WO2002005854A2

WO2002005854A2 - Bridged glycoconjugates

Info

Publication number: WO2002005854A2
Application number: PCT/EP2001/008246
Authority: WO
Inventors: Georg Orberger
Original assignee: Arimedes Biotechnology Gmbh
Priority date: 2000-07-17
Filing date: 2001-07-17
Publication date: 2002-01-24
Also published as: WO2002005854A3; AU2001289671A1; DE10034712A1

Abstract

The invention relates to glycoconjugates, in which a protein moiety is covalently bridged to an additional protein moiety, a nucleic acid, a lipid, or a micelle, using a carbohydrate chain. The bridging process takes place by means of a chemical group that does not occur naturally in the carbohydrate moiety of glycoproteins and that is coupled to a neuraminic acid group, a galactosamine group or a glucosamine group of the carbohydrate chain. The invention also relates to a method for producing said glycoconjugates. In a first step of said method, eukaryotic cells are cultivated in the presence of a non-physiological precursor carbohydrate and in a second step a covalent bridging of the glycoprotein isolated in the first step to the carbohydrate moiety of a glycoprotein, a nucleic acid, or a lipid, using a reactive chemical group that does not occur naturally there, takes place.

Description

Bridged glycoconjugates

The present invention relates to glycoconjugates in which a protein portion is covalently bridged via a carbohydrate chain with a further protein portion, a nucleic acid or a lipid or a micelle. The present invention further relates to methods for producing such glycoconjugates.

Metabolic labeling of proteins expressed in vivo by administering a precursor molecule to cells that express the protein and introduce the precursor molecule into the metabolic pathway and incorporate it into the protein during biosynthesis is a standard technique in cell biology. In the case of glycoproteins, for example, radiolabelled, physiological (i.e., naturally occurring in the metabolism of cells) carbohydrate building blocks can be used for metabolic labeling, which are incorporated into the oligosaccharide side chains of the glycoproteins, as a result of which these proteins are radiolabelled.

It is also possible to introduce chemically modified non-physiological carbohydrate derivatives into the biosynthetic pathway of cells, these derivatives being incorporated into the glycoproteins by the cells directly or after metabolic conversion of the basic carbohydrate structure.

Kayser et al. (J. Biol. Chem. 267 (1992), pp. 16934-16938) describe the labeling of glycoproteins in vivo by the administration of radioactively labeled N-propanoyl-D-mannosarnine and N-propanoyl-D-glucosamine to rats. It was found that the propanoylhexosamines were introduced into the metabolic pathway and were incorporated into the glycoproteins as N-propanoylneura acids.

In a series of subsequent investigations, this approach, ie the administration of different N-substituted mannosamine derivatives, was carried out according to the metabolic principle Conversion as sialic acid derivatives can be built into the membrane-bound glycoproteins of the plasma membrane, which modifies the cell surfaces of different cells.

For example, Keppler et al. (J. Biol. Chem. 270 (1995), pp. 1308-1314) investigated the sialic acid-dependent binding of various polyoma viruses to human B-lymphoma cells (BJA-B) and to Vero cells. For this purpose, various N-acylated mannosamine derivatives were used as sialic acid precursor molecules in the cell culture, the acyl group consisting of a propanoyl, a butanoyl or a pentanoyl residue. It was found that 50% of the physiological N-acetylneuraminic acid building blocks of the cell surface glycoproteins were replaced by the respective modified N-acylneuraminic acid building blocks.

Wieser et al. (FEBS Letters 395 (1996), pp. 170-173) examined the uptake of N-propanoyl-, N-butanoyl- and N-pentanoylmannosamine in human diploid lung fibroblasts. An altered growth behavior was observed, which was attributed to the altered cell surface sialylation twice.

Herrmann et al. (J. Virol. 71 (1997), pp. 5922-5931) examined the binding of a mouse polyoma virus to the cell line 3T6 (embryonic mouse fibroblasts). By replacing the physiological N-acetylneuraminic acids on the cell surface with N-propanoyl, N-butanoyl and N-pentanoylsialic acids, the infection rate by the viruses decreased considerably. This was attributed to the reduced binding of the viral receptor to the artificial neuraminic acids.

Also with the aim of modifying the cell surfaces, Mahal et al. (Science 276 (1997), pp. 1125-1128), various cell lines were cultivated in the presence of N-levulinoylmannosamine. After metabolic conversion, the cells exposed reactive ketone groups on the cell surface, which came from the levulinoyl residues of the non-physiological neuraminic acids. The ketone groups were then reacted with biotinamidocaproylhydrazide and detected by subsequent binding of fluorescein-avidin or used to bind a ricin-avidin conjugate to selectively kill the cells. By Yarema et al. (J. Biol. Chem. 273 (1998), p. 31168-31179) was also the surface manipulation of cells by metabolic reaction with N-levulinoylmannosamine and subsequent selective chemical reaction of the ketone groups expressed by the reaction on the cell surface with aminooxy galactose and a lactose hydrazide. The resulting additional galactose or lactose residues were then detected with fluorescein-labeled ricin.

Saxon and Bertozzi (Science 287 (2000), pp. 2007-2010) were able to show that after incubation of Jurkat cells in the presence of peracetylated azidoacetylmannosamine, free azido groups on the cell surface are exposed, which are treated with a biotin derivative that has a phosphine group , can be brought to reaction.

Recombinant glycoproteins which have naturally non-occurring neuraminic acids are described in DE-OS 198 52 729.2. According to the information in the description, the glycoproteins described there have changed pharmacokinetics, in particular an extended half-life in the blood circulation.

In addition to the above-mentioned methods of introducing reactive chemical groups into the oligosaccharide side chains of proteins by metabolic conversion, other methods are known with which such groups can be introduced into glycoproteins.

A known method in this regard is, for example, the limited periodate cleavage. In this method, sugars are split which have two cis-hydroxyl groups, the hydroxyl functions being converted into carbonyl functions. The resulting aldehyde groups can then be reacted with other molecules. In the case of glycoproteins, this reaction is possible in principle with mannose, galactose and with neuraminic acids, which have a primary vicinal diol at C8 and C9. A disadvantage of this technique is that the unspecific oxidation reaction is very difficult to control in practice, so that undesirable by-products often arise. In addition, Mannose and Galactose destroys the pyran ring by the oxidation, so that new structural elements are created within the oligosaccharide chain, which can have difficult biological consequences. Accordingly, the reaction is mostly used only for analytical purposes in that the aldehydes formed are reacted with a tritiated reducing agent after the oxidation with periodate.

US Pat. No. 5,635,603 describes the production of antibody fragments which have a carbohydrate residue which is not naturally present there around the amino acid position 18 of the variable region of the light chain, to which various organic molecules are coupled.

A known method for introducing a non-physiological chemical group into oligosaccharides is the enzymatic oxidation of galactose residues by galactose oxidase. In the case of terminal galactose residues, the enzyme oxidizes the hydroxyl group at C6 to an aldehyde function. A disadvantage of this method is that galactose is not a terminal structural element in many glycoproteins and can be substituted, for example, by sialic acids. Therefore, a neuraminidase treatment, for example, must be carried out before the actual oxidation in order to release terminal galactose residues. A further disadvantage is that galactose residues which are not converted represent a ligand for the asialoglycoprotein receptor in the liver. Preparations that have a large number of galactose residues are removed from the blood circulation very quickly and are accordingly ineffective.

These known methods have been used, for example, to immobilize glycoproteins on solid phases. For example, Keogh describes a method for modifying the surface of medical devices or implants in US Pat. No. 6,017,741. Here, 2-amino alcohols are oxidized by periodate to aldehydes, which can then form a Schiff base with free amines on the surface of the medical devices, which is then converted into a stable amine bond by reduction. US Pat. No. 5,989,552 describes immunogenic conjugates of a mucin polypeptide and oxidized mannan, aldehyde groups being formed by the oxidation and Schiff bases formed by the conjugation.

