JP6744738B2 - Glycosynthase - Google Patents

Description

The present invention relates to glycosynthase, which is an enzyme having transglycosylation activity, and DNA encoding the enzyme.

Specifically, a glycoprotein, particularly an N-linked sugar chain bound to an immunoglobulin G (IgG) antibody, is cleaved with an endoglycosidase such as endo-β-N-acetylglucosaminidase (ENGase) to give a sugar chain of the protein. A protein in which only N-acetylglucosamine is bound to the asparagine residue that is the addition site, or a protein in which α1,6-linked fucose called core fucose is added to the N-acetylglucosamine is used as an acceptor substrate, and oxazoline is used. The present invention relates to an enzyme (glycosynthase) having an activity of transferring a sugar chain derived from a donor substrate to this acceptor substrate when various sugar chains including a sugar chain are used as a donor substrate, and a DNA encoding the enzyme.
Generally, a sugar chain of a glycoprotein has a heterogeneous structure. However, by making the sugar chain have a uniform structure, the function and stability of the glycoprotein may be improved, and the quality may be maintained. Further, by creating a glycoprotein having a specific or novel sugar chain structure, the function of the glycoprotein can be improved or a new function can be added. Glycosynthase is useful for producing glycoproteins having a uniform sugar chain structure. Glycosynthase is useful for producing a glycoprotein having a desired sugar chain structure. That is, glycosynthase is useful for producing a glycoprotein having a free sugar chain structure.

Many proteins are glycosylated as a post-translational modification. There are two types of glycosylation modes. When a sugar chain is added to Asn in the sequence Asn-Xaa (amino acids other than Pro)-Ser/Thr in the protein, N-linked sugar chain, in the protein When a sugar chain is added via the hydroxyl groups of Ser and Thr of, it is called an O-linked sugar chain.

The sugar chain is necessary for the protein to take proper folding, and is involved in promoting secretion of the protein, expression of function, stabilization in blood, intermolecular interaction and the like. By the way, sugar chains are not the direct products of genes. In other words, what kind of structure the sugar chain is added to a specific sugar chain addition site of the protein is not determined by the gene, and the difference in the expression level of each glycosyltransferase and the abundance of the sugar nucleotide substrate. It is affected by such things and changes each time. Therefore, sugar chains having the same structure are not always added to the same sugar chain addition site of the same protein, but rather, sugar chains having various structures are generally added. That is, when looking at a group of certain glycoproteins, a group that is uniform from the viewpoint of the amino acid sequence of the protein is usually a heterogeneous group from the viewpoint of the sugar chain structure.

In the case of glycoprotein preparations such as antibody drugs, such heterogeneity of sugar chains may cause variations in quality. In particular, when the sugar chain is directly related to the function and stability of the glycoprotein preparation, if the content of such functional sugar chain is different, it is considered that the potency and stability are also affected. Such non-uniformity of sugar chains is a problem in performing sufficient quality control as a drug.

By the way, functional sugar chains involved in functional expression of glycoproteins and stabilization in blood have a specific sugar chain structure. For example, in the case of an N-linked sugar chain bound to the Fc region of an IgG1 antibody, an antibody having a sugar chain having no core fucose has a higher antibody-dependent cytotoxic activity than an antibody having a sugar chain having a core fucose. It is known to show (see Non-Patent Document 1). That is, in the case of IgG1 antibody, if the ratio of such core-fucose-deficient sugar chain-containing antibody is high, it is considered that high antibody-dependent cellular cytotoxicity can be expected accordingly.

However, actually, for example, when the structural composition of sugar chains bound to trastuzumab, which is an antibody drug, is examined, the ratio of such core fucose-deficient sugar chains is several percent (see Non-Patent Document 2). In order to obtain a more potent antibody drug, it is desirable to produce an antibody to which only a functional sugar chain having a specific structure such as a core fucose-deleted sugar chain is added.
As described above, glycoprotein preparations such as antibodies having a sugar chain structure that may be present in a small amount or may not exist in the natural form may be extremely effective, but it is difficult to make them, and therefore development is easy. Not.

From the viewpoint of antibody quality control and functional improvement, sugar chain remodeling in which a sugar chain having a uniform structure is bound to the antibody has been attempted (see Non-Patent Document 3 and Patent Document 1). According to this method, the N-linked sugar chain bound to the Fc region of the antibody is a type of endo-β-N-acetylglucosaminidase and is called EndoS which has the activity of specifically cleaving the sugar chain of the antibody. It is removed using an enzyme (see Non-Patent Document 4). At this time, one residue of N-acetylglucosamine (GlcNAc) or a sugar chain having core fucose bound thereto remains in asparagine which is a sugar chain addition site of the antibody. Using this antibody as an acceptor substrate and an oxazoline sugar chain in which the reducing end of the sugar chain is an oxazoline body as a donor substrate, the glycosylation-modified mutant that suppresses the sugar chain-cleaving activity of EndoS and enhances the sugar chain transfer activity (EndoS D233A mutant or D233Q mutant) was allowed to act to add a sugar chain having a uniform structure to the antibody (FIG. 1).

In general, glycosidases such as endo-β-N-acetylglucosaminidase are known to show a sugar chain-cleaving activity and a transglycosylation activity (transglycosylation activity) when the substrate concentration is high. Utilizing this, by introducing a mutation in the vicinity of the active center of the enzyme, the glycosylation activity is suppressed and the glycosyltransferase activity is enhanced, and endo-β-N-acetylglucosaminidase is converted to glycosynthase. Attempts have been made. For example, Endo-M, which is an endo-β-N-acetylglucosaminidase derived from Mucor hemalis, is converted to glycosynthase by mutation of N175Q (see Non-Patent Document 5).

The substrate specificity of such glycosynthase is considered to be closely related to the substrate specificity of the original endo-β-N-acetylglucosaminidase. That is, if a glycoprotein sugar chain that is a substrate of the endo-β-N-acetylglucosaminidase is highly likely to be a substrate (can be transglycosylated) in the corresponding glycosynthase, the endo-β-N -If the glycoprotein sugar chain cannot be cleaved by acetylglucosaminidase, it may not be a substrate even in the corresponding glycosynthase.

Due to the three-dimensional structure of the antibody molecule, some endo-β-N-acetylglucosaminidases cannot be cleaved using the N-linked sugar chain (antibody sugar chain) bound to the Fc region of the antibody as a substrate. In addition, antibody sugar chains include high-mannose type sugar chains, double-chain complex type sugar chains, and hybrid type sugar chains. Furthermore, the presence or absence of sialic acid, the presence or absence of core fucose, the presence or absence of branched N-acetylglucosamine, and the binding mode There can be different types of sugar chains, such as differences. This difference in sugar chain structure also affects whether endo-β-N-acetylglucosaminidase can be cleaved as a substrate. For example, Endo-M cannot cleave antibody sugar chains bound to core fucose, and EndoS cannot cleave high mannose type antibody sugar chains.

Currently, the glycosynases that can be used for the remodeling of antibody sugar chains are limited, and even those that can be used are limited in substrate specificity. In addition, a large amount of enzyme may be required during the reaction, but the enzyme preparation operation may be complicated, so an enzyme having a high specific activity that reacts in a smaller amount is desired. Therefore, it is necessary to develop a glycosynthase having a wide substrate specificity and a high specific activity. Furthermore, in order to secure the stability of the enzyme and the substrate and control the enzyme reaction, it is possible to carry out the enzyme reaction under various pH conditions, but a glycosynthase that exhibits high activity even in a wide pH range is desired. ing.

International Publication No. 2013/120066 International Publication No. 2013/037824

Toyohide Shinkawa et al. , J. Biol. Chem. 278, 3466-3473 (2003). Carala W. N. Damen et al. , J. Am. Soc. Mass Spectrom. 20, 2021-2033 (2009) Wei Huang et al. , J. Am. Chem. Soc. 134, 12308-12318 (2012) Mattias Collin et al. , EMBO J.; 20, 3046-3055 (2001) Midori Umekawa et al. , J. Biol. Chem. 285, 511-521 (2010) Jonathan Sjogren et al. , Biochem. J. 455, 107-118 (2013) Jonathan Sjogren et al. , Glycobiology 25, 1053-1063 (2015). Beatriz Trastoy et al. , Proc. Natl. Acad. Sci. USA 111, 6714-6719 (2014).

An object of the present invention is to provide a glycosynase that can be used for remodeling of an antibody sugar chain, has broader substrate specificity than a conventional enzyme, has a high specific activity, and exhibits activity even in a wider pH range. It is to be.

