The present invention relates to an oligosaccharide mixture whose individual constituents have an identical cone structure which is substituted by at least one sugar and in which all positional isomers are represented, and all isomers with respect to the substituent are in the form of α- or β-isomers, to a process for the preparation of said oligosaccharide mixture and to a process for the rapid isolation of a biologically active oligosaccharide.

Aside from their importance as energy carriers and structural units, carbohydrates have an important part to play as information carriers in intercellular communication and in inter¬ cellular recognition processes. In this respect their specific interaction with determinants of the cell surface causes signals to be passed on to the cell. For example, neurotrans- mitters, growth factors, protein hormones and cytokins are able - by means of carbo¬ hydrate-mediated interaction with the target cell - to transport signals to the cell that result in changes in metabolism. The carbohydrate-mediated adhesion of e.g. viruses, bac¬ teria, toxins, tumour cells and immune-competent cells such as lymphocytes and leuco¬ cytes are an important factor in pathological processes. The treatment of the illnesses arising from these processes with antagonists for the carbohydrate-binding receptors (lectin blocking) is a new method for the prevention and therapy of tumour metastasis, in¬ fection and inflammation. In this respect, the development of oligosaccharides having high affinity for carbohydrate-recognising receptors of the target cell is of great importance.

In the search for particularly active oligosaccharides having high affinity, it is common practice to synthesize individual oligosaccharides selectively and to assay them for possible biological activity. Processes for synthesizing oligosaccharides are known. Thus Ogawa, T., Nakabayashi, S., Kitajima, T., describe in Carbohydrate Research 114:225-236 (1983) a process for the preparation of a branched hexasaccharide in a number of inter¬ mediate steps. One of these steps is the reaction of a monomeric sugar that contains two free hydroxyl groups [C(3) and C(6)] with a completely protected activated monomeric sugar. Only substitution at the 6-OH and not at the 3-OH positions occurred. Paulsen, H., Lebuhn, R., disclose in Carbohydrate Research 125:21-45 (1984) a process for the pre¬ paration of penta- and octasaccharides in which only a simultaneous substitution of all free hydroxyl groups and only the linkage of β-glycosidic bonds takes place.

Surprisingly, it has now been found that the oligosaccharide mixtures of this invention enable the number of syntheses and bioassays to be reduced massively. These mixtures

make it possible to select core structures very rapidly.

This invention makes it possible for the first time to provide mixtures of different oligo¬ saccharides in which (a) the oligosaccharides consist of an identical core structure carry¬ ing at least one substituent, such that within a mixture the substituents are all identical but are located in different positions and are linked α- or β-glycosidically, and (b) all positio¬ nal isomers and stereoisomers are present in a mixture. The invention also makes it poss¬ ible to provide mixtures whose oligosaccharides have a uniform degree of substitution.

In one of its aspects, the invention relates to a mixture of different oligosaccharides which are derived from at least two sugar monomers, both individual components of said mixture having an identical core structure of identical or different sugar monomers, wherein

(1) the sugar monomers of the core structure are unprotected or partially protected, but are preferably unprotected,

(2) at least one substituent selected from the group consisting of unprotected and protected mono-, di- and trimeric sugars is linked α- or β-, -0-, -N-, -S- or -C-glycosidically to the core structure, with the proviso that all substituents are identical if there is more than one substituent,

(3) in the mixture all positional isomers with respect to the substituent are represented, and

(4) all isomers with respect to the substituent are present as α- or β-isomers.

Conveniently the oligosaccharides consist of 2 to 8, preferably 2 to 6, most preferably 2 to 3, sugar monomers. The sugar monomers are linked α- or β-(anomeric centre→n) glycosi- dically to one another, where n is a number from 1 to 15 and the term (anomeric centre-n) denotes in which positions of the two participating monomers the glycosidic bond is located. Within the scope of this invention, the core structure of all oligosaccharides of a mixture is built up from 1 to 5, preferably 1 to 3 and, most preferably, 1 or 2, identical or different unprotected sugar monomers. The core structure may be branched or unbranched.

At each core structure of a mixture, at least one unprotected or protected mono-, di- or trimeric sugar is attached to a hydroxyl group which does not participate in a bond within the core structure. The maximum number of sugars attached to a core structure will depend on the number of these hydroxyl groups. Those mixtures are preferred in which the core structures of their components are substituted by 1 to 6, more particularly by 1 to 3 and, most preferably, by 1 or 2 sugars.

The novel mixtures contain only components in which the same sugars are attached to the core structure. The components differ only in the position at which the sugar, or each of the sugars, is attached, and in their stereochemistry. In each inventive mixture all stereo- isomers are represented and also all α- and β-isomers with respect to the substituent Pre¬ ferred mixtures are those in which all components each have only one substituent attached to the core structure and the remaining hydroxyl groups are free.

For example, the following different components are present in a mixture whose com¬ ponents have a core structure containing three free hydroxyl groups: (1) substitution at hydroxyl group 1, (2) substitution at hydroxyl group 2, (3) substitution at hydroxyl group 3, (4) substitutions at hydroxyl group 1 and hydroxyl group 2, (5) substitutions at hydroxyl group 1 and hydroxyl group 3, (6) substitutions at hydroxyl group 2 and hydroxyl group 3, (7) substitutions at all hydroxyl groups, the variants (1) to (7) being present as α-, β-, αβ-, αα-, ββ-, βα-, ααα-, αββ-, αβα-, ααβ-, βαα-, βββ-, ββα- or βαβ-isomers. Useful mixtures are those in which the variants (1) to (6), preferably (1) to (3), with their respective stereo- isomers, are present. Very particularly preferred mixtures have a uniform degree of substi¬ tution. The degree of substitution will be understood as meaning the number of substi¬ tuents per core structure.

Within the scope of this invention, monomeric sugars will be understood as meaning all compounds whose structure conforms to the formula (CH ₂ 0) _m , where m is preferably a natural number from 3 to 15, as well as polyhydroxyaldehydes, polyhydroxyketones, polyhydroxyacids and polyhydroxyamines and derivatives thereof.

Sugar monomers are known from standard works of organic chemistry and biochemistry (e.g. Beyer/Walter, "Lehrbuch der Organischen Chemie", S. Hirzel Verlag Stuttgart, 19th edition, p. 393 et seq.; A.L. Lehninger, "Biochemie", 2nd edition, p. 201 et seq.; J.F. Kennedy, ed., "Carbohydrate Chemistry", Clarendon Press, Oxford, 1988, p. 4 et seq.)