Known methods for linking proteins or glycoproteins to one another are based on the chemical conversion of reactive amino acid side chains. Here, for example, amino groups of lysine residues or sulfhydryl groups of cysteine residues are covalently linked to one another by homo- or heterobifunctional linking molecules, di- or oligomeric protein units being formed. A large number of such reagents are commercially available and have been used in a wide variety of applications. A fundamental disadvantage of the linking method mentioned is that the function of the proteins involved is impaired or even completely destroyed either by the close contact of the proteins with one another or by the derivatization of individual amino acids. When using bifunctional linking molecules with long alkyl groups, which can create a spatial distance, there is also the risk that these structures have an immunogenic effect.

A method for linking proteins is described in US Pat. No. 5,919,758. Part of a protein is first reacted with a reagent that exposes free sulfhydryl groups. A second part of the protein is reacted with a reagent that exposes free maleimide groups. The two protein parts are then brought into contact, so that at least two proteins are linked to one another after the sulfhydryl has reacted with the maleimide group via a thioether group. In this way, erythropoietin oligomers could be produced. A disadvantage of this method is that it is limited to proteins that have no cysteine residues with free sulfhydryl groups. However, this is often not the case.

Molecular biological techniques allow the production of fusion proteins, it being possible for two or more proteins to be connected to one another via their N or C termini with or without the insertion of an amino acid spacer. One disadvantage of this

Technique is that the procedures for making these fusion proteins are general are very expensive. In addition, the N or C termini are not freely accessible within the three-dimensional structure for all proteins, so that a genetic modification of the amino acid sequence can lead to a loss of the functionally active three-dimensional structure. A further problem lies in the fact that the terms can also be directly involved in the biological function, so that this is lost when modified. The inserted new amino acid sequences can also, under certain circumstances, also form immunogenic structures.

Glycoproteins play an increasingly important role as therapeutic agents. These proteins can be produced recombinantly in large quantities using biotechnological techniques. They are usually used to replace or supplement certain endogenous glycoproteins that are not or only insufficiently formed due to a pathological condition, or to achieve a certain biological effect by administering these proteins. The biological effector function is usually found in the protein portion and can be, for example, in an enzymatic activity (t-PA, blood coagulation factors such as factor VIII, factor IX) or a hormonal effect in the sense of receptor-ligand interactions (erythropoietin (EPO) , Interleukins, interferons, GCSF's).

Recent studies indicate that the biological effects of such protein factors often only come into play when several factors act on a cell at the same time. A synergistic effect of GM-CSF and SCF ("stem cell factor") on the maturation of hematopoietic progenitor cells has been demonstrated (Lund et al., Exp. Hematol. 27 (1999), pp. 762-772).

US Pat. No. 5,919,758 describes homodimers of biologically active polypeptides, for example erythropoietin, which are linked to one another by a heterobifunctional cross-linking reagent, the compound of the reagent being linked to the first polypeptide via a sulfhydryl group and to the second polypeptide a maleimido group occurs. Accordingly, it is an object of the present invention to provide new methods for bridging proteins with other proteins or biomolecules which overcome the disadvantages of the prior art mentioned. Another task is to provide new bridged conjugates of proteins with other proteins or biomolecules.

In the context of the present invention, it was found that glycoproteins by means of reactive groups present in their carbohydrate portion, which are introduced into the carbohydrate portion by metabolic conversion of unphysiological precursor molecules in the cells that synthesize the proteins, with other glycoproteins or with nucleic acids or lipids or Can covalently bridge micelles in a relatively simple process and thus the above-mentioned objects can be achieved.

Accordingly, the present invention relates to glycoconjugates in which a protein portion is covalently bridged via a carbohydrate chain with a further protein portion, a nucleic acid or a lipid or a micelle, the bridging being via a naturally occurring non-acidic residue in the carbohydrate portion of glycoproteins and to a Neura acid residue. a galactosamine residue or a glucosamine residue of the carbohydrate chain coupled chemical group.

The above-mentioned protein portion of the glycoconjugates according to the invention is a protein or polypeptide which is glycosylated when expressed in eukaryotic cells, ie whose amino acid sequence contains the signals required for glycosylation in eukaryotic cells. Accordingly, it is preferably the protein portion of a glycoprotein or polypeptide normally expressed in eukaryotic cells or a protein or polypeptide derived therefrom. In this regard, a wide variety of glycosylated proteins or polypeptides come into question. Preferred among others are polypeptides which act as differentiation or growth factors, cytokines or hormones, for example erythropoietin, CSF (colony stimulating factor), G-CSF (granulocyte colony stimulating factor), GM-CSF (granulocyte-macrophage colony stimulating factor) . LIF (leukemia inhibitory factor), TNF (tumor necrosis factor), lymphotoxin, PDGF (platelet-derived growth factor), FGF (fibroblast growth factor), VEGF (vascular endothelial cell growth factor), EGF (epidermal growth factor), TGF- (transforming growth factor α), TGF-ß (transforming growth factor ß), thrombopoietin, SCF (stem cell factor), Oncostatin M, Amphiregulin, Mullerian-Inhibiting Substance, B-CGF (B-cell growth factor), MMIF (macrophage migration inhibitory factor) α-intereron, β-interferon, γ-interferon, IL- (interleukin) -1, IL-2, IL-3, jX-4, IL-5, IL-6, IL-7, IL- 8, IL-9, E -10, ΓL-11, JX-12, E -13, IL-14, B -15, IL-16 and IL-17. These proteins are readily familiar to the person skilled in the art (see the list of references by Sytkowski, US Pat. No. 5,919,758). Thrombolytically acting proteins or polypeptides, for example t-PA and urokinase, are also preferred; Thromboprophylactics, for example antithrombin III; and blood coagulation factors, e.g. factors VIII and IX. Also preferred are glycosylated envelope proteins of viruses, both alone and as part of complete, biologically active virus particles.

The protein content of the glycoconjugates according to the invention, e.g. the above-mentioned differentiation or growth factors, cytokines, hormones, thrombolytics, thromboprophylactics, coagulation factors and viral coat proteins, it can also be a recombinantly produced protein, i.e. a protein synthesized by cells transformed with an expression vector encoding the protein. Also suitable are proteins produced by recombinant methods which are modified from the natural amino acid sequence of the protein by amino acid exchanges or amino acid deletions or insertions or fusion with further protein sequences, for example fusion with glutathione-S-transferase or β-galactosidase or fragments thereof.

The carbohydrate chain, via which the said protein portion of the glycoconjugates according to the invention is bridged, comprises at least one oligosaccharide chain which is formed in the course of the production process according to the invention by metabolic glycosylation of the protein portion in eukaryotic cells, ie by covalently modifying the protein portion by appending the oligosaccharide chain a corresponding glycosylation signa! or motif within the amino acid sequence of the protein portion.

Accordingly, the oligosaccharide chain is either N-glycosidically bound to an asparagine located within the protein portion (N-glycan) or O-glycosidically bound to a serine or threonine located within the protein portion (O-glycan).

In the case of an N-glycosidic bond to the protein portion, the carbohydrate chain of the glycoconjugates according to the invention or the oligosaccharide chain (s) encompassed by them preferably comprises a relatively uniform “core” structure, which in the following is exemplary of a tetraantennary oligosaccharide of the complex Type and the oligomannosidic type structure is included, and which also have fewer antennas than the structures shown by way of example and can optionally be modified by adding further monosaccharides or oligosaccharides.

Preferred “core” structure of oligosaccharides of the N-glycans of glycoconjugates according to the invention:

= Neuraminic acid

Gal = galactose

GlcNAc = N-acetylglucosamine

Man = mannose

In the case of an O-glycosidic bond, the carbohydrate chain of the glycoconjugates according to the invention comprises O-glycans, which can have the following structures, among others:

Examples of oligosaccharides of O-glycans of glycoconjugates according to the invention:

O-Gal-β1-3-GalNAc-αO GlcNAc-β1-3-GalNAc-αO

GlcNAc schillings-sixth GlcNAc schillings-sixth

GalNAc-αO GalNAc-αO

Q-Gal-ßl-3 GlcNAc-ßl-3 '

Q-GalNAc-αl-3-GaϊNAc-αO

GalNAc = N-acetylgalactosamine

In the N-glycans contained in the carbohydrate chain of the glycoconjugates according to the invention, an “antenna” of the oligosaccharide side chain preferably has about 8 to 12 carbohydrate units. In the O-glycans contained in the carbohydrate chain of the glycoconjugates according to the invention, the antenna length can vary between 2 and 20 units. For both glycans - Types, the last sugar residue is preferably one Sialic acid. The neuraminic acid is preferably also α-glycosidically bound to Gal, GalNAc or GlcNAc.

