JPH1080276A - Control of bisect saccharide chain - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain a medicine which can control the proportion of the bisect saccharide chains in the N-bond type saccharide chains of glycoprotein and sustain the medicinal effect by controlling the relative activity ratio of β 1-4- galactose transferase to N-acetylglucosamine transferase III in the cells. SOLUTION: When glycoprotein having N-bonding type saccharide chain is intracellularly produced, the intracellular expression of β 1-4 galactose transferase is controlled to control the relative activity ratio of the β 1-4 galactose transferase to the N-acetylglucosamine transferase III and adjust the proportion of the bisect saccharide chain in the N-bonding type saccharide chain of glycoprotein and control the hydrophilic and hydrophobic properties of the glycoprotein saccharide chain. The cells described above is cultured to obtain the glycoprotein bearing N-bonding type saccharide chain.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a technique for producing a glycoprotein. Specifically, asparagine-linked (hereinafter referred to as N) having the sugar chain structural unit shown in FIG. 2 (FIG. 2) in the sugar chain structure
-Abbreviated as linked type) sugar chain (hereinafter referred to as bisect sugar chain)
When producing a glycoprotein that binds to β1-4 galactosyltransferase (EC 2.4.1.3) in cells,
8/90) (hereinafter abbreviated as GalT) to regulate the expression level of bisect sugar chains.

[0002]

2. Description of the Related Art In recent years, with the development of genetic engineering or cell engineering, it has become possible to obtain a large amount of glycoprotein which originally exists only in a trace amount in a living body. Along with this, a sugar chain portion of a glycoprotein, for example, a sugar chain linked to an asparagine residue (hereinafter referred to as an N-linked sugar chain) has a physicochemically and biologically important function for a living body. (Varki, A., Glycobiology, Vol. 3, p. 97, 1993; Makoto Takeuchi et al., Bioscience and Industry, Vol. 47, N.
o.1, 1989). From this, it is expected that providing a protein having a desired sugar chain structure to a protein used as a pharmaceutical has great industrial value.

[0003] Unlike in the case of DNA or protein, the formation of sugar chains (glycosylation) in cells is not dependent on a template, but is dependent on a specific monosaccharide and a plurality of enzymes (sugar hydrolase and glycosidase) specific to a binding mode. Post-transla by glycosyltransferase)
nation modification (post-translational modification) (Kornfeld,
R. and Kornfeld, S., Ann. Rev. Biochem., Vol. 54, p. 63
1, 1985). For this reason, the sugar chain provided to a specific sugar chain-binding site of the glycoprotein is not always in a uniform state.
Further, since there is no sugar chain template, it is impossible to modify the structure by directly modifying the template used in a technique such as genetic engineering.

[0004] N-linked sugar chains are composed of monosaccharides such as N-acetyl-D-glucosamine, D-mannose, D-galactose, fucose and sialic acid. The N-linked sugar chain includes a branched pentasaccharide structure (hereinafter, referred to as a trimannosyl core) shown in FIG. 1 (FIG. 1) as a common core structure. The α1-6-branched mannose and α1-3-branched mannose of the trimannosyl core are usually further linked to a side chain, and N-acetyl-D-glucosamine (hereinafter abbreviated as GlcNAc) is attached to the branched mannose. )
Are extended to sugar chains having a different number of side chains, such as a double chain or a triple chain, by bonding. The double-chain sugar chain is called a biantennary sugar chain. Further, apart from the diversity of the side chains, as shown in FIG. 2 (FIG. 2), GlcNA linked to the mannose at the center of the trimannosyl core by β1-4 bond.
The presence of c (referred to as bisecting GlcNAc) may or may not be present, and an N-linked sugar chain containing a hexasaccharide structure to which bisecting GlcNAc is added is called a bisect sugar chain. N-linked sugar chains to which bisecting GlcNAc is not added are called non-bisect sugar chains.

