WO1991002739A1

WO1991002739A1 - Synthesis of glycosides having predetermined stereochemistry

Info

Publication number: WO1991002739A1
Application number: PCT/US1990/004338
Authority: WO
Inventors: Samuel J. Danishefsky
Original assignee: Yale University
Priority date: 1989-08-11
Filing date: 1990-08-02
Publication date: 1991-03-07
Also published as: EP0485517A1; JPH05500057A; AU6333090A

Abstract

Methods of stereoselectively preparing a substituted 1,2-anhydrosugar wherein a substituted glycal is converted into a 1,2-anhydrosugar that is utilized to glycosylate a hydroxyl group containing glycal, are disclosed. Methods of preparing a saccharide multimer wherein a nucleophile is reacted with a substituted 1,2-anhydrosugar having a plurality of non-participating substituents, are disclosed. In addition, methods of preparing stereospecific halo-substituted saccharide multimers by haloglycosylation of a glycosyl donor substituted glycal and a glycosyl acceptor glycal derivative in the absence of water and in the presence of a halonium ion, are disclosed. Methods of preparing stereospecific particle-linked halo-substituted saccharide multimers, by haloglycosylation of a particle-linked nucleophilic hydroxyl group with a substituted glycal and a halonium ion, are also disclosed.

Description

SYNTHESIS OF GLYCOSIDES HAVING PREDETERMINED STEREOCHEMISTRY

BACKGROUND OF THE INVENTION

1. Field of the Invention The present invention contemplates the use of substituted glycals, solid phase-bound nucleotides and substituted 1,2-anhydrosugars in the preparation of glycosides of predetermined stereochemical configuration, and particularly in the preparation of oligosaccharides and halo-substituted oligosaccharides of predetermined stereochemical configuration using such glycals or 1,2-anhydrosugars as starting materials, and in the haloglycosylation of hydroxyl molecules that are linked to a solid phase support.

2. Description of the Prior Art

The possibility of using substituted 1,2- anhydrosugars as glycosyl donors in reactions has been recognized, but not extensively exploited. See Sondheimer, et al . , Carbohydr. Res. 74:327 (1979); Sharkey, et al . , Carbohydr. Res. 96:223 (1981); and Trumbo, et al . , Carbohydr. Res. 135:195 (1985); Lemieux, Can. J. Chem. 31:949 (1953); Le ieux, et al . , Can. J. Chem. 32:340 (1954); and Lemieux, et al . , J. Am. Chem. SOC. , 78:4117 (1956).

Equation I of Figure 1 illustrates the synthesis of anhydrosugars such as generalized Compound 1 in which R is any group from suitably functionalized and differentiated hexose derivatives has been accomplished in certain cases — see Brigl, Z. Physiol . Chem. 122:257 (1922), and Lemieux, et al . , Meth. Carbohydr. Chem. 2:400 (1963) — but has not been widely developed. Perhaps owing to a lack of broad availability of such oxiranes, the transformation implicit in Equation I, wherein NuH is a nucleophile, has not been systematically investigated or generalized — see Hickenbottom, J. Chem. Soc. 3140 (1928); Hardegger, et al . , Helv. Chim. Acta 31:221 (1948); Klein, et al., J. Am. Chem. Soc. 104:7362 (1982); and Bellosta, et al . , J. Chem. Soc. Chem. Commun. 199 (1989).

The use of glycals such as generalized Compound 2 of Equation II of Figure 1 as starting materials for glycoside formation has been studied, is discussed herein, and is broadly shown in Equation II. The symbol X in Equation II includes two possibilities.

In one instance, X represents a non-isolable (onium) intermediate arising from attack of E⁺ (an electrophilic reagent) upon Compound 2. Alternatively, X can represent an isolable compound that subsequently reacts with a nucleophile, NuH. With respect to the latter formulation, the possibility of direct epoxidation of glycals to produce substituted 1,2- anhydrosugars such as Compound 1 was considered.

Previous attempts to achieve this reaction using peracids did not lead to isolable epoxides. See Berg ann, et al . , Ber. 54:440 (1921); Levene, et al . , J. Biol . Chem. 88:513 (1930); Haworth, et al . , J. Chem. Soc. 2615 (1930); Haworth, et al . , J. Chem. Soc. 2637 (1930); Hirst, et al . , J. Chem. Soc. 1131 (1931); Levene, et al . , J. Biol . Chem. 93:631 (1931); Wood, et al . , J. Am. Chem. Soc. 79:3234 (1957); and Sweet, et al . , Can. J. Chem. 44:1571 (1966). Instead, there were obtained products that correspond to the reaction of an initially formed substituted 1,2-anhydrosugar that subsequently reacted with either solvent or carboxylic acid (derived from reduction of the percarboxylic acid) .

Virtually all current glycosylations conserve the oxidation level of both coupling components. See for example, Paulson, Angew. Chem. , Int. Ed. Engl . 21:155 (1988), and

Schmidt, Angew. Chem. , Int. Ed. Engl . 25:212 (1986). Consider the merger of two hexose residues as shown in the schematic equations of Figures 6a and 6b.

Typically, the glycosyl acceptor (A) enters the reaction with a single free hydroxyl group and four protected oxygen atom appendages (-OP and -OP¹; P = protecting group). The donor (D) must be equipped with a displaceable or leaving group (L) at its anomeric carbon and is presented for coupling with four masked (protected) hydroxylic, deoxy or other substituent centers.

If the AD disaccharide that is formed as shown in Figure 6a is itself eventually to function as a glycosyl donor, for elongation to an oligosaccharide, its reducing end must be furnished with glycosyl donating (i.e., a leaving group, L) capabilities. Provision of a leaving group capability, in the form of a unique blocking group P¹ at the anomeric center of the original A acceptor, is necessary. Conversely, if the AD disaccharide that is formed as shown in Figure 6b is itself eventually to function as a glycosyl acceptor, for elongation to an oligosaccharide, its oxidizing end must be furnished with glycosyl accepting (i.e., alcohol nucleophile group) capabilities. Provision of a nucleophilic group capability, in the form of a unique blocking group P¹ away from the anomeric center of the original donor, D, is then necessary. Thus, it would be convenient and preferable if a unique P¹ group function in a glycosyl acceptor molecule, A, or a glycosyl donor molecule, D, were suitable for conversion to the OL or OH group of AD in Figure 6a or 6b, respectively, so that the reaction could readily be reiterated.

BRIEF SUMMARY OF THE INVENTION The present invention relates to the preparation of a glycoside. Oligo- and polysaccharides, sometimes referred to herein generically as saccharide multimers, are particularly preferred glycosides. The prepared glycosides have a predictably predetermined stereoconfiguration about the formed glycosidyl bond, and preferably about each of the substituent bonds as well.

In one aspect, this invention contemplates the use of a substituted glycal in the preparation of the glycoside. A method aspect contemplates preparing a glycal-terminated oligosaccharide, and comprises converting a substituted glycal having only nonparticipating substituents in the glycal- containing ring into a corresponding substituted 1,2- anhydrosugar thereof. The substituted 1,2-anhydrosugar is utilized to glycosylate a glycal derivative having a single reactive, nucleophilic, hydroxyl group. The conversion and glycosylation steps are thereafter repeated seriatim to prepare a glycal-terminated oligosaccharide of desired length. The glycal-terminated oligosaccharide prepared by this method can itself be converted to a corresponding substituted 1,2-anhydrosugar that itself can be used to glycosylate the hydroxyl group of a substituted sugar derivative other than a glycal to form an oligosaccharide containing an additional glycoside bond.

The substituted glycal compounds utilized in this method and those that follow have five through nine carbon atoms in a chain, with hexoses and pentoses being preferred as are nonuloses for sialic acid derivative syntheses. Such glycals contain a ring having five or six atoms including oxygen, and have a plurality of non-participating substituents. Such substituted glycals are preferably converted to corresponding substituted 1,2-anhydrosugar intermediates by epoxidation with a dialkyl dioxirane having a total of two to about six carbon atoms in the two alkyl groups. The substituted 1,2-anhydrosugar intermediate is reacted with a nucleophile having a nucleophilic atom selected from the group consisting of oxygen, nitrogen and sulfur in the presence of a Lewis acid catalyst and absence of water to form an epoxide- opened corresponding glycoside reaction product.

The non-participating substituents of the substituted glycal are preferably 3-position hydroxyl, hydrogen, C^-Cg alkyl and O-ether substituents. In one aspect, the glycoside reaction product is an oligosaccharide, whereas in another, the substituted glycal is itself an oligosaccharide.

A substituted 1,2-anhydrosugar having a chain of five through nine carbon atoms, a ring that contains five or six atoms including oxygen, and a plurality of non-participating substituents can itself be used to form an oligosaccharide. That substituted 1,2-anhydrosugar is reacted with a hydroxyl group of a glycal derivative having five through nine carbon atoms in the ring, five, six or seven carbon atoms in a chain and a plurality of substituents to form an expoxide-opened glycosyl glycal reaction product.

The glycosyl glycal reaction product is converted to a corresponding glycosyl substituted 1,2-anhydrosugar intermediate by reaction with a dialkyl dioxirane such as dimethyl dioxirane as already discussed. The intermediate so formed is thereafter reacted with a nucleophile having a nucleophilic atom as discussed before in the presence of a Lewis acid and absence of water to form a corresponding glycosyl glycosyl glycoside reaction product in which the nucleophilic atom is bonded to the anomeric carbon atom at the terminal glycoside bond.

Where the substituted 1,2-anhydrosugar is itself an oligosaccharide, the above method extends or lengthens the chain of the original oligosaccharide by at least one saccharide unit. The original oligosaccharide chain is extended by at least two saccharide units where the nucleophile is a hydroxyl group of a sugar derivative.

Where the glycal derivative contains a participating substituent so that the corresponding glycosyl glycal reaction product formed in the glycosylation step also contains a participating substituent, any participating substituent is removed and replaced by a non-participating substituent prior to the conversion to a corresponding substituted 1,2- anhydrosugar is carried out.

In another aspect of the invention, a particle-linked glycal-terminated saccharide multimer such as an oligosaccharide is prepared. Here, a particle-linked substituted 1,2-anhydrosugar having only nonparticipating substituents, as can be converted from a corresponding particle-linked substituted glycal, is provided. That substituted 1,2-anhydrosugar is used to glycosylate a glycal derivative having a single reactive hydroxyl group. The corresponding substituted glycosyl glycal so formed is converted into a corresponding substituted 1,2-anhydrosugar thereof, and then used to glycosylate a glycal derivative having a single reactive, nucleophilic hydroxyl group. The converting and glycosylating steps are repeated seriatim . In preferred practice, each glycal derivative having a single reactive hydroxyl group discussed before also contains a plurality of non-participating groups. Still further, the glycal-terminated saccharide multimer prepared by the above method can be cleaved from the particle and recovered.

In a further aspect, this invention relates to the preparation of a glycoside by the reaction of a glycosyl donor substituted 1,2-anhydrosugar with a nucleophile that is preferably linked to a particulate solid phase support. In the preparation of a particle-linked glycoside, a solid phase particle-linked nucleophile having a single nucleophilic atom per molecule is glycosylated in a liquid composition free of water with a substituted 1,2-anhydrosugar having a plurality of non-participating substituents. A particle-linked epoxide ring-opened glycoside derivative is thereby formed. The solid and liquid phases are separated.

A hydroxyl group is formed at the 2-position of the formed (added) glycosyl ring, and that hydroxyl group is preferably protected prior to carrying out any further glycosylation reactions or the deprotection reaction discussed below. Preferably, one of the non-participating substituents of the substituted 1,2-anhydrosugar is a protected nucleophilic hydroxyl group whose protecting moiety can be selectively removed in the presence of any other substituent present. Exemplary of such protecting moieties are trisubstituted silyl moieties such as trimethylsilyl where the other protecting moieties are benzyl groups.

In that preferred situation, and where it is desired to extend the glycoside formed, the selectively removable protecting moiety is removed, i.e., the protected nucleophilic hydroxyl group is deprotected, and the resulting free, unprotected hydroxyl group is glycosylated with another substituted 1,2-anhydrosugar.

The nucleophilic atom of the particle-linked nucleophile can be oxygen, nitrogen or sulfur, although oxygen is preferred. Where the nucleophilic atom is oxygen, that atom is preferably present in a hydroxyl group, and that nucleophilic hydroxyl group is preferably present in a sugar derivative. Regardless of the nucleophile molecule used, that molecule is preferably linked to the support particles by a selectively severable bond or link.

The glycosylation reaction is preferably carried out in the presence of a Lewis acid catalyst, and the substituted 1,2-anhydrosugar is preferably utilized in excess as compared to the nucleophile.

Thus, a saccharide multimer linked to a solid phase particulate support can be prepared by glycosylating a particle-linked nucleophile as discussed before with a before- discussed glycosyl donor substituted 1,2-anhydrosugar having a plurality of non-participating substituents including a protected nucleophilic hydroxyl group having a selectively removable protecting moiety. The phases are separated. Protection of the formed 2-hydroxyl group is carried out and the protected nucleophilic hydroxyl group is selectively deprotected.

The free hydroxyl group resulting from that deprotection step is then glycosylated with a glycosyl donor substituted 1,2-anhydrosugar as discussed before, and the solid and liquid phases are separated. Repeating the 2-position protecting, nucleophilic hydroxyl deprotecting, glycosylation and phase separation steps seriatim provides a particle-linked saccharide multimer of the desired length. The present invention further contemplates use of different glycal compounds in the preparation of a saccharide multimer, i.e., an oligosaccharide or a polysaccharide, and the products thereby produced. The saccharide compounds produced have a predetermined stereochemical configuration of glycosidic bonds as well as of individual saccharide substituents. The invention further contemplates a method of preparing particle- linked glycosides, and particularly particle-linked oligo- and polysaccharides that are often referred to herein as saccharide multimers. In one aspect, a glycal-terminated halo-substituted saccharide multimer is prepared. In this method, a glycal derivative having a plurality of substituent groups and a single reactive (nucleophilic) hydroxyl group and a substituted glycal having non-participating electron donating groups are haloglycosylated to form a halo-substituted glycosyl glycal. The substituents on the glycal derivative are electron withdrawing relative to the substituents on the substituted glycal. The haloglycosylation is carried out in the presence of a halonium ion reagent and the absence of water. An electron withdrawing substituent is removed from the glycal portion of the haloglycosylation product and is replaced with an electron donating substituent. The haloglycosylation is repeated with another glycal derivative having a plurality of substituent groups and a single reactive hydroxyl group, the substituents and the other glycal derivative being electron withdrawing relative to the substituents on the glycal portion of the halo-substituted glycosyl glycal. The above prepared glycal-terminated halo-substituted saccharide multimer can itself be used to haloglycosylate an alcohol having a single reactive hydroxyl group other than a glycal. In this case, the product of the before-described reaction, a halonium ion reagent and an alcohol other than a glycal are haloglycosylated. The product of the reaction is preferably recovered. The non-glycal alcohol can be a sugar molecule alcohol, or an alcohol of a non-sugar molecule.

Another aspect contemplated is preparation of a halo-substituted oligosaccharide. Here, a substituted glycal containing five or six atoms in the ring, five through nine carbon atoms in the sugar backbone chain and having a plurality of non-participating electron donating substituents, a halonium ion reagent and a glycal derivative having five or six atoms in the ring, five through nine carbon atoms in the sugar backbone chain, a single reactive hydroxyl group and an electron withdrawing acyl group are reacted in the absence of water to form a halogenated glycal-terminated oligosaccharide reaction product. That reaction product has a halogen group at the 2-position of the glycosyl ring that is trans to a glycosyl bond.

The above reaction product is reacted with an alcohol and a halonium ion reagent in the absence of water to form a further reaction product having a further glycosidic bond and a halogen bonded on a carbon atom adjacent to that further glycosidic bond; i.e., at the 2-position; the further glycosidic bond and halogen are in a trans relation to each other. Preferably, both the substituted glycal and glycal derivative have rings containing six atoms, including the ring oxygenation. Preferred non-participating relatively electron donating substituents are hydrogen, C₁-C₈ alkyl and O-ether groups. Preferred halonium ion reagents are I(syjn-collidine)₂C10₄ and Br(syjn-collidine)C10₄.

A 2-position halogen group of a glycosyl ring can be removed and replaced with another substituent before or after a following haloglycosylation.

A haloglycosyl glycal is prepared in yet another aspect of the invention. Here, a before-described glycal derivative and substituted glycal are reacted in the absence of water and presence of a halonium ion reagent to form the haloglycosyl glycal in which the halide at the 2-position is trans to the glycoside bond. In a further aspect of the invention, a solid phase particle-linked nucleophilic molecule having a single reactive hydroxyl group per molecule, and free from other nucleophilic groups, is haloglycosylated with a liquid composition that is free of water and contains a glycosyl donor substituted glycal having a plurality of substituents and a halonium ion reagent. The substituted glycal and halonium ion reagent are present in about equimolar amounts and in excess relative to the moles of nucleophilic hydroxyl groups. The particle-linked glycoside formed has a halide group at the 2-position of the added (formed) glycosyl ring that is trans to the adjacent glycosidic bond. The solid and liquid phases are separated.

The glycosyl donor substituted glycal includes hydrogen .-Cg alkyl and O-ether substituent groups among its plurality of substituent groups, and preferably includes a protected hydroxyl group whose protecting moiety is selectively removable in the presence of any other substituent that is present, without affecting that substituent. Exemplary of such selectively removable protecting moieties are acyl and trisubstituted silyl moieties. Selective removal of such a protecting moiety (deprotection) to form a particle-linked glycoside having a free nucleophilic hydroxyl group followed by haloglycosylation with a substituted glycal and halonium ion reagent as before, phase separation and repeating of the deprotection, haloglycosylation and phase separation steps seriatim provides a particle-linked saccharide multimer of a desired length.

Here as well, the substituted glycal utilized has a ring of five or six atoms including the oxygen atom and a sugar backbone chain length of five through nine carbon atoms. Preferred halonium ion reagents are I(syjn-collidine),C10₄ and Br(sym-collidine)C10₄.

BRIEF DESCRIPTION OF THE DRAWINGS In the drawings forming a portion of this disclosure: Figure 1 contains two schematic equations for a generalized synthesis of a glycoside. In Equation I, a generalized substituted 1,2-anhydrosugar, in which R is a substituent, is reacted with a nucleophile, NuH, to form a glycosylated nucleophile. In equation II, a generalized glycal, in which R is a substituent, is reacted with an electrophile, E , to form an intermediate X, which can be an isolable intermediate or a non-isolable intermediate.

Intermediate X thereafter reacts with the nucleophile to glycosylate the nucleophile.

Figure 2 is a schematic representation of the reaction of a substituted glucal, in which R can be acetyl (Ac) , benzyl (Bn) or t-butyldimethylsilyl (Me_SiBu or TBS) , with dimethyl dioxirane (Compound 3) to form substituted 1,2- anhydrosugar Compounds 4, 5 or 6, respectively. The scheme goes on to show that when R was acetyl (Compound 4) , methanolysis yielded a mixture of glycosides (Compound 7) . Contrarily, similar methanolysis with the two ether-substituted glycals (Compounds 5 and 6) led to almost completely stereospecific glycoside products (Compounds 8 and 9) . Experimental procedures utilized in obtaining the various compounds shown in this figure and in those that follow are discussed in detail hereinafter.

Figure 3 contains three schematically drawn epoxidation reactions using Compound 3, and the stereospecificity of the reactions. Thus, the t-butyldimethylsilyl (TBS) etherified galactal Compound 10 gave the corresponding substituted -1,2-anhydrosugar 11 in 98 percent yield. The substituted allal Compound 12, which bore one TBS ether group and an acetal formed from benzaldehyde (in which the phenyl ring is shown as Ph) , was converted to the corresponding substituted β-l,2-anhydrosugar, Compound 13, in 98 percent yield. The similarly substituted glucal Compound 14 provided an approximately equal mixture of the corresponding α- and β-anomer substituted 1,2-anhydrosugars, Compounds 15 and 16.

Figure 4 schematically illustrates the results obtained when four of the substituted 1,2-anhydrosugars were used to glycosylate methanol, and the resulting reaction product glycosides were acetylated. Thus, Compound 8, wherein R was benzyl (Bn) provided the glycoside Compound 17 in 97 percent yield; Compound 9, wherein R was TBS, provided glycoside Compound 18 in 87 percent yield; Compound 11 provided glycoside Compound 19 in 87 percent yield; and Compound 13 provided Compound 20 in 95 percent yield. Figure 5 schematically illustrates the synthesis of the oligosaccharide compounds such as 22a, 24a and 26. In the upper scheme, the diisopropylidenesubstituted galactose Compound 21 was glycosylated with Compound 8 in tetrahydrofuran (THF) as solvent and ZnCl₂ catalyst at an initial temperature of -78 degrees C that was permitted to warm to room temperature (rt) to form glycoside Compound 22a. The free hydroxyl group of Compound 22a was thereafter acetylated to form Compound 22b.

The lower scheme illustrates the preparation of a substituted glycal-terminated disaccharide, Compound 24a, followed by the trisaccharide, Compound 26a. Here, the benzyl (Bn)-substituted glucal, Compound 23, was glycosylated with Compound 8 under the above conditions, to form the substituted glycal-terminated disaccharide, Compound 24a. The 2-position hydroxyl of Compound 24a was benzylated with benzyl bromide (BnBr) in the presence of sodium hydride (NaH) to convert the participating 2-hydroxyl group into a non-participating ether group. The resulting glycal was converted to a corresponding substituted 1,2-anhydrosugar Compound 25, using dimethyl dioxirane, Compound 3. Reiteration of the glycosylation step using Compound 23 and the before-described conditions yielded the glycal-terminated trisaccharide, Compound 26a.

Figure 6 is a schematic representation showing a generalized glycosylation of a glycosyl acceptor (A) having a free hydroxyl group (OH) and four protected oxygen appendages (OP and OP' in which P and P¹ are protecting groups) reacting with glycosyl donor D having similarly protected oxygen appendages. The disaccharide formed is shown to undergo transformation of one of its protecting groups, P*. into a leaving group L prior to trisaccharide formation.

Figure 7a is a schematic representation showing a generalized synthetic procedure described in detail herein in which a glycosyl donor substituted glycal, D, and a glycosyl acceptor glycal derivative, A, are reacted with an electrophile, E⁺, to form a glycosyl glycal that incorporates the electrophile at the 2-position of the glycosyl ring. The scheme further shows that the glycosyl glycal can be further reacted with an acceptor hydroxy glycal to form a trisaccharide reaction product. P is a protecting group as noted before. Figure 7b is a schematic representation showing a generalized synthetic procedure described in detail herein in which a glycosyl donor substituted glycal, D, and a particle- linked glycosyl acceptor derivative, A, are reacted with an electrophile, E, to form a particle-linked glycosyl glycoside that incorporates the electrophile at the 2-position of the glycosyl ring. The scheme further shows that the glycosyl glycal can be further transformed into a glycosyl acceptor by conversion of a single OP* group hydroxy into an OH group and then further reacted with a donator glycal to form a trisaccharide reaction product. P is a protecting group as noted before. Figure 8 is a further scheme that illustrates with more specificity a reiterative strategy utilized herein for preparing saccharide multimers such as oligosaccharides and polysaccharides. The substituted glycal glycosyl donor, Compound 101, is shown to contain three electron donating O-ether groups (R = alkyl) whereas the glycal derivative glycosyl acceptor. Compound 102, contains a single reactive hydroxyl group and two electron withdrawing O-acyl groups (R¹ = acyl) . The two glycals are shown to be reacted in the presence of an iodiniu ion to form a glycosy glycal, Compound 103. The O-acyl groups (R') of Compound 103 are removed and replaced with two alkyl groups in Compound 104 (R¹ = alkyl) , thereby transforming the substituents of the glycal ring portion of the glycosyl glycal from being relatively electron withdrawing (R' - acyl; Compound 103) into relative electron donating substituents (R' - alkyl; Compound 104). The iodoglycosylation (I⁺) reaction is run again using Compound 105 having a single reactive hydroxyl group and two relatively electron withdrawing substituents (R¹ ' = acyl) as the glycosyl acceptor to form the trisaccharide. Compound 106. Reiteration of the removal and replacement of the acyl groups on the glycal ring with alkyl groups and iodoglycosylation with another glycosyl acceptor forms a tetrasaccharide product. Each of the glycosyl bonds shown as forming an α-linkage and each iodo group of the glycosides is shown in a β-configuration to emphasize that the iodoglycosylation proceeds by a trans diaxial addition to the glycal double bond and the stereochemical purity of the formed glycosidic bond. Beta glyosidic compounds can also be formed. Figure 9 illustrates the structures of four glycal molecules (Compounds 7-10) and two substituted sugar derivative molecules (Compounds 11 and 12) utilized herein. In these compounds and those discussed hereinafter, Ac is acetyl, Bn is

X CH_g is isopropilidyl. Hydrogen r^~~_

~3 atoms not needed to show stereochemical configuration are omitted in these and the remaining formulas. Figure 10 schematically illustrates three iodoglycosylation reactions. In the first, Compounds 108 and 109 are reacted in 55 percent yield in the presence of (sy_π-collidine)₂I⁺C10₄~ and powdered 4A molecular sieves in dichloromethane as solvent at room temperature, Reaction (i) , to form glycosyl glycal Compound 113. In the upper reaction, Compound 113 was reacted with Compound 111 under the same conditions to form Compound 114 in 77 percent yield and in which X is iodide. The two iodide groups of Compound 114 were replaced with hydrogen atoms to form Compound 115 in 94 percent yield by reaction with triphenylstannane, azobis(isobutyro)- nitrile (AIBN) in refluxing benzene as solvent, Reaction (ii) . The lower reaction shows that Compound 113 was reacted with Compound 112 to form Compound 116 in 67 percent yield using the conditions of Reaction (i) . Details of these syntheses and those of the reaction that follow are provided hereinafter.

Figure 11 is a schematic illustration of the iodoglycosylation of Compounds 108 and 110 using Reaction (i) of Figure 10 to form Compound 117 in 57 percent yield. Compound 117 was thereafter reacted with Compound 111 under the same conditions to form Compound 118 in a 60 percent yield.

Figure 12 schematically illustrates the preparation of a tetrasaccharide, Compound 121. Compound 113 of Figure 10 was first reacted with sodium hydroxide (NaOH) in methanol (MeOH) , Reaction (iii) , to remove the benzoyl groups of the glycal rings. This was done in 86 percent yield. The resulting hydroxyl groups were then replaced with t-butyldimethylsilyl (TBS) ether groups by reaction with t-butyldimethylsilyl chloride in the presence of imidazole as catalyst and dimethyl formamide as solvent, Reaction (iv) , to form glycosyl donor Compound 119 in 90 percent yield. Compound 119 was thereafter reacted with glycosyl acceptor glycal derivative Compound 109 using the conditions of Reaction (i) of Figure 10 to form trisaccharide Compound 120 in 59 percent yield. Resulting Compound 120 was thereafter iodoglycosylated with Compound 111 using the conditions of Reaction (i) of Figure 10 to form Compound 121 in 74 percent yield. Figure 13 is a schematic illustration of the synthesis of tetrasaccharide 124 in which the reaction conditions for Reactions (i) , (iii) and (iv) are as discussed previously. Here, disaccharide glycosyl glycal Compound 117 shown in Figure 11 was converted in 99 percent yield to the diol using Reaction (iii) conditions, and that diol was formed into the di-TBS ether Compound 122 in 79 percent yield using the conditions of Reaction (iv) . Compound 122 was then iodoglycosylated with Compound 109 using the conditions of Reaction (i) to form di-iodo-substituted glycal-terminated trisaccharide Compound 123 in 64 percent yield. Compound 123 and Compound 111 were iodoglycosylated using the conditions of Reaction (i) to form the tri-iodo-substituted sugar derivative-terminated tetrasaccharide Compound 124 in 72 percent yield.