The chemical group which does not occur naturally in the carbohydrate portion of glycoproteins and via which the covalent bridging takes place in the glycoconjugates according to the invention is coupled to a neuraminic acid residue, a galactosamine residue or a glucosamine residue in the carbohydrate chain, preferably to the neuraminic acid residue. The coupling of the chemical group to the neuraminic acid residue, the galactosamine residue or the glucosamine residue is also preferably carried out via a nitrogen atom.

Particularly suitable chemical groups in this context are those selected from the group consisting of aldehyde, ketone, thiocarboxylate, hydrazide, aminooxy, thiosemicarbazide, β-aminothiol, α-halocarbonyl and azide groups. Ketone or aldehyde groups, in particular ketone groups, are particularly preferred.

In a preferred embodiment of the present invention, the chemical group takes part in the formation of a bridging group V which can be formed under reaction conditions which do not denature for proteins. Preferred bridging groups V in the context of the invention are in particular hydrazone, oxime, thiosemicarbazone, thiazolidine or thioester groups, or groups with the following structural formula:

The carbohydrate chain of the glycoconjugates according to the invention optionally additionally comprises a linker group L, which is connected via the bridging group V to a monosaccharide component of the carbohydrate chain, and via a bridging group V to the further protein portion, the nucleic acid or the lipid, where V has the same meanings as V can assume and V and V can be the same or different bridging groups.

In a preferred embodiment, the linker group L is an acid-cleavable linker group. Linker group L may also contain a biologically active group or be a biologically active group, e.g. a toxin or a steroid hormone, for example an androgen, a progestogen or a glucocorticoid.

Suitable as linker group L in the context of the invention are in particular linear or cyclic, unbranched or branched, saturated or mono- or polyunsaturated, mono- or polysubstituted by fluorine, chlorine, bromine, iodine, or a hydroxyl, epoxy, carbonyl, carboxyl or , Amino, mercaptan, phenyl, phenol or benzyl radical substituted groups with 2 to 18 carbon atoms, preferably 2 to 12 carbon atoms, very particularly preferably 2 to 8 or 2 to 6 carbon atoms.

A preferred embodiment of the glycoconjugates according to the invention are dimers in which a protein portion is bridged by a further protein portion via one or more carbohydrate chain (s). Glycoconjugates in which a protein portion is bridged via one or more carbohydrate chains with two or more further protein portions to form trimers, tetramers or oligomers with more than four bridged protein portions are also preferred for certain applications.

The glycoconjugates according to the invention are preferably in a purified form. The term "purified" in this context does not necessarily mean that the glycoconjugates must be virtually purified to homogeneity, although this is of course included in the meaning of this term. For the purposes of the invention, "purified" glycoconjugates are usually in a mixture of different Isomers and in a mixture (for example solution or lyophilisate) with other substances, for example with the constituents of buffer solutions or with carriers or carrier proteins such as bovine serum albumin (BSA) and the like. In this context, “cleaned” means that the glycoconjugates according to the invention are specifically enriched in relation to other constituents of the starting material for their production. The glycoconjugates of the invention are preferably purified in such a way that, apart from certain impurities, they are in a substantially defined mixture with other substances.

Preferred glycoconjugates of the present invention have the general formula

on, where

P is a mono-, di- or oligosaccharide residue with up to 40 glycosidically linked, optionally branched sugar residues, which represent furanose and / or pyranose rings and contain 5 to 230 carbon atoms and N- or O-glycosidic to the protein portion of the glycoconjugate are bound;

R _{Ϊ is} a radical of the general formula AVBX in which

A is a linear or cyclic, unbranched or branched, optionally mono- or polyunsaturated, hydroxylated and / or ketylated (C ₂ -C ₆ ) acyl radical, in particular a (C ₂ -C ₆ ) alkanoyl radical, or a linear or cyclic, unbranched or branched, optionally mono- or polyunsaturated, hydroxylated and / or ketylated (C ₂ -C ₆ ) hydrocarbyl radical, in particular a (C ₂ -C ₆ ) alkyl, alkenyl or alkynyl radical;

V has the meaning given above;

B is a mono-, di- or oligosaccharide residue with up to 40 glycosidically linked, optionally branched sugar residues, which are furanose and / or pyranose rings and contain 5 to 230 carbon atoms, and N- or O-glycosidically bonded to X. ; and

X is another protein portion;

or a radical of the general formula AVLV'BX in which

A, V, V, L, B and X have the meanings given above;

or a radical of the general formula AVX 'or AVLV'X' in which

A, V, V and L have the meanings given above, and

X 'is a nucleic acid or a lipid or a micelle;

R ₂ , R ₃ , R ₄ and R ₅ , which may be the same or different, each represent a hydrogen atom, a linear or branched alkyl radical having 1 to 20 carbon atoms (C “H _{2n + 2} , n = 1 to 20), preferably 1 up to 7 carbon atoms; a linear or branched alkenyl radical having 3 to 20 carbon atoms (C _n H _2n , n = 3 to 20, position of the double bond at C _n n = 2 to 19), preferably 3 to 10 carbon atoms; a linear or branched alkynyl radical having 3 to 20 carbon atoms n = 3 to 20, position of the triple bond at C _n n = 2 to 19), preferably 3 to 10 carbon atoms; an alkenyl or alkynyl radical with 2 or more double bonds or triple bonds with 4 to 20, preferably 7 to 12 carbon atoms; an aryl radical with 6 to 20 Carbon atoms, preferably a phenyl radical; a linear or branched, saturated or mono- or polyunsaturated acyl radical (-CO-Rβ) with a total of 1 to 20, preferably 1 to 7 carbon atoms, in particular acetyl radical, including its mono- or poly-hydroxylated analogs; an aroyl radical having 6 to 20, preferably 6 to 10, carbon atoms; one

Carbonylamide radical of the formula -CONH ₂ or -CONHRg; a linear or branched, saturated or mono- or polyunsaturated thioacyl radical (-CS-Rö) with a total of 1 to 20, preferably 1 to 7 carbon atoms; or a thiocarbamide radical of the formula -CS-NH ₂ or -CS-NHRe; where Rβ is in each case a hydrogen atom or a linear or branched, saturated or mono- or polyunsaturated alkyl radical having 1 to 20 carbon atoms and each of the aforementioned radicals apart from H can optionally be mono- or polysubstituted by halogen, in particular fluorine, chlorine, bromine or iodine, hydroxy, epoxy, amino, mercaptan, phenyl, phenol or benzyl groups.

In a particularly preferred embodiment, the radicals R _2, R _3, R ₄ and R ₅ each represent a hydrogen atom or CH ₃ or _(Cj. -C ₇₎ -acyl, especially acetyl, including its mono- or poly-hydroxylated analogues.

Furthermore, those glycoconjugates are particularly preferred in which A forms a levulinoyl radical together with the chemical group which takes part in the formation of the bridging group V and / or together with the chemical group which takes part in the formation of the bridging group V.

An advantage of the glycoconjugates according to the invention is that the oligosaccharides contained within the carbohydrate chain do not have a rigid conformation, as is known, for example, from amino acid chains within proteins. In connection with the present invention, it was found that the astonishingly free rotation of the covalent bonds between the individual carbohydrate residues gives the overall structure enormous flexibility and contributes to the fact that any steric effects of the bridging have practically no role for the activity of the biologically active components of the glycoconjugates according to the invention play. At the same time, the carbohydrate chain exercises Glycoconjugates according to the invention, due to their size and flexibility, have a shielding or masking function, so that large parts of the protein portion or the protein portions of molecules of the immune system are no longer recognized and accordingly cannot have an immunogenic effect.

