METHOD OF PREPARING PERACETYL OXAZOLINES

Field of the Invention This invention relates to a direct method of syn¬ thesizing peracetyl oxazolines from peracetyl saccha- rides.

Background of the Invention

Glyσoproteins are one of the major components of animal cell surfaces. The cell surfaces are usually coated with carbohydrate molecules which are attached to specific surface proteins to form glycoproteins. The glycoproteins are present in the outer protein layer of the cell's plasma membrane.

These cell surfaces contain specific recognition sites that interact with biological substances. For example, membrane surfaces contain certain areas cap¬ able of acting as antigens. Researchers are investi¬ gating how cells recognize other cells as "foreign" to

understand how to treat tissue rejection and autoim¬ mune diseases. N-linked glycan chains of glycopro¬ teins (N-glycoproteins) are considered to be respon¬ sible for many of the cell site's biological recogni¬ tion mechanisms. Thus, the understanding of the N- glycoprotein saccharide processing and function is of major interest. To study the N-glycoproteins, these compounds need to be first synthesized.

One of the steps in N-glycoprotein biosynthesis is the formation of "lipid intermediates." The oligosac- charide chains of the N-glycoproteins are assembled on a "lipid intermediate" prior to transfer to protein. Thus, oligosaccharide "lipid intermediates" are re¬ quired as exogenous glycosyl acceptors for studies of N-glycoprotein biosynthesis. Herscovics, et al. , FEBS Lett,. 156:298-302 (1983); Sasak et al., J. Biol. Che . , 259_:332-337 (1984).

Oligosaccharides with structures corresponding to those in the N-glycoprotein saccharide "core" may be isolated from the urine of animals with swainsonine- induced alpha-mannosidosis. Sadeh et al. , FEBS Lett. , 163:104-109 (1983); Daniel et al., Biochem. J. , 221:601-607 (1984). These oligosaccharides may be isolated also by chemical [M. Fukuda et al. J. Biochem

(Tokyo) 8J3: 1223-1232 (1976)] and enzymic [F. K. Chu, J. Biol. Chem., 261: 172-177 (1986); A. L. Tarentino et al. , Biochemistry, 24: 4665-4671 (1985)] degrada¬ tion of glycoproteins, such as ovalbumin or ribonucle- ase B.

These oligosaccharides may also be synthesized for use as lipid intermediates. Synthesis of these oligo¬ saccharides is accomplished by formation of a pera- cetylglycosyl phosphate, then coupling this compound with an "activated" derivative of dolichyl phosphate to produce a peracetyl diphosphate diester. This re¬ sulting compound is O-deacetylated to form the oligo- saccharide "lipid intermediate." Warren et al. , Car- bohydr. Res., 126:61-80 (1984). ^* Chemical synthesis is preferable over the isolation of natural compounds for obtaining suitable glycosyl acceptors since the syn¬ thesis ensures relatively large quantities of pure compounds having known structures.

The peracetylglycosyl phosphate, the first com¬ pound formed in the synthesis of the "lipid intermedi¬ ate," is produced from peracetyl oxazolines. The per¬ acetyl oxazolines are appropriate precursors of per¬ acetylglycosyl phosphates because (a) these compounds provide the alpha anomer in a reaction that involves

net retention of configuration and (b) phosphorylation occurs without any scission or modification of inter- residue glycosidic linkages. Warren et al. , Carbo- hydr. Res., 126:61-80 (1984); Warren et al., Carbo- hydr. Res., 1:181-196 (1978); and. Warren et al. , Carbohydr. Res., £2 ^ϊ85 -1.01 (1981).