Illustrative examples are sugar monomers selected from the group consisting of D- and L-aldopyranoses and D- and L-aldofuranoses, including glycerol aldehyde, erythrose, threose, ribose, arabinose, xylose, lyxose, allose, altrose, glucose, mannose, gulose, idose, galactose and talose, from the group consisting of D- and L-ketopyranoses and D- and L-ketofuranoses, typically including dihydroxyacetone, erythrulose, ribulose, xylulose, psicose, fructose, sorbose and tagatose, as well as from the group consisting of D- and L-diketopyranoses, typically pentodiulose and hexodiulose.

The term "sugar monomers" comprises also sugar monomers which represent substitutions of the cited examples. To those skilled in the art these sugar monomers typically include protected, partially protected or unprotected deoxy sugars of the D- and L-configuration, preferably 2-, 3-, 4-, 5- and 6-deoxyaldoses such as fucose, rhamnose and digitoxose, 1,2-dideoxyaldoses such as glucal, galactal and fucal, and 1-, 3-, 4-, 5- and 6-deoxyketo- ses, 1-, 3-, 4-, 5- and 6-deoxyazido, 2-, 3-, 4-, 5- and 6-deoxyamino sugars of the D- and L-configuration, typically glucosamine, mannosamine, galactosamine and fucosamine, deoxyacylamino sugars such as N-acylglucosamine, N-acylmannosamine, N-acylgalactos- amine and N-acylfucosamine, preferably the C _r C ₄ alkyl- and aryl esters thereof, such as chloromethyl, dichloromethyl, trichloromethyl, trifluoromethyl, benzyl, phenyl, 3-phenyl- ethyl, 2-, 3-, 4-, 5- and 6-deoxyalkoxycarbonylamino sugars of the D- and L-configura¬ tion, such as trichloroethoxy-, benzyloxy-, 9-fluoroenylmethoxy-, t-butyloxy- or allyloxy- carbonylamino, 2-, 3-, 4-, 5- and 6-deoxyimido such as N-phthalimido, N-dithiasuccin- imido, N-2,3-diphenyl maleimido, N-2,3-dimethyl pyrrole and N-l,l,4,4-tetramethyldi- silylazacyclopentane adduct.

Sugar monomers will also be understood as meaning aldonic, aldaric and uronic acids such as gluconic acid or glucuronic acid, as well as ascorbic acid, amino acid-carrying sugar monomers and those that carry lipid, phosphatidyl or polyol substituents.

Substituted sugar monomers will also be understood as meaning those having a carbon chain longer than 6 carbon atoms, typically heptoses, octoses, nonoses, heptuloses, octuloses and nonuloses, and also the representatives substituted in accordance with the foregoing criteria, for example ketodeoxyoctanoic acid, ketodeoxynonanoic acid, N-acyl- neuraminic acids and N-acylmuramic acids.

Within the scope of this invention, di- and trimeric sugars will be understood as meaning those derived from two or three identical or different monomers cited above. The linkage is preferably α- or β-O-glycosidic, but S-, N- and C-glycosidic linkages are also possible. All carbon atoms of the one participant of a linkage are suitable. Illustrative examples are in particular (1-2)-, (1-3)-, (1-4)-, (1-5), (1-6), (2-3)- and (2-6)glycosidic linkages. Typical examples of dimeric sugars are those selected from the group consisting of treha- lose, sophorose, kojibiose, laminaribiose, maltose, cellobiose, isomaltose, gentibiose, sac¬ charose and lactose. Illustrative examples of trimeric sugars are raffinose and melezitose.

It has been found that it is advantageous if a substituent is linked α- or β-O-, α- or β-N-, α- or β-S- or α- or β-C-glycosidically to the anomeric carbon atom of the core structure direct or through a linking group. This substituent influences a property of the oligosaccharide selected from the group consisting of detectability, separability, bonding and distribution behaviour.

Illustrative examples of substituents that influence the detectability of the oligosaccharide are UV-detectable groups such as p-nitrophenacyl (265 nm), trinitroanilido (342 n ), ω-(4-methoxyphenoxyl)alkylene, where alkylene = (CH^-, and n = 6 to 30 (292 nm) and fluorescing groups such as 5-dimethylamino-l-sulfonylnaphthalene. Examples of substi¬ tuents that influence the separability of the oligosaccharide are ligands for biopolymers, for example biotin as ligand for streptavidin and hapten as ligand for antibodies. Illustra¬ tive examples of substituents that influence the distribution behaviour of the oligo¬ saccharide are lipophilic groups such as ω-(4-methoxyphenoxyl)alkylene, where alkylene = (CH ₂ ) _n and n = 6 to 30, ω-carboalkoxyalkylene, where alkylene = (CH ₂ ) _n and n = 6 to 30, unbranched or branched alkyl or alkenyl groups, fatty acid esters, ceramides, sphingo- sines, steroids, triterpenes, vitamin E. These groups are particularly suitable as lipophilic anchor groups for chromatographic purification, coating lipophilic surfaces, incorporation in liposomes and micelles and transportation through membranes.

When these substituents are attached to the oligosaccharide through a linking group, they are protected by (a) groups that can be activated chemically, for example amines, carb- oxylic acids, phenols, aldehydes, mercaptans that are protected by conventional methods, reactive double bonds that are suitable for conversion into functional groups, or (b) groups that can be activated photochemically, for example arylazide and azirinyl groups.