In recent years, it has been revealed that changes in the sugar chain structure of glycoproteins are due to changes in the activity of glycosyltransferases. There are many attempts to modify this. Regarding the control of bisect sugar chains, attempts have been made to increase the expression of bisect sugar chains or to suppress the production of bisect sugar chains by genetic engineering techniques. Examples of increased production of bisect sugar chains include N-acetylglucosamine transferase III (EC 2.4.1.144) involved in the production of sugar chains having bisect sugar chains.
By cloning a glycosyltransferase gene called GnTIII (hereinafter abbreviated as GnTIII) and expressing the enzyme in large amounts in liver cancer cells by using the isolated gene,
There is a report that directly increased the production of bisect sugar chains in the protein of liver cancer cells (Miyoshi, E., Ihara, Y.,
Hayashi, N., Fusamoto, H., Kamada, T. and Taniguchi,
N., J. Biol. Chem., Vol.270, p.28311-28315, 1995).
On the other hand, as an example of suppressing GnTIII activity,
The GnTIII gene is disrupted by the
Report of the production of knockout mice with suppressed nTIII activity (Stanley, P., Raju, TSand Bhaumik, M., XIIIt)
h Int.Symp. on Glycoconjugates, Satellite Symp., "
Hillfest '95 ", August 16-19, Orcas Island, Wash.,
Abs. P. 14, 1995). However, there is no example of controlling GnTIII activity in a cultured cell that enables production of a useful protein, and thereby controlling a bisect sugar chain in a protein sugar chain. The properties of GalT and GnTIII have been studied in detail, and GalT and GnT
It is known that III shares a common sugar chain (agalactosyl biantennary sugar chain (FIG. 3)) as a substrate. During the reaction of both enzymes, GalT acts first to add galactose to both antenna ends and is converted into a galactosyl biantennary sugar chain (FIG. 4). GnTIII does not use the galactosyl biantennary sugar chain as a substrate. Cannot add bisecting GlcNAc (Kornfeld, S. et al.
in The Glycoconjugates vol.III, Part A, p.3-23, A
cademic Press, New York, 1982, Gleeson, PA and
Schachter, H., J. Biol. Chem., Vol. 258, 6162-6273,
1983). However, since the substrate specificity of such a glycosyltransferase is studied using sugar chains released from proteins,
It does not necessarily reflect the state in which the glycosyltransferase reacts to the actual protein. Glycosylation of proteins is performed in the Golgi apparatus of cells. When glycosyltransferase acts on proteins, the steric structure of glycoproteins affects the reactivity of glycosyltransferases. The actual structure of the chain does not always match the predicted sugar chain structure based on the properties of glycosyltransferases that have been revealed in studies using free sugar chains (Shigeru Fujii, Protein Nucleic Acid Enzyme, vol. .37, 1992). Therefore, a sugar chain structure control technique that can be used in actual protein production has not yet been established. Regarding the relationship between GalT activity and bisect sugar chains, study of IgG sugar chain structure and glycosyltransferase activity (Nishiura, T., et al, Cancer Re.
search vol.50, p.5345-5350, 1990).
Due to the influence of the reactivity based on the substrate specificity of T and GnTIII, the production of bisect sugar chains and the addition of galactose
There have been reports suggesting that the activity is influenced by the balance of GalT and GnTIII enzyme activities in cells. It is a comparison of the relationship between the sugar chain structures of IgG in the serum of each patient, and merely describes the physiological properties exhibited by myeloma patients. This report does not directly analyze and compare the enzymatic activity of cloned cells with the sugar chain structure of the glycoprotein produced by the cells.
It can be said that the I activity does not show direct evidence of affecting the sugar chain structure that binds to a protein actually produced by the same cell, and does not provide any knowledge on the control of the sugar chain structure. As described above, considering the state of the Golgi apparatus, which is the site of the actual modification of the sugar chain added to the protein, the actual sugar chain structure of the protein is not necessarily determined by the glycosyltransferase activity measured in vitro. Can't predict
Unless cloned cells can be used to change the glycosyltransferase activity of the cells and control the sugar chains of the protein actually produced, no technique for controlling the sugar chain structure has been established. As described above, it is determined whether the activity of GalT, an enzyme that is not directly involved in the addition reaction of bisecting GlcNAc, directly affects the amount of bisect sugar chains generated in the actual sugar chain structure added to the protein. I'm not sure. Furthermore, by actively regulating the expression level of GalT,
An example of controlling the composition of a bisect sugar chain in a sugar chain that binds to a protein actually produced has not yet been known.

[0006]

An object of the present invention is to provide a method for producing a desired bisect sugar chain or a glycoprotein having the same. Specifically, an object of the present invention is to provide a method for controlling the expression amount of GalT in a cell, thereby controlling the ratio of bisect sugar chains in N-linked sugar chains of glycoprotein.

[0007]

Means for Solving the Problems The present inventors have proposed GalT
In the process of studying the regulation of the expression level of glycerol, we discovered a phenomenon that was not considered by common sense in the field of glycotechnology research. That is, it was found that there was a correlation between the ratio of the expression levels of GalT and GnTIII and the production amount of bisect sugar chains actually added to the protein to be produced. We succeeded in developing a simple and innovative method.

[0008] When the GalT activity in the cell is increased, the addition of galactose to the agalactosyl biantennary sugar chain is promoted even when reacting with the actual protein, and the conversion to the galactosyl biantennary sugar chain becomes dominant. , GnT
III is a bisecting Glc for depriving substrate
The NAc addition reaction is suppressed. That is, in the composition of the sugar chain added to the produced protein, the ratio of the bisect sugar chain decreases. Conversely, when the intracellular GalT activity is reduced, the reaction of GnTIII, that is, the addition of bisecting GlcNAc to the agalactosyl biantennary sugar chain becomes dominant even when reacting with the actual protein. That is, in the composition of the sugar chain added to the produced protein, the ratio of the bisect sugar chain increases.
In this way, by adjusting the balance between the intracellular enzyme activities of GalT and GnTIII, it is possible to regulate the amount of bisecting GlcNAc added to the agalactosyl biantennary sugar chain in the actually produced protein. The ratio of sugar chains can be adjusted.

That is, the present invention controls the expression level of β1-4 galactosyltransferase in a cell when a glycoprotein having an N-linked sugar chain is produced in the cell,
By controlling the relative ratio of β1-4 galactosyltransferase activity to N-acetylglucosamine transferase III activity in the cells, the proportion of bisect sugar chains in the N-linked sugar chains of the glycoprotein is adjusted. It provides a method.

[0010]

BEST MODE FOR CARRYING OUT THE INVENTION The cells or cell lines that produce the desired glycoprotein that can be used in the present invention are of eukaryotic origin and particularly those that can produce the glycoprotein. Without limitation, animals, plants, cells of eukaryotic microorganisms and the like can be mentioned. In animals, both adherent cells and planktonic cells can be used. Examples of such cells include animal cells that produce and accumulate glycoproteins inside cells, and animal cells that secrete and produce glycoproteins extracellularly.
Specific examples include antibody-producing B cells, tissue plasminogen activator (hereinafter abbreviated as t-PA) -producing cells, and the like. Anti-Pseudomonas aeruginosa IgM antibody-producing B cell line MP
-5045 (this cell line has been deposited with the Ministry of International Trade and Industry at the National Institute of Bioscience and Human Technology, Accession No. FERM P-14976) and the like.

The glycoprotein according to the present invention refers to a protein having a sugar chain bound to at least one of the asparagine residues to which the sugar chain in the amino acid sequence of the protein may bind. Examples of such include immunoglobulin (Ig) M and G, t-PA and the like.
For example, as shown in FIG. 5 (FIG. 5), a monomer of glycoprotein IgM has a binding site of an N-linked sugar chain in the constant region of the heavy chain.
Location (hereinafter, 167, 330, 3
Sugar chains linked to asparagine, amino acids 93, 400 and 560, were N167, N330 and N39, respectively.
3, N400 and N560. ) Have (Sh
imizu, A., Patnam, FW and Paul, C., Nature NeW Biolo
gy, 231, p. 73, 1971).