A key step then in producing the oligosaccharide "lipid intermediates" is the preparation, in high yield, of a peracetyl oxazoline that can then be phos- phorylated to produce the peracetylglycosyl phosphate, the synthetic precursor of an oligosaccharide "lipid intermediate." The synthesis of a lipid intermediate from a peracetyl oxazoline is as follows (Equation 1) :

( 1 )

_♦ phosphorylation

peracetyl glycosyl phosphate

peracetyl diphosphate diester

lipid intermediate

wherein Ac = CH-CO wherein R = peracetyl (hexose). _._(N-acetylhexosamine) ₀ _

R* = CH-.C..H-

R" = (Hexose) _l _ _{2 (HexNAc) _Q _-

The peracetyl oxazolines are produced from per¬ acetyl oligosaccharides. The synthesis of a peracetyl oxazoline from a peracetyl oligosaccharide is diffi¬ cult due to the alpha (l-»6) linkages in the oligosac¬ charide. These are very labile to the acidic condi¬ tions normally employed for formation of peracetyl oligosaccharide halides, the usual precursor of the peracetyl oxazoline. Also, because the oligosaccha¬ rides contain a di-N-acetylchitobiose residue, any reagents employed must not adversely affect the aceta- mido groups, or cause significant hydrolysis of the beta-OM) linkage between the two glycosyland,noacetyl (GlcNAc) residues. Thus chloroacetolysis, the treat¬ ment of saccharides with HCl in acetyl chloride, which was successfully employed for the preparation of gly- cosyl chlorides from oligosaccharides, cannot be used. (Warren et al. Carbohydr. Res., 61, 92, and 126, supra. )

Alternative reagents for producing an oxazoline from a peracetyl saccharide are also problematic. For example, the use of ferric chloride as the reactive compound is of limited value since it only will react with the beta-D-anomer of a peracetyl saccharide, and the preferred starting material is the alpha peracetyl

saccharide, because this is the anomer readily avail¬ able by the action of pyridine-acetic anhydride on an oligosaccharide. (Matta et al., Carbohydr. Res. , 23.:460-464 (1972)). Further, stannic chloride, as the reactive compound with the alpha-D-anomer of peracetyl glucosamine, is efficient only when used with the monosaccharide. With oligosaccharides as the starting material, the reaction is incomplete, and side react¬ ions produce low yields. (Srivastava, Carbohydr. Res., 1013:286-292 (1982)).

Therefore, it would be desirable to develop a met¬ hod of synthesis for obtaining peracetyl oxazolines from peracetyl saccharides.

Summary of the Invention This invention relates to a method of producing peracetyl oxazolines from peracetyl saccharides. The method involves reacting the starting material, a per¬ acetyl saccharide, with a reactive compound, capable of generating the formation of an intermediate sac¬ charide acetoxonium ion to directly produce the per¬ acetyl oxazoline.

Detailed Description of the Invention In accordance with the invention, a peracetyl oxa¬ zoline can be prepared from a peracetyl saccharide by treatment with a reactive compound capable of genera¬ ting the formation of an intermediate saccharide ace- toxonium ion. Mono-, di-, and oligoperacetyl oxazo¬ lines can be prepared by the process of this inven¬ tion.

The reaction for producing peracetyl oxazolines from peracetyl saccharides, and the use of the per¬ acetyl oxazolines in the synthesis of lipid intermedi¬ ates to then form N-glycoproteins is as follows (Equa¬ tion 2) :

( II )

eracet l saccharide

+ reactive compound

peracetyl oxazoline

+ phosphorylation

peracetylglycosyl phosphate

wherein Ac = CH-CO wherein R = peracetyl (hexose). _- ₂ (N-acetylhexosamine) ₀ __

Any naturally occurring or synthetic mono-, di-, or oligosaccharide that contains a hexosamine or N- acetylhexosamine at the reducing terminus of the sac¬ charide may be used as the starting peracetyl saccha¬ ride in this invention. The peracetyl oligosaccha¬ ride that may be used in this invention, will typical¬ ly be less than 14 residue units. The oligosacchar¬ ides will typically contain 1 to 12 residues of neu¬ tral hexoses, and either 1 or more residues of an N-acetylhexosamine. In addition, the peracetyl sac¬ charide may be either the alpha or the beta anomer. (Warren et al. Carbohydrate Res., 82: 71-83 (1980); Auge et al. , Carbohyrdrate Res., 82: 85-95 (1980); Warren et al., Carbohydrate Res., 92: 85-101 (1981); Warren et al., Carbohyrdrate Res., 116: 171-182 (1983)) .