The reaction schemes illustrated in Figures 14 and 15 illustrate the synthesis of the chimeric drug, Compound 140, that is an iodo-di-TBS-ether derivative of cyclimycin 104. Here, as shown in Figure 14, L-fucal, Compound 125, was benzoylated with benzoyl chloride in dichloromethane as solvent containing 4-dimethylaminopyridine (DMAP) and triethylamine (Et₃N) at a temperature of -10 degrees C to room temperature (rt) as Reaction (v) to form Compound 126. L-3-Deoxy-rhamnal (Compound 127) was reacted at zero degrees C with sodium hydride (NaH) in THF and then benzylated with benzyl bromide (PhCH₂Br) at room temperature, collectively Reaction (vi) , to form Compound 128. Compounds 127 and 128 were halogylycosylated using the conditions of Reaction (i) , except dichloromethane was used as solvent and the temperature was initially zero degrees and was allowed to rise to room temperature to form the iodo-substituted glycosyl glycal Compound 129.

The iodide group of Compound 129 was converted to a hydrogen using Reaction (ii) of Figure 10 to form Compound 130. The benzoyl group of Compound 130 was then removed to provide a hydroxyl group by reaction with lithium aluminum hydride in diethyl ether as solvent at zero degrees C, Reaction (vii) to form Compound 131. The hydroxyl group Compound 131 was thereafter made into a TBS ester following Reaction (iv) of Figure 12 to form Compound 132.

Compound 132 was then haloglycosylated using Reaction (i) , as discussed above, in the presence of Compound 126 to form glycal-terminated trisaccharide

Compound 133. The iodide was removed from Compound 133 using Reaction (ii) , as discussed above, to form Compound 134.

Figure 15 illustrates that the benzoyl group of the glycal portion of Compound 134 of Figure 9 was removed using Reaction (vii) , above, to form the corresponding hydroxyl group of Compound 135. That hydroxyl group was thereafter converted into a TBS ether substituent using Reaction (iv) to form Compound 136. The sole benzyl group of Compound 136 was removed by reaction of that compound with the sodium in a solvent of ammonia-THF at a temperature of -78 degrees C to form hydroxyl group-containing Compound 137.

The hydroxyl group of Compound 137 was oxidized to a keto group using Dess-Martin periodinane, according to the provedure of Dess and Martin, J. Org. Chem. 48:4155 (1983), in dichloromethane at an initial temperature that was permitted to rise to room temperature as Reaction (ix) to form Compound 138.

Compound 139 (epsilon-pyrromycinone, a tetracycline aglycone obtained from marchellomycin) and Compound 138 were haloglycosylated using Reaction (i) as discussed in Figure 14 to form chimeric drug Compound 140.

DEFINITIONS The following words and phrases are utilized herein to have the meanings as are normally recognized in the art and as are described below.

A "glycal" is a cyclic enol ether derivative of a sugar having a double bond between carbon atoms at positions 1 and 2 of the ring.

A glycal useful herein contains a chain of 5-9 carbon atoms and is in the form of a cyclic ring having 5 or 6 atoms that are members of the ring.

The terms "sugar" and "sugar derivative" are utilized herein generically to mean a carbohydrate or carbohydrate derivative that contains 5-9 atoms in its backbone chain, and those of interest herein can be a pentose, hexose, heptose, octulose or nonulose.

A "monosaccharide" is a simple sugar that cannot be hydrolyzed into smaller units.

An "oligosaccharide" is a compound sugar that yields two to about 10 molecules of simple monosaccharide on hydrolysis.

A "polysaccharide" is a compound sugar that yields more than 10 molecules of simple monosaccharide on hydrolysis. The phrase "saccharide multimer" is utilized herein to embrace both an oligosaccharide and a polysaccharide.

A sugar molecule or its derivative typically contains a plurality of hydrogen, alkyl, hydroxyl, amine or mercaptan groups in free or protected form such as the N-phthalimido group or the S-acetyl group bonded to each of the carbon atoms of the molecular chain. Where the sugar molecule is in cyclic form, the oxygen atom of one of the hydroxyl groups is utilized as the oxygen that is part of the cyclic ring structure. A deoxy sugar contains a hydrogen atom or other substituent group in place of one of the hydroxyl groups.

Each of the substituents, other than the hydrogen of a deoxysugar, has a particular stereochemical configuration relative to the other substituents and relative to the plane of the cyclic ring. The chain length and stereochemical configuration of the substituent groups provide the basis for the names of the sugars. The sugar molecules and their derivatives utilized herein are of known stereochemical configuration and therefore can be referred to as having a "predetermined stereochemical configuration".

Position numbering in a sugar molecule begins with the aldehydic carbon for aldoses and the terminal carbon atom closest to the keto group for ketoses. Thus, for a glycal, the first carbon of the ethylenic unsaturation adjacent the ring oxygen is numbered "position 1", with the remaining positions being numbered around the ring away from the ring oxygen atom. The carbon atom at position 1 is also referred to as the "anomeric atom" or "anomeric carbon atom" due to the possible formation of anomers at that position. Alpha (c.) anomers are bonded below the plane of the ring in its usually drawn form, whereas in beta (β) anomers, the bond is above the plane of the ring when so drawn. A "glycoside" is a sugar derivative containing a substituent bonded to the anomeric carbon atom. The glycosidic bond can be between the anomeric carbon and an oxygen, nitrogen or sulfur atom and their appropriate other substituents other than hydrogen. A "substituted 1,2-anhydrosugar" is a sugar derivative having an epoxide linkage between the carbon atoms of the 1- and 2-positions of a cyclic sugar derivative. The substituted 1,2-anhydrosugars useful herein are isolable intermediates. The reactions that are described herein include the formation of a glycosyl bond or "glycosylation". In those reactions, one molecule provides the anomeric carbon atom of the glycosyl bond, and another, a nucleophile molecule, provides a nucleophilic atom that becomes bonded to the anomeric carbon atom. The molecule that provides the anomeric atom is referred to as a "glycosyl donor", and is a substituted 1,2-anhydrosugar herein. The nucleophile molecule that provides the nucleophilic atom that becomes bonded to the anomeric atom is referred to as a "glycosyl acceptor". Numerous glycosyl acceptor molecules are contemplated, and they are discussed in detail hereinafter.

The one molecule that provides the anomeric or 1-position carbon atom is referred to as the glycosyl donor, whereas the molecule that provides the nucleophilic group that becomes bonded to the anomeric atom is referred to as the glycosyl acceptor.

As is discussed in greater detail hereinafter, the glycals utilized herein are substituted or derivatized. Nevertheless, for ease in understanding, a glycal that is utilized as a glycosyl donor is usually referred to herein as a "substituted glycal", whereas a glycal that acts as the glycosyl acceptor is referred to herein as a "glycal derivative". In keeping with that nomenclature, the substituted 1,2-anhydrosugar formed from a substituted glycal is usually referred to as a "substituted 1,2-anhydrosugar".

A particularly preferred group of reactions discussed herein utilizes a glycal as the glycosyl acceptor and another glycal as the glycosyl donor. For ease in understanding, the glycal that is a glycosyl acceptor will usually be referred to herein as a "glycal derivative", whereas the glycal that is the glycosyl donor will usually be referred to herein as a "substituted glycal". Another particularly preferred group of reactions discussed herein utilize an alcohol of a sugar derivative as the glycosyl acceptor and a glycal as the glycosyl donor.

Disaccharides are named as a glycosyl glycoside for a nonreducing disaccharide and as a glycosyl glycose for a reducing disaccharide. Larger oligosaccharides are similarly named as a glycosyl glycosyl glycoside for a nonreducing trisaccharide, a glycosyl glycosyl glycosyl glycoside for a nonreducing tetrasaccharide, and a glycosyl glycosyl glycosyl glycose for a reducing tetrasaccharide, and the like. The product of a glycosyation reaction is often referred to as a "derivative" such as a glycosyl derivative or a glycoside derivative.

Thus, using the above glycosyl donor and glycosyl acceptor terminology and applying it to an oligosaccharide, a disaccharide is named by first reciting the glycosyl donor and then the glycosyl acceptor. A longer oligosaccharide is named by first reciting the first glycosyl donor and then the first acceptor. The first glycosyl acceptor or the glycosylated first glycosyl acceptor can also be viewed as a glycosyl donor for the third saccharide unit and so on through the chain until the last saccharide unit remains. That last saccharide unit that is not itself a glycosyl donor to another saccharide unit is referred to herein as the "terminal" saccharide. Thus, a "glycal-terminated" oligosaccharide is an oligosaccharide containing a glycal derivative as its last saccharide unit.

The anomeric carbon atom is thus free from a glycosidic bond to another saccharide unit. A substituted 1,2-anhydrosugar or sugar derivative can also occupy a "terminal" position of an oligosaccharide.

A "1,2-halonium species intermediate" discussed herein is an adduct formed between an electrophilic halogenating reagent (electrophile) and the glycal double bond. A 1,2-halonium species intermediate is formed by the reaction of a halonium ion reagent and a glycal. Such adducts are not capable of isolation under usual laboratory conditions. The glycosylation reactions discussed herein typically proceed via a 1,2-halonium species intermediate. As such, a glycosyl bond is formed at the 1-position of the glycosyl donor as discussed before and the halide of the halonium ion reagent becomes bonded to the 2-position of the glycosyl donor. As a result, these reactions are referred to herein generally as "haloglycosylation" reactions, and more specifically as "iodoglycosylation" or "bromoglycosylation", where I or Br are the halonium ions utilized.

When two substituents are on the same side of the plane of a sugar ring as drawn in a two-dimensional representation, those substituents are referred to as being ⁿsyn" to each other. When two substituents are on opposite sides of a similarly drawn sugar ring plane, they are in the "anti" configuration. The syn and anti nomenclature is usually utilized for substituents separated in a ring by at least one ring carbon atom. Where two adjacent carbons are discussed, ciε and trans are used for two substituents on the same and opposite sides of the ring plane, respectively.

The word "corresponding" is used herein in relation to substituted 1,2-anhydrosugars converted from substituted glycals and to glycosides prepared by using a substituted 1,2- anhydrosugar to glycosylate a nucleophile to mean that the substituents present on the substituted glycal and substituted 1,2-anhydrosugar prior to epoxidation and glycosylation, respectively, except for the glycal double bond and substituted 1,2-anhydrosugar epoxide are present in the same configuration after each respective reaction. In relation to glycosides, the word "corresponding" is used to mean that the substituents present on the substituted glycal prior to haloglycosylation, respectively, except for the glycal double bond and substituted 1,2-anhydrosugar epoxide are present in the same configuration after each respective reaction. Any substituents on the nucleophile except for a hydrogen bonded to the nucleophilic atom are similarly maintained after glycosylation.

DETAILED DESCRIPTION OF THE INVENTION

I. OVERVIEW

A. Introduction

The present invention contemplates the reaction of a substituted glycal derivative to form an isolable substituted 1,2-anhydrosugar that can be utilized in the preparation of a glycoside, and particularly in the preparation of an oligosaccharide or polysaccharide (saccharide multimer) of predetermined stereochemical conformation. Preparation of the isolable substituted 1,2-anhydrosugars was found to proceed readily and in high yield by reaction of a suitable substituted glycal with a dialkyl dioxirane whose alkyl substituents total two to about six carbon atoms as epoxidant. The substituted 1,2-anhydrosugar so formed is then reacted with a nucleophile to form an epoxide-opened glycoside reaction product in which the nucleophilic atom of the nucleophile is bonded to the anomeric atom, and in which the glycosyl bond is trans to the 2-position hydroxyl group formed during opening of the epoxide ring.

The stereoconfiguration of the anomeric atom is inverted by reaction with the nucleophile. The stereochemical configuration of the glycoside product is thus predictable from the known stereochemistry of the substituted 1,2-anhydrosugar.

A variety of nucleophiles can be utilized to form the glycoside. The nucleophilic atom of the nucleophile can be oxygen, nitrogen or sulfur. Oxygen is a preferred nucleophilic atom of a hydroxyl group (alcoholic) nucleophile. In particularly preferred practice, the alcoholic nucleophile is a portion of a sugar molecule so that an oligosaccharide is the glycoside that is formed. When the alcohol-containing sugar molecule is itself a glycal, a glycosyl glycal is formed that can be formed into another substituted 1,2-anhydrosugar and reacted again with a nucleophile to form another glycoside product. Those epoxidation and glycosylation steps can be repeated

(reiterated) over and over again where the nucleophile is a hydroxyl-containing glycal to form longer and longer oligosaccharides or polysaccharides.

The present invention also contemplates use of nucleophile as discussed before linked to a solid phase, particulate support such as resin beads. In such a reaction, the solid phase-bound (particulate-linked) nucleophile is reacted with a liquid composition, typically a solution, that contains a substituted 1,2-anhydrosugar. The substituted 1,2- anhydrosugar glycosylates the nucleophile to form a particle- linked ring-opened glycoside. The solid and liquid phases are separated.

In preferred practice, the substituted 1,2- anhydrosugar includes a protected nucleophilic hydroxyl group whose protecting moiety can be selectively removed

(deprotected) subsequent to the glycosylation step in the presence of any other substituents to form a particle-linked epoxide ring-opened glycoside having a free nucleophilic hydroxyl group. That nucleophilic hydroxyl group is then glycosylated with a substituted 1,2-anhydrosugar to form a ring-opened glycosyl glycoside linked to the particle. The solid and liquid phases are separated, and the glycosylation, deprotecting and separation steps are repeated seriatim as is desired. Thus, according to this invention, the glycoside such as an oligosaccharide is prepared in the direction from the terminal glycosidic bond toward the first glycosyl donor. This direction of synthesis can provide some advantages as are discussed in detail hereinafter. The glycoside compounds formed from the glycosylation reactions discussed herein are varied and have varied uses. For example, reaction of a suitably substituted galactal to form the corresponding α-1,2-anhydrosugar followed by glycosylation of a 4-hydroxy-1,2,3,6-tetra-substituted glucose can provide a substituted lactose whose substituents can be removed to form lactose itself.

Similarly, reaction of a suitably 3,4,6- trisubstituted glucal to form the corresponding substituted

1,2-anhydride derivative that is used to glycosylate 4-hydroxy- 3,6-disubstituted glucal forms a cellobiose derivative. Reiteration of the conversion of the cellobiose terminal glycal to a corresponding substituted 1,2-anhydrosugar and the glycosylation of another 4-hydroxy-3,6-disubstituted glucal provides cellotriose.

As a further example, conversion of a suitably 3,4,6- substituted glucal to form the corresponding 1,2-anhydrosugar can be used to glycosylate a hydroxyl group of a particle. Selective removal of the 4-position protecting moiety, followed by glycosylation with another, similarly substituted 1,2- anhydrosugar forms a particle-linked cellobiose derivative. Reiteration of the removal of the cellobiose 4-position protecting moiety and glycosylation of another similarly substituted glucal provides a substituted cellotriose.

The present invention also contemplates the reaction of polysubstituted glycal compounds to form non-isolable 1,2-halonium species intermediates, which are utilized in the preparation of glycosides, and particularly in the preparation of saccharide multimers such as oligosaccharides, of predetermined stereochemical conformation.

The non-isolable 1,2-halonium species intermediates useful herein provide a route to glycoside compounds via the acceptor (A)/donor (D) route of reaction discussed herein. Thus, the studies described hereinafter were in part organized around a new idea involving coupling of glycals as is shown illustratively in the equation shown in Figure 2.

Manipulations at the anomeric centers following the equations of Figures 2a and 2b were unnecessary since coupling is actuated by attack of the oxidant at the donor glycal.

Moreover, the next reiteration is straightforward since the AD disaccharide formed in Figure 2a is itself a glycal, ready for oxidative actuation as before, once its electron withdrawing groups are removed and replaced with electron donating groups. In the aspect of the invention represented by Figure 2b, the free hydroxyl function in the acceptor glycal needs only to be differentiated from two (rather than four) other hydroxyls or other substituents that must be protected. In this aspect, the next reiteration is also straightforward since the AD disaccharide formed in Figure 2b, upon conversion of an OP' protecting group into a reactive, nucleophilic hydroxyl group, is ready for oxidative reaction as before. The glycoside compounds formed from the haloglycosylation reactions described herein are varied and have varied uses. For example, where an iodonium species intermediate is utilized so the glycoside formed is an iodide, use of a radioactive form of iodine such as ⁴ⁱ³ι or ^OJ-ι in the reaction can provide a valuable tracer for following the metabolism of various biologically active materials such as the chimeric drug, Compound 40, discussed hereinafter, or a glycosylated drug.

Haloglycosylation of a suitably substituted glucal and 1,3,4,6-tetraacetyl fructose can provide a correspondingly substituted sucrose. That substituted sucrose can then be converted into sucrose itself. Maltose can similarly be prepared.

Cyclodextrins are cyclic c.-(l,4) glycosides of glucose that usually contain six to nine glucose units. Cyclodextrins are produced by Bacillus macerans.

Cyclodextrins are increasingly being used in drug delivery systems. However, because of the limited number of natural cyclodextrins available and because bulk reaction conditions permit only subsitutions at a statistical distribution to provide heterogeneous cyclodextrins, cyclodextrin derivatives having a precisely defined number of known substituents per molecule have heretofore not been possible to prepare. Thus, in one embodiment of the invention, haloglycosylation using 4-benzyl-3,6-di(t-butyldimethylsilyl)- glucal and 3,6-diacetylglucal, followed by removal and replacement of the acetyl groups with the above silyl ether (TBS) groups. Likewise, in another embodiment of the invention, haloglycosylation using 4-trimethylsilyl-3,6- dibenzylglucal and 3,6-dibenzyl glucosyl-particle, followed by removal and replacement of the trimethylsilyl group with a hydroxyl group. In each case, repeating the haloglycosylation and removal and replacement steps four, five, six or seven times provides a polyhalo-substituted glycal- or glucosyl- terminated particle-linked oligosaccharide from which the benzyl group can be removed, or, in the second case, that can be severed from the particle and from which the benzyl group can be removed. After removal of that benzyl group, self haloglycosylation provides a substituted cyclodextrin having 6- 9 glucoside units. Removal and replacement of the halide substituents provides the corresponding cyclodextrin. Selection of the replacing substituents in the above reaction scheme prior to each haloglycosylation or at the iodide replacement step permits one to prepare cyclodextrin molecules each of which contains a known number of known, preselected substituents per ring. Thus, homogeneous cyclodextrins can now be prepared.

It is well known that many viral, bacterial and mammalian proteins are glycosylated at serine or threonine hydroxyl groups or on the side chain amido nitrogen atoms of asparagine and glutanine. Such glycosyl groups can form antigenic determinants, and indeed, one of the first synthetic vaccines was prepared against a disaccharide, a cellobiuronic acid derivative. Goebel, Nature 143:77 (1939).

Heretofore, it has only been possible to prepare synthetic immunogens that were oligosaccharides as above, or synthetic polypeptides, in which the antigenic determinant was known. See Sutcliffe, et al . , Science 219:660 (1983). Antibodies could also be raised to glycoproteins, but the antigenic determinant bound by antibodies was unknown.

In accordance with the present invention, a glycal- terminated oligosaccharide of known size and structure can be prepared that can be used to glycosylate or haloglycosylate a protected amino acid such as N-t-BOC-serine methyl ester to form a glycosylated serine. That glycosylated serine can thereafter be utilized to form a glycosylated synthetic polypeptide immunogen whose antigenic determinant is known. In yet another embodiment, nucleosides and their analogues can be prepared. For example, α- and β-anomers of thymidyl riboside derivatives have been prepared, as have substituted glucoside, galactoside and alloside thymidyl analogues.

Glycosylated molecules are also known to be useful in enhancing the bioavailability of biologically active molecules. Thus, an oligosaccharide of known sequence and stereochemistry can be prepared as described herein and utilized to glycosylate the amino-terminal or an amine- or hydroxyl-derivatized carboxy-terminal amide residue or an internal serine, threonine, asparagine, glutamine, hydroxylysine or hydroxyproline residue of a biologically active molecule such as the position 61 and/or 62 (also referred to as positions 5 and 6) gluta ines of the tachykinin substance P, the 5-position serine or 7-position glutamine of kassinin, or the T cell differentiating polypeptide referred to as serum thymic factor at the 4-position serine and/or 5-position glutamine.

In view of the predictability and stereospecificity of the reactions described herein, the saccharide multimers and particularly the oligosaccharides produced by a described method can be utilized as an industry standard for such saccharide multimers and oligosaccharides. The cellobiose and cellotriose that can be used as such standards have already been noted.

The present invention can also be utilized to prepare blood group antigens such as the type A antigen discussed in U.S. Patent No. 4,195,174, whose disclosures are incorporated by reference. The type A antigen of that patent can be prepared as follows.

A substituted 1,2-anhydrosugar such as Compound 11 is reacted with a 3,6-diprotected glucal to form a corresponding substituted lactal. Formation of a triflate ester at the newly formed galactosyl 2-position followed by replacement with another nucleophile/leaving group such as chloride ion to invert the configuration at that position and then reaction of that leaving group with a protected nitrogen atom nucleophile such as phthalide ion provides the required 2-position amino group in protected form and proper configuration.

Conversion of the above-substituted lactal to its corresponding substituted α-l,2-anhydrosugar as discussed herein followed by glycosylation of an 8-methoxycarbonyloctyl- 3,6-diprotected-2-deoxy-2-N-protected-galactopyranose, a compound that can be prepared in a manner similar to compound XL of the above patent, provides the required second β-linkage, the third, terminal, saccharide unit and a means to link the ultimately formed antigen to an antigenic carrier molecules, i.e., the octanoylmethyl ester. The formed trisaccharide also includes an α-2-position hydroxyl group.

Reaction of the 2-hydroxyl group of the above trisaccharide with a 3,4-diprotected-L-fucal under the conditions described in Lemieux, et al . , Can. J. Chem. 43:2190 (1965) provides the tetrasaccharide of the desired type A blood group having a bromo or iodo group on the fucosyl ring, depending on the halonium perchlorate utilized. The halide group can be replaced with a hydroxyl group to form a protected type A antigen. Deprotection and replacement of the phthalimido groups with N-acetyl groups provides the completed antigen.

Also in accordance with the present invention, a serine, threonine, lysine, hydroxylysine or hydroxyproline residue linked to a solid support as part of an otherwise protected polypeptide can be reacted with a substituted 1,2- anhydrosugar to form a glycosyl polypeptide. Removal of a protecting moiety from the substituted glycosyl portion of the solid phase-linked glycosyl polypeptide and further glycosylation as described herein provides an oligosaccharide that is terminated with a polypeptide. That glycopeptide can then be deprotected and severed from the support to provide a useful glycopeptide antigen or immunogen whose antigenic determinant is known.

Thus, instead of forming an antigen, the glycosylated polypeptide can itself be biologically active, using the techniques discussed above, the 5-position of the serine of the tachykinin kassinin can be glycosylated as can the 4-position serine of the T cell differentiating polypeptide referred to as serum thymic factor.

In a similar manner, the N-terminal amine of a particle-linked protected polypeptide can serve as the particle-linked nucleophile that is glycosylated. For example, a biologically active molecule such as α- or β-endorphin can be prepared in protected form linked to a resin and having a t-butoxycarbonyl (t-BOC) amino-terminal residue, which in the case of both endorphins is a tyrosine residue. Removal of the t-BOC group, repeated deprotection and glycosylation as disclosed herein, followed by cleavage of the prepared glycopeptide from the particle supprt and deprotection of the saccharide and polypeptide portions of that formed molecule, provides a useful glycopeptide.

Thus, a glycal-terminated oligosaccharide of known size and structure can be prepared that can be used to haloglycosylate a protected amino acid such as N-t-BOC-serine methyl ester to form a glycosylated serine. That glycosylated serine can thereafter be utilized to form a glycosylated synthetic polypeptide immunogen whose antigenic determinant is known. Also, a serine, threonine, hydroxylysine or hydroxyproline residue linked to a solid support as part of an otherwise protected polypeptide can be haloglycosylated with a substituted glycal to form a glycosyl polypeptide. Removal of a protecting moiety from the substituted glycosyl portion of the solid phase-linked glycosyl polypeptide and further glycosylation as described herein provides an oligosaccharide that is terminated with a polypeptide. That glycopeptide can then be deprotected and severed from the support to provide a useful glycopeptide antigen or immunogen whose antigenic determinant is known.

Glycosylated molecules are also known to be useful in enhancing the bioavailability of the glycosylated acceptor portion of the molecules. Thus, an oligosaccharide of known sequence and stereochemistry can be prepared as described herein and utilized to haloglycosylate an internal serine, threonine, asparagine, glutamine, hydroxylysine or hydroxyproline residue of a biologically active molecule such as the 5-position serine of the tachykinin kassinin or the T cell differentiating polypeptide referred to as serum thymic factor at the 4-position serine. Also, in the case of glycosylated polypeptides, the glycosylated polypeptide can itself be biologically active instead of forming an antigen. Using the techniques discussed above, the 5-position serine of the tachykinin kassinin can be glycosylated as can the 4-position serine of the serum thymic factor. In view of the predictability and stereospecificity of the reactions described herein, the saccharide multimers and particularly the oligosaccharides produced by a described method can be utilized as an industry standard for such saccharide multimers and oligosaccharides. Yet another use for the present invention is in the preparation of what can be called chimeric drugs. Such chimeric drugs contain a terminal glycoside bond formed from an anthraσycline aglycone and a glycal-terminated oligosaccharide. These drugs are referred to as chimeric because the aglycone portion and oligosaccharide portions are usually not found together in a naturally occuring drug. That is not to say, however, that a naturally occuring drug could not be made following the techniques described herein.

An exemplary synthesis of such a chimeric drug is described herein and is shown schematically in Figures 9 and

10. Here, the anthracycline aglycone portion was obtained from the drug marchellamycin, whereas the oligosaccharide is an iodo-di-t-butyldimethylsilyl ether derivative of the oligosaccharide portion of the drug ciclimycin 4. Bieber, et al . , J. Antibiotics XL:1335 (1987).

B. General Glycoside Synthesis

As noted previously, the present invention is concerned in one aspect with the preparation of glycosides having predetermined stereochemical configurations. Exemplary of such glycosides are nucleosides wherein the glycosidic bond is formed between a nitrogen atom of a nucleic acid base such as guanine or thymine or a derivative of either and the anomeric carbon of a sugar such as ribose. A further type of useful glycoside is that formed from an azide ion or a side chain amido nitrogen of a glutamine or asparagine in a polypeptide, or blocked amino acid. Yet another glycoside is that formed between a sulfur atom of a nucleophile such as thiophenol or a blocked cysteine and a substituted glycal. Most of the glycosides used illustratively herein are formed between an alcohol hydroxyl group portion of a molecule as a glycosyl acceptor and the anomeric carbon atom of a substituted 1,2-anhydrosugar as glycosyl donor. A further type of useful glycoside is that formed from a nucleophilic nitrogen atom as is present on the amino-terminus of a particle-linked, protected polypeptide or a side chain amino nitrogen of a lysine residue in a polypeptide, or blocked amino acid. Yet another glycoside is that formed between a sulfur atom of a nucleophile such as an otherwise protected cysteine and a substituted 1,2-anhydrosugar.

The power of the synthetic approach that is embodied in this invention is illustrated particularly in the preparation of otherwise difficultly prepared oligosaccharides, such as those containing two, three and four saccharide units. Simpler glycosides are also illustrated using a wide variety of hydroxyl group-containing molecules, as well as in formation of a nucleoside derivative. The basic approach for the preparation of a glycoside having predetermined stereochemical configuration is as follows:

A glycal containing five, six, seven, eight or nine carbon atoms in its backbone chain having a ring that contains five or six atoms is provided. This glycal includes a plurality of substituents that include hydrogen, 3-position hydroxyl, protected hydroxyl, protected amino and protected mercapto groups, and each of those substituents has a predetermined stereochemical configuration relative to the glycal ring.

The glycal is admixed and reacted with a reagent to convert it into a stable, generally isolable substituted 1,2- anhydrosugar. That reagent is preferably a dialkyl dioxirane whose alkyl groups total two to about six carbon atoms.

A solid phase-linked nucleophile that contains a nucleophilic atom such as oxygen, nitrogen or sulfur is also provided. This nucleophile is glycosylated in a liquid composition containing a suitably substituted 1,2-anhydrosugar to form a particle-linked epoxide ring-opened glycoside. The 1,2-anhydrosugar utilized contains a plurality of non- participating substituents. The solid and liquid phases are thereafter separated.

The epoxide ring-opening glycosylation reaction forms a glycosidic bond between the anomeric carbon atom and the nucleophilic atom. That reaction also forms a hydroxyl group that is trans to the glycosidic bond at the adjacent 2-position of the glycosyl ring that is formed or added to the nucleophile.