Within the scope of this invention, a linking group will be understood as meaning a bivalent group of formula -(X-A) _p -X'- wherein X and X' are each independently of the other C _r C ₁₂ alkenylene, C ₁ -C ₁₂ alkynylene -(C _x H ₂ χO) _y -, C ₅ -C ₈ cyclo- alkylene, C ₆ -C ₁₂ arylene or C ₇ -Cι ₂ aralkylene, A is -0-, -S-, -S-S-, -NR ₁₀ -CO-NR ₁₀ -, -NR ₁₀ -CS-NR ₁₀ -, -NR ₁₀ -, -NR ₁₀ -C(O)-O-, -C(0)0-, -C(0)S-, -C(O)NR ₁₀ -, -C(S)S-, -C(S)0-, -C(S)NR ₁₀ -, -SO ₂ NR ₁₀ -, -S0 ₂ -, -P(0)(OH)0-, -P(S)(SH)S-, -P(S)(SH)0-, -P(S)(OH)0-, -P(0)(SH)S-, -P(0)(OH)S-, -P(0)(SH)0-, -P(O)(OH)-NR ₁₀ -, -P(S)(SH)-NR ₁₀ -, -P(S)(OH)-NR ₁₀ -, -P(O)(SH)-NR ₁₀ -, -HP(0)0-, -HP(S)S-, -HP(S)0-, -HP(0)S-, -HP(O)NR ₁₀ - or -HP(S)NR ₁₀ -, where R ₁₀ is H or C _r C ₆ alkyl, and p is an integer from 1 to 30. The linking group is conveniently α- or β-O-, α- or β-N-, α- or β-S-

or α- or β-C-glycosidically linked to the oligosaccharide.

Protective groups and processes for derivatizing the hydroxyl groups with such protective groups belong to the stock of common knowledge of sugar and nucleotide chemistry and are described, inter alia, by Greene, B.T., Protective Groups in Organic Synthesis, Wiley Interscience, New York (1991), by Sonveaux, E., Bioorganic Chemistry 14:274-325 (1986), or by Beaucage, S.L., Iyer, R., Tetrahedron 48:2223-2311 (1992). Illustrative examples of such protective groups are: benzyl, methylbenzyl, dimethylbenzyl, methoxy- benzyl, dimethoxybenzyl, bromobenzyl, 2,4-dichlorobenzyl; diphenylmethyl, di(methyl- phenyl)methyl, bis(dimethylphenyl)methyl, di(methoxyphenyl)methyl, bis(dimethoxy- phenyl)methyl, triphenylmethyl, tris-4,4',4"-tert-butylphenylmethyl, di-p-anisylphenyl- methyl, tri(methylphenyl)methyl, tris(dimethylphenyl)methyl, methoxyphenyl(diphenyl)- methyl, di(methoxyphenyl)phenylmethyl, tri(methoxyphenyl)methyl, tris(dimethoxy- phenyl)methyl; triphenylsilyl, alkyldiphenylsilyl, dialkylphenylsilyl and trialkylsilyl con¬ taining 1 to 20, preferably 1 to 12 and, most preferably, 1 to 8, carbon atoms in the alkyl moieties, for example trimethylsilyl, triethylsilyl, tri-n-propylsilyl, isopropyldimethylsilyl, tert-butyldimethylsilyl, tert-butyldiphenylsilyl, n-octyldimethylsilyl, (1,1,2,2-tetramethyl- ethyl)dimethylsilyl; -(C _r C ₈ alkyl) ₂ Si-0-Si(C _r Cgalkyl) ₂ -, wherein alkyl is typically methyl, ethyl, n- and isopropyl, n-, iso- or tert-butyl; -C^acyl, preferably C ₂ -C ₈ acyl, typically including acetyl, propanoyl, butanoyl, pentanoyl, hexanoyl, benzoyl, methyl- benzoyl, methoxy benzoyl, chlorobenzoyl and bromobenzoyl; R _S1 -S0 ₂ -, wherein R _S1 is C _r C ₁₂ alkyl, preferably C _r C ₆ alkyl, C ₅ cycloalkyl or C ₆ cycloalkyl, phenyl, benzyl, C _r C ₁₂ alkylphenyl and, preferably, C _r C alkylphenyl, or C _r C ₁₂ alkylbenzyl and, prefer¬ ably, Cι-C ₄ alkylbenzyl, or halophenyl or halobenzyl, typically methylsulfonyl, ethyl- sulfonyl, propylsulfonyl, butylsulfonyl, phenylsulfonyl, benzylsulfonyl, p-bromo-, p-meth- oxy- and p-methylphenylsulfonyl; C _r C ₁₂ alkoxycarbonyl, preferably Cι-C ₈ alkoxy- carbonyl, which is unsubstituted or substituted by F, Cl, Br, C _r C ₄ alkoxy, tri(C _r C ₄ alkyl)- silyl or C _r C ₄ alkylsulfonyl, for example methoxycarbonyl, ethoxycarbonyl, n- or isoprop- oxycarbonyl or n-, iso- or tert-butoxycarbonyl, 2-trimethylsilylethoxycarbonyl, 2-methyl- sulfonylethoxycarbonyl, allyloxycarbonyl, or phenoxycarbonyl or benzoxycarbonyl, each unsubstituted or substituted as indicated for alkoxycarbonyl, for example methyl- or meth¬ oxy- or chlorophenoxycarbonyl or -benzoxycarbonyl, as well as 9-fluoroenylmethoxy- carbonyl. The protective groups may be identical or different. Preferred protective groups are selected from the group consisting of linear and branched C _r C ₈ alkyl, preferably C ₁ -C ₄ alkyl, typically methyl, ethyl, n- and isopropyl, n-, iso- and tert-butyl; C ₇ -C ₁₂ aralkyl, typically benzyl; trialkylsilyl containing 3 to 20, preferably 3 tol2, carbon atoms, typically

trimethylsilyl, triethylsilyl, tri-n-propylsilyl, tri-i-propylsilyl, isopropyldimethylsilyl, tert-butyldimethylsilyl, tert-butyldiphenylsilyl, n-octyldimethylsilyl, (1,1,2^2-tetramethyl- ethyl)dimethylsilyl; substituted methylidene groups which are obtainable by acetal- or ketal-foÏ€nation of adjacent hydroxyl groups of the sugars or sugar derivatives with alde¬ hydes and ketones which preferably contain 2 to 12 and, respectively, 3 to 12, carbon atoms, for example C^C^alkylidene, preferably C]-C ₆ alkylidene and, preferably, C ₁ -C ₄ alkylidene such as ethylidene, 1,1- and 2,2-propylidene, 1,1- and 2,2-butylidene, benzylidene; unsubstituted and halogenated C ₂ -C ₁₂ acyl, preferably -C acyl, e.g. acetyl, propanoyl, butanoyl, pentanoyl, hexanoyl, pivaloyl and benzoyl; unsubstituted and halogenated R-S0 ₂ -, wherein R is C _j -C^alkyl, preferably C _r C ₆ alkyl, C ₅ -C ₆ cycloalkyl, phenyl, benzyl, C _j -C ₁₂ alkylphenyl, preferably preferably Cι-C ₄ alkylbenzyl, for example methylsulfonyl, ethylsulfonyl, propylsulfonyl, butylsulfonyl, phenylsulfonyl, benzylsulfonyl and p-methylphenylsulfonyl.