The peracetyl saccharide is reacted with a reac¬ tive compound capable of generating the formation of an intermediate saccharide acetoxonium ion. Compounds capable of generating the formation of an intermediate saccharide acetoxonium ion to produce the desired per¬ acetyl oxazoline include, but are not limited to, tri- fluoromethanesulfonic acid (triflie acid) , and deriva¬ tives of trifluoromethanesulfonic acid, trimethylsilyl

trifluoromethanesulfonate (TMS triflate), other tri- flates, such triflates of silver and sodium, tri- fluoromethanesulfonic anhydride, and trifluoromethane¬ sulfonyl chloride.

In the process according to this invention, the peracetyl saccharide reacts with the reactive com¬ pound, such as tri lie acid, to directly produce the peracetyl oxazoline via the acetoxonium ion as follows (Equation 3) :

( IH)

acetoxonium ion

peracetyl oxazoline

wherein Ac = CH,CO wherein R = peracetyl (hexose). .-(N-acetylhexosamine)-, -

The peracetyl saccharide is reacted with the reac¬ tive compound in an amount of from about 1:1 to about 1:2 moles per moles of starting material to reactive compound.

The reaction conditions include reaction tempera¬ ture of from about 20°C to about 50 C; reaction time of from about 13 to about 40 hours. The reaction can be followed by thin-layer chromatography on glass plates coated with silica gel in 10:1 (v/v) chloro- form- ethanol.

After the reaction is completed, the produced per¬ acetyl oxazoline can be recovered by means known in the art. In one embodiment, the reaction mixture is made slightly alkaline (pH 8) by addition of triethyl- amine, then the oxazoline is purified by column chro¬ matography on Merck Kieselgel 60 (230-400 mesh) with elution by 100:200:1 toluene—ethyl acetate—aceto- nitrile—triethylamine. The yield is approximately 90% based on the peracetyl saccharide.

The peracetyl oxazolines produced according to the process of this invention have various uses. The per¬ acetyl oxazolines can be used as glycosyl donors for oligosaccharide synthesis, for example as described in Warren et al. , Carbohydrate. Res., 9^:85-101 (1980).

The oxazolines can be used specifically for synthesis of alpha-D-glycosyl phosphates, en route to biosynthe- tic lipid intermediates, for example as described in Warren et al. , Carbohydr. Res., 126:61-80 (1984). The peracetyl oxazolines can be used in the synthesis of glycopeptides for the study of N-glycoprotein-saccha- ride processing and for the study for mammalian endo- beta-N-acetylglucosaminidase.

The peracetyl oxazolines of this invention are also useful for preparing other oligosaccharide deri¬ vatives in addition to N-glycoproteins. Treatment of the peracetyl oxazoline with azides, thiols, and alco¬ hols and other sugars, produce glycopeptides or amino- glycosides, thioglycosides, and O-glycosides, respec¬ tively. With this reaction, a B-glycosidic bond is formed at the reducing N-acetylglucosaminyl residue.

The peracetyl oxazolines of this invention can also be used in the synthesis of a glycopeptide. In this process, the peracetyl oxazoline is reacted with an azide to produce a glycosyl azide. The thus formed glycoside is th^n hydrogenated to produce a glycosyl- amine. The hydrogenation can be accomplished accord¬ ing to means known in the art, such as hydrogenation by acetic acid-water in the presence of palladium.

The glycosylamine is then coupled to an amino acid or a peptide via any exposed carboxylic group in the amino acid or peptide as described in Garg and Jean- loz, Carbohydr. Res., .23: 437-439 (1972).

The peracetyl oxazolines of this invention can further be used in a process for the synthesis of oligoglycosides by reacting a peracetyl oxazoline with a thiol in the presence of borontrifluoride (BF ) . The thiol compound can include alkyl, alkenyl, and aryl thiols and acetates and benzoates of thiols. The peracetyl oxazoline and thiol are reacted together in approximately equimolar equivalent proportions. The thiol compound is present in an amount of about 3-5 mole equivalents. Reaction temperature will depend upon the thiol compound and may be from 20 C to 100°C. The reaction may be followed by thin layer chromato- graphy as described in Ferrier and Furneaux, Methods Carbohydr. Chem, 8.: 251-253 (1980).

