. 13)

Description of the drawings: introduction of modifications to no2 to alter the isoelectric points of the respective chains

A polypeptide encoded by a polynucleotide inserted into an expression vector: NO3-Ml-H (SEQ ID NO: 59), NO3-Ml-L (SEQ ID NO: 60), NO2-Gh-H, NO3-Ghh-L (SEQ ID NO: 61)

D. Target molecule: no5 (FIG. 14)

Description of the drawings: introduction of modification promoting H chain heteroassociation and modification for purification of antibody heteroassociation with protein A into no3

A polypeptide encoded by a polynucleotide inserted into an expression vector: NO5-Ml-H (SEQ ID NO: 62), NO3-Ml-L, NO5-Gh-H (SEQ ID NO: 63), NO3-Ghh-L

E. Target molecule: no6 (FIG. 15)

Description of the drawings: introduction of a modification to no5 to promote association of target H chain with target L chain

A polypeptide encoded by a polynucleotide inserted into an expression vector: NO6-Ml-H (SEQ ID NO: 64), NO6-Ml-L (SEQ ID NO: 65), NO6-Gh-H (SEQ ID NO: 66), NO6-Ghh-L (SEQ ID NO: 67).

To the culture supernatant obtained by filtration through a filter of 0.22 μm was added rProtein A Sepharose Fast Flow resin (GE Healthcare) equilibrated with the medium, and purification was performed by batch elution. Since protein G binds to the Fab portion of the antibody, by using protein G, all antibodies present in CM can be purified regardless of affinity to protein a.

The expression pattern of the antibodies (no1, no2, no3, no5, no6) prepared was evaluated by cation exchange chromatography (IEC). In the cation exchange chromatography, a ProPac WCX-10 column (Dionex) as an analytical column was used, and the column was carried out at a flow rate of 0.5 mL/min under an appropriate gradient using 20mM MES-NaOH and pH6.1 for mobile phase A and 20mM MES-NaOH and 250 mM NaCl for mobile phase B, and pH 6.1. The results of IEC evaluation of each antibody are shown in fig. 16. In no1 of the natural anti-IL-6 receptor-anti-GPC 3 bispecific antibody, there were many peaks close to each other, and it was not possible to distinguish which peak was the target bispecific antibody. The same applies to no2 constructed by exchanging the VH domain and VL domain of anti-GPC 3 antibody in no 1. In no3 constructed by introducing a modification to no2 that alters the isoelectric points of the respective chains, the peak of the bispecific antibody of interest can be isolated for the first time. In no5 constructed by introducing a modification promoting H-chain heterologous association and a modification for purifying an antibody heteroassociated with protein a into no3, the ratio of the peak of the bispecific antibody of interest was significantly increased. In no6 constructed by introducing a modification that promotes the association of the target H chain and the target L chain into no5, the ratio of the peak of the target bispecific antibody was further increased.

Therefore, it was investigated whether it was possible to purify the target bispecific antibody with high purity using CM no6 and using a column for purification. CM filtered through a filter of 0.22 μm in diameter was loaded on a HiTrap protein A HP column (GE Healthcare) equilibrated with D-PBS, and washing 1, washing 2, and elution with a pH gradient of elution A and elution B shown in Table 18 were sequentially performed. pH gradient elution to elute a: elution B (100:0) → (35:65) was performed with a slope of the line for 40 minutes.

[ Table 18]

。

The pH gradient elution results for No6 are shown in FIG. 17. The H chain of the anti-GPC 3 antibody which did not bind to protein A was directly transferred to protein A, and it was found that: the peak eluting at position 1 was a heterologous antibody to the H chain of the anti-GPC 3 antibody and the H chain of the anti-IL-6 receptor antibody, and the peak eluting at position 2 was a homologous antibody to the H chain of the anti-IL-6 receptor antibody. Thereby confirming that: by substituting His at position 435 in the EU numbering with Arg, a heterologous antibody against the H chain of the GPC3 antibody and the H chain of the anti-IL-6 receptor antibody can be purified only by the protein a purification step.

