JP2007186522A - Humanized antibody against human tissue factor(tf) and method for producing the humanized antibody - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method for producing a natural humanized antibody. <P>SOLUTION: The natural humanized antibody has a nonhuman-derived complementarity determining region(CDR) and a natural human antibody-derived framework region(FR) and is reduced in immunogenicity. The method for producing the above antibody is also provided. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a human / mouse chimeric antibody comprising a variable region (V region) of a mouse monoclonal antibody against human tissue factor (TF) and a constant region (C region) of a human antibody, a light chain of a mouse monoclonal antibody against human TF ( A humanized antibody in which the compatibilities determining region (CDR) of the L chain) V region and the heavy chain (H chain) V region are transplanted into a human antibody, the L chain and H chain of the antibody, and the The present invention relates to a fragment of the V region constituting the L chain or H chain of an antibody. The present invention further relates to a method for producing a humanized antibody against human TF.

The present invention further relates to DNA encoding the above antibody, particularly a fragment of the V region thereof, and DNA encoding an L chain or H chain containing the V region. The present invention further relates to a recombinant vector containing the DNA and a host transformed with the vector.
The present invention further relates to a method for producing a chimeric antibody and a humanized antibody against human TF. The present invention further relates to a pharmaceutical composition comprising a humanized antibody against human TF as an active ingredient and a therapeutic agent for disseminated intravascular coagulation syndrome (DIC).

Tissue factor (TF) is a coagulation factor VII receptor expressed on the cell surface and plays an essential role in the activation of coagulation factor IX and factor X through complex formation with coagulation factor VII. It is positioned as a substantial initiation factor for blood clotting reactions.
TF is expressed in fibroblasts and smooth muscle cells that compose blood vessels, and is known to activate the coagulation system and perform a hemostatic function upon vascular injury.

  DIC is a disease in which thrombus occurs frequently mainly in small blood vessels throughout the body due to the activation of the coagulation system in blood vessels. When platelets and coagulation factors are consumed and lowered, bleeding often occurs, which is the opposite phenomenon to thrombus. In addition, microcirculatory failure of important organs due to frequent microthrombosis, which often causes irreversible dysfunction after onset, is recognized as an important disease because the prognosis of DIC is poor.

  Ministry of Health and Welfare Specific Diseases Blood Coagulation Disorder Research Group The proportion of basic diseases estimated from the research reports in FY 1990 and FY 2004 is about 30% for hematopoietic malignancies, about 20% for solid cancer, and about 15% for infectious diseases About 10% of obstetrical diseases, 6% of liver diseases, 5% of shocks, and 3% of cardiovascular diseases. The incidence of DIC is as high as about 15% for leukemia and 6-7% for malignant lymphoma, and about 3% for solid cancer.

  Although DIC develops in combination with these various diseases, the causative substance is common and it is TF. That is, in acute leukemia, malignant lymphoma, and solid cancer, abnormally increased TF production / expression in tumor cells, and in infections (especially Gram-negative bacterial sepsis), increased TF production / expression in monocytes / vascular endothelial cells, In fulminant hepatitis, TF flows from necrotic liver tissue into the blood, in aortic aneurysms / heart aneurysms / giant hemangiomas, TF formation on the inner surface of the blood vessels, and obstetric diseases (amniotic fluid embolism, normal placenta early detachment) The influx of TF into the blood is considered to be the onset mechanism of DIC in surgery, trauma, and burns.

Treatment of the primary disease (underlying disease) is the first, but in practice this is not easy.
Current DIC treatment methods include anticoagulant therapy and replacement therapy. Heparin preparations (unfractionated heparin, low molecular weight heparin) are at the center of anticoagulation therapy. Synthetic protease inhibitors (gabexate mesylate, nafamostat mesylate) and concentrated plasma preparations (antithrombin III) , Activated protein C preparations) are also used. Replacement therapy includes concentrated platelet plasma, fresh frozen plasma (supplement of fibrin), washed red blood cells, and the like.

  However, there are no current therapies that are sufficiently satisfactory in terms of efficacy and side effects, and in most cases, complete withdrawal from DIC is not possible. Has been.

  On the other hand, new DIC treatment attempts include thrombomodulin preparations, hirudin, and anti-PAF agents. TFPI (Tissue Factor Pathway Inhibitor) and FXa selective inhibitors are attracting attention as anticoagulant and antithrombotic agents that can be administered orally. Moreover, as neutralizing the activity of TF, WO88 / 07543 discloses a mouse anti-human TF monoclonal antibody, and WO96 / 40921 discloses a humanized anti-human TF antibody.

  A mouse anti-human TF monoclonal antibody can be expected to be a safe and effective therapeutic drug that does not show bleeding tendency associated with the main drug effect in DIC. However, mouse monoclonal antibodies are highly immunogenic in humans (sometimes referred to as “antigenic”), which limits the therapeutic value of mouse monoclonal antibodies in humans. For example, when a mouse antibody is administered to a human, it can be metabolized as a foreign substance. Therefore, the half-life of a mouse antibody in humans is relatively short, and the expected effect cannot be fully exhibited. Furthermore, human anti-mouse antibodies (HAMA) raised against administered mouse antibodies elicit immune responses that are inconvenient and dangerous for the patient, such as serum sickness or other allergic reactions. Therefore, mouse monoclonal antibodies cannot be administered frequently to humans.

  In order to solve these problems, a method for reducing the immunogenicity of non-human-derived antibodies, for example, mouse-derived monoclonal antibodies, has been developed. One of them is a method for producing a chimeric antibody in which the variable region (V region) of an antibody is derived from the original mouse monoclonal antibody and the constant region (C region) is derived from an appropriate human antibody.

  Since the resulting chimeric antibody contains the variable region of the original mouse antibody in its complete form, it can be expected to bind to the antigen with the same specificity as the original mouse antibody. Furthermore, the ratio of amino acid sequences derived from non-humans is substantially reduced in chimeric antibodies, and therefore is expected to be less immunogenic than the original mouse antibodies, but nevertheless against mouse variable regions. An immune response can occur (LoBuglio, A., F. et al. Proc. Natl. Acad. Sci. USA, 86, 4220-4224, 1989).

  The second method for reducing the immunogenicity of mouse antibodies is more complex, but is expected to further reduce the potential immunogenicity of mouse antibodies. In this method, only a complementarity determining region (CDR) is transplanted from a variable region of a mouse antibody into a human variable region to produce a “reshaped” human variable region. However, if necessary, in order to bring the CDR structure of the reshaped human variable region closer to that of the original mouse antibody, the partial amino acid sequence of the framework region (FR) supporting CDR is In some cases, the variable region of an antibody is transplanted into a human variable region. These humanized reshaped human variable regions are then linked to human constant regions. The part derived from the non-human amino acid sequence of the finally reconstituted humanized antibody is only CDR and a part of FR. CDRs are composed of hypervariable amino acid sequences, which do not show species-specific sequences.

  For humanized antibodies, see also Riechmann, L. et al. Nature, 332, 323-327, 1988; Verhoeye, M. et al. Science, 239, 1534-1536, 1998; Kettleborough, CA et al. Protein Engng., 4, 773 -783, 1991; Maeda, H., Human Antibodies and Hybridoma, 2, 124-134, 1991; Gorman, SD et al. Proc. Natl. Acad. Sci. USA, 88, 4181-4185, 1991; Tempest, PR, Bio / Technology, 9, 266-271, 1991; Co, MS et al. Proc. Natl. Acad. Sci. USA, 88, 2869-2873, 1991; Cater, P. et al. Proc. Natl. Acad. Sci. USA, 89, 4285-4289, 1992; Co, MS et al. J. Immunol., 148, 1149-1154, 1992; and Sato, K., et al. Cancer Res., 53, 851-856, 1993.

  In the conventional humanization technology, a part of the framework region (FR) includes an amino acid sequence grafted from a variable region of a mouse antibody to a human variable region. Therefore, when administered as a therapeutic agent in humans, there is a risk that an antibody against a site having an amino acid sequence that is one or several amino acids in the variable region but does not exist in humans can be produced. In order to avoid this danger, a third humanization technology was devised. That is, for four FRs (FR1 to 4) necessary to maintain the three-dimensional structure of three CDRs, a human antibody highly homologous to the mouse antibody FR present in the database, with one FR as a unit. It is a method of replacing FR. At this time, several types of FRs are selected from human antibodies existing on the database, and a highly humanized antibody is produced by sequentially replacing (shuffle).

  By doing so, it is possible to produce humanized antibodies having all amino acid sequences derived from human antibodies for all FRs except CDRs in the variable region. For this reason, humanized antibodies carrying mouse CDRs should no longer have stronger immunogenicity than human antibodies containing human CDRs.

  As described above, humanized antibodies are expected to be useful for therapeutic purposes, but there is no uniform method universally applicable to any antibody in the method for producing humanized antibodies. In order to prepare a humanized antibody exhibiting sufficient binding activity and neutralizing activity for a specific antigen, various devices are required (for example, Sato, K. et al., Cancer Res., 53, 851 -856, 1993).

  The present invention relates to a human / mouse chimeric antibody comprising a variable region (V region) of a mouse monoclonal antibody against human TF and a constant region (C region) of a human antibody, and a light chain (L chain) V of a mouse monoclonal antibody against human TF. A humanized antibody in which the region and the heavy chain (H chain) V region are determined in a human phase, a humanized antibody, the L chain and H chain of the antibody, and the L chain or H chain of the antibody It aims at providing the fragment | piece of the V area | region which comprises. It is another object of the present invention to provide a method for producing a humanized antibody against human TF.

  Another object of the present invention is to provide a DNA encoding the above-described antibody, particularly a V region fragment thereof, and a DNA encoding an L chain or H chain containing the V region fragment. It is another object of the present invention to provide a recombinant vector containing the DNA and a host transformed with the vector. Another object of the present invention is to provide a pharmaceutical composition comprising a humanized antibody against human TF as an active ingredient and a therapeutic agent for disseminated intravascular coagulation syndrome (DIC).

As a result of diligent research based on the above problems, the present inventors have succeeded in obtaining an antibody having reduced human immunogenicity of a mouse monoclonal antibody against human TF, and a new humanized antibody. The production method was developed and the present invention was completed.
That is, the present invention relates to a chimeric H chain comprising an H chain C region of a human antibody and a fragment of the H chain V region of a mouse monoclonal antibody against human TF. Examples of the H chain V region include those containing the amino acid sequence represented by SEQ ID NO: 9, and examples of the C region include those of the Cγ4 region.

Furthermore, the present invention relates to a chimeric L chain comprising a fragment of the L chain C region of a human antibody and the L chain V region of a mouse monoclonal antibody against human TF. Examples of the L chain V region include those containing the amino acid sequence represented by SEQ ID NO: 15, and examples of the L chain C region include those of the Cκ region.
Furthermore, the present invention relates to a human mouse chimeric monoclonal antibody against human TF comprising the chimeric H chain and the chimeric L chain.

  Furthermore, the present invention includes a framework region (FR) 1 to 4 of the H chain V region of a human antibody, and a complementarity determining region (CDR) 1 to 3 of the H chain V region of a mouse monoclonal antibody against human TF. It relates to a fragment of the H chain V region of a humanized antibody. As CDR1-3, what contains the amino acid sequence represented by sequence number 133-135, respectively is mentioned. Examples of human antibody H chain V region FR1 include human antibody FR1 having a homology of 40% or more with mouse antibody H chain V region FR1, and FR2 includes mouse antibody H chain V region FR2. Human antibody FR2 having a homology of 40% or more is mentioned, FR3 is a human antibody FR3 having a homology of 40% or more with the H chain V region FR3 of a mouse antibody, and FR4 is a mouse antibody H2. And human antibody FR4 having 40% or more homology with the chain V region FR4.

  Preferably, FR1 of the H chain V region of the human antibody includes a human antibody FR1 having 50% or more homology with the H chain V region FR1 of the mouse antibody, and FR2 includes the H chain V region of the mouse antibody. Human antibody FR2 with 70% or more homology with FR2 can be mentioned, FR3 can include human antibody FR3 with 65% or more homology with mouse antibody H chain V region FR3, and FR4 with mouse Examples include human antibody FR4 having 80% or more homology with the H chain V region FR4 of the antibody. As specific examples, FR1 of the H chain V region of a human antibody includes human antibody L39130, FR2 includes human antibody L39130, human antibody P01742 and human antibody Z80844, and FR3 includes Human antibody L39130, human antibody Z34963, human antibody P01825, human antibody M62723, human antibody Z80844, human antibody L04345, human antibody S78322, human antibody Z26827, human antibody U95239 and human antibody L03147, and FR4 includes human antibody L39130 Is mentioned.

  Preferred examples of human antibody H chain V region FR1 include human antibody L39130, FR2 includes human antibody L39130, and human antibody Z80844, and FR3 includes human antibody Z34963, human. Examples include antibody M62723 and human antibody U95239, and FR4 includes human antibody L39130. Further preferred examples include human antibody L39130 as FR1 of the H chain V region of human antibody, human antibody L39130 as FR2, and human antibody Z34963 and human antibody U95239 as FR3. FR4 includes human antibody L39130.

Furthermore, the present invention is according to Kabat's specification (Kabat, EA et al., US Dept. Health and Human Services, US Government Printing Offices, 1991) as a number in the framework area.
Furthermore, the present invention relates to a humanized version comprising any amino acid sequence represented by SEQ ID NOs: 30, 40, 42, 50, 52, 58, 60, 64, 70, 72, 76, 78, 82, 84. It relates to a fragment of the H chain V region of an antibody.

  Furthermore, the present invention relates to a fragment of the L chain V region of a humanized antibody comprising FR1 to FR4 of the L chain V region of a human antibody and CDR1 to 3 of the L chain V region of a mouse monoclonal antibody against human TF. As CDR1-3, what contains the amino acid sequence represented by sequence number 136-138, respectively is mentioned. Examples of the FR1 of the L chain V region of the human antibody include a human antibody FR1 having a homology of 40% or more with the L chain V region FR1 of the mouse antibody. Human antibody FR2 having a homology of 40% or more is mentioned, FR3 is a human antibody FR3 having a homology of 40% or more with the L chain V region FR3 of a mouse antibody, and FR4 is a mouse antibody L2. And human antibody FR4 having 40% or more homology with the chain V region FR4.

  Preferably, FR1 of the L chain V region of the human antibody includes a human antibody FR1 having a homology of 75% or more with the L chain V region FR1 of the mouse antibody, and FR2 includes the L chain V region of the mouse antibody. The human antibody FR2 has a homology with FR2 of 80% or higher, FR3 includes a human antibody FR3 with a homology with the L chain V region FR3 of a mouse antibody of 70% or higher, and FR4 has a mouse Examples include human antibody FR4 having 80% or more homology with the L chain V region FR4 of the antibody. Specific examples of the human antibody L chain V region FR1 include human antibody Z37332, FR2 includes human antibody Z37332 and human antibody X93625, and FR3 includes human antibody Z37332, human antibody. Examples include antibodies S68699 and P01607, and FR4 includes human antibody Z37332. More preferable examples include human antibody Z37332 as FR1 of the L chain V region of a human antibody, human antibody X93625 as FR2, and human antibody S68699 as FR3, and human as FR4. The antibody Z37332 is mentioned.

Furthermore, the present invention is according to Kabat's specification (Kabat, EA et al., US Dept. Health and Human Services, US Government Printing Offices, 1991) as a number in the framework area.
Furthermore, the present invention relates to a fragment of the L chain V region of a humanized antibody comprising the amino acid sequence represented by SEQ ID NO: 93, 99, 101, 107 or 109.

  Furthermore, the present invention relates to the H chain of a humanized antibody against human TF, comprising a fragment of the H chain V region of the humanized antibody and a fragment of the H chain C region of the human antibody. Here, C region is Cγ4 region, and human antibody-derived FR1-4 are human antibody L39130 (FR1), human antibody L39130 (FR2), human antibody Z34963 (FR3) or human antibody U95239 (FR3), human, respectively. Examples derived from the antibody L39130 (FR4) and CDR1-3 include those containing the amino acid sequences represented by SEQ ID NOs: 133-135, respectively.

Furthermore, the present invention relates to a humanized antibody L chain against human TF, comprising a fragment of the L chain V region of the humanized antibody and a fragment of the L chain C region of the human antibody. Here, C region is Cκ region, and human antibody-derived FR1-4 are derived from human antibody Z37332 (FR1), human antibody X93625 (FR2), human antibody S68699 (FR3), and human antibody Z37332 (FR4), respectively. And those containing the amino acid sequences represented by SEQ ID NOs: 136 to 138, respectively.
Furthermore, the present invention relates to a humanized antibody against human TF, comprising the L chain and H chain of the humanized antibody.

  Furthermore, the present invention relates to a method for producing a humanized antibody against human TF. The humanized preparation method relates to a method for selecting FR1 to FR4 that supports the structure of CDR1 to CDR3 that is an antigen recognition site of H chain or L chain. That is, using each FR as a unit, select multiple human antibody FRs that are highly homologous to mouse antibody-derived FRs, and sequentially replace the FRs to create humanized antibodies with the desired activity. Regarding the method.

More specifically, in an example of the method for producing a humanized antibody of the present invention, immunogenicity having a non-human-derived complementarity determining region (CDR) and a natural human antibody-derived framework region (FR) is reduced. In a method for producing a natural humanized antibody,
(1) Prepare a non-human monoclonal antibody reactive to the target antigen,
(2) preparing a plurality of human antibodies having high homology to the FR amino acid sequence in the monoclonal antibody of (1),
(3) Substituting four FRs of one type of human antibody in (2) above with corresponding FRs of the non-human monoclonal antibody of (1) to produce a first humanized antibody,
(4) measuring the ability of the humanized antibody prepared in (3) above to bind to the antigen or neutralize the biological activity of the antigen;
(5) One to three FRs in the humanized antibody prepared in (3) above correspond to human antibodies prepared in (2) that are different from those used in (3). Replace with FR to make a second humanized antibody,
(6) Ability to neutralize antigen binding or biological activity of the second humanized antibody prepared in (5) and the first humanized antibody obtained in (3) above And select humanized antibodies that exhibit favorable activity,
(7) Steps (3) to (6) above are performed for the humanized antibody selected in (6), and (8) the activity is equivalent to that of the non-human monoclonal antibody in (1). Repeating steps (3) to (6) until a humanized antibody is obtained,
It is characterized by that.

  Once a humanized antibody having a certain level of neutralizing activity against human TF is obtained, a homology search is performed for a specific FR in the V region of the H and L chains, and a human antibody with higher homology is obtained. You can choose. The human antibody obtained here is added to the plurality of human antibody groups in the above step (2), and steps (3) to (6) are further repeated to obtain a humanized antibody having a desired activity. I can do it.

  Furthermore, the present invention relates to a DNA encoding a fragment of the H chain V region or the L chain V region of a mouse monoclonal antibody against human TF. Examples of the amino acid sequences and DNAs of the H chain V region fragment and the L chain V region fragment include those containing the nucleotide sequence represented by SEQ ID NO: 9 or 15, respectively.

  Furthermore, the present invention relates to a DNA encoding the chimeric H chain or the chimeric L chain. Examples of the DNA encoding the H chain include those containing the base sequence represented by SEQ ID NO: 9, and examples of the DNA encoding the L chain include those containing the base sequence represented by SEQ ID NO: 15. .

  Furthermore, the present invention relates to a DNA encoding the H chain V region fragment or the L chain V region fragment of the humanized antibody. As the DNA encoding the fragment of the H chain V region, any one of the nucleotide sequences represented by SEQ ID NOs: 29, 39, 41, 49, 51, 57, 59, 63, 69, 71, 75, 77, 81 or 83 DNAs that encode fragments of the L chain V region include those containing the nucleotide sequence represented by SEQ ID NO: 92, 98, 100, 106 or 108.

Furthermore, the present invention relates to DNA encoding the H chain of a humanized antibody.
Furthermore, the present invention includes DNA encoding any amino acid sequence represented by SEQ ID NOs: 30, 40, 42, 50, 52, 58, 60, 64, 70, 72, 76, 78, 82, or 84. The invention relates to H-chain DNA of humanized antibodies. Examples of the DNA include those containing any one of the nucleotide sequences represented by SEQ ID NOs: 29, 39, 41, 49, 51, 57, 59, 63, 69, 71, 75, 77, 81, or 83.
Furthermore, the present invention relates to a DNA encoding the L chain of the humanized antibody.

Furthermore, the present invention is L-chain DNA of a humanized antibody encoding the amino acid sequence represented by SEQ ID NO: 93, 99, 101, 107 or 109. Examples of the DNA include those containing the base sequence represented by SEQ ID NO: 92, 98, 100, 106 or 108.
Furthermore, the present invention relates to a recombinant vector containing any one of the above DNAs.
Furthermore, the present invention relates to a transformant transformed with the recombinant vector.
Furthermore, the present invention provides a method for producing a chimeric antibody or humanized antibody against human TF, comprising culturing the transformant and collecting a chimeric antibody or humanized antibody against human TF from the resulting culture. About.
Furthermore, the present invention relates to a pharmaceutical composition or a DIC therapeutic agent comprising the humanized antibody as an active ingredient.

Hereinafter, the present invention will be described in detail. 1. Production of mouse monoclonal antibody against human TF
A mouse monoclonal antibody against TF is a clone that produces an antibody that specifically inhibits TF activity by preparing a hybridoma by cell fusion of an antibody-producing cell obtained from an animal immunized with an antigen and a myeloma cell. It is prepared by selecting.

  Specifically, hybridomas were prepared by fusing spleen cells of mice immunized with TF purified from human placenta as an antigen and myeloma cells. Hybridoma screening is based on cell-ELISA using the TF-expressing cell line J82 as the antibody's ability to bind to TF, and neutralizing ability against TF exhibits inhibitory activity against the activation of coagulation factor X (Factor X: FX). The measurement system was used as an index. As a result, we succeeded in establishing a hybridoma producing 6 types of antibodies that strongly inhibit the FX activation of TF / VIIa complex.

(1) Preparation of antigen Examples of TF used for immunization of animals include a partial peptide of TF amino acid sequence prepared by recombinant DNA method or chemical synthesis, or TF derived from human placenta. For example, TF derived from human placenta purified according to the method of Ito et al. (Ito T. et al. J. Biochem. 114, 691-696, 1993) can be used as an antigen.
The obtained human TF is mixed with an adjuvant and used as an antigen. Examples of adjuvants include Freund's complete adjuvant and Freund's incomplete adjuvant, and any of these may be mixed.

(2) Collection of immunity and antibody-producing cells The antigen obtained as described above is administered to non-human mammals such as mammals such as mice, rats, horses, monkeys, rabbits, goats and sheep. Any method can be used for immunization as long as it is an existing method, but it is mainly carried out by intravenous injection, subcutaneous injection, intraperitoneal injection or the like. Immunization intervals are not particularly limited, and immunization is performed at intervals of several days to several weeks, preferably at intervals of 4 to 21 days.
Antibody-producing cells are collected 2-3 days after the final immunization day. Examples of antibody-producing cells include spleen cells, lymph node cells, and peripheral blood cells. Generally, spleen cells are used. The immunization amount of the antigen is 0.1-100 μg per mouse at a time.

(3) Measurement of antibody titer In order to confirm the immune response level of the immunized animal and to select the target hybridoma from the cells after cell fusion treatment, the antibody titer in the blood of the immunized animal or the antibody producing cell The antibody titer in the culture supernatant is measured.
Examples of antibody detection methods include known techniques such as EIA (enzyme immunoassay), RIA (radioimmunoassay), ELISA (enzyme linked immunosorbent assay) and the like.

(4) Cell fusion As myeloma (myeloma) cells to be fused with antibody-producing cells, cell lines derived from various animals such as mice, rats, humans and generally available to those skilled in the art are used. As the cell line to be used, those that have drug resistance and have the property that they cannot survive in a selective medium (for example, HAT medium) in an unfused state but can survive only in a fused state. An 8-azaguanine resistant strain is generally used, and this cell line is deficient in hypoxanthine-guanine-phosphoribosyltransferase and cannot grow in hypoxanthine / aminopterin / thymidine (HAT) medium.

  Myeloma cells are known in various known cell lines such as P3 (P3x63Ag8.653) (J. Immunol., 123, 1548-1550, 1979), P3x63Ag8.1 (Current Topics in Micro biology and Immunology, 81, 1 -7, 1978), NS-1 (Kohler, G. and Milstein, C., Eur. J. Immunol. 6, 511-519, 1976), MPC-11 (Margulies, DH Cell, 8, 405-415, 1976), SP2 / 0 (Shulman, M. et al. Nature, 276, 269-270, 1978), F0 (de St. Groth, SF et al. J. Immunol. Methods, 35, 1-21, 1980), S194 (Trowbride , IS, J. Exp. Med., 148, 313-323, 1978), R210 (GGalfre, G. et al. Nature, 277, 131-133, 1979) and the like are preferably used.