The N-linked sugar chain according to the present invention is a sugar chain linked to an asparagine residue in the amino acid sequence of a protein portion constituting a glycoprotein, and comprises N-acetyl-glycan.
It is composed of monosaccharides such as D-glucosamine, D-mannose, D-galactose, fucose and sialic acid. A bisect sugar chain is a GlcNAc linked to a mannose at the center of a trimannosyl core by a β1-4 bond.
Is an N-linked sugar chain containing a hexasaccharide structure in the form of an N-linked sugar chain.

As a method of controlling the expression level of the GalT gene, for example, a general gene recombination technique can be used. As an expression suppression method, a gene targeting method,
An antisense method and the like can be mentioned. As a method for increasing the expression level, a method of introducing a GalT gene into cells using a general gene recombination technique to achieve high expression or a method of controlling the expression of GalT using a homologous recombination technique, for example, A method of replacing a promoter site with a promoter having a higher function than before is included.

In the gene targeting method, the following operations are performed according to a general gene recombination technique,
The gene disruption of the GalT gene can be performed. That is, GalT exon 2 of a human cell as a host and a portion containing introns before and after GalT are cloned, and the DN of GalT exon 2 sandwiched between introns before and after is cloned.
The A fragment needs to be isolated. For this purpose, a phage clone containing the desired GalT exon 2 can be identified from a commercially available human chromosome gene phage library by plaque hybridization using an appropriate probe, and the desired gene can be isolated. I can do it. As a probe to be used, a synthetic DNA which is a part of a known sequence encoding a GalT structural gene may be used, or a gene fragment prepared by a PCR method using this as a primer may be used. The foreign DNA to be inserted into exon 2 is not particularly limited as long as the inserted DNA results in the disruption of the function of the GalT protein after translation. However, the inserted DNA may contain a stop codon or a reading frame for codon. And at the same time, has a function of a selection marker such as neomycin (G418) or hygromycin, thereby providing a means for selecting transformed cells and facilitating isolation of the objective transformed cells. Are particularly preferred. Also, GalT
Applying selection pressure by ganciclovir by ligating a virus-derived thymidine kinase gene to both or one of the outer DNA fragments capable of depriving GalT expression function by being inserted into exon 2 Screening can be performed more efficiently. The gene fragment constructed in this way is
By utilizing the homologous recombination property of the cell and replacing it with the GalT gene of the original normal host, Ga
It is possible to obtain a cell line in which the IT expression function is stopped or suppressed. Specifically, by introducing the gene fragment for targeting constructed by the above-described method into cells and applying selective pressure with neomycin, ganciclovir, or the like, homologous recombination occurs from the transformed cell lines that have grown. It is possible to select cells in which expression of the GalT gene on the chromosome has stopped. As a specific method for introducing the DNA fragment into cells, in addition to the conventional method using calcium phosphate or the like, a method using a liposome-like gene transfer reagent, or a method of applying an electric pulse to cells (electroporation) ), A method of physically introducing the gene using a gene gun may be used.

In the antisense method, mT of the GalT gene
A gene region (antisense strand) having a sequence complementary to the RNA sequence (sense strand) can be supplied from inside or outside the cell to specifically suppress GalT expression.
As a method of supplying an antisense strand, an antisense RN sequence having an arbitrary length of an inverted sequence (antisense sequence) incorporated into a vector and endogenously expressed as RNA in cells is used.
A method and an "antisense DNA method" in which synthetic DNA having a length of 15 to 30 bases is directly added to a cell culture solution can be used.

In the method of introducing the GalT gene into cells using gene recombination technology to achieve high expression, generally known gene transfer technology can be applied. It is preferable that the vector has a function of a selection marker such as neomycin (G418) or hygromycin, thereby providing a means for selecting transformed cells and facilitating isolation of the desired transformed cells. Promoters for expressing genes include Simian virus 40 early (SV40
E) gene, mouse phosphoglycerate kinase gene, human β-actin gene promoter, cytomegalovirus major IE (CMV) promoter, rous sarcoma virus (RSV) long terminal repeat (LT)
R) Promoters and the like are used. As a method for introducing the prepared expression vector into cells, in addition to the conventional method using calcium phosphate or the like, a method using a liposome-like gene transfer reagent, or a method of applying an electric pulse to cells (electroporation) ), A method of physically introducing the gene using a gene gun may be used.

The cells or cell lines obtained by the method of the present invention can be cultured using various known methods, and there is no particular limitation as long as they do not inhibit the growth of the cell lines and the production of glycoproteins. . For example, suspension culture in a tank, adhesion culture in which cells are adhered to the surface of styrene microbeads or the inner wall of a roller bottle, stationary culture using a flask, and the like can be appropriately selected according to the cell strain.
The culturing time may be sufficient until the cells are sufficiently grown and the glycoprotein is sufficiently produced when culturing by the batch method.
Usually, it is about one week to six months. When performing continuous culture while aseptically replacing a part of the culture medium during the culture, the culture time is about one week to six months. In culturing, after inoculating a cell line that produces a glycoprotein, the cell line is cultured while maintaining an appropriate temperature, aeration, and pH of the medium. As a medium that can be used for culturing the cell line obtained by the method of the present invention, a medium obtained by adding an additive such as serum to a basic medium can be used. As the basic medium, a commercially available medium for cell culture can be used. For example, Eagle's minimum essential medium, RPMI-1640
A medium, a ham F12 medium, a modified Dulbecco's Eagle medium, or the like may be used alone or in an appropriate mixture. In addition, the culture of the cell line can be performed in two stages of increasing the number of cells and producing the glycoprotein, and two different types of culture media can be used. In this case, as the growth medium, for example, a nutrient medium obtained by adding fetal bovine serum (FCS) at a concentration of 1 to 30% to the above basic medium, and as the production medium, fetal bovine serum (FCS)
The above-mentioned basal medium containing no can be used.