If desired, that 2-position hydroxyl group can be utilized as the nucleophile in a subsequent glycosylation reaction. Usually, however, when further glycosylations are to be carried out, that 2-hydroxyl group is converted to a non- participating substituent, as is discussed hereinafter, and a protecting moiety present on a protected, nucleophilic hydroxyl group substituent is removed to provide a free nucleophilic hydroxyl group that is utilized in a subsequent glycosylation reaction.

The substituted 1,2-anhydrosugar is typically and preferably prepared from a corresponding substituted glycal.

As noted hereinafter, substituted 1,2-anhydrosugars prepared from glycals whose plurality of substituents are protected by O-ether linkages have been made in very high yield and very high stereochemical purity. Isolation and/or purification of a substituted 1,2-anhydrosugar is not however required, although at least isolation is preferred.

The fact that the substituted 1,2-anhydrosugars can be readily isolated cannot be over emphasized. Previously used oxidation conditions such as oxidation with a peracid have led to low yields because of subsequent reaction of the formed epoxide with the carboyxlic acid produced. Still further, use of O-acyl substituents as in Brigl's anhydride (3,4,6- triacetyl-l,2-anhydroglucose) that stabilize the epoxide ring can lead to mixed configurations due to acyl anchimeric assistance and also typically require more time to run due to the increased stability of the epoxide ring.

This method of stereospecifically preparing substituted 1,2-anhydrosugars is believed to be new in itself. In addition, the one step epoxidation of a glycal as discussed has never been reported. The substituted 1,2-anhydrosugar so prepared is then utilized to glycosylate a nucleophile to form an epoxide-opened reaction product in which the nucleophilic atom of the nucleophile is bonded to the anomeric atom of the reacted substituted 1,2-anhydrosugar. The nucleophile utilized can have an oxygen, nitrogen or sulfur nucleophilic atom, although oxygen and nitrogen are preferred. The resulting glycoside reaction product can illustratively be a simple 0-methyl glycoside where methanol is the nucleophile, a nucleoside where a nitrogen of a nucleic acid base was the nucleophilic atom, or the glycoside can be a substituted oligosaccharide.

A particularly preferred substituted oligosaccharide reaction product is a substituted glycosyl glycal whose substituted glycal portion can be converted to a substituted 1,2-anhydrosugar that is utilized to glycosylate a hydroxyl group of a substituted glycal to form a further substituted glycal-terminated oligosaccharide. Those conversion and glycosylation steps can be repeated (reiterated) again and again as is desired to prepare a continually lenthening oligosaccharide. Illustrative glycosides herein are also formed between an alcohol hydroxyl group of one glycal and the anomeric carbon atom of the substituted glycal that serves as the glycosylating agent. Thus, both the substituted glycal that serves as the glycosyl donor (D) and the hydroxyl- substituted glycal derivative that serves as the glycosyl acceptor are present in the reaction medium at the same time, and the substituents present on such glycal govern whether the reaction will proceed in a straightforward manner, as is discussed hereinafter. Alternatively, the glycosides are formed between a particle-linked nucleophilic alcohol hydroxyl group and the anomeric carbon atom of the substituted glycal that serves as the glycosylating agent. The power of the synthetic approach is illustrated in the preparation of otherwise difficultly prepared oligosaccharides, such as those containing two, three and four saccharide units.

The basic approach for the preparation of a glycoside having a predetermined stereochemical configuration is as follows:

A glycosyl donor substituted glycal is provided having a ring that contains five or six atoms and a backbone chain of 5 through 9 carbons. The provided substituted glycal includes a plurality of non-participating, relatively electron donating substituents that include hydrogen, C,-C₆ alkyl, phthalimido, S-acyl and protected hydroxyl, typically an ether substituent, and each of those substituents has a predetermined stereochemical configuration relative to the glycal ring. The substituents of the provided glycal are referred to as "protected" in that a protected substituent does not undergo a reaction during formation of glycoside, whereas an unprotected substituent can react, although it need not do so. A substituent is preferably protected with a relatively easily and relatively selectively removable protecting moiety so that the protected hydroxyl, amino or mercapto group can be regenerated as desired, and possibly reprotected again, all without substantial loss of the product glycoside or alteration of the configuration about the anomeric, glycosidic bond. Exemplary protecting moieties and their uses are discussed hereinafter.

A substituted glycal is haloglycosylated by reaction with a nucleophile such as a nucleophilic alcohol of a glycal, or a particle-linked nucleophile such as a particle-linked nucleophilic alcohol of a sugar derivative, preferably having relatively electron withdrawing substituents and a halonium ion reagent to form a haloglycoside derivative having a glycosidic bond to the nucleophilic oxygen atom of the alcohol and a halogen group from the halonium reagent at the 2-position of the formed glycoside derivative.

The synthesis route to glycoside compounds using a non-isolable 1,2-halonium species intermediate as illustrated in the Figure 7 equation, before, is thus examined.

For that equation to be viable, there must be available a menu of stereospecific oxidative coupling reactions, wherein E, the electrophile, is readily translatable to a relevant substituent group such as a hydroxyl group, ether group, a ine group or a hydrogen. Moreover, if the coupling is to occur via a fugitive intermediate (some version of an onium species) , rather than through an isolable compound, such as a 1,2-anhydrosugar, it is crucial that the two glycals or saccharide species that are present in the reaction medium assume strictly defined glycosyl donor and acceptor roles.

The reaction first considered for oxidative coupling was haloetherification. See Lemieux, et al . , Can. J. Chem. 43:2190 (1965); Lemieux, et al . , Can. J. Chem. 42:532, 539 (1964); and Lemieux, et al . , Can J. Chem. 43:1460 (1965). The oxidatively triggered addition of alcohols to glycals via presumed 1,2-iodonium ion formation was known. See Thiem, Synthesis , 696 (1978) ; Thiem, Chapter 8, ACS Symposium Series 386, Horton, et al. , eds., American Chemical Society (1989). The question remained, however, whether the nucleophile (i. e. , the glycosyl acceptor) could itself be a glycal. It was soon established that the order of presentation of two similar glycals (for example Compounds 107 and 109 of Figure 9) to the oxidizing agent was of no useful consequence. A complex mixture of products was obtained. Of course only that glycal derivative that bears a free hydroxy group can be the acceptor (A) component. In principle, however, either glycal can function as a donor. Fortunately, it was found that the glycosyl donating tendencies of the hydroxyl-bearing glycal derivative can be suppressed relative to the substituted glycal that lacks a free hydroxyl group.

This is accomplished when the intended glycosyl acceptor containing the free hydroxyl group contains substituents that are relatively electron withdrawing. Preferably, the acceptor glycal derivative has hydroxyl protecting groups such as acyl protecting groups, whereas the intended donor (with no free hydroxyl group) is substituted with relatively electron donating hydroxyl protecting groups such as ether functions. Preferably, a glycosyl acceptor glycal contains one reactive, nucleophilic hydroxyl group and all of its remaining substituents are electron withdrawing O-acyl substituents (esters) , whereas a preferred glycosyl donator glycal contains only relative electron donating substituents such as O-ether substituents. (The concept of arranging the nature of the protecting groups to control the susceptibility of N-pentenyl glycosides toward electrophilic attack in a different system free from glycals has recently been described. Mootoo, et al . , J. Am. Chem. Soc. 110:5583 (1988) . This methodology has been applied to a very interesting new construction of oligosaccharides.

When a 1:1 mixture of a preferred donor and a preferred acceptor glycal as discussed above was presented to the oxidizing agent, the resulting AD disaccharide was assembled with strict regiochemical and stereochemical control (Figure 8, Compound 101 + Compound 102 → Compound 103). To reiterate the scheme, i.e., repeat the reaction to lengthen the oligosaccharide with another glycal, the electron withdrawing, e . g. acyl, protecting groups were removed, and the resulting hydroxyl groups were reprotected as ethers (see Compound 103 → Compound 104) .

The resulting reprotected AD substituted glycal Compound 104 thereby became a glycosyl donor with respect to diacyloxymonohydroxy glycal Compound 105 in the formation of another glycal-terminated oligosaccharide (Figure 8, Compound 104 + Compound 105 → Compound 106) . In this way, a trisaccharide was readily produced.

Use of a particle-linked hydroxyl group as the initial nucleophilic glycosyl acceptor as described herein in such a reaction has not been reported.

The particle-linked intended glycosyl acceptor containing the free hydroxyl group also contains substituents that are relatively electron withdrawing or electron donating. Preferably, the acceptor derivative has hydroxyl protecting groups such as acyl or ether protecting groups, whereas the intended donor (with no free hydroxyl group) is substituted with relatively electron donating hydroxyl protecting groups such as ether functions. A particle-linked glycosyl acceptor contains one reactive, nucleophilic hydroxyl group and all of its remaining substituents are protected.

The halgoglycosylating coupling reactions, presumably involving a 1,2-iodonium ion intermediate, occur in a clean, trans 1,2-diaxial manner to afford substantially only one glycoside. The process can be iterated further following the deprotecting, and haloglycosylation steps discussed above to prepare longer multimer saccharides such as the cyclodextrins.

II. COMPOUNDS AND METHODS

A. The Glycals

A useful substituted 1,2-anhydrosugar is preferably, although not necessarily, prepared from a corresponding substituted glycal. That being the case, substituted glycals provide an appropriate starting place for discussion.

As already noted, the useful glycals are substituted cyclic sugar derivatives that contain a double bond between the carbon atoms at glycal ring positions 1 and 2. The glycal rings themselves contain five or six atoms including oxygen, whereas the chain that makes up the backbone of the sugar derivatives can have 5 through 9 carbon atoms. Naturally occurring sugars containing 5 through 9 carbon atoms in their backbones are exemplified by ribose, galactose, sedheptulose, manno-2-octulosic acid and sialic acid. Glycals containing five, six and nine carbon atoms in the chain are preferred.

The substituted glycals utilized herein are prepared by well known techniques. Some substituted glycals used herein such as Compounds 12 and 14 were known compounds, while others, such as Compound 107, are available commercially. See also Danishefsky, et al . , J. Am. Chem. Soc , 110:3929 (1988), and the citations therein. A useful glycal is referred to as being substituted. Cyclic sugar molecules are themselves hydroxyl-substituted tetrahydropyran or tetrahydrofuran derivatives. A useful glycal can therefore be considered to be a dihydropyran or a 2,3-dihydrofuran derivative. In view of the differing nomenclatures that can be utilized for sugar derivatives, and for clarity of expression, a hydroxyl group as would normally be present in a sugar molecule is referred to herein as a substituent of a glycal as well as a substituent of the various other sugar derivatives discussed herein. Similarly, hydrogen, which is normally present on a dihydropyran or 2,3- dihydrofuran, but is a substituent that replaces a hydroxyl of a sugar, is also referred to herein as a substituent.

Each glycal thus has a plurality of substituent groups. Exemplary, useful substituent groups include hydrogen, hydroxyl, C.-C₈ alkyl, as well as protected hydroxyl, protected mercaptan and protected amine groups. Of the above substituents, non-participating hydrogen, hydroxyl, C.-C₈ alkyl and protected hydroxyl (O-ether) groups are preferred. Protected substituents are those that are reacted with another reagent to form a substituent that does not undergo reaction under the conditions of conversion of a substituted glycal to a substituted 1,2-anhydrosugar or during the subsequent glycosylation step, as are discussed hereinafter, or the haloglycosylation conditions, as noted earlier. Most preferably, as where a glycoside such as an oligosaccharide, or a haloglycoside such as a halo-substituted oligosaccharide, is to be prepared that does not have protecting groups, a preferred protecting group is readily removable to provide the unprotected substituent with little or no alteration of the stereochemical configuration of the substituent or glycoside bonds.

Various types of ether-forming readily removable protecting groups are preferred for hydroxyl substituents, so that the substituents themselves are ethers. Particularly preferred readily removable ether linkages are benzyl or ring- substituted benzyl ethers having 7-10 carbon atoms, diaryl-C,- Cg alkylsilyl ethers such as diphenylmethylsilyl ether, aryl- di-C.-C₈ alkylsilyl ethers such as a phenyldimethylsilyl ether, and tri-C-_j-C₈ alkylsilyl ethers such as trimethy1silyl and t- butyldimethylsilyl ethers. Acetals and ketals are also considered to contain ether linkages since each contains the C-O-C bond of an ether. Acetals and ketals formed from aldehydes or ketones containing 1 to about 12 carbon atoms such as formaldehyde, acetone, cyclohexanone, 1-decanal and 5- nonanone or an aromatic aldehyde such as benzaldehyde or naphthaldehyde or an aromatic ketone such as acetophenone. Acetone, formaldehyde and benzaldehyde are preferred for the preparation of ketals and acetals, respectively. Additional useful readily removable protecting groups are discussed in Kunz, Angew. Chem. Int. Ed. Engl . 26:294 (1988), whose disclosures are incorporated by reference. The above-described readily removable nonparticipating substituents can be removed by a number of means well known in the art. For example, the benzyl ether- type protecting groups can be removed by hydrogenolysis over a palladium catalyst or by sodium or lithium is liquid ammonia. The various silyl ethers can be removed by reaction with tetrabutylammonium fluoride. The acetals and ketals can be removed with mild acids.

The protecting moiety of each of the above-described ether protecting group types can be removed in the presence of the other protecting group substituents. For example, a benzyl group can be removed to provide a free hydroxyl group in the presence of a silyl ether or an acetal or ketal. Tri-C,-C₆ alkylsilyl such as trimethylsilyl moieties are preferred protecting moieties for removal to provide a free nucleophilic hydroxyl group after a glycosylation or haloglycosylation step has been completed and before the next glycosylation or haloglycosylation is carried out. Where a 2-position hydroxyl is to be protected after a glycosylation step, it is preferred that that hydroxyl be benzylated and another protecting moiety such as a trimethylsilyl group be thereafter removed to provide the nucleophilic hydroxyl group. Where a 2-position halide is to be replaced after a haloglycosylation step, it is preferred that the halide be replaced prior to removal of the hydroxyl protecting moiety (deprotection) .

A useful substituted glycal can itself also be an oligosaccharide whose terminal saccharide unit is a substituted glycal. Thus, an ether linkage can also be present between the substituted glycal and another substituted monosaccharide or oligosaccharide. This ether linkage is preferably between the anomeric carbon atom of the substituent sugar derivative and the glycal, although such a glycosidic bond is not required. In this instance, the substituents on the glycal are as described before along with one or more substituted saccharide units as further substitutents; those substituted saccharide substituents also having substituents as described before. Thus, a substituted glycal can be a substituted glycal- terminated oligosaccharide.

For example, a di- or trisaccharide terminated with a glycal can itself be a starting substituted glycal useful herein. Compounds 24b and 24c, described hereinafter, exemplify a glycal-terminated disaccharides that were used to form a substituted 1,2-anhydrosugar-terminated disaccharide, Compound 25, that was itself thereafter used to glycosylate a hydroxyl group of a particle-linked nucleophile as well as another glycal to form a glycal-terminated trisaccharide derivative, Compound 26. Compound 119 exemplifies a glycal- terminated disaccharide that was haloglycosylated with the hydroxyl group of a glycal derivative to form a glycal- terminated trisaccharide derivative. Compound 120. A glycal-terminated oligosaccharide can be prepared by various techniques well known to skilled workers. It is neither always necessary nor desired that an ether substituent be readily removable. To that end, an ether can contain a C-_^-C-_^g alkyl, ^C ₆~C₁₀ aryl or substituted aryl or non-benzyl C-,-C₁₀ aralkyl group bonded through an oxygen atom to the ring of a derivatized sugar molecule. Exemplary of such groups are methyl, ethyl, isopropyl, cyclohexyl, lauryl and stearyl ethers, as well as phenyl, p-tolyl, 2-naphthyl, ethylphenyl, and 4-t-butylphenyl ethers. Oligosaccharides are another group of ether substituents that are not readily removed.

In another method, a substituted glycal glocosyl donor as described herein and a glycosyl acceptor glycal derivative having a ring containing five or six atoms, five through nine carbon atoms in the sugar backbone chain, a single reactive hydroxyl group and substituents that are relatively electron withdrawing such as are provided by a C-_^-C-_^g O-acyl group or an O-carbamyl ester group having a total of 1-14 carbon atoms such as is formed from ethyl, tolyl isocyanate or 1-(1-naphthyl)ethyl isocyanate are haloglycosylated as described herein. Removal and replacement of the electron withdrawing group with an electron donating group such as an O-ether linkage permits the substituted glycosyl glycal formed to function as a glycosyl donor for a further reaction with a glycosyl acceptor glycal derivative so that a desired glycal-terminated oligosaccharide can be prepared and utilized. The before-discussed ethers are collectively referred to herein as "O-ethers" or the molecules are referred to as having "O-ether linkages", "O-ether groups" or "O-ether substituents".

Other non-participating groups include C.-C₆ alkyl, hydrogen and 3-position hydroxyl groups. Exemplary C_^-C₈ alkyl groups include methyl, ethyl, isopropyl, sec-butyl, cyclopentyl and n-hexyl groups.

When the substituted glycal is oxidized to form a substituted 1,2-anhydrosugar, it is important that the substituents not also be oxidized. To that end, acyl protecting groups such as those formed by the reaction of a C-- C₁₈ alkanoic acid such as acetic, stearic or cyclohexanoic acids or an aromatic acid such as benzoic, or 1- naphthaleneacetic acid (or formed from an anhydride, acid chloride or activated ester such as a N-hydroxysuccinimido ester of such an acid); i.e., a

acyl group, and a hydroxyl, mercaptan or amine group are also useful. A cyclic imide containing a total of 4-10 carbon atoms and 5-7 atoms in the imido ring such as succinimido, substituted succinimido such as methylsuccinimido, phthalimido or a substituted phthalimide like 4-chlorophthalimido and a cyclic amide having 4-10 carbon atoms and 5-7 atoms in the amido ring such as pyrrolidinyl, valerolactamyl or caprolactamyl are also useful. It is noted, however, that the stereoselectivity of the glycosylating or haloglycosylating reaction discussed herein can be undermined or altered by using protecting groups on the glycal or donor glycal that can participate anchimerically. O-Acyl- and N-acyl protected substituents are such "participating" substitutents present particularly at the 4- and 6-positions of glucopyranose derivatives, as are the unprotected 3-, 4- and 6-hydroxyl groups themselves. If present on the glycal, a participating group is removed prior to the oxidative conversion and glycosylation steps, and is replaced with a "non-participating" substituent such as a protecting ether group as discussed herein. Participating groups are likewise removed prior to the haloglycosylation step, and are replaced with "non-participating" substituents such as ether linkages or cyclic imido groups such as phthalimido groups, respectively. A substituted glycal is thus free of participating groups on the glycal ring.

A glycosyl donor substituted glycal is free from particpating groups on the glycal ring at the time the glycal is converted into a corresponding substituted 1,2-anhydrosugar. Thus, the substituted glycal ring has only non-participating substituents. The same is true for a substituted 1,2- anhydrosugar. However, where the substituted glycal contains a saccharide substituent, that substituent can include substituents that might participate if they were bonded to the glycal ring. Acyl protecting groups and cyclic imide and cyclic amide protecting groups as discussed before are preferred protecting groups for mercaptan and amine substituents, respectively, during the oxidation step, including the oxidation step in which the substituted 1,2-anhydrosugar is prepared. Neither is considered a participating substitutent. Where an oxygen atom is linked to a glycal ring from which a substituted 1,2-anhydrosugar is to be prepared, it is preferable to remove any participating substituent such as C»- ₁₈ ^ac-⁷-¹- substituent, and replace it with a non-participating substituent such as one of the before discussed silyl ethers prior to converting the glycal to a substituted 1,2- anhydrosugar. Although a substituted glycal can include one or more

O-acyl substituents, it is preferred that such substituents be absent from a substituted glycal utilized in a haloglycosylation reaction with a glycal derivative. The reason for this preference is that O-acyl groups are relatively electron withdrawing so that the presence of one or more electron withdrawing O-acyl groups on the glycosyl donor substituted glycal tends to diminish the distinction between the donor and acceptor glycals present during haloglycosylation, or slow the rate of haloglycosylation. Substituents on a glycosyl acceptor glycal derivative, in addition to the reactive, nucleophilic, hydroxyl group, are preferably relatively electron withdrawing groups such as a before-described C,-C_lg acyl group. Such acyl groups are preferably present as O-acyl esters. The possibly "participating" character of such groups when present on a glycosyl donor glycal is not relevant for the glycosyl acceptor glycal.

Another relatively electron withdrawing substituent that can be present on a glycal derivative is an O-carbamate ester as is formed from an isocyanate having a total of 1 to about 14 carbon atoms and a ring hydroxyl group. Exemplary isocyanates include methyl, phenyl, tolyl and 1-(1-napthyl)ethyl isocyanates.

Additional substituents that can be present on the glycosyl acceptor glycal derivative are those that are present on the glycosyl donor. Those groups include hydrogen, C--C₈ alkyl, ^s~^cτ~^c ₃ acyl, phthalimido, and the like as discussed before, as well as CN, 2-furyl and O-carbamyl esters such as phenyl and ethyl carbamates as are formed from their respective isocyanates and a hydroxyl group.

It may thus appear that contrary to the previous discussion, the glycosyl donor and glycosyl acceptor glycals can be the same. That is not the case since the glycosyl acceptor glycal derivative must contain substituent groups that are relatively more electron withdrawing than are the substituents of the glycosyl donor substituted glycal, or one could also say that the substituents of the substituted glycal are relatively more electron donating than are the substitutents of the glycal derivative, including the single hydroxyl group of the glycal derivative.

Whether a glycal reacts cleanly as a glycosyl donor or a glycosyl acceptor in a haloglycosylation can be predicted by summing the Hammett sigma constants for para substituents for each glycal and then comparing the sums. Thus, a glycosyl donor (substituted glycal) in such a comparison has a sum of sigma constants that is relatively more negative than that for the sum of the sigma constants of the glycosyl acceptor (glycal derivative) . Put differently, the sum of the Hammett sigma constants for para substituents of a glycosyl acceptor glycal derivative is greater than is the sum of the sigma constants of the glycosyl donor substituted glycal. Thus, the substituents on one glycal are referred to as being relatively more or less electron donating or withdrawing because the sum of the electron donating or withdrawing effects is considered. A partial listing of published Hammett sigma constants for para substituents can be found in Hine, J. , Physical Organic Chemistry, 2nd ed. , McGraw-Hill Book Company, Inc. , New York, page 87 (1982) . Where a published sigma value for a particular substituent cannot be found, the skilled worker can approximate the value and its sign (positive or negative) . Hydrogen and N-acyl groups each are normally considered to be neither electron withdrawing nor donating since the Hammett sigma value system arbitrarily places hydrogen at zero and N-acyl groups typically also have zero sigma constant values for the para position. However, relative to an O-acyl group, both hydrogen and N-acyl are electron donators. It is surprising that the above relation is useful here since Hammett sigma values were developed for totally different systems, and those for para substituents are normally considered to include a resonance effect that should not be applicable here. However, use of such sums is predictive of the results obtained herein.

Although O-acyl groups can be present on the nucleophilic glycosyl acceptor, it is preferred that such groups be absent. This stems from the fact that O-acyl groups can be hydrolyzed during subsequent reactions. The preferred substituents of a glycosyl donor substituted glycal are the preferred substituents of a particle-linked nucleophilic glycosyl acceptor when that acceptor is a sugar derivative.

A useful glycosyl donor substituted glycal can therefore be represented by the chemical formula

wherein:

R is selected from the group consisting of OH, H,

^C1"^C ₆ ^alkY^A 2-furyl, OR⁴, NR⁵R⁶ and SR⁶;

R2 i.s selected from the group consi.sting of H, C,-Cg alkyl, 2-furyl, OR⁴, NR⁵R⁶ and SR⁶; or R 1 and R2, together form a cyclic acetal or ketal prepared from an aldehyde or ketone containing 1 to 12 carbon atoms;

R 3 is selected from the group consisting of H, C-,-Cg alkyl, (CH₂)_mOR⁴, 2-furyl, OR⁴, NR⁵R⁶, SR⁶,

(CH₂ ) _mOR⁴ , CH (OR⁴) CH (OR⁴ ) (CH₂OR⁴ )

HC CH

or R and R , or and R , form a cyclic acetal or ketal prepared from an aldehyde or ketone containing 1 to 12 carbon atoms; n is zero, 1, 2, 3 or 4 with the proviso that the number of carbon atoms in the glycal chain is not greater than 9, such that when m is zero, R is OR⁴, whereas when m is 1-4, m represents the number of methylene groups present; n is zero or 1 such that when n is zero, R 2CH is absent, and when n is 1, R CH is present;

R is selected from the group consisting of C--C,_g alkyl, C₈-C₁₀ aryl, C₇-C₁₀ aralkyl, tri-C-_^-C₈ alkylsilyl, diaryl-C-.-C₈ alkylsilyl, aryldi-C-- C_g alkylsilyl, and a substituted mono- or oligosaccharide; R is selected from the group consisting of C -C₁₈ alkyl, Cg-C,₀ aryl, _-C₁₀ aralkyl, tri-C-^Cg alkylsilyl, aryl di-C-_j-Cg alkylsilyl, diaryl C-- Cg alkylsilyl, and C_η-C.g acyl;

R is selected from the group consisting of H, C.-

C₁₈ alkyl, C₇-C₁₀ aralkyl, and ^-C^ acyl with the proviso that (a) at least one of R or R of NR⁵R⁶ and SR⁶ is C^C^ acyl, or (b) NR⁴R⁵ together form a cyclic amide or imide containing five to seven atoms in the ring and a total of 4-10 carbon atoms;

R 7 and R8 are i.ndependently H or C.-C₉ alkyl with the

7 proviso that the number of carbon atoms n R plus those of R is nine or fewer; and

R is selected from the group consisting of hydrogen (H) , or C-^Cg alkyl, C0₂R⁴, CN, CH₂OR⁴, 2-furyl,

R¹⁰ is selected from the group consisting of hydrogen

(H) , 2-furyl, or C^-C₈ alkyl. A glycosyl acceptor glycal derivative can also be represented by the above chemical formula with the additional features that R⁴ also includes H, and C_η-C_ηg acyl, R also includes CN, O-carbamyl ester having a total of 1-14 carbon atoms, CH(OH)CH₂0R⁴, CH(OR⁴)C__₂0H, R¹ and R² can also include CN and O-carbamyl ester having a total of 1-14 carbon atoms, and the proviso that at least one of R 1, R2 and R3 includes an OH group.

Examples of each of the above-described R groups, where the R can be more than a single group, are provided in the previous discussion. It is noted that for purposes of counting carbon atoms in a chain of a substituted glycal and 1,2-anhydrosugar, a C-,-C₈ alkyl and 2-furyl groups are considered to be substituents and not part of the sugar backbone chain.

In preferred practice, the stereochemical configuration of each of the substituent groups is known at the time the substituted glycal is either converted into the 1,2- anhydrosugar derivative or haloglycosylated. It is also

Q If) preferred that and R be hydrogen, and that all remaining substituent groups be non-participating. Those remaining substituents for the substituted glycal are preferably selected from the group consisting of hydrogen, C_^-Cg alkyl, 3-position hydroxyl and O-ether substituents, as defined before, whereas a glycal derivative or a hydroxyl-group-containing particle- linked nucleophile must include a hydroxyl group, with the preference for other substituents also being hydrogen, C,-C₈ alkyl and O-ether groups as discussed for the glycal donor. In the above formula, n is preferably 1 so that a preferred substituted glycal contains a ring that contains six atoms. This preference also extends to a substituted 1,2- anhydrosugar. Thus, a preferred substituted glycal and preferred substituted 1,2-anhydrosugar contain substituents at the 3-, 4- and 6-positions, and those substituents are preferably O-ether substituents. B. The Substituted 1,2-Anhydrosugars

A substituted 1,2-anhydrosugar useful herein is preferably prepared by oxidation of a substituted glycal discussed before with a dialkyl dioxirane as is discussed hereinafter. It is noted, however, that in a method aspect of the invention discussed hereinafter, one can begin utilizing an otherwise prepared substituted 1,2-anhydrosugar to glycosylate a hydroxyl-substituted glycal or a nucleophile, and proceed from the glycosyl glycal or glycosyl derivative so prepared. Thus, using well known techniques, the skilled worker can prepare a substituted 1,2-anhydrosugar by means other than the conversion of a substituted glycal. An exemplary reaction for formation of such a substituted 1,2-anhydrosugar is a dehydrohalogenation reaction as is discussed in Sondheimer, et al . , Carbohydr. Res. 74:327 (1979).