  Antibody-producing cells are obtained from spleen cells, lymph node cells, and the like. That is, the spleen, lymph nodes, etc. are removed or collected from the various animals, and these tissues are crushed. The obtained crushed material is suspended in a medium or buffer such as PBS, DMEM, RPMI1640, etc., filtered through a stainless mesh, etc., and then centrifuged to prepare the desired antibody-producing cells.

Next, the myeloma cell and the antibody-producing cell are fused.
Cell fusion is performed at 30-37 ° C. in a medium for animal cell culture such as MEM, DMEM, RPMI1640 medium, and myeloma cells and antibody-producing cells at a mixing ratio of 1: 1-1: 10 in the presence of a fusion promoter. In contact for 1-15 minutes. In order to promote cell fusion, a fusion promoter or fusion virus such as polyethylene glycol having an average molecular weight of 1,000 to 6,000, polyvinyl alcohol, or Sendai virus can be used. Alternatively, antibody-producing cells and myeloma cells can be fused using a commercially available cell fusion device utilizing electrical stimulation (for example, electroporation).

(5) Selection and cloning of hybridomas Select the desired hybridomas from the cells after cell fusion treatment. Examples of the method include a method utilizing selective growth of cells in a selective medium. That is, after the cell suspension is diluted with an appropriate medium, it is spread on a microtiter plate, a selective medium (HAT medium or the like) is added to each well, and thereafter, the selective medium is appropriately replaced and cultured. As a result, growing cells can be obtained as hybridomas.

Hybridoma screening is performed by limiting dilution, fluorescence excitation cell sorter, or the like, and finally a monoclonal antibody-producing hybridoma is obtained.
Selection of a hybridoma producing a monoclonal antibody can be performed by combining various measurement systems. For example, production of a monoclonal antibody having a desired activity by combining an antigen recognition measurement system such as Cell-ELISA, a TF neutralization activity measurement system using Factor Xa activity as an indicator, or a plasma clotting inhibition activity measurement system. Obtain a hybridoma. In this way, for example, monoclonal antibody-producing hybridomas such as ATR-2, ATR-3, ATR-4, ATR-5, ATR-7 and ATR-8 can be obtained.

(6) Collection of monoclonal antibody Examples of a method for collecting a monoclonal antibody from the obtained hybridoma include a normal cell culture method and ascites formation method.
In the cell culture method, the hybridoma is cultured in an animal cell culture medium such as RPMI1640 medium containing 10 to 20% fetal bovine serum, DMEM medium, or serum-free medium (for example, 37 ° C., 5% CO ₂ concentration). Incubate for 2 to 14 days, and obtain antibody from the culture supernatant.

In the ascites formation method, a hybridoma is administered into the abdominal cavity of an animal of the same kind as a mammal derived from myeloma cells, and the hybridoma is proliferated in large quantities. And ascites or serum is collected 1 to 4 weeks later.
When antibody purification is required in the antibody collection method, purification is performed by appropriately selecting a known method such as ammonium sulfate salting-out method, ion exchange chromatography, affinity chromatography, or a combination thereof. To do.

2. Cloning of DNA encoding mouse monoclonal antibody V region against human TF (i) Preparation of mRNA Recovered for cloning of DNA encoding mouse monoclonal antibody H chain and L chain V regions against human TF Methods known from hybridomas, such as guanidine-ultracentrifugation (Chirgwin, JM et al., Biochemistry, (1979), 18, 5294-5299), AGPC method (Chomczynski, P et al. (1987), 162, 156-159), etc. Prepare total RNA, and prepare mRNA using oligo (dT) -cellulose span column attached to mRNA Purification Kit (Pharmacia Biotech). In addition, mRNA can be prepared without extraction of total RNA by using QuickPrep mRNA Purification Kit (Pharmacia Biotech).

(Ii) Preparation and amplification of cDNA From the mRNA obtained in (i) above, cDNAs in the V region of the L chain and the H chain are respectively synthesized using reverse transcriptase. For synthesis of cDNA, an Oligo-dT primer or an appropriate primer that hybridizes with the L chain C region or the H chain C region (for example, a cDNA synthesis primer attached to the kit) can be used.
In the cDNA synthesis reaction, the mRNA and the primer are mixed, and the reaction is performed at 52 ° C. for 30 minutes in the presence of reverse transcriptase.

  Amplification of cDNA was performed using 5'-RACE method (Frohman, MA et al. Proc. Natl. Acad. USA 85, 8998-9002, 1988, using 5'-Ampli FINDER RACE kit (CLONTECH) for both L and H chains. Belyavsky, A. et al. Nucleic Acids Res. 17, 2919-2932, 1989) can be performed by PCR (polymerase chain reaction). That is, a cDNA adapter is ligated to both ends of the double-stranded cDNA synthesized above, and DNAs encoding the fragments of the H chain V region and L chain V region (hereinafter referred to as DNAs encoding the fragments of the L chain V region are referred to as “L PCR may be carried out for “strand V region DNA” or “DNA encoding L chain V region” (the same applies to H chain V region and the like).

  As primers for amplifying the H chain V region DNA, for example, the 5′-side primer includes Adapter primer 1 included in the kit, and the 3′-side primer includes the mouse antibody H chain constant region MHC-G1 primer (SEQ ID NO: 1) (ATR-2, ATR-3, ATR-4 and ATR-5) (Cγ1 region) or MHC-G2a primer (SEQ ID NO: 2) (ATR-7 and ATR-8) (Cγ2a region) (STJones et al. Biotechnology, 9, 88, 1991). In addition, as primers for amplifying the DNA of the L chain V region, for example, the 5′-side primer includes Adapter primer 1 attached to the kit and the 3′-side primer includes the mouse antibody L chain κ chain constant region (Cκ region). ) Primers (for example, MKC primers having the base sequence represented by SEQ ID NO: 3) can be used.

(Iii) DNA purification and nucleotide sequence determination
The PCR product is subjected to agarose gel electrophoresis according to a known method to cut out the target DNA fragment, and then the DNA is recovered and purified, and then ligated to vector DNA.
Purification of DNA is performed by extraction with phenol and chloroform (J. Sambrook, et al. “Molecular Cloning”, Cold Spring Harbor Laboratory Press, 1989) or using a commercially available kit (eg GENECLEAN II; BIO101). Known vector DNA (eg, pUC19, Bluescript, etc.) can be used for retaining the DNA fragment.

  The DNA and the vector DNA are ligated using a known ligation kit (Takara Shuzo) to obtain a recombinant vector. Next, the resulting recombinant vector is introduced into Escherichia coli JM109 competent cell (Nippon Gene) and then ampicillin resistant colonies are selected, and vector DNA is prepared based on a known method (J. Sambrook, et al. “Molecular Cloning”). ", Cold Spring Harbor Laboratory Press, 1989). The base sequence of the target DNA is determined by digesting the vector DNA with a restriction enzyme and then using a known method (eg, dideoxy method) (J. Sambrook, et al. “Molecular Cloning”, Cold Spring Harbor Laboratory Press, 1989). ). In the present invention, an automatic base sequencer (DNA Sequencer 373A, Perkin-Elmer) can be used.

(Iv) Complementarity determining region (CDR)
The H chain V region and the L chain V region form an antigen-binding site, and their general structures are similar to each other. That is, each of the four framework region (FR) portions is linked by three hypervariable regions, ie, complementarity determining regions (CDR). While the amino acid sequence of FR is relatively well conserved, the variability of the amino acid sequence of the CDR region is extremely high (Kabat, EA et al., “Sequence of Proteins of Immunological Interest” US Dept. Health and Human Services, 1983. ).

  Many portions of the four FRs have a β-sheet structure, so that the three CDRs form a loop. CDRs may in some cases form part of the β-sheet structure. Thus, the three CDRs are held sterically very close to each other by the FR, and the FR forms an antigen binding site with the three CDRs of the paired region.

  Based on these facts, the antibody amino acid sequence database created by Kabat et al. For the variable region of mouse monoclonal antibodies against human TF ("Sequence of Proteins of Immunological Interest" US Dept. Health and Human Services, 1983) It is possible to find a CDR region by applying homology and examining the homology.

  If the sequence of the CDR region is within a range in which the binding activity and neutralizing activity to human TF are maintained when a humanized antibody is produced, a modified form by insertion, substitution or deletion is also included in the present invention. Examples thereof include those having a sequence having 90 to 100% homology with each CDR region of SEQ ID NOs: 133-138 or with each CDR region in the V regions of SEQ ID NOs: 139-141, 143-144, 145-147, and 149-150. Preferable examples include sequences having a homology of 95 to 100%. More preferred is a sequence having a homology of 98 to 100%.

3. Preparation of Expression Vector for Chimeric Antibody Mouse L chain of mouse monoclonal antibody (hereinafter referred to as “mouse L chain” for mouse when referring to antibody L chain or H chain, “human H chain” for human antibody H chain) When DNA fragments encoding the H chain V region are cloned, the DNA encoding the mouse V region is linked to the DNA encoding the human antibody constant region. To obtain a chimeric anti-human TF antibody.

The basic method for making a chimeric antibody is to link the mouse leader sequence and V region sequence present in the cloned cDNA to the sequence encoding the human antibody C region already present in the mammalian cell expression vector. Comprising doing. Alternatively, the method comprises ligating a mouse leader sequence and V region sequence present in the cloned cDNA to a sequence encoding a human antibody C region, and then ligating it to a mammalian cell expression vector.
The fragment of the human antibody C region can be of any human antibody H chain C region and human antibody L chain C region, for example, for human H chain, Cγ1, Cγ2, Cγ3 or Cγ4, and For the L chain, Cλ or Cκ can be mentioned, respectively.

  For the production of chimeric antibodies, expression vectors containing DNA encoding mouse H chain V region and human H chain C region under expression control region such as enhancer / promoter system, and expression such as enhancer / promoter system A single expression vector (for example, refer to WO94 / 11523) containing DNA encoding mouse L chain V region and human L chain C region is prepared under the control of the control region. Next, a host cell such as a mammalian cell is co-transformed with this expression vector, and the transformed cell is cultured in vitro or in vivo to produce a chimeric antibody (see, eg, WO91 / 16928) . As a single expression vector, IgG1κ type antibody expression vector N5KG1 (V) and IgG4κ type antibody expression vector N5KG4P can be used.

(I) Construction of chimeric antibody H chain In the chimeric antibody H chain expression vector, a cDNA encoding the mouse H chain V region is introduced into an appropriate expression vector containing DNA encoding the H chain C region of the human antibody. Can be obtained. Examples of the H chain C region include Cγ1, Cγ2, Cγ3, and Cγ4 regions.

  Here, when inserting the cDNA encoding the mouse H chain V region into an expression vector, an appropriate base sequence can be introduced into the cDNA by PCR. For example, a Kozak consensus sequence (Kozak, M. et al., J. Co., Ltd.) has an appropriate restriction enzyme recognition sequence at the 5′-end of the cDNA and immediately before the start codon of the cDNA to improve transcription efficiency. Mol. Biol., 196, 947-950, 1987) and a PCR primer designed to have an appropriate restriction enzyme recognition sequence at the 3′-end of the cDNA. By carrying out PCR, these appropriate base sequences can be introduced into an expression vector.

  A chimeric H chain expression vector containing a DNA encoding the H chain C region (Cγ1 or Cγ4) after treating the cDNA encoding the mouse H chain V region thus constructed with an appropriate restriction enzyme and then inserting it into the expression vector. Build up.

(Ii) Construction of expression vector containing cDNA encoding chimeric antibody L chain κ chain The L chain expression vector of chimeric antibody is a cDNA encoding mouse L chain V region, and a DNA encoding human antibody L chain C region. Can be obtained by introducing it into an appropriate expression vector. Examples of the L chain C region include Cκ and Cλ regions.

  In constructing an expression vector containing cDNA encoding the mouse L chain V region, an appropriate base sequence can be introduced into the cDNA by PCR. For example, a PCR primer designed to have an appropriate restriction enzyme recognition sequence at the 5'-end of the cDNA and a Kozak consensus sequence to improve transcription efficiency, and suitable at the 3'-end These appropriate base sequences are introduced into the cDNA by performing PCR using PCR primers designed to have a recognition sequence for a restriction enzyme.

  A cDNA encoding the mouse L chain V region thus constructed is treated with an appropriate restriction enzyme and then inserted into the above expression vector to obtain a chimeric L chain expression vector containing a DNA encoding the L chain C region (Cκ region). To construct.

4. Production of humanized antibody (1) Homology search with human antibody In order to produce humanized antibody in which CDR of mouse monoclonal antibody is transplanted to human antibody, mouse monoclonal antibody FR and human antibody It is desirable that a high degree of homology exists with the FR. Therefore, the V regions of the H and L chains of the mouse anti-human TF monoclonal antibody are compared with the V regions of all known antibodies whose structures have been elucidated using the Data Bank. At the same time, Kabat et al. Classified human antibody subgroups (HSG: Human subgroup) (Kabat, EA et al., US Dep. Health and Human Services, US Government). Comparison with Printing Offices, 1991).

  In the case of the human H chain V region, it can be classified into HSGI to III by the HSG classification by Kabat et al., For example, the H chain V region of the mouse anti-human TF monoclonal antibody ATR-5 has a consensus sequence of HSGI and 67.8. % Homology. On the other hand, the human L chain κ chain V region can be classified into HSGI to IV according to HSG classification by Kabat et al., For example, the L chain κ chain V region of the mouse anti-human TF monoclonal antibody ATR-5 is HSGI. Has 72.3% homology with the consensus sequence.

  When mouse antibodies are humanized by conventional techniques, if necessary, mice that support CDRs are used to bring the CDR structure of the humanized V region closer to that of the original mouse antibody. A part of the amino acid sequence of the FR of the V region of the antibody may be transplanted into the FR of the human V region. However, there is no fixed rule as to which amino acids of the mouse antibody V region FR should be transplanted into the human antibody V region FR. Therefore, identifying the essential amino acids to retain the structure of the CDR requires a lot of effort.

  In addition, there is a risk that a human antibody against an amino acid sequence grafted from the V region of a mouse antibody to a human V region can be formed in a part of FR. In the present invention, in order to make all amino acid sequences excluding CDR in a humanized antibody to be amino acid sequences derived from human antibodies, four FRs (FR1 to FR4) necessary for retaining the three-dimensional structure of CDR are used. ), A human antibody FR having a high homology with a mouse antibody FR present in the database was searched using one FR as a unit. The database and homology search results for each FR of the V region of the H chain and L chain of the monoclonal antibody ATR-5 are shown below.

(2) Design of DNA encoding humanized antibody V region The first step in designing DNA encoding humanized antibody V region is to select each FR of the human antibody V region that is the basis of the design. is there. Further, in FR shuffle, it is necessary to select a human antibody V region FR having a high variety in each FR.

  In the present invention, regarding the monoclonal antibody ATR-5, as for the H chain, from the results of homology search between the entire mouse antibody H chain V region and each FR, three types of FR2 and ten types of human antibodies for FR3 V region FR was selected. For the L chain, three types of human antibody V regions FR can be selected for FR2 and three types of human antibody V regions FR can be selected from the results of homology search between the entire mouse antibody L chain V region and each FR.

  Both humanized H chain and L chain V region, as the first version (version a: ver.a), human antibody H chain having high homology with the H chain and L chain V region of mouse monoclonal antibody ATR-5, respectively L chain V region L39130 and Z37332 can be selected. In preparing these humanized antibodies, appropriate restriction enzyme recognition sites can be designed within each CDR and at appropriate sites in the FR so that FR shuffle can be easily performed. This makes it possible to easily replace only one of the FRs.

For example, humanized H chain CDR1 restriction enzyme EcoT22I recognition site, CDR2 restriction enzyme BalI recognition site, CDR3 restriction enzyme NcoI recognition site and FR3 restriction enzyme XhoI recognition site, for example, humanized L chain CDR1 restriction enzyme AflII recognition A restriction enzyme SpeI recognition site of CDR2, a restriction enzyme PstI recognition site of CDR3, and a restriction enzyme AccIII recognition site of FR3.
Based on the version a designed in this way, FR shuffle is performed on each FR, and a humanized antibody having a desired activity can be prepared.

(3) Preparation of humanized antibody V region fragment In the humanized antibody of the present invention, the C region of the antibody and the FR of the V region are derived from a human antibody, and the CDR of the V region is derived from a mouse antibody. belongs to. The fragment of the V region of the humanized antibody of the present invention can be prepared by a technique called CDR-grafting by PCR if a DNA fragment of a human antibody as a template is available. “CDR-grafting” refers to a technique in which a DNA fragment encoding a mouse-derived CDR is prepared and replaced with the CDR of a human antibody as a template.

  In addition, when a DNA fragment of a human antibody as a template is not available, a base sequence registered in a database can be synthesized with a DNA synthesizer, and a humanized antibody V region can be prepared by PCR. Furthermore, if only the amino acid sequence is registered in the database, the codon usage frequency of the antibody reported by Kabat, EA et al. (US Dep. Health and Human Services, US Government Printing Offices, 1991) based on the amino acid sequence. Based on the above, the entire base sequence can be inferred. This base sequence is synthesized with a DNA synthesizer, and a fragment of the humanized antibody V region can be prepared by PCR.

(I) Construction of DNA encoding humanized H chain V region and expression vector In the present invention, a gene encoding the H chain V region of a human antibody used as a template for a humanized antibody is obtained from a database, and human type The entire base sequence of DNA encoding the modified H chain V region can be synthesized by a DNA synthesizer and constructed by PCR. For example, L39130 having high homology with the H chain V region of the mouse anti-human TF monoclonal antibody ATR-5 can be prepared as a humanized H chain V region version “a”. In order to prepare the humanized H chain V region version “a”, for example, the five primers shown in SEQ ID NOs: 22 to 26 and the two external primers shown in SEQ ID NOs: 27 and 28 are used separately.

  CDR-grafting primers hR5Hv1S (SEQ ID NO: 22), hR5Hv2S (SEQ ID NO: 23) and hR5Hv4S (SEQ ID NO: 24) have sense DNA sequences, and CDR grafting primers hR5Hv3A (SEQ ID NO: 25) and hR5Hv5A (SEQ ID NO: 26) ) Has an antisense DNA sequence and each has a complementary sequence of 18-35 bp on each end of the primer. hR5Hv1S is designed to have a Kozak consensus sequence (Kozak, M, et al., J. Mol. Biol. 196, 947-950, 1987) and a SalI recognition site, and hR5Hv5A is designed to have an NheI recognition site. The external primers hR5HvPrS (SEQ ID NO: 27) and hR5HvPrA (SEQ ID NO: 28) have homology with the CDR grafting primers hR5Hv1S and hR5Hv5A.

Using PCR, five primers are assembled to synthesize full-length cDNA, and external primers are added to amplify the DNA. The assembly by PCR means that hR5Hv1S, hR5Hv2S, hR5Hv4S, hR5Hv3A and hR5Hv5A are annealed by their complementary sequences, and a full-length humanized H chain V region DNA is synthesized.
The human antibody H chain C region can be any human H chain C region, and examples thereof include human H chains Cγ1, Cγ2, Cγ3, and Cγ4.

  The humanized antibody H chain V region DNA constructed as described above can be linked to DNA of any human antibody H chain C region, for example, human H chain C region Cγ1 or Cγ4. As described in the construction of the chimeric antibody H chain, it is treated with an appropriate restriction enzyme and then linked to DNA encoding the human H chain C region under an expression control region such as an enhancer / promoter system. An expression vector containing the DNA of the modified H chain V region and the human H chain C region is prepared.

(Ii) Construction of DNA encoding humanized L chain V region and expression vector In the present invention, as in the case of DNA encoding H chain V region, the database of L chain V region gene of human antibody used as template And the entire nucleotide sequence of DNA encoding the humanized L chain V region can be synthesized by a DNA synthesizer and constructed by PCR. For example, Z37332 having high homology with the L chain V region of mouse anti-human TF monoclonal antibody ATR-5 can be prepared as a humanized L chain V region version “a”.

  To create humanized L chain V region version “a”, CDR-grafting primers h5Lv1S (SEQ ID NO: 85) and h5Lv4S (SEQ ID NO: 86) are sense DNA sequences, CDR grafting primers h5Lv2A (SEQ ID NO: 87), h5Lv3A (SEQ ID NO: 88) and h5Lv5A (SEQ ID NO: 89) have antisense DNA sequences and have 20 bp complementary sequences at both ends of each primer. Primer h5Lv1S has a Kozak consensus sequence (Kozak, M, et al., J. Mol. Biol. 196, 947-950, 1987) and a restriction enzyme BglII recognition site, and h5Lv5A appears to have a restriction enzyme SplI recognition site To design. Further, the external primers h5LvS (SEQ ID NO: 90) and h5LvA (SEQ ID NO: 91) have homology with the CDR grafting primers h5Lv1S and h5Lv5A.

As with the humanized H chain V region, PCR can be used to assemble 5 primers, synthesize full-length cDNA, and add external primers to amplify the DNA.
The human antibody L chain C region can be any human L chain C region, and examples thereof include human L chain Cλ and Cκ.
The humanized antibody L chain V region DNA constructed as described above can be linked to any human antibody L chain C region, for example, human L chain Cκ or Cλ region. After treatment with an appropriate restriction enzyme, it is ligated with DNA encoding human L chain κ chain C region under an expression control region such as an enhancer / promoter system, and humanized L chain V region and human L chain κ chain An expression vector containing DNA encoding the C region is prepared.

  Even if a V region fragment of a humanized antibody is prepared as described above, it is not always clear whether the V region fragment has activity as an antibody (antigen binding activity, neutralization activity, etc.). . Therefore, it must be expressed in animal cells such as COS-7 in combination with a humanized H chain and examined for the presence or absence of activity.

(Iii) FR shuffle of H chain and L chain V region of humanized antibody
The present inventor transiently expressed the humanized antibody containing the humanized H chain and L chain V region in animal cells such as COS-7, and examined the antigen binding activity and neutralizing activity. Although it has antigen-binding activity and neutralizing activity, it has been found that it is not sufficient as compared with the chimeric antibody.

  The present inventor can solve this problem by sequentially shuffling each FR of the humanized H chain and L chain V region. Human antibodies for FR shuffle can be selected from existing databases. The FR of the selected human antibody can be synthesized by a DNA synthesizer based on the nucleotide sequence revealed in the database. At this time, as described above, by adding a restriction enzyme recognition sequence designed to CDR or FR, it can be easily shuffled with the FRs of the H chain and L chain V regions of the humanized antibody prepared above. I can do it. By examining the activity of the humanized antibody thus produced, a humanized antibody having the desired antigen-binding activity and neutralizing activity can be obtained.

For example, humanized antibody V region H chain FR3 can be shuffled into FR3 derived from human antibody Z34963 (GenBank, Borretzen M. et al., Proc. Natl. Acad. Sci. USA, 91, 12917-12921, 1994). .
FR-shuffling primers F3RFFS (SEQ ID NO: 35) and F3RFBS (SEQ ID NO: 36) have a sense DNA sequence, and F3RFFA (SEQ ID NO: 37) and F3RFBA (SEQ ID NO: 38) have an antisense DNA sequence. FR-shuffling primers F3RFFS, F3RFBS, F3RFFA, F3RFBA can be synthesized with a DNA synthesizer.

  F3RFFS and F3RFFA, F3RFBS and F3RFBA were annealed and digested with BalI and XhoI, NcoI and XhoI, respectively. These are introduced into a plasmid hATR5Hva / CVIDEC (BalI / NcoI) prepared by digesting with BalI and NcoI, and a plasmid having the correct sequence can be obtained by confirming the base sequence. The plasmid containing the humanized antibody H chain thus obtained is named hATR5Hvb / CVIDEC, and the humanized H chain contained in the plasmid hATR5Hvb / CVIDEC is referred to as version “b”. The base sequence and the corresponding amino acid sequence are shown in SEQ ID NO: 39, and the amino acid sequence of version “b” is shown in SEQ ID NO: 40.

Similarly, FRs derived from other human antibody V region H chains and L chains selected from the database can be shuffled with FRs of humanized antibody V region H chains and L chains.
In order to select a more preferable human antibody for shuffling the FRs of the H chain V region and the L chain V region of the humanized antibody, it can be performed as follows. That is, the combination of the humanized antibody H chain version “b” and the chimeric antibody L chain has neutralizing activity equivalent to that of the chimeric antibody or mouse antibody. However, the combination of the humanized antibody H chain version “b” and the humanized antibody L chain version “a” has a lower neutralization activity than the chimeric antibody or mouse antibody.

In such a case, in order to select a human antibody as a candidate for shuffling FR, for example, humanized antibody H chain version “b” is homologous to FR3 (Accession No. Z34963: SEQ ID NO: 115). A sex search can be performed to select human antibodies highly homologous to this sequence. For example, in the human antibody H chain V region FR3 thus selected, U95239 (SEQ ID NO: 122) and L03147 (SEQ ID NO: 123) can be mentioned.
The amino acid sequences of the humanized antibody V region H chain thus prepared are shown in Tables 3 and 4, and the amino acid sequence of the humanized antibody V region L chain is shown in Table 5.