Moreover, the oxazolines can be used in the syn¬ thesis of intermediates for the glycosylation of pep- tides, for instance, to provide synthetic antigens. Pinto et al. , Carbohydr. Res., 124:313-318 (1983).

Further, the peracetyl oxazolines can be used for attachment to a solid support for affinity chromato-

graphy. Columns for affinity chromatography using the peracetyl oxazolines of this invention may be prepared by processes well known in the art, including the fol¬ lowing process: (1) the peracetyl oxazolines are reac¬ ted with the methyl ester of an omega-hydroxy fatty acid to produce a glycoside. The methyl ester groups of the glycoside are then saponified to expose a free acid group for coupling to the solid support.

Having now generally described this invention, the same will be better understood by reference to speci¬ fic examples, which are included herein for purposes of illustration only, and are not intended to be lim¬ iting unless otherwise specified _^ .

Example 1 Hydrogen chloride, previously employed by many workers for the preparation of glycosyl chlorides, was tried, with either peracetyl fXr ^M a -(l-?6)τ3-Man-(_r→4)--5 -Glc£NAc-(l-*4)-Glcp_NAc (Compound I) or peracetyl OC-Man-(l-)6 ⁾ -[σCc-Man-(l-_3 ⁾ ]-ofc-Man-(l-τ )-[oC-Man-(l-3) ] -β -Man-(i-^4)--^-Glc£NAc-(l—->4 )-Glcp_NAc, (Compound II), both predominantly the alpha anomers, as the starting compounds. This method was unsatisfactory because of inter-residue bond cleavage. The results are shown in

Table 1. A study was initiated, using 2-acetamido- 1,3,4,6-tetra-0-acetyl-2-deoxy-o-D-glucopyranose (Com¬ pound III) as a model compound, to try to identify a more satisfactory procedure. As can be seen from the results in Table 1, none of the reagents tried pro¬ duced a high yield of a glycosyl halide from the alpha anomer of the starting compound.

The formulas of compounds I, II, and III are shown below.

(I) peracetyl c<-Man-(1-^6)-^-Man-(l-v4)-^GlcpNAc-(1^4) ^■ Glcp_NAc

(II) peracetyl C< -Man- (l-»6 )-[ -Ma n- (l-→3 ) ]-c -Man- (l-»6)-_= -Man- ( l-*3 ) ( l-→4 ) -Glcp_NAc

(III) 2-acetamido-l ,3,4, 6-tetra-0_-acetyl-2-deoxy- o -D glucopyranose

Table 1

Formation of peracetylglycosyl halides from derivatives of 2-acetamido-2-deoxy-D-glucose.

Starting compound Reagent ^a Result

I HCl r40% yield of glycosyl chloride II HCl /v5-i5% yield of glycosyl chloride III TMS-C1 No reaction III TMS-Br* Low yield of glycosyl bromide, decomposition III TiCl ^■ 4** Mixture of compounds ⁰ . decomposition

III TiBr 4*** Mixture of compounds ⁰ . decomposition

* Gillard et al. , Tetrahedron Lett., 22:513-516 (1981). ** Nashed et al. , Carbohydr. Res., 8^:237-252 (1980). *** Paulsen et al. , Chem. Ber., 114:3079-3101 (1981). a. All reactions were conducted at room temperature in 1,2-dichloroethane and the products identified by t.1.c. b. Evidence of major side reactions involving cleavage of glycosidic bonds. c. Compounds included glycosyl halide, oxazoline, and starting material.