The present inventors have found that for EndoS2 (see Patent Document 2 and Non-Patent Document 6), which is an endo-β-N-acetylglucosaminidase derived from Streptococcus pyogenes, a DxxDxDxE motif (x is an arbitrary amino acid, in EndoS2) in its active site. The effect of introducing a mutation into D179 to E186) was investigated. As a result, they found that the transglycosylation activity was enhanced by introducing a mutation into D182 or D184. In order to suppress the sugar chain-cleaving activity, further mutagenesis was examined, and it was confirmed that the sugar chain-cleaving activity was sufficiently suppressed in the D182Q and D182A/D184A mutants. When we investigated the transglycosylation activity of these mutants, surprisingly, the D182Q mutant had a higher specific activity and a wider pH range than the D233Q mutant of the conventional glycosynthase EndoS. It was found that it is a glycosynthase that exhibits high activity. Unlike EndoS, EndoS2 can cleave almost all double-chain antibody sugar chains including high-mannose type sugar chains, and also can cleave double-chain sugar chains of α1-acid glycoprotein (Non-Patent Document 6, Non-Patent Document 7), it was predicted that a wide range of substrate specificity would be exhibited in the transglycosylation reaction, and the present invention was completed here.

That is, according to the present invention, a glycosynthase having the following substrate specificity is provided.
A glycosynthase having the following substrate specificity.
Substrate specificity: An antibody in which only N-acetylglucosamine as a sugar chain component is bound to the "glycosylation site existing in the Fc region of the antibody", or "fucose attached to the N-acetylglucosamine by α1,6-bonding" In a reaction system in which "antibody in a state" is used as an acceptor substrate and various sugar chains are used as donor substrates, sugar chains derived from the donor substrate are efficiently transferred to the acceptor substrate under conditions of pH 4 or more and 10 or less.

Preferably, the present invention provides a glycosynthase having any of the following amino acid sequences.
(1) Amino acid sequence shown in SEQ ID NO: 1 or 3 of the sequence listing; or (2) Deletion, substitution and/or deletion of 1 to several amino acids in the amino acid sequence shown in SEQ ID NO: 1 or 3 of the sequence listing. Alternatively, an amino acid sequence having an addition and an amino acid sequence having a glycosynthase activity that can be used for remodeling of an antibody sugar chain:

According to another aspect of the present invention, there is provided a glycosynthase gene encoding the amino acid sequence of the glycosynthase of the present invention described above. Preferably, the present invention provides a glycosynthase having any of the following nucleotide sequences.
(1) a nucleotide sequence specified by nucleotide numbers 1 to 2532 in the nucleotide sequence set forth in SEQ ID NO: 2 in the sequence listing;
(2) having a base sequence having deletions, substitutions and/or additions of 1 to several bases in the base sequence specified by base numbers 1 to 2532 in the base sequence described in SEQ ID NO: 2 of the sequence listing , A nucleotide sequence encoding a protein having glycosynthase activity:
(3) the nucleotide sequence specified by nucleotide numbers 1 to 2532 in the nucleotide sequence set forth in SEQ ID NO: 4 in the sequence listing;
(4) It has a base sequence having 1 to several deletions, substitutions and/or additions in the base sequence specified by base numbers 1 to 2532 in the base sequence set forth in SEQ ID NO: 4 of the Sequence Listing. , A nucleotide sequence encoding a protein having glycosynthase activity:

According to another aspect of the present invention, a recombinant vector (preferably an expression vector) containing the above-mentioned glycosynthase gene of the present invention; a transformant transformed with the above-mentioned recombinant vector; and the above-mentioned There is provided a method for producing glycosynthase, which comprises culturing a transformant and collecting the glycosynthase of the present invention from the culture.

Further, according to another aspect of the present invention, the N-acetylglucosamine in which one or both of the H chains has fucose α1,6-linked as a sugar chain component using the glycosynthase of the present invention described above is the only antibody. Of the antibody bound to the glycosylation site existing in the Fc region of “,” is used as the acceptor substrate, the high mannose type sugar chain is used as the donor substrate, and the high mannose type sugar chain derived from the donor substrate is transferred to the acceptor substrate A method for producing an antibody having a high mannose type sugar chain having one or two core fucose added thereto; and having a high mannose type sugar chain having one or two core fucose added thereto A featured antibody is provided.

It is a conceptual diagram of the remodeling method of an antibody sugar chain. It is a figure which shows the result of the comparison of the amino acid sequences of EndoS and EndoS2 derived from Streptococcus pyogenes. * Indicates the same amino acid. It is a structural schematic diagram of various oxazoline sugar chains. It is a figure which shows the result of the transglycosylation reaction by various EndoS2 mutants. The arrow indicates the transglycosylation reaction product. It is a figure which shows the result of the transglycosylation reaction by an EndoS2 mutant when the substrate concentration is increased. The arrow indicates the transglycosylation reaction product. It is a figure which shows the result of the sugar chain cutting reaction by various EndoS2 mutants. It is a figure which shows the result of the transglycosylation reaction by EndoS2 D182Q mutant, D182A/D184A mutant, and EndoS D233Q mutant. The arrow indicates the transglycosylation reaction product. It is a figure which shows the result of the influence of pH with respect to transglycosylation activity. The arrow indicates the transglycosylation reaction product. It is a figure which shows the result of a sugar chain transfer reaction when A2-oxazoline or oxazoline of a high mannose type sugar chain is made to act as a donor substrate, and an EndoS2 D182Q mutant or an EndoS D233Q mutant is made to act with respect to an antibody acceptor substrate. The arrow indicates the transglycosylation reaction product. It is a figure which shows the result of a transglycosylation reaction by an EndoS2 D182X mutant. The arrow indicates the transglycosylation reaction product. It is a figure which shows the result of a sugar chain cutting reaction by an EndoS2 D182X mutant. It is a figure which shows the domain structure of EndoS, EndoS2, and an EndoS2 deletion mutant. Numbers indicate amino acid numbers. It is a figure which shows the result of the sugar chain cutting reaction by an EndoS2 deletion mutant. The arrow indicates the substrate glycoprotein with the sugar chain cleaved. It is a figure which shows the result of the transglycosylation reaction by an EndoS2 D182Q deletion mutant.

Hereinafter, the present invention will be described, but the present invention is not limited to the following specific forms and can be arbitrarily modified within the scope of the technical idea.

<1> Enzyme and Protein of the Present Invention The glycosynthase of the present invention is characterized by having the following substrate specificity.
A glycosynthase having the following substrate specificity.
Substrate specificity: An antibody in which only N-acetylglucosamine as a sugar chain component is bound to the "glycosylation site existing in the Fc region of the antibody", or "fucose attached to the N-acetylglucosamine by α1,6-bonding" In a reaction system in which "antibody in a state" is used as an acceptor substrate and various sugar chains are used as donor substrates, sugar chains derived from the donor substrate are efficiently transferred to the acceptor substrate under conditions of pH 4 or more and 10 or less.

Here, the “various sugar chains” refer to sugar chains in which N-acetylglucosamine (GlcNAc) and a leaving compound are bound in that order to the ends of the sugar chains. The sugar chain structures and forms of various sugar chains are not particularly limited. The elimination compound is not particularly limited, but is preferably oxazoline.
Examples of the above-mentioned various sugar chains include sugar chains in which N-acetylglucosamine (GlcNAc) and a leaving compound are bound to the ends of the sugar chains in this order (hereinafter, may be abbreviated as “oxazoline sugar chain”). .. The sugar chain structure and form of the oxazoline sugar chain are not particularly limited.

The substrate specificity described above is a property demonstrated for a mutant of EndoS2 derived from Streptococcus pyogenes NZ131 (serotype M49) strain obtained in the examples described herein. The origin of the glycosynthase of the present invention is not limited to that derived from the same strain, and it is present in a closely related strain and is produced by modifying endo-β-N-acetylglucosaminidase showing a high degree of homology with EndoS2. All of the above glycosyl synthases having the substrate specificity also belong to the scope of the present invention.

Examples of such an enzyme include an enzyme obtained by modifying a microorganism-derived endo-β-N-acetylglucosaminidase by a gene recombination technique, an enzyme such as adding various tags to the enzyme, or a glycosynthase activity by natural mutation. The enzyme and the like which have acquired the above can be mentioned, and all of them are included in the scope of the present invention.