  Humanized antibody H chain and L chain V region versions constructed as described above are arbitrary human antibody H chain C region and L chain C region, for example, human H chain Cγ4 and human L chain Cκ region DNA, respectively. Can be linked to. After treatment with an appropriate restriction enzyme, it is ligated with DNA encoding human H chain Cγ4 and human L chain Cκ regions under an expression control region such as an enhancer / promoter system, and humanized H chain and L chain V regions. An expression vector comprising DNA encoding each version and DNA encoding human H chain Cγ4 and human L chain Cκ region, respectively, is prepared.

  In addition, a DNA encoding the humanized antibody H chain V region and human H chain C region constructed as described above, and a DNA encoding the humanized L chain V region and human L chain C region are simply used. An expression vector (see, for example, WO94 / 11523), and the host cell is transformed with the vector, and then the transformed host is cultured in vivo or in vitro. The humanized antibody can be produced.

5. Production of chimeric antibody and humanized antibody In order to produce a chimeric antibody or humanized antibody, DNA encoding H chain V region and H chain C region, and L chain V region and L chain C region are used. Encoding DNA can be ligated into a single vector, and appropriate host cells transformed to produce antibodies. That is, for the expression of a chimeric antibody, the mouse leader sequence and mouse H chain V region and human H chain C region present in the cloned cDNA are encoded under the control of an expression control region such as an enhancer / promoter system. DNA and DNA encoding a mouse leader sequence and mouse L chain V region and human L chain C region are introduced into a single expression vector (see, for example, WO94 / 11523).

  For expression of humanized antibody, DNA encoding humanized H chain V region and human H chain C region, and humanized L chain V under the control of an expression control region such as an enhancer / promoter system. The region and the DNA encoding the human L chain C region are introduced into a single expression vector (see, for example, WO94 / 11523). Then, host cells are transformed with these vectors, and then the transformed host is cultured in-vivo or in-vitro to produce the desired chimeric antibody or humanized antibody.

  In addition, two types of expression vectors each containing an H chain V region and an L chain V region can be prepared. That is, for chimeric antibodies, expression vectors containing DNA encoding mouse H chain V region and human H chain C region under the control of expression control regions such as enhancer / promoter system, and expression such as enhancer / promoter system An expression vector containing DNA encoding mouse L chain V region and human L chain C region is prepared under the control region, and humanized antibodies are controlled under the expression control region such as an enhancer / promoter system. An expression vector containing DNA encoding humanized H chain V region and human H chain C region, and humanized L chain V region and human L chain C region under an expression control region such as an enhancer / promoter system An expression vector containing DNA encoding is prepared.

  Alternatively, for chimeric antibodies, DNA encoding mouse H chain V region and human H chain C region, and mouse L chain V region and human L chain under the control of an expression control region such as an enhancer / promoter system. An expression vector comprising DNA encoding the C region is prepared. For humanized antibodies, the humanized H chain V region and human H chain are controlled under the control of an expression control region such as an enhancer / promoter system. An expression vector comprising DNA encoding the C region and DNA encoding the humanized L chain V region and human L chain C region is prepared.

Next, host cells such as mammalian cells are co-transformed with these expression vectors, and the transformed cells are cultured in-vitro or in-vivo to produce chimeric or humanized antibodies (eg, , See WO91 / 16928).
The transformant transformed with the gene encoding the target chimeric antibody or humanized antibody is cultured as described above, and the produced chimeric antibody or humanized antibody is uniformly separated from inside or outside the cell. Can be purified.

  Separation and purification of the chimeric antibody or humanized antibody that is the target protein of the present invention can be performed using a protein A agarose column. In addition, any other separation / purification method used for ordinary proteins may be used, and the method is not limited at all. For example, the chimeric antibody or humanized antibody can be separated and purified by appropriately selecting and combining various types of chromatography, ultrafiltration, salting out, dialysis and the like.

  Any expression system can be used to produce the chimeric or humanized antibody of the invention against human TF. For example, when eukaryotic cells are used, animal cells (for example, established mammalian cell lines), filamentous fungal cells, or yeast cells can be used. When prokaryotic cells are used, bacterial cells (such as E. coli cells) are used. can do. Preferably, the chimeric antibody or humanized antibody of the present invention is expressed in mammalian cells such as COS cells or CHO cells.

  In these cases, a conventional promoter useful for expression in mammalian cells can be used. For example, it is preferable to use a human cytomegalovirus immediate early (HCMV) promoter. Examples of expression vectors containing the HCMV promoter include HCMV-VH-HCγ1, HCMV-VL-HCK, etc., which are derived from pSV2neo (WO92-19759).

  In addition, examples of promoters for gene expression in mammalian cells that can be used for the present invention include viral promoters such as retrovirus, polyomavirus, adenovirus, and simian virus 40 (SV40), and human polypeptides. A promoter derived from a mammalian cell such as chain elongation factor 1α (HEF-1α) may be used. For example, when the SV40 promoter is used, the method of Mulligan et al. (Nature, 277, 108, 1979), and when the HEF-1α promoter is used, the method of Mizushima, S. et al. (Nucleic Acids Research, 18, 5322, 1990) can be easily implemented.

  The origin of replication can be derived from SV40, polyomavirus, adenovirus, bovine papillomavirus (BPV), etc., and the expression vector is selected for gene copy number amplification in the host cell system. Markers may include phosphotransferase APH (3 ') II or I (neo) gene, thymidine kinase (TK) gene, E. coli xanthine-guanine phosphoribosyltransferase (Ecogpt) gene, dihydrofolate reductase (DHFR) gene, etc. it can.

6. Evaluation of antigen-binding activity and neutralizing activity of chimeric antibody and humanized antibody (1) Measurement of antibody concentration by ELISA The concentration of the obtained purified antibody can be measured by ELISA.
An ELISA plate for antibody concentration measurement is prepared as follows. 100 μl of goat anti-human IgGγ antibody (BioSource) prepared at a concentration of, for example, 1 μg / ml in each well of a 96-well plate for ELISA (for example, Maxisorp, NUNC) is immobilized.

Blocking with 200 μl of dilution buffer (hereinafter referred to as DB, eg 50 mM Tris-HCl, 1 mM MgCl ₂ , 0.1 M NaCl, 0.05% Tween 20, 0.02% NaN ₃ , 1% bovine serum albumin (BSA), pH 7.2) Thereafter, the culture supernatant of COS-7 cells or CHO cells expressing the chimeric antibody or humanized antibody, or purified chimeric antibody or humanized antibody is serially diluted and added to each well. Next, 100 μl of alkaline phosphatase-conjugated goat anti-human IgG antibody was added, 1 mg / ml substrate solution (Sigma104, p-nitrophenyl phosphate, SIGMA) was added, and then the absorbance at 405/655 nm was measured using a microplate reader (Bio Rad ) To measure. IgG4κ (The Binding Site) can be used as a standard for concentration measurement.

(2) Measurement of antigen binding ability Cell ELISA plates for antigen binding measurement are prepared as follows. Human bladder cancer cells J82 (ATCC HTB-1) are seeded at a number of 1 × 10 ⁶ cells in 60 wells of a 96-well plate for cell culture. This is cultured for one day in a CO ₂ incubator (RPMI1640 medium containing 10% fetal bovine serum (GIBCO)), and the cells are allowed to adhere. Discard the culture and wash each well twice with 300 μl of PBS. 100 μl of PBS containing 4% paraformaldehyde (hereinafter referred to as PFA / PBS) is added to each well and left on ice for 10 minutes to immobilize the cells.

  PFA / PBS is discarded, and each well is washed twice with 300 μl of PBS and then blocked with 250 μl of DB. Culture supernatant containing chimeric antibody or humanized antibody, or purified chimeric antibody or humanized antibody serially diluted with DB, 100 μl added to each well, incubated at room temperature for 2 hours, washed with RB, Add 100 μl of alkaline phosphatase-conjugated goat anti-human IgGγ antibody (BioSource) diluted 1000-fold with DB. After incubating at room temperature for 1 hour and washing with RB, the substrate solution is added, and then the absorbance at 405/655 nm is measured with a microplate reader (Bio-Rad).

(3) Measurement of neutralizing activity The neutralizing activity of mouse antibodies, chimeric antibodies and humanized antibodies can be measured using the factor Xa production inhibitory activity by human placenta-derived thromboplastin and Thromborel S (Behringwerke AG) as an index. That is, add 10 μl of 1.25 mg / ml Thromborel S and 10 μl of antibody diluted to an appropriate concentration to 60 μl of buffer solution (TBS containing 5 mM CaCl ₂ and 0.1% BSA), and react for 1 hour at room temperature in a 96-well plate. Let

  To this, 10 μl each of 3.245 μg / ml human factor X (Celsus Laboratories) and 82.5 ng / ml human factor VIIa (Enzyme Research) are added, and further reacted at room temperature for 1 hour. Add 10 μl of 0.5 M EDTA to stop the reaction, add 50 μl of chromogenic substrate solution, and measure the absorbance at 405/655 nm with Microplate Reader (Bio Rad). The reaction is allowed to proceed at room temperature for 1 hour, and the absorbance at 405/655 nm is measured again. The neutralization activity is measured by calculating the remaining activity (%) from the change in absorbance for each hour with the change in absorbance for 1 hour without addition of antibody as 100% activity.

  For the chromogenic substrate solution, dissolve the test team chromogenic substrate S-2222 (Chromogenix) according to the package insert, dilute it twice with purified water, and then mix 1: 1 with polybrene solution (0.6 mg / ml hexadimethylin bromide, SIGMA). Prepare.

7. Kinetic analysis of the interaction between humanized antibody and soluble TF
BIACORE can measure the kinetics parameters of the anti-TF antibody of the present invention, that is, the dissociation constant (KD), the dissociation rate constant (kdiss), and the association rate constant (kass).
Recombinant protein G was immobilized on a sensor chip, an antibody was bound to this, and recombinant TF purified as an antigen (soluble TF with FLAG peptide tag attached to 1-219) (hereinafter referred to as soluble TF) The soluble TF prepared at various concentrations is used as the analyte. By calculating kinetic parameters (dissociation rate constant kdiss and association rate constant kass) from the obtained sensorgram, the dissociation constant can be obtained. For kinetic analysis, refer to "Kinetic analysis of monoclonal antibody-antigen interactions with a new biosensor based analytical system" (karlsson, R. et al. (1991) J. Immunol. Methods 145, 229-240.) can do.

The anti-TF antibody of the present invention is more preferable in that it has neutralizing activity as the dissociation constant (KD) is smaller. In the anti-TF antibody of the present invention, the KD value is preferably 2.30 × 10 ⁻⁸ [1 / M] or less, more preferably 2.30 × 10 ⁻⁹ [1 / M] or less, and 1.17 × 10 ⁻ Most preferable is ⁹ [1 / M] or less.
The KD value is determined from two parameters, a dissociation rate constant (kdiss) and a binding rate constant (kass) (KD = kdiss / kass). Therefore, it is clear that if the kdiss value is small and the kass value is large, the KD value is small.

Specifically, in the case of the anti-TF antibody of the present invention, the value of kdiss may be 9.52 × 10 ⁻³ [1 / sec] or less. Preferably, the value of kdiss is 9.52 × 10 ⁻⁴ [1 / sec] or less, and most preferably 6.35 × 10 ⁻⁴ [1 / sec] or less.
On the other hand, the value of kass may be 4.15 × 10 ⁴ [1 / M · sec] or more. Preferably, the kass value is 4.15 × 10 ⁵ [1 / M · sec] or more, and most preferably 4.65 × 10 ⁵ [1 / M · sec] or more.

Furthermore, an anti-TF antibody having a kdiss value of 9.52 × 10 ⁻³ [1 / sec] or less and a kass value of 4.15 × 10 ⁴ [1 / M · sec] or more is also preferable.
More specifically, the anti-TF antibody of the present invention has a KD value in the range of 1.09 × 10 ^{−10 to} 2.30 × 10 ⁻⁸ [1 / M], and 1.09 × 10 ^{−9 to} 2.30 × 10 ⁻⁹ [ 1 / M] is preferable, and those of 1.09 × 10 ^{−9 to} 1.39 × 10 ⁻⁹ [1 / M] are most preferable.

The kdiss value is in the range of 5.06 × 10 ^{−4 to} 9.52 × 10 ⁻³ [1 / sec], preferably 5.06 × 10 ^{−4 to} 9.52 × 10 ⁻⁴ [1 / sec], and 5.06 × 10 ^{-4 to} 6.49 × 10 ^-4 [1 / sec] is most preferable.
The kass values are in the range of ^{4.15 × 10 4 ~5.44 × 10 5} [1 / M · sec], preferably it has ^{4.15 × 10 5 ~5.44 × 10 5} [1 / M · sec], 4.65 × 10 Most preferred is ^{5 to} 5.44 × 10 ⁵ [1 / M · sec].
These KD value, kdiss value, and kass value can be obtained by Scatchard analysis or a surface plasmon resonance sensor such as BIACORE in addition to BIACORE, but are preferably obtained using BIACORE.

8. Measurement of reactivity of humanized antibody to human TF Using the dot blot hybridization method, the reactivity of the humanized antibody of the present invention to non-denatured TF, non-reduced denatured TF, and reduced denatured TF was examined. can do.

  TF can be examined using purified TF from human tissue or expressed and purified in mammalian cells of CHO cells. In addition, guanidine hydrochloride, SDS or the like can be used as the denaturing agent instead of urea, and an SH reducing reagent such as 2-mercaptoethanol can be used as the reducing agent instead of DTT. For the detection of humanized antibodies, anti-human IgG antibodies labeled with various substances can be used. Examples of the labeling substance here include radioactive substances, fluorescent substances such as biotin and FITC, and enzymes such as peroxidase and alkaline phosphatase. The anti-TF antibody of the present invention reacts with any of non-denatured TF, non-reduced modified TF and reduced-modified TF.

9. Pharmaceutical composition containing humanized antibody as an active ingredient and DIC therapeutic agent To confirm the therapeutic effect of humanized antibody against human TF, humanized anti-human TF antibody is applied to animals with high DIC symptoms. The therapeutic effect can be confirmed by administering and measuring the index of DIC.
The antibody used in the present invention is a humanized antibody against human TF. This antibody is an antibody that neutralizes the activity of human TF by binding to human TF, and particularly preferred is a humanized ATR5 antibody. Methods for producing humanized ATR5 antibodies are described in the examples.

  The antibody used in the present invention can be purified with high purity by combining ordinary purification means such as salting-out method, gel filtration method using HPLC, etc., affinity chromatography method using protein A column, etc. . The antibody purified in this manner can be purified with high precision by using conventional immunological means such as radioimmunoassay (RIA), enzyme immunoassay (EIA, ELISA), or immunofluorescence analysis. Can be confirmed.

  The pharmaceutical composition containing the humanized antibody against TF of the present invention and the DIC therapeutic agent can be administered parenterally systemically or locally. For example, intravenous injection such as infusion, intramuscular injection, intraperitoneal injection, and subcutaneous injection can be selected, and an administration method can be appropriately selected depending on the age and symptoms of the patient. The effective dose is selected in the range of 0.01 mg to 1000 mg per kg body weight at a time. Alternatively, a dose of 10 mg / body, preferably 1-1000 mg / body per patient can be selected.

  The pharmaceutical composition and the DIC therapeutic agent containing the humanized antibody against TF of the present invention as an active ingredient may contain both pharmaceutically acceptable carriers and additives depending on the administration route. Examples of such carriers and additives include water, pharmaceutically acceptable organic solvents, collagen, polyvinyl alcohol, polyvinyl pyrrolidone, carboxyvinyl polymer, sodium carboxymethylcellulose, sodium polyacrylate, sodium alginate, water-soluble dextran, Sodium carboxymethyl starch, pectin, methylcellulose, ethylcellulose, xanthan gum, gum arabic, casein, gelatin, agar, diglycerin, glycerin, propylene glycol, polyethylene glycol, petrolatum, paraffin, stearyl alcohol, stearic acid, human serum albumin (HSA), Examples thereof include mannitol, sorbitol, lactose, and surfactants acceptable as pharmaceutical additives. The additive to be used is selected appropriately or in combination from the above according to the dosage form of the present invention, but is not limited thereto.

  The present invention provides a chimeric antibody and a humanized antibody against human TF, and a method for producing a humanized antibody. These antibodies are useful as therapeutic agents because of their low antigenicity in humans.

Next, the present invention will be described more specifically with reference to examples.
Example 1. Human TF DNA cloning of encoding the V region of mouse monoclonal antibody against (1) mRNA preparation hybridomas ATR-2, ATR-3, ATR-4, ATR-5 (IgG1κ), ATR-7, MRNA from ATR-8 (IgG2aκ) was prepared using the Quick Prep mRNA Purification Kit (Pharmacia Biotech). According to the instructions attached to the kit, each hybridoma cell was completely homogenized with an extraction buffer, and mRNA was purified with an oligo (dT) -cellulose span column, followed by ethanol precipitation. The mRNA precipitate was dissolved in elution buffer.

(2) Preparation and amplification of the cDNA of the gene encoding the mouse antibody V region (i) Cloning of the H chain V region cDNA Cloning of the gene encoding the H chain V region of the mouse monoclonal antibody against human TF was performed using 5'-RACE. (Frohman, MAet al., Proc. Natl. Acad. Sci. USA, 85, 8998-9002, 1988; Belyavsky, A. et al., Nucleic Acid Res. 17, 2919-2932, 1989). For the 5′-RACE method, Marathon cDNA Amplification Kit (CLONTECH) was used, and the operation was performed according to the prescription attached to the kit.

  Using about 1 μg of the mRNA prepared as described above as a template, cDNA synthesis primer attached to the kit was added and reacted with reverse transcriptase at 42 ° C. for 60 minutes for reverse transcription into cDNA. This was reacted with DNA polymerase I, DNA ligase, and RNaseH at 16 ° C. for 1.5 hours, and with T4 DNA polymerase at 16 ° C. for 45 minutes to synthesize double-stranded cDNA. Double-stranded cDNA was extracted with phenol and chloroform and recovered by ethanol precipitation.

  By overnight reaction at 16 ° C. with T4 DNA ligase, a cDNA adapter was ligated to both ends of the double-stranded cDNA. The reaction mixture was diluted 50-fold with 10 mM Tricine-KOH (pH 8.5), 0.1 mM EDTA solution. Using this as a template, the gene encoding the H chain V region was amplified by PCR. Adapter primer 1 included in the kit is used for the 5'-side primer, and MHC-G1 primer (SEQ ID NO: 1) (ATR-2, ATR-3, ATR-4 and ATR-5) or MHC is used for the 3'-side primer. -G2a primer (SEQ ID NO: 2) (ATR-7 and ATR-8) (STJones, et al., Biotechnology, 9, 88-89, 1991) was used.

The PCR solution for the ATR-2, 3, 4 and 5 antibody H chain V regions was 120 mM Tris-HCl (pH 8.0), 10 mM KCl, 6 mM (NH ₄ ) ₂ SO ₄ , 0.1% Triton X-100 in 100 μl. , 0.001% BSA, 0.2 mM dNTPs (dATP, dGTP, dCTP, dTTP), 1 mM MgCl ₂ , 2.5 units of KOD DNA polymerase (Toyobo), 30-50 pmole adapter primer 1, MHC-G1 primer, and cDNA adapter Contains 1-5 μl of ligated cDNA reaction mixture.
Each PCR was performed 30 times with a DNA Thermal Cycler 480 (Perkin-Elmer) at a temperature cycle of 94 ° C. for 30 seconds, 55 ° C. for 30 seconds, and 74 ° C. for 1 minute.

(Ii) Cloning of L chain V region cDNA Cloning of the gene encoding the L chain V region of the mouse monoclonal antibody against human TF can be performed by the 5′-RACE method (Frohman, MAetal., Proc. Natl. Acad. Sci. USA, 85, 8998-9002, 1988; Belyavsky, A. et al., Nucleic Acid Res. 17, 2919-2932, 1989). For the 5′-RACE method, Marathon cDNA Amplification Kit (CLONTECH) was used, and the operation was performed according to the prescription attached to the kit. A cDNA synthesis primer was added using about 1 μg of the mRNA prepared as described above as a template, and reverse transcription to cDNA was performed by reacting with reverse transcriptase at 42 ° C. for 60 minutes.

  This was reacted with DNA polymerase I, DNA ligase, and RNaseH at 16 ° C. for 1.5 hours, and with T4 DNA polymerase at 16 ° C. for 45 minutes to synthesize double-stranded cDNA. Double-stranded cDNA was extracted with phenol and chloroform and recovered by ethanol precipitation. By overnight reaction at 16 ° C. with T4 DNA ligase, a cDNA adapter was ligated to both ends of the double-stranded cDNA. The reaction mixture was diluted 50-fold with 10 mM Tricine-KOH (pH 8.5) and 0.1 mM EDTA solution. Using this as a template, the gene encoding the L chain V region was amplified by PCR. Adapter primer 1 was used as the 5′-side primer, and MKC primer (SEQ ID NO: 3) (S. T. Jones, et al., Biotechnology, 9, 88-89, 1991) was used as the 3′-side primer.

The PCR solution was 120 mM Tris-HCl (pH 8.0), 10 mM KCl, 6 mM (NH ₄ ) ₂ SO ₄ , 0.1% Triton X-100, 0.001% BSA, 0.2 mM dNTPs (dATP, dGTP, dCTP, dTTP), 1 mM MgCl ₂ , 2.5 units of KOD DNA polymerase (Toyobo), 30-50 pmole of adapter primer 1 and MKC primer, and 1 μl of a reaction mixture of cDNA ligated with a cDNA adapter.

  PCR was performed 30 times with a DNA Thermal Cycler 480 (Perkin-Elmer) at a temperature cycle of 94 ° C. for 30 seconds, 55 ° C. for 30 seconds, and 74 ° C. for 1 minute.

(3) Purification and fragmentation of PCR product The PCR reaction mixture was extracted with phenol and chloroform, and the amplified DNA fragment was recovered by ethanol precipitation. The DNA fragment was digested with the restriction enzyme XmaI (New England Biolabs) at 37 ° C. for 1 hour. The XmaI digestion mixture was separated by agarose gel electrophoresis using 2% to 3% NuSieve GTG agarose (FMC BioProducts), containing DNA fragments approximately 500 bp long as the H chain V region and approximately 500 bp long as the L chain V region. Agarose pieces were cut out. The agarose piece was extracted with phenol and chloroform, and the DNA fragment was precipitated with ethanol, and then dissolved in 10 μl of 10 mM Tris-HCl (pH 8.0) and 1 mM EDTA solution (hereinafter referred to as TE).

The DNA ligation kit ver.2 (Takara Shuzo) was prepared by digesting with XmaI and the pUC19 plasmid vector containing the gene encoding the mouse H chain V region and L chain V region prepared as described above. ) And reacted at 16 ° C. for 1 hour according to the attached formulation.
This ligation mixture was added to 100 μl of E. coli JM109 competent cells (Nippon Gene) and allowed to stand on ice for 30 minutes and at 42 ° C. for 1 minute.

Next, 300 μl of Hi-Competence Broth (Nippon Gene) was added and incubated at 37 ° C. for 1 hour, and then LB agar medium containing 100 μg / ml ampicillin (Molecular Cloning: A Laboratory Manual, Sambrook, et al., Cold Spring Harbor Laboratory Press) 1989) (hereinafter referred to as LBA agar medium), this E. coli was seeded and incubated at 37 ° C. overnight to obtain an E. coli transformant.
This transformant is cultured overnight at 37 ° C in 3 ml or 4 ml of LB medium (hereinafter referred to as LBA medium) containing 50 µg / ml ampicillin, and plasmid DNA is obtained from the cell fraction using the QIAprep Spin Plasmid Kit (QIAGEN). Was prepared and the nucleotide sequence was determined.

(4) Determination of the nucleotide sequence of the gene encoding the mouse antibody V region Using the Dye Terminator Cycle Sequencing FS Ready Reaction Kit (Perkin-Elmer), the nucleotide sequence of the cDNA coding region in the above plasmid was determined using the DNA Sequencer 373A (Perkin-Elmer). ). Using M13 Primer M4 (Takara Shuzo) (SEQ ID NO: 4) and M13 Primer RV (Takara Shuzo) (SEQ ID NO: 5) as sequencing primers, the sequence was determined by confirming the base sequence in both directions.

  The plasmids containing the gene encoding the mouse H chain V region derived from the hybridomas ATR-2, ATR-3, ATR-4, ATR-5, ATR-7 and ATR-8 thus obtained were respectively ATR-xHv / Plasmids named pUC19 (x = 2, 3, 4, 5, 7 or 8) and containing the gene encoding the L chain V region were designated ATR-xLv / pUC19 (x = 2, 3, 4, 5), respectively. 7 or 8). The nucleotide sequence (including the corresponding amino acid sequence) of the gene encoding the H chain V region of each mouse antibody contained in the plasmid ATR-xHv / pUC19 (x = 2, 3, 4, 5, 7 or 8) No. 6 to 11, the nucleotide sequence (corresponding amino acid sequence) of the gene encoding the L chain V region of each mouse antibody contained in the plasmid ATR-xLv / pUC19 (x = 2, 3, 4, 5, 7 or 8) Are shown in SEQ ID NOs: 12 to 17, respectively.