The glycoprotein can be recovered from the culture obtained as described above by a usual method. For example, when the glycoprotein is secreted and produced extracellularly, the culture solution can be recovered by replacing the culture solution in a timely manner. When the glycoprotein is produced and accumulated in the cell, the culture solution is filtered or centrifuged. After collecting the cells by the method described above, the cells are disrupted to obtain a glycoprotein. Purification of the recovered glycoprotein may be performed by various known purification methods using, for example, ion exchange, biological affinity, adsorption or hydrophobicity, hydrophilicity, molecular size, ultrafiltration, etc., depending on the target glycoprotein. Is achieved in.

In order to further obtain a sugar chain from the glycoprotein obtained as described above, various known methods can be used. Examples thereof include a known hydrazinolysis method and a method in which a sugar chain and a protein portion are cleaved using various glycopeptidases. When only a sugar chain bound to a specific site of a glycoprotein is required, the glycoprotein is fragmented in advance using a protease such as trypsin, lysyl endopeptidase, or V8 protease, and purification of the glycoprotein is performed. After obtaining a peptide fragment containing the target sugar chain by a method according to the above, the sugar chain and the peptide may be cleaved by the same method as described above. In order to separate the sugar chain from the peptide or amino acid after the sugar chain is cut out from the glycoprotein or glycopeptide by the above-mentioned method, various purification methods utilizing the molecular weight, hydrophobicity or hydrophilicity can be performed. . A sugar-specific coloring method such as the orcinol-sulfuric acid method can be used for the detection of the fraction containing a sugar.

As a method for quantifying the activity of GalT and GnTIII, a method having high detection sensitivity, high quantification and high reproducibility is desirable. The activity can be determined by adding a cell lysate as a measurement sample to a reaction solution containing a receptor sugar chain and a sugar nucleotide serving as a substrate, and quantifying the remaining receptor sugar chain or generated sugar chain using HPLC. For the measurement conditions, a known method can be used. At this time, it is desirable that the system considers the presence of substances that degrade the acceptor sugar chains, products, and substrates that hinder the quantitativeness of the reaction.

In order to analyze and confirm the structure of the sugar chain obtained from the glycoprotein obtained in the present invention, various known methods can be appropriately used. However, since the detection sensitivity is increased,
It is desirable that the obtained sugar chain is fluorescently labeled with 2-aminopyridine and analyzed by high performance liquid chromatography (HPLC) as described later. HPL
In the separation and analysis of sugar chains using C, any known method can be used based on the charge, molecular weight, hydrophilicity, hydrophobicity, affinity for various lectins, etc. of the sugar chains. . As an example, while separating and quantifying sugar chains having different numbers of sialic acids by using an anion exchange column according to the difference in their charges, the neutral sugar chains from which sialic acid has been eliminated are separated by a reversed-phase column, The sugar chain of each peak component obtained by the reverse phase HPLC is further analyzed using a normal phase column and a reverse phase column, and the sugar chain structure of each component is determined by Takahashi et al.
Dimensional sugar chain map (Reiko Takahashi, Glycoprotein Glycoprotein Research Method, Biochemical Experimental Method 23, Gakkai Shuppan Center, 1989, and Reiko Takahashi et al., Extra number of protein nucleic acid enzymes, Kyoritsu Shuppan, 37 , p.2117, 1992
Year) fingerprinting. At this time, for the removal of sialic acid, for example, a method of performing acid hydrolysis under acidic conditions or a method of using sialidase derived from microorganisms can be used.

[0022]

EXAMPLES The following describes an anti-Pseudomonas aeruginosa IgM antibody producing B cell line
Glycoprotein IgM produced by culture of P-5045
An example is given of a method in which the sugar chain is modified. However, the present invention is not limited by these. FIG. 6 shows the relationship between the ratio of GalT activity / GnTIII activity and the ratio of non-bisect sugar chains / bisect sugar chains in the example of IgM sugar chain modification using MP-5045 strain.
(FIG. 6).

Test Example 1 1. Culture of anti-Pseudomonas aeruginosa IgM antibody-producing B cell line MP-5045 and production of IgM 3 × 10 ⁵ cells of anti-Pseudomonas aeruginosa IgM antibody-producing B cell line MP-5045 (hereinafter abbreviated as MP-5045 strain)
At a cell density of 20% / ml in a modified NYSF medium containing 20% FCS (described in JP-A-7-23780), and using a T-type flask at 37 ° C. and 5% CO _2. Stationary culture was performed. The medium was changed every three days, and each time the cells were subcultured to a cell density of 3 × 10 ⁵ cells / ml. The culture scale was expanded as the cells grew. The grown cells were seeded in a serum-free medium (modified NYSF) at a cell density of 1 × 10 ⁶ cells / ml in a T-type flask, and then inoculated in a CO ₂ incubator at 37 ° C.
The culture was allowed to stand for 3 days at 5 ° C. and 5% CO ₂ . A part of the culture solution was sampled and used for measuring the enzyme activity. In addition, the culture solution was centrifuged to remove the precipitated cells, and the supernatant was recovered and used for analyzing the sugar chain structure of the produced IgM.