An example of the glycosynthase of the present invention is a glycosynthase having any of the following amino acid sequences.
(1) Amino acid sequence described in SEQ ID NO: 1 or 3 of the sequence listing; or (2) Amino acid sequence described in SEQ ID NO: 1 or 3 of the sequence listing, deletion, substitution and/or deletion of 1 to several amino acids An amino acid sequence having an amino acid sequence having an addition and having a glycosynthase activity that can be used for remodeling of an antibody sugar chain:

Further, it should be understood that all proteins having a glycosynthase activity obtained by altering or modifying the active domain of the glycosynthase of the present invention or a part of the amino acid sequence thereof are included in the scope of the present invention. Preferred examples of such an active domain include the active domain of glycosynthase specified by 1 to 440 of the amino acid sequences shown in SEQ ID NOS: 1 and 3 of the sequence listing.

The range of "1 to several" in the "amino acid sequence having a deletion, substitution and/or addition of 1 to several amino acids" as used herein is not particularly limited, but is, for example, 1 to 20, preferably It means 1 to 10, more preferably 1 to 7, further preferably 1 to 5, and particularly preferably 1 to 3.

The method for obtaining the enzyme or protein of the present invention is not particularly limited, and may be a protein synthesized by chemical synthesis or a recombinant protein produced by a gene recombination technique.

When producing a recombinant protein, it is necessary to first obtain the DNA encoding the protein. An appropriate primer was designed by utilizing the information on the amino acid sequence and the nucleotide sequence described in SEQ ID NOS: 1 to 4 of the sequence listing of the present specification, and PCR was carried out using them by using an appropriate DNA library as a template. In some cases, the DNA encoding the enzyme of the present invention can be obtained by introducing mutation into the obtained DNA. Alternatively, the DNA encoding the enzyme of the present invention may be wholly synthesized by artificial gene synthesis.

For example, a method for isolating DNA encoding a glycosynthase having the amino acid sequences set forth in SEQ ID NO: 1 and SEQ ID NO: 3 is described in detail in the Examples below. However, the method for isolating the DNA encoding the glycosynthase of the present invention is not limited to these methods, and those skilled in the art can appropriately modify this method with reference to the methods described in Examples below. By changing it, the target DNA can be easily isolated. The enzyme of the present invention can be produced by introducing this DNA into an appropriate expression system. Expression in the expression system will be described later in this specification.

The glycosynthase of the present invention may become an insoluble protein after expression. In order to effectively utilize the above glycosynthase activity, it is possible to maintain the activity of the enzyme and produce a soluble form of the protein upon expression. Such proteins include fusion proteins of the glycosynthase of the present invention and glutathione-S-transferase (GST). However, the solubilization method is not limited to the above-mentioned method, and those skilled in the art can appropriately select a protein to be fused with glycosynthase or a peptide peptide serving as a tag, or can be solubilized by a chemical method. May be.

<2> Gene of the present invention According to the present invention, a glycosynthase gene encoding the amino acid sequence of the glycosynthase of the present invention is provided.
Specific examples of the glycosynthase gene encoding the amino acid sequence of glycosynthase of the present invention include genes having any of the following base sequences.
(1) a nucleotide sequence specified by nucleotide numbers 1 to 2532 in the nucleotide sequence set forth in SEQ ID NO: 2 in the sequence listing;
(2) having a base sequence having deletions, substitutions and/or additions of 1 to several bases in the base sequence specified by base numbers 1 to 2532 in the base sequence described in SEQ ID NO: 2 of the sequence listing , A nucleotide sequence encoding a protein having glycosynthase activity:
(3) the nucleotide sequence specified by nucleotide numbers 1 to 2532 in the nucleotide sequence set forth in SEQ ID NO: 4 in the sequence listing;
(4) It has a base sequence having 1 to several deletions, substitutions and/or additions in the base sequence specified by base numbers 1 to 2532 in the base sequence set forth in SEQ ID NO: 4 of the Sequence Listing. , A nucleotide sequence encoding a protein having glycosynthase activity:

The range of “1 to several” in the “base sequence having a deletion, substitution and/or addition of 1 to several bases” as used herein is not particularly limited, but is, for example, 1 to 60, preferably It means 1 to 30, more preferably 1 to 20, further preferably 1 to 10, and particularly preferably 1 to 3.

The method for obtaining the gene of the present invention is as described above. Further, methods for introducing a desired mutation into a predetermined nucleic acid sequence are known to those skilled in the art. For example, a mutation-containing DNA can be constructed by appropriately using a known technique such as a site-directed mutagenesis method, PCR using a degenerate oligonucleotide, exposure of a cell containing a nucleic acid to a mutagen or radiation. Such publicly known techniques are described in, for example, Molecular Cloning: A laboratory Manual, 2nd Ed. , Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. , 1989, and Current Protocols in Molecular Biology, Supplements 1-38, John Wiley & Sons (1987-1997).

<3> Recombinant Vector of the Present Invention The gene of the present invention can be used by inserting it into an appropriate vector. The type of vector used in the present invention is not particularly limited, and may be, for example, a vector that replicates autonomously (for example, a plasmid or a phage vector), or, when introduced into a host cell, it is integrated into the genome of the host cell, It may be replicated with the integrated chromosome.

Preferably, the vector used in the present invention is an expression vector. In the expression vector, the DNA of the present invention is operably linked with a DNA encoding a protein and elements necessary for transcription such as a promoter. The promoter is a DNA sequence that exhibits transcriptional activity in the host cell and can be appropriately selected depending on the type of host. Examples include E. coli lac, trp, or tac promoters and SV40 promoter that can operate in mammalian cells. In addition, these expression vectors may contain an appropriate selectable marker gene such as ampicillin resistance gene. Examples of the expression vector include, for example, pBluescript II SK+ vector (Stratagene), pcDNA3.1/Myc-His ver. A (Invitrogen) and the like can be mentioned, but the expression vector used in the present invention is not limited thereto.
It is well known to those skilled in the art how to ligate the gene (DNA) of the present invention, a promoter, a selectable marker gene, etc. and insert them into an appropriate vector.

<4> Transformant of the present invention A transformant can be prepared by introducing the gene (DNA) or recombinant vector of the present invention into an appropriate host.
The host cell into which the gene (DNA) or recombinant vector of the present invention is introduced may be any cell as long as it can express the DNA construct of the present invention, and examples thereof include bacteria, yeast, fungi and higher eukaryotic cells.

Examples of bacterial cells include Gram-positive bacteria such as Bacillus or Gram-negative bacteria such as Escherichia coli. Transformation of these bacteria may be performed by the protoplast method or the known method using competent cells.
Examples of mammalian cells include HEK293 cells and CHO cells. Methods for transforming mammalian cells and expressing the DNA sequences introduced into the cells are also known, and for example, electroporation, calcium phosphate method, lipofection method and the like can be used.
In addition, yeast cells, filamentous fungal cells, insect cells and the like can also be used as host cells, and a transformant can be obtained by introducing a DNA construct using a known transformation method suitable for each host.

<5> Method for Producing Enzyme of the Present Invention The method for producing glycosynase of the present invention is characterized by culturing the above transformant and collecting glycosynase from the culture.
The above transformants are cultured in a suitable nutrient medium under conditions that allow the expression of the introduced DNA construct. In order to isolate and purify the enzyme of the present invention from the culture of the transformant, ordinary protein isolation and purification methods may be used.
For example, when the enzyme of the present invention is expressed in a lysed state in cells, after the culture is completed, the cells are collected by centrifugation, suspended in an aqueous buffer solution, and then disrupted by an ultrasonic disruptor or the like. , To obtain a cell-free extract. From the supernatant obtained by centrifuging the cell-free extract, a usual protein isolation and purification method, that is, a solvent extraction method, a salting-out method using ammonium sulfate, a desalting method, a precipitation method using an organic solvent, Anion exchange chromatography using a resin such as diethylaminoethyl (DEAE) Sepharose, cation exchange chromatography using a resin such as sulfopropyl (SP) Sepharose, and hydrophobicity using a resin such as butyl sepharose and phenyl sepharose. A purified sample can be obtained by using, alone or in combination, a method such as a sex chromatography method, a gel filtration method using a molecular sieve, an affinity chromatography method, a chromatofocusing method, and an electrophoresis method such as isoelectric focusing. ..

The method for producing the protein of the present invention is not limited to the method using cells, and it can be produced by a cell-free protein synthesis system. Examples of the cell-free protein synthesis system include those prepared from wheat germ, Escherichia coli, rabbit reticulocytes, insect cells and the like. The DNA encoding the target protein is incorporated into a vector capable of reverse transcription, and the mRNA synthesized by the reverse transcription reaction is added to an appropriate cell-free protein synthesis system to synthesize the target protein. Then, the target protein is purified from the reaction solution in the same manner as above.