Example 2 Construction of Chimeric Antibody A chimeric ATR-5 antibody was prepared by linking the mouse ATR-5 antibody V region to the human antibody C region. A chimeric antibody expression vector was constructed by linking a gene encoding the ATR-5 antibody V region to an expression vector encoding the human antibody C region.

(1) Construction of chimeric antibody H chain V region The ATR-5 antibody H chain V region was modified by PCR to be linked to an expression vector encoding the human antibody H chain C region. The 5′-end primer ch5HS (SEQ ID NO: 18) hybridizes to the 5′-end of the DNA encoding the V region, and the Kozak consensus sequence (Kozak, M. et al., J. Mol. Biol., 196, 947-950, 1987) and a restriction enzyme SalI recognition sequence. The 3′-side primer ch5HA (SEQ ID NO: 19) was designed to hybridize to the 3′-end of DNA encoding the J region and to have a restriction enzyme NheI recognition sequence.

The PCR solution was 120 mM Tris-HCl (pH 8.0), 10 mM KCl, 6 mM (NH ₄ ) ₂ SO ₄ , 0.1% Triton X-100, 0.001% BSA, 0.2 mM dNTPs (dATP, dGTP, dCTP, dTTP), 1 mM MgCl ₂ , 2.5 units of KOD DNA polymerase (Toyobo), 50 pmole of ch5HS primer and ch5HA primer, and 1 μl of plasmid ATR5Hv / pUC19 as template DNA. PCR was performed 30 times with a DNA Thermal Cycler 480 (Perkin-Elmer) at a temperature cycle of 94 ° C. for 30 seconds, 55 ° C. for 30 seconds, and 74 ° C. for 1 minute.

  The PCR reaction mixture was extracted with phenol and chloroform, and the amplified DNA fragment was recovered by ethanol precipitation. The DNA fragment was digested with the restriction enzyme NheI (Takara Shuzo) at 37 ° C. for 1 hour, and then digested with the restriction enzyme SalI (Takara Shuzo) at 37 ° C. for 1 hour. This digestion mixture was separated by agarose gel electrophoresis using 3% NuSieve GTG agarose (FMC BioProducts), and an agarose piece containing a DNA fragment having a length of about 450 bp was excised. The agarose piece was extracted with phenol and chloroform, and the DNA fragment was precipitated with ethanol, and then dissolved in 20 μl of TE.

  As a cloning vector, a modified pUC19 vector (hereinafter referred to as CVIDEC) into which recognition sequences for restriction enzymes NheI, SalI, SplI, and BglII were introduced was used. Using the DNA ligation kit ver.2 (Takara Shuzo), the gene fragment encoding the mouse H chain V region prepared as described above and digested with NheI and SalI at 16 ° C. according to the attached prescription Ligated for 1 hour.

  This ligation mixture was added to 100 μl of E. coli JM109 competent cells (Nippon Gene) and allowed to stand on ice for 30 minutes and at 42 ° C. for 1 minute. Next, 300 μl of Hi-Competence Broth (Nippon Gene) was added and incubated at 37 ° C. for 1 hour, then this E. coli was seeded on a 100 μg / ml LBA agar medium and incubated overnight at 37 ° C. to obtain an E. coli transformant. . This transformant was cultured overnight at 37 ° C. in 3 ml of LBA medium, and plasmid DNA was prepared from the cell fraction using QIAprep Spin Plasmid Kit (QIAGEN).

  The nucleotide sequence of the cDNA coding region in the plasmid was determined by DNA Sequencer 373A (Perkin-Elmer) using Dye Terminator Cycle Sequencing FS Ready Reaction Kit (Perkin-Elmer). Using M13 Primer M4 (Takara Shuzo) and M13 Primer RV (Takara Shuzo) as sequencing primers, the sequence was determined by confirming the base sequence in both directions. A plasmid containing a gene encoding the ATR-5 antibody H chain V region and having a SalI recognition sequence and Kozak consensus sequence on the 5′-side and an NheI recognition sequence on the 3′-side was designated as chATR5Hv / CVIDEC.

(2) Construction of chimeric antibody L chain V region The ATR-5 antibody L chain V region was modified by PCR in order to be linked to an expression vector encoding the human antibody L chain C region. The 5′-end primer ch5LS (SEQ ID NO: 20) hybridizes to the 5′-end of the DNA encoding the V region, and the Kozak consensus sequence (Kozak, M. et al., J. Mol. Biol., 196, 947-950, 1987) and a restriction enzyme BglII recognition sequence. The 3′-side primer ch5LA (SEQ ID NO: 21) was designed to hybridize to the 3′-end of DNA encoding the J region and to have a restriction enzyme SplI recognition sequence.

The PCR solution was 120 mM Tris-HCl (pH 8.0), 10 mM KCl, 6 mM (NH ₄ ) ₂ SO ₄ , 0.1% Triton X-100, 0.001% BSA, 0.2 mM dNTPs (dATP, dGTP, dCTP, dTTP), 1 mM MgCl ₂ , 2.5 units of KOD DNA polymerase (Toyobo), 50 pmole of ch5LS primer and ch5LA primer, and 1 μl of plasmid ATR5Lv / pUC19 as template DNA. PCR was performed 30 times with a DNA Thermal Cycler 480 (Perkin-Elmer) at a temperature cycle of 94 ° C. for 30 seconds, 55 ° C. for 30 seconds, and 74 ° C. for 1 minute.

  The PCR reaction mixture was extracted with phenol and chloroform, and the amplified DNA fragment was recovered by ethanol precipitation. The DNA fragment was digested with the restriction enzyme SplI (Takara Shuzo) at 37 ° C. for 1 hour, and then with the restriction enzyme BglII (Takara Shuzo) at 37 ° C. for 1 hour. The digestion mixture was separated by agarose gel electrophoresis using 3% NuSieve GTG agarose (FMC BioProducts), and an agarose piece containing a DNA fragment having a length of about 400 bp was excised. The agarose piece was extracted with phenol and chloroform, and the DNA fragment was precipitated with ethanol and then dissolved in 20 μl of TE.

  Using the DNA ligation kit ver.2 (Takara Shuzo), the gene fragment encoding the mouse L chain V region prepared as described above and the CVIDEC vector prepared by digesting with SplI and BglII at 16 ° C. according to the attached prescription Ligated for 1 hour.

  This ligation mixture was added to 100 μl of E. coli JM109 competent cells (Nippon Gene) and allowed to stand on ice for 30 minutes and at 42 ° C. for 1 minute. Next, 300 μl of Hi-Competence Broth (Nippon Gene) was added and incubated at 37 ° C. for 1 hour, then this E. coli was seeded on a 100 μg / ml LBA agar medium and incubated overnight at 37 ° C. to obtain an E. coli transformant. . This transformant was cultured overnight at 37 ° C. in 3 ml of LBA medium, and plasmid DNA was prepared from the cell fraction using QIAprep Spin Plasmid Kit (QIAGEN).

  The nucleotide sequence of the cDNA coding region in the plasmid was determined by DNA Sequencer 373A (Perkin-Elmer) using Dye Terminator Cycle Sequencing FS Ready Reaction Kit (Perkin-Elmer). Using M13 Primer M4 (Takara Shuzo) and M13 Primer RV (Takara Shuzo) as sequencing primers, the sequence was determined by confirming the base sequence in both directions. A plasmid containing a gene encoding the ATR-5 antibody L chain V region and having a BglII recognition sequence and a Kozak consensus sequence on the 5′-side and a SplI recognition sequence on the 3′-side was designated as chATR5Lv / CVIDEC.

(3) Construction of chimeric antibody expression vector
A chimeric antibody expression vector was constructed using an antibody expression vector introduced from IDEC. IgG1 type antibody expression vector N5KG1 (V) and IgG4 type antibody expression vector N5KG4P were used as vectors. The gene encoding the ATR-5 H chain V region at the SalI-NheI site immediately before the human antibody H chain C region of the expression vector N5KG1 (V) or N5KG4P, and BglII- A chimeric ATR-5 antibody expression vector was prepared by linking a gene encoding the L chain V region of ATR-5 to the SplI site.

(I) Introduction of H chain V region Plasmid chATR5Hv / CVIDEC was digested with the restriction enzyme NheI (Takara Shuzo) at 37 ° C for 3 hours, and then with the restriction enzyme SalI (Takara Shuzo) at 37 ° C for 3 hours. The digestion mixture was separated by agarose gel electrophoresis using 1.5% NuSieve GTG agarose (FMC BioProducts), and an agarose piece containing a DNA fragment having a length of about 450 bp was excised. The agarose piece was extracted with phenol and chloroform, and the DNA fragment was precipitated with ethanol, and then dissolved in 20 μl of TE.

  Expression vectors N5KG1 (V) and N5KG4P were digested with the restriction enzyme NheI (Takara Shuzo) at 37 ° C. for 3 hours, and then with the restriction enzyme SalI (Takara Shuzo) at 37 ° C. for 3 hours. The digestion mixture was separated by agarose gel electrophoresis using 1.5% NuSieve GTG agarose (FMC BioProducts), and an agarose piece containing a DNA fragment having a length of about 9000 bp was excised. The agarose piece was extracted with phenol and chloroform, and the DNA fragment was precipitated with ethanol and dissolved in 60 μl of TE.

  Using the DNA ligation kit ver.2 (Takara Shuzo), the SalI-NheI DNA fragment containing the gene encoding the H chain V region prepared as described above and N5KG1 (V) or N5KG4P digested with SalI and NheI are attached. The reaction was allowed to proceed at 16 ° C. for 1 hour according to the formulation.

  This ligation mixture was added to 100 μl of E. coli JM109 competent cells (Nippon Gene) and allowed to stand on ice for 30 minutes and at 42 ° C. for 1 minute. Next, 300 μl of Hi-Competence Broth (Nippon Gene) was added and incubated at 37 ° C. for 1 hour, then this E. coli was seeded on a 100 μg / ml LBA agar medium and incubated overnight at 37 ° C. to obtain an E. coli transformant. . This transformant was cultured overnight at 37 ° C. in 3 ml of LBA medium, and plasmid DNA was prepared from the cell fraction using QIAprep Spin Plasmid Kit (QIAGEN). The plasmids containing the gene encoding the chimeric ATR-5 antibody H chain were designated chATR5Hv / N5KG1 (V) and chATR5Hv / N5KG4P, respectively.

(Ii) Introduction of L chain V region Plasmid chATR5Lv / CVIDEC was digested with restriction enzymes BglII (Takara Shuzo) and SplI (Takara Shuzo) at 37 ° C for 1.5 hours. The digestion mixture was separated by agarose gel electrophoresis using 1.5% NuSieve GTG agarose (FMC BioProducts), and an agarose piece containing a DNA fragment having a length of about 400 bp was excised. The agarose piece was extracted with phenol and chloroform, and the DNA fragment was precipitated with ethanol and then dissolved in 20 μl of TE.

  Plasmids chATR5Hv / N5KG1 (V) and chATR5Hv / N5KG4P were digested with restriction enzymes BglII (Takara Shuzo) and SplI (Takara Shuzo) at 37 ° C. for 1.5 hours. The digestion mixture was separated by agarose gel electrophoresis using 1.5% NuSieve GTG agarose (FMC BioProducts), and an agarose piece containing a DNA fragment of about 9400 bp was excised. The agarose piece was extracted with phenol and chloroform, and the DNA fragment was precipitated with ethanol, and then dissolved in 20 μl of TE.

  The DNA ligation kit ver.2 (Takara Shuzo) was prepared by using the SplI-BglII DNA fragment containing the gene encoding the L chain V region prepared as described above and chATR5Hv / N5KG1 (V) or chATR5Hv / N5KG4P digested with SplI and BglII. Used and reacted for 1 hour at 16 ° C. according to the attached recipe.

  This ligation mixture was added to 100 μl of E. coli JM109 competent cells (Nippon Gene) and allowed to stand on ice for 30 minutes and at 42 ° C. for 1 minute. Next, 300 μl of Hi-Competence Broth (Nippon Gene) was added and incubated at 37 ° C. for 1 hour, then this E. coli was seeded on a 100 μg / ml LBA agar medium and incubated overnight at 37 ° C. to obtain an E. coli transformant. . This transformant was cultured overnight at 37 ° C. in 1 l of 2 × YT medium containing 50 μg / ml ampicillin, and plasmid DNA was prepared from the bacterial cell fraction using Plasmid Maxi Kit (QIAGEN). The plasmids containing the genes encoding these chimeric ATR-5 antibodies were named chATR5 / N5KG1 (V) and chATR5 / N5KG4P, respectively.

(4) Transfection into COS-7 cells In order to evaluate the antigen-binding activity and neutralizing activity of the chimeric antibody, the expression plasmid was transfected into COS-7 cells to express the chimeric antibody transiently.
Plasmids chATR5 / N5KG1 (V) or chATR5 / N5KG4P were transduced into COS-7 cells by electroporation using a Gene Pulser apparatus (Bio Rad). Add 50 µg of plasmid to 0.78 ml of COS-7 cells suspended in Dulbecco's PBS (-) (hereinafter referred to as PBS) at a cell concentration of 1x10 ⁷ cells / ml to obtain a capacitance of 1,500 V and 25 µF. Gave a pulse.

After a 10-minute recovery period at room temperature, electroporated cells are suspended in DMEM medium (GIBCO) containing 5% Ultra Low IgG fetal calf serum (GIBCO) and CO is used in a 10 cm culture dish. _The cells were cultured in ₂ incubators. After 24 hours of culture, the culture supernatant was removed by aspiration, and a serum-free medium HBCHO (Irvine Scientific) was newly added. After a further 72 hours of culture, the culture supernatant was collected and cell debris was removed by centrifugation.

(5) Antibody purification
The chimeric antibody was purified from the culture supernatant of COS-7 cells using rProtein A Sepharose Fast Flow (Pharmacia Biotech) as follows.
The column was equilibrated by loading 1 ml of rProtein A Sepharose Fast Flow onto the column and flowing 10 times the amount of TBS. After the culture supernatant of COS-7 cells was applied to the equilibrated column, the column was washed with 10 times the amount of TBS.

Next, the adsorbed antibody fraction was eluted from the column by flowing 13.5 ml of 2.5 mM HCl (pH 3.0). Immediately after addition of 1.5 ml of 1 M Tris-HCl (pH 8.0), the eluate was added to the medium. It was summed up.
The purified antibody fraction is subjected to ultrafiltration using Centriprep 100 (Amicon) twice, so that a solvent is added to 50 mM Tris-HCl (pH 7.6) (hereinafter referred to as TBS) containing 150 mM NaCl. Replaced and finally concentrated to about 1.5 ml.

(6) Establishment of stable production cell line of CHO In order to establish a stable production cell line of chimeric antibody, the expression plasmid was introduced into CHO cells (DG44) conditioned to CHO-S-SFMII serum-free medium (GIBCO).

Plasmids chATR5 / N5KG1 (V) or chATR5 / N5KG4P were cleaved with restriction enzyme SspI (Takara Shuzo) to obtain linear DNA, extracted with phenol and chloroform, and then recovered by ethanol precipitation. The linearized plasmid was transduced into DG44 cells by electroporation using a Gene Pulser apparatus (Bio Rad). 10 μg of plasmid was added to 0.78 ml of DG44 cells suspended at a cell concentration of 1 × 10 ⁷ cells / ml in PBS, and pulsed at a capacitance of 1,500 V and 25 μF.

After a 10-minute recovery period at room temperature, the electroporated cells were suspended in CHO-S-SFMII medium (GIBCO) containing hypoxanthine / thymidine (GIBCO), and two 96-well plates (Falcon ) In a CO ₂ incubator. On the next day after the start of the culture, change to a selective medium of CHO-S-SFMII medium (GIBCO) containing hypoxanthine / thymidine (GIBCO) and 500 μg / ml GENETICIN (G418Sulfate, GIBCO), and select the cells into which the antibody gene has been introduced. did. After changing the selective medium, observe the cells under a microscope for about 2 weeks, and after normal cell growth is observed, measure the amount of antibody produced by the antibody concentration measurement ELISA described later, and select cells with a large amount of antibody production. did.

Example 3. Construction of humanized antibody (1) Construction of humanized antibody H chain (i) Construction of humanized H chain version “a” Humanized ATR-5 antibody H chain was converted to CDR by PCR. -Made by grafting. Seven PCR primers were used for the preparation of humanized ATR-5 antibody heavy chain version “a” with FRs derived from human antibody L39130 (DDBJ, Gao L. et al., Unpublished, 1995). CDR-grafting primers hR5Hv1S (SEQ ID NO: 22), hR5Hv2S (SEQ ID NO: 23) and hR5Hv4S (SEQ ID NO: 24) have sense DNA sequences, and CDR grafting primers hR5Hv3A (SEQ ID NO: 25) and hR5Hv5A (SEQ ID NO: 26) ) Has an antisense DNA sequence and each has a complementary sequence of 18-35 bp on each end of the primer.

hR5Hv1S was designed to have a Kozak consensus sequence (Kozak, M, et al., J. Mol. Biol. 196, 947-950, 1987) and a SalI recognition site, and hR5Hv5A was designed to have an NheI recognition site. The external primer hR5HvPrS (SEQ ID NO: 27) has homology with the CDR grafting primer hR5Hv1S and hR5HvPrA (SEQ ID NO: 28) with the CDR grafting primer hR5Hv5A.
CDR-grafting primers hR5Hv1S, hR5Hv2S, hR5Hv3A, hR5Hv4S and hR5Hv5A, and external primers hR5HvPrS and hR5HvPrA were synthesized and purified by Pharmacia Biotech.

For PCR, KOD DNA polymerase (Toyobo Co., Ltd.) was used. In 98 μl, 120 mM Tris-HCl (pH 8.0), 10 mM KCl, 6 mM (NH ₄ ) ₂ SO ₄ , 0.1% Triton X-100, 0.001% BSA, 0.2 Add 5 pmole with mM dNTPs (dATP, dGTP, dCTP, dTTP), 1 mM MgCl ₂ , 2.5 units of KOD DNA polymerase (Toyobo), CDR-grafting primers hR5Hv1S, hR5Hv2S, hR5Hv3A, hR5Hv4S and hR5Hv5A Using the solution at 94 ° C for 30 seconds, 50 ° C for 1 minute, 72 ° C for 1 minute 5 times, and adding 100 pmole of external primers hR5HvPrS and hR5HvPrA at the same temperature in a 100 μl system The cycle was performed 25 times. DNA fragments amplified by the PCR method were separated by agarose gel electrophoresis using 2% Nu Sieve GTG agarose (FMC Bio.Products).

  A piece of agarose containing a DNA fragment having a length of about 430 bp was cut out, 3 times the amount (ml / g) of TE was added, and the DNA fragment was purified by phenol extraction, phenol / chloroform extraction, and chloroform extraction. After purifying the purified DNA with ethanol, one third of the purified DNA was dissolved in 17 μl of water. The obtained PCR reaction mixture was digested with NheI and SalI, and ligated to the plasmid vector CVIDEC prepared by digesting with NheI and SalI using DNA ligation kit ver.2 (Takara Shuzo) according to the attached instructions.

  This ligation mixture was added to 100 μl of E. coli JM109 competent cells (Nippon Gene) and allowed to stand on ice for 30 minutes and at 42 ° C. for 1 minute. Next, 300 μl of Hi-Competence Broth (Nippon Gene) was added and incubated at 37 ° C. for 1 hour, then this E. coli was seeded on a 100 μg / ml LBA agar medium and incubated overnight at 37 ° C. to obtain an E. coli transformant. . This transformant was cultured overnight at 37 ° C. in 3 ml of LBA medium, and plasmid DNA was prepared from the cell fraction using QIAprep Spin Plasmid Kit (QIAGEN).

  The nucleotide sequence of the cDNA coding region in the plasmid was determined by DNA Sequencer 373A (Perkin-Elmer) using Dye Terminator Cycle Sequencing FS Ready Reaction Kit (Perkin-Elmer). Using M13 Primer M4 (Takara Shuzo) and M13 Primer RV (Takara Shuzo) as sequencing primers, the sequence was determined by confirming the base sequence in both directions.

  Since mutations and deletions were observed before or after the EcoT22I recognition site, fragments each having the correct sequence were ligated and subcloned into CVIDEC again to determine the nucleotide sequence. The plasmid with the correct sequence was named hATR5Hva / CVIDEC. The nucleotide sequence of the humanized H chain version “a” contained in the plasmid hATR5Hva / CVIDEC and the corresponding amino acid sequence are shown in SEQ ID NO: 29. The amino acid sequence of version “a” is shown in SEQ ID NO: 30.

(Ii) Construction of humanized heavy chain versions “b” and “c” Prepare versions “b” and “c” by replacing FR3 of version “a” with FR3 derived from another human antibody by FR-shuffling method. did. Version “b” replaces FR3 with one derived from human antibody Z34963 (DDBJ, Borretzen M. et al., Proc. Natl. Acad. Sci. USA, 91, 12917-12921, 1994), so that DNA primer encoding FR3 4 were produced. FR-shuffling primers F3RFFS (SEQ ID NO: 31) and F3RFBS (SEQ ID NO: 32) have a sense DNA sequence, and F3RFFA (SEQ ID NO: 33) and F3RFBA (SEQ ID NO: 34) have an antisense DNA sequence.

  F3RFFS and F3RFFA have complementary sequences to each other, and have BalI and XhoI recognition sequences at both ends. In version “c”, FR3 was replaced with one derived from human antibody P01825 (SWISS-PROT, Poljak RJ. Et al., Biochemistry, 16, 3412-3420, 1977), so four DNA primers encoding FR3 were prepared. FR-shuffling vectors F3NMFS (SEQ ID NO: 35) and F3NMBS (SEQ ID NO: 36) have sense DNA sequences, and F3NMFA (SEQ ID NO: 37) and F3NMBA (SEQ ID NO: 38) have antisense DNA sequences. F3RFBS and F3RFBA have complementary sequences, and have XhoI and NcoI recognition sequences at both ends.

  F3RFFS, F3RFBS, F3RFFA, F3RFBA, F3NMFS, F3NMBS, F3NMFA and F3NMBA were synthesized by Pharmacia Biotech. F3RFFS and F3RFFA, F3RFBS and F3RFBA were annealed and digested with BalI and XhoI, NcoI and XhoI, respectively. These were introduced into a plasmid hATR5Hva / CVIDEC (BalI / NcoI) prepared by digesting with BalI and NcoI, and the nucleotide sequence was determined. The plasmid with the correct sequence was named hATR5Hvb / CVIDEC. The nucleotide sequence of the humanized H chain version “b” contained in the plasmid hATR5Hvb / CVIDEC and the corresponding amino acid sequence are shown in SEQ ID NO: 39. The amino acid sequence of version “b” is shown in SEQ ID NO: 40.

  F3NMFS and F3NMFA, F3NMBS and F3NMBA were annealed and digested with BalI and XhoI, NcoI and XhoI, respectively. These were introduced into a plasmid hATR5Hva / CVIDEC (BalI / NcoI) prepared by digesting with BalI and NcoI, and the nucleotide sequence was determined. The plasmid with the correct sequence was named hATR5Hvc / CVIDEC. The nucleotide sequence of the humanized H chain version “c” contained in the plasmid hATR5Hvc / CVIDEC and the corresponding amino acid sequence are shown in SEQ ID NO: 41. The amino acid sequence of version “c” is shown in SEQ ID NO: 42.

(Iii) Construction of humanized H chain versions “d” and “e” Versions “d” and “e” were prepared by replacing FR3 of version “a” with FR3 derived from another human antibody by FR-shuffling method. did. Version "d" replaces FR3 with one derived from human antibody M62723 (DDBJ, Pascual V. et al., J. Clin. Invest., 86, 1320-1328, 1990), so four DNA primers encoding FR3 Produced. The FR-shuffling primer F3EPS (SEQ ID NO: 43) has a sense DNA sequence, F3EPA (SEQ ID NO: 44) has an antisense DNA sequence, and the 3'-end of the primer has an 18 bp complementary sequence.

  The external primers F3PrS (SEQ ID NO: 45) and F3PrA (SEQ ID NO: 46) have homology to the FR-shuffling primers F3EPS and F3EPA, and can be used for other FR3 shuffling. In version “e”, two DNA primers encoding FR3 were prepared to replace FR3 with one derived from human antibody Z80844 (DDBJ, Thomsett AR. Et al., Unpublished). The FR-shuffling primer F3VHS (SEQ ID NO: 47) has a sense DNA sequence, F3VHA (SEQ ID NO: 48) has an antisense DNA sequence, and the 3'-end of the primer has an 18 bp complementary sequence. F3EPS, F3EPA, F3PrS, F3PrA, F3VHS and F3VHA were synthesized by Pharmacia Biotech.

