1. Patent Search
  2. Details

METHOD FOR EXPRESSING AND PREPARING POLYVALENT MULTI-SPECIFIC ANTIBODY AND IMMUNE HYBRID PROTEIN

20200317819 ยท 2020-10-08

    Inventors

    Cpc classification

    International classification

    Abstract

    The present disclosure relates to the field of biological technologies, and discloses a method for expressing and preparing a polyvalent multi-specific antibody and an immune hybrid protein. By using the characteristics of protein Intein, a novel method for preparing a polyvalent specific antibody, a hybrid immune protein having an antibody fused to a cytokine, an immunotoxin having an antibody fused to a toxin, or an immune hybrid protein having an antibody fused to other active proteins is designed and developed. Each portion of a hybrid protein is respectively expressed in a suitable prokaryotic or eukaryotic cell system, separated and purified by high-performance affinity chromatography, and then spliced in vitro by trans-splicing mediated by the intein, to prepare a polyvalent specific antibody and an immune hybrid protein. The method has high production efficiency, and wide scope of application, and facilitates the separation and purification of the products.

    Claims

    1. A method for expressing and preparing a polyvalent multi-specific antibody, comprising the following steps: S1: splitting an expressed sequence of the polyvalent multi-specific antibody, to obtain several antibody portions including a portion A antibody and a portion B antibody, wherein the portion A antibody comprises a first light chain, a first heavy chain, and an Fc chain of a second heavy chain, in which the Fc chain has Ic fused to the N terminal; the portion B antibody comprises a second light chain and a VH+CH1 chain of the second heavy chain, in which the VH+CH1 chain has In fused to the C terminal; the first light chain and the first heavy chain are a first light chain and a first heavy chain of an antibody that specifically binds to a first antigen; and the second light chain and the second heavy chain are a second light chain and a second heavy chain of an antibody that specifically binds to a second antigen; S2: constructing an eukaryotic or prokaryotic expression vector by whole-gene synthesis, and expressing and preparing the several antibody portions including the portion A antibody and the portion B antibody by transient transfection or stable transfection; and S3: subjecting the portion A antibody and the portion B antibody to protein trans-splicing in vitro, or subjecting the portion A antibody, the portion B antibody, and other antibody portions to protein trans-splicing in vitro, to obtain the polyvalent multi-specific antibody.

    2. The method for expressing and preparing a polyvalent multi-specific antibody according to claim 1, wherein a knob is formed at an interface of the CH3 domain in the first heavy chain, which is capable of being located in an hole formed at an interface of the CH3 domain in the Fc chain of the second heavy chain having Ic fused to the N terminal; or a hole is formed at the interface of the CH3 domain in the first heavy chain, into which a knob formed at the interface of the CH3 domain in the Fc chain of the second heavy chain having Ic fused to the N terminal is capable of being located.

    3. The method for expressing and preparing a polyvalent multi-specific antibody according to claim 1, wherein in Step S2, the expression is expression by a mammalian cell expression system.

    4. The method for expressing and preparing a polyvalent multi-specific antibody according to claim 3, wherein in Step S2, the mammalian cell is 293E, 293F or CHO cells.

    5. The method for expressing and preparing a polyvalent multi-specific antibody according to claim 1, wherein the product expressed in Step S2 is obtained through purification by ProteinL affinity chromatography or by ProteinA/G chromatography; and the polyvalent multi-specific antibody in Step S3 is obtained through purification by ProteinA/G chromatography.

    6. The method for expressing and preparing a polyvalent multi-specific antibody according to claim 1, wherein in Step S3, the in-vitro trans-splicing is in-vitro trans-splicing mediated by a split intein in the presence of a sulfhydryl compound.

    7. The method for expressing and preparing a polyvalent multi-specific antibody according to claim 1, wherein in Step S1, the several antibody portions further comprise a portion C antibody, and the portion C antibody comprises a third single chain of an antibody that specifically binds to a third antigen, where the third single chain has In fused to one end; and the Fc chain of the first heavy chain has Ic fused to the C terminal, or the Fc chain of the second heavy chain has Ic fused to the C terminal.

    8. The method for expressing and preparing a polyvalent multi-specific antibody according to claim 1, wherein in Step S1, the several antibody portions further comprise a portion C antibody and a portion D antibody, where the portion C antibody comprises a third single chain of an antibody that specifically binds to a third antigen and the third single chain has In fused to one end; the portion D antibody comprises a fourth single chain of an antibody that specifically binds to a fourth antigen, and the fourth single chain has In fused to one end; and the Fc chain of the second heavy chain has Ic fused to the C terminal, and the Fc chain of the first heavy chain has Ic fused to the C terminal.

    9. A method for expressing and preparing an immune hybrid protein, comprising the following steps: A1: splitting an expressed sequence of the immune hybrid protein, to obtain a protein molecule and a portion A antibody or the portion A antibody and a portion B antibody, where the portion A antibody comprises a first light chain, a first heavy chain, and an Fc chain of a second heavy chain, in which the Fc chain has Ic fused to the N terminal; the portion B antibody comprises a second single chain having In fused to one end; and the protein molecule has In fused to one end, where the first light chain and the first heavy chain are a first light chain and a first heavy chain of an antibody that specifically binds to a first antigen; and the second heavy chain and the second single chain are a second heavy chain and a second single chain of an antibody that specifically binds to a second antigen; A2: constructing an eukaryotic or prokaryotic expression vector by whole-gene synthesis, and expressing and preparing the portion A antibody or the portion A antibody and the portion B antibody by transient transfection or steady transfection; and A3: subjecting the portion A antibody and the protein molecule to protein/trans-splicing in vitro, or subjecting the portion A antibody and the portion B antibody to protein/trans-splicing in vitro, to obtain the immune hybrid protein.

    10. The method for expressing and preparing an immune hybrid protein according to claim 9, wherein the protein molecule comprises cytokines, toxin polypeptides or active polypeptides.

    11. A method for expressing and preparing an immune hybrid protein, comprising the following steps: B1: splitting an expressed sequence of the immune hybrid protein, to obtain a protein molecule, a portion A antibody, and a portion B antibody, where the portion A antibody comprises a first light chain, a first heavy chain, and an Fc chain of a second heavy chain, in which the Fc chain has Ic fused to the N terminal; the portion B antibody comprises a second light chain and a VH+CH1 chain of the second heavy chain, in which the VH+CH1 chain has In fused to the C terminal; the first light chain and the first heavy chain are a first light chain and a first heavy chain of an antibody that specifically binds to a first antigen; the second light chain and the second heavy chain are a second light chain and a second heavy chain of an antibody that specifically binds to a second antigen; the protein molecule has In fused to one end, and at least one of the Fc chain of the second heavy chain and the Fc chain of the first heavy chain has Ic fused to the C terminal; B2: constructing an eukaryotic or prokaryotic expression vector by whole-gene synthesis, and expressing and preparing the portion A antibody and the portion B antibody by transient transfection or steady transfection; and B3: subjecting the portion A antibody, the portion B antibody, and the protein molecule to protein trans-splicing in vitro, to obtain the immune hybrid protein.

    12. The method for expressing and preparing an immune hybrid protein according to claim 11, wherein the protein molecule comprises cytokines, toxin polypeptides or active polypeptides.

    13. A method for preparing an immunotoxin, comprising the following steps: Step 1: splitting a structural sequence of the target immunotoxin into a structure I and a structure II, where the structure I is an antibody or a fragment thereof; and the structure II is a toxin portion; Step 2: expressing the structure I and the structure II respectively; and Step 3: ligating the structure I and the structure II by protein trans-splicing by a split intein, to obtain the target immunotoxin.

    14. A method for preparing a cytokine-antibody fusion protein, comprising the following steps: Step 1: splitting a structural sequence of the target cytokine-antibody fusion protein into a structure I and a structure II, where the structure I is an antibody or a fragment thereof; and the structure II is a cytokine portion; Step 2: expressing the structure I and the structure II respectively; and Step 3: ligating the structure I and the structure II by protein trans-splicing by a split intein, to obtain the target cytokine-antibody fusion protein.

    15. A method for preparing an ADC antibody, comprising the following steps: Step 1: splitting a structural sequence of the target ADC antibody into a structure I and a structure II, where the structure I is an antibody or a fragment thereof; and the structure II is a compound portion; Step 2: expressing the structure I and the structure II respectively; and Step 3: ligating the structure I and the structure II by protein/trans-splicing by a split intein, to obtain the target ADC antibody.

    Description

    BRIEF DESCRIPTION OF THE DRAWINGS

    [0078] Other features, objects, and advantages of the present disclosure will become apparent upon reading the detailed description of non-limiting embodiments that follow with reference to the accompanying drawings.

    [0079] FIG. 1 is a schematic view showing the protein trans-splicing mediated by a split intein;

    [0080] FIG. 2 is a schematic view showing the splitting of a bispecific antibody, in which A is a schematic view showing the splitting of the bispecific antibody into a portion A antibody having a knob-type heavy chain and a hole-type Fc chain, and a portion B antibody; and B is a schematic view showing the splitting of the bispecific antibody into a portion A antibody having a hole-type heavy chain and a knob-type Fc chain, and a portion B antibody;

    [0081] FIG. 3 is a schematic view showing a hybrid protein having an antibody portion (including hybrid cytokines, toxins, and single chains etc.);

    [0082] FIG. 4 is a flow chart showing the preparation of bispecific antibodies, where A shows the splicing of fragments expressed in two mammalian cells, B shows the splicing of fragments expressed in prokaryotic cells, C shows the splicing of two fragments in which one is expressed in a mammalian cell and the other is expressed in a prokaryotic cell, and D shows the splicing of four fragments which are expressed either in mammalian cells or in prokaryotic cells;

    [0083] FIG. 5A is a schematic view showing the construction of a fragment-type bispecific antibody; B is a schematic view showing the construction of an immune hybrid protein of an antibody fragment; and C is a schematic view showing the construction of an immune hybrid protein of a full-length antibody;

    [0084] FIG. 6 is a schematic view showing a light chain of a portion A antibody;

    [0085] FIG. 7 is a schematic view showing a knob-type heavy chain of the portion A antibody;

    [0086] FIG. 8 is a schematic view showing a hole-type Fc chain of the portion A antibody;

    [0087] FIG. 9 is a schematic view showing a heavy chain of a portion B antibody and IN;

    [0088] FIG. 10 is a schematic view showing a light chain of the portion B antibody;

    [0089] FIG. 11 is a schematic view showing a hole-type heavy chain of a portion A antibody;

    [0090] FIG. 12 is a schematic view showing a knob-type Fc chain of the portion A antibody;

    [0091] FIG. 13 is an SDS-PAGE electrophoretogram of a purified product co-transfected with three expression vectors of a portion A antibody of a bispecific antibody.