Another advantage of the glycoconjugates according to the invention is that the carbohydrate side chains have little or no direct influence on the biological function of the protein constituent or constituents. In the case of glycoproteins, the influence of the carbohydrate side chains is mostly limited to the phase of the biosynthesis of the protein portion, where they are often involved in the folding mechanism, or to the chemical-physical properties of the matured protein.

The present invention further relates to methods for producing the glycoconjugates described above. Accordingly, one aspect of the invention is formed by a process for the preparation of glycoconjugates, which comprises the following steps: a) culturing eukaryotic cells in the presence of a non-physiological precursor carbohydrate which is suitable for being naturally non-existent and attached to a neuramic acid residue, a galactosamine residue or a glucosamine residue to insert coupled reactive chemical group W into the carbohydrate portion of glycoproteins of the cells, and isolate one of the cultured ones

Cells expressed glycoprotein having said chemical group W in its carbohydrate portion; b) covalently bridging the glycoprotein isolated according to step a) by means of the chemical group W mentioned with the carbohydrate portion of a glycoprotein, with a nucleic acid or a lipid, via a reactive chemical group W 'which does not occur naturally there.

In principle, all eukaryotic cells are suitable for the purposes of the method according to the invention, for example primary cells, tissues or cell lines which express the desired glycoprotein to be isolated in step a). Cell lines which have been transiently or stably transformed with an expression vector for recombinant expression which codes for the glycoprotein to be isolated in step a) are particularly suitable. As preferred host cell lines for the recombinant expression are in particular NTH3T3 cells, COS cells, BHK cells, HeLa cells, HEK293 cells or CHO cells, but also insect cells, for example SF9 cells (Schneider cells). CHO cells and BHK cells are particularly preferred.

In step a) of the production process according to the invention, glycoproteins are obtained, e.g. recombinant glycoproteins which have one or more reactive chemical groups W which do not naturally occur in the carbohydrate portion of glycoproteins in their carbohydrate portion. The protein fraction of these glycoproteins then forms the protein fraction or a first protein fraction of the glycoconjugates obtained by the production process according to the invention. Accordingly, the glycoprotein which is isolated in step a) is preferably one of the eukaryotic glycoproteins mentioned as suitable or preferred in connection with the protein portion of the glycoconjugates. If transformed cells are used in step a), the expression vector used for the transformation is preferably selected so that it codes for the corresponding glycoprotein, the protein portion of which is then to form the recombinant protein portion or the recombinant protein portions of the glycoconjugate according to the invention to be produced.

As mentioned above, it can be advantageous to change the recombinantly produced protein portions in relation to their natural amino acid sequence. Another suitable change in this context is the introduction of additional N or O glycosylation sites or signals into the amino acid sequence or the removal or destruction of N or O glycosylation sites contained in the amino acid sequence. In this way, suitable carbohydrate chains can be introduced into protein portions within the scope of the method according to the invention which would otherwise have no or too few or unsuitable carbohydrate chains for bridging. In addition, undesired carbohydrate chains can be removed from a protein portion to be bridged. Step a) of the production process according to the invention is carried out in the presence of a non-physiological precursor carbohydrate which is added to the nutrient medium in which the cells are cultivated. The term “non-physiological precursor carbohydrate” means a carbohydrate building block which is accepted by the enzymatic machinery of the cells as a building block of the glycosylation of glycoproteins, ie is incorporated into the carbohydrate portion of glycoproteins of the cells in unchanged or slightly modified form, and one or more additional reactive chemical chemicals Contains groups that are not present in naturally occurring precursor building blocks of glycosylation. According to the invention, these are, in particular, precursor building blocks that are converted in the cells to neuraminic acid, galactosamine or glucosarine derivatives with reactive chemical groups, which in turn are converted into by the enzymatic machinery of the cells the carbohydrate content of glycoproteins.

Despite the cultivation of the cells in the presence of non-physiological precursor carbohydrates in the sense of the invention, the physiological sialylation pattern remains the same or essentially the same in comparison to untreated cells. For example, when using non-physiological precursor carbohydrates of neuraminic acid biosynthesis according to the invention, only very few free galactose residues (which indicate an incomplete incorporation of neuraminic acid building blocks) are expressed on the recombinant proteins isolated after step a).

By selecting suitable concentrations of non-physiological precursor carbohydrates, the stoichiometry or number of modified carbohydrate derivatives incorporated in glycoproteins, in particular the neuraminic acid, galactosamine or glucosamine derivatives with the corresponding reactive chemical groups, can be controlled efficiently and easily in the course of the production process according to the invention. After culturing the cells under these conditions, certain glycoproteins of interest can then be isolated from the culture supernatants or the cell extracts using standard methods known to those skilled in the art. The glycoproteins obtained in this way now naturally have one or more im Carbohydrate content of glycoproteins does not occur, reactive chemical group (s), depending on the non-physiological precursor carbohydrate used. The metaboic modification according to step a) of the method according to the invention thus represents a very simple and at the same time effective method for introducing reactive chemical groups into the carbohydrate portion of glycoproteins, which then continue with other biologically active molecules such as other glycoproteins, nucleic acids or lipids Corresponding reactive chemical groups as reactants can be used to form new glycoconjugates with advantageous properties.

The reactive chemical group W, which naturally does not occur in the carbohydrate portion of glycoproteins, is coupled to a neuraminic acid residue, a galactosamine residue or a glucosamine residue in the non-physiological precursor carbohydrates according to the invention, preferably via a nitrogen atom. Particularly preferred as chemical group W in this context are those groups which are selected from the group of chemical groups consisting of aldehyde, ketone, thiocarboxylate, hydrazide, aminooxy, thiosemicarbazide, β-aminothiol, α-halocarbonyl and azide group. Ketone or aldehyde groups, in particular ketone groups, are particularly preferred.

Carbohydrates of the general formula are particularly preferred as non-physiological precursor carbohydrates

in which Rg is a radical of the general formula AW, in which A and W have the meanings given above

and

R ₂ and R ₃ also have the meanings given above, and δ and R _{7 can have} the same meanings as R ₂ and R ₃ .

The radicals R ₂ , R ₃ , ^ and R ₇ in the non-physiological precursor carbohydrates according to the invention each represent a hydrogen atom, CH ₃ , or a linear or branched, saturated or mono- or polyunsaturated acyl radical with a total of 1 to 20, preferably 1 to 7 carbon atoms, particularly preferred for an acetyl radical, including its mono- or polydroxylated analogs.

Furthermore, precursor carbohydrates in which A and W together form a levulinoyl residue are particularly suitable.

Specific examples of particularly preferred non-physiological precursor carbohydrates are N-levulinoylmaimosamine, peracetylated N-levuHnoylmannosamine, N-azidoacetylmannosamine and peracetylated N-azidoacetylmannosamine.

In a further aspect, the present invention relates to the use of the non-physiological precursor carbohydrates mentioned for the production of the glycoconjugates according to the invention or the use in the production process according to the invention.

After culturing the eukaryotic cells in the presence of the non-physiological precursor carbohydrate, one (or more, for example in the case of cells which have been cotransformed with several expression vectors or which express several glycoproteins which are of interest for the production of the glycoconjugates according to the invention) is expressed by the cultivated cells Glycoproteins isolated. In this context, another advantage of the invention is shown Production method, since known and standardized methods for the isolation or purification of these glycoproteins can be used to isolate the now reactive groups W in their carbohydrate content glycoproteins. In step a), the glycoprotein is isolated to such an extent that it is in a form which allows efficient covalent bridging in accordance with step b) of the process according to the invention.

In step b) of the method according to the invention, the isolated glycoprotein is then covalently bridged by means of the chemical group W mentioned with the carbohydrate portion of a glycoprotein, with a nucleic acid or a lipid, each of which contains a reactive chemical group W 'which does not occur naturally there.