For the eluted fraction at position 1, it was added to a HiTrap SP Sepharose HP column (GE Healthcare) equilibrated with 20mM sodium acetate buffer (pH5.5), washed with the same solution, and then subjected to NaCl concentration gradient elution from 0mM to 500 mM. For the obtained main peak components, cation exchange chromatography analysis was performed in the same manner. The results are shown in FIG. 18. The results show that: the bispecific antibody of interest can be purified with extremely high purity.

Industrial applicability

In the present invention, the following methods are provided: a method for purifying or producing a polypeptide multimer (multispecific antibody) having binding activity to two or more antigens with high purity and efficiency by using only a protein A purification step by changing the binding force to protein A. By using the method of the present invention, a target polypeptide multimer can be purified or produced with high purity and efficiency without impairing the effect of other target amino acid modifications. Particularly, by combining the above-mentioned method with a method of controlling the association between two protein domains, a polypeptide multimer of interest can be purified or produced with higher purity and efficiency.

Claims

1. A method for producing a polypeptide multimer comprising a1 st polypeptide having antigen-binding activity and a2 nd polypeptide having antigen-binding activity or not, the method comprising:

(b) a step of recovering the expression product of step (a),

2. The method of claim 1, wherein, in step (b), protein a affinity chromatography is used to recover the expression product.

3. The method according to claim 1 or 2, wherein one or more amino acid residues in both or either of the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity or having no antigen-binding activity are modified so that a difference is produced between the pH of a solvent for eluting the 1 st polypeptide having antigen-binding activity from the protein A and the pH of a solvent for eluting the 2 nd polypeptide having antigen-binding activity or having no antigen-binding activity from the protein A.

4. The method according to any one of claims 1 to 3, wherein one or more amino acid residues in the 1 st polypeptide having antigen-binding activity or the 2 nd polypeptide having antigen-binding activity or having no antigen-binding activity are modified so that the binding force between the 1 st polypeptide having antigen-binding activity or the 2 nd polypeptide having antigen-binding activity or having no antigen-binding activity and protein A is increased or decreased.

5. The method according to any one of claims 1 to 4, wherein one or more amino acid residues in the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity or having no antigen-binding activity are modified so that the binding force between the protein A and the polypeptide of either the 1 st polypeptide having antigen-binding activity or the 2 nd polypeptide having antigen-binding activity or having no antigen-binding activity is increased and the binding force between the protein A and the polypeptide of the other is decreased.

6. The method according to any one of claims 1 to 5, wherein the purity of the recovered polypeptide multimer is 95% or more.

7. The method of any one of claims 1 to 6, wherein the 1 st polypeptide having antigen binding activity and the 2 nd polypeptide having antigen binding activity or having no antigen binding activity comprise an amino acid sequence of an antibody Fc region or an antibody heavy chain constant region.

8. The method according to claim 7, wherein at least one amino acid residue selected from the group consisting of positions 250 to 255, positions 308 to 317 and positions 430 to 436 in the EU numbering system in the amino acid sequence of the Fc region of the antibody or the constant region of the heavy chain of the antibody is modified.

9. The method according to any one of claims 1 to 8, wherein the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity comprise an amino acid sequence of an antibody heavy chain variable region.

10. The method of claim 9, wherein at least one amino acid residue in the amino acid sequences of FR1, CDR2 and FR3 in the variable region of an antibody heavy chain is modified.

11. The method according to any one of claims 1 to 10, wherein the polypeptide multimer comprises one or two 3 rd polypeptides having antigen-binding activity, and step (a) comprises expressing DNA encoding the 3 rd polypeptides having antigen-binding activity.

12. The method of claim 11, wherein the 3 rd polypeptide having antigen binding activity comprises the amino acid sequence of an antibody light chain.

13. The method of claim 11 or 12, wherein the polypeptide multimer further comprises a4 th polypeptide having antigen-binding activity, and step (a) comprises expressing a DNA encoding the 4 th polypeptide having antigen-binding activity.

14. The method of claim 13, wherein at least one of the 3 rd polypeptide and the 4 th polypeptide having antigen binding activity comprises an amino acid sequence of an antibody light chain.