2. Measurement of Intracellular GalT and GnTIII Enzyme Activity of MP-5045 Strain After counting the number of cells in the sampled culture solution with trypan blue, the cells were collected by centrifugation, washed twice with PBS, and counted as 2.0 cells. × 10 ⁵ cells were placed in a reaction vessel used for the GalT reaction in a volume of 5 μl of buffer (10 mM HE).
PES (pH 7.2), 1% Triton X-10
0). Using a bath-type ultrasonic crusher, crushing was performed three times in ice water at 30 second intervals for 30 seconds. 5
Take 1 μl of the cell lysate into a reaction vessel,
20 μl of the reaction solution for activity measurement (12.5 m
M receptor sugar chain: pyridylaminated agalactosyl biantennary sugar chain (type in which fucose is not bound to reducing terminal GlcNAc), 12.5 mM sugar nucleotide, 10
mM HEPES (pH 7.2), 12.5 mM Mn
Cl ₂ , 41.25 mM NaCl, 3.75 mM K
(Cl, 7.0 mM γ-lactone) was added to start the reaction, and the enzyme reaction was carried out at 37 ° C. As sugar nucleotides, UDP-
Galactose (Gal) was used for the measurement of GnTIII activity.
DP-N-acetylglucosamine (GlcNAc) was used. 5 μl was sampled with time from the reaction solution, and 50 μl of a reaction stop solution (10 mM HEPES (pH
7.2), 50 mM EDTA), and the reaction was stopped by boiling for 3 minutes. Shim-Pack CLC
The residual ratio of the acceptor sugar chain was determined over time by HPLC using an -ODS column (inner diameter 6 mm x length 150 mm, manufactured by Shimadzu Corporation). The logarithmic value of the receptor sugar chain remaining rate was plotted against the reaction time, and the obtained slope was converted into a value per unit cell number (2.0 × 10 ⁵ cells) to obtain an index of the reaction rate. The slope and Gn obtained by measuring the activity of GalT
The ratio of the slope obtained in the activity measurement of TIII was determined, and
The relative activity ratio of T and GnTIII was calculated. As a result, as shown in Table 1 (Table 1), the relative ratio of GalT activity to GnTIII activity in the MP-5045 strain was 1.3. This means that in the MP-5045 strain, the reactivity of GalT to the same substrate sugar chain (agalactosyl biantennary) is 1.3 times higher than that of GnTIII.

3. Monochrome produced by MP5045 strain
Analysis of sugar chain structure of null IgM 3-1. Purification of IgM After filtration of the culture supernatant of the MP5045 strain to remove insolubles, the filtrate was subjected to affinity chromatography using an anti-human IgM antibody-binding column (column capacity: 10 ml) (manufactured by NGK). Specifically, after passing the culture supernatant through a column to adsorb IgM, P
Washing was performed with BS (pH 7.6), and subsequently with washing solution A (manufactured by NGK). Then, 0.2M glycine-HC
IgM was eluted with 1 buffer. The protein was detected by measuring the absorbance at 280 nm of the eluate from the column. The eluted IgM fraction contains 1M Tris-HCl
After neutralization by adding a small amount of (pH 8.3) + 0.15M NaCl, the resultant was freeze-dried to obtain a purified IgM preparation.

3-2. Purification of Sugar Chain As shown in FIG. 5, five sugar chains of IgM are bound on the μ chain, three sugar chains on the amino terminal side are of a complex type, and two sugar chains on the carboxyl terminal side. The chains are of high mannose type (Shimizu, A., Patnam, FW and Paul, C., Nature NeW
Biology, 231, p. 73, 1971). Although the bisect sugar chain is generated only in the complex type portion, five sugar chains were cut out at a time and purified as a mixture. To reduce the SS bond of IgM, purified IgM was added to 10 M urea, 20 mM dithiothreitol, 150 mM NaCl, 200 mM
Dissolve in Tris-HCl buffer (pH 8.2)
Reduced at 7 ° C. overnight. Further, iodoacetic acid was added so as to have a concentration three times that of DTT, and the mixture was treated at 37 ° C. for 90 minutes to modify and protect the SH group. Then, 50 mM Tri
It was dialyzed against s-HCl (pH 8.2) and 2 mM CaCl ₂ to remove urea and iodoacetic acid. Next, thermolysin (manufactured by MERCK) was added to the solution of the peptide chain modified and reduced.
Was added and enzyme digestion was performed at 60 ° C. for 4 hours. The reaction was stopped by adding trifluoroacetic acid to 0.1%, filtered with an insoluble matter filter, and then 10 mM NH ₄
Sephadex G-25 using HCO ₃ as buffer
(Pharmacia) to recover the glycopeptide fraction by gel filtration. The orcinol-sulfuric acid method was used for the detection of glycopeptides. To cut sugar chains from glycopeptides, 1
In a 00 mM citrate-phosphate buffer (pH 5.0), almond-derived glycopeptidase A (Seikagaku Corporation) was added and reacted at 37 ° C. overnight. Separate the reaction solution
-It was applied to a Pak C18 cartridge (Waters), the sugar chain was eluted with a 0.1% TFA, 5% acetonitrile solution, and lyophilized to obtain a purified sugar chain sample. The obtained sugar chain is obtained by a known method using 2-aminopyridine (Ha
se, S. et al., J. Biochem. (Tokyo), 100 , p. 1, 1986), and converted to pyridylamino (hereinafter abbreviated as PA), followed by fluorescent labeling. Labeled sugar chains (hereinafter abbreviated as PA sugar chains)
Is Sephadex G-15 (Pharmacia)
Purified using gel filtration chromatography using as a carrier, and used for the following analysis.

3-3. Sugar chain structure analysis The PA-linked sugar chain is digested with Arthrobacter ureafaciens- derived sialidase (manufactured by Nacalai Tesque, Inc.), and the asialo (from which sialic acid has been eliminated) fraction is anion exchange column Mo.
It was fractionated by HPLC using noQ HR5 / 5 (inner diameter 5 mm x length 50 mm, manufactured by Pharmacia). The sugar chain from which sialic acid has been removed (asialo sugar chain) is analyzed using HPLC.
Reversed phase column (Shim-pack CLC-ODS, inner diameter 6 x length 150 mm, manufactured by Shimadzu Corporation) and amide adsorption column (TSKgel Amide-80, inner diameter 4.6)
mm × 250 mm length, manufactured by Tosoh Corporation), and the structure of each IgM sugar chain was estimated and quantified from these patterns. The sugar chain structure of each component is expressed as a glucose unit by calculating a relative glucose polymerization degree relative to an elution position of a PA-containing glucose oligomer (polymerization degree: 3 to 22 or 3 to 15, manufactured by Takara Shuzo). Structured by a three-dimensional sugar chain map (Reiko Takahashi, Glycoprotein Glycoprotein Research Method, Biochemical Experimental Method 23, Gakkai Shuppan Center, 1989, and Reiko Takahashi et al., Extra number of protein nucleic acid enzymes, Kyoritsu Shuppan, 37 , p.2117, 1992) Was estimated. In addition, in order to confirm each structure, the sugar chains of each component were converted to various exoglycosidases (α-L-fucosidase from bovine kidney (Boehringer Mannh).
eim), and β-D-galactosidase or β-N-acetylhexosaminidase (manufactured by Seikagaku Corporation) derived from common bean, and then the behavior of the digested product was similarly subjected to a two-dimensional sugar chain map. Analyzed above. In the case of a commercially available standard PA-linked sugar chain having the same structure as that of the sample sugar chain, by co-punching both with HPLC, a single peak is obtained.
Confirmed that The results of quantifying the sugar chain structure of IgM produced by the MP-5045 strain by HPLC are shown in Table 1 (Table 1). The abundance ratio of the biantennary glycan in the total composition of the glycans bound to 5 places on the μ chain of IgM is 56%
Met. Among the biantennary sugar chains, the presence rate of the bisect-type biantennary sugar chain to which the bisecting GlcNAc was added (the rate of addition of the bisecting GlcNAc to the biantennary sugar chain) was 62%.