In the event that the bridging takes place with the carbohydrate portion of a further glycoprotein, the reactive chemical group W 'is preferably coupled, like the reactive chemical group W, via a nitrogen atom to a neuraminic acid residue, a galactosamine residue or even glucosamine residue in the carbohydrate portion of the glycoprotein mentioned. The use of the glycoprotein isolated after step a) is particularly preferred for bridging with itself, covalently bridged homo-oligomers of the glycoprotein obtained in step a) being produced in this way. Homo-dimers of the glycoprotein isolated according to step a) are particularly preferred as homo-oligomers. Relevant examples of suitable homo-dimeric glycoconjugates for the purposes of the invention are CSF-CSF, SCF-SCF, G-CSF-G-CSF, GM-CSF-GM-CSF, LΓF-LIF, EPO-EPO, TNF-TNF, lymphotoxin Lymphotoxin, PDGF-PDGF, FGF-FGF, VEGF-VEGF, EGF-EGF, TGF-α-TGF-α, TGF-ß-TGF-ß, thrombopoietin-thrombopoietin, SCF-SCF, oncostatin M-oncostatin M, amphiregulin- Amphiregulin, Mullerian-Inhibiting Substance-Mullerian-Inhibiting Substance, B-CGF-B-CGF, MMJT-MMIF, α-interferon-α-interferon, ß-interferon-ß-interferon, γ-interferon-γ-interferon, JX- IL-1, IL-2-IL-2, JJ -3-IL-3, IL-4-IL-4, JJ -5-JX-5, JX-6-IL-6, IL-7- IL-7, IL-8-IL-8, D -9-IL-9, IL-10-IL-10, IL-II-IL-11, EL-12-JX-12, IL-13-IL- 13, IL-14-IL-14, E -15-IL-15, JJ -16-IL-16 and JJL-17-IL-17. GM-CSF-GM-CSF, EPO-EPO, α-interferon-α-interferon, β-interferon-β-interferon, γ-interferon-γ- are particularly preferred Interferon, E -2-IL-2, JX-15-IL-15, JX-16-IL-16 and JX-17-IL-17, t-PA-t-PA, urokinase urokinase, antithrombin III antithrombin III, factor VUI factor VUI and factor IX factor IX.

Preferred for the bridging step b) of the method according to the invention are, however, also other glycoproteins produced by the method according to step a), so that hetero-ougomers, preferably hetero-dimers, are formed. Relevant examples of suitable heterodimeric glycoconjugates in the sense of the invention are GM-CSF-SCF, GM-CSF-E -2, GM-CSF-IL-3, CSF-JX-1 and IL-2-IL-3.

In the event that the bridging with a nucleic acid or a lipid or a lipid takes place within a micelle composed of lipids, the reactive chemical group W 'can be present at any suitable position within the nucleic acid or the lipid which has the desired function of the nucleic acid or the lipid or the micelle containing the lipid within the glycoconjugate formed by the bridging not or not significantly impaired. Appropriate positions in this regard can be readily determined by a person skilled in the art. Furthermore, the person skilled in the art is aware of a large number of known methods for introducing or preferably reactive chemical groups W ′ which are suitable or preferred according to the invention into the nucleic acid or the lipid.

In the course of the production process according to the invention, the precursor carbohydrates in step a) are statistically incorporated into the carbohydrate side chains or their branches or “antennas”. This also applies in particular to proteins which have more than one glycosylation site and accordingly have several carbohydrate side chains, which in turn each have several antennas Accordingly, the exact stereochemistry of the glycoconjugates obtained in step b) cannot be predicted exactly, or a mixture of differently linked conjugates may be obtained. However, this plays no role or only a minor role in view of the biological function of the conjugates, since the carbohydrate side chains, especially the neuraminic acids, are usually not directly involved in the biological function of glycoproteins Oligosaccharide chains sufficiently long and flexible to ensure steric accessibility of the individual functionally active protein domains on the bridged protein (s). With regard to glycoproteins with multiple glycosylation sites, it is also possible to limit the number of possible isomers by removing or destroying glycosylation sites that are not critical to biosynthesis.

As mentioned above, the stoichiometry of the incorporation can also be efficiently controlled, inter alia, by the concentration of the non-physiological precursor carbohydrate in the cultivation medium in step a), so that the formation of homo- or hetero-dimers is effectively influenced compared to the formation of homo- or hetero-ougomers can be. A further control possibility in this regard is available to the person skilled in the art through the selection of suitable concentrations and reaction times and conditions in the bridging reaction according to step b), through which, as desired, more dimers or ougomers bridged over the carbohydrate content can be obtained as a percentage. Furthermore, further process steps and measures familiar to the person skilled in the art, e.g. by gel filtration or other chromatographic methods which separate the glycoconjugates obtained in step b) into the individual isomers, i.e. homo- or hetero-dimers or homo- or hetero-ougomers (i.e. trimers, tetramers or larger conjugates) of lower or higher molecular weight.

Like the reactive chemical group W mentioned above, the reactive chemical group W 'is preferably selected from the group of chemical groups consisting of aldehyde, ketone, thiocarboxylate, hydrazide, aminooxy, thiosemicarbazide, β-aminothiol, α- Halocarbonyl and azide group. The groups W and W can each be the same or different from one another. In the event that W and W 'are the same, the covalent bridging takes place according to step b) via a homobifunctional linker molecule. In the event that W and W 'are not the same, the covalent bridging can take place both directly by reacting W with W' and via a heterobifunctional linker molecule. In a particularly preferred embodiment, the reactive groups W and / or W '(in the case of bridging with one or more other protein components) are exposed on a neuraminic acid and are an aldehyde or ketone group. In this FaU, this group is preferably introduced into the glycoprotein which is isolated in this step by metabolic conversion according to step a) of the process according to the invention with a mannosamine derivative as a non-naturally occurring precursor carbohydrate which is N-acylated with an aldehyde or a ketone-containing group. Peracetylated N-levulinoyl-D-mannosamine is particularly preferred in this context. The cells metabolize the substance so that the newly synthesized glycoproteins expose the reactive group W or W 'as part of N-levuhnoyl neuraminic acid (s).

The bridging in step b) is preferably carried out under reaction conditions which do not denature for proteins. In this way, glycoconjugates can be obtained which retain the biological activity of the protein portion, which in turn is due to the covalent bridging with other biologically active biomolecules, i.e. Glycoproteins, nucleic acids or lipids are amplified or supplemented or modulated.

In the event that the bridging in step a) takes place without the use of a linker molecule, the reactive chemical groups W, W in the resulting glycoconjugate according to the invention jointly participate in the formation of a bridging group V. When a homo- or heterobifunctional linker molecule is used, the groups W, W 'do not react directly with one another, but instead each take part separately in the formation of bridging groups V, V, which they each form with reactive chemical groups Z, Z' present in the linker molecule.

The bridging groups V or V are preferably a hydrazone, an oxime, a thiosemicarbazone, a thiazolidine or a thioester group, or a group of the following formula:

where V and V can each be the same or different.

Preferred linker molecules in the sense of the invention are those of the formula ZLZ ', where Z and Z' are selected from the group of chemical groups consisting of aldehyde, ketone, thiocarboxylate, hydrazide, aminooxy, thiosemicarbazide, β-aminothiol and α-halocarbonyl group and the chemical group with the following formula:

and, as mentioned above, may be the same or different, where L has the meanings given above. In particular, the reactive chemical groups Z, Z 'are the same in the homobifunctional linker molecules used in step b), while they are different from one another in the heterobifunctional linker molecules which may be used. Particularly preferred as a homobifunctional linker molecule according to the invention is 1,4-bis (hydroxyamino) ethane, 1,4-bis (hydroxyamino) propane or 1,4-bis (hydroxyamino) butane, the last-mentioned linker molecule being particularly preferred , The person skilled in the art is readily able to select suitable reactive chemical groups W, W or Z, Z 'for the formation of the bridging groups V, V. A wide variety of reactive groups W, W ', Z, Z' come into question for the selective chemical coupling or bridging. The groups W, W, Z, Z 'are preferably selected such that their implementation leads specifically to bridging and that under the chosen conditions the protein portion (or in the case of bridging with the same or another glycoprotein, the protein portions) is not irreversible is (are) denatured and thereby the biological function is lost or impaired. An overview of currently known and preferred reactions which meet the criteria mentioned and represent correspondingly preferred combinations of reactive chemical groups in the sense of the invention is given in FIG. 1. On the basis of these examples and his specialist knowledge, the person skilled in the art is readily able to select suitable groups W, W, Z, Z 'in the sense of the present invention.

