14 Biological Chemistry Part (i) Monosaccharides By P. M. COLLINS Department of Chemistry Birkbeck College Malet Street London WC7 E 7HX 1 Introduction It is four years since monosaccharides were last reviewed in Annual Reports and for a comprehensive survey of the work published during that period the reader is referred to the Specialist Periodical Reports on carbohydrate chemistry the 10th volume of which appears this year. The present report is necessarily highly selec- tive. 2 Glycosides The search for efficient stereospecific syntheses of 0-glycosides continues. Angyal’s have shown that calcium or strontium chloride can affect the outcome of the Fischer glycosidation of sugars in which the two hydroxy-groups adjacent to the anomeric centre have the erythro-configuration.Complex forma- tion occurs with the anomer which has three oxygen atoms in an axial-equatorial- axial sequence on six-membered rings or in a cis-cis sequence on five-membered rings. In this way good yields of methyl aldosides’ and ketosides2 were obtained which are not otherwise readily available. Glycosidations utilizing halide displacements have undergone many refinements over the years and are now used with much success. Lemieux’s group proposed three conditions for the best yields of a-glycosides by halide-ion-catalysed glyco- sidations from glycosyl halides with non-participating groups at C-2. Firstly the concentration of halide ion should be such that anomerization of the glycosyl halide is faster than the displacement of the less reactive a-halide by the alcohol secondly the glycosyl halide should react with the alcohol without assistance from metal ions or highly polar solvents and finally the reaction conditions should minimize the formation of glycosidic products other than the desired a-gly~oside.~ Based upon these considerations several a-linked disaccharides have been synthesized in good yields and in a highly stereoselective manner by the reaction of perbenzylated glycopyranosyl bromides with suitably protected sugar derivatives in the presence of S.J. Angyal C. L. Bodkin,and F. W. Parrish Austral. J. Chem. 1975,28 1541. * S. J. Angyal C. L. Bodkin J. A. Mills and P. M. Pojer Austral. 1.Chem. 1977 30 1259. R. U. Lemieux. K. B. Hendriks R. V. Stick and K.James J. Amer. Chem. SOC..1975,97,4056. 343 344 P. M. Collins tetraethylammonium bromide. The Lea blood-group antigenic determinant 0-(a-L-fucopyranosy1)-(1 -* 4)-[ 0-(P-D-galactopyranosyl)-(l + 3)]-2-acetamido-2-deoxy-D-glucose was synthesized using similar principle^,^ as was 0-(a-L-fuco-pyranosy1)-(1 -+ 2)-[0-(a-D-galactopyranosyl)-(l +3)]-D-galaCtOSe which forms part of the antigenic determinant of blood-group B substance.' One of the key steps in the synthesis of this trisaccharide is the a-D-galactosylation of disaccharide (1) with (2). The D-galactopyranoside derivative (3) is an important intermediate in the synthesis of disaccharide (1) since after fucosylation the 3,4-0-iso-propylidene group was replaced by a 3,4-0-(ethoxyethylidene)group which was regiospecifically opened on partial hydrolysis to give the corresponding axial 4- acetate.P-Glycosides of 0-(a-L-fucopyranosy1)-(1 +4)-[O-(P-D-galacto-pyranosy1)-(1 +3)]-2-acetamido-2-deoxy-~-glucose have also been prepared using either 8-ethoxycarbonyl- or 8-methoxycarbonyl-octanol.6Semisynthetic antigens were then prepared from these esters by conversion into an 8-azidocarbonyfoctyl derivative which reacts with the free amino-groups in bovine serum albumin. Antibodies raised with the trisaccharide antigen precipitated blood-group Lea substance and agglutinated Lea red-blood cells. PhCHl? ,OCH,Ph PhCH,OG4 Br OCH,Ph Me@OCH2Ph PhCH20 CH,Ph (3) The 2-0-benzyl group is the most popular non-participating group in glyco- sylations although others have been used.For example 6-0-acetyl-2-azido-3,4- di-0-benzyl-P-D-glucopyranosyl chloride reacted with benzyl 2-azido-3,4-0- benzyl-0-D-glycopyranosideunder silver-perchlor ate-catalysedconditions to give an a-linked (1+6) disaccharide without participation from the azido-gro~p.~ Syntheses of 1,2-trans -linked disaccharides from glycopyranosyl bromides possessing participating groups at C-2 have been reported' which use silver trifluoromethanesulphonate as the catalyst and 1,1,3,3-tetramethylurea as the pro- ton acceptor. Thus glycosylations at hydroxy-groups situated individually at C-2 C-3 C-4 and C-6 in suitably protected sugar derivatives have been achieved by this method with 2,3,4,6-tetra-O-acetyl-a -D-glucopyranosyl bromide.The result- ing (1-2)- (1+3)- (1-4)- and (1+6)-@-linked disaccharide derivatives were 'R. U. Lemieux and H. Driguez 1.Amer. Chem. Soc. 1975,97,4063. ' R. U. Lemieux and H. Driguez I. Amer. Chem. Soc. 1975,97,4069. 'R. U. Lemieux D. R. Bundle and D. A. Barker J. Amer. Chem. Soc. 1975,97,4076. 'H. Paulsen and W. Stenzel Angew. Chem. Internat. Edn.,1975,14 558. S. Hanessian and J. Banoub Carbohydrate Res. 1977 53 C1-3. Biological Chemistry -Part (i) Monosaccharides 345 respectively formed in a high state of anomeric purity in preparatively significant yields and with much manipulative simplification compared with previously used methods. Glycosylating agents other than glycosyl halides have been studied in the past few years.The well-established orthoester procedure continues to attract attention and it has been shown that the basicity of the alcohol used influences the ratio of a-and @-glycosides formed.' Thus the ethyl derivative (4) gave only the @-glycoside whereas the chloroethyl derivatives (5) (6) and (7) yielded mixtures containing 16 50 and 67% respectively of the corresponding a-glycoside. Disaccharide derivatives are formed when 1-0-aryloxycarbonyl-pyranose derivatives which possess non-participating groups at C-2 are fused with partially blocked sugar derivatives as illustrated for example by the condensation of the 1-0-aryloxycarbonyl mannose derivative (11) with 1,2,3,4-tetra-O-acetyl-p-D-glucopyranose to give the a (1 +6)-linked disaccharide derivative (15)" When participating groups are present at C-2 then the reaction takes a different course and orthoesters are produced in good yields.AcOUO (11) R' =OCOZPh R2 = H Mc (12) R' =OH R2 = H (4) R=OEt (13) R' = C1 R2 = H (5) R = OCH2CEI2C1 (14) R' = H R2=CECH (6) R = OCH2CHC12 CH,O + (7) R = OCHzCC13 (8) R=SEt R'= ACO-(9) R = iEtCPh3 (15) R~=H 0 (10) R= tOCH,-( YOMe OAc oAc Sinay and his co-workers" have introduced what appears to be an excellent Q -glycosylating reagent. They have shown that 2,3,4,6-tetra-O-benzyl-a -D-glycopyranosyl chloride reacts with N-methylacetamide to give 1-0-(N-me thy1)ace timidyl-2,3,4,6 -te tra- 0-benzyl-P -D-glucopyranose in 88Yo yield. This compound readily glucosylates free hydroxy-groups in partially protected sugars [see compound (16)].For example with methyl 2,3,6-tri-O-benzyl-a -D-gluco- pyranoside in anhydrous benzene containing toluene-p-sulphonic acid it gave methyl hepta-0-benzyl-a -D-maltoside in 85% yield which is remarkably good for 'P. J. Garegg and I. Kvarnstrom Acta Chem. Scand. 