Embodiment 1 1. Preparation of GalT targeting vector pGEXN-Neo-TK General DNA manipulation methods are described in the following literature (Sambrook, J. et.a.
l., Molecular Cloning A Laboratory Manual.Second E
dition, Cold Spring Harbor Laboratory Press., (198
9)). pBluescript KS, pBluescript S
K, pMC1Neo polyA is from STRATAGENE, pHSV-106 is G
Each was purchased from IBCO BRL. pSTNeoB is JCRB
(Japanese Cancer Research Resources Bank). Cloning of the GalT gene from the chromosome was performed as follows. That is, using the plaque hybridization method, the Human Leukocyte Genomic Library (Cl
ontech, HL1111j, Lot; 3511, Vector; EMBL-3 / SP6 / T7
), A gene fragment containing exon 2 of the GalT chromosome gene was cloned. The probe labeled with Digoxigeninn-11-UTP has “Prob” at the top of FIG. 7 (FIG. 7).
Using the region shown as "e", it was prepared by the PCR method.
About 1 × 10 ⁶ phages were screened to obtain 2 positive clones. Of these, Lambda Gal-T-2
-2 was analyzed. When a restriction enzyme map of a gene fragment containing exon 2 of the inserted GalT chromosome gene was prepared, it was revealed that the structure was as shown in the upper part of FIG. 7 (FIG. 7). As shown in FIG. 7 (FIG. 7), pGEX1, pGENX
1 is prepared, the base sequence is analyzed, and GalT
The targeting vector pGEXN-Neo-TK was prepared according to the procedure shown in FIG. 8 (FIG. 8) and FIG. 9 (FIG. 9).

2. Culture of MP-5045 strain and MP-
GalT targeting vector pGEX for 5045 strains
Introduction of N-Neo-TK The culture of the MP-5045 strain was performed as follows. That is,
20% fetal calf serum (FCS) and 30
A modified NYSF medium containing mg / l kanamycin was used. The cell maintenance passage was performed in a CO ₂ incubator
In a static culture method using a 75T flask, at 37 ° C., 5% C
Performed under O ₂ conditions, every 3 days, cell density about 3 × 10 ⁵ / m
and replaced with fresh medium. Ga in which exon 2 of the GalT chromosome gene has been disrupted by the neomycin resistance gene
The IT targeting vector pGEXN-Neo-TK was transformed into 7.5 × 10 ⁸ MP using electroporation.
-5045 strain. Make a selection with G418,
About 1100 G418 resistant clones were obtained. pGEXN-Ne
o-TK has helpes simplex v for negative selection
The thymidine kinase (HSV-TK) gene from irus is connected. When such a vector is introduced into a cell and causes homologous recombination, as shown in FIG. 10 (FIG. 10), HSV-TK
The gene is not taken up. On the other hand, when heterologous recombination occurs and the vector is randomly integrated into the chromosome,
As shown in FIG. 1 (FIG. 11), it is highly probable that the HSV-TK gene is also incorporated. The guanine analog ganciclovir (G
ANC) is added to human thymidine kinase (TK).
The gene is unaffected because these drugs cannot be used as substrates for enzymatic reactions, but cells carrying the HSV-TK gene take up GANC during DNA synthesis, stop elongation of the DNA strand, and die. . This allows GANC
Cells that have become resistant are more likely to be homologous recombinants.
Based on such a theory, GANC selection is performed and 580
GANC resistant clones were obtained. Chromosomal DNA is prepared from each clone using a simple adjustment method, and PCR is performed.
Screening was performed.

3. Screening by PCR When pGEXN-Neo-TK shown in FIG. 12 (FIG. 12) undergoes homologous recombination with the GalT chromosomal gene shown in FIG. 13 (FIG. 13), the structure shown in FIG. 14 (FIG. 14) is obtained. It is considered to be. In this case, Ne on the neomycin resistance gene
o 5`primer and Gal T int 3`pr on GalT chromosome gene
If PCR is performed using imer, 11
An 80 bp band should be observed. As a result of performing PCR on 580 G418GANC-resistant clones, a 1180 bp band was observed for two clones. Each clone was prepared as Y6-17 strain, Y
The strain was -35. The Y6-17 strain has been accorded the accession number FERMP
-15743. Further, Y6-1
For 7 strains, the primer F01 of exon 2 of the GalT chromosome gene (FIG. 13 (FIG. 13)) and GalTint 3`prim
PCR was performed using er. 2210 bp band when homologous recombination occurs, 1065 bp for wild type
Band should be observed. As a result, 2210
A bp band was present, and a 1065 bp band was clearly observed. From this, it was inferred that the Y6-17 strain was a single knockout clone having one Galt gene that had undergone homologous recombination with the wild-type GalT gene. Next, in order to confirm this, analysis by Southern hybridization was performed.