Within the scope of this invention, a partially protected mono- or oligosaccharide will be understood as meaning one containing at least 2, preferably at least 3, free hydroxyl groups.

In another of its aspects, the invention relates to a mixture of different oligosaccharides which are derived from at least two sugar monomers, both individual components of said mixture having an identical core structure of identical or different sugar monomers, wherein

(1) the sugar monomers of the core structure are unprotected or partially protected, but are preferably unprotected,

(2) at least one substituent selected from the group consisting of unprotected and protected mono-, di- and trimeric sugars is linked α- or β-, -0-, -N-, -S- or -C-glycosidically to the core structure, with the proviso that all substituents are identical if there is more than one substituent,

(3) at least one pharmacologically active substituent is attached to the core structure or the complete oligosaccharide, with the proviso that all substituents are identical if there is more than one substituent,

(4) in the mixture all positional isomers are represented with respect to the sugar and the pharmacologically active substituent, and

(5) all isomers with respect to the sugar substituent are present as α- or β-isomers.

Within the scope of this invention, a pharmacologically active substituent will be under-

stood as meaning a negatively charged substituent which is introduced by reacting the free hydroxyl groups of the monosaccharides or oligosaccharides with carboxylic acids which in α-position carry a leaving group, for example a halide, triflate, tosylate, mesylate, brosylate or a diazo group, by reaction with a sulfuric acid derivative (q.v. Example 3) or by reacting the monosaccharides or oligosaccharides with derivatives of phosphoric acid, for example di-0-benzylphosphochloridate, or with derivatives of phosphorous acid such as di-0-benzyl-N,N-dialky-phosphoroamidite, followed by oxidation (I ₂ , tBuOOH) and, in both cases, followed by the hydrogenolytic removal of the benzyl protective group.

The maximum number of pharmacologically active substituents attached to an oligo¬ saccharide will depend on the number of free hydroxyl groups in the oligosaccharide mix¬ ture. Preferred mixtures are those in which the oligosaccharides are substituted by 1 to 6, preferably 1 to 3 and, most preferably, 1 to 2, pharmacologically active substituents.

The novel mixtures contain only oligosaccharides in which the pharmacologically active substituents are identical. The oligosaccharides differ only in the position at which the pharmacologically active substituent or each of the pharmacologically active substituents is attached. In each novel mixture all stereoisomers are represented.

A further object of the invention is a process for the preparation of the novel mixtures, which comprises reacting an unprotected or partially protected mono- or oligosaccharide which has a core structure of identical or different sugar monomers, in the presence of an aprotic polar solvent and optionally a promoter, with an activated, protected mono-, di- or trimeric sugar, and removing any protective groups present.

The reaction can be carried out in solution or in immobilized form. Either core structure or sugar substituent can be immobilized. Useful are all solid phase materials normally used in solid phase peptide synthesis.

By choosing different protective groups for the core structure and the sugar substituents, it is possible to remove these substituents independently of one another.

All aprotic polar solvents are suitable for use in the inventive process. Particularly useful solvents are those selected from the group consisting of nitriles, sulfoxides, sulfones, N-hydrocarbons, N-dialkylcarboxamides, N-alkyllactams, polyethylene glycol dialkyl ethers, cyclic ethers, N-alkylated cyclic amines and mixtures of said solvents with one an-

other. Representative examples of nitriles are acetonitrile, propionitrile, benzonitrile and benzyl nitrile. A suitable sulfoxide is typically dimethyl sulfoxide. Illustrative examples of sulfones are tetramethylene sulfone and dimethyl sulfone. N-Hydrocarbons are typically nitromethane and nitrobenzene. N-Dialkylcarboxamides are typically dimethyl form¬ amide and dimethyl acetamide. A suitable N-alkyllactam is is typically N-methylpyιτoli- done. A suitable polyethylene glycol dialkyl ether is typically diethylene glycol dimethyl ether. Cyclic ethers are typically tetrahydrofuran, dioxane and dioxolane. Typical examples of N-alkylated cyclic amines are N-methylmorpholine and N-methylpyrro- lidone. Preferred solvents are selected from the group consisting of acetonitrile, propio¬ nitrile, d methyl formamide and dimethyl formamide in admixture with acetonitrile. A very particularly preferred solvent is dimethyl formamide or a mixture of dimethyl form¬ amide and acetonitrile, the ratio of dimethyl formamide to acetonitrile advantageously being 1:10 to 10:1, preferably from 1:1 to 1:5, most preferably from 1:4.

In a preferred embodiment of the inventive process, the ratio of oligosaccharide to protected mono-, di- or trimeric sugar is from 10:0.1, preferably from 3:1 and, most preferably, from 2:1.

In a further preferred embodiment of the inventive process, the ratio of protected mono-, -di- or trimeric sugar to promoter is from 10:0.1 to 0.1: 10, the preferred ratio being 1:1.

The reaction is carried out in the temperature range from -40°C to 100°C, preferably from 0°C to 50°C, most preferably at room temperature.

Within the scope of this invention, activation means the introduction of an anomeric leaving or acceptor group into the mono-, di- or trimeric sugar.

The anomeric leaving group is conveniently selected from a subgroup of the group con¬ sisting of (a) -OH, (b) -halogen, preferably F, Cl, Br, I, (c) -OC(=NH)R, (d) -SR, (e) -SeR, (f) -SC(=S)R, (g) -SC(=S)OR, (h) -OP(=0)(OR) ₂ , (i) -OP(OR) ₂ , 0) -0P(=O)R ₂ , (k) -OPR ₂ , (1) -OP(=S)R ₂ , (m) -OC(=S)R, (n) -S-heterocycle, (o) -OP(=NTs)(Ar) ₂ , (p) -OSiR ₃ , (q) -OSnR ₃ , (r) -S(0)R, (s) -OS(=0) ₂ R, (t) -OC(=0)R, (u) -OC(=0)OR, (v) -OC(=0)NHR, (w) -OC(=0)CH ₂ C(=0)R, (x) -Oaryl, where aryl is phenyl or trimethyl- phenyl, (y) is -Oaryl, where aryl is OCH ₂ -o,p-dimethylphenyl, (z) -0(CH ₂ ) ₃ CH=CH ₂ , (zl) -OC(=0)CH ₂ CH ₂ CH=CH ₂ , (z2) -OC^CH^R, (z3) -OCH=CHR, (z4) heterocycles con¬ taining 4 to 6 ring atoms and substituted by O, N, S or Se, the hetero atom being selected

from the group consisting of N, S and P, and is linked through a double bond, and (z5)

C _r C ₁₀ haloalkyl, aryl, aralkyl, heterocycloalkyl or heteroaryl, preferably CF ₃ , CC1 ₃ , methyl, ethyl or benzyl.