    [0092] FIG. 14 is an SDS-PAGE electrophoretogram of a purified product co-transfected with a three-expression vector of a portion B antibody of a bispecific antibody.

    [0093] FIG. 15 is a schematic view showing the splicing (type I) of a portion A antibody and a portion B antibody mediated by a split intein;

    [0094] FIG. 16 is a schematic view showing the splicing (type II) of a portion A antibody and a portion B antibody mediated by a split intein;

    [0095] FIG. 17 is a schematic view showing the trans-splicing into a bispecific antibody induced by a split intein at various DTT concentrations (mM);

    [0096] FIG. 18 is a schematic view showing the trans-splicing into a bispecific antibody induced by a split intein at various temperatures ( C.);

    [0097] FIG. 19 is a schematic view showing the trans-splicing into a bispecific antibody induced by a split intein at various reaction times (min);

    [0098] FIG. 20 is an SDS-PAGE electrophoretogram of a bispecific antibody purified by ProteinA affinity chromatography; and

    [0099] FIG. 21 is a schematic view showing products prepared through methods for preparing a multi-specific antibody and an immune hybrid protein, in which A shows a product group A prepared through a method for preparing a multi-specific antibody and an immune hybrid protein (bispecific antibodies, antibody-directed immunotoxins, and other products); B shows a product group B prepared through a method for preparing a multi-specific antibody and an immune hybrid protein (hybrid proteins of antibody fragments with cytokines or toxins); and C shows a product group C prepared through a method for preparing a multi-specific antibody and an immune hybrid protein (bi- to tetra-specific antibodies, immune hybrid proteins, and other products).

    DETAILED DESCRIPTION

    [0100] Hereinafter, the present disclosure is described in detail by way of examples. The following examples are provided for better understanding of the present disclosure by those skilled in the art, but do not limit the present disclosure in any way. It should be pointed out that for those of ordinary skill in the art, several adjustments and improvements can be made without departing from the concept of the present disclosure, which are all contemplated in the protection scope of the present disclosure.

    [0101] Terms used in the present disclosure are defined below.

    [0102] Antibody refers to an intact monoclonal antibody. The intact antibody consists of two pairs of light chain (LC) and heavy chain (HC) (the light chain/heavy chain pair is abbreviated as LC/HC). The light and heavy chains of the antibody are polypeptides consisting of several domains. In intact antibodies, each heavy chain includes a heavy chain variable region (abbreviated as HCVR or VH) and a heavy chain constant region. The heavy chain constant region includes heavy chain constant domains CH1, CH2 and CH3 (antibody types IgA, IgD, and IgG) and, optionally, heavy chain constant domain CH4 (antibody types IgE and IgM). Each light chain includes a light chain variable domain VL and a light chain constant domain CL. The structure of a naturally occurring intact antibody, i.e., an IgG antibody, is shown, for example, in FIG. 1. The variable domains VH and VL can be further subdivided into hypervariable regions called complementarity determining regions (CDRs), with more conserved regions called framework regions (FR) distributed between them. VH and VL each consist of three CDRs and four FRs, arranged from the amino terminal to the carboxy terminal in the following order: FR1, CDR1, FR2, CDR2, FR3, CDR3, and FR4 (Janeway, C A, Jr. et al., Immunobiology, 5th Edition, Garland Publishing (2001); and Woof J, Burton D Nat Rev Immunol 4 (2004) 89-99). The two pairs of heavy and light chains (HC/LC) are capable of specifically binding to the same antigen. Thus, the intact antibody is a bivalent, monospecific antibody. The antibody includes, for example, a mouse antibody, a human antibody, a chimeric antibody, a humanized antibody, and a genetically engineered antibody (variant or mutant antibody), provided that their specific characteristics are retained. Human or humanized antibodies are particularly preferred, especially recombinant human or humanized antibodies. There are five types of heavy chains in mammalian antibodies, which are represented by Greek letters: , , , , and (Janeway, C A, Jr., et al., Immunobiology, 5th Edition, Garland Publishing (2001)). The types of heavy chains present define the types of antibodies. These chains are present in IgA, IgD, IgE, IgG, and IgM antibodies, respectively (Rhoades R A, Pflanzer R G (2002). Human Physiology, 4th Edition, Tom Thomson Learning). Different heavy chains vary in size and composition. Alpha and gamma types contain approximately 450 amino acids, while and type have approximately 550 amino acids. Each heavy chain has two regions, that is, a constant region and a variable region. The constant regions are identical in all antibodies of the same isotype, but differ in antibodies of different isotypes. The heavy chains , and have a constant region consisting of three constant domains CH1, CH2 and CH3 (on a line) and a hinge region for increasing the flexibility (Woof, J., Burton D Nat Rev Immunol 4 (2004) 89-99). The heavy chains and have a constant region consisting of four constant domains CH1, CH2, CH3 and CH4 (Janeway, C A, Jr., et al., Immunobiology, 5th Edition, Garland Publishing (2001)). The variable regions of the heavy chain vary in antibodies produced by different B cells, but are identical in all antibodies produced by a single type of B cells or B cell clone. The variable region of each heavy chain is approximately 110 amino acids in length and consists of a single antibody domain. In mammals, there are only two types of light chains, called and . The light chain has two consecutive domains: a constant domain CL and a variable domain VL. The approximate length of the light chain is 211-217 amino acids. Preferably, the light chain is a kappa light chain and the constant domain CL is preferably C.

    [0103] The Fc portion of an antibody is a term well known to the skilled artisan and is defined based on the cleavage of the antibody with papain. An antibody according to the present disclosure comprises, for example, an Fc portion, preferably a human derived Fc portion and preferably all other portions of a human constant region. The Fc portion of the antibody is directly involved in complement activation, C1q binding, C3 activation and Fc receptor binding. Although the effect of an antibody on the complement system depends on specific conditions, binding to C1q is caused by specific binding sites in the Fc portion. Such binding sites are known in the art and are described, for example, in Lukas, T J, et al., J. Immunol. 127 (1981) 2555-2560; Brunhouse, R., and Cebra, J J., Mol. Immunol. 16 (1979) 907-917; Burton, D R, et al., Nature 288 (1980) 338-344; Thommesen, J E, et al., Mol. Immunol.) 37 (2000) 995-1004; Idusogie, E E, et al., J. Immunol. 164 (2000) 4178-4184; Hezareh, M., et al., J. Virol. 75 (2001) 12161-12168; and Morgan, A., et al., Immunology 86 (1995) 319-324; and EP 0 307 434. The binding sites are, for example, L234, L235, D270, N297, E318, K320, K322, P331 and P329 (in accordance with Kabat's EU catalog number). Antibodies of subtypes IgG1, IgG2 and IgG3 typically exhibit the capabilities of complement activation, C1q binding and C3 activation, whereas IgG4 does not activate the complement system, does not bind to C1q and does not activate C3.

    [0104] Humanized antibody refers to an antibody in which the frameworks or complementarity determining region (CDRs) have been modified to include CDRs of immunoglobulin that differ in specificity compared to the specificity of the parent immunoglobulin. For example, murine CDRs are grafted into the framework regions of human antibodies to produce humanized antibodies. (Riechmann, L., et al., Nature 332 (1988) 323-327; and Neuberger, M. S., et al., Nature 314 (1985) 268-270).

    [0105] Human antibodies include antibodies having variable and constant regions derived from sequences of human immunoglobulin.

    [0106] Recombinant human antibodies refer to all human antibodies prepared, expressed, produced or isolated by recombination, such as antibodies isolated from host cells, such as NS0 or CHO cells, or antibodies isolated from transgenic animals (e.g., mice) with human immunoglobulin genes, or antibodies expressed by a recombinant expression vector transfected into a host cell. The recombinant human antibodies have a variable region and a constant region in a rearranged pattern.

    [0107] The variable region domain (the light chain (VL) variable region, and the heavy chain (VH) variable region) is each pair of light and heavy chains that are directly involved in the binding of an antibody to an antigen. The human light and heavy chain variable domains have the same general structure and each domain comprises four framework regions (FR), which have a sequence that is generally conserved, and are linked through 3 hypervariable regions (or Complementarity determining regions, CDRs). The framework regions take a beta-sheet conformation and the CDRs can form a loop that joins the beta-sheet structure. The CDRs in each chain maintain their three-dimensional structure through the framework regions and form an antigen binding site with the CDRs from the other chain.

    [0108] Bivalent bispecific antibody refers to an antibody as described above, where each of the two pairs of heavy and light chains (HC/LC) specifically binds to a different antigen, i.e., a first heavy chain and a first light chain (derived from an antibody against an antigen A) specifically bind to the antigen A, and a second heavy chain and a second light chain (derived from an antibody against an antigen B) specifically bind to the antigen B. The bivalent bispecific antibody can simultaneously specifically bind to two and no more than two different antigens, in contrast to a monospecific antibody capable of binding only one antigen on the one hand and a tetravalent tetra-specific antibody capable of simultaneously binding to four antigen molecules on the other hand, for example.