2 is an exemplary schematic representation of the bridging of isolated glycoprotein monomers according to step a) of the method according to the invention. As mentioned above, the monomers can be linked, for example, to homooligomers or two different glycoproteins to heterooligomers (in the case of FIG. 2 to homo- or heterodimers). The link can be made directly via the chemically reactive groups W, W '(Fig. 2a). An alternative to this is the indirect linking of the reactive groups W and W via homo- or heterobifunctional linker molecules (FIGS. 2b and 2c), where these bridging molecules can, as mentioned above, themselves be biologically active. The combinations of reactive chemical groups according to FIG. 2 are also suitable for bridging glycoproteins produced according to step a) of the method according to the invention with nucleic acids or lipids which carry the corresponding reactive groups.

As can be seen from the above explanations, the method according to the invention offers a number of advantages. The most diverse glycoproteins can work with one another or with other biologically active molecules using the modular principle linked, thereby combining different biological functions.

The bridging via the carbohydrate chain of the protein portion or the protein portions of the glycoconjugates according to the invention represents an ideal connection of the various functions embodied by the bridged molecules, since the ougosaccharides contained therein are extremely flexible and have little or no immunogen and generally no unspecific adsorption cause. Since the carbohydrate fractions are only slightly modified by the metabolic introduction of reactive groups according to step a) and are sufficiently long, two proteins can be linked to one another, for example, without their respective function being significantly impaired. Furthermore, one can fall back on already established methods for the isolation or purification of the glycoprotein to be linked or the glycoproteins to be linked.

Examples of preferred embodiments of the invention include multifunctional antibodies bridged by their carbohydrate content. With _. With the help of molecular biological techniques, it is now easily possible to produce "minimized" proteins, ie protein fragments or individual domains of proteins that have retained the functionality of the entire protein, or have one of these functions in isolation in the case of proteins with multiple functions. An example of this are so-called " single-chain antibody fragments ", which are significantly smaller than the naturally occurring antibody molecules. The disadvantage of the recombinant production of such proteins, however, is that, because of their small size, these molecules are removed from the blood circulation very quickly by renal filtration in vivo and are therefore no longer available for a therapeutic function.

In this regard, the method according to the invention offers the possibility of ougomerizing corresponding recombinant protein fragments, for example the “single-chain antibody fragments” mentioned, via their carbohydrate content. For this purpose, protein fragments which have, for example, an N-glycosylation sequence of the sequence Asn-X-Ser / Thr, in Mammalian cells are produced recombinantly in the presence of, for example, N-levulmoylmannosamine, as a result of which the glycoprotein fragments thus produced have reactive ketone groups in their carbohydrate content. After the isolation according to step a) of the method according to the invention, these glycoprotein fragments can then be covalently linked according to step b) by bridging to larger ougomer units. On the one hand, this means that the molecules are larger and, accordingly, are no longer removed from the blood circulation by renal filtration. On the other hand, for example in the case of bridged antibody fragments, the resulting ougomers can bind to antigenic structures much more efficiently than individual “single-chain fragments” due to the increased avidity of their binding.

Another problem of antibodies and recombinant fragments in "in v / vo" applications is often the lack of specificity. Since a specific antigenic structure is not only present on one cell type, the antibodies bind to a large number of cells. However, the specificity can be considerable by using, for example, heterospecific IgG-type antibodies that can bind to two different epitopes on a cell, whereas until now these antibodies have had to be prepared in a complex and lengthy process, the bridging technology of the present invention allows easy access Heteroougomers, namely heterougomers covalently bridged via their carbohydrate content, the individual stoichiometric composition of which can be specifically selected, making it easy to produce tailored antibody molecules with the desired binding properties.

In the absence of glycosylation motifs in the natural amino acid sequence, minimal antibodies can also be provided with an artificial glycosylation site, for example in accordance with WO 98/49198. These reactive chemical groups introduced by means of metabolic incorporation according to step a) of the method according to the invention can then be linked to themselves or to other minimal antibodies produced in this way. This results in oligomer conjugates according to the invention with tailor-made binding properties. Another area of application of the invention is polypeptide hormones. Many proteins or polypeptides with hormonal effects that occur naturally in the body are quickly removed from the blood circulation by renal filtration. These include proteins from the group of interleukins, interferons and growth hormones. However, in order to achieve a therapeutic effect in the sense of pharmacological application with these proteins, it is often desirable to achieve an increased concentration in the serum level over a longer period of time. An example of this is human recombinant EPO, which has a relatively short half-life in the blood as a monomer. In US Pat. No. 5,919,758 an EPO dimer is described which, by cross-linking via amino acid side chains, has an extended residence time in the blood circulation and thus an increased biological effectiveness.

The method according to the invention offers a simple alternative here. Thus, according to the invention, polypeptide hormones can be di- or oligomerized by bridging via their carbohydrate portion to glycoconjugates according to the invention, thus producing highly specific and extremely active therapeutic agents with advantageous pharmacokinetics. An example of a preferred glycoconjugate according to the invention is a homo-dimer of erythropoietin which has improved pharmacokinetics compared to the monomeric polypeptide.

With other small glycoproteins, too, the critical minimum size can be exceeded by di- or ougomerization according to the bridging process according to the invention, so that they are not removed from the blood circulation by renal filtration and thus have improved pharmacokinetics.

Another problem that frequently arises in the therapeutic use of polypeptide hormones is that several different polypeptide hormone molecules have to bind to their receptor on the cell surface at the same time in order to trigger a specific cellular signal cascade and thus a certain biological effect. Examples of this are the Colony Stimulating Factor- and Interleukin-1 D -l) -dependent experimentation of Microgüa cells, or the differentiation of hematopoietic Progenitor cells by GM-CSF and E -3. It can therefore be advantageous to combine functions of different polypeptide hormones in one molecule. In this regard, the method according to the invention allows simple covalent bridging of two or more different glycosylated polypeptide hormones via their carbohydrate content. There is also the possibility of using a specific linker molecule, for example a steroid hormone provided with suitable reactive groups, which itself has a hormonal effect. In this way, highly specific and extremely active chimeric therapeutic agents can be created. Examples of such advantageous chimeric conjugates according to the invention are those in which a GM-CSF polypeptide is bridged with a JX-2 or JL-3 polypeptide via reactive chemical groups which have been introduced into their carbohydrate content according to the method of the invention.

According to the method according to the invention, it is also possible to produce and isolate viral protein or whole virus particles which have reactive chemical groups in their carbohydrate content and then very easily in the bridging step with a "target" protein or another "target" molecule, for example a nucleic acid or a lipid or a micelle. For a variety of applications, viral expression vectors have been developed in recent years with which genetic information can be efficiently introduced into cells. However, the targeted infection of cells still poses a problem "Certain cell types either cannot be infected at all or the viruses have a very broad specificity, which leads to an uncontrollable spread of the infection, especially in vivo. To solve this problem, for example, alpha viruses were produced which have modified envelope proteins Expose iron and protein G to the virus surface. Any antibodies can now be bound to this protein via the Fc part of the virus, which then lead to a targeted attack of certain cell types. While the technique described can currently only be used for alpha viruses, the surfaces of all viruses which have glycoproteins in their shell can be modified in a targeted manner with the aid of the method according to the invention. For this purpose, the viruses are produced in a cell culture, for example in the presence of N-levuünoylmannosamine. Then the Virus particles isolated and reacted with a suitable linker molecule according to the invention and an excess of the desired "target" glycoprotein, which also has, for example, N-levuünoyl neuraminic acids. In this way, any target glycoprotein can be applied to a virus surface. If the "target" glycoprotein is For example, an antibody that recognizes certain antigenic structures on a cancer cell can be used to produce virus particles that preferentially affect cancer cells. According to the invention, it is also possible to immobilize a ligand on the virus particle for which only very specific cells have a receptor, so that only these cells are infected by the virus. Such conjugates are also suitable as highly specific ferries for the transfer of nucleic acids packed in virus particles, for example for the purpose of transfection of cells or tissues or for gene therapy.