1976 B30 655. 10 Y. Ishido S. Inaba H. Komura and A. Matsuno J.C.S. Chem. Comm. 1977 90. I' J.-R. Pougny J.-C. Jacquinet M. Nassr D. Duchet M.-L. Milat and P. Sinay J. Amer. Chem. Soc. 1977,99,6762. 346 P. M. Collins ,OCH,Ph the glycosylation of this 4-OH group. Eleven glycosides were prepared in 70-92% yield by this method and it was also extended to include a-L-fucosylation with obvious application in the synthesis of natural products of biological significance.Schuerch12 has extended the use of glycosyl sulphonates to the synthesis of disaccharides. Reactions related to this have been developed in which a free anomeric hydroxy-group in an otherwise fully protected sugar derivative is sulphonated and the product is used in situ to glycosylate the aglycone hydroxy- group. For example Perlin andco- worker^'^activated the anomeric hydroxy-groupof 2,3,4,6-tetra-0-benzyl-D-glucopyranose with trifluoromethanesulphonic anhydride and thence formed an LY-D-(~ -+6)-linked disaccharide in 60% yield by the addition of 1,2,3,4-tetra-O-acetyl-~-D-glucopyrznose. In a different appr~ach'~ to the activation of the anomeric position the hydroxy- group at C-1 in 2,3 5,6-di-O-isopropylidenernannofuranose(12) was treated with dichlorocarbene which was generated from chloroform and aqueous sodium hydroxide in benzene in the presence of the phase-transfer catalyst benzyl- triethylammonium chloride.This gave the a-furanosyl chloride (13) which could be isolated in reasonable yield or treated directly with simple alcohols or partially protected sugars to give glycosides or disaccharides respectively in yields in the range 65-85%. Presumably the dichlorocarbene inserts into the 0-H bond and the dichloromethyl aglycone so formed (or a hydrolysed form of it) undergoes nucleophilic displacement by chloride ion attack at the anomeric centre. Several groups have tried for improvements in glycoside synthesis by activation of the alcohol from which the aglycone is derived.Cyclic amide acetals derived from simple vicinal diols or carbohydrates have been used as a source of aglycones in glycoside and disaccharide syntheses." For example condensation of 1-0-acetyl-2,3,5-tri-O-benzoyl-P -D-ribofuranose with the 3,4-cyclic amide derived from 2-0-mesyl-P -D-arabinopyranoside (17) in the presence of stannic chloride gave the intermediate (18) which after hydrolysis yielded the formyl ester of the disaccharide (19). The elegance of this method stems from the regiospecific open- ing of the cyclic amide ring in which the disaccharide linkage is formed at the equatorially oriented hydroxy-group at C-3. Trialkylstannylation of alcohols increases their nucleophilicity. Therefore a good yield of methyl 2,3,4,6-tetra-0-acetyl-~-D-glucopyranoside was formed when the corresponding a-D-glucosyl bromide was treated with tributylstannyl methiodide l2 R.Eby and C. Schuerch Carbohydrate Res. 1976,50 203. l3 J. Leroux and A. S. Perlin Carbohydrate Res. 1976,41 C8. '4 P. Di Cesare and B. Gross Carbohydrate Res. 1977,58 C1. S. Hanessian and J. Banoub. Tetrahedron Letters 1976 661. 347 Biological Chemistry-Part (i)Monosaccharides OBz OBz + (18) X = =NMe2 C1- (19) X = =O and stannic chloride in dichloromethane. l6 Other glycosides were formed by treatment with the appropriate stannyl alkoxide. Alkoxides derived from primary alcohols gave -glycosides whereas those derived from secondary alcohols gave a-anomers.If the Lewis acid was omitted from the condensation and tetraethyl- ammonium bromide used in its place orthoesters were formed by participation from the 2-0-acetyl group upon the anomerized bromide. In this way several orthoesters were produced in excellent yields. Thus acetobromoglucose gave (10) almost quantitatively when treated with the tributylstannyl alkoxide of methyl 2,3-O-isopropylidene-~-D-ribopyranoside under these reaction conditions. Simple aglycones have been produced” from a variety of dialkyl acetals of dimethylformamide e.g. Me,NCH(OR), in which R = Me CHMe2 PhCH2 CH2CMe3 or cyclohexyl since these have been shown to react with peracetylated sugars in the presence of stannic chloride to give almost quantitative yields of 1,2-trans-gl ycosides.Tritylated sugars have been glycosylated.” For example in the presence of trityl perchlorate the anomeric carbon atom in the orthothioacetate (8) was nucleo- philically attacked by 0-6 of 1,2,3,4-tetra-O-acety1-6-0-tr~tyl-~-D-glucopyranose to give a /3(1 -+ 6)-linked disaccharide derivative. The reaction is thought to occur uia the sulphonium ion intermediate (9). Interest in the solid-phase method for the synthesis of oligosaccharides continues. For example 2,3,4-tri-O-benzyl- 1-thio-P-D-glucopyranose has been attached to a polystyrene support by a thioglycosidic linkage and the free hydroxy- group at C-6 has been gly cosyla ted with 6-0-acet yl-2,3,4-tri-0-benzyl-a-D-glucopyranosyl bromide.” The disaccharide was detached from the resin by sequential methylation and cleavage of the sulphonium salt with benzyl alcohol to give fully benzylated 6’-O-acetyl isomaltoside.The presence of the 6-0-acetyl residue in the resin-attached dimer offers future scope for selective deblocking and thence oligomer formation. The hydrolysis of 0-glycosides remains their most studied reaction and many investigations have been reported. One of interest2’ monitored the hydrolysis of p-nitrophenyl tetrahydropyran-2-yl ether as a model for glycosides with cyclo- hepta-amylose 2- 3- and 6-monophosphoric acids to determine their effectiveness l6 T. Ogawa and M. Matsui Carbohydrate Res. 1976 51 C13. ” S. Hanessian and J. Banoub Tetrahedron Letters 1976 657. N. N. Kochetkov L. V. Backinowsky and Y. E. Tsvetkov Tetrahedron Letters 1977 3681.l9 S.-H. L. Chiu and L. Anderson Carbohydrate Res. 1976 50 227. 2o B. Siegel A. Painter and R. Breslow J. Amer. Chem. Soc. 1977,99 2309. 