    [0109] Split intein is composed of a N-fragment of intein (In) and a C-fragment of intein (Ic). The gene expressing a precursor protein is split in two open reading frames. The split site is inside the intein sequence. The expression genes of the N-extein (En) and the In of the split intein form a fusion gene, and the fusion protein formed after translation is called an N precursor protein. The expression genes of the Ic of the split intein and the C-extein (Ec) form a fusion gene, and the fusion protein produced after translation is called a C precursor protein. The In or Ic of the split intein alone does not have a protein splicing function. However, after protein translation, the In in the N precursor protein and the Ic in the C precursor protein recognize and bind to each other by a non-covalent bond to form a functional intein, and thus can catalyze the protein trans-splicing to link two separated exteins (En, EC) with a peptide bond (Ozawa. T., Nat Biotechbol 21 (2003) 287-93).

    [0110] Protein trans-splicing refers to a protein splicing reaction mediated by a split intein. In this type of splicing process, the In and Ic of the split intein first recognize and bind to each other by a non-covalent bond (FIG. 1). After binding, the structure is properly folded, and the split intein with a re-constructed active center completes the protein splicing reaction following a typical protein splicing route, to link the extein at both sides (Saleh. L., Chemical Record 6 (2006) 183-193).

    [0111] IN refers to an N-fragment of a split intein alone.

    [0112] IC refers to a C-fragment of a split intein alone.

    [0113] Transient transfection is one of the ways to introduce DNA into eukaryotic cells. In transient transfection, a recombinant DNA is introduced into a cell line with high transfection potential to obtain a transient but high level of expression of the gene of interest. The transfected DNA does not have to be integrated into the chromosome of the host, the transfected cells can be harvested in a shorter time than stable transfection, and the expression of the gene of interest in the cell culture supernatant is detected.

    [0114] The present disclosure relates to methods for expressing and preparing a novel polyvalent multi-specific antibody and an immune hybrid protein. In the present disclosure, a bispecific antibody is split into an antigen A binding portion and an antigen B binding portion for the first time, as shown (in FIGS. 2A and 2B), which are expressed separately, and then ligated into an intact antibody by protein trans-splicing by a split intein. The portion A comprises a light chain of an antibody A, an intact heavy chain of the antibody A, and an Fc chain having Ic fused to the N terminal. The B comprises a light chain of an antibody B and a VH+CH1 chain of the antibody B having In fused to the C terminal In the present disclosure, the trans-splicing function of the split intein is combined with the construction of bispecific antibodies for the first time, and portion A and B antibodies expressed and purified separately are linked to form an intact antibody by means of the trans-splicing function of the split intein. This kind of specific antibodies are similar in structure to naturally occurring antibody molecules, thereby avoiding the instability of antibody molecules due to structural differences and the high immunogenicity in vivo. Firstly, an expressed sequence of the obtained antibody is analyzed and split, a mammalian cell expression vector is constructed by whole gene synthesis, and the purified vector is transiently transfected into mammalian cells such as 293E, 293F, and CHO, etc., or stably transfected into mammal cells such as CHO. The fermentation liquors are separately collected and purified by proteinL affinity chromatography. The purified portions A and B are subjected to trans-splicing in vitro, and the spliced product is purified by proteinA affinity chromatography to obtain a relatively pure bispecific antibody. The process flow is shown in FIG. 4.

    [0115] The method of the present disclosure can also be used to construct a bispecific antibody in which the Fc fragment is defective, for example, only a portion of the CH2 region in the Fc region is retained, or an intact CH2 region and a portion of the CH3 region are retained. In addition, the method is useful in the linkage of any two types of antibody fragments into a novel bispecific antibody. As shown in (FIG. 5), any form of an antibody fragment of a portion C can be trans-spliced with any form of an antibody fragment of a portion D by the split intein.

    [0116] The method for expressing and preparing a hybrid protein of a novel bivalent bispecific antibody provided in the present disclosure comprises the following steps

    [0117] 1. Construction of expression vector

    [0118] For the construction of expression vectors, general information about the nucleotide sequences of light and heavy chains of human immunoglobulin is provided in Kabat, E A, et al., Sequences of Proteins of Immunological Interest, 5th Edition, Public Health Services, National Institutes of Health, Bethesda, Md. (1991) and in the drugbank database. The amino acids in the antibody chain are numbered and referenced according to the EU numbering (Edelman, G. M., et al., Proc. Natl. Acad. Sci. USA 63(1969)78-85; Kabat, E. A., et al., Sequences of Proteins of Immunological Interest, 5th Edition, Public Health Service, National Institutes of Health, Bethesda, Md. (1991)). The desired gene segments are prepared by oligonucleotides prepared through chemical synthesis. The 600-1800 bp long gene segment is assembled by annealing and ligation of PCR-amplified oligonucleotides, and then cloned into an expression vector via the indicated restriction sites such as KpnI/BamHI. The DNA sequence of the subcloned gene fragment is verified by DNA sequencing. Infomax's VectorNTI Advance suite version 8.0 is used for sequence construction, mapping, analysis, annotation, and description.

    [0119] 1.1. In order to solve the problem of mispairing of heavy chains, Knobs-into-Holes is introduced and the VH and CH1 regions of one heavy chain are removed and IC (C-fragment of the split intein) is fused to the N-hinge region of CH2. Thus, the heavy-chain homodimer component formed by the heavy chain that cannot be purified and removed is completely prevented. In order to introduce the Knobs-into-Holes structure, (threonine) at position 366 in a CH3 region is mutated to W (tryptophan) to form a Knobs structure. T (threonine) at position 366 in a CH3 region of another heavy chain is mutated to S (serine), L (leucine) at position 368 is mutated to A (alanine), and Y (tyrosine) at position 407 is mutated to V (valine), to form a Holes structure. In addition, in order to enhance the binding stability of the CH3 regions, S (serine) at position 354 of the Knobs chain is mutated to C (cysteine), and Y (tyrosine) at position 349 on the Holes chain is mutated to C (cysteine) to enhance the stability between heavy chains by introducing a pair of inter-heavy chain disulfide bonds.

    [0120] 1.2. In order to introduce the split intein, the heavy chain of an antibody B is split into an Fc region and a VH+CH1 region in the heavy chain hinge region of the antibody B, IN (N-fragment of the split intein) is fused to the C-terminal of the CH1 region, and IC (C-fragment of the split intein) is fused to the N terminal of CH2.

    [0121] 1.3.a. As shown (in FIG. 6), the sequence of the light chain of the portion A antibody is a light chain sequence of natural antibody A. As shown (in FIG. 7), in the CH3 region of the heavy chain of the portion A antibody, T (threonine) at position 366 is mutated to W (tryptophan) to form a Knobs structure; meanwhile S (serine) at position 354 is mutated to C (cysteine). As shown (in FIG. 8), in the CH3 region of the IC+Fc (Fc having C-fragment of the split intein fused to the N terminal) region of the portion A antibody, T (threonine) at position 366 is mutated to S (serine), L (leucine) at position 368 is mutated to A (alanine), and Y (tyrosine) at position 407 is mutated to V (valine) to form a Holes structure; meanwhile, Y (tyrosine) at position 349 is mutated to C (cysteine). The heavy chain VH+CH1+IN (the heavy chain variable region of the antibody+CH1 region having N-fragment of the split intein fused to the C terminal) of the portion B antibody is as shown (in FIG. 9). As shown (in FIG. 10), the light chain of the portion B antibody is a light chain sequence of natural antibody B.

    [0122] 1.3.b. As shown (in FIG. 6), the sequence of the light chain of the portion A antibody is a light chain sequence of natural antibody A. As shown (in FIG. 11), in the CH3 region of the heavy chain of the portion A antibody, T (threonine) at position 366 is mutated to S (serine), L (leucine) at position 368 is mutated to A (alanine), and Y (tyrosine) at position 407 is mutated to V (valine), to form a Holes structure; meanwhile Y (tyrosine) at position 349 is mutated to C (cysteine). As shown (in FIG. 12), in the CH3 region of the IC+Fc (Fc having C-fragment of the split intein fused to the N terminal) region of the portion A antibody, T (threonine) at position 366 is mutated to W (tryptophan) to form a Knobs structure; meanwhile, S (serine) at position 354 is mutated to C (cysteine). The heavy chain VH+CH1+IN (the heavy chain variable region of the antibody+CH1 region having N-fragment of the split intein fused to the C terminal) of the portion B antibody is as shown (in FIG. 9). As shown (in FIG. 10), the light chain of the portion B antibody is a light chain sequence of natural antibody B.

    [0123] 1.3.c. Construction of expression vectors of small fragment antibodies as shown in FIG. 5, any one of the antibody fragments is selected from the group C, and N-fragment of the split intein is fused at the position IN shown in FIG. 5. Also, any one of the antibody fragments is selected from the group D, and C-fragment of the split intein is fused at the position IC shown in FIG. 5

    [0124] 1.4. The gene sequences designed in the above 1.3 are subjected to whole-gene chemical synthesis. Restriction endonuclease cleavage sites such as KpnI/BamHI are added at the two sides of the start codon and the stop codon by polymerase chain reaction (PCR). The genes are respectively inserted into a mammalian cell expression vector containing CMV promoter, the subclones are sequenced, and the plasmids are extracted. For transient transfection, a larger amount of plasmid is prepared with a plasmid preparation (omega) from transformed E. coli culture. In addition to the antibody expressing region, the vector includes an origin of replication which allows the plasmid to replicate in E. coli and the -lactamase gene, which confers ampicillin resistance in E. coli. The transcription unit of an antibody gene consists of a unique restriction site at the 5 end, an immediate early enhancer and promoter from human cytomegalovirus, followed by an intron A sequence, a 5 untranslated region of a human antibody gene, a signal peptide sequence of an immunoglobulin light chain (or other signal peptide sequence), a 3 untranslated region with a signal sequence A, and a unique restriction site at the 3 end, in the case of cDNA construction.

    [0125] 2. For example, standard cell culture techniques described in Current Protocols in Cell Biology (2000), Bonifacino, J S, Dasso, M., Harford, J B, Lippincott-Schwartz, J. and Yamada, K M (ed.), John Wiley & Sons, Inc can be used. The portions A and B antibodies are expressed by transiently co-transfecting HEK293-E cells grown in suspension or HEK29-F cells grown in suspension with various expression vectors, as described below.