4. Analysis of Single Knockout Clones by Southern Hybridization Amersham Hybond N + was used as the filter membrane. The probe was prepared by incorporating Digoxigenin-11-dUTP (Boehringer Mannheim) by PCR. Int F1 primer and Int R1 primer were used as primers (FIG. 13 (FIG. 13)). As shown in FIG. 15 (FIG. 15), the wild-type GalT chromosome gene was replaced with the restriction enzyme HindII.
I and XhoI digestion and GalT int 2-2 probe (FIG. 13 (FIG. 1)
The Southern hybridization using 3)) yielded a fragment of about 1.9 kb, and the GalT chromosomal gene that had undergone homologous recombination was treated with about 3.
A 0 kb fragment should be observed. The GalT chromosomal gene was isolated from the parent strains MP-5045 strain, heterologous recombinant, and homologous recombinant Y6-17, respectively.
GalT int 2 extracted using the TurboGEN Genomic DNA Isolation Kit (Invitrogen) and labeled with Digoxigenin
Southern hybridization was performed using -2 probe. As shown in FIG. 15 (FIG. 15), the parent strain MP-
In the 5045 strain and the non-homologous recombinant, only a fragment of about 1.9 kb was observed.
In the strain, both a fragment of about 1.9 kb and a fragment of about 3.0 kb could be observed. This revealed that the homologous recombinant Y6-17 strain was a single knockout clone in which only one GalT chromosome gene was disrupted.

5. Analysis of sugar chain structure of GalT single knockout clone Y6-17 strain IgM was produced by Y6-17 strain in the same manner as in Test Example 1. GalT and GnTIII in cells during IgM production
The activity was measured in the same manner as in Test Example 1, and the produced I
Purification and structural analysis of the gM sugar chain were performed in the same manner as in Test Example 1. As a result, the GalT enzyme activity of the Y6-17 strain was
The value was about 60% lower than that of the above-mentioned MP-5045 strain (Table 1 (Table 1)). Thereby, GnT of GalT
The relative ratio to III activity was reduced to 0.7. Also, I
The sugar chain composition of gM is shown in Table 1 (Table 1). IgM μ
In the total composition of the sugar chains linked to the five positions on the chain, the abundance ratio of the biantennary sugar chains was 50%, which was almost the same as the parent strain MP-5045. Among the biantennary sugar chains, the ratio of the bisect-type biantennary sugar chain to which bisecting GlcNAc was added was 74%, indicating that the parent strain MP-
It increased from 5045.

[0033]

[Table 1]

Embodiment 2 1. Construction of GalT expression vector pCXN2-GalT General DNA manipulation methods are described in the literature (Sambrook, J. et. Al., Mo.
lecular Cloning A Laboratory Manual.Second Editio
n, Cold Spring Harbor Laboratory Press., (1989))
Was performed according to GalT gene is cDNA derived from human
(Masri, KA, Appert, HE and Fukuda, MN, Biochem.
Biophys. Res. Commun., 157, p657-663, 1988) was used. PCXN2 (Niwa, H., Yamamura, K. and Mi
yazaki, J., Gene, 108, p.193-200, 1991) was used. The GalT expression vector-pCXN2-GalT was converted to Fig. 16 (Fig. 16).
It was created by the method shown in.

2. Culture of MP-5045 strain and MP-
Introduction of GalT expression vector pCXN2-GalT into 5045 strain The culture of MP-5045 strain uses a modified NYSF medium containing 20% fetal calf serum (FCS) as a medium,
In a CO ₂ incubator, static culture was performed using a T-type flask, and the cells were passaged every three days by replacing the cells with a fresh medium at a cell density of about 3 × 10 ⁵ / ml. The GalT expression vector pCXN2-GalT was prepared by electroporation.
Introduced into 2.3 × 10 ⁷ MP-5045 strains. G4
Selection was performed to obtain about 170 G418 resistant clones.

3. Screening of GalT-highly expressing strain The GalT activity of the obtained G418-resistant strain was measured in the same manner as in Test Example 1, and the Gal50 activity of the parent strain MP5045 was determined.
Cells showing GalT activity higher than T activity were selected. Among cells selected as GalT highly active strains, I
Since many did not produce gM, the amount of IgM produced was measured by ELISA, and cells in which IgM producing ability was not reduced were selected. GalT-20 and GalT-50 were obtained as high expression strains of GalT that produce IgM.

4. Analysis of Sugar Chain Structure of GalT Highly Expressing Strain GalT-20 and GalT-50 strains produced IgM in the same manner as in Test Example 1. Intracellular GalT and GnTIII activities during IgM production were measured in the same manner as in Test Example 1,
Further, purification and structural analysis of the produced IgM sugar chain were performed in the same manner as in Test Example 1. As a result, the amount of GalT enzyme activity of the GalT high expression strain increased several times as compared with the parent strain MP-5045 (Table 1 (Table 1)). Thereby, GalT
The relative ratio of activity to GnTIII activity was GalT-2
0 and 5.8 for GalT-50, and the parent strain MP-
It increased from 5045. In addition, the sugar chain composition of IgM is shown in (Table 1 (Table 1)). Of the total composition of the sugar chains binding to 5 places on the IgM μ chain, the abundance ratio of the biantennary sugar chain is as follows:
All the cell lines were about 55%, and there was almost no difference from the parent line MP-5045. Among the biantennary sugar chains, the ratio of the bisect-type biantennary sugar chain to which bisecting GlcNAc is added is GalT-20.
13%, GalT-50 24%, parent strain MP-50
It was significantly reduced from 45.