348 P. M. Collins as general acid catalysts for the hydrolysis of glycosides. Only the 3-phosphoric acid showed any net catalysis. The 2- and 6-phosphoric acids were not able to overcome the catalytic suppressing effect which binding to the p -cyclodextrin causes. The synthesis of C-glycosides continues to attract much attention because of their potential as precursors for C-nucleosides. Buchanan and co-workers21 report that ethynylmagnesium bromide reacts with 2,3 :5,6-di-O-isopropylidene-D-man-nofuranose (12) to give a 65% yield of 1,2-dideoxy-4,5 :7,8-di-O-isopropylidene-D-glycero-D-tab-oct- 1-ynitol (20); this efficiently ring-closed to the p-D-manno- furanosylethyne derivative (14) upon monotosylation via the 3-sulphonate (21).On the other hand sulphonation of the 3-benzoate (22) followed by methoxide- catalysed debenzoylation caused ring closure in the opposite sense yielding the talosylethyne derivative (23). Derivatives of this type are of value because they can be readily modified by cycloadditions to the acetylene residue. For example two group^^^,^^ have pre- pared the ribofuranosyl-pyrazole derivative (25) by diazomethane addition to the ribosylethyne (24) and another group24 prepared the closely related derivative (27) by addition of the reagent to (26). Me,C/O '04 (20) R'=R~=H (21) R'=TS,R~=H (22) R' = Bz R2 = H C0,Me C /"NH C0,Me R30YY ORZ OR' (24) R' = R2= R3=CH2Ph (25) R' = R2 = R3 = CH2Ph (26) R'R2 =CMe2 R3 = CPh3 (27) R'R2 = CMe2 R3=CPh3 Formylaminomethylenation of lactones provides C-glycosyl precursors.the mannonolactone derivative (28) condenses with ethyl isocyanoacetate in the J. G. Buchanan A. D. Dunn and A. R. Edgar J.C.S. Perkin I 1976 68. 22 C. M. Gupta G. H. Jones and J. G. Moffatt J. Org. Chem. 1976,41 3000. 23 J. G. Buchanan A. R. Edar M. J. Power and G. C. Williams Carbohydrate Res. 1977 55 225. 24 F. G. de Las Heras S. Y.-K. Tam R. S. Klein and J. J. Fox J. Org. Chem. 1976,41 84. 25 R. H. Hall K. Bischofberger S. J. Eitelman and A. Jordaan J.C.S. Perkin I 1977,743. Biological Chemistry-Part (i)Monosaccharides 349 presence of potassium hydride to yield the manno-octenoate (29).This can be hydrogenated to give the amino-acid derivative (30) which is related to the nucleoside amino-acids found in the polyoxin complex of antifungal agents. If in the condensation the potassium hydride is omitted and 1,5-diazabicyclo-[4,3,0]non-5-ene is used in its place then the oxazole (31) is produced.26 Both condensation products are formed from the oxazolide anion (32) which rearranges either by route (a)to give after protonation (29) or by route (b) to give (31). (28) R=O= (30) Et02C \ (29) R= C= / OHCNH C0,Et 3 Unsaturated Sugars 2,3-Unsaturated hexopyranose derivatives are of value as synthetic intermediates and there is active interest in their synthesis.Fraser-Reid's group2' has recom- mended a high-yielding route to ethyl 2,3,6-trideoxy-a-D-erythro-hex-2-enopy-ranoside (34) in which the 6-deoxy position was formed after the introduction of the double bond. Thus ethyl a-D-erythro-hex-2-enopyranoside(33) was deoxy- genated at C-6 via its primary tosylate which was nucleophilically displaced with sodium iodide in ethyl methyl ketone containing pyridine; the product was reduced to (34) with Raney nickel without detriment to the double bond. Pyridine was essential during the iodide displacement since in its absence the product rearranged to 2-(2-iodo-l-hydroxyethyl)furan.The enone (35) could be readily prepared from (34) but when it was the compound of interest it proved more economical to oxidize the allylic hydroxy-group in (33) iodinate the enone so formed directly at C-6 with triphenoxyphosphonium methodide and follow this by selective hydro- genolysis.A new method which augurs well for the preparation of this class of compounds has recently been introduced by Barton's group.28 They showed that vicinal diol " R. H. Hall K. Bischofberger S. J. Eitelman and A. Jordaan J.C.S. Perkin I 1977,2236. 27 M. B. Yunker S. Y.-K.Tam D. R. Hicks and B. Fraser-Reid Cunud.J. Chem. 1976,54,2411. 28 A. G. M. Barrett D. H. R. Barton R. Bielski and S. W. McCombie J.C.S. Chem. Comm. 1977,866. 350 P.M. Collins Me Me (33) (34) (35) bisdithiocarbonates reacted with tributylstannane in refluxing toluene to give olefins.Thus the glycopyranoside derivative (36) gave in 59% yield the hex-2- enoside derivative (37). The corresponding mannoside also gave (37) in 79% yield. The reaction which appears to be of general applicability occurs by the free- radical mechanism outlined in (36). Nucleosides containing a 2-enopyranosyl unit occur in antibiotics such as blasti-cidin S and they also serve as potential synthetic intermediates for amicetin and plicacetin. Consequently the preparation of the 2,3-dideoxyglyc-2- enopyranosylthymine derivative (40)from the 2,3-di-O-mesylthymine derivative (38) is of interest. The best overall yield of 56% was obtained when the 2,2'- anhydride (39) prepared from the dimesylate (38) by the action of sodium benzo- ate in DMF was isolated and heated with sodium iodide in DMF.Treatment of either (39) or (38) under the more usual Tipson-Cohen conditions (e.g. with added zinc) gave (40) in lower yield. 0 ,Me 0 Ph OMS (38) (39) (40) Methods based upon the Diels-Alder addition of n-butyl glyoxylate (Bu02CCHO) to dienes have been reported for the preparation of 2,3-unsaturated uronate derivative^.^"^' If mono-buta-l,3-dienyl ethers derived from sugars are used as dienes in this reaction then disaccharide derivatives containing a uronate are formed. By reducing the ester group and functionalizing the double bond in 29 T. Yamazaki K. Matsuda H. Sugiyama S. Seto and N. Yam'aoka J.C.S. Perkin I 1977,1654. 30 R.R. Schmidt and R. Angerbauer Angew. Chem. Infernat.Edn. 1977,10,783. 31 S.David A.Lubineau and J.-M. Vatkle J.C.S. Perkin I 1976 1831. Biological Chemistry -Part (i) Monosaccharides 351 these adducts David et aL3’ have developed this reaction into a route to unusual disaccharides. Furanoid glycal derivatives are not easily produced by zinc reduction of furanosyl halides. Therefore the reported3’ conversion of 2,3 :5,6-di-O-isopropylidene-a -D-mannofuranosyl bromide into the unsaturated disaccharide derivative (42) in 59% yield by sodium in tetrahydrofuran could prove of value. The anion (41) which would be formed from the mannosyl halide by bromine atom abstraction presumably rearranges with elimination of acetone to an 0-3 alkoxide derivative which attacks another mannosyl bromide molecule. An unusual reaction which leads to 172-unsaturated pyranoid derivatives has been observed33 to occur between phenyl tetra-O-benzoyl- l-thio-P-D-gluco- pyranoside (43) and an excess of N-bromosuccinimide in refluxing carbon tetra- chloride under visible light; it gives phenyl 2,4,6-tri-O-benzoyl-l-thio-D-erythro-hex- l-enopyranosid-3-ulose (47) in 76% yield.The galacto-isomer of (43) gave the threo-epimer of (47) in 83% yield when similarly treated. The reaction is thought to occur uia a