    [0126] 2.1. Transient transfection of HEK293-E system: the portions A and B of a bispecific antibody are produced by co-transfecting HEK293-E cells (human embryonic kidney cell line 293 expressing Epstein-Barr virus nuclear antigen; American Type Culture Center, accession number ATCC#CRL-10852, Lot.959 218) respectively with three expression vectors and two expression vectors. The cells are cultured in SFX4HEK293 medium (HyClone) and Gibco Freestyle 293 medium (Gibco) in a ratio of 1:1 to which 100 g/ml geneticin (Gibco) is added, and the cells are diluted to 1.5-2.510.sup.6 cells/ml with fresh medium one day before transfection and incubated at 37 C. and 120 rpm in 5% CO.sub.2 for transfection on the following day. Taking a 1 L shaking flask (Coming) as an example, the cells are collected by centrifugation at 500-2000 rpm for 5-10 min on the following day, and then washed several times with (10-50 ml) Gibco Freestyle 293 medium. The cells are collected by centrifugation at 500-2000 rpm for 5-10 min, and then resuspended in 150 ml Gibco Freestyle 293 medium to a cell density of 2-610.sup.6 cells/ml in a new 1 L shaking flask (Coming). Plasmids for co-transfection are used in an amount of 0.25-1.5 g DNA per 10.sup.6 cells at equimolar ratio of the vectors of genes encoding various chains, and the DNAs are diluted with Gibco Freestyle 293 medium to (40 ng/L). DNA: PEI (polyscience cationic transfection reagent)=1:2-1:6 are added to the uniformly mixed DNAs and incubated for 5-20 min at room temperature. The cell suspension is added, mixed, and transfected for 4 hours at 37 C. and 120 rpm, in 5% CO.sub.2. Equal volume of pre-warmed SFX4HEK293 medium is added after 4 hours, and then 100 g/ml geneticin (Gibco) is added and incubated at 37 C. and 120 rpm, in 5% CO.sub.2 for 5-10 days. The supernatant is directly collected for purification or the supernatant is collected and stored at 80 C.

    [0127] 2.1.a. PEI-mediated co-transfection of HEK293-E cells with three expression vectors of portion A antibody SFX4HEK293 medium (HyClone) and Gibco Freestyle 293 medium (Gibco) are added in a ratio of 1:1, 100 g/ml geneticin (Gibco) is added, and the cells are diluted to 1.5-2.510.sup.6 cells/ml with fresh medium one day before transfection and incubated at 37 C. and 120 rpm in 5% CO.sub.2 for transfection on the following day. Taking a 1 L shaking flask (Coming) as an example, the cells are collected by centrifugation at 500-2000 rpm for 5-10 min on the following day, and then washed several times with (10-50 ml) Gibco Freestyle 293 medium. The cells are collected by centrifugation at 500-2000 rpm for 5-10 min, and then resuspended in 150 ml Gibco Freestyle 293 medium to a cell density of 2-610.sup.6 cells/ml in a new 1 L shaking flask (Coming). The three expression vectors containing genes encoding the portion A antibody are mixed uniformly at an equimolar ratio in an amount of 0.25-1.5 g DNA per 10.sup.6 cells, and the DNAs are diluted with Gibco Freestyle 293 medium to (40 ng/L). DNA: PEI (polyscience cationic transfection reagent)=1:2-1:6 are added to the uniformly mixed DNAs and incubated for 5-20 min at room temperature. The cell suspension is added, mixed, and transfected for 4 hours at 37 C. and 120 rpm, in 5% CO.sub.2. Equal volume of pre-warmed SFX4HEK293 medium is added after 4 hours, and then 100 g/ml geneticin (Gibco) is added and incubated at 37 C. and 120 rpm, in 5% CO.sub.2 for 5-10 days, to obtain the portion A antibody. The supernatant is directly collected for purification or the supernatant is collected and stored at 80 C.

    [0128] 2.1.b. PEI-mediated co-transfection of HEK293-E cells with two expression vectors of portion B antibody SFX4HEK293 medium (HyClone) and Gibco Freestyle 293 medium (Gibco) are added in a ratio of 1:1, 100 g/ml geneticin (Gibco) is added, and the cells are diluted to 1.5-2.510.sup.6 cells/ml with fresh medium one day before transfection and incubated at 37 C. and 120 rpm in 5% CO.sub.2 for transfection on the following day. Taking a 1 L shaking flask (Coming) as an example, the cells are collected by centrifugation at 1000 rpm for 5 min on the following day, and then washed several times with 50 ml Gibco Freestyle 293 medium. The cells are collected by centrifugation at 1000 rpm for 5 min, and then resuspended in 150 ml Gibco Freestyle 293 medium to a cell density of 2-610.sup.6 cells/ml in a new 1 L shaking flask (Coming). The two expression vectors containing genes encoding the portion A antibody are mixed uniformly at an equimolar ratio in an amount of 0.25-1.5 g DNA per 10.sup.6 cells, and the DNAs are diluted with Gibco Freestyle 293 medium to (40 ng/L). DNA: PEI (polyscience cationic transfection reagent)=1:2-1:6 are added to the uniformly mixed DNAs and incubated for 5-20 min at room temperature. The cell suspension is added, mixed, and transfected for 4 hours at 37 C. and 120 rpm, in 5% CO.sub.2. Equal volume of pre-warmed SFX4HEK293 medium is added after 4 hours, and then 100 g/ml geneticin (Gibco) is added and incubated at 37 C. and 120 rpm, in 5% CO.sub.2 for 5-10 days, to obtain the portion B antibody. The supernatant is directly collected for purification or the supernatant is collected and stored at 80 C.

    [0129] 3. Protein L affinity purification of antibody in fermentation liquor: The protein is purified from the filtered cell culture supernatant following a standard procedure. Briefly, the antibody is loaded to protein L affinity chromatography (GE healthcare) and washed with PBS (containing 20 mM phosphate, 150 mM NaCl pH 6.8-7.4). The impurity components are washed off with 100 mM citrate buffer at pH 5.0, and the antibody is eluted with 100 mM citrate buffer at pH 3.0, and then immediately neutralized with 1 M tris-Hcl buffer at pH 9.0. A portion of the sample is provided for subsequent protein analysis by for example, SDS-PAGE. The monomeric antibody components are pooled for subsequent in-vitro trans splicing mediated by the intein. If necessary, the monomeric antibody components are concentrated using the MILLIPORE Amicon Ultra (30 MWCO) centrifugal concentrator, frozen and stored at 20 C. or 80 C.

    [0130] 3.1. Protein L affinity purification of portion A antibody in fermentation liquor co-transfected with three expression vectors. The protein is purified from the filtered cell culture supernatant following a standard procedure. The supernatant from which the cells are filtered off is mixed with PBS (containing 20 mM phosphate, and 150 mM NaCl pH 6.8-7.4), flow through a Protein L affinity chromatographic column pre-equilibrated with PBS, and washed with PBS after loading. The impurity components are washed off with 100 mM citrate buffer at pH 5.0, and the antibody is eluted with 100 mM citrate buffer at pH 3.0, and then immediately neutralized with 1 M tris-Hcl buffer at pH 9.0. A portion of the sample is provided for subsequent protein analysis by, for example, SDS-PAGE. As shown in (FIG. 13), in the non-reduced sample, assembled portion A antibody of the bispecific antibody appears at around 103 KD. In the reduced sample, the heavy chain of 55 KD, the IC+Fc chain of 40 KD, and the light chain of 25 KD appear. The monomeric antibody components are pooled for subsequent in-vitro trans splicing mediated by the intein. If necessary, the monomeric antibody components are concentrated using the MILLIPORE Amicon Ultra (30 MWCO) ultrafiltration centrifuge tube, frozen and stored at 20 C. or 80 C.

    [0131] 3.2. Protein L affinity purification of portion B antibody in fermentation liquor co-transfected with two expression vectors. The protein is purified from the filtered cell culture supernatant following a standard procedure. The supernatant from which the cells are filtered off is mixed with PBS (containing 20 mM phosphate, and 150 mM NaCl pH 6.8-7.4), run through a Protein L affinity chromatographic column pre-equilibrated with PBS, and washed with PBS after loading. The impurity components are washed off with 100 mM citrate buffer at pH 5.0, and the antibody is eluted with 100 mM citrate buffer at pH 3.0, and then immediately neutralized with 1 M tris-Hcl buffer at pH 9.0. A portion of the sample is provided for subsequent protein analysis by, for example, SDS-PAGE. As shown in (FIG. 14), in the non-reduced sample, assembled portion B antibody of the bispecific antibody appears at around 60 KD. In the reduced sample, the VH+CH1+IN of 35 KD and the light chain of 25 KD appear. The monomeric antibody components are pooled for subsequent in-vitro trans splicing mediated by the intein. If necessary, the monomeric antibody components are concentrated using the MILLIPORE Amicon Ultra (30 MWCO) ultrafiltration centrifuge tube, frozen and stored at 20 C. or 80 C.

    [0132] 4.1. The in-vitro trans-splicing mediated by the split intein of the portions A and B is as shown in (in FIGS. 15 and 16). The portion A and B antibodies purified in Step 3 are mixed at a molar ratio of 1:1, and 0.05 mM to 2 mM DTT or -mercaptoethanol is added. As shown in (FIG. 17), the final concentration of DTT is 0.01 mM, 0.05 mM, 1 mM, 2 mM respectively. The results show that DTT can induce the occurrence of split intein-mediated trans-splicing at a concentration of 0.05 mM, and an obvious band of the bispecific antibody appears at 150 KD. Trans-splicing mediated by the split intein is induced to occur by a sulfhydryl compound such as TCEP. 1 mM DTT or TCEP is added to the splicing reaction system at 4-37 C., and stood respectively at 4, 22, and 37 C. As shown (in FIG. 18), the reaction occurs at 4 C., the reaction efficiency is higher at 22 and 37 C., and an obvious band of the bispecific antibody appears at 150 KD. 1 mM DTT is added to the splicing reaction system, and stood at 37 C. for 5 min, 15 min, 30 min, 60 min, and 120 min, respectively. As shown in (FIG. 19), a bispecific antibody is produced at 5 min, and the reaction reaches a plateau at 60 min. At the end of the reaction, the sulfhydryl compound needed to be removed, and the sulfhydryl compound could be removed by adding an oxidizing agent such as hydrogen peroxide, or removed by dialysis. Further, the sulfhydryl compound might be diluted to below a working concentration by high-fold dilution with a buffer to achieve the purpose of terminating the reaction. The reaction is terminated and a sample is taken for detection by non-reducing SDS-PAGE.