The anomeric acceptor group is conveniently selected from a subgroup of the group

cons 5iisstitinngg o ι f (z6) and (z9) , where R is as defined above.

Depending on the participating leaving group, the promoter is preferably selected from the group consisting of

(α) mineral and organic acids such as HC1, H ₂ S0 ₄ , H ₃ P0 ₄ , H ₂ S0 ₄ , trifluoromethane- sulfonic acid, p-toluenesulfonic acid and methanesulfonic acid;

(β) NR ₄ Hal, wherein R is alkyl, cycloalkyl, heterocycloalkyl, aryl or heteroaryl and Hal is fluoro, chloro, bromo or iodo;

(χ) halophilic complexes of B, Ag, Hg, Sn, Zn, Hf, Zr, Ga, Ti, Cu, Pd, Al, Mg, Mn with counterions selected from the group consisting of F ^" , Cl ^" , Br ^" , I ^" , trifluoromethane sulfo- nate, N0 ₃ \ C0 ₃ ", HC0 ₃ ", RS0 ₂ ", where R is alkyl, aryl, heterocycloalkyl or heteroaryl, CN ^" , O ^2" , C10 ₄ ^" , salicylate, cyclopentadienyl, BF ₄ ^" , BenzoylO ^' and S0 ₄ ^2" ;

(δ) BF ₃ and the complexes thereof, preferably BF ₃ -diethylether, trialkylsilyl- or triarylsilyltrifluoromethanesulfonate, preferably trimethyl-, triethyl-, triisopropyltriflu- oromethanesulfonate; alkyl- and arylsulfonic acids, preferably CF ₃ S0 ₃ H and ArS0 ₃ H;

(e) compounds which carry electron-deficient carbonyl groups, e.g. CCl ₃ CHO;

(φ) N-alkylating reagents of the structure RX, where R is alkyl or aryl and X is a leaving group, preferably methyl iodide, methyl trifluoromethane sulfonate, trityl perchlorate, RCOX and RS0 ₂ X;

(γ) thiophilic and phosphophilic reagents such as BF ₃ -diethylether, NOBF ₄ , trialkylsilyl trifluoromethane sulfonate, wherein R is alkyl or aryl, dimethylmethylthiosulfonium tetrafluoroborate, methylsulfenyl bromide, methylsulfenyl trifluoro methane sulfonate, dimethylmethylthiosulfonium trifluoromethane sulfonate, S0 ₂ Cl ₂ /trifluoromethane sulfonic acid, phenyltrifluoromethane sulfonate, Cl ₂ , Br ₂ , 1 ₂ , Cl ⁺ , Br ⁺ , I ⁺ , preferably iodonium collidinium perchlorate and iodonium collidinium trifluoromethane

sulfonate, N-chloro-, N-bromo- or N-iodosuccinimide and their products from the reaction with strong acids (e.g. trifluoromethane sulfonic acid), sulfur and phosphor- alkylating reagents of the general structure RX, where X is a leaving group (e.g. methyl iodide, methyl trifluoromethane sulfonate, trityl perchlorate), alkyl and aryl selenium and sulfenium cations and their precursors (e.g. phenylselenyl trifluoromethane sulfonate);

(η) electron transfer reagents such as tribromophenylammonium hexachloroantimonate;

(i) RX, where R = alkyl and X is a leaving group (e.g methyl trifluoromethane sulfonate).

The inventive process can be carried out under the action of heat, light or electrical fields. The procedure in the case of photochemical activation is described by Hashimoto, S., Kurimoto, I., Fujii, Y., Noyori, R., in J. Amer. Chem. Soc. 107:1427 (1985) and, in the case of electrochemical activation, in J. Org. Chem. 4320 (1993).

It is advantageous to carry out the reaction of the mono-, di- or trimeric sugar carrying an anomeric leaving group of the subgroup(s)

(a) in the presence of a promoter as described in group (α);

(b) in the presence of a promoter as described in group (β) or (χ) or thermally;

(c) in the presence of a promoter as described in group (e), (φ), (δ) or (χ),

(d) to (o) in the presence of a promoter as described in group (χ), (γ) or (η);

(p) and (q) in the presence of a promoter as described in group (α), (χ), (e) or (δ);

(r) in the presence of a promoter as described in group (i);

(s) in the presence of a promoter as described in group (χ) or thermally;

(t) to (w) in the presence of a promoter as described in group (δ) or (χ);

(x) and (y) photochemically or electrochemically;

(z) to (z3) in the presence of a promoter as described in group (γ);

(z4) in the presence of a promoter as described in group (α), (χ), (δ), (γ), (η), (φ) or (i);

(z5) in the presence of a promoter as described in group (χ) or thermally;

(z6) in the presence of a promoter as described in group (α), (δ) or (γ);

(z7) and (z8) in the presence of a promoter as described in group (α), (δ) or (χ); and

(z9) in the presence of a promoter as described in group (α) or (δ).

To obtain an equal number of substituents at the individual oligosaccharides of a mixture it has been found useful to discontinue the glycosylation reaction prematurely. It is pre¬ ferred to discontinue the reaction at a conversion of 10 to 90 %, preferably of 20 to 80 % and, most preferably, of 30 to 70 %, of the mono- or oligosaccharide to be substituted.

In yet another of its aspects, the invention relates to a process for the rapid isolation of a biologically active oligosaccharide by

(a) preparing inventive oligosaccharide mixtures,

(b) determining the biological activity of the mixtures, and

(c) isolating from the active mixtures active oligosaccharides or active oligosaccharide mixtures by chromatography.