halide intermediate (45) which is reasonable since free-radical bromination at C-1 would be facilitated by the enhanced radical-stabilizing character of sulphur. Elimination of hydrogen bromide from (45) to give the lY2-unsaturated derivative (46) followed by its allylic bromination gives after loss of acyl bromide from C-3 the enone (47).The a-anomer of the glucoside (43) also gave enone (47) but a ten-fold increase in reaction time was required. The orientation of the anomeric hydrogen atoms in the thioglucosides is responsible for this rate difference since radical attack on axial anomeric hydrogen atoms is stereoelectronically favoured. The tetra-O-acetylated analogue (44) gave the corresponding enone (48) in lower yield accompanied by the 2-bromoacetyl derivative (49). &Ti””’ {3SPh R2((-ph RO OR OBr OBz 0 OR ’ (47) R’= R2= Bz (43) R = Bz (45) (46) (48)R’ = R2 = Ac (44)R=Ac (49) R’ = COCH’Br R2 = Ac 4 Dicarbonyl Sugars A mild photochemical method for the preparation of keto- and aldehydo-dicar- bony1 sugars from pyruvate esters of secondary or primary hydroxy-groups of ’2 S.3. Eitelman and A. Jordaan J.C.S.Chem. Comm. 1977 552. 33 R.J. Ferrier and R. H. Furneaux J.C.S. Perkin I 1977,1993. 352 P. M. Collins suitably blocked sugars has been For example 1,2 :5,6-di-O-iso-propylidene-a-D-glucofuranose was esterified with pyruvoyl chloride in benzene containing pyridine and the product irradiated in benzene with U.V. light either directly or after isolation which gave the furanos-3-ulose derivative in 71% yield. Several other isopropylidenated sugars have been similarly oxidized as have suitably blocked sugar acetates. Thus 1,3,4,6-tetra-O-acetyl-a-D-glucopyranose 2-pyruvate (50) gave upon irradiation the very unstable fully acetylated pyranos- 2-ulose derivative (51) which readily rearranged with loss of acetic acid to kojic acid diacetate (52).Thus this oxidation method appears to be one of the few mild enough to oxidize hydroxy-groups in partially acylated sugars. The method has also been applied with success to nucleosides the 5’-O-trityl and 5’-O-benzoyl deriva- tives of 3’-O-pyruvoylthymidine being converted in good yields into the cor-responding 3’-ulose derivatives. These ulose derivatives were also unstable since in pyridine or on silica gel the heterocyclic base was eliminated and a 1-ene-3-one formed. The oxidation reaction appears to be an example of the Norish Type I1 reaction as shown in (50). Microbiological oxidation of methyl a-D-xylopyranoside to the pentopyranosid- 4-ulose (53) has been to occur with Acetobacter suboxydans during four days in 76% yield.The p-anomer on the other hand was 28% oxidized during 30 days. Other pentosides were oxidized only to a small extent and hexopyranosides showed little oxidation. 6 Ac@tOAc OAc Ac0 6:) OAc .xb 0 0 0 OAc OH MeNo (5 1) (52) (53) (50) A base-catalysed degradation of the 3-ulose derivative (54) has been developed3’ into a useful synthesis of the rare deoxyketose derivative (56). A two-phase system of diethyl ether and water containing lithium hydroxide was found to bring the conversion about in 55% yield. The transformation is thought to occur by initial opening of the pyranose ring uia (57) rather than by elimination of methanol and hence via methyl 4,6-O-benzylidene-a-~-erythro-hex-l-en-3-ulose.34 R. W. Binkley J. Org. Chem. 1977. 42 1216. ’’ R. W. Binkley D. G. Hehemann and W. W. Binkley Carbohydrate Res. 1977,58 C10. 36 W. A. Szarek and G. W. Schnarr Carbohydrate Res. 1977,55 C5. 37 J.-C. Fischner D. Horton and W. Weckerle Canad. J. Chem. 1977,55,4078. Biological Chemistry-Part (i) Monosaccharides 353 Several groups have investigated the chemistry of a!-keto-oxirans. Methyl 3,4-anhydro-6-deoxy-a -~-lyxo -hexopyranosid-2-ulose was found to give the chloro-enone (58)when treated with lithium chloride in tetrahydr~furan~~ or with hydrochloric acid in acetone.39 Catalytic hydrogenation over palladium on char- coal selectively opened the an hydro-ring yieIding methyl 3,6-dideoxy-a-~-threo -' hexopyranosid-2-ulose.The sequential oxidation-reduction of partially protected sugars has been adop- ted as a standard procedure for epimerization. However a recent report4' indicates that the reduction step is not always straightforward as illustrated by methyl 2-0-acetyl-4,6-0-benzylidene-c~-~-ribo-hexopyranosid-3-ulose (55) which with sodium borodeuteride in moist methanol or propan-2-01 gave the expected alloside derivative (59) labelled with deuterium at C-3 whereas borodeu- teride reduction in dry propan-2-01 gave a 1 1 mixture of (59) and the [2- 2H]glucopyranoside(60). The formation of (60) may involve an unusual nucleo- philic attack upon C-2 of a 2,3-enediol intermediate. (59) (60) Deoxygenation of hexosulose derivatives by the Wolff-Kishner reaction has been examined again.41 1,5-Anhydro-4,6-0-benzylidene-2-deoxy-~-eryth~o-hex-3-ulose was smoothly converted into 1,5-anhydr0-4,6-0-benzylidene-2,3-dideoxy-erythro-hexitol in 69*/? yield.However the reaction was less satisfactory with glycosidulose derivatives since extensive degradation occurred. Deoxygenation of sugar ketones has also been achieved by reduction of their gem -dithio and related derivatives which are formed by the action of phosphorus pentasulphide in pyridine upon protected keto-s~gars.~' Based upon these findings 1,6-anhydro-3,4-0-isopropylidene-~-~-lyxo-hexopyranosulose gave the bridged derivative (61) whereas 1,2 5,6-di-O-isopropylidene-a -D-ribo-hexofuranos-3-ulose gave the disulphide (62). Reductive desulphurization of these products gave respectively the 2-deoxy derivative corresponding to (61) and 3-deoxy-l,2 5,6-di-0-isopropylidene-a!-D-xylo-hexofuranose. An overall inversion of configuration at C-4 occurred in the transformation of the 3-ulose to the 3-deoxy-derivative ..(41) 38 H. Paulsen and K. Eberstein Chem. Ber. 1976,109 3907. 39 G. S. Hajivarnava W. G. Overend and N. R. Williams Carbohydrate Res. 1976 49 93. 40 D. C. Baker J. Defaye A. Gadelle. and D. Horton J. Org. Chem. 1976,41 3834. 41 D. Horton W. Weckerle L. Odier and R. J. Sorenson J.C.S. Perkin I 1977 1564. 42 P.Koll R.-W. Rennecke and K. Heyns Chem. Ber. 1976,109 2537. 