[0038]

According to the present invention, the expression level of GalT in cells is controlled, and the relative ratio of β1-4 galactose transferase activity to N-acetylglucosamine transferase III activity in cells is controlled. Thus, a desired bisect sugar chain or a glycoprotein having the same can be produced. Sugar chains are involved in various life phenomena, such as generation, recognition and adhesion between cells, cell differentiation, and canceration.By artificially controlling the sugar chain structure, the mechanism of such life phenomena is understood. Can be elucidated. In addition, if a desired sugar chain structure is provided to a glycoprotein used as a pharmaceutical, it is possible to conduct research for modifying the function of the protein, and as a result, it is expected that this will lead to the development of an excellent glycoprotein preparation. For example, it is known that bisect sugar chains have high hydrophobicity, and if the production amount of bisect sugar chains is adjusted by the method of the present invention, the hydrophilicity and hydrophobicity of glycoprotein sugar chains can be adjusted. it is conceivable that. When the hydrophilicity of the glycoprotein sugar chain is improved, for example, in the case of a high molecular weight glycoprotein, the glycoproteins are prevented from agglomerating and insolubilizing with each other, and the stability in an aqueous solution, and thus in blood, is increased. It is useful in using glycoproteins as pharmaceuticals. Conversely, when using low-molecular-weight glycoproteins as pharmaceuticals, increasing their hydrophobicity increases their interaction with other proteins or cell membranes,
The characteristic that the drug stays in the blood or the body for a long time and the drug effect is maintained can be obtained. As described above, the present invention has great academic and industrial value.

[Brief description of the drawings]

FIG. 1 shows a pentasaccharide structure (trimannosyl core) commonly present in N-linked sugar chains.

FIG. 2 shows a minimum hexasaccharide structure representing a bisect sugar chain.

FIG. 3 shows an agalactosyl biantennary sugar chain.

FIG. 4 shows a galactosyl biantennary sugar chain.

FIG. 5 shows binding sites of glycoprotein IgM monomers and N-linked sugar chains.

FIG. 6: Anti-Pseudomonas aeruginosa IgM antibody producing B cell line MP-504
5 shows the relationship between the ratio of GalT activity / GnTIII activity and the ratio of non-bisect sugar chains / bisect sugar chains in an example in which the sugar chain of glycoprotein IgM produced by the culture of Culture No. 5 is modified.

FIG. 7 shows a restriction map near the exon 2 of the GalT gene and a plasmid pGEX1 containing a gene fragment containing the exon 2 of GalT gene and a plasmid pGENX1 containing a gene fragment containing a fragment of the exon 2 of GalT gene.

FIG. 8 shows the first half of the procedure for synthesizing the GalT targeting vector.

FIG. 9 shows the latter half of the procedure for synthesizing the Gal8 targeting vector.

FIG. 10: GalT targeting vector pGEXN-Ne
The GalT chromosome gene when homologous recombination occurs with o-TK is shown.

FIG. 11. GalT targeting vector pGEXN-Ne
2 shows a GalT chromosome gene when heterologous recombination with o-TK occurs.

FIG. 12: GalT targeting vector pGEXN-Ne
The structure of the o-TK gene and the primers used for PCR are shown.

FIG. 13 shows the structure of the GalT chromosome gene and primers used for PCR.

FIG. 14: GalT targeting vector pGEXN-Ne
The structure of the GalT chromosome gene that has undergone homologous recombination with o-TK and the primers used for PCR are shown.

FIG. 15 is an electrophoretic photograph showing the results of Southern hybridization analysis of the parent strain MP-5045, the heterologous recombinant, and the homologous recombinant Y6-17.

FIG. 16 shows a procedure for synthesizing a GalT expression vector.

[Explanation of symbols]

GalT 2ND Exon: Exon 2 of human GalT chromosome gene GalT 2ND Exon ': Fragment of exon 2 of human GalT chromosomal gene pGENK: Plasmid containing a fragment of exon 2 of human GalT chromosomal gene pGESK: Exon 2 of human GalT chromosomal gene Plasmid containing fragment pGEXK: plasmid containing exon 2 fragment of human GalT chromosomal gene pHSV-106: plasmid containing HSV-TK gene pUC-TK: plasmid containing HSV-TK gene pGE-TK: human GalT Plasmid pGEXN-TK containing exon 2 fragment of chromosomal gene and HSV-TK gene: plasmid pM containing exon 2 of human GalT chromosomal gene and HSV-TK gene 1Neo polyA: Plasmid pGEXN-Neo-TK containing a neomycin resistance gene: neomycin gene by Gal that destroyed exon 2 of GalT chromosomal gene
T targeting vector pCXN2-GalT: Human GalT gene expression vector

 ──────────────────────────────────────────────────続 き Continuing from the front page (72) Inventor Reiko Abe 1144 Togo, Mogoshi, Chiba Mitsui Toatsu Chemical Co., Ltd. (72) Inventor Tomoko Yokomatsu 1144, Togo Togo, Mobara, Chiba Mitsui Toatsu Chemicals In 72) Inventor Tadashi Makino 1144 Togo, Mobara-shi, Chiba Mitsui Toatsu Chemicals Co., Ltd.

Claims (5)

[Claims] When a glycoprotein having an N-linked sugar chain is produced in a cell, the expression level of β1-4 galactosyltransferase in the cell is controlled, and the N-glycoprotein is expressed in the cell. Β1-4 against acetylglucosamine transferase III activity
A method of controlling the ratio of bisect sugar chains in the N-linked sugar chains of the glycoprotein by controlling the relative ratio of galactose transferase activity. 2. The method according to claim 1, wherein the expression of β1-4 galactosyltransferase is reduced and the relative ratio of β1-4 galactosyltransferase activity to N-acetylglucosamine transferase III activity in the cells is reduced. The method according to claim 1, wherein the proportion of bisect sugar chains in the N-linked sugar chains of the protein is increased. 3. Increasing the expression level of β1-4 galactosyltransferase and increasing the relative ratio of β1-4 galactosyltransferase activity to N-acetylglucosamine transferase III activity in the cells, 2. The method according to claim 1, wherein the proportion of bisect sugar chains in the N-linked sugar chains of the protein is reduced. 4. The method according to claim 1, wherein the expression level of β1-4 galactosyltransferase in the cell is reduced by a gene targeting method. 5. A method for producing a glycoprotein having an N-linked sugar chain, comprising controlling the expression level of β1-4 galactosyltransferase in a cell and transferring N-acetylglucosamine in the cell. By controlling the relative ratio of β1-4 galactosyltransferase activity to enzyme III activity, the proportion of bisect sugar chains in the N-linked sugar chains of glycoproteins produced by the cells was adjusted, and the cells were cultured. Recovering the glycoprotein from the cells and / or culture obtained by the method.