    [0133] 4.2. The in-vitro trans-splicing mediated by the split intein of the portions C and D is as described in 4.1.

    [0134] 5.1. Protein A purification of product obtained after trans-splicing mediated by split intein of portions A and B The protein is purified from the reaction mixture of Step 4 following a standard procedure. The sample is mixed with PBS (containing 20 mM phosphate, and 150 mM NaCl pH 6.8-7.4), run through a Protein A affinity chromatographic column pre-equilibrated with PBS, and washed with PBS after loading. The impurity components are washed off with 100 mM citrate buffer at pH 5.0, and the antibody i eluted with 100 mM citrate buffer at pH 3.0, and then immediately neutralized with 1 M tris-Hcl buffer at pH 9.0. A portion of the sample is provided for subsequent protein analysis by, for example, SDS-PAGE. As shown in (FIG. 20), in the non-reduced sample, an obvious band of a bispecific antibody formed by trans-splicing mediated by the split intein appears at 150 KD and the purity is high. In the reduced sample, only a heavy chain of about 50 KD and a light chain of about 25 KD appear. The monomeric antibody component is pooled. If necessary, the monomeric antibody component is concentrated using the MILLIPORE Amicon Ultra (30 MWCO) ultrafiltration centrifuge tube, frozen and stored at 20 C. or 80 C.; or purified by, for example, ion exchange chromatography, hydrophobic chromatography, and molecular exclusion chromatography, to achieve a higher purity.

    [0135] 5.2. Purification of product formed by trans-splicing of portions C and D mediated by split intein for the product obtained by trans-splicing of the portions C and D, purification by recombinant protein purification methods such as ion exchange chromatography, hydrophobic chromatography and size exclusion chromatography is required.

    [0136] Specific applications are shown in the following examples.

    Example 1: Construction of CD3Her2 Bispecific Antibody

    [0137] 1.1. Construction of Expression Vectors

    [0138] For the construction of expression vectors, general information about the nucleotide sequences of light and heavy chains of human immunoglobulin is provided in Kabat, E A, et al., Sequences of Proteins of Immunological Interest, 5th Edition, Public Health Services, National Institutes of Health, Bethesda, Md. (1991) and in the drugbank database. The amino acids in the antibody chain are numbered and referenced according to the EU numbering (Edelman, G. M., et al., Proc. Natl. Acad. Sci. USA 63(1969)78-85; Kabat, E. A., et al., Sequences of Proteins of Immunological Interest, 5th Edition, Public Health Service, National Institutes of Health, Bethesda, Md. (1991)). The CD3 antibody sequence was derived from humanized OKT3 antibody sequence and the desired gene segments were prepared by oligonucleotides prepared through chemical synthesis. The 600-1800 bp long gene segment was assembled by annealing and ligation of PCR-amplified oligonucleotides, and then cloned into an expression vector via the indicated restriction sites such as KpnI/BamHI. The DNA sequence of the subcloned gene fragment was verified by DNA sequencing. Infomax's VectorNTI Advance suite version 8.0 was used for sequence construction, mapping, analysis, annotation, and description. In order to solve the problem of mispairing of heavy chains, Knobs-into-Holes was introduced, the VH and CH1 regions of one heavy chain are removed, and IC (C-fragment of the split intein) was fused to the N-hinge region of CH2. Thus, the heavy-chain homodimer component formed by the heavy chain that cannot be purified and removed was completely prevented. In order to introduce the Knobs-into-Holes structure, (threonine) at position 366 in a CH3 region of the CD3 antibody was mutated to W (tryptophan) to form a Knobs structure. T (threonine) at position 366 in a CH3 region of the Her2 antibody was mutated to S (serine), L (leucine) at position 368 was mutated to A (alanine), and Y (tyrosine) at position 407 was mutated to V (valine), to form a Holes structure. In addition, in order to enhance the binding stability of the CH3 regions, S (serine) at position 354 of the Knob chain was mutated to C (cysteine), and Y (tyrosine) at position 349 on the Hole chain was mutated to C (cysteine) to enhance the stability between heavy chains by introducing a pair of inter-heavy chain disulfide bonds.

    [0139] 1.1.a. The CD3 antibody was used as the portion A antibody, and the expression vector of each chain was respectively designed on the basis of the following. A light chain of the portion A antibody was designed as shown (in FIG. 6), a knob heavy chain of the portion A antibody was designed as shown (in FIG. 7), and a hole Fc chain of the portion A antibody was designed as shown (in FIG. 8). The Her2 antibody was used as the portion B antibody, and the expression vector of each chain was respectively designed on the basis of the following. A heavy chain IN of the portion B antibody was designed as shown (in FIG. 9) and a light chain of the portion B antibody was designed as shown (in FIG. 10).

    [0140] 1.1.b. The CD3 antibody was used as the portion A antibody, and the expression vector of each chain was respectively designed on the basis of the following. A light chain of the portion A antibody was designed as shown (in FIG. 6), a hole heavy chain of the portion A antibody was designed as shown (in FIG. 11), and a knob Fc chain of the portion A antibody was designed as shown (in FIG. 12). The Her2 antibody was used as the portion B antibody, and the expression vector of each chain was respectively designed on the basis of the following. A heavy chain IN of the portion B antibody was designed as shown (in FIG. 9) and a light chain of the portion B antibody was designed as shown (in FIG. 10).

    [0141] 1.1.c. The Her2 antibody was used as the portion A antibody, and the expression vector of each chain was respectively designed on the basis of the following. A light chain of the portion A antibody was designed as shown (in FIG. 6), a knob heavy chain of the portion A antibody was designed as shown (in FIG. 7), and a hole Fc chain of the portion A antibody was designed as shown (in FIG. 8). The CD3 antibody was used as the portion B antibody, and the expression vector of each chain was respectively designed on the basis of the following. A heavy chain IN of the portion B antibody was designed as shown (in FIG. 9) and a light chain of the portion B antibody was designed as shown (in FIG. 10).

    [0142] 1.1.d. The Her2 antibody was used as the portion A antibody, and the expression vector of each chain was respectively designed on the basis of the following. A light chain of the portion A antibody was designed as shown (in FIG. 6), a hole heavy chain of the portion A antibody was designed as shown (in FIG. 11), and a knob Fc chain of the portion A antibody was designed as shown (in FIG. 12). The CD3 antibody was used as the portion B antibody, and the expression vector of each chain was respectively designed on the basis of the following. A heavy chain IN of the portion B antibody was designed as shown (in FIG. 9) and a light chain of the portion B antibody was designed as shown (in FIG. 10).

    [0143] 1.2. Expression of Transiently Transfected HEK-293E Cells

    [0144] Transient transfection of HEK293-E system. The portions A and B of a bispecific antibody were produced by co-transfecting HEK293-E cells (human embryonic kidney cell line 293 expressing Epstein-Barr virus nuclear antigen; American Type Culture Center, accession number ATCC#CRL-10852, Lot.959 218) respectively with three expression vectors and two expression vectors. The cells were cultured in SFX4HEK293 medium (HyClone) and Gibco Freestyle 293 medium (Gibco) in a ratio of 1:1 to which 100 g/ml geneticin (Gibco) was added, and the cells were diluted to 1.5-2.510.sup.6 cells/ml with fresh medium one day before transfection and incubated at 37 C. and 120 rpm in 5% CO.sub.2 for transfection on the following day. Taking a 1 L shaking flask (Coming) as an example, the cells were collected by centrifugation at 1000 rpm for 5 min on the following day, and then washed once with (50 ml) Gibco Freestyle 293 medium. The cells were collected by centrifugation at 1000 rpm for 5 min, and then resuspended in 150 ml Gibco Freestyle 293 medium to a cell density of 410.sup.6 cells/ml in a new 1 L shaking flask (Coming). Plasmids for co-transfection were used in an amount of 0.5 g DNA per 10.sup.6 cells at equimolar ratio of the vectors of genes encoding various chains, and the DNAs were diluted with Gibco Freestyle 293 medium to (40 ng/L). DNA: PEI (polyscience cationic transfection reagent)=1:3 were added to the uniformly mixed DNAs and incubated for 20 min at room temperature. The cell suspension was added, mixed, and transfected for 4 hours at 37 C. and 110 rpm, in 5% CO.sub.2. Equal volume of pre-warmed SFX4HEK293 medium was added after 4 hours, and then 100 g/ml geneticin (Gibco) was added and incubated at 37 C. and 130 rpm, in 5% CO.sub.2 for 10 days. The supernatant was directly collected for purification or the supernatant was collected and stored at 80 C.

    [0145] 1.2.a. PEI-mediated co-transfection of HEK293-E cells with three expression vectors of portion A antibody constructed in 1.1.a. The cells were cultured in SFX4HEK293 medium (HyClone) and Gibco Freestyle 293 medium (Gibco) in a ratio of 1:1 to which 100 g/ml geneticin (Gibco) was added, and the cells were diluted to 1.5-2.510.sup.6 cells/ml with fresh medium one day before transfection and incubated at 37 C. and 120 rpm in 5% CO.sub.2 for transfection on the following day. Taking a 1 L shaking flask (Coming) as an example, the cells were collected by centrifugation at 1000 rpm for 5 min on the following day, and then washed once with (50 ml) Gibco Freestyle 293 medium. The cells were collected by centrifugation at 1000 rpm for 5 min, and then resuspended in 150 ml Gibco Freestyle 293 medium to a cell density of 410.sup.6 cells/ml in a new 1 L shaking flask (Coming). Plasmids for co-transfection were used in an amount of 0.5 g DNA per 10.sup.6 cells at equimolar ratio of the vectors of genes encoding various chains, and the DNAs were diluted with Gibco Freestyle 293 medium to (40 ng/L). DNA: PEI (polyscience cationic transfection reagent)=1:3 were added to the uniformly mixed DNAs and incubated for 20 min at room temperature. The cell suspension was added, mixed, and transfected for 4 hours at 37 C. and 110 rpm, in 5% CO.sub.2. Equal volume of pre-warmed SFX4HEK293 medium was added after 4 hours, and then 100 g/ml geneticin (Gibco) was added and incubated at 37 C. and 130 rpm, in 5% CO.sub.2 for 10 days. The supernatant was directly collected for purification or the supernatant was collected and stored at 80 C.