354 P. M. Collins presumably because hydrogen added to the double bond in (62) from its least hindered face.5 Esters Acylations which do not occur readily in pyridine have been to proceed' efficiently in tetrahydrofuran containing tetra-n-butylammonium fluoride. Thus when 5'-0-p-methoxytritylthymidinewas treated with pivalic anhydride in pyri- dine the 3'-pivalate was formed in only 9% yield whereas 95% of the ester was produced from a reaction carried out under the recommended conditions. Improved selectivity in tosylations of diols under phase-transfer conditions has been achieved,44 as shown by methyl 4,6-0-benzylidene-a-D-glucopyranoside which gave the 2-tosylate derivative in 78% yield when treated with tosyl chloride in a mixture of dichloromethane and aqueous sodium hydroxide containing tetra- butylammonium hydrogen sulphate.The mannose derivative exhibited even greater selectivity yielding the 2-tosylate in 95% yield. The use of di-n-butyltin oxide first introduced by Moffa~t~~ in nucleoside chem- istry has been used to selectively esterify pyranosides and their derivative^.^^ With this reagent in methanol methyl a-D-glucopyranoside gave a quantitative yield of the 2,3-0-dibutylstannylene derivative (63) which was converted into methyl 2-0-benzoyl-a-D-glucopyranoside in 80-90% yield by subsequent treatment with benzoyl chloride. Similarly (63) and myristoyl chloride or tosyl chloride gave the corresponding 2-esters. In contrast the P-glucoside was esterified only at its primary hydroxy-group. The method also gave 2-esters with the methyl 4,6-0- benzylidene pyranosides of a -D-gluCoSe a-D-galactose and a-D-allose.It failed with the corresponding derivatives of P -D-glucose and a-D-mannose. This has led to the proposal that the success of this reaction depends in part upon a stabilizing interaction between the tin atom and a suitably positioned methoxy-group as shown in (63). In a related process further selective acylations of secondary positions in hexo- pyranosides have been achieved through the regioselective enhanced nucleophili- city of one hydroxy-group by trialkyl~tannylation.~' In this way methyl a-D-glucopyranoside gave after sequential treatment with bis(tributylstanny1) oxide and benzoyl chloride its 2,6-dibenzoate in 82% yield. The selectivity is thought to arise because stannylation occurred preferentially at an equatorial hydroxy-group which has an alkoxy- (or hydroxy-) group cis to it.Stabilization of the type depicted in (64) is thought to accrue from such an arrangement. The methyl pyranosides of a-D-mannose and P-D-galactose give after similar treatment excellent yields of their 3,6-dibenzoates. The 2,3-and 3-4-cis-hydroxy-groups in these pyranosides would be respectively involved in the formation of the stannyl derivative. It is however not yet clear why the equatorially disposed 3-hydroxy-group should be esterified from both arrangements. 43 S. L. Beaucage and K. K. Ogilvie TetrehedronLetters 1977 1691. 44 P. J. Garegg T. Iversen and S. Oscarson Carbohydrate Res. 1977 53 C5. " D. Wagner J. P. H. Verheyden and J. G. Moffatt J. Org.Chem. 1974,39 24. "R. M. Munavu and H. H. Szmant J. Org. Chem. 1976,41 1832. 47 T. Ogawa and M. Matsui Carbohydrate Res. 1977 56 C1. Biological Chemistry -Part (i) Monosaccharides ,OH HO 0 ,OMe \ ,' Bu/\ Bu,Sn Bu (64) The order of reactivity of hydroxy-groups in carbohydrate derivatives towards acylation continues to attract attention. It has been that the hydroxy- groups in methyl p-lactoside react with benzoyl chloride in pyridine in the order 6'>3' > 6> 2 >2',4'> 3. Nucleophilic displacement remains a major interest in carbohydrate chemistry. Accordingly there have been several studies using trifluoromethanesulphonates (triflates) as good leaving group^.^^.^^ It was found for example that the 3-triflate (65) underwent ready nucleophilic displacement with sodium benzenethiolate at 5 "C with inversion of configuration.Reduction of the 3-phenylthioalloside (66)so formed gave the 3-deoxy-sugar derivative (67). C0,CH :Ph (65) R' = OSO;CF, R2 = H (66) R' = H R2 = SPh (67) R'=R~=H A series of ap-unsaturated esters have been prepared51 as temporary blocking groups for the 5'-position in nucleosides which can be removed by treatment with hydrazine under mild conditions. Deblocking occurs via a hydrazino addition product which cyclizes to a pyrazolidin-3-one with release of the deblocked nucleoside. Carbohydrate ester derivatives provide excellent models for examining the importance of stereochemistry in acyloxonium ion chemistry.Paulsen and his co-workers5* have reported examples of five-membered acetoxonium ions fused trans to conformationally locked pyranose rings. For example treatment of 2,3,4- tri-O-acetyl-1,6-anhydro-~-D-idopyranose with trifluoromethanesulphonic acid gave two such ions fused at the 2,3- and 3,4-positions. The same group have showns3 that 1,3-rearrangernent processes of acetoxonium ions compete with the more common 1,2-rearrangements. They found this by *' R. E. Bhatt L. Hough and A. C. Richardson 1.C.S.Perkin I 1977 2001. 49 L. D. Hall and D. C. Miller Carbohydrate Res. 1976,47 299. so T. H. Haskell P. W. K. Woo and D. R. Watson J. Org. Chem. 1977 42 1302. 51 R. Arentzen and C. B. Reese J.C.S. Chem. Comm. 1977 270. '* H. Paulsen H. Hohne and P.L. Durette Chem. Ber. 1976,109 597. 53 H. Paulsen and 0.Brauer Chem. Ber. 1977 110 331. 356 P.M. Collins observing the racemization of 2,3,4- tri- 0-acety1-1,5-an hydro-D-arabini to1 in hydrogen fluoride containing trifluoromethanesulphonic acid. Racemization can only occur uia the two 1,3-rearrangements shown in Scheme 1. 6 Ethers and Acetals New methods are being applied to carry out traditional carbohydrate etherifications. methyl at ion^^^*^^ and benzylationsss of partially protected sugars have been achieved with methyl and benzyl trifluoromethanesulphonate using the sterically hindered weak base 2,6-di-t-butylpyridine. The former reaction was carried out in refluxing dichloromethane whereas the latter was accomplished at -7O"C the benzyl triflate being prepared in situ from benzyl alcohol and the sulphonic acid anhydride.Sugars have been built into macrocyclic ethers as a source of asymmetry. Several chiral asymmetric 18-crown-6 macrocycles incorporating D-glucose and D-galac- tose residues have been prepareds6 by the condensation of disodio derivatives of the methyl 4,6-0-benzylidene-2,3-di-O-(2-hydroxyethyl)-~-~-hexopyranosides with triethylene glycol bis(tosy1ate). The ability of these crowns to form dia- stereoisomeric complexes with primary and secondary alkylammonium cations in solution has been st~died.