Regardless of its method of preparation, a substituted 1,2-anhydrosugar can be represented by the chemical formula

0

wherein R 1, R2 , R3 , R9 and R10 are as before described for a glycosyl donor substituted glycal, i.e., a substituted 1,2- anhydrosugar is free of participating substituent groups.

Preferably, however, a before-described substituted glycal is converted into a substituted 1,2-anhydrosugar by reaction with a dialkyl dioxirane having a total of two to about six carbon atoms in the dialkyl groups; i.e., the sum of the carbon atoms in both alkyl groups is 2 to about 6. Exemplary dialkyl dioxiranes include dimethyldioxirane, diethyldioxirane, methyl isopropyldioxirane, methyl propyldioxirane, ethyl sec-butyldioxirane and the like. 3,3-Dimethyldioxirane, Compound 3, (also referred to herein as dimethyl dioxirane) contains a total of two carbon atoms in its alkyl groups and is utilized illustratively herein. Compound 3 is particularly preferred as the dialkyl dioxirane because its reaction product, acetone, is relatively readily removable from the reaction mixture, and because the side reactions that occur with peroxides such as peracetic acid have not been observed. This finding was in accord with the results of others using the dioxirane with non-sugar derivative ethylenic derivatives. See Montgomery, J. Am. Chem. Soc. 96:7820 (1974); Edwards, et al . , Photochem. Photobiol . 30:63 (1979); Curci, et al . , Org. Chem. 45:4758 (1980); Murray, et al . , J. Org. Chem. 50:2847 (1985); Baumstark, et al . , Tetrahedron Lett. 28:3311 (1987); Baumstark, et al . , J. Org. Chem. 53:3437 (1988); and Adam, et al . , J. Org. Chem. 52:2800 (1987).

The stereochemistry of this conversion appears to proceed by adding the oxygen atom of the formed epoxide ring cis to the olefinic unsaturation, and from the less sterically hindered side of the glycal ring. Thus, the 1,2-epoxide ring can be or β to the substituted saccharide ring that is formed.

As will be seen hereinafter, the reaction with a substituted glycal and a dialkyl dioxirane can proceed in almost quantitative yield to provide a single product. In some instances, both - and β- forms of the substituted 1,2- anhydrosugar are formed.

The stereochemical purity of the formed substituted 1,2-anhydrosugar appears to be a function of steric hindrance to approach of the dialkyl dioxirane. Thus, where the upper side of the substituted glycal is relatively hindered as is case for Compounds 6 and 10, only the α-epoxides were formed, whereas for Compound 12, where the upperside of the glycal ring is relatively unhindered and the lower side is relatively hindered, the β-epoxide was formed.

Where both sides of the ring are about equally hindered, about equal amounts of CL- , β-epoxides were formed as was the case for Compound 14. Where relative hindrance is unequal, as in Compound 5, the conversion of a substituted glycal to the corresponding substituted 1,2-anhydrosugar proceeds to form one isomer in greater abundance than the other (α:β, 20:1). A similar finding was made with a 5-membered ring derivative that led to an approximately 6:1 ratio of β- to - anomers of 3-t-butyldimethylsilyl-l,2-anhydroribose when the 3- position hydroxyl was left unprotected as compared to a substantially only β-epoxide when both the 3- and 5-position hydroxyl groups were protected with t-butyldiphenylsilyl groups.

Hindrance of one or the other side of a glycal can be predicted by the configuration of the various ring substituents. Thus, relatively large, bulky groups tend to take up an equatorial configuration relative to the ring.

As a consequence, for a substituted glycal such as Compounds 5 or 6, the substituents at the 4-and 5-positions are trans to each other and tend toward the trans diequatorial conformation that inhibits attack of the epoxidant from the upper side of the ring. That conformation permits epoxide formation from the lower side, thereby providing the relative stereochemical purity observed in a substituted 1,2- anhydrosugar such as Compounds 8 and 9, respectively.

The conversion of a substituted glycal to the analogous substituted 1,2-anhydrosugar is carried out at a temperature such as at about -40 to about +20 degrees C, and typically at about zero degrees C. A non-reactive (inert) solvent such as acetone, methylene chloride-acetone or the like is utilized. Preferably the solvent boils at less than about 100 degrees C so as to facilitate isolation and recovery of the substituted 1,2-anhydrosugar by merely removing the solvent, as with a stream of dry nitrogen or under reduced pressure and at a temperature below about 20 degrees C so that the otherwise reactive epoxide ring does not react prematurely. The substituted 1,2-anhydrosugar is utilized to glycosylate a nucleophile. As such, it is desirable, though not necessary, to recover the substituted 1,2-anhydrosugar. The substituted 1,2-anhydrosugar should, however, be reacted with the nucleophile in the absence of the dialkyl dioxirane, since that reagent can sometimes oxidize the nucleophile.

Thus, if the substituted 1,2-anhydrosugar is not to be isolated, it is preferable to use stoichiometric amounts of substituted glycal and dialkyl dioxirane, or a slight excess of the former reactant. If an isolation step is utilized, an excess of the dialkyl dioxirane is typically used.

Where a substituted glycal to be converted to a substituted 1,2-anhydrosugar includes a participating substituent, that substituent should removed and replaced with a non-participating substituent prior to the reaction of the substituted 1,2-anhydrosugar with the nucleophile to glycosylate the nucleophile. Preferably, the participating substituent is removed and replaced prior to conversion of the substituted glycal to the substituted 1,2-anhydrosugar. Thus, a substituted 1,2-anhydrosugar contains a plurality of only the before-discussed non-participating substituents when reacted with the nucleophile.

C. The Halonium Reagent

Reactants used to form a useful 1,2-halonium species intermediate are well known in the art. The halogens utilized herein are bromine and iodine that form a bromonium and iodonium ions, respectively. It is preferred to utilize an iodomium ion intermediate formed by the reaction of

(2,4,6-collidine)₂IC10₄ or a bromonium ion intermediate formed by the reaction of (2,4,6-collidine)BrC10₄ in a solvent inert to the reaction conditions and that also is relatively readily removable such as dichloromethane or chloroform. 2,4,6-Collidine is also referred to as sy-n-collidine. N-Bromo- or N-iodosuccinimide can also be utilized as the halonium ion reagent as can other well known halonium ion reagents in a similar solvent. The halonium ion reagent utilized illustratively herein is (sy -collidine)₂IC10₄. The nucleophilic alcohol and halonium ion add in a trans diaxial manner across the double bond of the substituted glycal to form a 2-deoxy-2-(halo-substituted) glycoside such as a 2-deoxy-2-bromo- or 2-deoxy-2-iodoglycoside in which the glycosyl bond and the bond to the halide are in a trans diaxial orientation.

Inasmuch as the 1,2-halonium species intermediate is formed under oxidative conditions and in the presence of the nucleophile, the nucleophile must not react substantially under those oxidizing conditions. Alcohol hydroxyl groups, and particularly aliphatic hydroxyl groups, are substantially inert to the conditions utilized herein for formation of a 1,2-halonium species intermediate. Many amines and most mercaptans react under the conditions of 1,2-halonium species intermediate formation, and amines and mercaptans are consequently not utilized as the nucleophile for formation of a glycoside when the glycosidic bond to be formed is formed from a 1,2-halonium species intermediate.

D. Glycosylation

The substituted 1,2-anhydrosugar is thus an active intermediate that is utilized to glycosylate the nucleophilic atom of a nucleophilic molecule. As already noted, nucleophilic atoms contemplated herein are oxygen, nitrogen and sulfur. In preferred practice, there is a single nucleophilic atom per nucleophile molecule.

Nitrogen- and sulfur-containing nucleophiles are typically utilized as a glycosyl acceptor in the reaction with a terminal substituted 1,2-anhydrosugar. These nucleophiles are usually reacted with a terminal 1,2-anhydrosugar. When a substituted 1,2-anhydrosugar is formed by oxidation of a substituted glycal, the reaction can be carried out separately from the reaction to form the glycoside derivative, and nitrogen and sulfur atoms present in protected form or as the nucleophiles to be glycosylated can be present during the oxidation reaction. This fact provides one of the advantages of the present invention.

Exemplary nucleophiles are the nitrogen atoms of nucleic acid bases such as adenine, guanine, uracil, cytosine, thymine and hypoxanthine and their protected derivatives, and the amido nitrogen atoms of asparagine and glutamine, and their protected derivatives as are utilized in solid phase peptide synthesis, such as N-t-butoxycarbonyl (t-BOC) asparagine methyl ester or N-carbobenzoxy (CBZ) glutamine methyl ester. Further exemplary nucleophiles are the nitrogen atoms of amino- terminated amino acid residues of particle-linked polypeptides and epsilon-amino groups of lysine residues present in otherwise blocked particle-linked polypeptides. Sulfur- containing nucleophiles include thiophenol and cysteine derivatives as are used in polypeptide synthesis such as t- BOC-Cys or CBZ-Cys as their methyl esters, as well as cysteine residues and particle-linked polypeptides.

The usually used nucleophilic atom is an oxygen atom of a nucleophilic alcohol. Substantially any nucleophilic alcohol can be utilized, and that wide variety is exemplified herein by methanol, cholesterol and menthol, as well as the hydroxyl groups of otherwise protected sugar derivatives. The hydroxyl groups of serine, threonine, hydroxylysine and hydroxyproline, and their protected derivatives as are used in solid phase peptide synthesis and noted before are also useful, and are contemplated nucleophiles. A preferred nucleophilic hydroxyl is that of a sugar derivative, and most preferably of a nucleophilic hydroxyl of a substituted glycal or a particle-linked glycoside derivative. The most nucleophilic hydroxyl groups are generally primary hydroxyls as compared to secondary or tertiary hydroxyl groups, and that generality holds for the hydroxyl groups of sugar derivatives.

Reducing sugars contain a single primary hydroxyl group that is located on the carbon atom that is furthest down the chain from the aldehydic (anomeric) carbon atom; i . e. , on the 5-, 6- and 7-position carbon atoms of a pentose, hexose and heptose, respectively. Nonreducing sugars (ketoses) contain two primary alcohols, at the 1-position and at the 6-, 7-, 8-, and 9-positions, respectively, for hexuloses, heptuloses, octuloses and nonuloses, such as sialic acid derivatives. Thus, the primary hydroxyl group of a sugar derivative is a particularly useful and preferred nucleophile. However, a secondary hydroxyl group of a sugar derivative is also nucleophilic, and can also be glycosylated as contemplated herein. Where a branched oligosaccharide is desired, two or three hydroxyl groups of an individual sugar can be glycosylated separately or at once by use of one substituted 1,2-anhydrosugar or different substituted 1,2-anhydrosugars used as the glycosylating agent. When such branching is desired, and different glycosylating agents are utilized, it is preferred to use at least one selectively removable protecting group on the glycosyl acceptor sugar so that selectivity of the products can be achieved.

A particularly preferred nucleophilic hydroxyl group for certain aspects of this invention is that of a glycal derivative because glycosylation of a glycal permits preparation of a terminal glycosyl glycal that can be converted to a correspondingly substituted 1,2-anhydrosugar that can be used to glycosylate another nucleophile such as a further glycal to thereby prepare a lengthened oligosaccharide. This process of conversion of the terminal glycal of an oligosaccharide into a substituted 1,2-anhydrosugar followed by glycosylation of a glycal hydroxyl group can be reiterated or repeated several times as is discussed further hereinafter.

It is preferred that there be only one reactive nucleophilic atom on the nucleophile that is glycosylated by the substituted 1,2-anhydrosugar, regardless of the identity of the nucleophilic atom. Thus, for example, where a hydroxyl group of a sugar derivative is glycosylated, that sugar derivative contains only one reactive hydroxyl group, and any other hydroxy groups that may be present on the glycosylated sugar are present as protected hydroxyl groups, as have been previously discussed.

The protecting groups of the glycosylated molecule such as a sugar derivative can be participating or non- participating substituents, whereas all of the substituents on the substituted 1,2-anhydrosugar are non-participating. Thus, any of the previously mentioned substituents that could be present on a substituted glycal can be present on the nucleophile at the time of glycosylation.

A substituted 1,2-anhydrosugar intermediate is preferably reacted with the nucleophile in the presence of a Lewis acid catalyst in an appropriate solvent that is inert to the reactions used and at a temperature of from about minus 100 degrees C to about 40 degrees C, and preferably at a temperature of from about -78 degrees C to about room temperature (e. g. , about 22 degrees C) . Lewis acids are well known to those skilled in the chemical arts, and useful Lewis acids include SnCl₄, AgC10₄, BF₃, trimethylsilyl trifluoromethanesulfonate (triflate), Zn(triflate)₂, Mg(triflate)₂, MgCl₂, A1C1-,, ZnBr₂ and ZnCl₂ (zinc chloride), with zinc chloride being preferred. Tri-n-butyltin salts of the glycal alcohol glycosyl acceptor have also been successfully utilized.

As is also well known, Lewis acid catalysts are typically utilized in ethereal or chlorinated solvents such as diethylether (ether) , tetrahydrofuran (THF) and dimethoxyethane, or methylene chloride and chloroform.

Care is also taken to avoid conditions in which the substituted 1,2-anhydrosugar polymerizes, or at least to minimize any polymerization that does occur. See Sharkey, et al . , Carbohydr. Res. 96:223 (1981). Care is also taken to avoid reaction conditions in which the Lewis acid catalyst removes a protecting group as can occur with BC1-,.

A Lewis acid catalyst appears to be required for most neutral nucleophiles such as alcohols, amines and mercaptans, whereas negatively charged nucleophiles such as azide ion or trimethylsilyl thiophenoxide did not require Lewis acid catalysis. Methanol used as a solvent could be utilized without an added Lewis acid catalyst. However, reactions using iso-propanol, t-butanol, benzyl alcohol and the sugar hydroxyl groups discussed herein required Lewis acid catalysis.

Post glycosylation manipulations of substituents including inversion of configuration and exchange of one substituent for another are also contemplated herein. For example, as is shown in Figure 5, the 2-position hydroxyl of Compound 24a was protected with a benzyl group with retention of configuration to provide Compound 24b that was subsequently converted to its corresponding substituted 1,2-anhydrosugar. In another example, the configuration of the 2- position hydroxyl group formed from opening the epoxide ring can be inverted or substituted with another substituent after a glycosylation reaction is completed. Thus, the 2-position hydroxyl of the glucosyl portion of Compound 24a can be trifluoromethylated and then reacted with phthalimide to form a nitrogen-containing substituent and a mannosyl derivative that could thereafter be converted to a corresponding substituted 1,2-anhydrosugar and used to glycosylate Compound 8. A 2-position or other hydroxyl group can also be eliminated from a glycoside using well known techniques. One such technique is the so-called Barton deoxygenation reaction in which the hydroxyl group to be removed is made into a xanthate ester that is thereafter reduced off the sugar ring as with tributylstannane. See, for example, Barton, et al . ,

Tetrahedron 42:2329 (1986); and Hartwig, Tetrahedron 39:2609 (1983) .

Still futher, the substituted allal Compound 13 can be used to form an α-glycoside. If, however, a substituted mannosyl α-glycoside were desired, the t-butyldimethylsilyl ether at the 3-position can be removed with tetrabutylammonium fluoride. The resulting hydroxyl group can then be made into a triflate ester, and that ester replaced with a hydroxyl that can be reprotected, if desired, to provide the desired mannosyl α-glycoside.

E. Haloglycosylation

The substituted 1,2-halonium species intermediate is thus an active intermediate that is utilized to glycosylate the nucleophilic atom of a nucleophilic molecule. As already noted, the nucleophilic atom contemplated herein is oxygen, there is preferably a single nucleophilic oxygen atom per molecule of nucleophile.

The usually used nucleophilic atom is an oxygen atom of a nucleophilic alcohol. Substantially any nucleophilic alcohol can be utilized such as methanol, cholesterol and menthol, as well as the hydroxyl groups of otherwise protected sugar derivatives. The hydroxyl groups of serine, threonine, hydroxylysine and hydroxyproline, and their protected derivatives as are used in peptide synthesis such as

N-t-butoxycarbonyl(t-BOC)serine methyl ester or N-carbobenzoxy (CBZ) threonine methyl ester are contemplated nucleophiles with which to terminate a glycoside. A preferred nucleophilic hydroxyl is that of a sugar derivative, and most preferably a nucleophilic hydroxyl of a substituted glycal so that a saccharide multimer can be prepared. The most nucleophilic hydroxyl groups are generally primary hydroxyls as compared to secondary or tertiary hydroxyl groups, and that generality holds for the hydroxyl groups of sugar derivatives.

Reducing sugars contain a single primary hydroxyl group that is located on the carbon atom that is furthest down the chain from the aldehydic (anomeric) carbon atom; i . e. , on the 5-, 6- and 7-position carbon atoms of a pentose, hexose and heptose, respectively. Nonreducing sugars (ketoses) contain two primary alcohols, at the 1-position and at the 7-, 8- and 9-positions, respectively, for heptuloses, octuloses and nonuloses such as sialic acid.

Thus, the primary hydroxyl group of a sugar derivative is a particularly useful and preferred nucleophile. However, a secondary hydroxyl group of a sugar derivative is also nucleophilic, and can also be haloglycosylated as contemplated herein.

Where a branched oligosaccharide is desired, two or three hydroxyl groups of an individual sugar can be haloglycosylated separately or at once by use of one substituted glycal or different substituted glycals used as the glycosyl donor. When such branching is desired, and different glycosyl donor molecules are utilized, it is preferred to use at least one selectively removable protecting group on the glycosyl acceptor sugar so that selectivity of the products can be achieved. A particularly preferred nucleophilic hydroxyl group is that of a glycal derivative because glycosylation of a glycal permits preparation of a halogenated terminal glycosyl glycal that can be used to haloglycosylate another nucleophile such as a further glycal to thereby prepare a lengthened oligosaccharide. This process of haloglycosylating a glycal hydroxyl group with the terminal glycal of an oligosaccharide can be reiterated or repeated several times as is discussed further hereinafter. It is preferred that there be only one reactive nucleophilic atom on the nucleophile that is glycosylated by the substituted glycal. Thus, for example, where a hydroxyl group of a sugar derivative is glycosylated, that sugar derivative contains only one reactive hydroxyl group, and any other hydroxyl groups that may be present on the glycosylated sugar are present as protected hydroxyl groups, as have been previously discussed, or are used to link the sugar derivative to the particle, which linkage also acts as a protecting group. Post haloglycosylation manipulations of substituents including inversion of configuration and exchange of one substituent for another are also contemplated herein. For example, removal and replacement of a participating electron withdrawing O-acyl group with a non-participating electron donating O-ether substituent has already been noted.

In another example, the configuration of the 2-position halo group formed from a haloglycosylation can be inverted or substituted with another substituent after a haloglycosylation reaction is completed. For example, the 2-position iodide of the mannosyl portion of a glycoside such as Compound 113 can be reacted with phthalimide to form a nitrogen-containing substituent and an α-glucoside derivative that can thereafter be deprotected to form a glucosamine derivative. A 2-position halide can also be readily converted into a hydrogen substituent by reaction of the haloglycoside with tributyltin hydride in the presence of azobis(isobutyro)nitrile in refluxing benzene.

As noted previously, haloglycosylation proceeds by 1,2-trans diaxial addition of the nucleophile and halogen across the double bond of the substituted glycal. This reaction proceeds in high yield (typically about 60 to about 90 percent) to provide a single product.

The stereochemical purity of the corresponding glycoside thus formed appears to be a function of steric hindrance to approach of the nucleophile and oxidant. Thus, where the upper side of the substituted glycal is relatively hindered as is case for Compounds 108, 113, 118 and 120, and only the α-anomers were formed, whereas for the L-fucal and L- 3-deoxy-rhamnal compounds discussed hereinafter, where the upperside of the glycal ring is relatively unhindered and the lower side is relatively hindered, the β-glycoside is formed. Hindrance of one or the other side of glycal can be predicted by the configuration of the various ring substituents. Thus, relatively large, bulky groups tend to take up an equatorial configuration relative to the ring.

For example, large substituents at the 4- and 5- positions, such as in substituted glycals such as Compounds 8 or 13, tend to assume a trans diequatorial conformation that inhibits attack of the nucleophile from the upper side of the ring. That conformation permits glycoside formation from the lower side, thereby providing the relative stereochemical purity observed in reaction products such as Compounds 13 and 16, respectively, and of α-anomer formation.

Haloglycosylation is carried out in a solvent that is inert to the reaction conditions employed such as dichloromethane, chloroform or an ether such as diethyl ether or tetrahydrofuran. Water is absent from the reaction and the solvents are therefore dry. It is preferred that an excess of the substituted glycal and halonium ion reagent be present relative to the nucleophilic hydroxyl group. A wide ratio of about 2:1 to about 10:1 of glycal:alcohol is preferred, with the halonium ion reagent and glycal being present in about equimolar amounts. Molecular sieves (4A) are preferably utilized in the reaction medium formed by the solvent, substituted glycal and nucleophile to assist in keeping water absent, but their use is not mandatory. The reaction is run at a temperature of about -20 to about +40 degrees C, and preferably at a temperature of about zero to about ambient room temperature, i. e. , about 22 degrees C.

F. Solid Phase Particles

The glycosylation reactions of this invention preferably are carried out with a nucleophile linked to a solid phase such as are utilized in the solid phase syntheses of oligo- and polypeptides or oligo- and polynucleotides. The solid supports contemplated are particulate materials. The nucleophile is linked directly to the solid phase support. That linkage can be through a direct bond and the linkage must, of course, be inert to the reaction conditions, but is preferably capable of being cleaved or severed when desired so that the produced nucleophile-terminated glycoside can be separated from the support.

Benzyl ether linkages are preferred routes of bonding the nucleophile to the solid phase support, particularly where the nucleophile is a sugar hydroxyl group, and reactions used to form such linkages in particles are well known. Linking groups between the nucleophile and particles are also contemplated. One such group that is also useful for sugar nucleophiles is the 3-aminopropanol group that can be reacted with a benzyl halide-containing particle as discussed below to provide a primary hydroxyl group that can be haloglycosylated with a substituted glycal as discussed herein or by other well known means. Terminal glocosides such as that which is formed are readily cleavable, as is well known.

Several solid supports containing covalently linked reactive functionalities have been described in the chemical and biochemical literature, and any such support can be utilized so long as the solid support is insoluble in water, and in the organic solvents utilized, and is substantially chemically inert to the reaction conditions utilized. The solid support preferably swells in the solvents utilized during synthesis due to physical, rather than chemical processes.

The solid supports typically fall into one of three general types, each of which is discussed briefly below in view of the well known nature of these materials. Perhaps the most utilized particles are polymerized resins. The polymerized resins are generally in the form of porous beads.

Of the resins, the hydrophobic polymerized styrene cross-linked with divinyl benzene (typically at about 0.5 to about 2 weight percent) resins are exemplary. The resin beads so prepared are further reacted to provide a known quantity of a benzyl moiety as a portion of the polymerized resin. The benzyl moiety contains a reactive functional group through which the glycal or substituted 1,2-anhydrosugar can be covalently linked by a selectively severable bond. The before- noted linkers are also readily utilized with benzyl halide- substituted resins. Although the reactive benzyl moieties are typically added after the resin bead has been synthesized by reaction of a polymerized styrene moiety, such resins are generally described as polymerized styrene crosslinked with divinyl benzene and including a known amount of polymerized vinyl benzyl moiety. The reactive functionality of the benzyl moiety is typically halobenzyl such as chlorobenzyl. Polymerized, cross- linked styrene resins containing chlorobenzyl moieties are sometimes referred to in the art as chloromethyl styrene resins. Resins containing a known amount of chlorobenzyl moieties can be purchased from Sigma Chemical Co., St. Louis, Missouri, U.S.A., under the trademark name Merrifield's Peptide Resin (chloro ethylated co-polystyrene divinylbenzene) . Such materials are typically supplied containing about 0.1 to about 2 milliequivalents of chlorine per gram of particle, and have a particle size of about 200-400 mesh. The size of the particles utilized can also be larger such as 40 or 60 mesh, U.S. Standard Sieve Series.

A second group of solid supports is based on silica- containing particles such as porous glass beads and silica gel. For example, Parr and Grohmann, Angew. Chem. Internal . Ed. 11:314 (1972) reported on the use of the reaction product of trichloro-[3-(4-chloromethylphenyl]propylsilane and porous glass beads (sold under the trademark PORASIL E by Waters Associates, Framingham, Massachusetts, U.S.A.) as solid support for polypeptide syntheses. The above solid support is seen to utilize a reactive benzyl moiety through which the sugar derivative is bonded to the particle.

The third general type of useful solid support may be termed composites in that they are constituted by two major ingredients, a resin and another material that is also substantially inert to the organic synthesis reaction conditions employed. One exemplary composite reported by Scott, et al . , J. Chrom. Sci . 9:577 (1971) utilized glass particles coated with hydrophobic, polymerized, cross-linked styrene containing reactive chloromethyl groups and was supplied by Northgate Laboratories, Inc., of Hamden, Connecticut, U.S.A. This particulate support also utilizes a halobenzyl group for attachment to the sugar derivative.

The nucleophile or the linking group is reacted with the particulate support under usual benzylation conditions to form a particulate support-linked substituted nucleophile. The nucleophile is linked to the linking group, thereafter, as is appropriate.

Any remaining halomethylbenzyl groups of the support are reacted as with a primary alcohol such as methanol or a tertiary amine such as triethylamine. The particle-linked nucleophile is thereafter ready to use.

G. Methods

The present invention can thus be seen to contemplate several methods, some of which revolve about the conversion of a substituted glycal into a corresponding substituted 1,2- anhydrosugar thereof, typically followed by the glycosylation of a nucleophile with substituted 1,2-anhydrosugar that has only non-substituents on the ring to form a corresponding glycoside. Others revolve about the glycosylation of a nucleophile that is preferably linked to a solid phase support, preferably by a severable bond or link, with a substituted 1,2- anhydrosugar that has only non-participating substituents on the epoxide-containing ring to form a corresponding epoxide ring-opened glycoside. The products of all of those reactions are preferably recovered, although product can be utilized without recovery per se . Unless other wise specified, the nucleophile is any nucleophile discussed herein, as is the substituted 1,2-anhydrosugar any substituted 1,2-anhydrosugar discussed herein, although those prepared using a dialkyl dioxirane are preferred.

One aspect of the invention contemplates a method of preparing a glycoside derivative. In accordance with that method, a substituted glycal having a ring containing 5 or 6 atoms and a plurality of nonparticipating substituents is converted into a substituted 1,2-anhydrosugar intermediate by reacting that substituted glycal with a dialkyl dioxirane containing a total of two to about six carbon atoms in the alkyl groups. A nucleophile is glycosylated by being reacted with the epoxy ring of the substituted 1,2-anhydrosugar. That nucleophile has a nucleophilic atom selected from the group consisting of oxygen, nitrogen and sulfur. The glycosylation is carried out in the absence of water, and preferably in the presence of a Lewis acid catalyst to form an epoxideopened glycoside reaction product in which the nucleophilic atom is bonded to the formed glycosyl anomeric atom; i. e. , the carbon atom at the 1-position of the reacted glycosyl donating substituted 1,2-anhydrosugar derivative.

Another aspect contemplates a method of preparing a particle-linked glycoside derivative. In accordance with that method, a particle-linked nucleophile is glycosylated by being reacted in a liquid composition with the epoxy ring of a substituted 1,2-anhydrosugar. That nucleophile has a nucleophilic atom selected from the group consisting of oxygen, nitrogen and sulfur. The substituted 1,2-anhydrosugar has a ring containing 5 or 6 atoms, a plurality of non-participating substituents and a chain of 5-9 carbon atoms. The glycosylation is carried out in the absence of water, and preferably in the presence of a Lewis acid catalyst to form an epoxide-opened glycoside reaction product in which the nucleophilic atom is bonded to the formed glycosyl anomeric atom; i.e., the carbon atom at the 1-position of the reacted glycosyl donating substituted 1,2-anhydrosugar derivative. The solid and liquid phases are thereafter separated.