For PCR, KOD DNA Polymerase (Toyobo) was used, and 1 μM FR-shuffling primer F3EPS and F3EPA or F3VHS and F3VHA 5 μl, 0.2 mM dNTPs, 1.0 mM MgCl ₂ , 2.5 U in 100 μl reaction mixture, respectively. 5 times in a temperature cycle of 94 ° C for 30 seconds, 50 ° C for 1 minute, and 74 ° C for 1 minute using the attached buffer under the conditions including KOD DNA polymerase, and 100 pmole of external primer F3PrS and F3PrA was added and the same temperature cycle was performed 25 times.

  DNA fragments amplified by the PCR method were separated by agarose gel electrophoresis using 2% Nu Sieve GTG agarose (FMC Bio. Products). A piece of agarose containing a 424 bp long DNA fragment was cut out, 3 times the amount (ml / g) of TE was added, and the DNA fragment was purified by phenol extraction, phenol / chloroform extraction, and chloroform extraction. After purifying the purified DNA with ethanol, one third of the purified DNA was dissolved in 14 μl of water. The obtained PCR reaction mixture was digested with BalI and NcoI, and introduced into plasmid hATR5Hva / CVIDEC (BalI / NcoI) prepared by digesting with BalI and NcoI, and the nucleotide sequence was determined.

  Plasmids with the correct sequence were named hATR5Hvd / CVIDEC and hATR5Hve / CVIDEC. The nucleotide sequence and the corresponding amino acid sequence of the humanized H chain version “d” contained in the plasmid hATR5Hvd / CVIDEC are shown in SEQ ID NO: 49, and the amino acid sequence of the version “d” is shown in SEQ ID NO: 50. In addition, the base sequence and the corresponding amino acid sequence of the humanized H chain version “e” contained in the plasmid hATR5Hve / CVIDEC are shown in SEQ ID NO: 51, and the amino acid sequence of the version “e” is shown in SEQ ID NO: 52.

(Iv) Construction of humanized H chain versions “f” and “g” Versions “f” and “g” were prepared by replacing FR3 of version “a” with FR3 derived from another human antibody by the FR-shuffling method. did. Version “f” is FR3 derived from human antibody L04345 (DDBJ, Hillson JL. Et al., J. Exp. Med., 178, 331-336, 1993), version “g” is S78322 (DDBJ, Bejcek BE. Et al., Two primers encoding FR3 were synthesized in order to replace FR3 from Cancer Res., 55, 2346-2351, 1995). Version “f” of the FR-shuffling primer F3SSS (SEQ ID NO: 53) has a sense DNA sequence, F3SSA (SEQ ID NO: 54) has an antisense DNA sequence, and the primer 3′-end is a complementary sequence of 18 bp. Have

  Version “g” of the FR-shuffling primer F3CDS (SEQ ID NO: 55) has a sense DNA sequence, F3CDA (SEQ ID NO: 56) has an antisense DNA sequence, and the primer 3′-end is a complementary sequence of 18 bp. Have F3SSS, F3SSA, F3CDS and F3CDA were synthesized and purified by Pharmacia Biotech. PCR was performed using KOD DNA polymerase (Toyobo Co., Ltd.), 1 μM FR-shuffling primer F3SSS and F3SSA or F3CDS and F3CDA 5 μl each in 100 μl reaction mixture, 0.2 mM dNTPs, 1.0 mM MgCl2, 2.5 U Performed 5 times with a temperature cycle of 94 ° C for 30 seconds, 50 ° C for 1 minute, and 74 ° C for 1 minute using the attached buffer under conditions including KOD DNA polymerase, and 100 pmole external primers F3PrS and F3PrA And the same temperature cycle was performed 25 times.

  DNA fragments amplified by the PCR method were separated by agarose gel electrophoresis using 2% Nu Sieve GTG agarose (FMC Bio. Products). A piece of agarose containing a 424 bp long DNA fragment was cut out, 3 times the amount (ml / g) of TE was added, and the DNA fragment was purified by phenol extraction, phenol / chloroform extraction, and chloroform extraction. After purifying the purified DNA with ethanol, one third of the purified DNA was dissolved in 14 μl of water. The obtained PCR reaction mixture was digested with BalI and NcoI, and introduced into plasmid hATR5Hva / CVIDEC (BalI / NcoI) prepared by digesting with BalI and NcoI, and the nucleotide sequence was determined.

  The plasmids with the correct sequence were named hATR5Hvf / CVIDEC and hATR5Hvg / CVIDEC. The nucleotide sequence of the humanized H chain version “f” and the corresponding amino acid sequence and the version “f” amino acid sequence contained in the plasmid hATR5Hvf / CVIDEC are shown in SEQ ID NOs: 57 and 58, respectively. Further, the nucleotide sequence of the humanized H chain version “g” and the corresponding amino acid sequence contained in the plasmid hATR5Hvg / CVIDEC, and the amino acid sequence of version “g” are shown in SEQ ID NOs: 59 and 60.

(V) Construction of humanized H chain version “h” Version “h” was prepared by replacing FR3 of version “a” with FR3 derived from another human antibody by FR-shuffling method. Version “h” was synthesized with two primers encoding FR3 to replace FR3 derived from human antibody Z26827 (DDBJ, Van Der Stoep et al., J. Exp. Med., 177, 99-107, 1993). Version “h” of the FR-shuffling primer F3ADS (SEQ ID NO: 61) has a sense DNA sequence, F3ADA (SEQ ID NO: 62) has an antisense DNA sequence, and the primer 3′-end is a complementary sequence of 18 bp. Have

F3ADS and F3ADA were synthesized and purified by Pharmacia Biotech. For PCR, use KOD DNA polymerase (Toyobo), 100 μl reaction mixture with 1 μM FR-shuffling primers F3ADS and F3ADA, 5 μl each, 0.2 mM dNTPs, 1.0 mM MgCl ₂ , 2.5 U KOD DNA polymerase Using the attached buffer under conditions including 94 ° C for 30 seconds, 50 ° C for 1 minute, 74 ° C for 1 minute, 5 times, and adding 100 pmole external primers F3PrS and F3PrA, The same temperature cycle was performed 25 times. DNA fragments amplified by the PCR method were separated by agarose gel electrophoresis using 2% Nu Sieve GTG agarose (FMC Bio.Products).

  A piece of agarose containing a 424 bp long DNA fragment was cut out, 3 times the amount (ml / g) of TE was added, and the DNA fragment was purified by phenol extraction, phenol / chloroform extraction, and chloroform extraction. After purifying the purified DNA with ethanol, one third of the purified DNA was dissolved in 14 μl of water. The obtained PCR reaction mixture was digested with BalI and NcoI, and introduced into plasmid hATR5Hva / CVIDEC (BalI / NcoI) prepared by digesting with BalI and NcoI, and the nucleotide sequence was determined. The plasmid with the correct sequence was named hATR5Hvh / CVIDEC. The nucleotide sequence of the humanized H chain version “h” contained in the plasmid hATR5Hvh / CVIDEC and the corresponding amino acid sequence are shown in SEQ ID NO: 63. The amino acid sequence of version “h” is shown in SEQ ID NO: 64.

(Vi) Construction of humanized heavy chain versions “i” and “j” Versions “i” and “j” were prepared by replacing FR3 of version “a” with FR3 derived from another human antibody by the FR-shuffling method. did. Version “i” is FR3 derived from human antibody U95239 (DDBJ, Manheimer-Lory AJ., Unpublished), version “j” is L03147 (DDBJ, Collet TA. Et al., Proc. Natl. Acad. Sci. USA, 89, Two primers encoding FR3 were synthesized in order to replace FR3 derived from 10026-10030, 1992). Version “i” of the FR-shuffling primer F3MMS (SEQ ID NO: 65) has a sense DNA sequence, F3MMA (SEQ ID NO: 66) has an antisense DNA sequence, and the primer 3′-end is a complementary sequence of 18 bp. Have

Version “j” of the FR-shuffling primer F3BMS (SEQ ID NO: 67) has a sense DNA sequence, F3BMA (SEQ ID NO: 68) has an antisense DNA sequence, and the primer 3′-end is a complementary sequence of 18 bp. Have F3MMS, F3MMA, F3BMS and F3BMA were synthesized and purified by Pharmacia Biotech. PCR was performed using Ampli Taq Gold (Perkin-Elmer), and 100 μl of the reaction mixture was mixed with 1 μM FR-shuffling primers F3MMS and F3MMA, or F3BMS and F3BMA, 5 μl each, 0.2 mM dNTPs, 1.5 mM MgCl ₂ , Perform 5 cycles at 94 ° C for 30 seconds, 50 ° C for 1 minute, 74 ° C for 1 minute using the attached buffer with 2.5U Ampli Taq Gold, and 100 pmole external primer F3PrS and F3PrA were added and the same temperature cycle was performed 25 times.

  DNA fragments amplified by the PCR method were separated by agarose gel electrophoresis using 2% Nu Sieve GTG agarose (FMC Bio.Products). A piece of agarose containing a 424 bp long DNA fragment was cut out, 3 times the amount (ml / g) of TE was added, and the DNA fragment was purified by phenol extraction, phenol / chloroform extraction, and chloroform extraction. After purifying the purified DNA with ethanol, one third of the purified DNA was dissolved in 14 μl of water. The obtained PCR reaction mixture was digested with BalI and NcoI, and introduced into plasmid hATR5Hva / CVIDEC (BalI / NcoI) prepared by digesting with BalI and NcoI, and the nucleotide sequence was determined.

  Plasmids with the correct sequence were named hATR5Hvi / CVIDEC and hATR5Hvj / CVIDEC. The nucleotide sequence of the humanized H chain version “i” and the corresponding amino acid sequence and the version “i” amino acid sequence contained in the plasmid hATR5Hvi / CVIDEC are shown in SEQ ID NOs: 69 and 70, respectively. In addition, the nucleotide sequence of the humanized H chain version “j” and the corresponding amino acid sequence contained in the plasmid hATR5Hvj / CVIDEC and the amino acid sequence of version “j” are shown in SEQ ID NOs: 71 and 72, respectively.

(Vii) Construction of humanized heavy chain versions “b1” and “d1” Versions “b1” and “d1” are FR2s derived from different human antibodies by FR-shuffling method. It was made by replacing with. Two DNA primers encoding FR2 were prepared to replace those derived from human antibody P01742 (SWISS-PROT, Cunningham BA. Et al., Biochemistry, 9, 3161-3170, 1970). FR-shuffling vector F2MPS (SEQ ID NO: 73) has a sense DNA sequence, and F2MPA (SEQ ID NO: 74) has an antisense DNA sequence. Moreover, it has a mutually complementary sequence and has EcoT22I and BalI recognition sequences at both ends.

  F2MPS and F2MPA were synthesized and purified by Pharmacia Biotech. F2MPS and F2MPA were annealed and digested with EcoT22I and BalI. This was introduced into plasmids hATR5Hvb / CVIDEC (EcoT22I / BalI) and hATR5Hvd / CVIDEC (EcoT22I / BalI) prepared by digesting with EcoT22I and BalI, and the nucleotide sequence was determined. Plasmids with the correct sequence were named hATR5Hvb1 / CVIDEC and hATR5Hvd1 / CVIDEC. The nucleotide sequence of humanized H chain version “b1” contained in plasmid hATR5Hvb1 / CVIDEC and the corresponding amino acid sequence and version “b1” amino acid sequence are shown in SEQ ID NOs: 75 and 76, respectively. In addition, the nucleotide sequence of the humanized H chain version “d1” and the corresponding amino acid sequence contained in the plasmid hATR5Hvd1 / CVIDEC and the amino acid sequence of version “d1” are shown in SEQ ID NOs: 77 and 78, respectively.

(Viii) Construction of humanized heavy chain versions “b3” and “d3” Versions “b3” and “d3” are FR2s derived from different human antibodies by FR-shuffling method. It was made by replacing with. In order to replace FR2 derived from human antibody Z80844 (DDBJ, Thomsett AR. Et al., Unpublished), two DNA primers encoding FR2 were prepared. FR-shuffling vector F2VHS (SEQ ID NO: 79) has a sense DNA sequence and F2VHA (SEQ ID NO: 80) has an antisense DNA sequence. Moreover, it has a mutually complementary sequence and has EcoT22I and BalI recognition sequences at both ends. F2VHS and F2VHA were commissioned to Pharmacia Biotech for synthesis and purification.

  F2VHS and F2VHA were annealed and digested with EcoT22I and BalI. This was introduced into plasmids hATR5Hvb / CVIDEC (EcoT22I / BalI) and hATR5Hvd / CVIDEC (EcoT22I / BalI) prepared by digesting with EcoT22I and BalI, and the nucleotide sequence was determined. Plasmids with the correct sequence were named hATR5Hvb3 / CVIDEC and hATR5Hvd3 / CVIDEC. The nucleotide sequence of the humanized H chain version “b3” contained in the plasmid hATR5Hvb3 / CVIDEC, the corresponding amino acid sequence, and the version “b3” amino acid sequence are shown in SEQ ID NOs: 81 and 82, respectively. In addition, the nucleotide sequence of the humanized H chain version “d3” and the corresponding amino acid sequence contained in the plasmid hATR5Hvd3 / CVIDEC and the amino acid sequence of version “d3” are shown in SEQ ID NOs: 83 and 84, respectively.

(2) Construction of humanized antibody L chain V region (i) Version “a”
Humanized ATR5 antibody L chain was prepared by CDR-grafting by PCR. 7 for preparation of humanized antibody light chain (version “a”) with framework region derived from human antibody Z37332 (DDBJ, Welschof M et al., J. Immunol. Methods, 179, 203-214, 1995) PCR primers were used.

  CDR-grafting primers h5Lv1S (SEQ ID NO: 85) and h5Lv4S (SEQ ID NO: 86) are sense DNA sequences, and CDR grafting primers h5Lv2A (SEQ ID NO: 87), h5Lv3A (SEQ ID NO: 88) and h5Lv5A (SEQ ID NO: 89) are anti-senses. It has a sense DNA sequence and a 20 bp complementary sequence at each end of each primer. External primers h5LvS (SEQ ID NO: 90) and h5LvA (SEQ ID NO: 91) have homology with CDR grafting primers h5Lv1S and h5Lv5A. CDR-grafting primers h5Lv1S, h5Lv4S, h5Lv2A, h5Lv3A, h5Lv5A, h5LvS and h5LvA were synthesized and purified by Pharmacia Biotech.

The PCR solution was 120 mM Tris-HCl (pH 8.0), 10 mM KCl, 6 mM (NH ₄ ) ₂ SO ₄ , 0.1% Triton X-100, 0.001% BSA, 0.2 mM dNTPs (dATP, dGTP, dCTP, dTTP), 1 mM MgCl ₂ , 2.5 units of KOD DNA polymerase (Toyobo), 5pmole CDR grafting primers h5Lv1S, h5Lv2A, h5Lv3A, h5Lv4S, and h5Lv5A.

  PCR is performed using DNA Thermal Cycler 480 (Perkin-Elmer), and 5 CDR grafting is performed by performing 5 temperature cycles of 94 ° C for 30 seconds, 50 ° C for 1 minute, and 72 ° C for 1 minute. Primers were assembled. 100 pmole external primers h5LvS and h5LvA were added to this reaction mixture, and the assembled DNA fragments were obtained by performing 30 temperature cycles of 94 ° C for 30 seconds, 52 ° C for 1 minute, and 72 ° C for 1 minute. Amplified.

  The PCR reaction mixture was separated by agarose gel electrophoresis using 3% NuSieve GTG agarose (FMC BioProducts), and an agarose piece containing a DNA fragment having a length of about 400 bp was excised. The agarose piece was extracted with phenol and chloroform, and the DNA fragment was recovered by ethanol precipitation. The recovered DNA fragment was digested with restriction enzymes SplI (Takara Shuzo) and BglII (Takara Shuzo) at 37 ° C. for 4 hours. This digestion mixture was extracted with phenol and chloroform, and the DNA fragment was precipitated with ethanol and then dissolved in 10 μl of TE. The DNA ligation kit ver.2 (Takara Shuzo) was prepared by digesting the SplI-BglII DNA fragment containing the gene encoding the humanized antibody L chain V region prepared as described above with SplI and BglII. Used and reacted for 1 hour at 16 ° C. according to the attached recipe.

  This ligation mixture was added to 100 μl of E. coli JM109 competent cells (Nippon Gene) and allowed to stand on ice for 30 minutes and at 42 ° C. for 1 minute. Next, 300 μl of Hi-Competence Broth (Nippon Gene) was added and incubated at 37 ° C. for 1 hour, then this E. coli was seeded on a 100 μg / ml LBA agar medium and incubated overnight at 37 ° C. to obtain an E. coli transformant. . This transformant was cultured overnight at 37 ° C. in 3 ml of LBA medium, and plasmid DNA was prepared from the cell fraction using QIAprep Spin Plasmid Kit (QIAGEN).

  The nucleotide sequence of the cDNA coding region in the plasmid was determined by DNA Sequencer 373A (Perkin-Elmer) using Dye Terminator Cycle Sequencing FS Ready Reaction Kit (Perkin-Elmer). Using M13 Primer M4 (Takara Shuzo) and M13 Primer RV (Takara Shuzo) as sequencing primers, the sequence was determined by confirming the base sequence in both directions. A plasmid containing the gene encoding the humanized antibody L chain V region and having a BglII recognition sequence and Kozak sequence on the 5′-side and a SplI recognition sequence on the 3′-side was designated as hATR5Lva / CVIDEC. The nucleotide sequence (including the corresponding amino acid) of the humanized L chain version “a” is shown in SEQ ID NO: 92. The amino acid sequence of version “a” is shown in SEQ ID NO: 93.

(Ii) Versions “b” and “c”
Versions “b” and “c” were made by replacing FR3 of version “a” (FR-shuffling). Version “b” contains FR3 from human antibody S68699 (DDBJ, Hougs L et al., Exp. Clin. Immunogen et., 10, 141-151, 1993), and version “c” contains human antibody P01607 (SWISS-PROT FR3 from Epp O et al., Biochemistry, 14, 4943-4952, 1975).

  Primers F3SS (SEQ ID NO: 94) and F3SA (SEQ ID NO: 95) encoding FR3 of version “b” or primers F3RS (SEQ ID NO: 96) and F3RA (SEQ ID NO: 97) encoding FR3 of version “c” It has a complementary sequence and has recognition sequences for restriction enzymes KpnI and PstI at both ends. F3SS, F3SA, F3RS, and F3RA were commissioned to Pharmacia Biotech for synthesis and purification. Each 100 pmole of F3SS and F3SA, or F3RS and F3RA was annealed by treatment at 96 ° C. for 2 minutes and at 50 ° C. for 2 minutes to prepare double-stranded DNA fragments.

  These double-stranded DNA fragments were digested with the restriction enzyme KpnI (Takara Shuzo) at 37 ° C. for 1 hour, and then with the restriction enzyme PstI (Takara Shuzo) at 37 ° C. for 1 hour. The digestion mixture was extracted with phenol and chloroform, and the DNA fragment was precipitated with ethanol and then dissolved in TE.

  Plasmid hATR5Lva / CVIDEC was digested with restriction enzyme KpnI (Takara Shuzo) at 37 ° C. for 1 hour, and then with restriction enzyme PstI (Takara Shuzo) at 37 ° C. for 1 hour. The digestion mixture was separated by agarose gel electrophoresis using 1.5% NuSieve GTG agarose (FMC BioProducts), and agarose pieces containing about 3000 bp long DNA fragments were excised. The agarose piece was extracted with phenol and chloroform, and the DNA fragment was precipitated with ethanol and then dissolved in TE.

  The KpnI-PstI DNA fragment encoding version “b” or “c” of FR3 prepared as described above and the hATR5Lva / CVIDEC vector from which FR3 was removed by digestion with KpnI and PstI were used as DNA ligation kit ver.2 ( Takara Shuzo) was used for reaction for 1 hour at 16 ° C. according to the attached prescription.

  This ligation mixture was added to 100 μl of E. coli JM109 competent cells (Nippon Gene) and allowed to stand on ice for 30 minutes and at 42 ° C. for 1 minute. Next, 300 μl of Hi-Competence Broth (Nippon Gene) was added and incubated at 37 ° C. for 1 hour, then this E. coli was seeded on a 100 μg / ml LBA agar medium and incubated overnight at 37 ° C. to obtain an E. coli transformant. . This transformant was cultured overnight at 37 ° C. in 3 ml of LBA medium, and plasmid DNA was prepared from the cell fraction using QIAprep Spin Plasmid Kit (QIAGEN).

  The nucleotide sequence of the cDNA coding region in the plasmid was determined by DNA Sequencer 373A (Perkin-Elmer) using Dye Terminator Cycle Sequencing FS Ready Reaction Kit (Perkin-Elmer). Using M13 Primer M4 (Takara Shuzo) and M13 Primer RV (Takara Shuzo) as sequencing primers, the sequence was determined by confirming the base sequence in both directions.

  Plasmids containing the gene encoding version “b” or version “c” in which FR3 of humanized antibody L chain version “a” was replaced were named hATR5Lvb / CVIDEC and hATR5Lvc / CVIDEC, respectively. The nucleotide sequence of the humanized L chain version “b” contained in the plasmid hATR5Lvb / CVIDEC, the corresponding amino acid sequence, and the version “b” amino acid sequence are shown in SEQ ID NOs: 98 and 99. In addition, the nucleotide sequence of humanized L chain version “c” and the corresponding amino acid sequence and amino acid sequence of version “c” contained in plasmid hATR5Lvc / CVIDEC are shown in SEQ ID NOs: 100 and 101, respectively.

(Iii) Version “b1” and “b2”
Versions “b1” and “b2” were created by replacing FR2 of version “b”. Version “b1” contains FR2 derived from human antibody S65921 (DDBJ, Tonge DW et al., Year Immunol., 7, 56-62, 1993), and version “b2” contains human antibody X93625 (DDBJ, CoxJP et al., Eur. FR2 from J. Immunol., 24, 827-836, 1994) was used.

  Primers F2SS (SEQ ID NO: 102) and F2SA (SEQ ID NO: 103) encoding FR2 of version “b1”, or primers F2XS (SEQ ID NO: 104) and F2XA (SEQ ID NO: 105) encoding FR2 of version “b2” are mutually It has a complementary sequence and has recognition sequences for restriction enzymes AflII and SpeI at both ends. F2SS, F2SA, F2XS and F2XA were synthesized by Pharmacia Biotech. Each 100 pmole of F2SS and F2SA or F2XS and F2XA was annealed by treatment at 96 ° C. for 2 minutes and at 50 ° C. for 2 minutes to prepare double-stranded DNA fragments.

These double-stranded DNA fragments were digested with restriction enzymes AflII (Takara Shuzo) and SpeI (Takara Shuzo) at 37 ° C. for 1 hour. The digestion mixture was extracted with phenol and chloroform, and the DNA fragment was precipitated with ethanol and then dissolved in TE.
Plasmid hATR5Lvb / CVIDEC was digested with restriction enzymes AflII (Takara Shuzo) and SpeI (Takara Shuzo) at 37 ° C. for 1 hour. The digestion mixture was separated by agarose gel electrophoresis using 1.5% NuSieve GTG agarose (FMC BioProducts), and agarose pieces containing about 3000 bp long DNA fragments were excised. The agarose piece was extracted with phenol and chloroform, and the DNA fragment was precipitated with ethanol and then dissolved in TE.

  The AflII-SpeI DNA fragment encoding the FR2 of version “b1” or “b2” prepared as described above and the hATR5Lvb / CVIDEC vector from which FR2 was removed by digestion with AflII and SpeI were used as a DNA ligation kit ver.2 ( Takara Shuzo) was used for reaction for 1 hour at 16 ° C. according to the attached prescription.

  This ligation mixture was added to 100 μl of E. coli JM109 competent cells (Nippon Gene) and allowed to stand on ice for 30 minutes and at 42 ° C. for 1 minute. Next, 300 μl of Hi-Competence Broth (Nippon Gene) was added and incubated at 37 ° C. for 1 hour, then this E. coli was seeded on a 100 μg / ml LBA agar medium and incubated overnight at 37 ° C. to obtain an E. coli transformant. . This transformant was cultured overnight at 37 ° C. in 4 ml of LBA medium, and plasmid DNA was prepared from the bacterial cell fraction using QIAprep Spin Plasmid Kit (QIAGEN).

  The nucleotide sequence of the cDNA coding region in the plasmid was determined by DNA Sequencer 373A (Perkin-Elmer) using Dye Terminator Cycle Sequencing FS Ready Reaction Kit (Perkin-Elmer). Using M13 Primer M4 (Takara Shuzo) and M13 Primer RV (Takara Shuzo) as sequencing primers, the sequence was determined by confirming the base sequence in both directions.

  Plasmids containing the gene encoding version “b1” or version “b2” in which FR2 of humanized antibody L chain version “b” was replaced were named hATR5Lvb1 / CVIDEC and hATR5Lvb2 / CVIDEC, respectively. The nucleotide sequence of the humanized L chain version “b1” contained in the plasmid hATR5Lvb1 / CVIDEC and the corresponding amino acid sequence and version “b1” amino acid sequence are shown in SEQ ID NOs: 106 and 107, respectively. Further, the nucleotide sequence of the humanized L chain version “b2” and the corresponding amino acid sequence and the amino acid sequence of version “b2” contained in the plasmid hATR5Lvb2 / CVIDEC are shown in SEQ ID NOs: 108 and 109, respectively.