~' One unexpected observation was that the galacto-crowns which have one face more sterically hindered on account of the cis-ring junction at C-4 and C-5 formed stronger complexes than the corresponding gluco-crowns.The reason for this is that in for example the diastereoisomeric complex (68) the 0-4 can participate with the macrocyclic ether oxygen atoms in hydrogen bonding and/or electrostatic stabilization of the quaternary ammonium hydrogens. (68) '' J. Arnarp L. Kenne B. Lindberg and J. Lonngren Carbohydrate Res. 1975,44 C5. '' J. M. Berry and L. D. Hall Carbohydrate Res. 1976,47 307. s6 D. A. Laidler and J. F. Stoddart Carbohydrate Res. 1977 55 C1. '' D. A. Laidler and J. F. Stoddart J.C.S. Chem. Comm. 1977,481. Biological Chemistry -Part (i) Monosaccharides The role of acetals in carbohydrate chemistry has been changing during the past decade because synthetically useful reactions have been introduced which open cyclic acetals to give partially blocked sugars.Developments in this area have continued. For example Klemer and her co-workersS8 have extended their studies of the reaction of butyl-lithium on 0-benzylidene-sugar derivatives to various mono- and di- 0-isopropylidene derivatives. Their general findings are illustrated by the conversion of 1,2:3,4-di-O-isopropylidene-6-O-methyl-a -D-galacto-pyranose into the (E)-and (2)-isomers of the unsaturated polyol(72). Abstraction of the anomeric proton is thought to initiate this transformation and the anion (69) so formed subsequently rearranges with loss of acetone to (70) which then eli- minates another acetone molecule to give after protonation at 0-4 the unsaturated lactone (71). Further attack on (71) gives the isomeric products (72) in low yield.rOMe rOMe rOMe HO{ OH CH20Me (72) The synthetic value of this reaction has been studieds9 with methyl 2,3-0-benzylidene-a -L-rhamnopyranoside (73) and its 4-0-methyl derivative (74). Butyl-lithium smoothly converted (74) at -30 "C into the 3-ulose derivative (75) whereas the unmethylated compound (73) reacted only at 0 "C to give a mixture of elimination products. The vigorous reaction conditions needed to abstract a carbon-bonded proton from the 0-4 oxyanion of (73) were thought to be respon- sible for the difference in this case. Me 0 Ox0H (75) Ph (73) R=H (74) R=Me 58 G. Rodemeyer and A. Klemer Chern. Ber. 1976,109,1708. 59 D.M. Clode D. Horton. and W.Weckerle Carbohydrate Res. 1976,49,305. 358 P M. Collins Liptak's studies with the lithium aluminium hydride-aluminium trichloride reducing agent have revealed that the regiospecificity observed in the hydro- genolytic ring cleavage of 2,3-60 and 3,4-0-benzylidene6' derivatives is dependent upon the stereochemistry at the acetal centre. Accordingly the exo-isomer of (73) afforded methyl 3-0-benzyl-a -L-rhamnopyranoside whereas the endo-isomer of (73) gave the 2-0-benzyl derivative. The high degree of stereoselectivity does not appear to arise from steric interaction with the phenyl group in the endo-isomer of (73) since the spin-lattice relaxation times for all carbon atoms in endo- and exo- (73) were found6* to be similar which indicates that rotation of the phenyl groups occurs to the same extent in both isomers.By use of this reagent 0-benzylidene residues have been employed as temporary blocking groups in oligosaccharide ~yntheses.~~ Cleavage of the 1,3-dioxan ring in 4,6-O-benzylidene derivatives with this reagent is less predictable.60 This was also unfortunately found to be the case vith photogenerated ketyl radicals.64 Thus U.V. irradiation of acetophenone for exam- ple in an oxygenated solution of 1,2,3-tri-O-acetyl-4,6-0-benzylidene-@-~-glucopyranose gave a mixture of the 4-0-benzoyl and 6-0-benzoyl derivatives of 1,2,3-tri-O-acetyl-P-D-glucopyranose, plus an adduct formed by ketyl radical addition to the benzylidene acetal carbon atom 7 Halo-and Thio-sugars and other Inorganic Sugar Derivatives Two new to the synthesis of halo-sugars have been introduced which are particularly satisfactory for primary iodo- and chloro-derivatives.The precursors for the iodo-compounds are the readily available cyclic thiocarbonates which have been found to undergo opening of the thiocarbonate ring when treated with methyl Hence the 4,6-thiocarbonate (76) when treated with methyl iodide gave a high yield of 6-deoxy-6-iodo-4-0-methylthiocarbonyl glucoside (78).66 The ring opening occurred via the intermediate (77) which underwent regiospecific attack by iodide at the least hindered primary position. As would be expected from this mechanism methyl 4,6-O-benzylidene-a-~-gluco-pyranoside 2,3-thiocarbonate opened non-regio~pecifically~~ to give a mixture of the 2-and 3-iodo-glucoside derivatives with presumed rnanno- and allo-configura- tions respectively.6o A. Liptik P. Fugedi and P. Ninasi Carbohydrate Res.. 1976 51,C19. 61 A. Liptlk Tetrahedron Letters 1976 355 1. 62 A. NeszmClyi A. Liptlk and P. Nlnisi Carbohydrate Res. 1977,58 C7. 63 A Liptik and P.Nlnisi Tetrahedron Letters 1977 921. 64 M. Suzuki T. Inai and R. Matsushima Bull. Chem. SOC.Japan 1976,49 1585. 65 D. H. R. Barton and R. Subramanian J.C.S. Perkin I 1977 1718. 66 D. H. R. Barton and R. V. Stick J.C.S.Perkin I 1975 1773. 359 Biological Chemistry -Part (i) Monosaccharides The chloro-sugars were prepared6' by treating partially protected sugar deriva- tives possessing an unblocked I ,2- or 1,3-diol structural unit with NN-dimethyl- + a-chlorobenzylideneammonium chloride [Ph(Cl)C=NMe2 C1-1.In this way the pair of hydroxy-groups at C-5 and C-6 in 1,2-0-isopropylidene-3-0-methyl-a-D-glucofuranose react with this reagent to give in almost quantita- tive yield 5-0-benzoyl-6-chloro-6-deoxy-l,2-0-isopropylidene-3-O-methyl-a-D-glucofuranose. The reaction occurs via the cyclic intermediate (79) with ring opening taking place by chloride ion attack at C-6. Chloro-derivatives were not obtained from methyl 4,6-0-benzylidene-a -D-glucopyranoside and this reagent. 6-Bromo-6-deoxy-sugars are often prepared by opening dioxan rings in 4,6-0- benzylidene-pyranosides with N-bromosuccinimide but there are fewer cases of this reagent being used to cleave dioxolan rings in 0-benzylidene-sugar derivatives.However two recent report^^**^^ showed the reaction to be regiospecific since methyl 2,3-0-benzylidene-4-O-methyl-a -L-rhamnoside (74) gave68 a 3-bromo-3- deoxy 2-benzoate and methyl 3,4-0-benzylidene-P-D-fucopyranosidegave69 a 3-bromo-3-deoxy 4-benzoate. The biological properties of fluorinated saccharides have led to continued inter- est in this class of compounds and new fluorinating reagents are frequently applied to sugar derivatives. One such compound diethylaminosulphur trifluoride con- verted the primary hydroxy-group in 1,2,3,4-tetra-O-acetyl-P-D-glucopyranose into a 6-deoxy-6-fluoro-group in high yield." Free keto- or aldehydo-carbonyl groups in otherwise fully blocked sugars are converted" into gem-difluorides by this reagent as shown by methyl 2,3-0-isopropylidene-~-D-eryfhro-pentopy-ranosid-4-ulose which gave the 4-deoxy-4,4-difluoro-derivative(80)'l when so treated.Ferrier and Furneaux7* have reported radical bromination at C-5 in some methyl hexopyranuronates which yield glycopyranosyl bromides of unusual structure. Thus methyl glycopyranuronate (81) and its 2,6-anhydro analogue (83) gave upon treatment with N-bromosuccinimide in refluxing carbon tetrachloride under bright light the monobromo-derivatives (82) and (84) respectively. Acetolysis of (84) followed by deacylation and subsequent treatment with hydrogen chloride gave L-ascorbic acid. C0,Me F FGMe0' AcO OAc OAc (81) R'=OAC,R2 = H (82) R' =OAc R2 = Br (83) R' =R2 = H (84)R' =H R2 = Br 67 T.G. Black D. H. R. Barton andB. L. Rao J.C.S. PerkinI 1977 1715. 