The non-participating substituents of the glycal that are also present on the substituted 1,2-anhydrosugar are preferably those discussed previously; i. e. , 3-position hydroxyl, hydrogen, C.,-C₈ alkyl and O-ether substituents. The nucleophile that is glycosylated is also preferably as discussed previously. Most preferably, the nucleophile is the hydroxyl group of a sugar derivative so that the reaction product glycoside is an oligosaccharide. The starting substituted glycal or 1,2-anhydrosugar can itself be an oligosaccharide such as a glycosyl glycal or glycosyl 1,2-anhydrosugar, or a glycosyl glycosyl glycal or glycosyl glycosyl 1,2-anhydrosugar, in which case the substituted glycal or substituted 1,2- anhydrosugar is an oligosaccharide derivative, in the latter case a substituted 1,2-anhydrosugar-terminated oligosaccharide. Another embodiment contemplates preparation of an oligosaccharide. Here, a substituted 1,2-anhydrosugar derivative having a ring containing five or six atoms and a plurality of non-participating substituents is reacted with a hydroxyl group of a glycal derivative having five or six atoms in the ring and a plurality of substituents to form an epoxide- opened glycosyl glycal reaction product.

The resulting glycosyl glycal reaction product is converted to a glycosyl substituted 1,2-anhydrosugar derivative intermediate by reaction with a dialkyl dioxirane as discussed previously such as dimethyl dioxirane. That resulting glycosyl substituted 1,2-anhydrosugar intermediate is thereafter reacted with a nucleophile to form an epoxide-opened glycoside reaction product.

The non-participating substituents are those preferred groups discussed previously, as is the nucleophile preferably a hydroxyl group of a sugar derivative. The starting substituted 1,2-anhydrosugar can itself be a terminal saccharide unit of an oligosaccharide, in which case this method provides a lengthened oligosaccharide, or can be said to extend the chain of the original oligosaccharide.

Where the hydroxyl group-containing glycal derivative contained a participating substituent, that substituent is carried into the glycal portion of the glycosyl glycal reaction product. That participating substituent is removed and replaced with a nonparticipating group prior to reaction with the dialkyl dioxirane.

It is also noted that the epoxide ring-opening, glycosylation, reaction creates a participating hydroxyl group at the 2-positiσn of the glycosyl portion of the reaction product. That participating group is also removed by converting it to an O-ether substituent as with t- butyldimethylsilyl chloride under appropriate conditions. Thus, any participating substituent present in the glycosyl glycal is removed and replaced by a non-participating substituent prior to expoxidation of the glycosyl glycal to form the glycosyl substituted 1,2-anhydrosugar intermediate. Yet another aspect of the present invention is a method of preparing a glycal-terminated oligosaccharide. Here, a substituted glycal having only non-participating substituents is converted to a corresponding substituted 1,2-anhydrosugar thereof. A glycal derivative having a single reactive, nucleophilic, hydroxyl group is glycosylated with the substituted 1,2-anhydrosugar. The converting and glycosylation steps are thereafter repeated seriatim. Thus, a glycal- terminated oligosaccharide of any desired length is prepared.

As was the case in previously discussed methods, the glycal derivative that is glycosylated can contain participating as well as non-participating substituents. When this is the case, the participating substituents are removed and replaced with non-participating substituents prior to reiterating the conversion of the terminal glycal into a terminal substituted 1,2-anhydrosugar, as discussed previously. An oligosaccharide having a terminal saccharide unit that is other than a glycal can be prepared from the glycal- terminated oligosaccharide. Here, the glycal-terminated reaction product obtained from the last repeat of the above method is again converted to its corresponding substituted 1,2- anhydrosugar derivative. That substituted 1,2-anhydrosugar derivative is then reacted with the hydroxyl group of a substituted sugar derivative other than a glycal to glycosylate that hydroxyl group and form a corresponding oligosaccharide terminated with a substituted sugar derivative. The present invention also contemplates several methods which revolve about the haloglycosylation of a substituted glycal that has relative electron donating non- participating substituents with a glycal derivative having a single reactive alcohol and relative electron withdrawing substituents, or a particle-linked nucleophile having a single reactive alcohol substituent, to form a corresponding halo- substituted glycosyl glycal or a halo-substituted glycoside. The products of all of those reactions are preferably recovered, although product can be utilized without recovery per se .

In the methods that are discussed below, the glycals or substituted glycals utilized are those discussed previously. The substituted glycal glycosyl donor molecule thus contains a plurality of non-participating relative electron donating substituents, whereas the glycal derivative glycosyl acceptor molecule contains relatively electron donating or withdrawing substituents that typically include at least one O-acyl substituent and a single reactive, nucleophilic alcohol, although the presence of electron withdrawing substituent is not required in the nucleophile. The halonium ion reagent is also as discussed herein.

The product of each reaction step is preferably recovered. However, recovery between steps is not required nor is it required after the last haloglycosylation as that product can be further reacted without purification, as by simply removing the acyl substituents.

In one aspect, a glycal-terminated halo-substituted saccharide multimer is prepared. Here, a glycal derivative having a plurality of substituents and having a single reactive hydroxyl group is haloglycosylated with a substituted glycal having non-participating, electron donating substituents, the substituents on the glycal derivative being electron withdrawing relative to the substituents on the substituted glycal, and a halonium ion reagent. The reaction is carried out in the absence of water. The electron withdrawing substituents such as O-acyl groups on the formed haloglycosylation product are removed and replaced with non-participating electron donating substituents such as O-ether linkages, and the haloglycosylation is repeated.

In another aspect of the invention, a method of preparing a particle-linked glycoside is contemplated. Here, a solid phase particle-linked nucleophilic molecule having a single reactive hydroxyl group per molecule is haloglycosylated with a liquid composition free of water that contains a substituted glycal having a plurality of substituents and a halonium ion substituent. The particle-linked glycoside formed has a halide group at the 2-position of the formed (added) glycosyl ring that is adjacent and trans to the formed glycoside bond. The solid and liquid phases are then separated. In preferred practice, one of the plurality of substituents is a protected nucleophilic hydroxyl group whose protecting moiety can be selectively removed in the presence of any other substituent present. Exemplary of such protecting moieties are the silyl and benzyl moieties of ether groups, and acyl moieties of ester groups. Such selective removal provides a nucleophilic hydroxyl group in unprotected form. That hydroxyl group is then haloglycosylated as discussed before to form a particle-linked glycosyl glycoside. The solid and liquid phases are again separated. It is preferred that the nucleophilic oxygen atom of the hydroxyl group be a hydroxyl group of a sugar derivative. It is also preferred that the link between the particle and nucleophilic molecules be selectively severable as is the case when that link is a benzyl ether linkage. In another aspect, the substituted glycal contains an oligosaccharide substituent and is itself a substituted glycal-terminated oligosaccharide. In another embodiment, a solid phase particle-linked nucleophile having a single nucleophilic hydroxyl group per molecule (and no other nucleophilic group as is the case for all of the particle-linked nucleophiles herein) is haloglycosylated with a liquid composition free of water containing an excess of a substituted glycal as discussed before whose plurality of substituents includes a protected nucleophilic hydroxyl group whose protecting moiety can be selectively removed in the presence of any other substituent present and an excess of a halonium ion reagent to form a particle-linked glycoside having a halogen at the 2-position of the formed or added glycosyl ring. The solid and liquid phases are separated. The selectively removable protecting moiety is removed to provide a free nucleophilic hydroxyl group. That hydroxyl group is haloglycosylated as described above, the solid and liquid phases separated, and the selective removal (deprotecting) , haloglycosylation, and the phase separation steps are repeated seriatim .

It is to be understood that the single hydroxyl group-containing nucleophile linked to the particle can itself be a mono- or oligosaccharide. Thus, if a monosaccharide is initially linked to the particle and is haloglycosylated to form a particle-linked glycosyl glycoside (or glycosyl glycoside-particle) , removal of a protecting moiety from a protected hydroxyl group of the glycosyl portion of the particle-linked disaccharide provides another single nucleophilic hydroxyl group that is linked to the solid phase particles.

The above sequence of steps yields a dihalo- trisaccharide, when the starting substituted glycal is a monosaccharide. The reaction can be repeated as often as desired to prepare a larger oligosaccharide or a polysaccharide, i.e., a saccharide multimer. Where the glycosyl donor substituted glycal is itself an oligosaccharide, correspondingly larger oligosaccharides or polysaccharides are formed, depending upon the number of times the reaction sequence is repeated or reiterated. Preferably, the reaction product is an oligosaccharide.

If desired, the electron withdrawing substituents such as acyl groups can be removed without replacement with non-participating electron donating substituents. Here, the product formed by such a removal is a halo-polyhydroxyl- substituted glycal-terminated saccharide.

In each of the methods described herein, it is preferred that the haloglycosylation be carried out in the presence of powdered 4A molecular sieves as are commercially available. An amount about equal in weight to the glycosyl donor is utilized. The powdered molecular sieves are about the size of bath powder particles. Use of the molecular sieves helps to assure the absence of water from the reaction system. The halide substituents such as iodide formed on the glycoside reaction products can be replaced with other substituents using reaction conditions well known for such replacements. Hydroxyl groups and hydrogen are usual replacing substituents, although nitrogen- and sulfur-containing replacements are also contemplated. Preferably, the halide group is replaced prior to the carrying out of another haloglycosylation reaction, although a plurality of halide groups can also be replaced in a single reaction step. The substituted glycal-terminated halo-substituted saccharide prepared in accordance with an above method can itself be a glycal donor for an alcohol that is other than a glycal. In such an instance, any acyl substituent present need not be formed by the last haloglycosylation step, as would be the case if the haloglycosylated acceptor alcohol were a glycal. It is noted, however, that such a removal and replacement of substituents can be carried out if desired. The glycal donor that was formed by the above method that still contains an electron withdrawing group such as an O-acyl group is then used to haloglycosylate an alcohol of other than a glycal under the haloglycosylation conditions discussed herein. Thus, differentiation between two glycals as discussed before is unnecessary since only a single glycal is present.

The alcohol, or particle-linked alcohol where appropriate, that is glycosylated can be any nucleophilic alcohol as already discussed. Preferably, the nucleophilic alcohol is a portion of a substituted sugar molecule such as a substituted monosaccharide. Non-sugar molecules can also be utilized. Examples are simple alcohols like methanol or ethanol, the alcohol groups of the afore-mentioned amino acid derivatives, and more complex alcohols such as menthol and cholesterol, and those present on a tetracycline aglycone that itself can be severably-1inked to the particle by another hydroxyl or other aglycone substituent. Still another method of this invention contemplates carrying out a before-described method with the starting substituted glycal or substituted 1,2-anhydrosugar linked to a solid phase such as are utilized in the solid phase syntheses of oligo- and polypeptides or oligo- and polynucleotides. The solid supports contemplated are particulate materials.

The substituted glycal or substituted 1,2- anhydrosugar derivative in certain aspects of the invention is linked directly through one of its unprotected hydroxyl groups to the solid phase support. The nucleophile in other aspects of the invention is linked directly to the solid phase support. That linkage can be through a direct bond and the linkage must, of course, be inert to the reaction conditions, but capable of being cleaved when desired so that the produced glycoside or nucleophile-terminated glycoside can be separated from the support. Benzyl ether linkages are preferred routes of bonding the substituted glycal or substituted 1,2-anhydrosugar to the solid phase support. Benzyl ether linkages are also preferred routes of bonding the nucleophile to the solid phase support, particularly where the nucleophile is a sugar hydroxyl group, and reactions used to form such linkages in particles are well known. Linking groups between the nucleophile and particles are also contemplated. One such group that is also useful for sugar nucleophiles is the 3-aminopropanol group that can be reacted with a benzyl halide-containing particle as discussed below to provide a primary hydroxyl group that can be haloglycosylated with a substituted glycal as discussed herein or by other well known means. Terminal glycosides such as that which is formed are readily cleavable, as is well known.

The substituted glycal, nucleophile or linking group is reacted with the particulate support under usual benzylation conditions to form a particulate support-linked substituted glycal or nucleophile. A precursor to the substituted 1,2- anhydrosugar derivative such as a substituted glycal or another sugar derivative that can be made into a substituted 1,2- anhydrosugar is linked to the resin via an O-benzyl ether linkage.

Any remaining halomethylbenzyl groups of the support are reacted as with a primary alcohol such as methanol or a tertiary amine such as triethylamine. The particle-linked substituted 1,2-anhydrosugar is thereafter prepared as is discussed previously to form a particulate support-linked substituted 1,2-anhydrosugar. The particle-linked nucleophile is ready to use.

In preferred practice in certain aspects of the invention, one of the plurality of non-participating substituents of the 1,2-anhydrosugar is a nucleophilic hydroxyl group whose protecting moiety can be selectively removed in the presence of any other substituent that is present, such as a tri-substituted silyl group like trimethylsilyl. In such practice, the selectively removable protecting moiety is selectively removed to provide the deprotected nucleophilic hydroxyl group in free or unprotected form. That nucleophilic hydroxyl group is glycosylated with a liquid composition containing a substituted 1,2-anhydrosugar having a plurality of non-participating substituent groups to form a particle-linked epoxide ring-opened glycosyl glycoside derivative. The solid and liquid phase are again separated.

It is also preferred that the 2-position hydroxyl formed after each glycosylation be protected prior to deprotecting the nucleophilic hydroxyl to be glycosylated in a further glycosylation reaction. Benzyl groups are particularly useful for forming 2-benzyl ether substituents from 2-hydroxyl groups.

In a particular aspect of the invention, a particle-linked saccharide multimer glycoside derivative is prepared. Here, a solid phase particle-linked nucleophile such as the amino-terminal amine of a protected polypeptide or a hydroxyl group of a sugar derivative is glycosylated with a liquid composition free of water that contains a Lewis acid catalyst and an excess of a substituted 1,2-anhydrosugar, preferably in stoichiometric excess. That substituted 1,2- anhydrosugar has a plurality of non-participating substituents that include a protected nucleophilic hydroxyl group whose protecting moiety can be selectively removed in the presence of any other substituent that is present. A particle-linked epoxide ring-opened glycoside derivative having a hydroxyl group at the 2-position of the formed or added glycosyl ring is thus prepared. The solid and liquid phases are separated. The 2-position hydroxyl group formed is protected. The protecting moiety noted before is selectively removed in the presence of the other substituents present that are not removed to form a particle-linked glycoside derivative having a free nucleophilic hydroxyl group.

That particle-linked free hydroxyl group is then glycosylated in the absence of water with a liquid composition containing a Lewis acid catalyst and a substituted 1,2-anhydrosugar (preferably in excess) containing a chain of 5 through 9 carbon atoms, a ring containing 5 or 6 atoms and a plurality of non-participating substituents to form a particle-linked epoxide ring-opened glycosyl glycoside derivative having a hydroxyl group at the 2-position of the formed (added) glycosyl ring. The solid and liquid phases are separated.

The above 2-position hydroxyl group protecting, selectively removing the protecting moiety from the nucleophilic hydroxyl group (deprotecting) , glycosylating and solid-liquid phase separation steps are carried out seriatim until a multimer of desired length is prepared.

The various reactions as described before are thereafter carried out to prepare a desired product except that any unreacted reagents, solvents and the like are removed from the particle-linked reaction product after each step as by rinsing. Once the desired reaction product is prepared, it is separated from the particle support by cleavage of the particle benzyl ether-reaction product bond to provide a hydroxyl group on the reaction product where the particle had been linked, and free particles. Usual benzyl ether cleaving reagents are utilized for this step such as by use of trifluoroacetic acid or lithium in liquid ammonia. The cleaved reaction product is thereafter recovered.

The techniques described in U.S. Patent No. 4,631,211, whose disclosures are incorporated by reference, for the preparation of polypetides and polynucleotides are also useful herein. The disclosures of that patent are particularly useful where a number of oligo- or polysaccharides having similar sequences of saccharide units is to be prepared.

Thus, a plurality of foraminous containers containing particle-linked substituted glycal or nucleophile molecules as discussed before is provided. Conversion of the substituted glycal to a substituted 1,2-anhydrosugar and glycosylation of a glycal derivative, or glycosylation with a substitute 1,2- anhydrosugar (glycosyl donor) , or haloglycosylation, as described herein is carried out on a group of those containers using a single glycosyl acceptor (glycal derivative) or donor, as appropriate, that is common to the sequence of each saccharide to be made. One or more containers used for preparing a saccharide not containing the common saccharide unit are withheld from the group reaction and glycosylated separately.

The glycosylation, or conversion and glycosylation, or haloglycosylation steps are thereafter carried out using another group of containers and a single glycosyl acceptor or a single glycosyl donor (substituted 1,2-anhydrosugar), as appropriate, common to the sequences of the saccharide to be prepared in each container. Again, any container used to prepare a saccharide that does not include the common glycosyl acceptor or donor is removed from the group, and its growing saccharide is glycosylated separately. The above-described group of conversions, glycosylations, removal of containers with sequences not having a common glycosyl acceptor or donor and separate glycosylation are repeated as necessary until the separate saccharides having similar sequences are prepared. Where a terminal saccharide unit other than a glycal is desired, that unit is typically added by the last glycosylation step.

It is noted that although the above method is particularly useful for preparing a plurality of saccharide molecules having similar sequences, it need not be so limited. In this instance, it is only necessary that the saccharides produced have a common saccharide unit that is to be added. Thus, a group of containers having different particle-linked glycal-terminated or hydroxyl-containing oligo- or polysaccharides is used to glycosylate a common glycosyl acceptor, or glycosylated with a common glycosyl donor. After that reaction, the containers are separated from each other, regrouped and reacted again with another common glycosyl acceptor or donor, reacted separately with a glycosyl acceptor or the completed saccharide is cleaved from the particle as is required for each molecule being prepared.

The liquid composition discussed herein in connection with certain embodiments of the invention contains the substituted 1,2-anhydrosugar, the Lewis acid catalyst, when used, is free of water, and a before-discussed solvent. It is also noted that each of the deprotecting and protecting steps discussed herein is carried out using an appropriate liquid composition that contains reagents appropriate for the reaction to be carried out, and that the solid and liquid phases are preferably separated after each reaction is completed.

It is preferred in any of the glycosylation reactions discussed herein to use the substituted 1,2-anhydrosugar in excess over the nucleophile. Mole ratios on the order of about 2:1 to about 10:1 are utilized. A reason for this preference is that in initial studies with liquid phase reactions, the substituted 1,2-anhydrosugar compounds are prepared in very high yield. However, those compounds are substantially consumed during each glycosylation reaction, but the yield of glycoside derivative ranges between about 50 and 60 percent, when the nucleophile is present in excess. Using an excess of the relatively less expensive substituted 1,2-anhydrosugar can help to overcome possible deficiencies in reactivity of the nucleophile and the fact that much of the substituted 1,2- anhydrosugar is consumed in a side reaction.

It is further noted that although it is preferred to carry out all of the before-described reactions using solid phase support particles, each of the reactions not involving linkage of the nucleophile to the particle or the particle- requiring reactions similar to those described in U.S. Patent No. 4,631,211 can also be carried out in the liquid phase. In such a situation, products are separated and collected by usual means, and the phase separation steps are omitted. Indeed, one such preliminary reaction has been carried out. In that case, a substituted 1,2-anhydrosugar was used to glycosylate a sugar derivative. The 2-position hydroxyl group that was formed was then used as a glycosyl acceptor for glycosylation with another substituted 1,2-anhydrosugar glycosyl donor.

Ill. RESULTS

Dimethyldioxirane, Compound 3, prepared according to the procedure of Murray and co-workers — Murray, et al . , J. Org. Chem. 50:2847 (1985) — was used as a solution in acetone. Reaction of tri-O-acetylglucal, Compound 4, with Compound 3, in methylene chloride/acetone at zero degrees C followed by solvolysis in methanol afforded a mixture of products that was not separated. It was reasoned that selectivity might be improved both in epoxidation and in subsequent glycosylation reactions if the protecting groups were non-participating; i. e. , groups that could not provide anchimeric assistance at the reaction site, such as ether groups as compared to acyl groups. Toward this end, glycal Compounds 5 and 6 were employed as substrates as is shown in the scheme of Figure 2. A stereospecific reaction ensued with each of Compounds 5 and 6 as substrates to form Compounds 8 and 9, respectively.

The workup simply consisted of evaporation of the volatiles. Although the NMR spectra of the crude products did not rigorously define the stereochemical sense of epoxidation, this point was established by a sequence of methanolysis followed by acetylation.

In a similar way, the galactal-derived Compound 10, tri-O-tert-butyldimethylsilylgalactal was prepared in about 80% overall yield from tri-O-acetylgalactal by the two-step sequence that follows: a) methanol with a catalytic amount of sodium methoxide, followed by b) reaction with t- butyldimethylsilyl chloride, and Et₃N in DMF, cleanly gave rise to the α-oxirane Compound 11 in about 98% yield. This epoxidation is shown in the scheme of Figure 3.

The first synthesis of a β-epoxide by direct epoxidation using the above Murray system, Murray, et al . , J. Org. Chem. 50:2847 (1985), was carried out with the allal derivative Compound 12 that was generously provided by Dr. David A. Griffith of Yale University. See Lemieux, et al . , Can. J. Chem. 46:61 (1968). Resulting Compound 13 was obtained in 98% yield. A mixture (about 1:1) of oxirane Compounds 15 and 16 was similarly generated by epoxidation of the gulal derivative Compound 14, using the procedure of Lemieux, et al . , Can. J. Chem. 46:61 (1968). These reactions are also illustrated in Figure 3. An early study by Henvest, et al . , Bull . Soc. Chem.

France 1365 (1960) , designed to probe the possibility of hydroxyl directivity on peracid epoxide formation of cyclic allylic alcohols indicated relatively little effect. Thus, reaction of the C~-free hydroxyl version of Compound 14 gave only a small difference in the product ratio obtained relative to that with the protected substituent system.

With a general route to these epoxides in hand, their value as glycosyl donors was probed. The first studies involved methyl glycoside formation by dissolution of the epoxide in methanol. Reactions were complete within two hours at room temperature. The resultant methoxyhydrins were acetylated with acetic anhydride in pyridine at zero degrees C and the products were then analyzed. In each instance, methanolysis had occurred with clean inversion of configuration in excellent yield. Results for the acylated methoxyhydrins prepared as illustrative compounds are shown in Table 1 below and are illustrated in Figure 4 in which the structure of each product is shown above its Compound number.

TABLE 1

Exemplary Methyl Glycoside Formation

Compound numbers are as discussed in the specific examples that follow.

The usefulness of the before-described oxiranes as glycosyl donors in the construction of exemplary oligosaccharides has also been examined. This type of reaction had been studied in the past in a limited way, by Hickenbottom, et al . , J. Chem. Soc. 3140, (1928); Hardegger, et al . , Helv. Chim. Acta 31:221 (1948); Klein, et al . , J. Am Chem. Soc. 104:7362 (1982); Bellosta, et al . , J. Chem. Soc. Chem. Commun. 199 (1989); Tietze, et al . , Tetrahedron Lett. , 18:3441, (1977); and Tietz, et al . , Chem Ber. 2441 (1978), and had been attended by low yields.

In the early studies related to the present invention, anhydrous zinc chloride was used as the Lewis acid catalyst in tetrahydrofuran (THF) as solvent at temperatures from -78 degrees C to room temperature using a reaction time of 24 hours. Two encouraging cases have already been demonstrated, and these are shown schematically in Figure 5.

Reaction of epoxide Compound 8 with the diisopropylidene galactose derivative, Compound 21, afforded cleanly disaccharide Compound 22a. Similar reaction of Compound 8 with differentiated glucal Compound 23, according to the procedure of Blackburne, et al . , Aus. J. Chem. 29:381 (1976) afforded Compound 24a. The yields based on consumed glycosyl acceptor (e. g. , alcohol of the galactose and glucal derivatives) were 80-90%. However, the yields based on reacted epoxide were about 50- 58%. In a separate study, it was shown that epoxide Compound 8 is substantially consumed upon exposure to anhydrous zinc chloride under these conditions (-78 degrees C in THF) .

In each case, the glycosylation reaction was essentially stereospecific with formation of a β-glycosidic bond. This was firmly proven by H NMR analysis of acetylated derivatives Compounds 22b and 24b.

Benzylation of Compound 24a to form Compound 24c, followed by reiteration (serially repeating) of the conversion and glycosylation reaction sequence gave epoxide Compound 25 and trisaccharide Compound 26a. The much greater success of these glycosylation reactions both in terms of stereospecificity and yield relative to the classical Brigl's anhydride case, as reported by Brigl, Z. Physiol . Chem. 122:257 (1922); Hickenbottom, et al., J. Chem. Soc. 3140, (1928); Hardegger, et al . , Helv. Chim. Acta 221 (1948); Klein, et al . , J. Am Chem. Soc. 104:7362 (1982); Bellosta, et al . , J. Chem. Soc. Chem. Commun . 199 (1989), is thought to be due to the absence of participating groups (e . g. hydroxyl or acyl hydroxyl) and the presence of only non-participating groups in the epoxide substrate.

In addition to their utility in oligosaccharide construction, the substituted 1,2-anhydrosugars synthesized as described above are useful glycosylating agents for the formation of glycolipid conjugates. For instance, reaction of Compound 8 with menthol. Compound 27, give Compound 28a.

Likewise, the cholesterol conjugate Compound 30a was formed upon reaction of Compound 8 with cholesterol, Compound 29. As before, the structures of the glycosides were proven by analysis of the respective acetates Compounds 28b and 30b. In further exemplary studies, a variety of nucleophiles was reacted in the presence of zinc chloride and absence of water with a number of substituted 1,2-anhydrosugars formed by conversion of corresponding glycals using dimethyl dioxirane, Compound 3. Thus, tribenzyl glucal was converted to its corresponding substituted 1,2-anhydrosugar, and that substituted 1,2-anhydrosugar was used to cleanly glycosylate azide ion and N-t-BOC-threonine methyl ester and N-t-BOC-serine methyl ester. Exemplary nucleosides were prepared using bis(di- trimethylsilyl)thymine as the nucleophile. The corresponding substituted 1,2-anhydrosugar converted from tri-benzylgalactal with Compound 3 as well as similarly converted compounds, Compounds 6 and 12, cleanly glycosylated bis(di- trimethylsilyl)thymine to form the β-, β- and α-anomers, respectively. Use of 3-t-butyldimethylsilyl-5-t-butyldiphenyl- silylribal and the above thymine derivative led to the - anomer, whereas use of the 3-hydroxyl-5-t-butyldiphenyl- silylribal led to a 6:1 ratio of β: anomers. The previously discussed zinc chloride and absence of water reaction condition discussed herein was utilized for the glycosylation reactions.

In summary, the first general, high yield, one step conversion of substituted glycals to substituted 1,2- anhydrosugars of predetermined stereochemistry is described and illustrated herein. Displacement of the epoxide linkage at the anomeric center with clean inversion of configuration to form oligosaccharides and other conjugates with α- and β-glycosidic linkages at temperatures as low as -78 degrees C has been demonstrated.

Glycals, Compounds 107, 108, 109 and 110, readily derivable via D-glucal, were principally employed in this investigation. Diacetone galactose and diacetone glucose derivative Compounds 111 and 112 were used as terminating sugars to cap the halo-oligosaccharides. Structures for Compounds 107 through 112 are shown in Figure 9.

The scheme shown in Figures 10 and 11 illustrate the synthesis of trisaccharide Compounds 114, 115, 116 and 118. The exclusive formation of bicyclic glycal Compound 113 from the iodoglycosylation of Compounds 108 and 109 is illustrative of the power of the method. No other glycals or stereoisomers of Compound 113 were detected.

Oxidative coupling of Compound 113, this time to the "terminating" hexoses Compounds 111 and 112 could be achieved in the presence of the benzoyloxy groups; i . e. , electron withdrawing O-acyl groups. Of course in these cases, Compound 113 can only be a glycosyl donor and Compounds 111 or 112 could only be the glycosyl acceptor. Compounds 114 and 116, respectively, were obtained in clean stereospecific reactions. For illustrative purposes, it was demonstrated that Compound 114 could be doubly de-iodinated to provide trisaccharide Compound 115. It would be expected that the electron donating power of a glycal bearing an acyloxy group at the 3-position would be particularly suppressed relative to one in which there is an alkoxy function at this same center. It was important to establish whether the free alcohol of the intended glycosyl acceptor could be situated at the 3-position; i. e. , whether acyloxyl substitution at the 4- and 6-positions would suffice to direct the oxidizing agent to the triether glycal.