15. The method of claim 13, wherein the 1 st polypeptide having antigen binding activity comprises the amino acid sequences of an antibody light chain variable region and an antibody heavy chain constant region, the 2 nd polypeptide having antigen binding activity comprises the amino acid sequence of an antibody heavy chain, the 3 rd polypeptide having antigen binding activity comprises the amino acid sequences of an antibody heavy chain variable region and an antibody light chain constant region, and the 4 th polypeptide having antigen binding activity comprises the amino acid sequence of an antibody light chain.

16. The method of any one of claims 1 to 15, wherein the polypeptide multimer is a multispecific antibody.

17. The method of claim 16, wherein the multispecific antibody is a bispecific antibody.

18. The method of any one of claims 1 to 8, comprising a1 st polypeptide having antigen binding activity and a2 nd polypeptide having no antigen binding activity, wherein the 1 st polypeptide having antigen binding activity comprises the antigen binding domain of the receptor and the amino acid sequence of the antibody Fc region, and the 2 nd polypeptide having no antigen binding activity comprises the amino acid sequence of the antibody Fc region.

19. The method of any one of claims 7 to 18, wherein the antibody Fc region or antibody heavy chain constant region is derived from a human IgG.

20. A polypeptide multimer prepared by the method of any one of claims 1-19.

21. A method for purifying a polypeptide multimer comprising a1 st polypeptide having antigen-binding activity and a2 nd polypeptide having antigen-binding activity or not, the method comprising:

22. The method of claim 21, wherein one or more amino acid residues in the 1 st polypeptide having antigen binding activity or the 2 nd polypeptide having antigen binding activity or not are modified to increase or decrease the binding force of the 1 st polypeptide having antigen binding activity or the 2 nd polypeptide having antigen binding activity or not to protein a.

23. The method according to claim 20 or 21, wherein one or more amino acid residues in the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity or having no antigen-binding activity are modified so that the binding force between the protein A and the polypeptide of either the 1 st polypeptide having antigen-binding activity or the 2 nd polypeptide having antigen-binding activity or having no antigen-binding activity is increased and the binding force between the protein A and the polypeptide of the other is decreased.

24. The method of any one of claims 21 to 23, wherein the purity of the recovered polypeptide multimer is 95% or more.

25. The method of any one of claims 21-24, wherein the 1 st polypeptide having antigen binding activity, and the 2 nd polypeptide having antigen binding activity or not having antigen binding activity comprise the amino acid sequence of an antibody Fc region or an antibody heavy chain constant region.

26. The method of claim 25, wherein at least one amino acid residue selected from positions 250-255, 308-317, and 430-436 of the EU numbering in the amino acid sequence of the Fc region or the constant region of a heavy chain of the antibody is modified.

27. The method of any one of claims 21 to 26, wherein the 1 st polypeptide having antigen binding activity and the 2 nd polypeptide having antigen binding activity comprise an amino acid sequence of an antibody heavy chain variable region.

28. The method of claim 27, wherein at least one amino acid residue in the amino acid sequences of FR1, CDR2 and FR3 in the variable region of the antibody heavy chain is modified.

29. The method of any one of claims 21 to 28, wherein the polypeptide multimer comprises one or two 3 rd polypeptides having antigen-binding activity, and step (a) comprises expressing DNA encoding the 3 rd polypeptides having antigen-binding activity.

30. The method of claim 29, wherein the 3 rd polypeptide having antigen binding activity comprises the amino acid sequence of an antibody light chain.

31. The method of claim 29 or 30, wherein the polypeptide multimer further comprises a4 th polypeptide having antigen-binding activity, and step (a) comprises expressing DNA encoding the 4 th polypeptide having antigen-binding activity.

32. The method of claim 31, wherein at least one of the 3 rd polypeptide and the 4 th polypeptide having antigen binding activity comprises an amino acid sequence of an antibody light chain.