Furthermore, according to a preferred embodiment of the method according to the invention, the bridging according to step b) can be carried out with a lipid under conditions which are below the critical micellar concentration of the lipid in question. After the bridging reaction, the concentration of the lipid in question can then be increased or further lipids added, so that micelles form in which the glycoconjugates formed are incorporated with their lipid component. The conditions are preferably chosen so that the protein portion of the glycoconjugates contained in the micelles points outwards. The micelles containing glycoconjugates preferably contain biologically or pharmaceutically active substances, e.g. Toxins, RNA or DNA or cytostatic or cytotoxic agents. The outward-pointing protein portion of the glycoconjugates contained in the micelles is also preferably selected so that it has a certain cell or tissue specificity. In this way, cell-specific or tissue-specific transport reagents can be manufactured on the basis of micelles, which are suitable, for example, for the transfection of toes or for tumor or gene therapy.

The glycoconjugates according to the invention are particularly suitable for use as effective medicaments or for their formulation. Accordingly, the present invention also relates to a pharmaceutical composition that contains a portion of contains one or more of the glycoconjugates according to the invention. The glycoconjugates are preferably present in the pharmaceutical compositions of the invention in a form which is adapted to the particular dosage form and are optionally converted into a salt or derivative which is suitable in this regard. The pharmaceutical compositions of the invention may further contain suitable pharmaceutically acceptable carriers or other additives.

Example 1: Preparation of Peracetylated N-Levuünoylmannosamine

4.2 ml of triethylamine (30.2 mmol) are added to a solution of 2.85 ml of levuenic acid (27.8 mmol) in 77 ml of anhydrous tetrahydrofuran. The mixture is then stirred for 30 minutes under a nitrogen atmosphere. Then 3.31 ml of isobutyl chloroformate (25.5 mmol) are added dropwise. The mixture is then stirred at RT for 3 h, a white precipitate of levuoyoyl isobutylcarboxylic anhydride being formed. For the reaction with mannosamine, 5 g of mannosamine hydrochloride (23 mmol) are added to 154 ml of a mixture of water and tetrahydrofuran (5: 4), 3.6 ml of triethylamine (25.5 mmol) are added and the mixture is stirred for 5 min. The slurried levuünoyüsobutylcarboxylic anhydride is then added dropwise to the reaction mixture and the batch is stirred for 24 hours. The solvent is then removed in vacuo.

To the Peracetyüerang 100 mg of the resulting N-levuünoylmannosamine (0.38 mmol) are dissolved in 8 ml of pyridine and mixed with 4 ml of acetic anhydride (44 mmol) and a catalytic amount of 4-dimethylaminopyridine. The reaction mixture is cooled to 0 ° C. and stirred for 12 h. After warming to room temperature, 100 ml of dichloromethane are added. The organic phase is then washed three times with 50 ml of IM HC1, once with 50 ml of saturated sodium bicarbonate solution and once with 50 ml of 1 M sodium chloride solution. The organic phase is dried with sodium sulfate and then concentrated in vacuo. The crude product is then purified by chromatography on silica gel. Example 2: Preparation of erythropoietin (EPO) containing N-levuünoyl neuraminic acid by metabolic conversion

To obtain N-levuünoyl neuraminic acid-containing EPO, CHO cells stably transfected with an EPO expression vector are cultivated in RPMI medium to a density of 1 × 10 ⁶ cells per ml of culture medium in a spinner apparatus. Then 400 ul / 100 ml of culture medium are added to a solution of peracetyüert N-levulinoylmannosamine in a mixture of ethanol and DMSO (10: 1) so that the resulting solution contains 0.5 mM N-levuünoylmannosamine. After 6 h, the medium is changed, the cells being separated by centrifugation and standing in medium containing peracetyuated N-levuünoylmannosamine (0.5 mM) to a density of 1 × 10 ⁶ cells per ml of culture medium. After four days of culture at 37 ° C. under a 5% CO ₂ atmosphere, the eyes are separated by centrifugation and the culture supernatants containing the secreted recombinant EPO are collected.

Example 3: Cleaning EPO

The combined culture supernatants from Example 2 are centrifuged at 10,000 x g. The particle-free solution is then applied at a flow rate of 0.5 ml / min to an immunoaffinity column (3 × 1.8 cm) which contains a covalently bound monoclonal anti-EPO antibody. The column is then washed with 5 column volumes of PBS. The bound EPO is eluted in a column volume with 100 mM glycine / HCl at pH 2.9. The eluate is immediately neutralized by adding 1 M TRIS pH 7.4. The eluate is then exhaustively dialyzed against PBS and concentrated in ultrafiltration tubes (Centricon 30, Amicon).

Example 4:. Bridging of EPO and separation of the ougomers

To bridge the EPO monomers, 2 ml of the protein solution (1 mg / ml EPO in PBS) are mixed with 1,4-bis (hydroxyamino) butane (2 mmol). The mixture is at RT for 2 h incubated. Gel filtration is carried out to remove the unreacted linker molecules and to separate the resulting bridged ougomers, which consist of one, two or more EPO units. For this purpose, a Sephacryl column (Pharmacia) with PBS as the eluent is used. 0.5 ml of the reaction mixture are applied at a flow rate of 0.5 ml / min. The eluate is collected in 1 ml fractions. The protein fractions thus obtained are then examined by SDS-PAGE with subsequent Coomassie staining in order to identify the fractions in which the EPO dimers are contained.

Claims

claims

1. Glycoconjugate, in which a protein portion is covalently bridged via a carbohydrate chain with a further protein portion, a nucleic acid or a lipid or a micelle, the bridging being via a in the carbohydrate portion of

Glycoproteins naturally occurring chemical group coupled to a neuraminic acid residue, a galactosamine residue or a glucosamine residue of the carbohydrate chain.

2. Glycoconjugate according to Claim 1, the glycoconjugate being present in a purified form.

3. Glycoconjugate according to claim 1 or 2, the chemical group being coupled via a nitrogen atom to the neuraminic acid residue, the galactosamine residue or the glucosamine residue.

4. glycoconjugate according to one of claims 1 to 3, wherein the chemical group is selected from the group of chemical groups consisting of aldehyde, ketone, thiocarboxylate, hydrazide, aminooxy, thiosemicarbazide, ß-aminothiol, α-halocarbonyl - and azide group.

5. Glycoconjugate according to one of claims 1 to 4, in which the chemical group coupled to the neuraminic acid residue, the galactosamine residue or the glucosamine residue participates in the formation of a bridging group V which can be formed under reaction conditions which do not denature for proteins.

6. Glycoconjugate according to Anspmch 5, in which the bridging group V is a hydrazone, an oxime, a thiosemicarbazone, a thiazoüdin or a thioester group, or a group of the following formula:

7. Glycoconjugate according spoke 5 or 6, in which the carbohydrate chain further comprises a linker group L, which is connected via the bridging group V to a monosaccharide component of the carbohydrate chain, and via one

Bridging group V is connected to the further protein portion, the nucleic acid or the lipid, where V can have the same meanings as V and V and V can each be the same or different bridging groups.

8. Glycoconjugate according to spoke 7, in which the linker group L is an acid-cleavable linker group.

9. Glycoconjugate according to spoke 7 or 8, in which the linker group L is a biologically active group.

10. glycoconjugate according to claim 9, wherein the linker group L is a toxin or a steroid.

11. Glycoconjugate according to one of claims 7 to 10, wherein the linker group L is a linear or cycüsche, unbranched or branched, saturated or one or more unsaturated and optionally one or more times by fluorine, chlorine, Bromine, iodine, or a hydroxyl, epoxy, carbonyl, carboxyl, amino, mercaptan, phenyl, phenol or benzyl radical substituted group having 2 to 18 carbon atoms.