68 C. Monneret J.-C. Florent N. Gladieux and Q. Khuong-Huu Carbohydrate Res. 1976 50 35. 69 K. Eklind P. J. Garegg and B. Gotthammer Actu Chem. Scund. 1975 B29,633. 70 M. Sharma and W. Korytnyk Tetrahedron Letters 1977 573. " R. A. Sharma I. Kavai Y. L. Fu and M. Bobek Tetrahedron Letters 1977 3433. 72 R. J. Ferrier and R. H. Furneaux J.C.S. Perkin I 1977 1996. 360 P. M. Collins N.m.r. of relevance to glycopyranosyl chlorides have been carried out on chloromethyl methyl ether. At -182°C the methylene protons give rise to a distinct symmetrical doublet consistent with a gauche conformation about the carbon-xygen bond. Line-shape calculations on the coalescing methylene proton resonance at -180 “C suggested that two chiral gauche forms were exchanging with a rate constant of 250 s-l from which a value of 2-3 kcal mol-’ for the anomeric effect in a-chloro-ethers was calculated.1-Thioglycosides are usually produced by the action of thiols on 1,2-trans-glycosyl acetates but dithioacetals formed by opening of the glycosidic ring often contaminate the products. It has been that formation of dithioacetal derivatives can be suppressed by the use of tributylstannyl sulphide (TBSS). Based upon this finding 2,3,4,6-tetra-O-acetyl- 1-thio-P-D-glycopyranosidewas formed in good yield when the corresponding a -D-glucosyl 1-bromide was treated with this reagent. Unfortunately only 45% of the bromo-derivative reacted under the usual reaction conditions because the nucleophilicity of the sulphur in TBSS is attenuated.However upon addition of stannic chloride to the reaction a quan- titative yield of an anomeric mixture of the thioglucoside derivatives was produced. The isomers were formed by the stannic-chloride-catalysed anomerization of the &isomer. The propensity of sulphur to participate in displacement reactions and in ring closures is well known and interest in both continues. For example 5-thio-D- ribose and 5-thio-D-xylose both gave upon acid-catalysed acetonation 1,2 :3,4-di-0-isopropylidene-pyran~ses.~’ Furanose products (oxygen in the ring) were not detected thus demonstrating the strong preference for those diacetals with a six-membered sulphur-containing ring despite the strain engendered by a cyclic acetal derived from a vicinal trans-diol in the xylopyranose compound and by a cis-syn-cis-arrangement of cyclic acetal rings in the ribopyranose derivative.Neighbouring group participation by a thioether residue has been ~tilized’~ in the key step in the conversion of tubercidin (85) into 2’-deoxytubercidin (87). Thus when the 3’-S-benzyl 2’-mesylate (88) which was obtained from 2‘,3’-anhy-drotubercidin was heated with sodium benzoate in DMF the thiobenzyl group migrated to the 2’-position to give after deblocking (86). Reductive desul- phurization of this product gave (87). NH, I HNBz I OH R‘ OMS (85) R’ =OH R2= H (86) R’= H R2= SCHzPh (88) (87)R’=R~=H 73 F.A. L. Anet and I. Yauari J. Amer. Chem. SOC.1977,99,6752. 74 T. Ogawa and M..Matsui Carbohydrate Res. 1977 54 C17. 75 N. A. Hughes and C. J. Wood Carbohydrate Res. 1976,49 225. 76 M. J. Robins and W. H. Muhs J.C.S. Chem. Comm. 1976 269. Biological Chemistry -Part (i) Monosaccharides Reports of sugar rings containing phosphorus have appeared. Inokawa's group" have for example synthesized 5-ethylphosphinyl-5-deoxy-D-xylopyranose (89) the phosphorus being introduced by displacement at C-5 in 3-0-acetyl-S-deoxy-S- iodo- 1,2- 0-isopropylidene-a-D-xy lofuranose with 00-trie thylphosphorite. Compound (go) containing both phosphorus and oxygen in the six-membered ring was obtained when 2,3 :5,6-di-O-isopropylidene-D-mannofuranose was sub- jected to the Abramov reaction with dimethyl ph~sphite.~' (89) (90) Application of tin-containing reagents to carbohydrates has now become quite common so it was inevitable that stannane carbohydrate derivatives would even- tually appear.This new class of sugar derivative containing a carbon-tin bond has been prepared79 by triphenyltin lithium displacements in sugar tosylates or opening of oxiran rings in sugar epoxides. For example with methyl 2,3-anhydro-4,6-0- benzylidene-a -D-allopyranoside this reagent gave (methyl 4,6-0 -benzylidene-a -D-altropyranosid-2-y1)triphenylstannanein 75% yield. 8 Deoxy-sugars and Branched-chain Sugars Several new methods for preparing deoxy-sugars have recently been reported. Primary and secondary iodo-sugar derivatives have been in high yields by the Cr*+-thiol procedure.For example the 6-deoxy-6-iodo-glucoside derivative (78) was reduced with chromium(I1) acetate in DMF containing butane- l-thiol to give after de-esterification methyl 6-deoxy-2,3-di-O-methyl-a-D-glucopyranoside in excellent yield. A mild method for deoxygenation of secondary alcohols has been successfully applied to sugars.*' In this method NN-dimethyl-a -chlorobenzylideneammonium chloride reacts with one free hydroxy-group in an otherwise fully blocked sugar derivative to give an intermediate which is converted into a thionobenzoate by hydrogen sulphide. Reduction of the thiono-ester with tributylstannane gives the corresponding deoxy-sugar in high yield. In addition to preparing iodo-sugars from cyclic thiocarbonates Barton and co-w~rkers~~ have reductively ring-cleaved them.Thus the 5,6-thiocarbonate (91) 07 ,I 0-C (91) Me* 77 F. Seo and S. Inokawa Bull. Chem. SOC.Japan 1975,48,1237. ''J. Thiem M. Gunther H. Paulsen and J. Kopf Chem. Ber. 1977 110,3190. 79 L. D. Hall D. C. Miller and P. R. Steiner Carbohydrate Res. 1976 52 C1. D. H.R. Barton and S. W. McCornbie. J.C.S. Perkin I 1975 1574. 