Example 2

Because of the problems described in Example 1, a new procedure was developed, involving reaction of a peracetyl oligosaccharide with trifluoromethanesulfonic acid (triflic acid) . This reaction resulted in a direct formation of the oxazoline via the acetoxonium ion. Triflic acid was replaced by trimethylsilyl trifluoro- methanesulfonate (TMS triflate), without any loss of yield (Table 2). Indeed, preliminary H-N.M.R. evi¬ dence indicated that triflic acid was the reactive spe¬ cies when the latter reagent was employed. This method was greatly superior to the use of stannic chloride (Srivastava, Carbohydr. Res., 103:286-292.(1982) ) which was found to be unsatisfactory for the efficient syn¬ thesis of oligosaccharide oxazolines.

When the TMS-triflate procedure was applied to Com¬ pound I, as shown in Example 1, R _p 0.27 (20:1, v/v, chloroform—methanol), the tetrasaccharide oxazoline (R=peracetyl Man_GlcNAc) , R-. 0.31 (same t.l.c. solvent) was obtained in 74% yield. The identity of the product was confirmed by the H-N.M.R. spectrum (S5.89 ppm, J, ^Hz ' H-l) , and by hydrolysis at room tempera¬ ture, with a dilute solution of p-toluenesulfonic acid in acetonitrile, followed by O-deacetylation with sod-

ium methoxide in methanol, reduction with sodium boro- hydride, and comparison by high-pressure liquid chroma- tography (5 um Amino-Spherisorb column, acetonitrile —water 7:3) with an authentic specimen of the alditol derived from fX-Man-(l-→6)- ?-Man-(1→4)--_? -GlcβNAc-α-→ 4)-Glcp_NAc.

Similarly, when the TMS-triflate procedure was ap¬ plied to Compound II, as shown in Example 1, R _p 0.56 (10:1, v/v, chloroform — methanol), the heptasaccha- ride oxazoline (R=peracetyl Man ₅ GlcNAc) , R _p 0.60 (same t.l.c. solvent) was obtained in 90% yield. In neither case was there any t.l.c. evidence of formation of low molecular weight oxazolines indicative of glycosidic bond cleavage.

An important advantage of this new procedure for the synthesis of oligosaccharide oxazolines is that it works equally well with the alpha or beta anomer of the starting peracetyl compound, unlike the ferric chloride method (Matta et al., Carbohydr. Res., 21:460-464 (1972)) which can only utilize the relatively inaces- sible beta anomer. The oxazoline derived from Compound I has been converted into a tetrasaccharide phosphate and employed for the synthesis of a "lipid intermedi¬ ate." The oxazoline derived from compound II has been converted into a glycosyl azide and employed for the synthesis of a heptasaccharide-asparagine derivative.

Table 2

Formation of 2-methyl-(3,4,6-tri-0-acetyl-l,2-dideoxy-c>_r D-glucopyrano)-[2,l-d]-2-oxazoline from 2-acetamido- 1,3, ,6-tetra-0-acetyl-2-deoxy- o-D-glucopyranose (Compound III)

anomer of reagent ^a time yield Compound III (hours) alpha TMS-triflate 16 95% al 1pha ^b Triflic acid 12 97% beta ^c TMS-triflate 0.5 100%

a

A solution of the starting compound (0.1 mmol) in 1,2-dichloroethane was stirred at 50 C with 1.1 equiv. reagent. When t.l.c. (20:1, v/v chloroform—methanol) showed complete reaction, the reaction mixture was made slightly alkaline with excess triethylamine, applied to a column of silica gel (Merck Kieselgel 60; 230-400 mesh) and eluted with 1:2:0.01 toluene—ethyl _Q acetate — riethylamine. The product had Rr„ 0.43, [a] _ U+ll

(C 1-.35, chloroform) and was pure according to t.l.c. and H-N.M.R. spectrum.

^b R_, 0.37, [a]p°+91° (C 1.4, chloroform).

^c R _p 0.34, [a]p°+3° (C 1.75, chloroform).

Although the foregoing invention has been described in some detail by way of illustration and example for purposes of clarity and understanding, it will be obvi¬ ous that certain changes and modifications may be prac¬ ticed within the scope of the invention, as limited only by the scope of the appended claims.