<6> Method for producing sugar chain-modified antibody using enzyme of the present invention and reaction product thereof The method for producing an antibody of the present invention uses the above glycosyl synthase, and one or both H chains have fucose α1 as a sugar chain component. , 6-linked N-acetylglucosamine alone is used as the acceptor substrate for the antibody bound to the “glycoside addition site existing in the Fc region of the antibody”, and the high mannose type sugar chain is used as the donor substrate. A method for producing an antibody, which comprises a high-mannose type sugar chain to which one or two core fucose is added, which comprises a reaction step of transferring a type sugar chain to an acceptor substrate.

Further, the antibody of the present invention is characterized by having a high-mannose type sugar chain to which one or two core fucose is added.

An antibody in which only N-acetylglucosamine as a sugar chain component is bound to "a sugar chain addition site existing in the Fc region of the antibody", or an antibody in a state in which fucose bound to the N-acetylglucosamine by α1,6-bond is added Is used as an acceptor substrate and various sugar chains are used as donor substrates, to produce an antibody having a sugar chain composition different from that of the antibody which is a raw material of the acceptor substrate when the glycosidase of the present invention is allowed to act. You can Furthermore, the glycosynthase of the present invention can be used to produce an antibody having a non-natural sugar chain or an antibody having a sugar chain which is not present in ordinary antibodies.

For example, core fucose is not bound to the natural high mannose type sugar chain. However, an antibody having a sugar chain to which core fucose is bound is reacted with endo-β-N-acetylglucosaminidase to cleave the sugar chain, and the resulting antibody is used as an acceptor substrate, which contains 5 (M5) or 6 (M6) mannose. An antibody having a non-natural high mannose type sugar chain (M5 or M6) to which core fucose is bound can be produced when the glycosidase of the present invention is allowed to act using an oxazoline mannose type sugar chain as a donor substrate. it can. That is, by combining an acceptor substrate with or without core fucose and an oxazoline donor substrate of various sugar chains, it is possible to produce an antibody having a sugar chain that does not exist in ordinary antibodies. In this case, the donor substrate used is a sugar chain having a structure not included in the sugar chain of an antibody prepared from a glycoprotein other than an antibody, or a chemically synthesized sugar chain (including a non-natural sugar chain) in an oxazoline form. May be.

The present invention will be described in more detail with reference to the examples described below, but the present invention is not limited to these examples.

<Example 1>
Preparation of EndoS2 mutant In this example, the transglycosylation activity of the EndoS2 mutant derived from Streptococcus pyogenes was examined. EndoS2 is a kind of endo β-N-acetylglucosaminidase consisting of 843 amino acids belonging to the GH18 family in the CAZy (Carbohydrate-Active enZYmes) database (Fig. 2), which is an N-linked sugar linked to the Fc region of IgG antibody. It has been reported that the N-linked sugar chain bound to the chain or α1-Acid glycoprotein is cleaved (see Non-Patent Document 6).

First, using Streptococcus pyogenes NZ131 (serotype M49) strain-derived genomic DNA (American Type Culture Collection) as a template, primers NG1370+ and NG1371- (SEQ ID NOS: 5 and 6) and KAPAHiSTiAxHiATHiSHiHit are sequenced with KAPAHiSHiSHiHit. Amplification was performed, and the full-length DNA of the EndoS2 gene amplified here was cloned into the EcoRV site of pBluescript II SK+ vector (Stratagene). From this, a DNA fragment cut out with BamHI and XhoI was inserted into the BamHI-XhoI site of pGEX-6P-1 (GE Healthcare) to obtain a wild-type EndoS2 (WT) expression vector. In addition, when this expression vector is used, EndoS2 is expressed in a form fused to the C-terminal side of glutathione-S-transferase (GST). This GST can be cut by processing with PreScission Protease (GE Healthcare).

Next, an attempt was made to produce an EndoS2 mutant that suppressed the glycosylation activity and enhanced the transglycosylation activity. In the GH18 family enzymes, the motif DxxDxDxE (x is an arbitrary amino acid) is conserved in the amino acid sequence near the active site of the enzyme. It has been reported that acidic amino acids such as D and E play an important role in activity expression, and that introduction of a mutation at this site may lead to a decrease in glycosylation activity and a marked transglycosylation activity. (See Non-Patent Documents 3 and 5). Therefore, D179A mutant in which D179 in the motif (D179 to E186) of EndoS2 is replaced with Ala, D182A mutant in which D182 is replaced with Ala, D182Q mutant in which D182 is replaced with Gln, D184A mutation with D184 replaced by Ala , D184Q mutant in which D184 was replaced by Gln, D184N mutant in which D184 was replaced by Asn, E186A mutant in which E186 was replaced by Ala, and D182A/D184A mutant in which D182 and D184 were replaced by Ala were prepared.

In order to replace D179 with Ala, a combination of the primers NG1390+ and NG1391-(SEQ ID NOS: 7 and 8) and KAPA HiFi HotStart ReadyMix (KAPA BIOSYSTEMS) were used to prepare the pBluescript containing the EndoS2 gene. Was used as a template to introduce mutations by PCR. In order to replace D182 with Ala, mutations were introduced by PCR in the same manner as above using the combination of each primer of the primers NG1392+ and NG1393-(Sequence ID No. 9, 10). In order to replace D182 with Gln, mutation was introduced by PCR in the same manner as above using the combination of each primer of the primers NG1451+ and NG1452− (SEQ ID NOS: 11 and 12 in Sequence Listing). In order to replace D184 with Ala, a combination of primers NG1374+ and NG1375− (SEQ ID NOS: 13 and 14 in the Sequence Listing) was used, and mutation was introduced by PCR in the same manner as above. In order to replace D184 with Gln, a combination of primers NG1381+ and NG1382- (SEQ ID NOS: 15 and 16 in the Sequence Listing) was used, and mutation was introduced by PCR in the same manner as above. In order to replace D184 with Asn, a combination of primers NG1383+, NG1384- (SEQ ID NO: 17, 18 in the Sequence Listing) was used, and mutation was introduced by PCR in the same manner as above. Further, in order to replace E186 with Ala, a combination of primers NG1385+ and NG1386− (SEQ ID NOS: 19 and 20 in Sequence Listing) was used, and mutation was introduced by PCR in the same manner as above. Furthermore, in order to replace D182 and D184 with Ala, a combination of primers NG1419+ and NG1420− (SEQ ID NOS: 21 and 22 in the Sequence Listing) was used, and mutation was introduced by PCR in the same manner as above. In order to replace D182 with Ala and D184 with Gln, a combination of primers NG1455+ and NG1456- (SEQ ID NOS: 23 and 24 in the Sequence Listing) was used, and mutation was introduced by PCR in the same manner as above. In order to replace D182 with Gln and D184 with Ala, a combination of primers NG1453+ and NG1454-(SEQ ID NO: 25, 26 in the Sequence Listing) was used, and mutation was introduced by PCR in the same manner as above. In order to replace D182 and D184 with Gln, a combination of primers NG1457+ and NG1458− (SEQ ID NOS: 27 and 28 in Sequence Listing) was used, and mutation was introduced by PCR in the same manner as above. Further, in order to replace D184 and E186 with Ala, a combination of primers NG1387+ and NG1388- (SEQ ID NOs: 29 and 30 in the sequence listing) was used, and mutation was introduced by PCR in the same manner as above.
In either case, the introduction of mutation was confirmed by DNA sequence. The mutated EndoS2 gene was inserted into the BamHI-XhoI site of pGEX-6P-1 (GE Healthcare) to give an expression vector for each mutant.

Escherichia coli BL21(DE3) strain was transformed with each expression vector constructed as described above. The obtained transformant was shake-cultured overnight at 37° C. in 2 mL of liquid LB medium supplemented with ampicillin at 50 μg/mL.
Next, 1 mL of this seed culture was taken, and 100 mL of ampicillin-containing liquid LB medium was newly inoculated and shake-cultured at 37° C. for 3 hours. Here, Isopropyl-β-D-thiogalactoside (IPTG) was added so that the concentration was 0.1 mM, and the mixture was further shake-cultured at 37° C. for 3 hours, and then the cells were collected by centrifugation. To this bacterial cell, 5 mL of phosphate buffered saline (PBS) was added, stirred, and further sonicated to disrupt the bacterial cell. 50 μL of Triton X-100 was added to the disrupted cell suspension, and the mixture was shaken at room temperature for 30 minutes. Then, ultrasonic treatment was performed again, and the mixture was further separated into a soluble fraction and an insoluble fraction by centrifugation.