(3) Construction of humanized antibody expression vector (i) Combination of humanized H chain and chimeric L chain
The plasmid hATR5Hva / CVIDEC containing the H chain V region is digested with NheI and SalI, and the cDNA fragment of the humanized H chain V region is recovered, and the chATR-5 antibody expression plasmid vector, chATR5 / N5KG4P is digested with NheI and SalI. Was introduced into chATR5 / N5KG4P (SalI / NheI). The thus prepared plasmid was designated as hHva-chLv / N5KG4P. Plasmid hATR5Hvb / CVIDEC containing H chain V region is digested with NheI and SalI, cDNA fragment of humanized H chain V region is recovered, chATR-5 antibody expression plasmid vector, chATR5 / N5KG4P is digested with NheI and SalI Was introduced into chATR5 / N5KG4P (SalI / NheI). The thus prepared plasmid was designated as hHvb-chLv / N5KG4P.

  Plasmids hATR5Hvc / CVIDEC, hATR5Hvd / CVIDEC and hATR5Hve / CVIDEC containing the H chain V region were digested with NheI and SalI, and the cDNA fragment of the humanized H chain V region was recovered, chATR-5 antibody expression plasmid vector, chATR5 / N5KG4P was introduced into chATR5 / N5KG4P (SalI / NheI) prepared by digesting with NheI and SalI. The plasmids thus prepared were designated as hHvc-chLv / N5KG4P, hHvd-chLv / N5KG4P and hHve-chLv / N5KG4P.

  Plasmids hATR5Hvf / CVIDEC and hATR5Hvh / CVIDEC containing the H chain V region were digested with NheI and SalI, and the cDNA fragment of the humanized H chain V region was recovered, and the chATR-5 antibody expression plasmid vector, chATR5 / N5KG4P was converted to NheI and It was introduced into chATR5 / N5KG4P (SalI / NheI) prepared by digestion with SalI. The plasmids thus prepared were named hHvf-chLv / N5KG4P and hHvh-chLv / N5KG4P.

  The plasmids hATR5Hvi / CVIDEC and hATR5Hvj / CVIDEC containing the H chain V region were digested with NheI and SalI, and the cDNA fragment of the humanized H chain V region was recovered. The chATR-5 antibody expression plasmid vector, chATR5 / N5KG4P It was introduced into chATR5 / N5KG4P (SalI / NheI) prepared by digestion with SalI. The plasmids thus prepared were named hHvi-chLv / N5KG4P and hHvj-chLv / N5KG4P.

  Plasmids hATR5Hvb1 / CVIDEC and hATR5Hvd1 / CVIDEC containing the H chain V region were digested with NheI and SalI, and a cDNA fragment of the humanized H chain V region was recovered, and chATR-5 antibody expression plasmid vector, chATR5 / N5KG4P It was introduced into chATR5 / N5KG4P (SalI / NheI) prepared by digestion with SalI. The plasmids thus prepared were named hHvb1-chLv / N5KG4P and hHvd1-chLv / N5KG4P.

(Ii) Combination of humanized L chain and chimeric H chain Humanized L chain was evaluated by expressing humanized antibody in combination with chimeric H chain using antibody expression vector N5KG4P .

  Plasmids hATR5Lva / CVIDEC, hATR5Lvb / CVIDEC, hATR5Lvc / CVIDEC, hATR5Lvb1 / CVIDEC, hATR5Lvb2 / CVIDEC were digested with restriction enzymes BglII (Takara Shuzo) and SplI (Takara Shuzo) for 2 to 3 hours. The digestion mixture was separated by agarose gel electrophoresis using 1.5% or 2% NuSieve GTG agarose (FMC BioProducts), and agarose strips containing about 400 bp long DNA fragments were excised. The agarose piece was extracted with phenol and chloroform, and the DNA fragment was precipitated with ethanol and then dissolved in TE.

A DNA ligation kit ver.2 (Takara Shuzo Co., Ltd.) was used in accordance with the attached prescription using SplI-BglII DNA fragment containing the gene encoding each version of the humanized L chain V region and chATR5Hv / N5KG4P digested with SplI and BglII. The reaction was allowed to proceed at 0 ° C. for 1 hour and ligated.
The ligation mixture was added to 100 μl of E. coli JM109 competent cells (Nippon Gene), and allowed to stand on ice for 30 minutes and at 42 ° C. for 1 minute. Next, 300 μl of Hi-Competence Broth (Nippon Gene) was added and incubated at 37 ° C. for 1 hour, then the E. coli was plated on 100 μg / ml LBA agar medium and incubated overnight at 37 ° C. to obtain an E. coli transformant.

  This transformant was cultured overnight at 37 ° C. in 250 ml or 500 ml of LBA medium, and plasmid DNA was prepared from the bacterial cell fraction using Plasmid Maxi Kit (QIAGEN). Plasmids into which genes encoding these chimeric H chain and humanized L chain were introduced were chHv-hLva / N5KG4P, chHv-hLvb / N5KG4P, chHv-hLvc / N5KG4P, chHv-hLvb1 / N5KG4P and chHv-hLvb2 / N5KG4P, respectively. Named.

(Iii) Combination of humanized H chain and humanized L chain
The plasmid hATR5Hva / CVIDEC containing the H chain V region is digested with NheI and SalI, and the cDNA fragment of the humanized H chain V region is recovered and contains the sequence of the humanized ATR-5 antibody L chain version “a” cDNA. The plasmid chHv-hLva / N5KG4P was introduced into hLva / N5KG4P (SalI / NheI) prepared by digesting with NheI and SalI. The plasmid thus prepared was designated as hHva-hLva / N5KG4P.

  Plasmids hATR5Hvb / CVIDEC and hATR5Hvc / CVIDEC containing the H chain V region are digested with NheI and SalI, and the cDNA fragment of the humanized H chain V region is recovered, and the humanized ATR-5 antibody L chain version “a” cDNA The plasmid chHv-hLva / N5KG4P containing the sequence was introduced into hLva / N5KG4P (SalI / NheI) prepared by digesting with NheI and SalI. The plasmids thus prepared were named hHvb-hLva / N5KG4P and hHvc-hLva / N5KG4P.

  Plasmids hATR5Hvb / CVIDEC, hATR5Hvd / CVIDEC and hATR5Hve / CVIDEC containing H chain V region are digested with NheI and SalI, and the cDNA fragment of humanized H chain V region is recovered, and humanized ATR-5 antibody L chain version The plasmid chHv-hLvb / N5KG4P containing the sequence of “b” cDNA was introduced into hLvb / N5KG4P (SalI / NheI) prepared by digesting with NheI and SalI. The plasmids thus prepared were designated as hHvb-hLvb / N5KG4P, hHvd-hLvb / N5KG4P and hHve-hLvb / N5KG4P.

  Plasmids hATR5Hvf / CVIDEC, hATR5Hvg / CVIDEC and hATR5Hvh / CVIDEC containing H chain V region are digested with NheI and SalI to recover cDNA fragment of humanized H chain V region, and humanized ATR-5 antibody L chain version The plasmid chHv-hLvb / N5KG4P containing the sequence of “b” cDNA was introduced into hLvb / N5KG4P (SalI / NheI) prepared by digesting with NheI and SalI. The plasmids thus prepared were designated as hHvf-hLvb / N5KG4P, hHvg-hLvb / N5KG4P and hHvh-hLvb / N5KG4P.

  Plasmids hATR5Hvi / CVIDEC and hATR5Hvj / CVIDEC containing the H chain V region are digested with NheI and SalI, and the cDNA fragment of the humanized H chain V region is recovered, and the humanized ATR-5 antibody L chain version “b” cDNA The plasmid chHv-hLvb / N5KG4P containing the sequence was introduced into hLvb / N5KG4P (SalI / NheI) prepared by digesting with NheI and SalI. The plasmids thus prepared were named hHvi-hLvb / N5KG4P and hHvj-hLvb / N5KG4P.

  Plasmids hATR5Hvb1 / CVIDEC and hATR5Hvd1 / CVIDEC containing the H chain V region are digested with NheI and SalI, and the cDNA fragment of the humanized H chain V region is recovered, and the humanized ATR-5 antibody L chain version “b” cDNA The plasmid chHv-hLvb / N5KG4P containing the sequence was introduced into hLvb / N5KG4P (SalI / NheI) prepared by digesting with NheI and SalI. The plasmids thus prepared were designated as hHvb1-hLvb / N5KG4P and hHvd1-hLvb / N5KG4P.

  Plasmids hATR5Hvb3 / CVIDEC and hATR5Hvd3 / CVIDEC containing the H chain V region are digested with NheI and SalI, and the cDNA fragment of the humanized H chain V region is recovered, and the humanized ATR-5 antibody L chain version “b” cDNA The plasmid chHv-hLvb / N5KG4P containing the sequence was introduced into hLvb / N5KG4P (SalI / NheI) prepared by digesting with NheI and SalI. The plasmids thus prepared were designated as hHvb3-hLvb / N5KG4P and hHvd3-hLvb / N5KG4P.

  Plasmid hATR5Hvb / CVIDEC containing H chain V region is digested with NheI and SalI, and cDNA fragment of humanized H chain V region is recovered, and humanized ATR-5 antibody L chain version “b1” and “b2” cDNA Were introduced into hLvb1 / N5KG4P (SalI / NheI) and hLvb2 / N5KG4P (SalI / NheI) prepared by digesting the plasmids chHv-hLvb1 / N5KG4P and chHv-hLvb2 / N5KG4P with the sequences of NheI and SalI. The plasmids thus prepared were designated as hHvb-hLvb1 / N5KG4P and hHvb-hLvb2 / N5KG4P.

  Plasmid hATR5Hvi / CVIDEC containing H chain V region is digested with NheI and SalI, and cDNA fragment of humanized H chain V region is recovered, and humanized ATR-5 antibody L chain version “b1” and “b2” cDNA Were introduced into hLvb1 / N5KG4P (SalI / NheI) and hLvb2 / N5KG4P (SalI / NheI) prepared by digesting the plasmids chHv-hLvb1 / N5KG4P and chHv-hLvb2 / N5KG4P with the sequences of NheI and SalI. The plasmids thus prepared were named hHvi-hLvb1 / N5KG4P and hHvi-hLvb2 / N5KG4P.

(4) Transfection into COS-7 cells In order to evaluate the antigen-binding activity and neutralizing activity of the humanized antibody, the expression plasmid was transiently expressed in COS-7 cells.
The constructed expression plasmid vector was transduced into COS-7 cells by electroporation using a Gene Pulser apparatus (Bio-Rad). 50 μg or 20 μg of plasmid was added to 0.78 ml of COS-7 cells suspended at a cell concentration of 1 × 10 ⁷ cells / ml in PBS, and pulses were applied at a capacitance of 1,500 V and 25 μF.

After a 10-minute recovery period at room temperature, electroporated cells are suspended in DMEM medium (GIBCO) containing 5% Ultra Low IgG fetal calf serum (GIBCO) and placed in 10cm or 15cm culture dishes. And cultured in a CO ₂ incubator. After 24 hours of culture, the culture supernatant was removed by aspiration, and a serum-free medium HBCHO (Irvine Scientific) was newly added. After a further 72 or 96 hours of culture, the culture supernatant was collected and cell debris was removed by centrifugation.

(5) Antibody purification
Purification of antibodies from the culture supernatant of COS-7 cells was performed using AffiGel ProteinA MAPSII kit (Bio-Rad) or rProtein A Sepharose FastFlow (Pharmacia Biotech). Purification using the AffiGel Protein A MAPSII kit was performed according to the instructions attached to the kit. Purification using rProtein A Sepharose Fast Flow was performed as follows.

The column was equilibrated by loading 1 ml of rProtein A Sepharose Fast Flow onto the column and flowing 10 times the amount of TBS. After the culture supernatant of COS-7 cells was applied to the equilibrated column, the column was washed with 10 times the amount of TBS. Next, the adsorbed antibody fraction was eluted from the column by flowing 13.5 ml of 2.5 mM HCl (pH 3.0). The eluate was neutralized by adding 1.5 ml of 1M Tris-HCl (pH 8.0).
The purified antibody fraction was subjected to ultrafiltration using Centriprep 30 or 100 (amicon) 2 to 3 times to replace the solvent with TBS and finally concentrated to about 1.5 ml.

Example 4. Antibody quantification and activity evaluation (1) Measurement of antibody concentration by ELISA An ELISA plate for antibody concentration measurement was prepared as follows. A goat prepared in each well of a 96-well plate for ELISA (Maxisorp, NUNC) to a concentration of 1 μg / ml with a solid phase buffer (0.1 M NaHCO ₃ , 0.02% NaN ₃ , pH 9.6) (hereinafter referred to as CB) Immobilized with 100 μl of anti-human IgGγ antibody (BioSource) and 200 μl of dilution buffer (50 mM Tris-HCl, 1 mM MgCl ₂ , 0.1 M NaCl, 0.05% Tween20, 0.02% NaN ₃ , 1% bovine serum albumin (BSA), pH 8 .1) After blocking in (hereinafter referred to as DB), the culture supernatant or purified antibody of COS-7 cells in which the antibody was expressed was serially diluted in DB and added to each well.

After incubation at room temperature for 1 hour and washing with Dulbecco's PBS (hereinafter referred to as RB) containing 0.05% Tween 20, 100 μl of alkaline phosphatase-conjugated goat anti-human IgGγ antibody (BioSource) diluted 1000 times with DB was added. After incubating at room temperature for 1 hour and washing with RB, Sigma104 (p-nitrophenyl phosphate, SIGMA) was dissolved in substrate buffer (50 mM NaHCO ₃ , 10 mM MgCl ₂ , pH 9.8) to 1 mg / ml. (Hereinafter referred to as substrate solution) was added, and the absorbance at 405/655 nm was measured with a microplate reader (Bio Rad). IgG4κ (The Binding Site) was used as a standard for concentration measurement.

(2) Measurement of antigen binding ability A Cell ELISA plate for measuring antigen binding was prepared as follows. The cells used were human bladder cancer cells J82 (ATCC HTB-1). 1 × 10 ⁶ J82 cells were seeded in 60 holes of a 96-well plate for cell culture. This was cultured in a CO ₂ incubator for 1 day (RPMI1640 medium containing 10% fetal bovine serum (GIBCO)), and the cells were allowed to adhere. The culture solution was discarded and each well was washed twice with 300 μl of PBS. 100 μl of PBS containing 4% paraformaldehyde (hereinafter referred to as PFA / PBS) was added to each well and allowed to stand on ice for 10 minutes to immobilize the cells.

  PFA / PBS was discarded, and each well was washed twice with 300 μl of PBS and then blocked with 250 μl of DB. The culture supernatant or purified antibody was serially diluted with DB and 100 μl was added to each well. After incubating at room temperature for 2 hours and washing with RB, 100 μl of alkaline phosphatase-conjugated goat anti-human IgGγ antibody (BioSource) diluted 1000 times with DB was added. After incubating at room temperature for 1 hour and washing with RB, the substrate solution was added, and then the absorbance at 405/655 nm was measured with a Microplate Reader (Bio-Rad).

(3) Measurement of neutralizing activity The neutralizing activity of mouse antibody, chimeric antibody and humanized antibody was measured using the factor Xa production inhibitory activity by human placenta-derived thromboplastin and Thromborel S (Behringwerke AG) as an index. That is, 10 μl of 1.25 mg / ml Thromborel S and 10 μl of antibody diluted to an appropriate concentration were added 60 μl of buffer solution (TBS containing 5 mM CaCl ₂ and 0.1% BSA), and the reaction was allowed to proceed for 1 hour at room temperature in a 96-well plate. I let you. To this were added 10 μl each of 3.245 μg / ml human factor X (Celsus Laboratories) and 82.5 ng / ml human factor VIIa (Enzyme Research), and further reacted at room temperature for 1 hour.

The reaction was stopped by adding 10 μl of 0.5 M EDTA. 50 μl of the chromogenic substrate solution was added thereto, and the absorbance at 405/655 nm was measured with a Microplate Reader (Bio Rad). The reaction was performed at room temperature for 1 hour, and the absorbance at 405/655 nm was measured again. The change in absorbance for 1 hour without addition of antibody was defined as 100% activity, and the residual activity (%) was calculated from the change in absorbance.
For the chromogenic substrate solution, dissolve the test team chromogenic substrate S-2222 (Chromogenix) according to the package insert, dilute it twice with purified water, and then mix 1: 1 with polybrene solution (0.6 mg / ml hexadimethylin bromide, SIGMA). Prepared.

(4) Activity evaluation (i) Combination of humanized H chain version “a” and chimeric L chain An antibody (a-ch) combining humanized H chain version “a” and chimeric L chain was prepared. When the antigen binding ability was examined by cell ELISA, the binding amount to the antigen decreased at the high concentration side (FIG. 1). Antigen neutralization ability by inhibiting FXa production was also weaker than that of the positive control chimeric antibody (ch-ch) (Fig. 2). Therefore, we decided to upgrade the humanized H chain by FR-shuffling. The chimeric antibody used here was evaluated using an antibody expressed and purified in COS-7 cells.

(Ii) Combination of humanized L chain version “a” and chimeric H chain An antibody (ch-a) combining humanized L chain version “a” and chimeric H chain is prepared and antigens are obtained by cell ELISA. When the binding ability was examined, antigen binding activity equal to or higher than that of the chimeric antibody was observed (FIG. 1). On the other hand, the antigen neutralizing ability was weaker than that of the positive control chimeric antibody (FIG. 2). Therefore, we decided to upgrade the humanized L chain by FR-shuffling. The chimeric antibody used here was evaluated using an antibody expressed and purified in COS-7 cells.

(Iii) Combination of humanized H chain version “a” and humanized L chain version “a” Antibody combining humanized H chain version “a” and humanized L chain version “a” (aa ) And the antigen binding ability was examined by cell ELISA. As a result, the binding amount to the antigen decreased at the high concentration side (FIG. 3). Antigen neutralization ability due to FXa production inhibition was also considerably weaker than that of the positive control chimeric antibody (FIG. 4). Therefore, we decided to upgrade the humanized H and L chains by FR-shuffling. The chimeric antibody used here was evaluated using an antibody expressed and purified in COS-7 cells.

(Iv) Combination of humanized H chain version “b”, “c” and “d” with chimeric L chain
Prepare antibodies ("b-ch", "c-ch", and "d-ch") that combine humanized H chain and chimeric L chain that have been upgraded by FR-shuffling, and use cell ELISA to generate antigen When the binding ability was examined, “d-ch” showed antigen-binding activity equivalent to that of the chimeric antibody, and “b-ch” and “c-ch” showed slightly inferior antigen-binding activity (FIG. 5, 6). On the other hand, the antigen neutralizing ability was almost the same for “b-ch” and slightly weaker for “d-ch” compared to the positive control chimeric antibody. Version “c-ch” was much weaker than the chimeric antibody (FIG. 7). Therefore, the humanized H chain versions “b” and “d” are versions that are considered to exhibit high activity in the humanized H chain.

(V) Combination of humanized H chain version “b” and humanized L chain version “a”
An antibody (ba) was prepared by combining humanized H chain version “b” and humanized L chain version “a”, which were upgraded by FR-shuffling, and antigen binding ability was examined by cell ELISA. The amount of binding to the antigen decreased with the concentration (FIG. 5). On the other hand, the antigen neutralizing ability was considerably weaker than that of the positive control chimeric antibody (FIG. 8). Therefore, “ba” was a version showing higher activity than “aa”. The chimeric antibody used here was evaluated using an antibody expressed and purified in COS-7 cells.

(Vi) Combination of humanized L chain versions “b” and “c” with chimeric H chain Antibodies combining humanized L chain versions “b” and “c” with chimeric H chain (each “ch- When b "and" ch-c ") were prepared, both antibodies showed activity equivalent to that of the chimeric antibody in both antigen binding ability and antigen neutralizing ability (Figs. 9 and 10). Therefore, the versions “b” and “c” were selected as humanized antibody L chain candidates. It is considered that the version “b” having one amino acid residue derived from a mouse antibody is superior in antigenicity to the version “c”. The chimeric antibody used here was evaluated using an antibody expressed and purified in CHO cells DG44, and this antibody was used as a positive control in the subsequent evaluations.

(Vii) Combination of humanized H chain version “b” and humanized L chain versions “b” and “c” Humanized H chain version “b” is replaced with humanized L chain versions “b” and “b”. Antibodies combined with “c” (“bb” and “bc”, respectively) were prepared, and antigen binding ability and antigen neutralizing ability were measured. Both antibodies showed slightly inferior activity to the chimeric antibody in both antigen binding ability and antigen neutralizing ability (FIGS. 11 and 12).

(Viii) Combination of humanized heavy chain version “b” and “d” with humanized light chain version “b”
An antibody (“bb” and “db”), which is a combination of humanized heavy chain and humanized light chain version “b” upgraded by FR-shuffling, was prepared, and the antigen binding ability was examined by cell ELISA. However, “db” showed an antigen-binding activity equivalent to that of the chimeric antibody, and “bb” showed a slightly inferior antigen-binding activity at a high concentration (FIG. 13). On the other hand, the antigen neutralizing ability “bb” was slightly weaker than the positive control chimeric antibody, and “db” was significantly weaker than the chimeric antibody (FIG. 14). Therefore, it was shown that “bb” is a version with high antigen activity neutralizing ability and “db” is a version with high antigen binding ability.

(Ix) Combination of humanized H chain version “e” with chimeric L chain and humanized L chain version “b” Humanized L chain version “e” is converted into chimeric L chain and humanized version “b” Antibody ("e-ch" and "eb", respectively) was produced. The antigen-binding ability of "e-ch" showed the same activity as that of the chimeric antibody. And the antigen binding ability was almost lost (FIG. 15). The antigen activity neutralizing ability of “e-ch” was considerably weaker than that of the chimeric antibody (FIG. 16). Therefore, the combination of the H chain version “e” and the L chain version “b” was considered to be bad.

(X) Combination of humanized H chain version “f”, “g” and “h” with humanized L chain version “b” Humanized H chain version “f”, “g” and “h” Was prepared with the humanized L chain version “b” (“fb”, “gb” and “hb” respectively), and the antibody expression level of “fb” and “hb” was very high. It was low. For versions “f” and “h”, an antibody combined with a chimeric L chain was also produced, but was not expressed. “Gb” reached saturation from a low concentration and showed weaker antigen-binding ability than the chimeric antibody (FIG. 17). The antigen neutralizing ability of “gb” was considerably weaker than that of the chimeric antibody (FIG. 18).

(Xi) Combination of humanized H chain version “b1” and “d1” with humanized L chain version “b” Humanized H chain version “b1” and “d1” are humanized L chain version When antibodies combined with “b” (“b1-b” and “d1-b”, respectively) were prepared, both antibodies were hardly expressed. For these, antibodies combined with chimeric L chains were also produced, but were not expressed.

(Xii) Combination of humanized H chain versions “b3” and “d3” with humanized L chain version “b” Humanized H chain versions “b3” and “d3” are humanized L chain versions When the antibodies combined with “b” (“b3-b” and “d3-b”, respectively) were prepared, the antigen-binding ability of “d3-b” was slightly inferior to that of the chimeric antibody. The antigen binding ability of was further inferior (FIG. 19). Although the antigen neutralizing ability of “b3-b” exceeded that of “bb”, it did not reach the activity of the chimeric antibody, and “d3-b” remained at the same level as “bb” (Fig. 20).

(Xiii) Combination of humanized H chain version “i” and “j” with chimeric L chain and humanized L chain version “b” Humanized H chain version “i” and “j” are combined with chimeric L chain Antibodies ("i-ch" and "j-ch" respectively) combined with humanized L chain version "b"("ib" and "jb" respectively) Antigen neutralizing ability was measured. The antigen-binding ability of each antibody was almost the same as that of the chimeric antibody (FIGS. 21 and 22). “I-ch” showed an antigen neutralizing ability exceeding the activity of the chimeric antibody, and “j-ch” had a much weaker activity than the chimeric antibody (FIG. 23). “Ib” showed activity equivalent to that of the chimeric antibody, and “jb” showed considerably weaker activity than the chimeric antibody (FIG. 24).

(Xiv) Humanized light chain versions “b1” and “b2”
When humanized L chain versions “b1” and “b2” were combined with the chimeric H chain (“ch-b1” and “ch-b2”, respectively), both antibodies were equivalent to the chimeric antibody. Antigen binding ability was shown (FIG. 25). Regarding the antigen neutralizing ability, “ch-b1” showed an activity equivalent to that of the chimeric antibody, and “ch-b2” showed an activity slightly higher than that of the chimeric antibody at the high concentration side (FIG. 26). Both versions “b1” and “b2” can be candidates for humanized antibody L chain, but version “b2” is superior in that it has stronger activity.