In particular, this process is one by means of which it is possible to identify and isolate very rapidly particularly active oligosaccharides by comparing the activities of very many oligosaccharide mixtures with one another and carrying out further processing only with the most active mixtures and isolating therefrom the most active oligosaccharide.

The pharmacological activities of the mixtures of this invention are determined typically by assessing the inhibition of a molecular or cellular response induced by them. Within the scope of this invention, a mixture is always held to be active when it leads to a significant reduction of the molecular or cellular response.

Chromatographic methods of isolating oligosaccharides are known. In these methods the oligosaccharides are separated in accordance with their size, their electrical charge and their binding properties. The methods employed are typically affinity, molecular sieve and ion exchange chromatography. HPLC (high pressure liquid chromatography) is especially useful.

The following Examples illustrate the invention.

Preparation of the oligosaccharide mixtures

Example 1: Preparation of l-(4-methoxyphenoxyl)-oct-8-yl 2-acetamido-2-de- oxy-X-0-(Y-D-galactopyranosyl)-β-D-glucopyranoside (X: 3,4,6; Y: α,β)

H ₂ , Pd/C, MeOH

For galactosylation, 43.8 mg of the imidate 1 (0.064 mol) are added to the N-acetyl glucosamine derivative 2 (14.6 mg, 0.032 mmol) in 0.2 ml of dimethyl formamide and 1 ml of acetonitrile at room temperature in the presence of a 0.813 molar solution of BF ₃ -diethyl ether (7.8 μl, 6.4 μm) in dichlorethane and the mixture is stirred for 3 hours. The reaction is then quenched by the addition of 0.2 ml of triethylamine and the solvent is removed by rotary evaporation. Unreacted 2 is separated by chromatography on silica gel

(toluene/ethanol = 7:1) and the resultant mixture 3 is deprotected for 14 hours by hydrogenolysis (H ₂ , 10 % Pd/C, methanol). 3: l-(4-methoxyphenoxyl)-oct-8-yl 2-acet- amido-2-deoxy-X-0-(tetra-0-benzyl-Y-D-galacto-pyranosyl)-β- D-glucopyranoside (X: 3,4,6; Y: α,β). After purification on C18 Sep-Pak (3x5 ml CH ₂ Cl ₂ /methanol/H ₂ 0 = 68:15:2), the disaccharide mixture 4 is obtained (37 % conversion according to HPLC). 4: l-(4-methoxyphenoxyl)-oct-8-yl 2-acetamido-2-deoxy-X-0-(Y-D-galactopyranosyl)- - ^β -D-glucopyranoside (X: 3,4,6; Y: α,β).

HPLC data of the disaccharide mixture 4 (column: Partisil 5 PAC, 11cm x 4.7mm; mobile phase: CH ₃ CN/H 0 = 95:5; flow rate: 2 ml/min; monitoring: 292 nm): Retention: min (yield): glucosamine derivative 2: peak 1: 2.38(65%); 6 disaccharides 4 (37%): peak 2: 8.53(1.1%); peak 3: 8.95(13.3%); peak 4: 9.80(11.4%); peak 5: 10.97(2.6%); peak 6: 16.87(25.6%) and peak 7: 17.47(47%).

(Of the 6 possible disaccharides 4, all 6 are formed).

Example 2: Preparation of l-(4-methoxyphenoxyl)-oct-8-yl 2-acetamido-

2-deoxy-X-0(α-L-fucopyranosyl)-4-O-(β-D-galactopyranosy l)- β-D-glucopyranoside (X: 3,6,2\3',4\6'; Y: α,β)

For fucosylation, a 0.0765 mmolar solution of the imidate 4 (44.6 mg, 0.077 mmol) is added to the N-acetyl lactosamine derivative 3 (23.7 mg, 0.038 mmol) in 0.5 ml of dimethyl formamide and 2 ml of acetonitrile at room temperature in the presence of a 0.33 mol solution of BF ₃ -diethyl ether (23.7 μl, 7.7 μmol) and the mixture is stirred for 3 hours. The reaction is then quenched by addition of 0.2 ml of triethylamine and the

solvent is removed by rotary evaporation. The product mixture 5 l-(4-methoxy-phenox- yl)-oct-8-yl 2-acetamido-2-deoxy-X-0(tri-0-benzyl-Y-L-fucopyranosyl)-4-0- (β-D-gala- ctopyranosyl)-β-D-glucopyranoside (X: 3.6, 2',3',4',6'; Y: α,β) is subsequently dissolved in 2 ml of methanol and the benzyl protective groups are removed by hydrogenolysis (10% Pd/C, H ₂ ). After purification on an C18 Sep-Pak (3x5 ml of methanol), the tri- saccharide mixture 6 is obtained (28.3% conversion according to HPLC, 32% conversion according to H-NMR). (6: l-(4-methoxyphenoxyl)-oct-8-yl 2-acetamido-2-deoxy- X-0(α-L-fucopyÏ„anosyl)-4-0-(β-D-galactopyranosyl)-β-D-gl ucopyranoside (X: 3,6,2\3',4',6'; Y: α,β)

HPLC data of the trisaccharide mixture 6 (column: Partisit 5 PAC 11 cm x 4.7 mm; mobile phase: CH ₃ CN/H ₂ 0 = 9:1; flow rate: 1 ml/ in; monitoring: 292 nm): retention: min (yield): N-acetyl-lactosamine derivative 3: peak 1: 9.2(72%); 12 trisaccharides (28%): peak 2: 14.4(3.9%); peak 3: 15.4(22.6%); peak 4: 17.9(4.6%); peak 5: 18.8(9.1%); peak 6: 26.3(4.8%); peak 7: 26.8(4.5%); peak 8: 29.1(3.5%); peak 9: 29.9(4.4%); peak 10: 32.4(16.3%); peak 11: 33.4(9.4%); peak 12: 51.2(10.4%); and peak 13: 53.5 min(5.8%).

(Of the 12 possible trisaccharides 6, all 12 are formed).