An appropriate amount (-100 μL) of Glutatione Sepharose (GE Healthcare) was added to the soluble fraction obtained here, and the mixture was mixed overnight at 4° C. to adsorb the GST-EndoS2 mutant fusion protein. The Glutation Sepharose was collected by centrifugation and washed with PBS several times. Then, a cleavage reaction buffer solution (50 mM Tris-HCl, 150 mM NaCl, 1 mM EDTA, 1 mM DTT, pH 7.0) by PreScission Protease and PreScission Protease were added, and the GST cleavage reaction was performed at 4° C. for 16 hours.
Next, Glutatione Sepharose was removed by centrifugation, and the resulting reaction solution was centrifugally concentrated with VIVASPIN 500 (Sartorius Stedim), and PBS was added to perform buffer exchange. In this way, an enzyme solution of each EndoS2 mutant was obtained.

In addition, EndoS, another endoglycosidase derived from Streptococcus pyogenes, shows 37% homology with EndoS2 in amino acid sequence comparison (FIG. 2), but refer to the literature for enzyme solutions of this wild type and D233Q mutant. Was prepared (see Non-Patent Document 3).

<Example 2>
Glycosyl transfer reaction Using the enzyme solution obtained above, a glycosyl transfer reaction was carried out on a human IgG antibody. This transfer reaction requires a donor substrate that serves as a sugar chain donor and an acceptor substrate to which the sugar chain is transferred. As the acceptor substrate, the major part of the N-linked sugar chain bound to the Fc region of the antibody drug trastuzumab is removed by combining endoglycosidases such as EndoS, EndoD, EndoLL, and only N-acetylglucosamine is bonded as the sugar chain component. (When trastuzumab <herein referred to as HerB> produced in silkworm as a raw material is used) or an antibody obtained by adding α1,6-linked fucose called core fucose to its N-acetylglucosamine ( As a raw material, trastuzumab <herein, referred to as HerC> produced in CHO cells was used. Various sugar chains, more specifically, oxazoline derivatives of various sugar chains shown in FIG. 3 were used as a donor substrate serving as a sugar chain donor. The details of the method for preparing the acceptor substrate are currently pending (Japanese Patent Application No. 2014-0911157, Japanese Patent Application No. 2014-222191, Japanese Patent Application No. 2015-082388).

The enzyme reaction solution was prepared so as to contain 1 μg of the acceptor substrate, 1 μg of the donor substrate, and 1 μg of the EndoS2 mutant in 10 μL of 50 mM Tris-HCl (pH 7.4). This was reacted at 30° C. and sampled over time. The enzymatic reaction was stopped by adding a sample buffer for SDS polyacrylamide gel electrophoresis (SDS-PAGE). This sample solution was boiled and then subjected to SDS-PAGE. After completion of electrophoresis, the cells were stained with Quick-CBB (Wako Pure Chemical Industries), and the presence or absence of transfer products was confirmed by focusing on the H chain of the antibody.

FIG. 4 shows an antibody obtained by cleaving a sugar chain of trastuzumab (HerC) produced in CHO cells with various endoglycosidases as an acceptor substrate (HerC acceptor: with core fucose), A2-oxazoline (A2-Oxa) or G2-oxazoline ( G2-Oxa) is a result of carrying out a transglycosylation reaction with an EndoS2 mutant using G2_Oxa) as a donor substrate. The number on the horizontal axis in FIG. 4 indicates the time from the start of the enzymatic reaction (unit: hour). The arrow indicates the transglycosylation reaction product.
In the D182A mutant, D184A mutant, D184N mutant, and D184Q mutant, a band considered to be a transfer product was detected together with an unreacted acceptor substrate band. In some mutants, the band thought to be the transfer product disappeared due to the long-term reaction, which is probably because the sugar chain of the transfer product was cleaved because the sugar chain-cleaving activity of the mutant was not sufficiently suppressed. Was given.
On the other hand, in the D179A mutant and the E186A mutant, such a band considered as a transfer product was not detected. From the above results, in the case of EndoS2, in order to enhance the transglycosylation activity, it is effective to introduce a mutation into D182 or D184 among the DxxDxDxE motifs, and depending on the amino acid to be substituted, the glycosylation activity can be suppressed. It was thought that there was a difference.

By the way, the reaction conditions shown in FIG. 4 were used to screen whether or not each mutant had transglycosylation activity, and the substrate concentration was low. Therefore, it was decided to investigate the transglycosylation activity of the EndoS2 mutant under a more optimized reaction condition by increasing the substrate concentration. Referring to the reaction conditions described in Non-Patent Document 3, 10 μg of the acceptor substrate and 150 times the molar ratio of the donor substrate and EndoS2 were added to the enzyme reaction solution in 2 μL of 50 mM Tris-HCl (pH 7.4). Each mutant was prepared so as to contain 0.2 μg (2.1 pmol). This was reacted at 30° C. and sampled over time. The enzymatic reaction was stopped by adding the sample buffer for SDS-PAGE. This sample solution was boiled and then subjected to SDS-PAGE. After completion of electrophoresis, the cells were stained with Quick-CBB (Wako Pure Chemical Industries), and the presence or absence of transfer products was confirmed by focusing on the H chain of the antibody.

FIG. 5 shows the results of transglycosylation reaction with the EndoS2 mutant using HerC acceptor as the acceptor substrate and A2-oxazoline as the donor substrate. The number on the horizontal axis in FIG. 5 indicates the time from the start of the enzymatic reaction (unit: hour). The arrow indicates the transglycosylation reaction product.
Different from the case of FIG. 4, in the case of any of D182A mutant, D184A mutant, D184N mutant, D184Q mutant, any of the mutants, glycosyl transfer was observed in 80 to 90% or more of the acceptor substrate in 30 minutes of reaction time. Made However, as the reaction time progressed, the amount of transfer products decreased. That is, although these mutants exhibited transglycosylation activity, the transglycosylation activity was not sufficiently suppressed, and therefore it was considered that the transfection product glycosylation is cleaved if the reaction time is long.

Therefore, an attempt was made to screen for an EndoS2 mutant in which the glycosylation activity was sufficiently suppressed. Regarding the reaction conditions, 1 μg of human serum-derived IgG and 0.3 μg of EndoS2 mutant were added to 20 μL of PBS and reacted at 37° C. for 16 hours. Then, the enzyme reaction was stopped by adding the sample buffer for SDS-PAGE. This sample solution was boiled and then subjected to SDS-PAGE. After the electrophoresis, it was stained with Quick-CBB (Wako Pure Chemical Industries).
As a result, among the investigated mutants, sugars of D182Q mutant, D182A/D184A mutant, D182A/D184Q mutant, D182Q/D184A mutant, D182Q/D184Q mutant, E186A mutant, and D184A/E186A mutant. The strand scission activity was suppressed (Fig. 6). When the transglycosylation activity of these mutants was examined, the D182Q mutant showed the highest transglycosylation activity, and then the D182A/D184A mutant showed the high transglycosylation activity. The transglycosylation activity was observed in the D182A/D184Q mutant, the D182Q/D184A mutant, and the D182Q/D184Q mutant, but the E186A mutant and the D184A/E186A mutant had no transglycosylation activity. Therefore, the glycosyl transfer activity of the D182Q mutant (SEQ ID NOS: 1 and 2 in the sequence listing) and the D182A/D184A mutant (SEQ ID NOS: 3 and 4 in the sequence listing) was examined in more detail.

The enzyme reaction solution was prepared so that 2 μL of 50 mM Tris-HCl (pH 7.4) contained 10 μg of the acceptor substrate, a 150-fold molar ratio of the donor substrate, and 0.2 μg of the EndoS2 mutant. A similar enzymatic reaction solution was prepared for the D233Q mutant of EndoS. These were reacted at 30° C. and sampled over time. The enzymatic reaction was stopped by adding the sample buffer for SDS-PAGE. This sample solution was boiled and then subjected to SDS-PAGE. After completion of electrophoresis, the cells were stained with Quick-CBB (Wako Pure Chemical Industries), and the presence or absence of transfer products was confirmed by focusing on the H chain of the antibody.