(Xv) Combination of humanized H chain version “b” and humanized L chain version “b2” Antibody combining humanized H chain version “b” with humanized L chain version “b2” (“ b-b2 ") was prepared and the antigen binding ability and antigen neutralizing ability were measured. The antigen binding ability was slightly inferior to that of the chimeric antibody (FIG. 27). Although the antigen neutralizing ability exceeded the activity of “bb”, it did not reach the activity of “ib” (FIG. 28).

(Xvi) Combination of humanized H chain version “i” and humanized L chain version “b1” or “b2” Humanized H chain version “i” or humanized L chain version “b1” or Antibodies combined with “b2” (“i-b1” and “i-b2”, respectively) were prepared, and antigen binding ability and antigen neutralizing ability were measured. The antigen-binding ability of “i-b2” was almost the same as that of the chimeric antibody, and “i-b1” was slightly inferior (FIG. 29). In addition, the antigen neutralizing ability of “i-b1” and “i-b2” was higher than that of the chimeric antibody and “ib”, and was stronger in the order of “i-b2”> “i-b1” (FIG. 30). .

Example 5. Preparation and activity evaluation of CHO cell-producing humanized antibody (1) Establishment of CHO stable-producing cell line To establish stable-producing cell lines of humanized antibodies (bb, ib and i-b2) The antibody expression gene vector was introduced into CHO cells (DG44) conditioned to serum medium.

Plasmid DNA, hHvb-hLvb / N5KG4P, hHvi-hLvb / N5KG4P and hHvi-hLvb2 / N5KG4P were cleaved with restriction enzyme SspI (Takara Shuzo), extracted with phenol and chloroform, and purified by ethanol precipitation. The linearized expression gene vector was introduced into DG44 cells using an electroporation apparatus (Gene Pulser; Bio Rad). DG44 cells were suspended in PBS at a cell density of 1 × 10 ⁷ / ml, 10 or 50 μg of the above DNA was added to about 0.8 ml of this suspension, and pulses were given at a capacitance of 1,500 V and 25 μF. .

After a recovery period of 10 minutes at room temperature, the treated cells were suspended in CHO-S-SFMII medium containing hypoxanthine-thymidine (GIBCO) (hereinafter HT), and two 96-well flat bottom plates (Falcon ) Was seeded at 100 μl / well and cultured in a CO ₂ incubator. 8-9 hours after the start of culture, add CHO-S-SFMII medium containing HT and 1 mg / ml GENETICIN (GIBCO) to 100 μl / well, convert to 500 μg / ml GENETICIN selective medium, and introduce the antibody gene Selected. A medium in which 1/2 volume of medium was replaced with fresh medium once every 3 to 4 days, and when cell growth was observed after 4 to 5 days after the conversion to selective medium after about 2 weeks. A part of the culture supernatant was collected. The antibody concentration expressed in the culture supernatant was measured by the above-described antibody concentration measurement ELISA, and cells with high antibody production were selected.

(2) Mass purification of humanized antibody 500ml of humanized antibody ("bb", "ib" and "i-b2") expressing DG44 cell lines selected as described above using 2L roller bottle (CONING) / After culturing in a bottle of CHO-S-SFMII medium for several days, the culture solution was collected, fresh CHO-S-SFMII medium was added, and the culture was repeated. The culture solution was centrifuged to remove cell debris and filtered through a 0.22 μm or 0.45 μm filter. This was repeated to obtain a culture supernatant of about 2 L in total. The antibody was purified from the obtained culture supernatant using a ConSep LC100 system (Millipore) connected to a Protein A affinity column (Poros).

(3) Measurement of antibody concentration by ELISA An ELISA plate for antibody concentration measurement was prepared as follows. Each well of a 96-well plate for ELISA (Maxisorp, NUNC) was immobilized with 100 μl of goat anti-human IgGγ antibody (BioSource) prepared with CB to a concentration of 1 μg / ml, blocked with 200 μl of DB, and then the antibody was expressed. COS cell culture supernatant or purified antibody was serially diluted with DB and added to each well.

  After incubating at room temperature for 1 hour and washing with RB, 100 μl of alkaline phosphatase-conjugated goat anti-human IgGγ antibody (BioSource) diluted 1000 times with DB was added. After incubating at room temperature for 1 hour and washing with RB, 100 μl of the substrate solution was added, and the absorbance at 405/655 nm was measured with a microplate reader (Bio Rad). IgG4κ (The Binding Site) was used as a standard for concentration measurement.

(4) Measurement of antigen binding ability Cell ELISA plates for antigen binding measurement were prepared as follows. The cells used were human bladder cancer cells J82 (ATCC HTB-1). 1 × 10 ⁶ J82 cells were seeded in a 96-well plate for cell culture. This was cultured for 1 day in a CO2 incubator (RPMI1640 medium containing 10% fetal bovine serum (GIBCO)) to adhere the cells. The culture solution was discarded and each well was washed twice with PBS. 100 μl of PFA / PBS was added to each well and allowed to stand on ice for 10 minutes to immobilize the cells.

  PFA / PBS was discarded, and each well was washed twice with 300 μl of PBS and then blocked with 250 μl of DB. Based on the above measurement results, the purified antibody was serially diluted with DB at a common ratio of 2 from 10 μg / ml, and 100 μl was added to each well. After incubating at room temperature for 2 hours and washing with RB, 100 μl of alkaline phosphatase-conjugated goat anti-human IgGγ antibody (BioSource) diluted 1000 times with DB was added. After incubating at room temperature for 1 hour and washing with RB, 100 μl of the substrate solution was added, and then the absorbance at 405/655 nm was measured with a Microplate Reader (Bio-Rad).

(5) Measurement of TF neutralization activity (factor Xa production inhibitory activity) Factor Xa production inhibitory activity of humanized antibodies was measured using the factor Xa production inhibitory activity of human placenta-derived thromboplastin and Thromborel S (Behringwerke AG) as an index. . That is, 10 μl of 5 mg / ml Thromborel S and 10 μl of antibody were added 60 μl of a buffer solution (TBS containing 5 mM CaCl ₂ and 0.1% BSA), and reacted in a 96-well plate at room temperature for 1 hour. The antibody was serially diluted in buffer with a common ratio of 5 from 200 μg / ml.

10 μl each of 3.245 μg / ml human factor X (Celsus Laboratories) and 82.5 ng / ml human factor VIIa (Enzyme Research) were added thereto, and the mixture was further reacted at room temperature for 45 minutes. The reaction was stopped by adding 10 μl of 0.5 M EDTA. 50 μl of the chromogenic substrate solution was added thereto, and the absorbance at 405/655 nm was measured with a Microplate Reader (Bio Rad). The mixture was reacted at room temperature for 30 minutes, and the absorbance at 405/655 nm was measured again. The change in absorbance for 30 minutes with no antibody added was defined as 100% activity, and the residual activity (%) was calculated from the change in absorbance.
The chromogenic substrate solution was prepared by dissolving test team chromogenic substrate S-2222 (Chromogenix) according to the package insert and mixing 1: 1 with polybrene solution (0.6 mg / ml hexadimethylin bromide, SIGMA).

(6) Measurement of TF neutralizing activity (factor X binding inhibitory activity) The factor X binding inhibitory activity of humanized antibodies is a complex of TF and Factor VIIa using human placenta-derived thromboplastin, Thromborel S (Behringwerke AG). The factor X binding inhibitory activity was measured using the factor Xa production inhibitory activity of the complex as an index. That is, 10 μl of Thromborel S 5 mg / ml and 10 μl 82.5 ng / ml human Factor VIIa (Enzyme Research) 60 μl of buffer solution (TBS containing 5 mM CaCl 2 and 0.1% BSA) were added to a 96-well plate at room temperature. For 1 hour in advance.

10 μl of the antibody solution was added thereto and reacted at room temperature for 5 minutes, and then 10 μl of 3.245 μg / ml human Factor X (Celsus Laboratories) was added, and further reacted at room temperature for 45 minutes. The antibody was serially diluted with a buffer solution at a common ratio of 2 from 200 μg / ml. The reaction was stopped by adding 10 μl of 0.5 M EDTA. 50 μl of the chromogenic substrate solution was added thereto, and the absorbance at 405/655 nm was measured with a Microplate Reader (Bio Rad). The mixture was reacted at room temperature for 30 minutes, and the absorbance at 405/655 nm was measured again. The change in absorbance for 30 minutes with no antibody added was defined as 100% activity, and the residual activity (%) was calculated from the change in absorbance.
The chromogenic substrate solution was prepared by dissolving test team chromogenic substrate S-2222 (Chromogenix) according to the package insert and mixing 1: 1 with polybrene solution (0.6 mg / ml hexadimethylin bromide, SIGMA).

(7) Measurement of TF neutralizing activity (plasma coagulation inhibitory activity) The humanized antibody's TF neutralizing activity (plasma coagulation inhibitory activity) is measured using prothrombin time using human placenta-derived thromboplastin and Thromborel S (Behringwerke AG) as an index. It was measured. Specifically, 100 μl of human plasma (Cosmo Bio) was placed in a sample cup, 50 μl of antibody diluted to various concentrations was added thereto, and the mixture was heated at 37 ° C. for 3 minutes. 50 μl of 1.25 mg / ml Thromborel S previously heated to 37 ° C. was added to initiate plasma clotting. This coagulation time was measured with Amelung KC-10A (both MC Medical) connected with Amelung CR-A.

The antibody was serially diluted with TBS containing 0.1% BSA (hereinafter referred to as BSA-TBS) at a common ratio of 2 from 80 μg / ml. The measured clotting time without addition of antibody was defined as 100% TF plasma clotting activity, and TF residual activity was calculated from each clotting time when the antibody was added by a calibration curve plotting the Thromborel S concentration and clotting time.
A calibration curve was prepared by measuring various Thromborel S concentrations and their clotting time. Appropriately diluted Thromborel S, add 50 μl of BSA-TBS to 50 μl, heat at 37 ° C. for 3 minutes, add 100 μl of human plasma pre-warmed to 37 ° C. to start clotting and measure clotting time did. Thromborel S was serially diluted with Hank's buffer (GIBCO) containing 25 mM CaCl2 at a common ratio of 2 from 6.25 mg / ml. The Thromborel S concentration is plotted on the horizontal axis and the clotting time is plotted on the logarithmic graph on the vertical axis, which is used as a calibration curve.

(8) Evaluation of activity All humanized antibodies of “bb”, “ib” and “i-b2” had an activity equal to or higher than that of the chimeric antibody (FIG. 31). In the factor Xa production inhibitory activity, factor X binding inhibitory activity and plasma coagulation inhibitory activity, the humanized antibodies “bb”, “ib” and “i-b2” have an activity equal to or higher than that of the chimeric antibody, The activity was strong in the order of “i-b2”>“ib”> “bb” (FIGS. 32, 33 and 34).

Example 6. Kinetic analysis of interaction between TF and anti-TF antibody using BIACORE
BIACORE was used to analyze the kinetics of antigen-antibody reaction. Recombinant Protein G was immobilized on a sensor chip, and an antibody was bound thereto. Recombinant TF purified as an antigen (soluble TF with FLAG peptide tag added to 1-219) (hereinafter referred to as soluble TF) and analytes of soluble TF prepared at various concentrations It was. Kinetic parameters (dissociation rate constant kdiss and association rate constant kass) were calculated from the obtained sensorgram. Regarding kinetic analysis, we referred to “Kinetic analysis of monoclonal antibody-antigen interactions with a new biosensor based analytical system” (karlsson, R. et al. (1991) J. Immunol. Methods 145, 229-240.). .

(1) Immobilization of protein G on the sensor chip Protein G (ZYMED) is immobilized on the sensor chip CM5 (BIACORE).
HBS-EP buffer (0.01 M HEPES pH 7.4, 0.15 M NaCl, 3 mM EDTA, 0.005% polysorbate 20 (v / v)) (BIACORE) was used as the running buffer, and the flow rate was 5 μL / min. The carboxyl group of carboxymethyldextran on the sensor chip CM5 is 100 μL of 0.05 M N-hydroxysuccinimide (NHS) /0.2 M hydrochloric acid N-ethyl-N ′-(3-dimethylaminopropyl) -carbodiimide (EDC) Activated by the JECT.

  Subsequently, 10 μL of 50 μg / mL Protein G was injected, and this was performed 3 times to immobilize. Protein G was dissolved in 10 mM Tris-HCl buffer (pH 7.5) to a concentration of 10 mg / ml, and diluted with 50 mM / mL with 10 mM sodium acetate buffer (pH 4.0). Further, 100 μL of 1.0 M ethanolamine-hydrochloric acid (pH 8.5) was injected to block excess active groups. To this, 10 μL of 0.1 M glycine-hydrochloric acid buffer (pH 2.5) and 10 μL of 10 mM hydrochloric acid were injected, and the non-covalently bound substance was washed. This was carried out for each flow cell, and 10 μL of 72 nM humanized anti-TF antibody version “ib2” was injected and confirmed to bind about 1000 RU.

(2) Interaction between immobilized anti-TF antibody and human TF Human TF was purified by expressing in CHO cells a FLAG peptide linked to the C-terminus of amino acid sequence 1-219. This was used as soluble human TF.

Using HBS-EP buffer as a running buffer, 10 μL of 72 nM antibody solution was injected at a flow rate of 20 μL / min to immobilize the antibody. Antibody dilution was performed using HBS-EP buffer. To this, 40 μL of soluble human TF solution with various concentrations was injected at a flow rate of 30 μL / min, and analysis was performed with 80 seconds of injection as the binding phase, and then switched to HBS-EP buffer for 120 seconds of dissociation phase. . After completion of the dissociation phase, the sensor chip was regenerated by injecting 10 μL of 20 mM hydrochloric acid. This binding / dissociation / regeneration was regarded as one cycle of analysis, and sensorgrams were obtained for various antibodies. The soluble human TF solution was prepared using HBS-EP buffer at concentrations of 250 nM, 125 nM, 62.5 nM, 31.3 nM, 15.6 nM, 7.81 nM and 3.91 nM. As a blank, a sensorgram obtained by injecting the HBS-EP buffer used for dilution was used.
The above was performed for each of the flow cells 1 to 3.

(3) Kinetic analysis of the interaction The target data file was read, and the reaction pattern of the target reaction region was compared by overwriting using the HBS-EP buffer sensorgram as a baseline. In addition, BIACORE's analysis application software “BIAevaluation 2.1” (Pharmacia) that calculates kinetic parameters (association rate constant kass and dissociation rate constant kdiss) by curve fitting is used to analyze the kinetics of the interaction. went. When determining the binding rate constant kass, analysis model type 4 was used (BIAevaluation 2.1 Software Handbook, A1 to A5). From the values calculated from the respective flow cells, kinetic parameters of various antibodies were obtained. Table 6 shows the results (average value ± standard deviation of values calculated from each flow cell).

Example 7. Measurement of reactivity of humanized anti-TF antibody to human TF Dot blot hybridization method ("Revised protein experiment method for molecular biology" (Yodosha) Tadaomi Takenoha / Ed. P .101), reactivity to non-modified TF, non-reduced modified TF, and modified modified TF was examined. The TF used was one in which the extracellular region with a FLAG tag was expressed in CHO cells and purified (shTF). shTF is added to the following three buffer solutions (Buffer A: 10 mM Tris-HCl, pH 8.0; Buffer B: 10 mM Tris-HCl, pH 8.0, 8 M urea; Buffer C: 10 mM Tris-HCl, pH 8) .0, 8M urea, 5 mM DTT). Those treated with buffer A were non-denaturing TF, while non-reducing denatured TF was treated with buffer B, and reducing denatured TF was prepared with buffer C. Each sample was treated at room temperature for 24 hours. After the treatment, the sample was blotted on a nitrocellulose membrane (Bio-Rad).

Samples were blotted onto 0.5 μl, 1 μl and 2 μl (3 μg / ml) membranes and the membranes were air dried. Blocked with DB (50 mM Tris-HCl, pH 8.1, 0.15 M NaCl, 1 mM MgCl ₂ , 0.05% (v / v) Tween 20, 0.02% (w / v) NaN ₃ , 1% (w / v) BSA) . The membrane was reacted with DB containing humanized anti-TF antibody or DB (control). The plate was washed with PBS containing 0.05% (v / v) Tween 20, and reacted with DB containing peroxidase-labeled anti-human IgG antibody (DAKO). After washing with PBS containing 0.05% (v / v) Tween 20, it was treated with ECL Western Blotting reagent (Amersham) and exposed to X-ray film for 30 seconds.

  As shown in FIG. 35, the chimeric anti-TF antibody and the humanized anti-TF antibody (versions “bb”, “ib” and “ib2”) react to all non-denaturing TF, non-denaturing modified TF, and reducing denatured TF. did.

Example 8. Confirmation of antithrombotic activity in rat acute DIC model The antithrombotic activity of the anti-TF antibody was confirmed in a thromboplastin-induced DIC model using rats. Specifically, a DIC model was prepared by continuously injecting human thromboplastin solution into SD male rats at a dose of 40 mg / kg over 3 hours intravenously. Anti-TF antibody (chimeric and humanized anti-TF antibody i-b2) was administered intravenously at a dose of 0.2 mg / kg, 5 minutes before the start of infusion of the thromboplastin solution. Citrated blood was collected from the abdominal aorta 15 minutes after the end of continuous infusion of thromboplastin solution. Platelet count, activated partial thromboplastin time (aPTT), fibrinogen concentration (Fib), soluble fibrin monomer complex (sFMC) concentration, thrombin / Antithrombin III complex (TAT) concentration was measured.

As a result, as shown in Table 7, decreased platelet count, prolonged aPTT, decreased fibrinogen concentration, and increased sFMC and TAT concentrations were observed by continuous infusion of thromboplastin, and a clear hypercoagulable state was exhibited. In contrast, both chimeric and humanized anti-TF antibodies strongly suppressed these changes almost similarly.
From these results, it was shown that the humanized anti-TF antibody is useful as an antithrombotic agent.

Reference Example 1. Preparation of anti-TF monoclonal antibody
1. Purification of human TF Purification of TF from human placenta was performed according to the method of Ito et al. (Ito, T. et al. J. Biochem. 114, 691-696, 1993). That is, human placenta was homogenized in Tris buffered saline (TBS, pH 7.5) containing 10 mM benzamidine chloride, 1 mM phenylmethylsulfonyl chloride, 1 mM diisopropylfluorophosphate and 0.02% sodium azide, and the precipitate was cooled with acetone. The defatted powder obtained was suspended in the above buffer containing 2% Triton X-100 to solubilize TF.

From this supernatant, affinity chromatography was performed using a Concanavalin A-Sepharose 4B column (Pharmacia) and a Sepharose 4B column (Pharmacia) conjugated with an anti-TF antibody to obtain purified TF. This was concentrated with an ultrafiltration membrane (PM-10, Amicon) and stored at 4 ° C. as a purified sample.
The TF content in the purified preparation was quantified with a Sandwich ELISA combining a commercially available anti-TF monoclonal antibody (American Diagnostica) and a polyclonal antibody (American Diagnostica) with recombinant TF as a standard.
The purity of the purified sample was confirmed by silver staining of SDS-PAGE using 4-20% concentration gradient polyacrylamide gel.

2. Immunization and hybridoma preparation Purified human TF (approximately 70 μg / ml) is mixed with an equal volume of Freund's complete adjuvant (Difco), then subcutaneously in the abdomen of a 5-week-old Balb / c male mouse (Charles River Japan). Then, immunization was performed so that TF was 10 μg / mouse. On days 12, 18, and 25, TF mixed with Freund's incomplete adjuvant was boosted subcutaneously to 5 μg / mouse, and the final immunization was intraperitoneally with 5 μg / mouse of TF solution diluted with PBS on the 32nd. Administered.

Three days after the final immunization, spleen cells were prepared from 4 mice and fused with the mouse myeloma cell line P3U1 whose cell number was 1/5 using the polyethylene glycol method. The fused cells were suspended in RPMI-1640 medium containing 10% fetal calf serum (hereinafter referred to as RPMI-medium) (Lifetech oriental) and seeded in a 96-well plate with 400 holes per mouse (about 400 cells / hole). . After the fusion, on the 1st, 2nd, 3rd and 5th days, by replacing half of the medium with RPMI-medium (hereinafter referred to as HAT-medium) containing HAT (Dainippon Pharmaceutical) and condimed H1 (Boehringer Mannheim GmbH), Hybridoma HAT selection was performed.
Hybridomas selected by the screening method described below were cloned by performing two limiting dilutions.

  For limiting dilution, 0.8 cells were seeded per well in two 96-well plates. The clones were selected by measuring the TF binding activity and TF neutralization activity shown below for the holes that were confirmed to be a single colony by microscopic examination. The obtained clone was acclimated from HAT-medium to RPMI-medium, and after confirming that there was no decrease in antibody production ability due to acclimation, limiting dilution was performed again and complete cloning was performed. By the above operation, hybridomas producing 6 types of antibodies (ATR-2, 3, 4, 5, 7, and 8) that strongly inhibit the binding between the TF / factor VIIa complex and factor X were established.

3. Production of ascites and purification of antibody Production of ascites of the established hybridoma was performed according to a conventional method. That is, the hybridoma ¹⁰⁶ which were passaged in vitro, and transplanted in advance mineral oil intraperitoneally Balb / c male mice that had been administered in two intraperitoneal. Ascites fluid was collected from mice whose abdomen was enlarged 1-2 weeks after transplantation.
Antibody purification from ascites was performed using a ConSepLC100 system (Millipore) equipped with a Protein A column (NGK).

4.Cell-ELISA
A human bladder cancer cell line J82 (Fair DS et al., J. Biol. Chem., 262, 11692-11698, 1987), which is known to express TF at a high level, was introduced from ATCC. It was subcultured and maintained under the conditions of ℃, 5% CO ₂ and 100% humidity.
Cell-ELISA plates are seeded at a concentration of 10 ⁵ cells / well in a 96-well plate at a concentration of 10 ⁵ cells / well. After culturing for 1 day under the above conditions, the medium is removed and washed twice with phosphate buffered saline (PBS). Then, a 4% paraformaldehyde solution (PFA) was added, and the mixture was allowed to stand for 10 minutes under ice-cooling to prepare it. After removing PFA and washing with PBS, Tris buffer (Blocking buffer) containing 1% BSA and 0.02% sodium azide was added and stored at 4 ° C. until use.

  Cell-ELISA was performed as follows. That is, the blocking buffer solution was removed from the plate prepared as described above, and an anti-TF antibody solution or a hybridoma culture supernatant was added and reacted at room temperature for 1.5 hours. After washing with PBS containing 0.05% Tween20, alkaline phosphatase-conjugated goat anti-mouse IgG (H + L) (Zymed) was reacted for 1 hour, and after washing, 1 mg / ml p-nitrophenyl phosphate disodium (Sigma) The amount of anti-TF antibody bound to J82 cells was quantified by measuring the absorbance at 405 nm 1 hour after addition.

5. TF neutralization activity measurement system using factor Xa activity as an index
10 μl human placenta-derived thromboplastin solution (5 mg / ml) (Thromborel S) (Boehring) and 10 μl factor VIIa solution in Tris buffered saline (TBS: pH 7.6) containing 50 μl 5 mM CaCl ₂ and 0.1% bovine serum albumin (82.5 ng / ml) (American Diagnostica) was added and allowed to react at room temperature for 1 hour to form a TF / Factor VIIa complex, and then diluted to a predetermined concentration of 10 μl of anti-TF antibody solution or hybridoma culture supernatant Then, 10 μl of Factor X solution (3.245 μg / ml) (Celsus Laboratorise) was added and reacted for 45 minutes, and the reaction was stopped by adding 10 μl of 0.5 M EDTA. 50 μl of 2 mM S-2222 solution (Daiichi Kagaku) was added thereto, and the change in absorbance at 405 nm for 30 minutes was taken as Factor Xa production activity of TF. In this method, the activity of an antibody that inhibits the binding between TF / Factor VIIa complex and Factor X can be measured.

6. Plasma coagulation inhibitory activity measurement system Using commercially available normal human plasma (Kohjin Bio), mix 50 μl of an appropriately diluted anti-TF antibody solution with 100 μl and react at 37 ° C. for 3 minutes, then 50 μl of human placenta The derived thromboplastin solution (1.25 mg / ml) was added, and the time until the plasma coagulated was measured with a plasma clotting time measuring device (CR-A: Amelung). 7. Determination of antibody isotype For the hybridoma culture supernatant or purified antibody, the antibody isotype was confirmed using a mouse monoclonal antibody isotyping kit (manufactured by Amersham), and the results are shown below.

Reference Example 2 Production Method of Soluble Human TF Soluble human TF (shTF) was produced as follows.
A gene encoding a human TF penetrating region (amino acid at position 220) and below substituted with FLAG peptide M2 is inserted into an expression vector for mammalian cells (including neomycin resistance gene and DHFR gene) and inserted into CHO cells. Introduced. The cDNA sequence of human TF was referred to the report of James H. Morrissey et al. (Cell (1987) 50, 129-135). The gene sequence and amino acid sequence of this soluble human TF are shown in SEQ ID NO: 151. Drug selection with G418, selection of expression cells, and further amplification of expression with methotrexate established shTF expression cells.