    [0146] 1.2.b. PEI-mediated co-transfection of HEK293-E cells with three expression vectors of portion A antibody constructed in 1.1.b. The cells were cultured in SFX4HEK293 medium (HyClone) and Gibco Freestyle 293 medium (Gibco) in a ratio of 1:1 to which 100 g/ml geneticin (Gibco) was added, and the cells were diluted to 1.5-2.510.sup.6 cells/ml with fresh medium one day before transfection and incubated at 37 C. and 120 rpm in 5% CO.sub.2 for transfection on the following day. Taking a 1 L shaking flask (Coming) as an example, the cells were collected by centrifugation at 1000 rpm for 5 min on the following day, and then washed once with (50 ml) Gibco Freestyle 293 medium. The cells were collected by centrifugation at 1000 rpm for 5 min, and then resuspended in 150 ml Gibco Freestyle 293 medium to a cell density of 410.sup.6 cells/ml in a new 1 L shaking flask (Coming). Plasmids for co-transfection were used in an amount of 0.5 g DNA per 10.sup.6 cells at equimolar ratio of the vectors of genes encoding various chains, and the DNAs were diluted with Gibco Freestyle 293 medium to (40 ng/L). DNA: PEI (polyscience cationic transfection reagent)=1:3 were added to the uniformly mixed DNAs and incubated for 20 min at room temperature. The cell suspension was added, mixed, and transfected for 4 hours at 37 C. and 110 rpm, in 5% CO.sub.2. Equal volume of pre-warmed SFX4HEK293 medium was added after 4 hours, and then 100 g/ml geneticin (Gibco) was added and incubated at 37 C. and 130 rpm, in 5% CO.sub.2 for 10 days. The supernatant was directly collected for purification or the supernatant was collected and stored at 80 C.

    [0147] 1.2.c. PEI-mediated co-transfection of HEK293-E cells with two expression vectors of portion B antibody constructed in 1.1.c. The cells were cultured in SFX4HEK293 medium (HyClone) and Gibco Freestyle 293 medium (Gibco) in a ratio of 1:1 to which 100 g/ml geneticin (Gibco) was added, and the cells were diluted to 1.5-2.510.sup.6 cells/ml with fresh medium one day before transfection and incubated at 37 C. and 120 rpm in 5% CO.sub.2 for transfection on the following day. Taking a 1 L shaking flask (Coming) as an example, the cells were collected by centrifugation at 1000 rpm for 5 min on the following day, and then washed once with (50 ml) Gibco Freestyle 293 medium. The cells were collected by centrifugation at 1000 rpm for 5 min, and then resuspended in 150 ml Gibco Freestyle 293 medium to a cell density of 410.sup.6 cells/ml in a new 1 L shaking flask (Coming). Plasmids for co-transfection were used in an amount of 0.5 g DNA per 10.sup.6 cells at equimolar ratio of the vectors of genes encoding various chains, and the DNAs were diluted with Gibco Freestyle 293 medium to (40 ng/L). DNA: PEI (polyscience cationic transfection reagent)=1:3 were added to the uniformly mixed DNAs and incubated for 20 min at room temperature. The cell suspension was added, mixed, and transfected for 4 hours at 37 C. and 110 rpm, in 5% CO.sub.2. Equal volume of pre-warmed SFX4HEK293 medium was added after 4 hours, and then 100 g/ml geneticin (Gibco) was added and incubated at 37 C. and 130 rpm, in 5% CO.sub.2 for 10 days. The supernatant was directly collected for purification or the supernatant was collected and stored at 80 C.

    [0148] 1.2.d. PEI-mediated co-transfection of HEK293-E cells with two expression vectors of portion B antibody constructed in 1.1.d. The cells were cultured in SFX4HEK293 medium (HyClone) and Gibco Freestyle 293 medium (Gibco) in a ratio of 1:1 to which 100 g/ml geneticin (Gibco) was added, and the cells were diluted to 1.5-2.510.sup.6 cells/ml with fresh medium one day before transfection and incubated at 37 C. and 120 rpm in 5% CO.sub.2 for transfection on the following day. Taking a 1 L shaking flask (Coming) as an example, the cells were collected by centrifugation at 1000 rpm for 5 min on the following day, and then washed once with (50 ml) Gibco Freestyle 293 medium. The cells were collected by centrifugation at 1000 rpm for 5 min, and then resuspended in 150 ml Gibco Freestyle 293 medium to a cell density of 410.sup.6 cells/ml in a new 1 L shaking flask (Coming). Plasmids for co-transfection were used in an amount of 0.5 g DNA per 10.sup.6 cells at equimolar ratio of the vectors of genes encoding various chains, and the DNAs were diluted with Gibco Freestyle 293 medium to (40 ng/L). DNA: PEI (polyscience cationic transfection reagent)=1:3 were added to the uniformly mixed DNAs and incubated for 20 min at room temperature. The cell suspension was added, mixed, and transfected for 4 hours at 37 C. and 110 rpm, in 5% CO.sub.2. Equal volume of pre-warmed SFX4HEK293 medium was added after 4 hours, and then 100 g/ml geneticin (Gibco) was added and incubated at 37 C. and 130 rpm, in 5% CO.sub.2 for 10 days. The supernatant was directly collected for purification or the supernatant was collected and stored at 80 C.

    [0149] 1.3. Protein L Affinity Purification of Antibody in Fermentation Liquor

    [0150] The protein was purified from the filtered cell culture supernatant following a standard procedure. Briefly, the antibody was subjected to protein L affinity chromatography (GE healthcare) and washed with PBS (containing 20 mM phosphate, 150 mM NaCl pH 6.8-7.4). The impurity components were washed off with 100 mM citrate buffer at pH 5.0, and the antibody was eluted with 100 mM citrate buffer at pH 3.0, and then immediately neutralized with 1 M tris-Hcl buffer at pH 9.0. A portion of the sample was provided for subsequent protein analysis by for example, SDS-PAGE. The monomeric antibody components were pooled for subsequent in-vitro trans splicing mediated by the intein. If necessary, the monomeric antibody components were concentrated using the MILLIPORE Amicon Ultra (30 MWCO) centrifugal concentrator, frozen and stored at 20 C. or 80 C.

    [0151] 1.3.a. Protein L affinity purification of antibody in cell fermentation liquor obtained in Step 1.2.a. Protein L affinity purification of portion A antibody in fermentation liquor co-transfected with three expression vectorsThe protein was purified from the filtered cell culture supernatant following a standard procedure. The supernatant from which the cells were filtered off was mixed with PBS (containing 20 mM phosphate, and 150 mM NaCl pH 6.8-7.4), run through a Protein L affinity chromatographic column pre-equilibrated with PBS, and washed with PBS after loading. The impurity components were washed off with 100 mM citrate buffer at pH 5.0, and the antibody was eluted with 100 mM citrate buffer at pH 3.0, and then immediately neutralized with 1 M tris-Hcl buffer at pH 9.0. A portion of the sample was provided for subsequent protein analysis by, for example, SDS-PAGE. As shown in (FIG. 13), in the non-reduced sample, assembled portion A antibody of the bispecific antibody appears at around 103 KD. In the reducing electrophoresis, the heavy chain of 55 KD, the IC+Fc chain of 40 KD, and the light chain of 25 KD appear. The monomeric antibody component was pooled, which might be purified to obtain a purified product comprising mainly portion A antibody for subsequent in-vitro trans-splicing mediated by the intein. If necessary, the monomeric antibody components were concentrated using the MILLIPORE Amicon Ultra (30 MWCO) ultrafiltration centrifuge tube, frozen and stored at 20 C. or 80 C.

    [0152] 1.3.b. Protein L affinity purification of antibody in cell fermentation liquor obtained in Step 1.2.b. Protein L affinity purification of portion A antibody in fermentation liquor co-transfected with three expression vectorsThe protein was purified from the filtered cell culture supernatant following a standard procedure. The supernatant from which the cells were filtered off was mixed with PBS (containing 20 mM phosphate, and 150 mM NaCl pH 6.8-7.4), run through a Protein L affinity chromatographic column pre-equilibrated with PBS, and washed with PBS after loading. The impurity components were washed off with 100 mM citrate buffer at pH 5.0, and the antibody was eluted with 100 mM citrate buffer at pH 3.0, and then immediately neutralized with 1 M tris-Hcl buffer at pH 9.0. A portion of the sample was provided for subsequent protein analysis by, for example, SDS-PAGE. As shown in (FIG. 13), in the non-reduced sample, assembled portion A antibody of the bispecific antibody appears at around 103 KD. In the reducing electrophoresis, the heavy chain of 55 KD, the IC+Fc chain of 40 KD, and the light chain of 25 KD appear. The monomeric antibody component was pooled, which might be purified to obtain a purified product comprising mainly portion A antibody for subsequent in-vitro trans-splicing mediated by the intein. If necessary, the monomeric antibody components were concentrated using the MILLIPORE Amicon Ultra (30 MWCO) ultrafiltration centrifuge tube, frozen and stored at 20 C. or 80 C.