For this purpose, glycal Compound 110 proved to be particularly instructive. Here, iodoglycosylation of a mixture of Compounds 110 and 108 cleanly provided the "AD glycal"

Compound 117. Again, no evidence of the formation of products from an alternative coupling mode, or of stereoisomers of Compound 117, could be gleaned. As before, reiteration of the scheme with hexose Compound 111 as the terminating group was possible without modification of the benzoyloxy groups.

Trisaccharide Compound 118 was produced stereospecifically in 60% yield. These reactions are shown in the scheme of Figure 11.

The scheme shown in Figure 12 illustrates how this methodology was readily applied to the synthesis of tetrasaccharides. starting with disaccharide Compound 113, and following the logic set forth in the scheme shown in Figure 8, as discussed before, it was first necessary to convert the diester arrangement in the glycal segment of Compound 113 into a diether. The resultant product could be relied upon to function strictly as a glycosyl donor in the oxidative coupling with respect to a diacyloxymonohydroxy glycal. Accordingly, Compound 113 was converted as shown, to ether-substituted glycal Compound 119. Glycal Compound 119 was subjected to "i⁺" mediated coupling (iodoglycosylation) with glycal Compound 109 to afford Compound 120. As above, termination of the sequence with non-glycal Compound 111 occurred smoothly and stereospecifically to provide Compound 121. This was again effected without removal and replacement of the benzoyl groups (O-acyl) of Compound 120.

By the same logic and protocols, the previously described disaccharide Compound 117 was converted to tetrasaccharide Compound 124 via the agency of bis ether Compound 122 and trisaccharide Compound 123. This is illustrated in the scheme of Figure 13.

Preliminary results for the formation of what may be referred to as a chimeric drug, that is a drug having an aglycone portion of one drug and an oligosaccharide of another is shown in the schemes illustrated in Figures 14 and 15. In the example illustrated, the polysaccharide is an iodo-di-TBS ether derivative of cyclimycin 104, whereas the anthracycline aglycone with which it was haloglycosylated was from the naturally occurring drug marchellamycin, both of which exhibit in vitro and in vivo chemotherapeutic effects against a variety of non-solid tumor cancers such as lymphomas, and leukemias, such as Hodgkin's desease. In this particular example, the anthracycline aglycone of marchellamycin and ciclimycin 104 happen to be the same compound, epsilon-pyrromycinone. Compound 140 is nevertheless considered to be a chimeric drug, as that term is used herein.

Thus, L-fucal, Compound 125, was onobenzoylated to form the glycosyl acceptor Compound 126, and L-3-deoxy- rhamnal, Compound 127, was monobenzylated to form glycosyl donor Compound 128. Those two compounds were thereafter haloglycosylated as discussed herein to form iodinated Compound 129. The iodide at glycosyl position-2 was replaced to form Compound 130, and the benzoyl group of Compound 30 was removed to form Compound 131. Compound 131 was etherified to form the t-butyldimethyl ether (TBS) derivative Compound 132 that was then used to haloglycosylate Compound 126 to yield Compound 133, from which the iodide was removed as before to yield Compound 134.

Removal of the benzoyl group from Compound 134 yielded the corresponding alcohol, Compound 135, that was etherified to form another TBS ether linkage of Compound 136. The single benzyl group of Compound 136, was removed to form the corresponding alcohol, Compound 137, and that alcohol was oxidized to form the keto-substituted glycal-terminated trisaccharide, Compound 138.

Haloglycosylation of Compound 138 and epsilon- pyrromycinone (Compound 139) , an anthracycline aglycone obtained from marchellamycin, yielded the chimeric drug 140 that exhibits in vitro anticancer activity.

BEST MODE FOR CARRYING OUT THE INVENTION In the disclosures that follow, a generalized synthetic procedure for the preparation of each compound type is discussed. That procedure is followed by a listing of compounds prepared by that procedure and analytical data for each of those compounds. Those analytical data are as follows: optical rotation (specific rotation) using the sodium D line at 25 degrees C, [_-]_D with the molar concentration and solvent utilized in parentheses; major infrared absorbance peaks (IR) are listed following the solvent utilized in parentheses; proton nuclear magnetic resonance spectroscopic data (¹H NMR) with the field strength in megahertz

(MHz) , the solvent utilized in parentheses and the shift relative to tetramethylsilane (_") , the type of peak observed and the number of protons under the peak are listed as are coupling constants (J) in hertz; mass spectral data (MS) at 20 electron volts or by fast atom bombardment (FAB ms) with one or more peaks (m/e) being provided; calculated carbon and hydrogen percentages and those values found from combustion analysis are given for some compounds as are 13C NMR peaks.

General Procedure for Epoxidation: The glycal (0.1 mmol) was dissolved in 1.0 mL CH₂C1₂ and the resulting solution was cooled to 0 degrees C. A solution of dimethyldioxirane, Compound 3, in acetone (1.2 equivalents, approximately 0.05 M) was added dropwise. The reaction mixture was stirred at 0 degrees C for 1 hour or until thin layer chromatography (TLC) indicated complete consumption of the glycal. The solution was evaporated with a stream of dry nitrogen (N₂) and the residue was dried in vacua to afford the substituted 1,2-anhydrosugar(s) in quantitative yield.

l,2-Anhydro-3,4,6-tri-0-Benzyl-c.-D-glucopyranose: Compound 8

[α]_D ²⁵ +29.2 " (C 0.96, CHC1₃); IR (CHCl₃) 2990, 2900, 2840, 1714, 1440, 1245, 1090 cm^"1; ^XH NMR (250 MHz, CDC1-.) δ 7.4-7.1 (m, 15H) , 5.00 (br d, IH, J=1.96 Hz) , 4.88-4.50 (m, 6H) , 3.99 (d, IH, J=7.62 Hz), 3.85-3.60 (m. 4H) , 3.10 (d, IH, J=1.96 Hz); MS (20 eV) m/e 432, 341, 325, 181, 91.

1,2-Anhydro-3,4,6-tri-O-(tert-butyldimethylsilyl)-α-D- glucopyranose: Compound 9

[α]_D ²⁵ +23.6° (c 0.87, CHC1₃); IR (CHC1₃) 2958, 2935, 2860, 1475, 1262, 1145, 1110, 845 cm^"1; ~~E NMR (250 MHz, CDCl₃) 64.90 (d, IH, J=2.20 Hz) , 3.98 (d, IH, J=6.52 Hz) , 3.82-3.76 (m, 2H) , 3.60-3.43 (m, 2H) , 2.92 (d, IH, J=2.45 Hz) , 0.95 (s, 9H) , 0.91 (S,9H), 0.89 (s, 9H),0.21 (s, 3H) , 0.15 (s, 3H) , 0.11(S, 3H) , 0.10 (s, 3H) , 0.08 (s, 6H) ; MS (20 eV) m/e 505, 447, 374, 315, 241, 171, 129.

1,2-Anhydro-3,4,6-tri-O- (tert-butyldimethylsilyl)-α-D- galactopyranose: Compound 11 -^α-D^{25 +25} _'°'^C (^c 1-01, CHC1₃); IR (CHC1₃) 2960, 2935,

2860, 1472, 1260, 1184, 1160, 1148, 1100, 845 cm^"1; ¹H NMR (250 MHz, CDC1₃) δ 4.89 (d, IH, J=2.63 Hz) , 3.85 (br S, 2H) , 3.80- 3.40 (m, 3H) , 2.92 (br S, IH) , 0.96 (s, 9H) , 0.90 (s, 9H) , 0.88 (s, 9H) , 0.18 (s, 3H) , 0.16 (s, 3H) , 0.11 (s, 3H) , 0.09 (s, 3H) , 0.06 (s, 3H) , 0.05 (s, 3H) ; MS (20 eV) m/e 505, 487, 469, 447, 445, 374, 373, 357, 331, 316, 315, 273, 241, 213, 211, 171. 1,2-Anhydro-4 ,6-0-benzylidene-3-0(tert-butyldimethylsilyl)-β- D-altropyranose: Compound 13

[α]_D ²⁵ +29.0° (C 0.94, CHC1-.) ; IR (CHC1₃) 2930, 2905, 2835, 1458, 1420, 1245, 1150, 1135, 1110, 1090, 1010, 830, 819 cm^"1; H NMR (250 MHz, CDC1₃) ό^" 7.52-7.30 (m, 5H) , 5.50 (S, IH) , 4.91 (d, IH, J=2.52 Hz), 4.63 (t, IH, J=2.50 Hz) , 4.31 (dd, IH, J=10.03, 5.25 Hz) , 3.92 (dd, IH, J=9.85, 3.09 Hz) , 3.61 (t, IH, J=10.26 Hz) , 3.21 (t, IH, J=2.53 Hz) , 0.92 (s, 9H) , 0.11 (S, 3H) , 0.06 (s, 3H) ; MS (20 eV) m/e 364, 307, 215, 202, 201, 183, 171, 155, 145, 143, 129, 117, 105.

General Procedure for Methanolysis and Acetylation:

The substituted 1,2-anhydrosugar (0.1 mmol) was dissolved in 1.0 mL anhydrous methanol (MeOH) and the solution was stirred at room temperature for 2 hours or until TLC indicated complete consumption of the starting material. The MeOH was removed in vacuo to afford a quantitative yield of methyl glycoside(s) . The methyl glycoside(s) was dissolved in 1.0 mL pyridine and the resulting solution was cooled to 0 degrees C. . Acetic anhydride (1.0 mL) was added dropwise and the mixture was then stirred at 0 degrees C for 3 hours or until TLC indicated the disappearance of starting material. The mixture was added slowly to 20 mL saturated NaHC0₃ followed by extraction with 3 x 10 L ethyl acetate (EtOAc) The combined organic extracts were dried over MgSO., filtered, and concentrated to afford the acetylated products, which were purified by silica gel chromatography.

Methyl 2-0-acetyl-3,4,6-tri-O-benzyl-β-D-glucopyranoside: Compound 17

[α]_D ²⁵ +4.90° (C 0.49, CHC1₃); IR(CHCl₃) 2990, 1730,

1440, 1360, 1225, 1080, 1050 cm^"1; H NMR (250 MHz, CDC1-.) δ

7.42-7.10 (m, 15H) , 5.00 (t, IH, J=8.58 Hz), 4.88-4.76 (m, 2H) ,

4.73-4.50 (m, 4H) , 4.30 (d, IH, J=7.85 Hz) , 3.83-3.62 (m, 4H) , 3.49 (S, 3H) , 3.55-3.43 (m, IH) , 1.98 (s, 3H) ; MS (20 eV) m/e

469, 415, 383, 309, 277, 217, 205, 187, 181, 163, 127, 91;

Analysis calculated for ^C3o^H34°7^{: c} 71.13; H, 6.76; found: C,

71.51; H, 6.95. Methyl 2-0-acetyl-3,4,6-tri-O-(tert-butyldimethylsilyl)-β-D- glucopyranoside: Compound 18

[α]_D ²⁵ -23.9 ' (C 1.05, CHC1₃) ; IR (CHC1₃) 2940, 2915, 2845, 1735, 1465, 1255, 1110, 845 cm^'1; ~~H NMR (250 MHz, CDC1₃) δ A .12 (dd, IH, J=6.90, 3.40 Hz), 4.60 (d, IH, J=6.66 Hz) , 3.92-3.65 (m, 5H) , 3.47 (s, 3 H) , 2.07 (s, 3H) , 0.92 (s, 9H) , 0.90 (s, 9H) , 0.89 (S, 9H) , 0.14 (s, 3H) , 0.10 (s, 3H) , 0.09 (S, 6H) , 0.07 (S, 9H) ; MS (20 eV) m/e 521, 489, 461, 447, 429, 415, 401, 374, 373, 357, 347, 379, 315, 301, 273, 255, 231, 213, 175.

Methyl 2-0-acetyl-3,4,6-tri-O-(tert-butyldimethylsilyl)-β-D- galactopyranoside: Compound 19 -^α-D^{25 _0}-⁴°^e (^c 1-00, CHC1₃); IR (CHC1₃) 2940, 2910,

2840, 1740, 1470, 1370, 1255, 1110, 1080, 840 cm^"1; i NMR (250 MHZ, CDC1₃) δ 5.28 (dd, IH, J=9.70, 7.39 Hz) , 4.30 (d, IH, J=7.74 HZ), 3.98 (d, IH, J=1.81 Hz) , 3.80-3.70 (m, 2H) , 3.63 (dd, IH, J=9.92, 2.27 Hz) , 3.45 (s, 3H) , 3.39 (t, IH, J=7.09 Hz), 2.09 (S, 3H) , 0.93 (s, 9H) , 0.91 (s, 9H) , 0.90 (s, 9H) , 0.17 (s, 3H) , 0.10 (S, 9H) , 0.07 (s, 6H) , MS (20 eV) m/e 549, 521, 461, 429, 389, 375, 347, 315, 301, 287, 273, 261, 255, 245, 231, 229, 213, 197, 189, 175, 147, 117, 89; Analysis calculated for ^C2₇ ^H58°7^S13^{: C}' ⁵⁶-⁰¹' ^H» 10.10; found: C, 56.28, H, 9.99.

Methyl 2-0-acetyl-4,6-0-benzylidene-3-O-(tert- butyldimethylsilyl)-α-D-altropyranoside: Compound 20

[α]_D ²⁵ +71. ~ (c 0.59, CHC1₃); IR (CHC1₃) 3005, 2910, 2840, 1735, 1370, 1240, 1143, 1105, 1045, 1030 cm^"1; ~~E NMR (250 MHz, CDC1₃) δ 7.56-7.15 (m, 5H) , 5.59 (S, IH) , 4.86 (d, IH, J=3.00 HZ), 4.57 (s, IH) , 4.45-4.28 (m, 2H) , 4.07 (br t, IH, J=1.95 HZ), 3.90-3.72(m, 2H) , 3.37 (s, 3H) , 2.14 (s, 3H) , 0.93 (s,9H) ,0.11(s, 3H),0.04 (s, 3H) ; MS (20 eV) m/e 381, 349, 322, 321, 289,275, 243, 233, 215, 185, 169, 159, 149, 145, 121, 117, 105. 0-(3,4,6-Tri-O-benzyl-β-D-glucopyranosyl)-(1→6)-1,2:3,4-di-O- isopropylidene-c-D-galactopyranose: Compound 22a

Epoxide Compound 8 (37.7 mg, 0.0873 mmol) was dissolved in 0.15 mL tetrahydrofuran (THF) and the resulting solution was cooled to 78 degrees C. A solution of 34.0 mg (0.136 mmol) of Compound 21 in 0.15 mL THF was added followed by the dropwise addition of 0.15 mL of a 1.0 M solution of ZnCl₂ in ether. The reaction mixture was stirred at -78 degrees C for 1.5 hours and then warmed to room temperature over the course of 2 hours. After stirring at room temperature for 20 hours, the mixture was added to 25 mL saturated NaHC0₃ and extracted with 3 x 10 mL EtOAc. The combined organic extracts were dried over MgS0₄, filtered, and concentrated. The residue was chomatographed on silica gel, eluting with hexanes/EtOAc 7:3, to afford 34.8 mg of disaccharide Compound 22a (58%, 82% based on Compound 21) as well as 19.7 mg of unreacted Compound 21:

[α]_D ²⁵ -44.4° (c 3.25, CHC1₃); IR (CHCl₃) 3010, 2905, 1496, 1388, 1258, 1170, 1080, 1010 cm^"1; ^XH NMR (490 MHz, CDC1₃) δ 7.50-7.13 (m, 15H) , 5.56 (d, IH, J=4.96 Hz) , 5.03 (d, IH, J=11.27), 4.85 (d, IH, J=10.80 Hz) , 4.81 (d, IH, J=11.17 Hz), 4.67-4.59 (m,2H), 4.58-4.50 (m, 2H) , 4.40-4.30 (m, 2H) , 4.23 (dd, IH, J=7.96, 1.69 Hz), 4.10 (dd, IH, J=11.15, 3.44 Hz), 4.04(m, IH) , 3.80-3.68 (m, 3H) , 3.67-3.58 (m, 3H) , 3.50 (m, IH) , 2.97 (s, IH) , 1.55 (s, 3H) , 1.46 (s,3H), 1.35 (s, 3H) , 1.34 (s, 3H) ; ¹³C NMR (62.5 MHz, CDC1₃) 138.90, 138.23, 138.17, 128.29, 127.90, 127.90, 127.79, 127.62, 127.55, 127.46, 109.50, 108.77, 104.03, 96.29, 84.61, 77.32, 77.18, 75.17, 74.97, 74.82, 73.49, 71.20, 70.74, 70.47, 69.36, 68.94, 67.85, 26.01, 25.95, 24.92, 24.37; FAB MS m/e 692, 677, 447, 433, 405, 351, 325, 313, 253, 235, 224, 223, 221, 191, 181, 177; analysis calculated for ^C ₃₉ ^H ₄₈°n^{: C}' ⁶⁷-⁶¹∑ ^H 6-98; found: C, 6.79; H, 7.28.

O-(2-0-Acetyl-3,4,6-tri-O-benzyl-β-D-glucopyranosyl)-(1→6)- l,2:3,4-di-0-isopropylidene-α-D-galactopyranose: Compound 22b.

Disaccharide compound 22a (36 mg, 0.0520 mmol) was dissolved in 1.0 mL pyridine and the solution was cooled to 0 degrees C. Acetic anhydride (0.5 L) was added and the solution was stirred at zero degrees C for 3 hours after which time it was added slowly to 25 mL saturated NaHC0₃ and extracted with 3 x 10 mL EtOAc. The combined organic extracts were dried over MgS0₄, filtered, concentrated, and flash chromatographed on silica gel (eluting with hexanes/EtOAc 8:2) to afford 32.1 mg (84%) of Compound 22b.

[α]_D ²⁵ -30.1° (c 1.02, CHC1₃); IR (CHCl₃) 3010, 2920, 1745, 1455, 1380, 1235, 1080 cm^"1; ^XH NMR (490 MHz, CDC1₃) δ 7.40-7.18 (m, 15H) , 5.52 (d, IH, J=4.96 Hz) , 5.03 (t, IH) ,

J=8.54 Hz), 4.83-4.77 ( , 2H) , 4.69 (d, IH, J=11.40 Hz), 4.64 (d, IH, J=12.17 HZ), 4.60-4.55 (m, 3H) , 4.45 (d, IH, J=8.02 Hz), 4.29 (dd, IH, J=4.90, 2.40 Hz), 4.20 (dd, IH, J=7.97, 1.75 HZ), 4.07 (dd, IH, J=11.26, 3.52 Hz), 3.94 (m, IH) , 3.78-3.67 (m, 4H) , 3.64 (dd, IH, J=11.25, 7.39 Hz) , 3.49 (dt, IH, J=9.17, 3.13 Hz), 2.04 (s, 3H) , 1.53 (s, 3H) , 1.45 (s, 3H) , 1.35 (s, 3H) , 1.34 (S. 3H) ; FAB MS m/e 757 (M + Na⁺) .

0-(3,4,6-Tri-O-benzyl-β-D-glucopyranosyl)-(1→6)-1,5-anhydro- 3,4-di-0-benzyl-2deoxy-D-arabinohex-l-enopyranose: Compound 24a Epoxide Compound 8 (49.4 mg, 0.114 mmol) was dissolved in 0.15 mL THF and the resulting solution was cooled to -78 degrees C. A solution of glycal: Compound 23 (55.8 mg, 0.171 mmol) in 0.15 mL THF was added followed by the dropwise addition of 0.25 mL of a 1.0 M solution of ZnCl₂ in ether. The mixture was stirred at -78 degrees C for 1 hour and then allowed to warm to room temperature over the course of 1 hour. After stirring at room temperature for 18 hours the mixture was added to 25 mL saturated NaHC0₃, which was then extracted with 3 x 10 mL EtOAc. The combined organic extracts were dried over MgS0₄, filtered, and concentrated. The residue was chromatographed on silica gel (eluting with hexanes/EtOAc 7:3) to give 48.7 mg of Compound 24a (56%, 81% based on Compound 23) along with 30.0 mg unreacted Compound 23. I 3D²⁵ "⁵*⁹° (^c 2.97, CHC1₃); IR (CHC1₃) 3015, 3000,

2860, 1645, 1490, 1450, 1353, 1228, 1070, 700 cm^"1; ^XH NMR (490 MHZ, CDC1₃) δ 7.41-7.17 (m, 25H) , 6.42 (dd, IH, J=6.11, 0.91 Hz), 4.95 (d, IH, J=11.29 Hz),4.91 (dd, IH, J=6.19, 2.91 Hz) , 4.87-4.80 (m,3H), 4.72-4.52 (m, 6H) , 4.29 (d, IH, J=7.28 Hz) , 4.21-4.16 (m, 3H) , 3.87 (dd, IH, J=11.92, 7.14 Hz), 3.77 (dd, IH, J=7.49, 5.75 Hz), 3.76-3.68 (m, 2H) , 3.64- 3.57 (m, 3H) , 3.46 (m, IH) , 2.51 (s, IH) 73C NMR (62.5 MHZ, CDC1₃) δ 144.43, 138.75, 138.23, 128.55, 128.43, 128.35, 127.93, 127.87, 127.76, 127.70, 127.64, 127.55, 103.55, 99.88, 84.53, 76.29, 75.33, 75.03, 74.97, 74.74, 74.65, 74.41, 73.53, 73.29, 70.41, 68.97, 68.68; FAB MS m/e 758, 652, 651, 545, 447, 443, 433, 398, 346, 325, 307.

0-(2-0-Acetyl-3,4-di-O-benzyl-β-D-glucopyranosyl)-(1→6)-1,5- anhydro-3,4-di-0-benzyl-2-deoxy-D-arabino-hex-l-enopyranose: Compound 24b

Compound 24a (21.2 mg, 0.0280 mmol) was dissolved in 1.0 mL pyridine and cooled to 0 degrees C. Acetic anhydride

(1.0 mL) was added and the mixture was stirred for 1 hour at 0 degrees C followed by 2 hours at room temperature. The mixture was added slowly to 25 mL saturated NaHC0₃, which was then extracted with 3 x 10 mL EtOAc. The combined extracts were dried over MgS0₄, filtered, and concentrated. Flash chromatography on silica gel, eluting with hexanes/EtOAc 8:2, gave 16.5 mg (74%) of Compound 24b.

[α]_D ²⁵ +7.3° (C 1.65, CHC1₃) ; IR (CHC1₃) 3020, 3000, 2860, 1740, 1642, 1490, 1450, 1360, 1230, 1060, cm^"1; ^XH NMR (490 MHZ, CDC1₃) δ 7.40-7.18 (m, 25H) , 6.38 (dd, IH, J=6.17, 0.90 HZ), 5.05 (t, IH, J=8.54 Hz) , 4.87 (dd, IH, J=6.19 , 2.93 Hz), 4.82-4.78 (m, 3H) , 4.71-4.52 (m, 7H) , 4.43 (d, IH, J=7.94 Hz), 4.15-4.08 (m, 3H) , 3.84 (dd, IH, J=11.37, 6.13 Hz), 3.78- 3.64 (m, 5H) , 3.48 (m, IH) , 1.93 (s, 3H) FAB MS m/e 823 (M + Na⁺) analysis calculated for ^C4g^H52⁰ιo^{: C}' ⁷³«⁴⁸'" ^H 6.54; found: C, 73.61; H, 6.55.

O-(2,3,4,6-Tetra-O-benzyl-β-D-glucopyranosyl)-(1→6)-1,5- anhydro-3,4-di-0-benzyl-2-deoxy-D-arabino-hex-l-enopyranose: Compound 24c

A solution of disaccharide Compound 24a (30.2 mg, 0.0398 mmol) in 0.2 mL dimethylformamide (DMF) was added dropwise to a suspension of 60% NaH (3.0 mg) in 0.1 mL DMF at 0 degrees C. After stirring at 0 degrees C for 1 hour, 10 mL of benzyl bromide was added, and the mixture stirred for another 1 hour at 0 degrees C followed by 1 hour at room temperature. The mixture was added to 15 mL saturated NaHC0₃, which was then extracted with 3 x 5 mL EtOAc. The combined organic extracts were dried over MgS0₄, filtered, and concentrated. Purification by flash chromatography on silica gel (eluting with hexanes/EtOAc 8:2) gave 29.6 mg (88%) of Compound 24c. [α]_D ²⁵ +13.9° (C 0.825, CHC1₃); IR (CHC1₃) : 3010, 3000, 2860, 1680, 1495, 1453, 1230, 1070 cm^"1; ¹H NMR (490 MHz, CDC1₃) δ 7.40-7.15 (m, 30H) , 6.42 (d, IH, J=6.17 Hz), 5.02 (d, IH, J=10.95 HZ), 4.94 (d, IH, J=10.90Hz), 4.90 (dd, IH, J=6.18, 3.03 Hz), 4.85-4.77 ( , 2H) 4.76-4.72 (m, 2H) 4.65-4.59 (m, 3H) 4.5:7-4.51 (1, 3H) , 4.42 (d, IH, J=7.80 Hz) , 4.26-4.18 (m, 2H) , 4.15 (m, IH) , 3.89 (dd, IH, J=10.96, 6.40 Hz) , 3.79 (dd, IH, J=7.30, 5.67 HZ), 3.76-3.58 (m, 4H) , 3.50 (t, IH, J=8.24 Hz), 3.43 (m, IH) ; FAB MS m/e 849 (M + Na⁺) analysis calculated for ^C5₄ ^H ₅₆°₉ ^{: C}' ⁷⁶ _' ³⁹' ^{H 6}-65^; found: C, 75.91; H, 6.93.

0-(2,3,4,6-Tetra-O-benzyl-β-D-glucopyranosyl)-(1→6)-0-3,4-di- O-benzyl-β-D-glucopyranosyl)-(1→6)-1,5-anhydro-3,4-di-O-benzyl- 2-deoxy-D-arabino-hex-l-enopyranose: Compound 26a

Disaccharide Compound 24c (42.3 mg, 0.0499 mmol) was dissolved in 1.0 mL CH₂C1₂ and the resulting solution was cooled to 0 degrees C. A solution of Compound 3 (0.10 M, 0.75 mL) was added dropwise. After stirring at 0 degrees C for 0.5 hour the volatiles were removed in vacuo, giving 44.9 mg of Compound 25. Compound 25 exists as a 9:1 mixture of _:β epoxides and was used without purification. Epoxide Compound 25 (40.0 mg, 0.0463 mmol) was dissolved in 0.20 mL THF and cooled to -78 degrees C. A solution of glycal Compound 23 in 0.15 mL THF was added followed by a solution of ZnCl₂ in ether (1.0 M, 0.12 mL) . The mixture was stirred at - 78 degrees C for 1 hour and allowed to warm to room temperature over the course of 2 hours. After stirring for 10 hours at room temperature the mixture was added to 25 mL saturated NaHC0₃ and extracted with 3 x 10 mL EtOAc. The combined extracts were dried over MgS0₄, filtered, and concentrated. Flash chromatography on silica gel gave 17.4 mg of trisaccharide Compound 26a (32%, 57% based on 16.0 mg of recovered Compound 23) .

[α]_D ²⁵ +3.88° (C 0.825, CHC1₃); IR (CHC1₃) 3690, 3035, 3010, 1240, 1072, 813, 710 cm^"1; ^XH NMR (490 MHz, CDC1₃) δ

7.45-7.14 (m, 40H) , 6.38 (d, IH, J=6.20 Hz), 4.98-4.93 (m, 2H) , 4.89 (d, IH, J=10.97 Hz) , 4.86 (dd, IH, J=6.14, 2.90 Hz) , 4.84- 4.72 (m, 7H) , 4.65-4.50 (m, 7H) , 4.44 (d, IH, -1=1 .11 Hz), 4.26 (m, IH) , 4.20 (dd, IH, J=11.13, 1.40 Hz) , 4.16-4.12 (m, 2H) , 4.02 (m, IH) , 3.78-3.68 (m, 4H) , 3.65-355 (m, 5H) , 3.53-3.40 (m, 3H) , 2.45 (s, IH) FAB MS m/e 1123 (M + Na⁺ - Bn) .