  The cells were cultured in a serum-free medium CHO-S-SFMII (GIBCO) to obtain a culture supernatant containing shTF. Dilute 2-fold with the same volume of 40 mM Tris-HCl buffer (pH 8.5) and add to a Q-Sepharose Fast Flow column (100 mL, Pharmacia Biotech) equilibrated with 20 mM Tris-HCl buffer (pH 8.5). After washing with the same buffer containing 0.1M NaCl, the NaCl concentration was adjusted to 0.3M, and shTF was eluted from the column. Ammonium sulfate was added to the obtained shTF fraction to a final concentration of 2.5 M, and contaminating proteins were precipitated by centrifugation (10,000 rpm, 20 minutes). The supernatant was added to Butyl TOYOPEARL (30 mL, TOSOH) and washed with 50 mM Tris-HCl buffer (pH 6.8) containing 2.5 M ammonium sulfate.

  In a 50 mM Tris-HCl buffer (pH 6.8), the ammonium sulfate concentration was linearly decreased from 2.5 M to 0 M to elute shTF. The peak fraction containing shTF was concentrated with Centri-Prep 10 (Amicon). The concentrated solution was added to a TSKgel G3000SWG column (21.5 × 600 mm, TOSOH) equilibrated with 20 mM Tris-HCl buffer (pH 7.0) containing 150 mM NaCl, and the peak fraction of shTF was collected. This was sterilized by filtration through a 0.22 μm membrane filter to obtain soluble human TF (shTF). The sample concentration was calculated assuming that the molar extinction coefficient of the absorbance of the sample at 280 nm was ε = 40,130 and the molecular weight was 43,210.

The contents described in <223> of the sequence listing free text sequence listing are as follows.
SEQ ID NO: 1: Primer MHC-G1
SEQ ID NO: 2: Primer MHC-G2a
SEQ ID NO: 3: Primer MKC
SEQ ID NO: 4: M13 primer M4
SEQ ID NO: 5: M13 primer RV
SEQ ID NO: 6: Amino acid sequence of H chain V region of anti-TF mouse monoclonal antibody ATR-2 and nucleotide sequence encoding the same SEQ ID NO: 7: Amino acid of H chain V region of anti-TF mouse monoclonal antibody ATR-3 Sequence and base sequence encoding it

SEQ ID NO: 8: Amino acid sequence of the H chain V region of the anti-TF mouse monoclonal antibody ATR-4 and the nucleotide sequence encoding the same SEQ ID NO: 9: Amino acid of the H chain V region of the anti-TF mouse monoclonal antibody ATR-5 Sequence and nucleotide sequence encoding it SEQ ID NO: 10: amino acid sequence of the H chain V region of anti-TF mouse monoclonal antibody ATR-7 and nucleotide sequence encoding it SEQ ID NO: 11: anti-TF mouse monoclonal antibody ATR- Amino acid sequence of H chain V region of 8 and nucleotide sequence encoding it SEQ ID NO: 12: Amino acid sequence of L chain V region of anti-TF mouse monoclonal antibody ATR-2 and nucleotide sequence encoding it SEQ ID NO: 13: Amino acid sequence of the L chain V region of the anti-TF mouse monoclonal antibody ATR-3 and the base sequence encoding it

SEQ ID NO: 14: Amino acid sequence of the L chain V region of the anti-TF mouse monoclonal antibody ATR-4 and the nucleotide sequence encoding the same SEQ ID NO: 15: Amino acid of the L chain V region of the anti-TF mouse monoclonal antibody ATR-5 Sequence and nucleotide sequence encoding the same SEQ ID NO: 16: amino acid sequence of the L chain V region of anti-TF mouse monoclonal antibody ATR-7 and the nucleotide sequence encoding the same SEQ ID NO: 17: anti-TF mouse monoclonal antibody ATR- Amino acid sequence of L chain V region of 8 and nucleotide sequence encoding it SEQ ID NO: 18: Primer ch5HS
SEQ ID NO: 19: primer ch5HA
SEQ ID NO: 20: Primer ch5LS
SEQ ID NO: 21: Primer ch5LA

SEQ ID NO: 22: CDR grafting primer hR5Hv1S
SEQ ID NO: 23: CDR grafting primer hR5Hv28
SEQ ID NO: 24: CDR grafting primer hR5Hv4S
SEQ ID NO: 25: CDR grafting primer hR5Hv3A
SEQ ID NO: 26: CDR grafting primer hR5Hv5A
SEQ ID NO: 27: Primer hR5HvPrS
SEQ ID NO: 28: Primer hR5HvPrA
SEQ ID NO: 29: Amino acid sequence of humanized H chain V region version “a” and the nucleotide sequence encoding it Amino acid sequence of humanized H chain V region version “a”

SEQ ID NO: 31: FR shuffling primer F3RFFS
SEQ ID NO: 32: FR shuffling primer F3RFBS
SEQ ID NO: 33: FR shuffling primer F3RFFA
SEQ ID NO: 34: FR shuffling primer F3RFBA
SEQ ID NO: 35: FR shuffling primer F3NMFS
SEQ ID NO: 36: FR shuffling primer F3NMBS
SEQ ID NO: 37: FR shuffling primer F3NMFA
SEQ ID NO: 38: FR shuffling primer F3NMBA
SEQ ID NO: 39: Amino acid sequence of humanized H chain V region version “b” and the nucleotide sequence encoding the amino acid sequence SEQ ID NO: 40: Amino acid sequence of humanized H chain V region version “b” SEQ ID NO: 41: Amino acid sequence of humanized H chain V region version “c” and nucleotide sequence encoding it

SEQ ID NO: 42: Amino acid sequence of humanized H chain V region version “c” SEQ ID NO: 43: FR shuffling primer F3EPS
SEQ ID NO: 44: FR shuffling primer F3EPA
SEQ ID NO: 45: Primer F3PrS
SEQ ID NO: 46: Primer F3PrA
SEQ ID NO: 47: FR shuffling primer F3vHS
SEQ ID NO: 48: FR shuffling primer F3vHA
SEQ ID NO: 49: Amino acid sequence of humanized H chain V region version “d” and the nucleotide sequence encoding the amino acid sequence SEQ ID NO: 50: Amino acid sequence of humanized H chain V region version “d” SEQ ID NO: 51: Amino acid sequence of humanized H chain V region version “e” and its base sequence

SEQ ID NO: 52: Amino acid sequence of humanized H chain V region version “e” SEQ ID NO: 53: FR shuffling primer F3SSS
SEQ ID NO: 54: FR shuffling primer F3SSA
SEQ ID NO: 55: FR shuffling primer F3CDS
SEQ ID NO: 56: FR shuffling primer F3CDA
SEQ ID NO: 57: Amino acid sequence of humanized H chain V region version “f” and the nucleotide sequence encoding the amino acid sequence SEQ ID NO: 58: Amino acid sequence of humanized H chain V region version “f” SEQ ID NO: 59: Amino acid sequence of humanized H chain V region version “g” and its nucleotide sequence SEQ ID NO: 60: Amino acid sequence of humanized H chain V region version “g” SEQ ID NO: 61: FR shuffling primer F3ADS
SEQ ID NO: 62: FR shuffling primer F3ADA
SEQ ID NO: 63: Amino acid sequence of humanized H chain V region version “h” and base sequence encoding it

SEQ ID NO: 64: Amino acid sequence of humanized H chain V region version “h” SEQ ID NO: 65: FR shuffling primer F3MMS
SEQ ID NO: 66: FR shuffling primer F3MMA
SEQ ID NO: 67: FR shuffling primer F3BMS
SEQ ID NO: 68: FR shuffling primer F3BMA
SEQ ID NO: 69: Amino acid sequence of humanized H chain V region version “i” and the nucleotide sequence encoding the amino acid sequence SEQ ID NO: 70: Amino acid sequence of humanized H chain V region version “i” SEQ ID NO: 71: Amino acid sequence of humanized H chain V region version “j” and the nucleotide sequence encoding the amino acid sequence SEQ ID NO: 72: Amino acid sequence of humanized H chain V region version “j”

SEQ ID NO: 73: FR shuffling primer F2MPS
SEQ ID NO: 74: FR shuffling primer F2MPA
SEQ ID NO: 75: Amino acid sequence of humanized H chain V region version “b1” and the nucleotide sequence encoding the amino acid sequence SEQ ID NO: 76: Amino acid sequence of humanized H chain V region version “b1” SEQ ID NO: 77: Amino acid sequence of humanized H chain V region version “d1” and its nucleotide sequence SEQ ID NO: 78: Amino acid sequence of humanized H chain V region version “d1” SEQ ID NO: 79: FR shuffling primer F2VHS
SEQ ID NO: 80: FR shuffling primer F2VHA
SEQ ID NO: 81: Amino acid sequence of humanized H chain V region version “b3” and the nucleotide sequence encoding it

SEQ ID NO: 82: Amino acid sequence of humanized H chain V region version “b3” SEQ ID NO: 83: Amino acid sequence of humanized H chain V region version “d3” and the nucleotide sequence encoding the same SEQ ID NO: 84: Amino acid sequence of humanized H chain V region version “d3” SEQ ID NO: 85: FR shuffling vector Lv1S
SEQ ID NO: 86: FR shuffling vector h5Lv4S
SEQ ID NO: 87: FR shuffling vector h5Lv2A
SEQ ID NO: 88: FR shuffling vector h5Lv3A
SEQ ID NO: 89: FR shuffling primer h5Lv5A
SEQ ID NO: 90: Primer h5LvS SEQ ID NO: 91: Primer h5LvA

SEQ ID NO: 92: Amino acid sequence of humanized L chain V region version “a” and the nucleotide sequence encoding the amino acid sequence SEQ ID NO: 93: Amino acid sequence of humanized L chain V region version “a” SEQ ID NO: 94: FR shuffling primer F3SS
SEQ ID NO: 95: FR shuffling primer F3SA
SEQ ID NO: 96: FR shuffling primer F3RS
SEQ ID NO: 97: FR shuffling primer F3RA
SEQ ID NO: 98: Amino acid sequence of humanized L chain V region version “b” and the base sequence that encodes it SEQ ID NO: 99: Amino acid sequence of humanized L chain V region version “b”

SEQ ID NO: 100: Amino acid sequence of humanized L chain V region version “c” and the nucleotide sequence encoding the amino acid sequence SEQ ID NO: 101: Amino acid sequence of humanized L chain V region version “c” SEQ ID NO: 102: FR shuffling primer F2SS
SEQ ID NO: 103: FR shuffling primer F2SA
SEQ ID NO: 104: FR shuffling primer F2XS
SEQ ID NO: 105: FR shuffling primer F2XA
SEQ ID NO: 106: Amino acid sequence of humanized L chain V region version “b1” and the nucleotide sequence encoding the same SEQ ID NO: 107: Amino acid sequence of humanized L chain V region version “b1” SEQ ID NO: 108 : Amino acid sequence of humanized L chain V region version “b2” and the nucleotide sequence encoding the same SEQ ID NO: 109: Amino acid sequence of humanized L chain V region version “b2”

SEQ ID NO: 110: Amino acid sequence of FR1 of all versions of humanized H chain V region SEQ ID NO: 111: Amino acid sequence of FR2 of humanized H chain V region versions “a” to “j” SEQ ID NO: 112: Human RF2 amino acid sequence of typed H chain V region versions “b1” and “d1” SEQ ID NO: 113: Amino acid sequence of RF2 of humanized H chain V region versions “b3” and “d3” SEQ ID NO: 114: human FR3 amino acid sequence of typed H chain V region version “a” SEQ ID NO: 115: FR3 amino acid sequence of humanized H chain V region versions “b”, “b1” and “b3” SEQ ID NO: 116: FR3 amino acid sequence of humanized H chain V region version “c”

SEQ ID NO: 117: Humanized H chain V region version “d”, “d1” and “d3” FR3 amino acid sequence SEQ ID NO: 118: Humanized H chain V region version “e” FR3 amino acid SEQ ID NO: 119: Amino acid sequence of FR3 of humanized H chain V region version “f” SEQ ID NO: 120: Amino acid sequence of FR3 of humanized H chain V region version “g” SEQ ID NO: 121: Human type FR3 amino acid sequence SEQ ID NO: 122: humanized H chain V region version “j” FR3 amino acid sequence SEQ ID NO: 123: humanized H chain V region version “j” FR3 amino acid sequence

SEQ ID NO: 124: Humanized H chain V region all versions of FR4 amino acid sequence SEQ ID NO: 125: Humanized L chain V region all versions of FR1 amino acid sequence SEQ ID NO: 126: Humanized L chain V region FR2 amino acid sequence of version “a”, “b” and “c” SEQ ID NO: 127: humanized L chain V region version FR2 amino acid sequence of version “b1” SEQ ID NO: 128: humanized L chain V region Amino acid sequence of FR2 of version “b2” SEQ ID NO: 129: Amino acid sequence of FR3 of humanized L chain V region version “a” SEQ ID NO: 130: Humanized L chain V region version “b”, “b1” And the amino acid sequence of FR3 of “b2” SEQ ID NO: 131: The amino acid sequence of FR3 of humanized L chain V region version “c”

SEQ ID NO: 132: Amino acid sequence of humanized L chain V region all version FR4 SEQ ID NO: 133: Amino acid sequence of humanized H chain V region all version CDR1 SEQ ID NO: 134: All versions of humanized H chain V region The amino acid sequence of CDR2 of SEQ ID NO: 135: The amino acid sequence of CDR3 of all versions of humanized H chain V region SEQ ID NO: 136: The amino acid sequence of CDR1 of all versions of humanized L chain V region SEQ ID NO: 137: Human type Amino acid sequence of CDR2 of all versions of humanized L chain V region SEQ ID NO: 138: Amino acid sequence of CDR3 of all versions of humanized L chain V region SEQ ID NO: 139: H chain V region of anti-TF mouse monoclonal antibody ATR-2 Amino acid sequence

SEQ ID NO: 140: Amino acid sequence of the H chain V region of the anti-TF mouse monoclonal antibody ATR-3 SEQ ID NO: 141: Amino acid sequence of the H chain V region of the anti-TF mouse monoclonal antibody ATR-4 SEQ ID NO: 142: Anti -TF mouse monoclonal antibody ATR-5 H chain V region amino acid sequence SEQ ID NO: 143: Anti-TF mouse monoclonal antibody ATR-7 H chain V region amino acid sequence SEQ ID NO: 144: Anti-TF mouse monoclonal antibody ATR -8 H chain V region amino acid sequence SEQ ID NO: 145: Anti-TF mouse monoclonal antibody ATR-2 L chain V region amino acid sequence SEQ ID NO: 146: Anti-TF mouse monoclonal antibody ATR-3 L chain V Amino acid sequence of the region SEQ ID NO: 147: amino acid sequence of the L chain V region of the anti-TF mouse monoclonal antibody ATR-4 SEQ ID NO: 148: amino acid sequence of the L chain V region of the anti-TF mouse monoclonal antibody ATR-5

SEQ ID NO: 149: Amino acid sequence of L chain V region of anti-TF mouse monoclonal antibody ATR-7 SEQ ID NO: 150: Amino acid sequence of L chain V region of anti-TF mouse monoclonal antibody ATR-8 SEQ ID NO: 151: Yes Amino acid sequence of soluble human TF and nucleotide sequence encoding it SEQ ID NO: 152: Amino acid sequence of soluble human TF

FIG. 1 is a graph comparing the antigen binding activities of an H chain chimeric / L chain chimeric antibody, an H chain humanized a version / L chain chimeric antibody, and an H chain chimeric / L chain humanized a version antibody. FIG. 2 compares the neutralizing activity against human TF of H chain chimeric / L chain chimeric antibody, H chain humanized a version / L chain chimeric antibody, and H chain chimeric / L chain humanized a version antibody. It is a graph. FIG. 3 is a graph comparing the antigen binding activities of the H chain chimeric / L chain chimeric antibody and the H chain humanized a version / L chain humanized a version antibody. FIG. 4 shows the neutralizing activity of anti-TF-mouse monoclonal antibody ATR-5, H chain chimeric / L chain chimeric antibody, and H chain humanized version a / L chain humanized version a antibody against human TF. It is the graph compared. FIG. 5 compares the antigen binding activities of the H chain chimeric / L chain chimeric antibody, the H chain humanized version b / L chain chimeric antibody, and the H chain humanized version b / L chain humanized version a antibody. It is a graph. FIG. 6 is a graph showing the antigen binding activity of H chain chimeric / L chain chimeric antibody, H chain humanized version c / L chain chimeric antibody, and H chain humanized version d / L chain chimeric antibody. FIG. 7 shows H chain chimeric / L chain chimeric antibody, H chain humanized version b / L chain chimeric antibody, H chain humanized version c / L chain chimeric antibody, H chain humanized version d / L chain chimera It is the graph which compared the neutralizing activity with respect to human TF of an antibody. FIG. 8 is a graph comparing the neutralizing activity against human TF of the H chain chimeric / L chain chimeric antibody and the H chain humanized version b / L chain humanized version a antibody. FIG. 9 is a graph comparing the antigen binding activities of the H chain chimeric / L chain chimeric antibody, the H chain chimeric / L chain humanized version b antibody, and the H chain chimeric / L chain humanized version c antibody. FIG. 10 shows a comparison of neutralizing activity against human TF of H chain chimeric / L chain chimeric antibody, H chain chimeric / L chain humanized version b antibody, and H chain chimeric / L chain humanized version c antibody. It is a graph to show. FIG. 11 shows antigen binding of H chain chimeric / L chain chimeric antibody, H chain humanized version b / L chain humanized version b antibody, and H chain humanized version b / L chain humanized version c antibody. It is the graph which compared activity. FIG. 12 shows humans of H chain chimeric / L chain chimeric antibody, H chain humanized version b / L chain humanized version b antibody, and H chain humanized version b / L chain humanized version c antibody. It is the graph which compared the neutralization activity with respect to TF. FIG. 13 shows antigen binding of H chain chimeric / L chain chimeric antibody, H chain humanized version b / L chain humanized version b antibody, and H chain humanized version d / L chain humanized version b antibody. It is the graph which compared activity. FIG. 14 shows humans of H chain chimeric / L chain chimeric antibody, H chain humanized version b / L chain humanized version b antibody, and H chain humanized version d / L chain humanized version b antibody. It is the graph which compared the neutralization activity with respect to TF. FIG. 15 compares the antigen binding activities of the H chain chimeric / L chain chimeric antibody, the H chain humanized version e / L chain chimeric antibody, and the H chain humanized version e / L chain humanized version b antibody. It is a graph. FIG. 16 is a graph comparing the neutralizing activity against human TF of the H chain chimeric / L chain chimeric antibody and the H chain humanized version e / L chain chimeric antibody. FIG. 17 is a graph comparing the antigen binding activities of the H chain chimeric / L chain chimeric antibody and the H chain humanized version g / L chain humanized version b antibody. FIG. 18 is a graph comparing the neutralizing activity against human TF of H chain chimeric / L chain chimeric antibody and H chain humanized version g / L chain humanized version b antibody. FIG. 19 shows antigen binding of H chain chimeric / L chain chimeric antibody, H chain humanized version b3 / L chain humanized version b antibody, and H chain humanized version d3 / L chain humanized version b antibody. It is the graph which compared activity. FIG. 20 shows humans of H chain chimeric / L chain chimeric antibody, H chain humanized version b3 / L chain humanized version b antibody, and H chain humanized version d3 / L chain humanized version b antibody. It is the graph which compared the neutralization activity with respect to TF. FIG. 21 is a graph comparing the antigen binding activities of the H chain chimeric / L chain chimeric antibody, the H chain humanized version i / L chain chimeric antibody, and the H chain humanized version j / L chain chimeric antibody. FIG. 22 shows antigen binding of H chain chimeric / L chain chimeric antibody, H chain humanized version i / L chain humanized version b antibody, and H chain humanized version j / L chain humanized version b antibody. It is the graph which compared activity. FIG. 23 compares the neutralizing activity against human TF of the H chain chimeric / L chain chimeric antibody, the H chain humanized version i / L chain chimeric antibody, and the H chain humanized version j / L chain chimeric antibody. It is a graph. FIG. 24 shows an H chain chimeric / L chain chimeric antibody, an H chain humanized version b / L chain humanized version b antibody, an H chain humanized version i / L chain humanized version b antibody, and an H chain It is the graph which compared the neutralization activity with respect to human TF of the humanized version j / L chain humanized version b antibody. FIG. 25 is a graph comparing the antigen binding activities of H chain chimeric / L chain chimeric antibody, H chain chimeric / L chain humanized version b1 antibody, and H chain chimeric / L chain humanized version b2 antibody. FIG. 26 compares the neutralizing activity against human TF of the H chain chimeric / L chain chimeric antibody, the H chain chimeric / L chain humanized version b1 antibody, and the H chain chimeric / L chain humanized version b2 antibody. It is a graph. FIG. 27 is a graph comparing the antigen binding activities of the H chain chimeric / L chain chimeric antibody and the H chain humanized version b / L chain humanized version b2 antibody. FIG. 28 shows an H chain chimeric / L chain chimeric antibody, an H chain humanized version i / L chain humanized version b antibody, an H chain humanized version b / L chain humanized version b antibody, and an H chain. It is the graph which compared the neutralization activity with respect to human TF of the humanized version b / L chain humanized version b2 antibody. FIG. 29 shows antigen binding of H chain chimeric / L chain chimeric antibody, H chain humanized version i / L chain humanized version b1 antibody, and H chain humanized version i / L chain humanized version b2 antibody. It is the graph which compared activity. FIG. 30 shows an H chain chimeric / L chain chimeric antibody, an H chain humanized version i / L chain humanized version b antibody, an H chain humanized version i / L chain humanized version b1 antibody, and an H chain It is the graph which compared the neutralization activity with respect to human TF of the humanized version i / L chain humanized version b2 antibody. FIG. 31 shows an H chain chimeric / L chain chimeric antibody, an H chain humanized version b / L chain humanized version b antibody, an H chain humanized version i / L chain humanized version b antibody, and an H chain It is the graph which compared the antigen binding activity of the humanized version i / L chain humanized version b2 antibody. FIG. 32 shows an H chain chimeric / L chain chimeric antibody, an H chain humanized version b / L chain humanized version b antibody, an H chain humanized version i / L chain humanized version b antibody, and an H chain It is the graph which compared the neutralization activity (factor Xa production inhibitory activity of TF) with respect to human TF of the humanized version i / L chain humanized version b2 antibody. FIG. 33 shows an H chain chimeric / L chain chimeric antibody, an H chain humanized version b / L chain humanized version b antibody, an H chain humanized version i / L chain humanized version b antibody, and an H chain. It is the graph which compared the neutralization activity (factor X binding inhibitory activity) with respect to human TF of the humanized version i / L chain humanized version b2 antibody. FIG. 34 shows an H chain chimeric / L chain chimeric antibody, an H chain humanized version b / L chain humanized version b antibody, an H chain humanized version i / L chain humanized version b antibody, and an H chain It is the graph which compared the neutralization activity (plasma coagulation inhibitory activity of TF) with respect to human TF of the humanized version i / L chain humanized version b2 antibody. FIG. 35 shows H chain chimeric / L chain chimeric antibody, H chain humanized version b / L chain humanized version b antibody, H chain humanized version i / L chain humanized version b antibody, H chain human It is the figure which compared the reactivity to the human TF processed on various conditions of the typed version i / L chain humanized version b2 antibody.

Claims (4)

In a method for producing a natural humanized antibody with reduced immunogenicity having a non-human-derived complementarity determining region (CDR) and a framework region (FR) derived from a natural human antibody,
(1) Prepare a non-human monoclonal antibody reactive to the target antigen,
(2) preparing a plurality of human antibodies having high homology to the FR amino acid sequence in the monoclonal antibody of (1),
(3) Replacing four FRs of one type of human antibody in (2) with corresponding FRs of the non-human monoclonal antibody of (1) to produce a first humanized antibody,
(4) measuring the ability of the humanized antibody prepared in (3) above to bind to the antigen or neutralize the biological activity of the antigen;
(5) 1 to 3 FRs in the humanized antibody prepared in (3) above correspond to human antibodies prepared in (2) that are different from those used in (3). Making a second humanized antibody by substitution with FR,
(6) Ability to neutralize antigen binding or biological activity of the second humanized antibody prepared in (5) and the first humanized antibody obtained in (3) above And select humanized antibodies that exhibit favorable activity,
(7) Steps (3) to (6) above are performed for the humanized antibody selected in (6), and (8) the activity is equivalent to that of the non-human monoclonal antibody in (1). Repeating steps (3) to (6) until a humanized antibody is obtained,
A humanized antibody obtained by the method characterized by the above.   The humanized antibody according to claim 1, wherein the target antigen is human tissue factor (TF).   A humanized antibody in which four FRs contained in the variable region of an antibody are derived from two or more types of human antibodies.   The humanized antibody according to claim 3, wherein no mutation is introduced into the human FR sequence.