Example 3: Preparation of l-(methoxyphenoxyl)-oct-8-yl 2-acetamido-2-de- oxy-X-0-(β-L-fucopyranosyI)-4-O-(β-D-galactopyranosyl)- β-D-glucopyranoside Z-O-sulfate ester (X: 3, 6, 2', 3', 4', 6'; Y: α,β; Z ≠ X, 3, 6, 2\ 3 4\ 6')

The trisaccharide mixture 7 (98 mg, 0.095 mmol) is charged to dry pyridine (5 ml) at room

temperature and sulfur trioxide-pyridine complex (15 mg, one equivalent) is added. The reaction mixture is then stirred for 14 hours at 50°C.

The reaction is discontinued by addition of 0.5 ml of methanol. After concentration of the reaction mixture on a rotary evaporator, the residual pyridine is removed as an azeotrope with toluene (3x5 ml). The resultant crude product (117 mg), l-(4-methoxyphenoxyl)-oct- 8-yl 2-acetamido-2-deoxy-X-0-(tribenzyl-Y-L-fucopyranosyl)-4-0-(Î ²-D-galactopyrano- syl)-β-D-glucopyranoside Z-O-sulfate ester (X: 3, 6, 2', 3', 4', 6'; Y: α,β; Z ≠ X, 3, 6, 2', 3', 4', 6'), is dissolved in 5 ml of methanol and 100 μl of glacial acetic acid and deprotected overnight by hydrogenolysis (10% Pd/C, H^). The reaction mixture is afterwards filtered over Celite and the filter product is cautiously washed with 250 ml of methanol and the combined solutions are concentrated to dryness. The crude product is charged to an ion exchange column (0.5 g DEAEA-50), the nonsulfated products are eluted with deionised water and the sulfated products with 1 M NaCl. The sulfated products are charged to a C-18 Sep Pack column and the column is washed with 120 ml of deionised water and the sulfated products 10 (20.2 mg) l-(methoxyphenoxyl)-oct-8-yl 2-acetamido-2-deoxy-X-0-(β-L-fucopyranosyl)-4-0-(β-D-galac topyranosyl)-β-D-gluco- pyranosid-Z-O-sulfate ester (X: 3, 6, 2\ 3', 4', 6'; Y: α,β; Z ≠ X, 3, 6, 2', 3', 4', 6') are eluted with methanol (120 ml). The detection for the ion exchange chromatography and the C-18 chromatography is effected by UV detection at 292nm.

Examples 4 to 13:

The following mixtures are prepared in general accordance with the foregoing Examples:

l-(4-methoxyphenoxyl)-oct-8-yl 2-acetamido-2-deoxy-X-0-(Y-L-fucopyranosyl)-6-0-(β- D-galactopyranosyl)-β-D-glucopyranoside Z-O-sulfate ester 11 (X: 3,4,2',3',4',6'; Y: α,β; Z ≠ X: 3,4,2',3',4',6').

l-(4-methoxyphenoxyl)-oct-8-yl 2-acetamido-2-deoxy-X-0-(Y-L-fucopyranosyl)-4-0- (α-D-galactopyranosyl)-β-D-glucopyranoside Z-O-sulfate ester 12 (X: 3,6,2',3',4',6'; Y-. a& Z≠X-. SAT^'AW ,

l-(4-methoxyphenoxyl)-oct-8-yl 2-acetamido-2-deoxy-X-0-(Y-L-fucopyranosyl)-6-0- (α-D-galactopyranosyl)-β-D-glucopyranoside Z-O-sulfate ester 13 (X: 3,4,2',3',4',6'; Y: α,β; Z ≠ X: 3,4,2',3',4',6'),

l-(4-methoxyphenoxyl)-chct-8-yl 2-acetamido-2-deoxy-X-0-(Y-L-fucopyranosyl)- 3-0- (α-D-galactopyranosyl)-β-D-glucopyranoside Z-O-sulfate ester 14 (X: 4,6,2',3 ^, ,4 ^, ,6'; Y: α,β; Z ≠ X: 4,6,2',3',4',6'),

l-(4-methoxyphenoxyl)-oct-8-yl X-0-(Y-L-fucopyranosyl)-6-0-(α-D-galactopyr--nosyl)- β-D-glucopyranoside Z-O-sulfate ester 15 (X: 2,3,4,2',3',4',6'; Y: α,β; Z ≠ X: 2,3,4,2\3',4\6'),

l-(4-methoxyphenoxyl)-oct-8-yl X-0-(Y-L-fucopyranosyl)-6-0-(α-D-glucopyranosyl)- ^β -D-glucopyranoside Z-O-sulfate ester 16 (X: 2,3,4,2',3',4',6'; Y: α, ^β ; Z ≠ X: 2,3,4,2',3',4',6'),

l-(4-methoxyphenoxyl)-oct-8-yl X-0-(Y-L-fucopyranosyl)-4-0-(β-D-glucopyranosyl)- β-D-glucopyranoside Z-O-sulfate ester 17 (X: 2,3,6,2',3',4',6'; Y: α,β; Z ≠ X: 2,3,6,2\3',4',6'),

l-(4-methoxyphenoxyl)-oct-8-yl X-0-(Y-L-fucopyranosyl)-4-0-(α-D-glucopyranosyl)- ^β -D-glucopyranoside Z-O-sulfate ester 18 (X: 2,3,6,2',3',4',6'; Y: α,β; Z ≠ X: 2,3,6,2',3',4',6'),

l-(4-methoxyphenoxyl)-oct-8-yl 2-acetamido-2-deoxy-X-0-(Y-L-fucopyranosyl)- 3-0-(β-D-galactopyranosyl)-β-D-glucopyranoside Z-O-sulfate ester 19 (X: 4,6,2',3',4',6'; Y: α,β; Z ≠ X: 2,3,6,2',3\4\6'), and

l-(4-methoxyphenoxyl)-oct-8-yl X-0-(Y-L-fucopyranosyl)-4-0-(β-D-galactopyranosyl)- β-D-glucopyranoside Z-O-sulfate ester 20 (X: 2,3,6,2',3',4\6' Y: α,β; Z ≠ X: 2,3,6,2', 3'A', 6')

Biological Examples

Example Bl: Primary assay to investigate the activity of the oligosaccharide mixtures

This assay demonstrates the activity of the oligosaccharide mixtures on the interaction of the ligands sialyl Lewis A or sialyl Lewis X with selectins (E, P, L). The bonding

participants are genetically engineered soluble fusion proteins from each of the extracyto- plasmic domains of E-, P- and L-selectin and the constant region of a human immuno- globulin of the subclass IgG 1.