33. The method of claim 31, wherein the 1 st polypeptide having antigen binding activity comprises the amino acid sequences of an antibody light chain variable region and an antibody heavy chain constant region, the 2 nd polypeptide having antigen binding activity comprises the amino acid sequence of an antibody heavy chain, the 3 rd polypeptide having antigen binding activity comprises the amino acid sequences of an antibody heavy chain variable region and an antibody light chain constant region, and the 4 th polypeptide having antigen binding activity comprises the amino acid sequence of an antibody light chain.

34. The method of any one of claims 21 to 33, wherein the polypeptide multimer is a multispecific antibody.

35. The method of claim 34, wherein the multispecific antibody is a bispecific antibody.

36. The method of any one of claims 25 to 35, wherein the antibody Fc region or antibody heavy chain constant region is derived from a human IgG.

37. A polypeptide multimer comprising a1 st polypeptide having antigen-binding activity and a2 nd polypeptide having antigen-binding activity or not,

38. The polypeptide multimer of claim 37, wherein the pH of the solvent that elutes the 1 st polypeptide having antigen binding activity from protein a is different from the pH of the solvent that elutes the 2 nd polypeptide having antigen binding activity or not from protein a.

39. The polypeptide multimer of claim 37 or 38, wherein the 1 st polypeptide having antigen-binding activity or the 2 nd polypeptide having antigen-binding activity or having no antigen-binding activity comprises an amino acid sequence of an antibody Fc region or an antibody heavy chain constant region,

40. The polypeptide multimer of any of claims 37-39, wherein the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity or not comprise an amino acid sequence of an antibody Fc region or an antibody heavy chain constant region,

41. The polypeptide multimer of any of claims 37-40, wherein the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity or not comprise an amino acid sequence of an antibody Fc region or an antibody heavy chain constant region,

42. The polypeptide multimer of any of claims 37 to 41, wherein the 1 st polypeptide having antigen-binding activity and the 2 nd polypeptide having antigen-binding activity comprise an amino acid sequence of an antibody heavy chain variable region,

43. The polypeptide multimer of any of claims 37-42, wherein the polypeptide multimer further comprises one or two 3 rd polypeptides having antigen-binding activity.

44. The polypeptide multimer of claim 43, wherein the 3 rd polypeptide having antigen-binding activity comprises the amino acid sequence of an antibody light chain.

45. The polypeptide multimer of claim 43 or 44, wherein the polypeptide multimer further comprises a4 th polypeptide having antigen-binding activity.

46. The polypeptide multimer of claim 45, wherein at least one of the 3 rd and 4 th polypeptides having antigen binding activity comprises an amino acid sequence of an antibody light chain.

47. The polypeptide multimer of claim 45, wherein the 1 st polypeptide having antigen-binding activity comprises the amino acid sequences of an antibody light chain variable region and an antibody heavy chain constant region, the 2 nd polypeptide having antigen-binding activity comprises the amino acid sequence of an antibody heavy chain, the 3 rd polypeptide having antigen-binding activity comprises the amino acid sequences of an antibody heavy chain variable region and an antibody light chain constant region, and the 4 th polypeptide having antigen-binding activity comprises the amino acid sequence of an antibody light chain.

48. The polypeptide multimer of any of claims 37-47, which is a multispecific antibody.

49. The polypeptide multimer of claim 48, wherein the multispecific antibody is a bispecific antibody.

50. The polypeptide multimer of any of claims 37-41, comprising a1 st polypeptide having antigen binding activity and a2 nd polypeptide having no antigen binding activity, wherein the 1 st polypeptide having antigen binding activity comprises the antigen binding domain of the receptor and the amino acid sequence of the antibody Fc region, and the 2 nd polypeptide having no antigen binding activity comprises the amino acid sequence of the antibody Fc region.

51. The polypeptide multimer of any of claims 39-50, wherein an antibody Fc region or an antibody heavy chain constant region is derived from a human IgG.

52. A nucleic acid encoding a polypeptide constituting the polypeptide multimer of any one of claims 20, 37-51.

53. A vector into which the nucleic acid of claim 52 is inserted.

54. A cell comprising the nucleic acid of claim 52 or the vector of claim 53.

55. A pharmaceutical composition comprising the polypeptide multimer of any one of claims 20 and 37-51 as an active ingredient.