FIG. 7 shows the results of measuring the transglycosylation activity of the D182Q mutant and the D182A/D184A mutant when the oxazoline derivative of each sugar chain was used as the donor substrate and the HerC acceptor or HerB acceptor was used as the acceptor substrate.
The number on the horizontal axis in FIG. 7 indicates the time from the start of the enzymatic reaction (unit: hour). In FIG. 7, the circled number 1 is the result of the glycosyl transfer reaction by the D182Q mutant of EndoS2, the circled number 2 is the result of the sugar chain transfer reaction by the D182A/D184A mutant of EndoS2, and the circled number 3 is the D233Q mutation of EndoS. The result of the transglycosylation reaction by the body is shown. The band in FIG. 7 is the antibody H chain, and the band shifted to the polymer side shown by the arrow is the antibody H chain to which the sugar chain transfer has been performed.

When HerC acceptor was used as the acceptor substrate, the transglycosylation activity was higher in the order of EndoS2 D182Q mutant>EndoS D233Q mutant>EndoS2 D182A/D184A mutant, regardless of the type of donor substrate (FIG. 7). .. A similar tendency was observed when an antibody obtained by cleaving the sugar chain of trastuzumab (HerB) produced in silkworm with various endoglycosidases was used as the acceptor substrate (HerB acceptor: no core fucose), but depending on the substrate, EndoS2 In some cases, the D182A/D184A mutant of EndoS had higher transfer efficiency than the D233Q mutant of EndoS. In the case of HerB acceptor, an unreacted acceptor substrate-derived band is always observed in addition to the glycosyl transfer product. This is because the acceptor substrate originally contains an antibody to which no sugar chain is bound. This is because

Next, the buffer solution used for the reaction was changed, and the transglycosylation activity of these mutants at various pHs was examined. 50 mM citrate buffer for pH 3.5 to 6.0, 50 mM Tris-HCl for pH 7.4 to 8.8, 50 mM glycine-NaOH buffer for pH 10.0 and pH 11.0, and the same procedure as above. The enzyme reaction was carried out.
The results are shown in Fig. 8. The number on the horizontal axis in FIG. 8 indicates the time from the start of the enzymatic reaction (unit: hour). Further, the circled numbers in FIG. 8 indicate the results of the transglycosylation reaction by the same enzyme as in FIG. 7. The band in FIG. 8 is the antibody H chain, and the band shifted to the polymer side shown by the arrow is the antibody H chain to which the sugar chain transfer has been performed.

As a result, when HerC acceptor was used as the acceptor substrate, all the mutants exhibited transglycosylation activity between pH 4 and 10, but high activity was obtained between pH 6 and 8 (FIG. 8). .. Among them, the D182Q mutant of EndoS2, unlike other mutants, was confirmed to have a clear transfer product between pH 4 and 10 even at a reaction time of 30 minutes, and had the highest transglycosylation activity. When HerB acceptor was used as the acceptor substrate, transglycosylation activity was confirmed between pH 6 and 8.8 in all the mutants. In this case as well, the glycosyl transfer activity of the D182Q mutant of EndoS2 was confirmed. It was the highest.

<Example 3>
Glycosylation reaction of high mannose type sugar chain Regarding the cleavage of antibody sugar chain, EndoS does not cleave high mannose type sugar chain, although EndoS does not cleave high mannose type sugar chain because of the difference in enzymatic properties between EndoS and EndoS2. It has been reported (see Non-Patent Document 7). In the transglycosylation reaction by the above EndoS mutant and EndoS2 mutant, transfer of a complex type sugar chain (A2-Oxa, G2-Oxa, G0-Oxa) and a pouch mannose type sugar chain (M3-Oxa) was investigated, and both Both were shown to have transfer activity for these sugar chains, but the transfer activity for high-mannose type sugar chains was not clear. Therefore, it was decided to investigate whether the EndoS D233Q mutant and the EndoS2 D182Q mutant have a difference in the transfer activity of the high-mannose type sugar chain as well as the sugar chain-cleaving activity.

The high-mannose type sugar chain was prepared by treating ribonuclease B derived from bovine pancreas with EndoH (New England Biolabs), which is a type of endo-β-N-acetylglucosaminidase, and converted it into an oxazoline body by a chemical method. The obtained product was used as a donor substrate (high M-Oxa) for the transglycosylation reaction. In addition, when the composition of the high mannose type sugar chain was examined by mass spectrometry, it was found that M5 sugar chain consisting of 5 mannose was most contained, followed by M6 sugar chain consisting of 6 mannose, and other mannose. The amount of M7, M8, and M9 sugar chains consisting of 7 to 9 was very small.

The enzyme reaction solution for the transglycosylation reaction using a high-mannose type sugar chain was 10 μg of an acceptor substrate (prepared from HerC or HerB) in 2.5 μL of 50 mM Tris-HCl (pH 7.0) and a donor substrate. Was prepared to contain 0.2 mM of EndoS D233Q mutant or EndoS2 D182Q mutant. This was reacted at 30° C. and sampled over time. The enzymatic reaction was stopped by adding the sample buffer for SDS-PAGE. This sample solution was boiled and then subjected to SDS-PAGE. After completion of the electrophoresis, the cells were stained with Quick-CBB (Wako Pure Chemical Industries), and the presence or absence of transfer products was confirmed by focusing on the H chain of the antibody. As a control, when A2-oxazoline (A2-Oxa) was used as a donor substrate, the reaction was performed and analyzed in the same manner as above.

FIG. 9 is a result of analyzing the above reaction product by SDS-PAGE. The number on the horizontal axis in FIG. 9 indicates the time from the start of the enzymatic reaction (unit: hour). The arrow indicates the transglycosylation reaction product. In the case of the EndoS D233Q mutant, when A2-oxazoline is used as a donor substrate, the transfer efficiency is lower than that of the EndoS2 D182Q mutant, but both HerC acceptor and HerB acceptor transfer sugar chains to most acceptors within 5 hours. It was The band observed at the position of the HerB acceptor even after the transfer reaction is originally derived from an antibody in which no sugar chain is bound. On the other hand, when mannose-type sugar chain oxazoline was used as the donor substrate, no clear sugar chain transfer was observed for both HerC acceptor and HerB acceptor. Mass spectrometry was performed on the reaction product after 1 hour of reaction, but no transfer product was detected. From this, it was considered that the EndoS mutant has no ability to transfer high-mannose type sugar chains.

In the case of the EndoS2 D182Q mutant, when A2-oxazoline was used as the donor substrate, sugar chains were transferred to most of the HerC acceptor and HerB acceptor within 1 hour. When a high-mannose type sugar chain oxazoline is used as the donor substrate, the HerC acceptor transfers the sugar chain to about half of the acceptors within 1 hour, and even after extending the reaction time, the amount of transfer product is increased. There was no change. Regarding the HerB acceptor, the transglycosylation efficiency was 5 to 6% in 1 hour of the reaction, but the amount of the transferred product did not change significantly even if the reaction time was extended thereafter.

As for the HerB acceptor, M5 sugar chain was detected by mass spectrometry as the main transfer product because there was no core fucose. On the other hand, in the case of HerC acceptor, since 90% or more thereof contains core fucose, most of the above-mentioned transfer products are contained as core fucose-linked sugar chains. In fact, mass spectrometry detected M5 sugar chains with fucose and M6 sugar chains with fucose on the antibody. Usually, no core fucose is bound to the high-mannose type sugar chain, which indicates that in the present transfer reaction, surprisingly, a high-mannose type sugar chain having an unnatural type core fucose was produced.

From the above results, it can be said that the mannose-type sugar chain oxazoline has a lower efficiency of transglycosylation than the case of A2-oxazoline, but the EndoS2 D182Q mutant has a high mannose-type sugar chain transfer ability. Furthermore, a non-natural sugar chain was generated in this transfer reaction. In this respect, the present invention is an enzyme having a wider substrate specificity than EndoS D233Q, which is a conventional glycosynthase, and a versatile enzyme capable of adding a sugar chain which a normal antibody does not have to an antibody. It can be said that there is.

<Example 4>
Analysis of EndoS2 D182X mutant As a mutant in which D182 of EndoS2 was substituted, mutants of D182A and D182Q were prepared as described above. On the other hand, by substituting this residue with another amino acid, a mutant with higher transglycosylation activity or suppressed residual hydrolytic activity may be obtained. Therefore, it was decided to prepare a mutant in which D182 was replaced with an amino acid other than Ala or Gln, and to investigate its transglycosylation activity and glycosylation activity.