    [0153] 1.3.c. Protein L affinity purification of antibody in cell fermentation liquor obtained in Step 1.2.c. The protein was purified from the filtered cell culture supernatant following a standard procedure. The supernatant from which the cells were filtered off was mixed with PBS (containing 20 mM phosphate, and 150 mM NaCl pH 6.8-7.4), run through a Protein L affinity chromatographic column pre-equilibrated with PBS, and washed with PBS after loading. The impurity components were washed off with 100 mM citrate buffer at pH 5.0, and the antibody was eluted with 100 mM citrate buffer at pH 3.0, and then immediately neutralized with 1 M tris-Hcl buffer at pH 9.0. A portion of the sample was provided for subsequent protein analysis by, for example, SDS-PAGE. As shown in (FIG. 14), in the non-reduced sample, assembled portion B antibody of the bispecific antibody appears at around 60 KD. In the reduced sample, the VH+CH1+IN of 35 KD and the light chain of 25 KD appear. The monomeric antibody component was pooled, which might be purified to obtain a purified product comprising mainly portion B antibody for subsequent in-vitro trans-splicing mediated by the intein. If necessary, the monomeric antibody components were concentrated using the MILLIPORE Amicon Ultra (30 MWCO) ultrafiltration centrifuge tube, frozen and stored at 20 C. or 80 C.

    [0154] 1.3.d. Protein L affinity purification of antibody in cell fermentation liquor obtained in Step 1.2.d. The protein was purified from the filtered cell culture supernatant following a standard procedure. The supernatant from which the cells were filtered off was mixed with PBS (containing 20 mM phosphate, and 150 mM NaCl pH 6.8-7.4), run through a Protein L affinity chromatographic column pre-equilibrated with PBS, and washed with PBS after loading. The impurity components were washed off with 100 mM citrate buffer at pH 5.0, and the antibody was eluted with 100 mM citrate buffer at pH 3.0, and then immediately neutralized with 1 M tris-Hcl buffer at pH 9.0. A portion of the sample was provided for subsequent protein analysis by, for example, SDS-PAGE. As shown in (FIG. 14), in the non-reduced sample, assembled portion B antibody of the bispecific antibody appears at around 60 KD. In the reduced sample, the VH+CH1+IN of 35 KD and the light chain of 25 KD appear. The monomeric antibody component was pooled, which might be purified to obtain a purified product comprising mainly portion B antibody for subsequent in-vitro trans-splicing mediated by the intein. If necessary, the monomeric antibody components were concentrated using the MILLIPORE Amicon Ultra (30 MWCO) ultrafiltration centrifuge tube, frozen and stored at 20 C. or 80 C.

    [0155] 1.4. In-Vitro Trans-Splicing Mediated by the Split Intein of the Portions A and B

    [0156] As shown (in FIGS. 15 and 16), the portion A and B antibodies purified in Step 1.3 were mixed at a molar ratio of 1:1, and 0.05 mM to 2 mM DTT or -mercaptoethanol was added. As shown in (FIG. 17), the final concentration of DTT is 0.01 mM, 0.05 mM, 1 mM, and 2 mM respectively. The results show that DTT can induce the occurrence of split intein-mediated trans-splicing at a concentration of 0.05 mM, and an obvious band of the bispecific antibody appears at 150 KD. Trans-splicing mediated by the split intein was induced to occur by a sulfhydryl compound such as TCEP. 1 mM DTT or TCEP was added to the splicing reaction system at 4-37 C., and stood respectively at 4, 22, and 37 C. As shown (in FIG. 18), the reaction occurs at 4 C., the reaction efficiency is higher at 22 and 37 C., and an obvious band of the bispecific antibody appears at 150 KD. 1 mM DTT was added to the splicing reaction system, and stood at 37 C. for 5 min, 15 min, 30 min, 60 min, and 120 min, respectively. As shown in (FIG. 19), a bispecific antibody is produced at 5 min, and the reaction reaches a plateau at 60 min. At the end of the reaction, the sulfhydryl compound needed to be removed, and the sulfhydryl compound could be removed by adding an oxidizing agent such as hydrogen peroxide, or removed by dialysis. Further, the sulfhydryl compound might be diluted to below a working concentration by high-fold dilution with a buffer to achieve the purpose of terminating the reaction. The reaction was terminated and a sample was taken for detection by non-reducing SDS-PAGE.

    [0157] 1.5. Protein a Purification of Product Obtained after Trans-Splicing Mediated by Split Intein of Portions A and B

    [0158] The protein was purified from the reaction mixture of Step 4 following a standard procedure. The sample was mixed with PBS (containing 20 mM phosphate, and 150 mM NaCl pH 6.8-7.4), run through a Protein A affinity chromatographic column pre-equilibrated with PBS, and washed with PBS after loading. The impurity components were washed off with 100 mM citrate buffer at pH 5.0, and the antibody was eluted with 100 mM citrate buffer at pH 3.0, and then immediately neutralized with 1 M tris-Hcl buffer at pH 9.0. A portion of the sample was provided for subsequent protein analysis by, for example, SDS-PAGE, as shown in (FIG. 20). FIG. 20 shows Coomassie blue staining of the product eluate from rProteinA chromatography in SDS-PAGE, in which M. marker; 1. before loading (N); 2. eluate from Ni column (N); 3. eluate 1 from rProteinA chromatography (N); 4. eluate 2 from rProteinA chromatography (N); 5. eluate 3 from rProteinA chromatography; 6. empty; 7. before loading (R); 8. eluate from Ni column (R); 9. eluate 1 from ProteinA chromatography (R) 10. eluate 2 from rProteinA chromatography (R), where N denote Nonreducing, and R denote Reducing. It can be known from FIG. 20, in the non-reduced sample, an obvious band of a bispecific antibody formed by trans-splicing mediated by the split intein appears at 150 KD and the purity is high. In the reduced sample, only a heavy chain of about 50 KD and a light chain of about 25 KD appear. The monomeric antibody component was pooled. If necessary, the monomeric antibody component was concentrated using the MILLIPORE Amicon Ultra (30 MWCO) ultrafiltration centrifuge tube, frozen and stored at 20 C. or 80 C.; or purified by, for example, ion exchange chromatography, hydrophobic chromatography, and molecular exclusion chromatography, to achieve a higher purity. FIG. 21 is a schematic view showing products prepared through methods for preparing a multi-specific antibody and an immune hybrid protein, in which A shows a product group A prepared through a method for preparing a multi-specific antibody and an immune hybrid protein (bispecific antibodies, antibody-directed immunotoxins, and other products); B shows a product group B prepared through a method for preparing a multi-specific antibody and an immune hybrid protein (hybrid proteins of antibody fragments with cytokines or toxins); and C shows a product group C prepared through a method for preparing a multi-specific antibody and an immune hybrid protein (bi- to tetra-specific antibodies, immune hybrid proteins, and other products).

    [0159] In summary, in the present disclosure, in order to solve the problem of mispairing of heavy chains, Knobs-into-Holes is introduced, the VH and CH1 regions of one heavy chain are removed, and IC (C-fragment of the split intein) is fused to the N-hinge region of CH2. Thus, the heavy-chain homodimer component formed by the heavy chain that cannot be purified and removed is completely prevented. In order to introduce the Knobs-into-Holes structure, (threonine) at position 366 in a CH3 region is mutated to W (tryptophan) to form a Knob structure. T (threonine) at position 366 in a CH3 region of another heavy chain is mutated to S (serine), L (leucine) at position 368 is mutated to A (alanine), and Y (tyrosine) at position 407 is mutated to V (valine), to form a Hole structure. In addition, in order to enhance the binding stability of the CH3 regions, S (serine) at position 354 of the Knob chain is mutated to C (cysteine), and Y (tyrosine) at position 349 on the Hole chain is mutated to C (cysteine) to enhance the stability between heavy chains by introducing a pair of inter-heavy chain disulfide bonds. Also, more importantly, an intact Knob heavy chain and Hole Fc chain are co-expressed. Due to the high difference in the properties of the Knob heavy chain homodimer and the Hole Fc homodimer from the target product, a heterodimer of the Knob heavy chain and the Hole Fc, separation and purification can be carried out simply. Therefore, the problem of mispairing of heavy chains can be completely avoided in the final product.

    [0160] In the present disclosure, a bispecific antibody is split into an antigen A binding portion and an antigen B binding portion for the first time, as shown (in FIGS. 2 and 3), which are expressed separately, and then ligated into an intact antibody by protein trans-splicing by a split intein. The two light chains do not exist at the same time, and the two VH+CH1 chains do not exist at the same time, so there is no case where the light chain of A binds to the heavy chain of B, either the case where the light chain of B binds to A. Therefore, the situation of mispairing of light chains is avoided completely.

    [0161] In the present disclosure, the trans-splicing function of the split intein is combined with the construction of bispecific antibodies for the first time, and portion A and B antibodies expressed and purified separately are linked to form an intact antibody by means of the am-splicing function of the split intein. This kind of specific antibodies is similar in structure to naturally occurring antibody molecules, thereby avoiding the instability of antibody molecules due to structural differences and the high immunogenicity in vivo.

    [0162] In the present disclosure, recombinant gene expression technology is used to produce a bispecific antibody, and the sequence used may be a humanized antibody sequence or a fully human antibody sequence, to finally obtain a humanized or fully human bispecific antibody. This will greatly reduce the immunogenicity of the bispecific antibody in vivo, laying a foundation for the use of the bispecific as a drug.

    [0163] Since the portion A antibody retains the entire Fc region, the bispecific antibody obtained by trans-splicing mediated by the intein also retains the entire Fc region, thus retaining the effector functions of the antibody, such as complement-dependent cytotoxicity (CDC) or antibody dependent cellular cytotoxicity (ADCC) and extended half-life of binding to FcRn (Fc receptor) in vascular endothelium.

    [0164] In the present disclosure, both portion A and B antibodies are expressed in a mammalian cell expression system, for example, by transiently transfecting 293E, 293F, or CHO cells, and stably transfecting CHO cells. The products expressed by mammalian cells are glycosylated and more similar in structure to natural antibody molecules. The bispecific antibodies obtained by intein-mediated trans-splicing are well glycosylated and have well maintained stability of the bispecific antibody molecules, and antibody effects such as ADCC, CDC, etc., the in vivo half-life is prolonged, and the duration of the effect of action of the drug is increased.