A. Preparation of selectin/IgG fusion proteins (selectin/IgG chimeras)

The selectin/IgG fusion proteins (selectin/IgG chimeras) are prepared by linking the soluble domains of the different selectins to the carboxy-terminal end of the first constant region (CHI) of human IgGl. The construction is carried out in a manner similar to that described by Walz et al. [Walz, G., Aruffo, A., Kolanus, W., Bevilaqua, B., Seed, B., Science 250:1132-1135 (1990)].

To prepare the E-selectin/IgG chimeras, the complete cDNA of E-selectin available from British Biotechnology (Product No. BBG 57) as well as genomic DNA coding for human IgGl are used. The DNA fragments for E-selectin and genomic DNA of human IgGl are amplified by polymerase chain reaction (PCR) and subsequently fused by SOE-PCR [Horton, R.M., Hunt, H.D., Ho, S.N., Pullen, J.K., Pease, L.R., Gene 77:61-68 (1989)]. The fused DNA fragment coding for the soluble portion of E-selectin (amino acids 1-548, including signal sequence) and constant portions of human IgGl [amino acids 219-478, numbering according to Kabat, E.A., Wu, T.T., Perry, H.M., Gottesmann, K.S:, Foeller, C., Sequences of Proteins of Immunological interest, 5th edition, U.S. Department of Health and Human Services, NEH Publication No. 91-3242 (1991)], is inserted into the suction expression vector pcDNAI/NEO (Invitrogen) and transfected to CHO Kl cells (ATCC CCL-61). After selection with 0.5 mg ml of G418 (GIBCO), stable transfected cell clones are isolated and tested with ELISA for the production of E-selectin/IgG fusion protein. The cell clone with the highest production rate is used for the production of larger amounts of E-selectin/lgG. The transferred CHO-K1 cells are either reproduced in mono- layer culture or in a hollow-fibre system in OptiMEM medium and 2% (v/v) foetal calf-serum and the secreted E-selectin/IgG from the supernatant is purified by affinity chromatography on protein A-sepharose.

The construction of the L-selectin/IgG chimera is carried out in analogous manner. The starting basis is the complete cDNA of L-selectin isolated by PCR of HL-60 (ATCC CCL 240) cDNA as well as the above described genomic DNA of human IgGl. The expression of the L-selectin/IgG chimera is effected in a manner similar to that described for E-selectin/IgG using the same vectors, cell lines, selection and purification methods.

For the construction of the P-selectin/IgG chimera, P-selectin is partially amplified by polymerase chain reaction from the cDNA bank of human lung (KatNr. HL3004b; Clon- tech). The resultant DNA fragment codes for a shortened form of P-selectin with six repeat domains (amino acids 1-568 including signal sequence). After DNA sequencing, and correction of two point mutations by SOE-PCR [Ho, S.N., Hunt, H.D., Horton, R.M., Pullen, J.K., Pease, L.R., Gene 77:51-59 (1989)], the P-selectin fragment - as described for the two other selectins - is fused with genomic DNA of human IgGl. The construct is inserted into pcDNA3 (invitrogen) and transfected to CHO Kl cells. The selection of cell clones producing P-selectin/IgG and chimera purification are likewise effected as described above.

B. Competitive cell adhesion assay for determining the inhibition of the attachment of human HL-60 cells to E-selectin/IgG

1. 96 Microtitration test plates (Duo-Streifen Immulon, Dynatech Produkte AG, Embrach-Embraport, CH, Product No. M17916B) are incubated with 100 μl of purified E-selectin/IgG (100 ng) in PBS for 2 hours at room temperature. The E-selectin/IgG covering solution is then removed.

2. For the blockage of free binding sites, 300 μl of blocking buffer (0.1% w/v BSA in PBS) are pipetted into the wells and left for 30 minutes at 37°C. Afterwards the blocking buffer is removed.

3. The HL-60 cells are prepared parallel thereto: HL-60 cells in suspension culture are centriguged for 5 minutes at 350 xg. After removing the medium, the cells are resuspended in RPMI 1640. The cell suspension contains lxlO ⁶ cells/ml. To the HL-60 cell suspension is added 1/5 volume of freshly prepared MTT solution (thiazolyl blue tetrazolium bromide, Fluka, Buchs, CH, Catalogue No. 88415, 5 mg/ml in PBS) and mixed by pipetting. The marking batch is then incubated for 30 minutes at 37°C. After centrifuging for 5 minutes at 350 xg, the marked cells are washed with 2x20 ml of binding buffer [0.1% of BSA and lOμg/ml of human immunoglobulin (isotype: gamma/lambda; sigma; Catalogue No. 1-2511) in HBSS)] and finally resuspended in binding buffer. The concentration of marked HL60 cells is lxlO ⁶ cells/ml.

4. 20 μl each of the test substance and of l-carbomethoxy-oct-8-yl sialyl Lewis x as

comparison substance are pipetted into the wells.

5. 100 μl each of MTT-marked HL-60 cells in binding buffer are added to the wells and the microtitration plate is agitated for 30 minutes at 37°C on a rotating shaker at 100 rpm.

6. The wells are washed with 2x200 μl of binding buffer.

7. For cell lysis and solubilisation of the MTT, 100 μl of 1-propanol are added to the wells and briefly shaken.

8. The determination of the number of bonded HL-60 cells is made by measuring the absorption of MTT at a wavelength of 600 nm.

C. Examples of the competitive assay

The sulfate esters 10 and 19 exhibit better inhibition than the comparison substance l-carbomethoxy-oct-8-yl sialyl Lewis x:

10: l-(4-Methoxy-phenoxyl)-oct-8-yl 2-acetamido-2-deoxy-X-0-(Y-L-fucopyranosyl)- 4-0-(β-D-galacto-pyτanosyl)-β-D-glycopyranoside Z-O-sulfate ester (X: 3,6,2', 3',4',6'; Y: α,β; Z ≠ X: 3,6,2',3\4',6')

19: l-(4-Methoxy-phenoxyl)-oct-8-yl 2-acetamido-2-deoxy-X-0-(Y-L-fucopyranosyl)- 3-0-(β-D-galacto-pyranosyl)-β-D-glucopyranoside Z-O-sulfate ester (X: 4,6,2',3',4',6'; Y: α,β; Z ≠ X: 4,6,2',3',4',6')