Preparation of mutants and preparation of mutant enzymes were performed in the same manner as the above-mentioned method. In order to replace D182 with Phe, a combination of the primers NG1520+ and NG1521- (SEQ ID NOS: 31 and 32 in the Sequence Listing) was used. In order to replace D182 with Leu, a combination of the primers NG1522+ and NG1523-(Sequence ID Nos. 33 and 34) was used. In order to replace D182 with Ile, a combination of the primers NG1524+ and NG1525- (SEQ ID NO: 35, 36 in Sequence Listing) was used. In order to replace D182 with Met, a combination of the primers NG1526+ and NG1527- (SEQ ID NOs: 37 and 38 in the Sequence Listing) was used. In order to replace D182 with Ser, a combination of primers NG1530+ and NG1531-(SEQ ID NO: 39, 40 in the Sequence Listing) was used. In order to replace D182 with Pro, a combination of primers NG1532+ and NG1533− (SEQ ID NOS: 41 and 42 in the Sequence Listing) was used.

In order to replace D182 with Thr, each primer combination of primers NG1534+ and NG1535− (SEQ ID NO: 43, 44 in Sequence Listing) was used. In order to replace D182 with Tyr, a combination of primers NG1536+ and NG1537- (SEQ ID NO: 45, 46 in the Sequence Listing) was used. In order to replace D182 with His, a combination of primers NG1538+ and NG1539− (SEQ ID NOS: 47 and 48 in the Sequence Listing) was used. In order to replace D182 with Asn, a combination of the primers NG1540+ and NG1541-(Sequence ID No. 49, 50) was used. In order to replace D182 with Lys, a combination of the primers NG1542+ and NG1543-(Sequence ID Nos. 51 and 52) was used. In order to replace D182 with Glu, a combination of the primers NG1544+ and NG1545-(Sequence ID No. 53, 54) was used.

In order to replace D182 with Cys, a combination of primers NG1546+ and NG1547− (SEQ ID NOS: 55 and 56 in Sequence Listing) was used. In order to replace D182 with Trp, a combination of primers NG1548+ and NG1549− (SEQ ID NOs: 57 and 58 in Sequence Listing) was used. In order to replace D182 with Arg, a combination of the primers NG1550+ and NG1551-((SEQ ID NO: 59, 60 in Sequence Listing) was used. In order to replace D182 with Gly, each primer combination of primers NG1552+ and NG1553− (SEQ ID NO: 61, 62 in Sequence Listing) was used. In order to replace D182 with Val, each primer combination of primers NG1554+ and NG1555- (SEQ ID NO: 63, 64 in Sequence Listing) was used.

The enzyme reaction solution for the transglycosylation reaction was prepared so that 2.5 μL of 50 mM Tris-HCl (pH 7.0) contained 10 μg of HerC acceptor, 6 mM of A2-oxazoline, and 0.2 μg of EndoS2 mutant. did. These were reacted at 30° C. and sampled over time. The enzymatic reaction was stopped by adding the sample buffer for SDS-PAGE. This sample solution was boiled and then subjected to SDS-PAGE. After completion of electrophoresis, the cells were stained with Quick-CBB (Wako Pure Chemical Industries), and the presence or absence of transfer products was confirmed by focusing on the H chain of the antibody.

FIG. 10 shows the results of measuring the transglycosylation activity of each EndoS2 mutant. The number on the horizontal axis in FIG. 10 indicates the time from the start of the enzymatic reaction (unit: hour). The band in FIG. 10 is the antibody H chain, and the band shifted to the polymer side shown by the arrow is the antibody H chain to which sugar chain transfer has been performed. From this result, each mutant is roughly classified into three groups. That is, the circled number 1 is a group of mutants (D182C, D182H, D182L, D182M, D182N, D182Q, D182S, D182T, D182V) having a high transglycosylation activity and suppressed glycosylation activity, and a circled number. 2 is a group of mutants in which the sugar chain-cleaving activity remains (the sugar chain of the transglycosylation product is cleaved by a long-term reaction. D182A, D182E, D182G, D182I and wild-type EndoS2), and the circled number 3 is This is a group of mutants having low transglycosylation activity (D182F, D182K, D182P, D182R, D182W, D182Y). Among them, the group of circled number 3 having a low transglycosylation activity was composed of a mutant in which D182 was replaced with a basic amino acid such as Lys or Arg, or an amino acid having an aromatic ring such as Phe, Trp, or Tyr, or Pro. ..

Next, the sugar chain-cleaving activity of each mutant was examined. The reaction conditions were as follows: 1 μg of human serum-derived IgG and 0.2 μg of EndoS2 mutation in 10 μL of buffer (50 mM citrate buffer (pH 6.0), 50 mM Tris-HCl (pH 7.0), or PBS). The body was added and reacted at 30° C. for 15 hours. Then, the enzyme reaction was stopped by adding the sample buffer for SDS-PAGE. This sample solution was boiled and then subjected to SDS-PAGE. After the electrophoresis, it was stained with Quick-CBB (Wako Pure Chemical Industries).

The result is shown in FIG. The band in FIG. 11 is the antibody H chain, the position of the band in the lane with hIgG is in the state where the sugar chain is bound, and the position of the band in the lane with the PNGaseF is in the state where the sugar chain is cleaved. belongs to. Circled numbers 1 to 3 are based on the grouping of FIG. In this case, the mutant in the group of circled number 3 had no sugar chain-breaking activity. Most of the mutants in the groups of circled numbers 1 and 2 showed sugar chain-cleaving activity in the 15-hour reaction, but the sugar chain-cleaving activity was suppressed depending on the reaction conditions (D182H, D182L, D182M, D182Q, D182A). That is, it can be said that a more efficient glycosyl transfer reaction can be performed than before by using a mutant having a high transglycosylation activity and performing the transglycosylation reaction under a reaction condition that suppresses the glycosylation activity.

<Example 5>
Analysis of EndoS2 Deletion Mutant A crystal structure of EndoS has been reported (see Non-Patent Document 8), and amino acid numbers 1 to 36 of EndoS (full length 995 amino acids) are signal peptides (SP), amino acid numbers 37 to. 97 is a putative coiled coil region (CC), amino acid numbers 98 to 445 are catalytic domains (Catalytic D), amino acid numbers 446 to 631 are leucine rich repeat domains (LRR), and amino acid numbers 632 to 764 are hybrid Ig domains (Hybrid Ig D). ), amino acid numbers 765 to 923 are carbohydrate-binding modules (CBM), and amino acid numbers 924 to 995 are three-helix bundle domains (3H) (FIG. 12).

By analogy with the homology of the amino acid sequence of the domain structure of EndoS2 (full length 843 amino acids), amino acid numbers 1 to 36 were deleted from the signal peptide, and the portion corresponding to the putative coiled coil region was deleted, and amino acid numbers 37 to 384 were catalytic. Domains, amino acids Nos. 385 to 547 are LRRs, amino acids Nos. 548 to 677 are hybrid Ig domains, amino acids Nos. 678 to 843 are CBM, and a part corresponding to the three-helix bundle domain is deleted (FIG. 12). We investigated how the enzyme activity would be affected if such a domain structure was deleted.

As the EndoS2 deletion mutant, CBM-deleted EndoS2 Δ (679-843), Hybrid Ig domain and CBM-deleted EndoS2 Δ (548-843), and LRR, Hybrid Ig domain, and CBM are deleted. The missing EndoS2 Δ(386-843) was constructed (FIG. 13). The DNA encoding EndoS2Δ(679-843) was subjected to PCR using primers NG1370+, NG1560− (SEQ ID NOS: 5 and 65 in Sequence Listing) and KAPA HiFi HotStart ReadyMix (KAPA BIOSYSTEMS) using the DNA encoding the full-length EndoS2 as a template. After amplification and synthesis, the amplified DNA was cloned into the EcoRV site of pBluescript II SK+ vector (Stratagene). From this, a DNA fragment cut out with BamHI and XhoI was inserted into the BamHI-XhoI site of pGEX-6P-1 (GE Healthcare) to give an EndoS2Δ(679-843) expression vector.

DNA encoding EndoS2Δ(548-843) was subjected to PCR using primers NG1370+, NG1559− (SEQ ID NO: 5, 66 in the sequence listing) and KAPA HiFi HotStart ReadyMix (KAPA BIOSYSTEMS), using the DNA encoding the full-length EndoS2 as a template. After amplification and synthesis, the amplified DNA was cloned into the EcoRV site of pBluescript II SK+ vector (Stratagene). From this, a DNA fragment cut out with BamHI and XhoI was inserted into the BamHI-XhoI site of pGEX-6P-1 (GE Healthcare) to obtain an EndoS2Δ(548-843) expression vector.