12. Glycoconjugate according to one of claims 1-11, which has the following general formula:

in which

Ri is a radical of the general formula AVBX in which

A is a linear or cyclic, unbranched or branched, optionally mono- or polyunsaturated, hydroxylated and / or ketyüerte (C ₂ -Ce) acyl radical, in particular a (C ₂ -C6) alkanoyl radical, or a non-linear or cyclic, unbranched or branched, optionally mono- or polyunsaturated, hydroxylated and / or ketylated (C ₂ -C ₆ ) hydrocarbyl radical, in particular a (C ₂ -C δ) alkyl, alkenyl or alkynyl radical;

V has the meaning given above; B is a mono-, di- or ohgosaccharide residue with up to 40 glycosidically linked, optionally branched sugar residues, which are furanose and / or pyranose rings and contain 5 to 230 carbon atoms, and N- or O-glycosidically bound to X. ; and

X is another protein portion;

or a radical of the general formula AVLV'BX in which

A, V, V, L, B and X have the meanings given above;

or a radical of the general formula AVX 'or AVLV'X' in which

A, V, V and L have the meanings given above, and

X 'is a nucleic acid or a lipid or a micelle;

and

R ₂ , R ₃ , R and R ₅ , which may be the same or different, are each one

Hydrogen atom, a linear or branched alkyl radical having 1 to 20 carbon atoms

n = 1 to 20); a non-linear or branched alkenyl radical having 3 to 20 carbon atoms (CJHb _n , n = 3 to 20, position of the double bond at C _n n = 2 to 19); a non-linear or branched alkynyl radical having 3 to 20 carbon atoms (C _n H _2n - ₂ , n = 3 to 20, position of the

Triple bond at C „n = 2 to 19); an alkenyl or alkynyl radical with 2 or more double bonds or triple bonds with 4 to 20 carbon atoms; an aryl group of 6 to 20 carbon atoms; a non-linear or branched, saturated or mono- or polyunsaturated acyl radical (-CO-Rβ) with a total of 1 to 20 carbon atoms, including its mono- or polyhydroxy-analogues; an aroyl group of 6 to 20 carbon atoms; a carbonylamide radical of the formula -CONH ₂ or -CONHRe; a linear or branched, saturated or mono- or polyunsaturated thioacyl radical (- CS-Rβ) with a total of 1 to 20 carbon atoms; or a thiocarbamide radical of the formula -CS-NH ₂ or -CS-NHR _< s; where Rβ each represents a hydrogen atom or a linear or branched, saturated or mono- or polyunsaturated alkyl radical having 1 to 20 carbon atoms and each of the abovementioned radicals apart from H can optionally be mono- or polysubstituted by halogen, hydroxyl, epoxy or , Amino, mercaptan, phenyl, phenol or benzyl groups.

13. glycoconjugate according to claim 12, in which the radicals R ₂ , R ₃ , j and R ₅ each represent a hydrogen atom, CH ₃ or (-C-C ₇ ) acyl, in particular AcetyL including its mono- or polydehydrogenated analogs.

14. Glycoconjugate according to claim 12 or 13, in which A together with the chemical group which takes part in the formation of the bridging group V and / or together with the chemical group which takes part in the formation of the bridging group V forms a levuünoyl residue.

15. A method for producing glycoconjugates, comprising the following steps: a) culturing eukaryotic cells in the presence of a non-physiological

Precursor carbohydrate, which is suitable for introducing a first reactive chemical group W, which does not naturally occur there and is coupled to a neuraminic acid residue, a galactosamine residue or a glucosamine residue, into the carbohydrate portion of glycoproteins of the toes, and isolating a glycoprotein expressed by the cultured cells, the so-called chemical Group W has in its carbohydrate portion; b) covalently bridging the glycoprotein isolated according to step a) by means of the chemical group W mentioned with the carbohydrate portion of a

Glycoproteins, with a nucleic acid or a lipid, via a further reactive chemical group W ^' which does not naturally occur there.

16. The method according spoke 15, wherein the eukaryotic toes are selected from the group of cells consisting of NTH3T3 cells, COS cells, BHK cells, HeLa cells, HEK293 cells, CHO cells and SF9 cells.

17. The method according spoke 15 or 16, wherein the chemical group W is coupled via a nitrogen atom to a neuraminic acid residue, a galactosamine residue or a glucosamine residue in the carbohydrate portion of the glycoprotein iso-etherated according to step a).

18. The method according to any one of claims 15 to 17, wherein the covalent bridging in step b) is carried out with the carbohydrate portion of a glycoprotein.

19. The method of claim 18, wherein the chemical grappe W 'is coupled via a nitrogen atom to a neuraminic acid residue, a galactosamine residue or a glucosamine residue in the carbohydrate portion of the glycoprotein.

20. The method according to any one of claims 15 to 20, wherein the chemical grapple W and W 'are each selected from the group of chemical grapple consisting of aldehyde, ketone, thiocarboxylate, hydrazide, aminooxy, thiosemicarbazide, ß-aminothiol -, α-halocarbonyl and azide group, and may be identical or different from one another, where in the case that W and W 'are the same, the covalent bridging takes place via a homobifunctional linker molecule, and where in the case that W and W' are not the same, the covalent bridging optionally takes place via a heterobifunctional linker molecule.

21. The method according to any one of claims 15 to 20, wherein the chemical grapples W and W 'after covalent bridging jointly participate in the formation of a bridging group V or separately from each other in the formation of bridging groups V, V, and the bridging under for Proteins are not denaturing reaction conditions.

22. The method according to any one of claims 15 to 21, wherein the bridging group V or V is a hydrazone, an oxime, a thiosemicarbazone, a thiazolidine or a thioester group, or a group of the following formula :

23. The method according to any one of claims 15 to 22, in which the covalent bridging takes place via a linker molecule of the formula ZLZ ',

in which

Z and Z 'are selected from the grappa of chemical grappa consisting of aldehyde, ketone, thiocarboxylate, hydrazide, aminooxy, thiosemicarbazide, ß-aminothiol and α-halocarbonyl groups and the chemical group with the following formula:

and can be the same or different; and L can assume the meanings given above.

24. Glycoconjugate or method according to one of claims 1 to 23, in which the glycoprotein, the protein portion of which is bridged, or in which the glycoprotein isomerized in step a) is a recombinantly produced glycoprotein.

25. glycoconjugate or method according to any one of claims 1 to 24, in which it is in the glycoprotein whose protein portion is bridged, or in the glycoprotein isolated in step a), a glycoprotein from the Grappe of the differentiation or growth factors , Cytokines, hormones, thrombolytics, thromboprophylactics, coagulation factors, viral proteins, including complete bioactive viral particles, and antibodies, including bioactive antibody fragments.

26. Glycoconjugate or method according to spoke 24 or 25, in which the glycoprotein forms a homo-dimer through the bridging.

27. Glycoconjugate or method according to any one of claims 24 to 26, wherein the glycoprotein is selected from the grappe consisting of erythropoietin, CSF, G-CSF, GM-CSF, LIF, TNF, lymphotoxin, PDGF, FGF, VEGF, EGF , TGF-α, TGF-ß, thrombopoietin, SCF, oncostatin M, amphiregulin, MuUerian-inhibiting substance, B-CGF, MLF, α-interferon, ß-interferon, γ-interferon, E -1, IL-2, E -3, JJL-4, IL-5, D -6, JJL-7, EL-8, JJL-9, IL-10, D -11, IL-12, IL-13, JJL-14, IL-15 , IL-16, EL-17, t-PA, urokinase, antithrombin III, factor VIII and factor IX.

28. Glycoconjugate or method according to spoke 24 or 25, in which the glycoprotein forms a hetero-dimer through the bridging.

29. Glycoconjugate or method according to spoke 28, in which the hetero-dimer is a hetero-dimer which is selected from the Grappe consisting of GM-CSF-SCF,

GM-CSF-E -2, GM-CSF-IL-3, CSF-IL-1 and EL-2-IL-3.

30. A pharmaceutical composition containing a glycoconjugate according to one of claims 1 to 14 or 24 to 29, optionally in the form of a galenically suitable salt or derivative, and one or more pharmaceutically acceptable carriers or additives.

31. Use of a precursor carbohydrate of the general formula

in which

R 1 is a radical of the general formula AW, in which A and W have the meanings given above

and

R ₂ and R ₃ also have the meanings given above, and Rg and R _{7 can have} the same meanings as R ₂ and R ₃ ,

for the production of a glycoconjugate or in a process according to one of claims 1 to 29.