    [0165] In the method for preparing a bispecific antibody according to the present disclosure, the purification process is easy and convenient in operation. First, both portions A and B can be recovered by chromatography with a high recovery rate, such as ProteinL or ProteinA/G affinity chromatography. The bispecific antibody obtained by intein-mediated trans-splicing can be recovered by chromatography with a high recovery rate, such as ProteinA/G affinity chromatography, which can facilitate the subsequent hydrophobic chromatography, or ion exchange chromatography, and other operations. Therefore, the difficulty of purification is greatly reduced and a high-quality product can be obtained.

    [0166] Since the hinge region between CH1 and CH2 of the antibody heavy chain is flexible and the primary Fab region in the Fc region of the antibody is substantially identical in structure, the method is applicable to the production of any bispecific antibodies with no need to perform property analysis based on the nature of each antibody. The present disclosure is fully applicable to the production of bispecific antibodies of any of the antibody subtypes (IgG, IgA, IgM, IgD, IgE, IgM, and light chain kappa and lambda), thus having broad versatility.

    Example 2: Preparation of Immunotoxin Herceptin-PE38KDEL by Taking Advantage of the Trans-Splicing Function of Npu DnaE

    [0167] 2.1. Construction of her HC Expressing Vector pCEP4-her HC-Nn Having N-Fragment of Split Intein Npu DnaE Fused

    [0168] Using the primers shown in Table 1, and using the synthesized heavy-chain nucleic acid molecule of Herceptin comprising a coding signal peptide gene as a template, the gene encoding the heavy chain of Herceptin was cloned. Using the synthesized nucleic acid molecule comprising Npu DnaE as a template, the gene encoding the N-fragment of Npu DnaE was cloned. The multiplication was carried out using the PrimerStar Max available from TaKaRa, under PCR conditions including 30 cycles of 10 s at 94 C., 10 s at 55 C., and 10 s at 72 C. The obtained fragments were recovered by agarose gel electrophoresis, and the gene encoding the heavy chain of Herceptin and the gene encoding the N-fragment of Npu DnaE were sequentially synthesized by overlap PCR, in which the N-fragment of Npu DnaE was at the C terminal of the fusion peptide. PCR conditions included 30 cycles of 10 s at 94 C., 10 s at 55 C., and 10 s at 72 C. The gene fragment was treated with HindIII and BamHI, and ligated to pCEP4 that was also treated with HindIII and BamHI. The plasmid structure was as shown in the figure. The ligated product was transformed into E. coli DH5a competent cells, and the transformed cells were plated on an agar plate containing 50 g/mL ampicillin overnight. The monoclonal clones grown on the plate were picked and cultured in 5 mL of LB medium containing 50 g/mL ampicillin with agitation overnight, and the plasmid was extracted and sequenced. The sequencing results indicated that the constructed Her HC-Nn sequence was correct.

    TABLE-US-00001 TABLE1 Enzymatic cleavage Primername Primersequence site Sig-Her-For 5-AAAAAGCTTATGGGT HindIII (SEQIDNO.1) TGGAGCTGCATCATCTTG TTCCTCG-3 Her-HC-OL-Nn-Rev 5-GCCACCACCTGATCC (SEQIDNO.2) GCCTCCGCCTGACTTCCC GGGGCTCAGG-3 Nn-OL-HerHC-For 5-GGCGGATCTGGTGGT (SEQIDNO.3) GGCGGATCTTATTGTTTA AGCTATGAAACGGAAATA TTG-3 Nn-Rev-BamHI 5-AAAGGATCCTCAGTG BamHI (SEQIDNO.4) GTGGTGGTGGTGGTGGTG GTGGTGGTGATTCGGCAA ATTATCAACCCG-3

    [0169] 2.2. Expression and Purification of Fusion Polypeptide Herceptin-Nn

    [0170] The fusion peptide Herceptin-Nn was expressed using a transient expression system in the HEK293-E system. HEK293-E cells (human embryonic kidney cell line 293 expressing Epstein-Barr virus nuclear antigen; American Type Culture Center, accession number ATCC#CRL-10852, Lot.959 218) were cultured in SFX4HEK293 medium (HyClone) and Gibco Freestyle 293 medium (Gibco) in a ratio of 1:1 to which 100 g/ml geneticin (Gibco) was added. The cells were diluted to 1.5-2.5106 cells/ml with fresh medium one day before transfection and incubated at 37 C. and 120 rpm in 5% CO.sub.2 for transfection on the following day. On the day of transfection, pCEP4-Her HC-Nn was mixed with the constructed Herceptin light chain expression vector pCEP4-Her LC at an equal weight ratio in an amount of 0.5 g DNA per 10.sup.6 cells. The DNAs were diluted with Gibco Freestyle 293 medium to (40 ng/L), and DNA: PEI=1:3 were added to the uniformly mixed DNAs and incubated for 20 min at room temperature for later use. The cells were collected by centrifugation at 1000 rpm for 5 min, washed once with Gibco Freestyle 293 medium, collected by centrifugation at 1000 rpm for 5 min, and then resuspended at a cell density of 410.sup.6 cells/ml in 150 ml Gibco Freestyle 293 medium in a new 1 L shaking flask (Coming). The incubated DNA-PEI complex was added to the cells, and transfected at 37 C. and 110 rpm in 5% CO.sub.2 for 4 hours. Subsequently, an equal volume of pre-warmed SFX4 HEK293 medium, and 100 g/ml geneticin (Gibco) were added, and the incubation was continued for 10 days at 37 C. and 130 rpm, in 5% CO.sub.2. The supernatant was directly collected for purification or the supernatant was collected and stored at 80 C.

    [0171] The collected supernatant was 1:1 mixed with PBS (20 mM PBS, 150 mM NaCl, pH 6.8-7.4), and loaded onto a Protein A affinity chromatographic column equilibrated with PBS in advance. After loading, the column was washed with 10 times volume of PBS, the antibody was eluted with 100 mM citrate buffer (pH 3.0), and the collected eluted sample was immediately neutralized with 1 M Tris-HCl buffer (pH 9.0). A small sample was taken for SDS-PAGG analysis. The non-reduced sample showed assembled Herceptin-Nn at around 170 kD, and a Her HC-Nn chain of about 70 kD and a light chain of 25 KD appeared in the reduced sample. Samples containing the protein of interest were pooled for subsequent split intein-mediated in-vitro splicing If necessary, the monomeric antibody components were concentrated using the MILLIPORE Amicon Ultra (30 MWCO) ultrafiltration centrifuge tube, frozen and stored at 20 C. or 80 C.

    [0172] 3. Preparation of Herceptin-PE38KDEL by Taking Advantage of the Trans-Splicing Function of Npu DnaE

    [0173] The obtained fusion polypeptide Nc-PE38KDEL and the fusion polypeptide Herceptin-Nn were mixed at a molar ratio of 1:1, and DTT was added at a final concentration of 1 mM, and incubated at 25 C. for 60 min. Samples were taken for analysis by SDS-PAGE and Western Blot.

    [0174] 4. Separation and Purification of Herceptin-PE38KDEL Produced by Npu DnaE-Mediated Trans-Splicing

    [0175] Herceptin-PE38KDEL was purified by using Ni.sup.2+ NTA to capture the fusion polypeptide in Step 3 that was not amenable to trans-splicing reaction and the by-product produced by trans-splicing. A Ni.sup.2+ gravity column was packed, the volume of the column was 1 ml, washed with 5 column volumes of water, and then equilibrated with 10 column volumes of a binding buffer (20 mM PBS, 500 mM NaCl, 20 mM imidazole, pH 7.5). The reaction system obtained in Step 3 was loaded onto the column. The flow rate was 1 ml/min, and the flow-through was collected. After loading, the column was washed with 5 column volumes of an elution buffer containing 40 mM imidazole (20 mM PBS, 500 mM NaCl, 40 mM imidazole, pH 7.5), and the washing was collected, followed by washing with 10 column volumes of an elution buffer containing 150 mM imidazole (20 mM PBS, 500 mM NaCl, 150 mM imidazole, pH 7.5). The protein samples collected for SDS-PAGE were, if necessary, frozen at 20 C. or stored at 80 C., or purified to achieve a higher purity, by for example ion exchange chromatography, hydrophobic chromatography, and size exclusion chromatography.

    [0176] There are many shortcomings in the traditional methods for producing immunotoxins. For example, chemical reagents are required in the chemical coupling method, and the product is not uniform due to the large number of modification sites in the targeting polypeptide portion, and the chemical coupling bond is prone to breakage, leading to the leakage of toxicity. In the strategy of directly expressing the immunotoxin fusion protein, if a prokaryotic expression system is used, the target protein is often in the form of an inclusion body, the refolding efficiency is low and the steps are complicated. If an eukaryotic expression system is used, the expression of an immunotoxin may be limited by its natural toxicity to host cells. The method of the present disclosure achieves many advantages: 1. In the method of the present disclosure, only different targeting polypeptide portion and toxic polypeptide portion are needed to be prepared, and then combined to produce an immunotoxin which can target a different target and possess a different toxicity mechanism, with significant diversity and flexibility. 2. In the method of the present disclosure, the targeting polypeptide portion and the toxic polypeptide portion can be separately expressed in a suitable host cell. Where a special folding environment, especially advanced post-translational modifications, is/are needed, the expression can be carried out in mammalian cells. Where the polypeptide has no too much modification requirements, the expression can take place in E. coli. By means of the separate expression of the polypeptide of interest in a suitable expression system, higher production, yield and purity can be achieved. 3. In the method of the present disclosure, the linkage between the targeting polypeptide portion and the toxic polypeptide portion is site-specific, such that no by-products are produced, and the resulting product is highly homogeneous. 4. In the method of the present disclosure, the targeting polypeptide portion and the toxic polypeptide portion are linked via a peptide bond by means of the trans-splicing effect of intein, so good stability is exhibited compared with other connection models such as chemical coupling. 5. In the method of the present disclosure, the trans-splicing reaction conditions are mild, the reaction is efficient and can be easily integrated with other process and scaled up, and no harmful substance are needed to be added during the reaction.

    [0177] Specific embodiments of the present disclosure are described above. It should be understood that the present disclosure is not limited to the above specific embodiments, and various variations or modifications can be made by those skilled in the art without departing from the scope of the claims, which do not affect the essence of the present disclosure.