0-(2,3,4,6-Tetra-O-benzyl-β-D-glucopyranosyl) -(1→6)-0-(2-0- acetyl-3,4-di-O-benzyl-β-D-glucopyranosyl)-(1→6)-1,5-anhydro- 3,4-di-0-benzyl-2-deoxy-D-arabino-hex-l-enopyranose: Compound 26b

Trisaccharide Compound 26a (16.0 mg, 0.0136 mmol) was dissolved in 0.5 mL pyridine and cooled to 0 degrees C. Acetic anhydride (0.5 mL) was added and the mixture was stirred at 0 degrees C for 3 hours. The solution was added slowly to 15 mL saturated NaHC0₃ and extracted with 3 x 5 mL EtOAc. The combined extracts were dried over MgS0₄, filtered, concentrated, and filtered through silica to afford 16.1 mg (96%) of Compound 26b. .^a~ v^{25 +7}-0° (c 0.99, CHC1₃) ; IR (CHC1₃) 3005, 2992,

2860, 1742, 1453, 1358, 1240, 1070 cml; H HMR (490 MHz, CDC-) δ 7.40-7.17 (m, 40H) , 6.35 (dd, IH, J=6.14, 0.94 Hz) , 5.07 (dd, IH, J=9.38, 8.02 Hz) , 4.95 (d, IH, J=11.06 Hz) , 4.90 (d, IH, J=10.94 HZ), 4.84-4.80 (m, 2H) , 4.79-4.65 (m, 5H) , 4.64-4.49 (m, 8H) , 4.48 (d, IH, J=7.81 Hz), 4.38 (d, IH, J=7.93 Hz), R.19 (br d, IH, J=10.42 Hz), 4.14-4.09 ( , 2H) , 3.93 (m, IH) , 3.78- 3.55 (m, 10H) , 3.48-3.42 (m, 2H) , 1.89 (s, 3H) FAB MS m/e 1255 (M + Na+)

1-Menthyl 3,4,6-tri-O-benzyl-β-d-glucopyranoside: Compound 28a Epoxide Compound 8 (43.1 mg, 0.10 mmol) and 1-menthol, Compound 27, (23.4 mg, 0.15 mmol) were dissolved in 0.25 mL THF and the resulting solution was cooled to -78 degrees C. A solution of ZnCl₂ in ether (1.0 M, 0.20 mL) was then added dropwise. The solution was stirred at -78 degrees C for 1 hour, then allowed to warm to room temperature over a period of 2 hours. After stirring at room temperature for 10 hours the mixture was added to 25 mL saturated NaHC0₃ and extracted with 3 x 10 mL EtOAc. The combined organic extracts were dried over MgS0₄, filtered, and concentrated. Flash chromatography on silica gel (eluting with hexanes/EtOAc 85:15) gave 25.8 mg (43%) of Compound 28a. -<*-_D ²⁵ -41.7° (c 2.57, CHC1₃) ; IR (CHC1₃) ; 2995, 2935,

2900, 2845, 1493, 1452, 1350, 1110, 1060 cm^"1; ¹H NMR (490 MHz, CDC1₃) δ 7.43-7.20 (m, 15H) , 4.96 (d, IH, J=11.40 Hz) , 4.89- 4.82 (m, 2H) , 4.64-4.53 (m, 3H) , 4.34 (d, IH, J=7.75 Hz) 3.75- 3.69 (m, 2H) , 3.66-3.58 (m, 2H) , 3.56-3.44 (m, 3H) , 2.31 (d sept, IH, J=6.97, 2.14 Hz) , 2.07 (br d, IH, J=12.28 Hz) , 1.71- 1.64 (m, 2H) , 1.44-1.33 (m, IH) , 1.31-1.22 (m, IH) , 1.08-0.85 (m, 10H) , 0.82 (d, 3H, J=6.88 Hz) ; FAB MS m/e 588 (M⁺) , 587, 547, 443, 433, 415, 369, 346, 341, 325, 313, 271, 253, 235, 225; analysis calculated for C₃₇H₄₈0₆: C, 75.48; H, 8.22; found: C, 74.94; H, 8.19.

1-Menthyl 2-0-acetyl-3,4,6-tri-O-benzyl-β-D-glucopyranoside: Compound 28b

To a solution of Compound (16.9 mg, 0.0287 mmol) in 1.0 mL pyridine at 0 degrees C was added 0.5 mL acetic anhydride. The mixture was stirred at 0 degrees C for 2 hours after which it was slowly added to 25 mL saturated NaHC0₃ and extracted with 3 x 10 mL EtOAc. The combined organic extracts were dried over MgS0₄, filtered, and concentrated. Flash chromatography on silica gel (eluting with hexanes/EtOAc 9:1) gave 15.9 mg (88%) of Compound 28b.

[α]_D ²⁵ -22.64° (c 0.795, CHCl₃) ; IR (CHC1₃) 3020, 3005, 2955, 2920, 2865, 1740, 1452, 1352, 1230, 1060 cml; ^XH NMR (490 MHz, CDC1₃) δ 7.30-7.21 (m, 15H) , 4.94 (t, IH, J=8.57 HZ), 4.83-4.78 (m, 2H) , 4.68 (d IH, J-11.41 Hz) , 4.65-4.60 (m, 2H) , 4.55 (d. IH, J=12.12 Hz), 4.41 (d, IH, J=8.02 Hz) , 3.76- 3.64 (m, 4H) , 3.46 (m, IH) , 3.39 (dt, IH, J=10.61, 4.23 Hz) , 2.36-2.27 (d sept, IH, J=6.96, 2.44 Hz) , 1.96 (s, 3H) , 2.68- 2 . 59 (m, 2H) , 2 . 40-2 . 28 (m, 2H) , 2 . 25-2 . 18 (m, IH) , 0 . 91 (d, 3H, J=6 . 54 Hz ) , 0 . 88 (d, 3H, J=7 . 09 Hz ) , 0 . 79 (d, 3H, J=6 . 85 Hz) FAB MS m/e 653 (M + Na⁺) .

Cholesteryl 3,4,6-tri-O-benzyl-β-D-glucopyranoside: Compound 30a

Cholesterol, Compound 29, (60.8 mg, 0.157 mmol) and epoxide Compound 8 (45.4 mg, 0.105 mmol) were dissolved in 0.3 mL THF and the resulting solution was cooled to -78 C. A solution of ZnCl₂ in ether (1.0 M, 0.3 L) was added dropwise and the mixture was stirred at -78 degrees C for 1 hour. The mixture was allowed to warm to room temperature over a period of 2 hours, and then stirred for a further 6 hours. The mixture was added to 25 L saturated NaHC0₃ which was then extracted with 3 x 10 mL EtOAc. The combined organic extracts were dried over MgS0₄, filtered, and concentrated. Flash chromatography on silica gel (eluting with hexanes/EtOAc 8:2) gave 44.3 mg of Compound 30a (52%, 83% based on 35.7 mg of recovered Compound 29) . l -D²⁵ "¹⁵-²° (^c -55 CHC1₃) ; IR (CHCl₃) 3570, 3010,

3000, 2930, 2860, 1710, 1470, 1455, 1365, 1273, 1115, 1068 cml; ^XH NMR (490 MHZ, CDCl₃) ό^" 7.41-7.14 (m, 15H) , 5.40-5.35 (m, IH) , 4.95 (d, IH, J=11.31 Hz), 4.87-4.82 (m, 2H) , 4.64-4.54 (m, 3H) , 4.36 (d, IH, J=7.64 Hz) , 3.75 (dd, IH, J=10.78, 1.83 Hz) , 3.67 (dd, IH, J=10.82, 5.04 Hz) , 3.64-3.47 (m, 5H) , 2.36 (dd, IH, J=4.70, 2.14 Hz), 2.31-2.25 (m, 2H) , 2.06-1.96 (m, 3H) , 1.90-1.81 (m, 2H) , 1.71-1.22 (m, 14H) , 1.21-1.15 (m, 7H) , 1.02 (s, 3H) , 0.93 (d, 3H, J=6.52 Hz) , 0.90-0.87 (m, 6H) , 0.69 (s, 3H) ; FAB MS m/e 818 (M⁺)

Cholesteryl 2-0-acetyl-3,4 ,6-tri-O-benzyl-β-D-glucopyranoside: Compound 30b

To a solution of compound 30a (19.3 mg, 0.0236 mmol) in 1.0 mL pyridine at 0 degrees C was added 0.5 L acetic anhydride. The mixture was stirred at 0 degrees C for 1 hour and then at room temperature for 2 hours. The reaction mixture was slowly added to 15 mL saturated NaHC0₃, which was then extracted with 3 5 mL EtOAc. The combined organic extracts were dried over MgS0₄, filtered, and concentrated. Flash chromatography on silica gel (eluting with hexanes/EtOAc 9:1) gave 14.2 mg (70%) of Compound 30b.

[α]_D ²⁵ +2.70° (c 0.705, CHC1₃) ; IR (CHC1₃) 3020, 2950, 1740, 1450, 1380, 1230, 1060 cm^"1; ~~H NMR (490 MHz, CDC1₃) 5 7.40-7.17 (m, 15H) , 5.34 (m, IH) , 4.97 (t, IH, J=8.20 Hz), 4.84-4.78 (m, 2H) , 4.67 (d, IH, J=11.41 Hz) , 4.62 (d, IH, J=12.20 HZ), 4.60-4.56 (m, 2H) , 4.44 (d, IH, J=7.97 Hz) , 3.75 (dd, IH, J=10.88, 1.85 Hz), 3.71-3.63 (m, 3H) , 3.53-3.45 (m, 2H) , 2.28-2.17 (m, 2H) , 2.08-1.92 (m, 6H) , 1.89-1.81 (m, 2H) , 1.68-1.23 (m, 14H) , 1.20-0.99 (m, 12H) , 0.97-0.83 (m, 12H) , 0.69 (s, 3H) ; FAB MS m/e 883 (M + Na⁺) .

General Procedure for I(sy-n-collidine)₂C10₄-Mediated Coupling: To a solution of glycal and alcohol (1.1 equivalents) in dry CH₂C1₂ (0.04 M in glycal) was added powdered 4A molecular sieves (approximately equal weight to that of glycal) . The resulting mixture was stirred at room temperature for 30 minutes and then I(syjn-collidine) ₂C10₄ was added as a solid. When TLC analysis indicated completion of the reaction (typically 1-2 hours) , the mixture was filtered, washing with CH₂C1₂. The resulting filtrate was washed with 10% aqueous Na₂S₂0₃, dried over MgS0₄, and concentrated. Chromatography of the residual oil on silica gel (hexanes:ethyl acetate, 4:1 to 5:1 v/v) provided the coupled product.

Glycal Compound 113:

3,4,6-Tri-O-benzyl-D-glucal, Compound 107, (563.8 mg) and 3,6-di-O-benzoyl-D-glucal, Compound 109, (527.7 mg) gave 704.1 mg (58%) of Compound 113 as a colorless oil:

[α]_D ²³ -18.5° (c=0.48, CHC1₃); IR (CHCl₃); IR (CHC1₃) 3010, 1717, 1645, 1450, 1270, 1105 cm^"1; ^XH NMR (250 MHz, CDC1₃) δ 3.30 (dd, IH, J=3.6, 8.4 Hz) , 3.62 (d,lH, J=10.9 HZ), 3.77 (brd, IH, J=10.9 Hz), 3.96 (m, 2H) , 4.31-4.70 (10H) , 4.82 (d, IH, J=10.7 Hz) , 5.00 (dd, IH, J=3.7, 6.4 Hz) , 5.53 (t, IH, J=3.7 HZ), 5.65 (d, IH, J=1.4 Hz) , 6.53 (dd, IH, J=1.5, 6.4 HZ), 7.14-7.61 (21H) , 8.01-8.06 (m, 4H) ; ¹³C NMR (63 MHz, CDC1₃) δ 32.7, 62.3, 68.7, 68.8, 71.2, 72.9, 73.4, 74.5, 75.0, 75.8, 76.8, 98.3, 102.6, 127.4, 127.5, 127.6, 127.8, 128.0, 128.2, 128.3, 128.4, 128.6, 129.5, 129.7, 133.0, 133.4, 137.6,

138.2, 145.9, 165.9, 166.0; FAB MS, m/e 895 (M - H)⁺; Analysis calculated for ^C47^H45^I0ιo^{: c} _> ⁶2.95; H, 5.06; Found: C, 63.20; H, 4.87.

Trisaccharide Compound 114:

Glycal Compound 113 (149.8 mg) and 1,2,3,4-di-O- isopropylidene-D-galactopyranose, Compound 111 (47.8 mg) gave 169.8 mg (79%) of Compound 114 as a colorless oil:

[α]_D ²³ +21.8° (c=0.74, CHC1₃); IR (CHCl₃) 3020, 2930, 1720, 1270, 1075 cm^"1; ^XH NMR (250 MHz, CDC1₃) 5:1.30, 1.34, 1.43 and 1.55 (s each, 3H each), 3.15 (dd, J=3.9, 8.0 Hz, IH) , 3.54 (d, J=10 Hz, IH) , 3.71-4.01 (6H) , 4.19-4.76 (15H) , 4.91 (dd, J=4.2, 9.0 HZ, IH) , 5.28 (s) , 5.53 (d, J=5.0 Hz, IH) , 5.61 (s, IH) 7.08-7.12 (m, 2H) , 7.21-7.34 (12H) , 7.41-7.65 (7H) , 8.12-8.18 (m, 4H) ; ¹³C NMR (63 MHz, CDC1₃) δ '. 24.6, 25.0, 26.0, 26.2, 30.4, 32.5, 63.4, 663., 67.3, 68.6, 70.2, 70.8, 760.9, 71.1, 72.2, 73.5, 73.7, 74.9, 75.6, 75.7, 68.8, 96.4, 101.3, 104.0, 108.7, 109.6, 127.3, 127.5, 127.6, 127.8, 128.0, 128.2,

128.3, 128.4, 128.8, 129.1, 130.0, 130.1, 130.2, 132.9, 133.8, 137.7, 138.4, 138.6, 165.2, 166.2; FAB MS, m/e 1283 (M + H)⁺; Analysis calculated for 5g^H64^I2⁰16^{: C}' ⁵⁵*²⁴' ^H' 5.03; Found; C, 55.04; H, 5.04.

Trisaccharide Compound 115:

A solution of Compound 114 (71.6 mg, 5.6X10^"5 mol) , triphenyltin hydride (58.8 mg, 3 equivalents) and a catalytic amount of azobis(isobutyro)nitrile (AIBN) in benzene (3 mL) was refluxed for 15 minutes and then concentrated. Chromatography of the residual oil on silica gel (hexanes followed by hexanes:ethyl acetate, 2:1 v/v) provided Compound 115 (54.1 mg, 94%) as a foam:

[α]_D ²³ +38.7° (c=0.63, CHC1₃) ; IR (CHC1₃) 3020, 1720, 1275, 1115, 1075 cm^"1; ^XH NMR (250 MHz, CDC1₃) _^": 1.37, 1.38, 1.47 and 1.60 (s each, 3H each), 1.5 (IH buried under Me signals), 1.92 (dt, J=3.6, 12.4 Hz, IH) , 2.09 (dd, J=4.8, 12.8 Hz, IH) , 2.47 (dd, J=5.2, 12.7 Hz, IH) , 3.50-4.75 (20H) , 4.83 (d, J=10.9 Hz, IH) , 5.06 (d, J=2.5 Hz, IH) , 5.40 (d, J=2.5 Hz, IH) , 5.57 (d, J=4.9 Hz, IH) , 5.62 (ddd, J=5.1, 8.7, 11.9 Hz, IH) , 7.10-7.70 (21H) , 8.13 (m, 4H) , ¹³C NMR (63 MHZ, CDC1₃ δ : 24.6, 25.0, 26.0, 26.2, 35.1, 64.1, 66.1, 66.2, 68.6, 69.1, 70.8, 71.1, 71.7, 72.2, 72.9, 73.5, 74.6, 76.3, 77.9, 96.4,

97.0, 99.6, 108.6, 109.3, 127.4, 127.6, 127.8, 128.2, 128.4,

128.6, 129.6, 129.8, 130.0, 130.3, 132.8, 133.4, 138.3, 138.6,

165.7, 166.3; FAB MS, m/e 1032 (M + H)⁺; Analysis calculated for ^C59^H66°16^{: C}' ⁶⁸*⁷²» H, 6.45; Found: C, 68.61; H, 6.58.

Trisaccharide Compound 116:

Glycal Compound 113 (140.0 mg) and 1,2,4,6- di-Oisopropylidene-D-glucofuranose, Compound 112 (44.7 mg) gave 167.9 mg (84%) of Compound 116 as a colorless oil: [^α-D^{23 +28}-⁸° (c=0.51, CHC1₃) ^β: 1.26, 1.36, 1.40 and

1.40 (s each, 3H each), 3.18 (dd, IH, J=3.9, 7.7 Hz), 3.62 (d, J=10.8 HZ, IH) , 3.76 (br d, J=10.3 Hz, IH) , 3.92-4.77 (21H) , 4.84 (dd, J=4.2, 8.9 Hz, IH) , 5.54 (s, IH) , 5.59 (d, J=1.6 Hz, IH) , 5.95 (d, J=3.6 HZ, IH) , 7.09-7.74 (m, 2H) , 7.19-7.65 (19H), 8.12-8.16 (m, 4H) ; ¹³C NMR (63 MHz, CDC1₃) _": 25.4, 26.2, 26.7, 26.9, 29.1, 32.0, 63.7, 65.8, 68.1, 68.5, 70.9, 71.8, 72.6, 73.5, 73.6, 74.9, 75.6, 75.9, 76.7, 81.6, 81.9,

84.1, 102.4, 104.2, 105.4, 109.5, 112.1, 127.4, 127.7, 127.8, 127.9, 128.0, 128.2, 128.3, 128.4, 128.8, 128.9, 130.0, 130.1, 133.0, 133.9, 137.6, 138.3, 138.4, 165.2, 166.2; Analysis calculated for ^C5g^H64^I2°i6^{: C}' ⁵⁵* 4; H, 5.03; Found: C,55.10; H, 5.05.

Glycal Compound 117: 3,4,6 Tri-O-benzyl-D-glucal, Compound 18 (69.9 mg) and

4,6-di-O-benzoyl-D-glucal, Compound 110 (65.4 mg) gave 114.4 mg (76%) of Compound 117 as a colorless oil:

[α]_D ²² +0.53° (c=0.76, CHC1₃) ; IR (CHC1₃) 3020, 1720,

1270, 1115 cm^"1; ^XH NMR (250 MHz, CDC1₃) δ 3.30 (dd, IH, J=4.1, 8.5 HZ), 3.73-4.08 (4H) , 4.30-4.72 (10H) , 4.86 (d, IH, J=10.7 Hz), 5.01 (dd, IH, J=3.8, 6.2 Hz), 5.47-5.52 (m, 2H) , 6.46 (d, IH, J=6.2 HZ), 7.17-7.68 (21H) , 8.02-8.10 (m, 4H) ; ¹³C NMR (63 MHZ, CDC1₃) δ 32.9, 61.9, 68.9, 69.1, 70.6, 71.0, 72.8, 73.4, 73.6, 75.1, 75.9, 76.7, 100.3, 102.0, 127.4, 127.5, 127.6, 127.8, 128.0, 128.1, 128.2, 128.3, 128.46, 128.52, 129.2, 129.6, 129.7, 129.8, 133.2, 133.5, 137.6, 138.2, 138.3, 144.6, 165.3, 166.0 FAB MS, m/e 897 (M + H)⁺; Analysis calculated for ^C47^H45^I0ιo^{: C}' 62.95; H, 5.06: Found: C, 63.10; H, 5.07.

Trisaccharide Compound 118:

Glycal compound 117 (70.0 mg) and 1,2,3,4- di-O-isopropylidene-D-galactopyranose, Compound 111 (22.0 mg) gave 68.5 mg (67%) of Compound 118 as a colorless oil:

[α]_D ²² -6.8° (c=0.47, CHC1₃) δ : 1.32, 1.34, 1.41 and 1.58 (s each, 3H each), 3.19 (dd, J=4.0, 8.3 Hz, IH) , 3.65- 3.87 (6H) , 3.96 (brt, J=6 Hz, IH) , 4.09-4.08 (12H) , 5.22 (s, IH) , 5.43 (s, IH) , 5.51 (d, J=5.0 Hz, IH) , 5.76 (t, J=9.6 Hz, IH) , 7.08-7.65 (21H) , 8.00-8.17 (m, 4H) ; ¹³C NMR (63 MHz,

CDC1₃) δ 24.6, 24.9, 26.0, 26.2, 32.3, 32.9, 63.1, 66.4, 67.2, 69.0, 69.6, 70.6, 70.8, 70.9, 71.1, 73.4, 73.8, 74.8, 75.7, 76.0, 76.9, 96.4, 101.4, 104.2, 108.7, 109.6, 127.4, 127.5, 127.6, 127.7, 128.0, 128.3, 128.7, 129.2, 129.8, 129.9, 130.0, 132.9, 133.6, 137.5, 138.3, 138.4, 165.2, 166.2; FAB MS, m/e 1282 M⁺; Analysis calculated for ^C59H₆₄I₂0₁₆: ^c 55.24; H, 5.03; Found: C, 55.69; H, 5.33.

Glycal Compound 119: To a stirred solution of glycal Compound 113 (134.6 mg, 0.15 mmol) in methanol:ether (10:1, 5 mL) was added 1 mL of a 1% (w/w) solution of NaOH in methanol. The resulting solution was stirred at room temperature for 1 hour and then concentrated. To the residual material was added water (10 mL) and CH₂C1₂ (10 mL) . The aqueous phase was extracted with

CH₂CL₂ (3 x 15 mL) . The combined organics were dried (Na₂S0₄) and concentrated.

The residual oil was purified by chromatography on silica gel (hexanes:ethyl acetate, 1:1 v/v) to provide 88.7 mg (86%) of the diol which was then dissolved in dimethyl formamide (DMF, 1 mL) . To this stirred solution was added imidazole (44 mg, 5 equivalents and t-butyl-dimethylsilyl chloride (TBSC1) (49 mg, 2.5 equivalents). After 15 hours, water (15 mL) was added and the resulting mixture was extracted with ether (5 x 10 mL) . Drying (Na₂S0₄) , concentration, and chromatography on silica gel (hexanes:ethyl acetate, 5:1 v/v) provided 105.8 mg (90%) of Compound 119 as a colorless oil: -^α-D^{23 +5}-⁶° (C=0.46, CHC1₃); IR (CHC1₃) 3020, 2930,

1650, 915 cm^"1; H NMR (250 MHz CDC1₃) 3020, 2930, 1650, 915 cm^"1; H NMR (250 MHz CDC1₃) δ 0.044, 0.046, 0.114 and 0.118 (s each, 3H each), 0.88 and 0.91 (s each, 9H each), 3.29 (dd, IH, J=4.0, 8.0 Hz), 3.65-4.11 (9H) , 4.47-4.86 (8H) , 5.58 (s, IH) , 6.30 (d, IH, J=6.2 Hz), 7.12-7.41 (15H) ; ¹³C NMR (63 MHz,

CDC1₃) δ -5.2, - 4.5, -4.1, 17.9, 18.4, 25.8, 26.0, 33.9, 62.0, 66.0, 68.7, 71.1, 73.0, 73.5, 74.5, 75.0, 75.9, 76.9, 78.0, 101.1, 101.8, 127.3, 127.5, 127.6, 127.7, 127.9, 128.2, 128.3, 1378.8, 138.5, 143.3; FAB MS, m/e 915 (M-H)⁺; Analysis calculated for C₄₅Hg₅IO_gSi₂: C, 58.94; H, 7.14; Found: C, 59.40; H, 7.29.

Glycal Compound 120:

Glycal Compound 119 (338.0 mg) and 3,6-di-0-benzoyl- D-glucal, Compound 109 (143.6 mg) gave 303.9 mg (59%) of Compound 120 as a colorless glass:

[α]_D ²³ +5.5" (c=1.46, CHC1₃) ; IR (CHC1₃) 2950, 2920, 1715, 1650, 1265, 1105, 840 cm^"1; H NMR (250 HMz, CDC1₃) δ - 0.02, 0.01, 0.09 and 0.15 (s each, 3H each), 0.84 and 0.94 (s each, 9H each), 3.23 (dd, IH, J=3.9, 8.7 Hz), 3.43 (m, IH) ,

3.66-3.86 (6H) , 4.02 (t, IH, J=8.3 Hz) , 4.07 (t, IH, J=9.2 Hz) ,

4.17 (t, IH, J=2.8 Hz), 4.29 (t, IH, J=5.9 Hz) , 4.46-4.88

(10H) , 5.02 (dd, IH, J=3.5, 6.1 Hz) , 5.49 (s, IH) , 5.53 (t, IH,

J=3.8 HZ), 5.66 (brs, IH) , 6.53 (d, IH, J=6.1 Hz) , 7.16-7.62 (21H) , 8.04 (d, 4H, J=7.4 Hz) ; ¹³C NMR (63 MHz, CDCl₃) δ -

5.3, -5.0, -4.2, -3.9, 14.0, 18.1, 18.4, 22.6, 25.6, 26.1, 26.2, 31.5, 33.7, 62.6, 69.0, 69.4, 71.2, 72.5, 73.5, 73.6, 74.4, 74.8, 74.9, 75.1, 76.1, 98.6, 101.9, 102.0, 127.2, 127.3, 127.5, 127.6, 127.7, 127.8, 128.1, 128.2, 128.3, 128.4, 128.5, 129.7, 129.8, 133.0, 133.3, 138.0, 138.7, 145.7, 165.8, 166.0;

Analysis calculated for ^C65^H82^I2°14^Sx2^: ' ⁵⁵*⁸⁷' ^H' 5.91; Found: C, 56.19; H, 6.11.

Tetrasaccharide Compound 121: Glycal Compound 120 (212.0 mg) and l,2,3,4-di-0- isopropylidene-D-glactopyranose, Compound 111 (43.4 mg) gave 248.8 mg (92%) of Compound 121 as a colorless glass:

[α]_D ²³ +32.1° (c=0.90, CHC1₃) ; IR (CHCl₃) 2930, 1720, 1265, 1110, 1070, 845 cm^"1; ¹H NMR (250 MHZ , CDC1₃) δ -0.12, -0.08, 0.06 and 0.05 (s each, 3H each), 0.81 and 0.84 (s each, 9H each), 1.34 (s, 6H) , 1.43 and 1.59 (s each, 3H each), 3.18- 3.31 (m, 2H) , 3.55-4.14 (12H) , 4.20-4.90 (16H) , 5.30 (s, IH) , 5.40 (br s, IH) , 5.53 (d, J=5.0 Hz, IH) , 5.66 (br s, IH) , 7.15-7.68 (21H) , 8.08-8.18 (m, 4H) ; ¹³C NMR (63 MHz, CDC1₃) δ -5.4, -4.9, -4.5, -4.1, 18.0, 18.5, 24.5, 25.0, 26.0, 26.07, 26.13, 30.4, 33.8, 61.8, 63.2, 66.3, 67.2, 68.7, 70.0, 70.6, 70.7, 70.9, 71.0, 72.2, 73.2, 73.6, 74.1, 74.2, 74.9, 75.8, 96.3, 101.1, 101.7, 108.6, 109.4, 127.2, 127.3, 127.6, 127.7, 128.0, 128.1, 128.3, 128.4, 128.8, 129.9, 130.0, 133.0, 133.8, 138.0, 138.6, 165.0, 166.0; FAB MS, m/e 1783 (M + H)⁺; Analysis calculated for ^C77^Hιoi^I3°20^Sx2^{: C}' 51.86; ^H' 5.71; Found: C, 51.97; H, 5.75.

Glycal Compound 122:

Glycal compound 117 (334.1 mg) was converted into the corresponding diol (253.9 mg, 99%) and then into 264.7 mg (79%) of Compound 122 by following the procedure for the preparation of glycal Compound 119. This colorless oil exhibited: -^α-D²² "¹⁴-⁹° (C=0.52, CHC1₃) ; IR (CHCl₃) 3010, 1650,

1210, 1120, 840 cm^"1; ^XH NMR (250 MHz, CDC1-.) δ 0.09 and 0.15 (s each, 3H each), 0.10 (s, 6H) , 0.92 and 0.94 (s each, 9H each), 3.30 (dd, IH, J=4.2, 8.6 Hz), 3.69-3.94 (8H) , 4.05-4.10 (m, 2H) , 4.48-4.72 (6H) , 4.87 (d, IH, J=10.7 Hz) , 4.98 (dd, IH, J=2.7, 6.1 Hz), 5.39 (s, IH) , 6.21 (d, IH, J=6.1 Hz) , 7.17-

7.48 (15H) ; ¹³C NMR (63 MHz, CDC1₃) <_ -5.2, -5.0, -4.7, -4.2,

18.1, 18.5, 25.9, 26.0, 33.5, 61.7, 68.1, 71.1, 72.7, 73.5,

75.2, 77.2, 79.4, 79.5, 100.8, 103.4, 127.4, 127.5, 127.6, 127.7, 127.8, 128.0, 128.3, 128.4, 137.9, 138.4, 138.5, 144.6; FAB MS, m/e 917 (M + H)⁺.