The DNA encoding EndoS2Δ(386-843) was subjected to PCR using primers NG1370+, NG1558− (SEQ ID NOs: 5 and 67 in Sequence Listing) and KAPA HiFi HotStart ReadyMix (KAPA BIOSYSTEMS) using the DNA encoding the full-length EndoS2 as a template. After amplification and synthesis, the amplified DNA was cloned into the EcoRV site of pBluescript II SK+ vector (Stratagene). From this, a DNA fragment cut out with BamHI and XhoI was inserted into the BamHI-XhoI site of pGEX-6P-1 (GE Healthcare) to give an EndoS2Δ(386-843) expression vector.

In addition, a glycosynthase corresponding to each EndoS2 deletion mutant was also constructed. In this case, PCR amplification was carried out in the same manner as above using the DNA encoding EndoS2 D182Q as a template, and EndoS2 D182Q Δ(679-843), EndoS2 D182Q Δ(548-843), EndoS2 D182Q Δ(386-843). Each expression vector of was prepared.
Escherichia coli BL21(DE3) strain was transformed with each of the expression vectors thus prepared, and each deletion mutant was expressed and purified in the same manner as described above to prepare an enzyme solution.

Using the enzyme solution of each EndoS2 deletion mutant obtained above, each of human serum-derived IgG (hIgG), human plasma-derived α1-acid glycoprotein (α1-AGP), and bovine pancreas-derived ribonuclease B (RNase B) was used. Experiments for cutting glycoprotein sugar chains were performed. As the enzyme reaction solution, 1 μg of the substrate glycoprotein and 0.2 μg of the EndoS2 deletion mutant were added to 10 μL of 50 mM citrate buffer (pH 6.0), and the mixture was reacted at 30° C. for 20 hours. As a control, a reaction with Peptide:N-Glycosidase F (PNGase F, Takara Bio) was also performed. Then, the enzyme reaction was stopped by adding the sample buffer for SDS-PAGE. This sample solution was boiled and then subjected to SDS-PAGE. After the electrophoresis, it was stained with Quick-CBB (Wako Pure Chemical Industries).

The result is shown in FIG. Among the EndoS2 deletion mutants, EndoS2Δ(679-843) and EndoS2Δ(548-843) had the activity of cleaving the hIgG and α1-AGP sugar chains, as in the case of wild-type EndoS2. The band indicated by the arrow is the substrate protein in which the sugar chain has been cleaved. In the case of α1-AGP, digestion with EndoS2, EndoS2Δ(679-843), and EndoS2Δ(548-843) resulted in only partial cleavage of the sugar chain, as compared with digestion with PNGase F. Moreover, unlike EndoS2, EndoS did not cleave the α1-AGP sugar chain. Since EndoS2 Δ(386-843) did not cleave the sugar chains of hIgG and α1-AGP, EndoS2 requires the entire LRR region or a part of the region to exert the sugar chain-cleaving activity. Can be said. Note that these deletion mutants did not cleave the sugar chain of RNase B, which is not a substrate of EndoS2 originally.

Next, the transglycosylation activity of the deletion mutant of EndoS2 D182Q was examined. The enzyme reaction solution for the glycosyl transfer reaction was 2.5 μL of 50 mM Tris-HCl (pH 7.0) containing 10 μg of HerC acceptor, 6 mM of A2-oxazoline, and 0.2 μg of a deletion mutant of EndoS2 D182Q, respectively. It was prepared to contain. These were reacted at 30° C. and sampled over time. The enzymatic reaction was stopped by adding the sample buffer for SDS-PAGE. This sample solution was boiled and then subjected to SDS-PAGE. After completion of electrophoresis, the cells were stained with Quick-CBB (Wako Pure Chemical Industries), and the presence or absence of transfer products was confirmed by focusing on the H chain of the antibody.

FIG. 14 shows the results of measuring the transglycosylation activity of each deletion mutant of EndoS2 D182Q. The number on the horizontal axis in FIG. 14 indicates the time from the start of the enzymatic reaction (unit: hour). The band in FIG. 14 is the antibody H chain, and the band seen on the column of EndoS2 D182Q, which shifts to the high molecular side with the elapse of the reaction time, is the antibody H chain to which the sugar chain has been transferred. Also in each column of EndoS2 D182Q Δ(679-843), EndoS2 D182Q Δ(548-843), and EndoS2 D182Q Δ(386-843), a small amount of the band shifted to the polymer side was detected as the reaction time passed. .. However, even in the column without enzyme (No enzyme), a small amount of the band shifted to the polymer side was detected after 3 and 5 hours. It is considered that this is because A2-oxazoline was nonspecifically bound to the antibody H chain. At the present time, is the band shifted to the polymer side detected in each column of EndoS2 D182Q Δ(679-843), EndoS2 D182Q Δ(548-843) and EndoS2 D182Q Δ(386-843) a transfer product? Although it is unclear whether it is a non-specific reaction product, it can be said that when EndoS2 D182Q is deleted domain by domain from the C-terminal side, its transglycosylation activity is remarkably reduced. This is something to be careful of in constructing a deletion mutant type glycosynthase.

As described above, the glycosynthase of the present invention was shown to be an enzyme exhibiting a higher transglycosylation activity in a wider pH range than the conventional glycosynthase EndoS D233Q mutant. Regarding the substrate specificity, it also had a high mannose-type sugar chain transfer activity that was not found in the EndoS D233Q mutant. Since the transglycosylation reaction was efficiently confirmed when the oxazoline sugar chain was used as the donor substrate in the examples, the same effect can be obtained when various sugar chains other than the oxazoline sugar chain are used as the donor substrate. When such glycosynthase is used for sugar chain remodeling of sugar chains of glycoproteins, the sugar chain transfer reaction using various sugar chains including an oxazoline sugar chain as a donor substrate proceeds efficiently because of high activity.

INDUSTRIAL APPLICABILITY The glycosynthase of the present invention is useful for producing a glycoprotein to which a sugar chain having a uniform structure is bound from a glycoprotein such as an antibody to which a sugar chain having a heterogeneous structure is bound. There is a possibility that the function and stability of proteins will be improved and the quality will be maintained. In addition, it is useful for producing a glycoprotein that exhibits high functionality by having a specific sugar chain structure.

Claims (8)

A glycosynthase having any of the following amino acid sequences.
(1) the amino acid sequence described in SEQ ID NO: 1 or 3 of the sequence listing; or (2) the deletion, substitution and/or deletion of 1 to 20 amino acids in the amino acid sequence described in SEQ ID NO: 1 or 3 of the sequence listing. Alternatively, an amino acid sequence having an addition and an amino acid sequence having a glycosynthase activity that can be used for remodeling of an antibody sugar chain: A glycosynthase gene, which encodes the amino acid sequence of the glycosynthase according to claim 1 . The glycosynthase gene according to claim 2 , which has any of the following nucleotide sequences.
(1) a nucleotide sequence specified by nucleotide numbers 1 to 2532 in the nucleotide sequence set forth in SEQ ID NO: 2 in the sequence listing;
(2) has a base sequence having deletions, substitutions and/or additions of 1 to several bases in the base sequence specified by base numbers 1 to 2532 in the base sequence shown in SEQ ID NO: 2 of the sequence listing , A nucleotide sequence encoding a protein having glycosynthase activity:
(3) the nucleotide sequence specified by nucleotide numbers 1 to 2532 in the nucleotide sequence set forth in SEQ ID NO: 4 in the sequence listing;
(4) having a base sequence having deletions, substitutions and/or additions of 1 to several bases in the base sequence specified by base numbers 1 to 2532 in the base sequence described in SEQ ID NO: 4 of the sequence listing , A nucleotide sequence encoding a protein having glycosynthase activity: A recombinant vector comprising the glycosynthase gene according to claim 2 or 3 . The recombinant vector according to claim 4 , which is an expression vector. A transformant transformed with the recombinant vector according to claim 4 or 5 . A method for producing glycosynthase, which comprises culturing the transformant according to claim 6 and collecting the glycosynthase from the culture. Using the glycosynthase according to claim 1, only N-acetylglucosamine having α1,6-bonded fucose as a sugar chain component to one or both H chains is “a sugar chain addition site present in the Fc region of an antibody”. One or two core fucose is added, which includes a reaction step in which the antibody bound to is the acceptor substrate, the high-mannose type sugar chain is the donor substrate, and the high-mannose type sugar chain derived from the donor substrate is transferred to the acceptor substrate. Having a high mannose type sugar chain described above.