Glycal Compound 123: Glycal Compound 122 (120.7 mg) and 3,6-di-O-benzoyl- D-glucal, Compound 19 (51.3 mg) gave 117.2 mg (64%) of Compound 123 as a colorless glass:

[α]_D ²³ -9.2° (c=0.54, CHC1₃); ^XH; NMR (250 MHz, CDCL₃) δ 0.01, 0.07, 0.08m and 0.13 (s each, 3H each), 0.91 and 0.92 (s each, 9H each), 3.20-3.30 (m, 2H) , 3.56-3.84 (5H) , 4.00 (t, IH, J=9.3 Hz), 4.10 (t, IH, J=8.7 Hz) , 4.23 (m, IH) , 4.30 (t, IH, J=5.3 Hz), 4.49-4.78 (10H) , 4.87 (d, IH, J=11.0 Hz) , 5.02 (dd, IH, J=3.7, 6.2 Hz), 5.42 (t, IH, J=3.9 Hz) , 5.51 (s, IH) , 5.52 (s, 1H0, 6.53 (d, IH, J=6.2 Hz) , 7.18-7.66 (21H) , 8.03- 8.08 (m, 4H) ; ¹³C NMR (63 MHz, CDC1₃) δ -5.5, -4.9, -4.5, - 3.9, 18.1, 18.3, 25.7, 25.9, 26.0, 33.5, 61.3, 62.3, 68.5, 68.7, 69.0, 71.2, 71.3, 73.4, 74.3, 74.4, 75.0, 75.9, 76.0, 76.9, 79.0, 98.3, 101.6, 127.4, 127.5, 127.8, 128.0, 128.2, 128.4, 128.5, 129.6, 129.7, 133.2, 137.7, 138.4, 138.5, 145.8, 165.7, 166.0; Analysis calculated for ^cg5^H82^I2⁰14^S 2^{: C}' 55.87; H, 5.91; Found: C, 56.28; H, 6.07.

Tetrasaccharide Compound 124: Glycal Compound 123 (84.3 mg) and 1,2,3,4-di-O- isopropylidene-D-galactopyranose, Compound 111 (17.3 mg) gave 77.8 mg (72%) of Compound 124 as a colorless glass:

[C.]_D ²³ +2.8^β (C=0.86, CHC1₃); 3010, 2930, 1720, 1265, 1075 cm^"1; H NMR (250 MHz, CDCl₃) ό": -0.06, -0.05, 0.05 and 0.14 (s each, 3H each), 0.88 and 0.91 (s each, 9H each), 1.36, 1.38, 1.47 and 1.62 (s each, 3H each), 3.15-3.28 (m, 3H) , 3.49- 3.60 (m, 3H) , 3.68-4.90 (24H) , 5.34 (s, IH) , 5.45 (s, IH) , 5.52 (s, IH) , 5.56 (D, j=5.0 Hz, IH) , 7.10-7.64 (21H) , 8.04-8.23 (m, 4H) ; ¹³C NMR (63 MHz, CDC1₃) δ -5.5, -4.9, -4.6, -4.0, 18.1, 18.3, 24.6, 25.0, 26.0, 26.1, 26.3, 30.4, 33.3, 60.6, 63.3, 66.2, 67.4, 68.1, 69.8, 70.7, 70.9, 71.2, 72.9, 73.3, 74.4, 74.9, 75.8, 76.7, 78.6, 96.3, 101.4, 103.7, 105.2, 108.8, 109.4, 127.3, 127.4, 127.6, 127.7, 127.8, 128.0, 128.1, 128.2, 128.4, 128.8, 129.9, 130.0, 130.1, 133.1, 133.8, 137.8, 138.6, 165.0, 166.1; Analysis calculated for ^C77^Hιni^I3°20^Sx2^{: C}' 51.86; H, 5.71; Found: C, 52.39; H, 5.98.0.

The present invention has been described with respect to preferred embodiments. It will be clear to those skilled in the art that modifications and/or variations of the disclosed subject matter can be made without departing from the scope of the invention set forth herein.

Claims

WHAT IS CLAIMED IS:

1. A method of preparing a glycal-terminated saccharide multimer that comprises converting a substituted glycal having only non-participating substituents in the glycal ring into a substituted 1,2-anhydrosugar derivative thereof, glycosylating a glycal derivative having a single reactive hydroxyl group with said substituted 1,2-anhydrosugar derivative, and thereafter repeating said converting and glycosylating steps seriatim.

2. The method according to claim 1 wherein the substituents of the terminal glycal are non-participating substituents when the individual substituted 1,2-anhydrosugars are formed.

3. The method according to claim 2 wherein said substituted 1,2-anhydrosugar is formed by the reaction of said glycal with a dialkyl dioxirane having a total of 2 to about 6 carbon atoms in the alkyl groups.

4. The method according to claim 1 including the further step of recovering the glycal-terminated saccharide multimer.

5. A method of preparing an oligosaccharide comprising the steps of:

(a) converting the terminal substituted glycal portion of the product of claim 1 into a terminal substituted 1,2-anhydrosugar;

(b) reacting said terminal substituted 1,2- anhydrosugar with the hydroxyl group of a substituted sugar derivative other than a glycal derivative to form a reaction product that is an oligosaccharide terminated with a substituted sugar derivative; and

(c) recovering the reaction product of step (b) .

6. A method of preparing a glycoside derivative that comprises the steps of:

(a) reacting a substituted glycal having a ring containing five or six atoms and a plurality of non- participating substituents with a dialkyl dioxirane containing a total of two to about six carbon atoms in the alkyl groups to form an intermediate substituted 1,2-anhydrosugar; and

(b) reacting said intermediate substituted 1,2- anhydrosugar with a nucleophile having a nucleophilic atom selected from the group of consisting of oxygen, nitrogen and sulfur in the presence of a Lewis acid and absence of water to form an epoxide-opened glycoside reaction product.

7. The method according to claim 6 wherein said glycal has a ring containing six atoms.

8. The method according to claim 6 wherein said non- participating substituents are selected from the group consisting of 3-position hydroxyl, hydrogen, C,-Cg alkyl and O- ether substitutents.

9. The method according to claim 8 wherein said glycoside reaction product is an oligosaccharide.

10. The method according to claim 6 wherein said substituted glycal is an oligosaccharide derivative.

11. The method according to claim 6 wherein said Lewis acid is selected from the group consisting of SnCl₄, AgC10₄, BF₃, trimethylsilyl triflate, Zn(triflate)₂,

Mg(triflate) ₂, MgCl₂, A1C1₃, ZnBr₂ and ZnCl₂.

12. The method according to claim 6 wherein said dialkyl dioxirane is dimethyl dioxirane.

13. A method of preparing an oligosaccharide derivative comprising the steps of:

(a) reacting a substituted 1,2-anhydrosugar derivative having a ring containing five or six atoms and a plurality of non-participating substituents with a hydroxyl group of glycal derivative having five or six atoms in the ring and a plurality of substituents to form an epoxide-opened glycosyl glycal reaction product; (b) reacting said glycosyl glycal reaction product with a dialkyl dioxirane containing a total of two to about six carbon atoms in the alkyl groups to form a glycosyl substituted 1,2-anhydrosugar intermediate; and (c) reacting said glycosyl substituted 1,2- anhydrosugar intermediate with a nucleophile having a nucleophilic atom selected from the group consisting of oxygen, nitrogen and sulfur in the presence of a Lewis acid and absence of water to form an epoxide- opened glycosyl glycosyl glycoside reaction product in which said nucleophilic atom is bonded to the anomeric atom at the terminal glycoside bond.

14. The method according to claim 13 wherein said non-participating substitutents are selected from the group consisting of hydrogen, C^-Cg alkyl, a 3-position hydroxyl and O-ether groups.

15. The method according to claim 13 wherein said substituted 1,2-anhydrosugar of step (a) is itself an oligosaccharide derivative and said method extends the chain of said original oligosaccharide.

16. The method according to claim 13 including the further steps of removing any participating substituent present in said glycosyl glycal and replacing that participating substituent with a non-participating substituent prior to step (b).

17. The method according to claim 13 wherein said substituted 1,2-anhydrosugar and said substituted glycal of step (a) each have six atoms in their respective rings, and each of said anhydrosugar and glycal contains an O-ether substituent at the 3-, 4- and 6-positions of its ring.

18. The method according to claim 13 wherein said nucleophile of step (c) is a sugar derivative whose nucleophilic atom is oxygen.

19. The method according to claim 13 including the further step of recovering said epoxide-opened glycoside reaction product of step (c) .

20. A method of preparing an oligosaccharide derivative that comprises the steps of:

(a) reacting a substituted 1,2-anhydrosugar having a ring containing five or six atoms and a plurality of non-participating substituents selected from the group consisting of hydrogen, C.-Cg alkyl, a 3-position hydroxyl and O-ether groups in the presence of a Lewis acid and absence of water with the hydroxyl group with a glycal derivative having five or six atoms in the ring and a plurality of participating or non-participating substituents to form an epoxide- opened glycosyl glycal reaction product having a plurality of substituents;

(b) removing any participating substituent that may be present in said glycosyl glycal reaction product and replacing said removed participating substituent with a non-participating substituent to form a glycosyl glycal reaction product containing a plurality of substituents that are all non- participating;

(c) reacting said glycosyl glycal reaction product containing a plurality of substituents that are all non-participating with a dialkyl dioxirane containing a total of two to about six carbon atoms in the alkyl groups to form a corresponding glycosyl substituted 1,2-anhydrosugar intermediate; (d) reacting the glycosyl-substituted 1,2-anhydrosugar with the hydroxyl group of a sugar derivative having a ring containing five or six carbon atoms and a plurality of participating and nonparticipating substituents in the presence of a Lewis acid and absence of water to form an epoxide- opened reaction product in which the oxygen atom of the hydroxyl of said sugar derivative is bonded to the anomeric atom of the reacted substituted 1,2- anhydrosugar portion of said glycosyl substituted 1,2-anhydrosugar.

21. The method according to claim 20 including the further step of recovering the reaction product formed in step (d).

22. The method according to claim 21 wherein said dialkyl dioxirane is dimethyl dioxirane.

23. The method according to claim 20 wherein the sugar derivative of step (d) is a glycal derivative.

24. A method of forming a substituted 1,2- anhydrosugar comprising reacting a substituted glycal having five or six atoms in the ring and whose substituents are non- participating with a dialkyl dioxirane having a total of two to six carbon atoms in the alkyl groups.

25. The method according to claim 24 including the further step of isolating said substituted 1,2-anhydrosugar.

26. The method according to claim 25 wherein said dialky dioxirane is dimethyl dioxirane.

27. The method according to claim 26 wherein said substituted 1,2-anhydrosugar formed is selected from the group consisting of an ether substituted glucopyranose, ether substituted galactopyranose and ether substituted altropyranose.

28. A method of preparing a particle-linked glycal- terminated saccharide multimer that comprises providing a particle-linked substituted 1,2-anhydrosugar having only non- participating substituents, glycosylating a glycal derivative having a single reactive hydroxyl group to form a corresponding substituted glycosyl glycal, converting the corresponding substituted glycosyl glycal formed into a corresponding substituted 1,2-anhydrosugar thereof, glycosylating a glycal derivative having a single reactive hydroxy group with said corresponding substituted 1,2-anhydrosugar, and repeating said converting and glycosylating steps seriatim.

29. The method according to claim 28 wherein said glycal-terminated saccharide is a glycal-terminated oligosaccharide.

30. The method according to claim 29 including the further step of recovering said glycal-terminated oligosaccharide.

31. The method according to claim 30 wherein each glycal derivative having a single reactive hydroxyl group also contains a plurality of non-participating substituents.

32. A method of preparing a glycal-terminated saccharide multimer that comprises cleaving the glycal- terminated saccharide multimer formed according to claim 28 from the particle, and recovering said saccharide multimer.

33. A particle-linked substituted glycal.

34. The particle-linked substituted glycal according to claim 33 wherein said substituted glycal contains a chain of five or six carbon atoms.

35. The particle-linked substituted glycal according to claim 33 wherein said substituted glycal contains a ring of six atoms.

36. A particle-linked substituted 1,2-anhydrosugar.

37. The particle-linked substituted 1,2-anhydrosugar according to claim 36 wherein said substituted 1,2-anhydrosugar contains a chain of five or six carbon atoms.

38. The particle-linked substituted 1,2-anhydrosugar according to claim 36 wherein said substituted 1,2-anhydrosugar contains a ring of six atoms.

39. A method of preparing a substituted 1,2- anhydrosugar that comprises oxidizing a substituted glycal with a dialkyl dioxirane having a total of 2 to about 6 carbon atoms.

40. The method according to claim 39 wherein said dialkyl dioxirane is dimethyl dioxirane.

41. A method of preparing a glycal-terminated halo- substituted saccharide multimer that comprises haloglycosylating a glycal derivative having a plurality of substituent groups and a single reactive hydroxyl group with a substituted glycal having non-participating electron donating substituents, the substituents on said glycal derivative being electron withdrawing relative to the substituents on said substituted glycal and a halonium ion reagent in the absence of water, removing an electron withdrawing substituent and replacing said substitutent with an electron donating non- participating substituent, and repeating said haloglycosylation.

42. The method according to claim 41 wherein said glycal-terminated halo-substituted saccharide is an oligosaccharide.

43. The method according to claim 42 including the further step of recovering the glycal-terminated halo- substituted oligosaccharide.

44. The method according to claim 43 wherein each halo-substituted glycal-terminated oligosaccharide is recovered prior to each repeated haloglycosylation.

45. The method according to claim 41 wherein said halonium ion reagent is I(sym-collidine)₂C10₄ or Br(sy_n- collidine)₂C10₄.

46. The method according to claim 41 including the further steps of removing an electron withdrawing substituent after the last haloglycosylation.

47. A method of preparing a halosubstituted saccharide that comprises haloglycosylating an alcohol other than a glycal and that has a single reactive hydroxyl group with the substituted glycal-terminated halo-substituted saccharide multimer prepared in claim 1 and a halonium ion reagent in the absence of water, and recovering the product of the reaction.

48. The method according to claim 47 wherein said alcohol is a portion of a substituted sugar derivative.

49. The method according to claim 48 wherein said alcohol is a portion of a non-sugar molecule.

50. A method of preparing a halosubstituted oligosaccharide comprising the steps of: (a) reacting a substituted glycal containing five or six atoms in the ring and having a plurality of non-participating electron donating substituents with a halonium ion reagent and a glycal derivative having five or six atoms in the ring, a single reactive alcohol and an electron withdrawing acyl substituent in the absence of water to form a halogenated glycal-terminated oligosaccharide reaction product having a halogen group at the 2-position of a glycosyl ring trans to a glycosyl bond; and

(b) reacting said halogenated glycal-terminated oligosaccharide reaction product with an alcohol and a halonium ion reagent in the absence of water to form a further reaction product having further a glycoside bond and a halogen on a carbon adjacent to said further glycoside bond in which said further glycosidic bond and the halogen bonded to the adjacent carbon atom are trans to each other.

51. The method according to claim 50 wherein said substituted glycal has a ring containing six atoms.

52. The method according to claim 50 wherein said glycal derivative has a ring containing six atoms.

53. The method according to claim 50 wherein said non-participating electron donating substituents are selected from the group consisting of hydrogen, C.-Cg alkyl and O-ether groups.

54. The method according to claim 53 wherein said substituted glycal contains an O-ether substituent at the 3-, 4- and 6-positions of its ring, and said glycal derivative contains an O-acyl substituent at two of the 3-, 4- and 6-positions of its ring.

55. The method according to claim 50 wherein said substituted glycal is an oligosaccharide derivative.

56. The method according to claim 50 wherein said alcohol of step (b) is a hydroxyl group of a non-sugar molecule.

57. The method according to claim 50 wherein the alcohol is a portion of a sugar derivative.

58. The method according to claim 57 wherein said sugar derivative is a derivative of glucose having a single reactive alcohol.

59. The method according to claim 50 including the further step of recovering said further reaction product.

60. The method according to claim 50 wherein said halonium ion reagent is I(sym-collidine)₂C10₄.

61. The method according to claim 50 including the further step of removing the 2-position halogen formed in step (a) and replacing it with another substituent prior to step (b).

62. A method of preparing an iodosubstituted oligosaccharide comprising the steps of:

(a) reacting a substituted six-carbon glycal containing six atoms in the ring and having non- participating electron donating substituents that are hydrogen, O-ether or C,-Cg alkyl substituents at the 3-, 4-, and 6-positions with a six carbon glycal derivative containing six atoms in the ring, a single hydroxyl group and having O-acyl substituents at two of the 3-, 4- and 6-positions and I (sym- collidine)₂C10₄ in the absence of water to form an iodinated glycal-terminated oligosaccharide reaction product having an iodide group at the 2-position of a glycosyl ring that is trans to a glycosidic bond; (b) removing said O-acyl substituents present in said iodinated glycosyl glycal and replacing the removed substituents with electron donating substituents to form an iodinated substituted glycal- terminated oligosaccharide free of electron withdrawing substituents;

(c) reacting said iodinated substituted glycal- terminated oligosaccharide with an alcohol and I(sy_n- collidine)₂C10₄ in the absence of water to form a further reaction product containing a further glycosidic bond and an iodide group bonded to the carbon atom adjacent to said further glycosidic bond and in a configuration trans thereto; and

(d) recovering said further reaction product.

63. The method according to claim 62 wherein the 6-position of said substituted glycal contains a O-ether substituen .

64. The method according to claim 63 wherein each of said O-ether substituents in said substituted glycal and said iodinated substituted glycal-terminated oligosaccharide are selected from the group consisting of O-benzyl, tri-C-,-Cg alkylsilyl, phenyl-di-C-^Cg alkyl silyl and diphenyl C -Cg alkyl silyl groups.

65. The method according to claim 63 wherein the O-ether substituent at the 6-position of said substituted glycal is a substituted sugar derivative.

66. The method according to claim 63 wherein said nucleophilic alcohol of step (c) is a portion of a sugar derivative.

67. A method of preparing a haloglycosyl glycal comprising haloglycosylating (a) a substituted glycal containing 5-9 carbon atoms in a chain, a ring containing five or six atoms and a plurality of nonparticipating electron donating substituents with (b) a glycal derivative containing 5-9 carbon atoms in a chain, a ring containing five or six atoms, a reactive hydroxyl group and electron withdrawing substitutents and (c) a halonium ion reagent to form a halosubstituted glycosyl glycal having a halide at the 2- position of the glycosyl ring that is trans to the glycoside bond.

68. The method according to claim 67 wherein said non-participating electron donating substitutents are hydrogen, C^-C₈ alkyl and O-ether groups.

69. The method according to claim 67 wherein said electron withdrawing substituents include one O-acyl group.

70. A method of preparing a particle-linked glycoside that comprises glycosylating a solid phase particle-linked nucleophile having a single nucleophilic atom with a liquid composition free of water containing a substituted 1,2-anhydrosugar having a plurality of non-participating substituents to form a particle-linked epoxide ring-opened glycoside derivative, and separating the solid and liquid phases.

71. The method according to claim 70 wherein one of said plurality of non-participating substituents is a protected nucleophilic hydroxyl group whose protecting moiety can be selectively removed in the presence of any other substituent that is present.

72. The method according to claim 71 including the further steps of selectively removing said protecting moiety to provide said nucleophilic hydroxyl group in unprotected form, glycosylating said nucleophilic hydroxyl group with a liquid composition free of water containing a substituted

1,2-anhydrosugar having a plurality of non-participating substituent groups to form a particle-linked epoxide ring- opened glycosyl glycoside, and separating the solid and liquid phases.

73. The method according to claim 70 wherein the link between said nucleophile and said particle is selectively severable.

74. The method according to claim 70 wherein said nucleophilic atom is the oxygen of a hydroxyl group.

75. A method of preparing a particle linked saccharide multimer comprising the steps of:

(a) glycosylating a solid phase-linked nucleophile having a nucleophilic atom with a liquid composition free of water containing an excess of a substituted 1,2-anhydrosugar containing a ring of 5 or 6 atoms, 5 through 9 carbon atoms in the chain, and a plurality of non-participating substituents that include a protected nucleophilic hydroxyl group whose protecting moiety can be selectively removed in the presence of any other substituent that is present to form a particle-linked epoxide ring-opened glycoside having a hydroxyl group at the 2-position of the formed glycosyl ring; (b) separating the solid and liquid phases;

(c) selectively removing said protecting moiety in the presence of any other substituents present to form a particle-linked glycoside having a free nucleophilic hydroxyl group; (d) glycosylating said free hydroxyl group of the particle-linked glycoside of step (c) with a liquid composition free of water containing a substituted 1,2-anhydrosugar containing a chain of 5 through 9 carbon atoms, a ring containing 5 or 6 atoms and a plurality of non-participating substituents to form a particle-linked epoxide ring-opened glycosyl glycoside having a hydroxyl group of the 2-position of the formed glycosyl ring; and (e) separating the solid and liquid phases.

76. The method according to claim 75 wherein said nucleophile is the terminal amino group of a protected polypeptide.

77. The method according to claim 75 wherein said glycosylation reaction of step (d) is carried out in the presence of a Lewis acid catalyst.

78. The method according to claim 77 wherein said nucleophile is a hydroxyl group of a sugar derivative.

79. The method according to claim 78 including the further steps of protecting the 2-position hydroxyl group formed in step (a) to form a 2-protected hydroxyl substituent prior to carrying out step (c) .

80. The method according to claim 75 wherein said nucleophile is linked to said particle by a selectively severable bond.

81. The method according to claim 75 wherein said substituted 1,2-anhydrosugar of step (d) includes a saccharide unit as a substituent.

82. A method of preparing a particle-linked saccharide multimer glycoside derivative comprising the steps of:

(a) glycosylating a solid phase-linked nucleophile having a single nucleophilic atom per molecule with a liquid composition free of water containing a Lewis acid catalyst, an excess of a substituted 1,2-anhydrosugar containing a chain of 5 through 9 carbon atoms, a ring having 5 or 6 atoms, and a plurality of non-participating substituents that include a protected nucleophilic hydroxyl group whose protecting moiety can be selectively removed in the presence of any other substituent that is present to form a particle-linked epoxide ring-opened glycoside derivative having a hydroxyl group at the 2-position of the formed glycosyl ring; (b) separating the solid and liquid phases;

(c) protecting the 2-position hydroxyl group formed to form a protected 2-position hydroxyl substituent;

(d) selectively removing said protecting moiety in the presence of any other substituents present to form a particle-linked glycoside derivative having a free nucleophilic hydroxyl group;

(e) glycosylating said free hydroxyl group of the particle-linked glycoside with a liquid composition free of water containing a Lewis acid catalyst and an excess of a substituted 1,2-anhydrosugar containing a chain of 5 through 9 carbon atoms, a ring containing 5 or 6 atoms and a plurality of non-participating substituents to form another particle-linked epoxide ring-opened glycoside derivative having a hydroxyl group at the 2-position of the formed glycosyl ring;

(f) separating the solid and liquid phases; and

(g) repeating steps (c) , (d) , (e) and (f) seriatim.

83. The method according to claim 82 wherein said substituted 1,2-anhydrosugar of step (a) includes a saccharide unit as a substituent.

84. The method according to claim 82 wherein said nucleophile is the hydroxyl group of a substituted sugar.

85. A method of preparing a particle-linked glycoside that comprises haloglycosylating a solid phase particle-linked nucleophile molecule having a single reactive hydroxyl group per molecule with a liquid composition free of water containing a substituted glycal having a plurality of substituents and a halonium ion reagent to form a particle-linked glycoside, and separating the solid and liquid phases.

86. The method to claim 85 wherein one of said plurality of substituents is a protected nucleophilic hydroxyl group whose protecting moiety is selectively removable in the presence of any other substituent that is present.

87. The method according to claim 86 the further steps of selectively removing said protecting moiety to provide said nucleophilic hydroxyl group in unprotected form, haloglycosylating said unprotected nucleophilic hydroxyl group with a liquid composition free of water containing a substituted glycal having a plurality of substituent groups to form a particle-linked glycosyl glycoside, and separating the solid and liquid phase.

88. The method according to claim 85 wherein the link between said nucleophile and said particle is selectively severable.

89. The method according to claim 85 wherein said nucleophilic atom is the oxygen of a hydroxyl group of a sugar derivative.

90. A method of preparing a particle-linked saccharide multimer comprising the steps of;

(a) haloglycosylating a solid phase particle- linked nucleophile having a single hydroxyl group per molecule with a liquid composition free of water containing a chain of 5 or 6 atoms, 5 through 9 carbon atoms in the chain, and a plurality of substituents that include a protected nucleophilic hydroxyl group whose protecting moiety can be selectively removed in the presence of any other substituent that is present and an excess of halonium ion reagent to form a particle-linked glycoside having a halogen group at the 2-position of the formed glycosyl ring; (b) separating the solid and liquid phases;

(c) selectively removing said protecting moiety in the presence of any other substituents present to form a particle-linked glycoside having a free nucleophilic hydroxyl group;

(d) haloglycosating said free hydroxyl group of the particle-linked glycoside of step (c) with a liquid composition free of water containing a halonium ion reagent and a substituted glycal containing a chain of 5 through 9 carbon atoms, a ring containing 5 or 6 atoms and a plurality of substituents to form a particle-linked glycosyl glycoside having a halogen group at the 2-position of the formed glycosyl ring; and (e) separating the solid and liquid phases.

91. The method according to claim 90 wherein said haloglycosylation reactions of steps (a) and (d) are carried out in the presence of powdered 4A molecular sieves.

92. The method according to claim 91 wherein said nucleophilic hydroxyl group of step (a) is a hydroxyl group of a sugar derivative.

93. The method according to claim 92 including the further steps of removing the 2-position halogen group with a protected substituent prior carrying out step (c) .

94. The process according to claim 92 wherein said nucleophile is linked to said particle by a selectively severable bond.

95. The process according to claim 90 wherein said substituted glycal of step (d) is a substituted glycal- terminated oligosaccharide.

96. A method of preparing a particle linked saccharide multimer glycoside comprising the steps of: (a) haloglycosylating a solid phase particle- linked nucleophile having a single nucleophilic hydroxyl group per molecule with a liquid composition free of water containing an excess of a substituted glycal having a backbone chain of 5 through 9 carbon atoms, a ring having 5 or 6 atoms, and a plurality of substituents that include a protected nucleophilic hydroxyl group whose protecting moiety can be selectively removed in the presence of any other substituent that is present and an excess of a halonium ion reagent to form a particle-linked glycoside having a halogen group at the 2-position of the formed glycosyl ring;

(b) separating the solid and liquid phases; (c) selectively removing said protecting moiety in the presence of any other substituents present to form a particle-linked glycoside having a free nucleophilic hydroxyl group;

(d) haloglycosylating said free hydroxyl group of the particle-linked glycoside of step (c) with a liquid composition free of water containing an excess of a substituted glycal having a containing 5 or 6 atoms and a plurality of substituents that include a protected nucleophilic hydroxyl group whose protecting moiety can be selectively removed in the presence of any other substituent that is present and an excess of a halonium ion reagent;

(e) separating the solid and liquid phases; and

(f) repeating steps (c) , (d) and (e) seriatim.

97. The method according to claim 96 wherein said substituted glycal of step (a) is a substituted glycal- terminated oligosaccharide.

98. The method according to claim 96 wherein said nucleophile is the hydroxyl group of a substituted sugar.

99. The method according to claim 96 wherein said halonium ion reagent of step (d) is I(syra-collidine)C10₄.

100. The method according to claim 96 including the further steps of removing the 2-position halogen group formed in step (a) and replacing it with another substituent prior to step (c) .