362 P. M. Collins upon refluxing with tributylstannane in toluene containing aa'-azoisobutyronitrile gave after alkaline hydrolysis the corresponding derivative of the 5-deoxy-hexose in 67% yield as the only deoxy-sugar product. The regiospecificity arises because the radical-initiated ring cleavage leads to the secondary radical at C-5,as shown in (91).The 4,tj-thiocarbonate (76) similarly gave the corresponding derivative of the 4-deoxy-hexose. Specificity was absent in the reaction of methyl 4,6-0-ben- zylidene-a-D-glucopyranoside 2,3-thiocarbonate with the tin hydride since this gave 3-deoxy- and 2-deoxy-glycoside derivatives in 60 and 30% yields respec- tively. Common primary and secondary esters of sugars have been photochemically reduced in aqueous hexamethylphosphortriamide(HMPT). Accordingly 1,2 :5,6-di-0-isopropylidene-a-D-glucofuranose3-a~etate~~*~* or 3-~ivalate,'~ upon U.V. irradiation in aqueous HMPT gave the corresponding 3-deoxy-hexose derivative in 70% yield. Di- and tri-deoxy-sugars were obtained from the corresponding di- and tri-esters in one step and in this way the readily available 2,3,6-tri-0-acetyl-a- D-glucopyranoside was converted into methyl LY -Damicetoside in reasonable yield.82 The branched-chain sugar L-dendroketose (96) has been synthesized using the well-established aldol-condensation procedure to generate the branch points3 from the u2dehydo-arabinose derivative (92) and formaldehyde.The branched alditol derivative (93) so formed was deblocked and oxidized by Acetobacter suboxyduns at the C-2 position as expected from the Bertrand-Hudson rule to give (96). Acetonation of (96) gave 1,2:3,4-and 2,3 :4,4'-di-O-isopropylidene-~-dendroketose derivatives which were used in 13C n.m.r. spectroscopic studies to assign the configuration at C-4.'" CH,OH M I A X A-R20 c=o LMC2 Me .-"$ '0 CH,0R3 Ez+CH20H CHO CH20R3 CH,OH (93) R'R2 = CMe2 R3= H (96) (94) R' = R3= Bz,R2 = H (95) R'= CPh3 R2= H R3= CHzPh In an earlier attempts5 to prepare dendroketose compound (94) was oxidized with a chromium trioxide-pyridine complex to give the fully protected branched- chain sugar.However base-catalysed debenzoylation of this material was accom- panied by isomerization at C-3. This problem has been overcomes6 by use of compound (95) which upon oxidation yielded a dendroketose derivative from which the blocking groups could be removed without recourse to base. J.-P. Pete C. Portelk C. Monneret J.-C. Florent and Q. Khuong-Huu Synthesis 1977 774. P. M. Collins and V. R. N. Munasinghe J.C.S. Chem. Comm. 1977 927. 83 W. A. Szarek G.W. Schnarr H. C. Jarrell and K. J. N.Jones Carbohydrate Res. 1977 53 101. 84 D. M. Vyas H. C. Jarrell and W. A. Szarek Cunad. J. Chem. 1975 53 2748. " H. C. Jarrell W. A. Szarek J. K. N. Jones A. Dmytraczenko and E. B. Rathbone Carbohydrate Res. 1975,45 151. E. B. Rathbone and G. R. Woolard Carbohydrate Res. 1976,46 183. 363 Biological Chemistry -Part (i) Monosaccharides A new route to branched-chain sugar derivatives has been reported87 which involves the condensation shown in (97) between 2,3-O-isopropylidene-D-gly-ceraldehyde and 4,5-dimethyldioxaphospholento give the corresponding dioxo- phospholan. The addition is remarkable because the two new chiral centres are generated stereospecifically. Hydrolysis of the dioxophospholan followed by Fischer glycosidation gave a methyl pyranoside and furanoside of 1-deoxy-3-C-methyl+ -D-ribo -hexulose in 20% and 30% yields respectively.(97) Anions derived from 1,3-dithians are frequently added to ulose derivatives as the key step in syntheses of branched-chain sugars The monobenzoyl derivative of pillarose (99) has been prepared in this way.nn Thus the dianion of 1,3-dithian-2-methanol reacted with methyl 2,3,6-trideoxy-a -~-glycero-hexopyranosid-4-ulose to give the addition product (98) which was monobenzoylated at the primary hydroxy-group and deblocked at the carbonyl group to give the pillarose derivative (99) in overall 10'/0 yield. Although additions to ulose derivatives remain one of the most common routes to branched-chain sugars 1,4-additions to carbohydrate enones are beginning to attract attention.One such example is the addition of lithium dimethylcuprate to the hex-2-enopyranosid-4-ulose(100) to give (in 84% yield) the 2-C-methyl derivative (102).89 The addition occurred to the least hindered side of the pyranose ring trans to the aglycone. Lithium dimethylcuprate vinylmagnesium bromide and 2-lithio-2-ethoxycar- bonyl- 1,3-dithiolan have been similarly added" to methyl hex-2-enopyranosid-4- ulose (101) to give the 2-C-branched derivatives shown in (104). <>>Rt o{E$ 0 R4 R2 OEt (100) R'=R4=H,R2=OEt (102) R=Me (98) R= R3= CH20CPh3 (103) R = CHzOH (101) R' = OMe R2 = R3 = H, R4= Me (99) R = BzOCH~CO 87 Ltpine G. Aranda and G. Vass J.C.S. Chem. Comm. 1976 747.S. David M.-C. 88 H. Paulsen K. Roden V. Sinnewell and W. Koebernick Chem. Ber. 1977,110,2146. 89 M. B. Yunker D. E. Plaumann and B. Fraser-Reid Canad. J. Chem. 1977,5S 4002. H. padsen W. Koebernick and H. Koebernick Tetrahedron Letters 1976 2297. 364 P. M. Collins R (104) R = Me or CH2=CH Photochemically induced conjugate additions of simple a~etals,~~ and aldehydes to this class of carbohydrate derivative have been achieved. Thus U.V. irradiation of benzophenone in methanol containing the enone (100) gave (103) presumably by the 1,4-addition of photogenerated hydroxymethyl radicals. The addition was stereospecific and the configuration at C-2 in (103) was verified by ready formation of a 2,3-cyclopropane derivative with the D-lyxo-structure.Photoexcited 1,2 :4,6-diacetalated hexopyranosid-3-uloses have been found93 to react with methanol to give isomeric mixtures of the corresponding derivatives of 3-C-hydroxymethyl branched-chain sugars the configurations of which were determined by 13Cn.m.r. spectroscopy. A photochemical synthesis of DL-apiose (108) has been achieved94 by appli- cation of the Patterno-Buchii reaction to 1,3-dihydroxy-2-propanonediacetate (105) and 1,3-dioxolen-2-one (106) which gave the oxetan (107) and upon subsequent a1 kaline hydrolysis DL-apiose. OAc CHO 9 Amino-sugars The occurrence of amino-sugars that possess unusual structures in antibiotics has been responsible for the continuing interest in the synthesis of this class of compounds.Two synthetic to the aminodeoxy-octose lincosamine which is a constituent of lincomycin have been reported. One highly stereoselec- tive route9' commences with the unsaturated nitro-sugar (109) epoxidation of which gave the key intermediate (110). This was opened with benzylamine to give (1 11) and subsequently reduced and converted into the N-acetyl derivative of lincosamine (112). The other route96 to (112) which uses the diazoketone deriva- tive (1 13) as the starting material contained steps that were not stereospecific. 91 B. Fraser-Reid N. L. Holder D. R. Hicks and D. L. Walker Canad. J. Chem. 1977,55 3978. 92 B. Fraser-Reid R.C. Anderson D. R. Hicks and D. L. Walker Canad. J. Chem. 1977 55 3986. 93 P.M. Collins V. R. N. Munasinghe and N. N.Operaeche J.C.S. Perkin I 1977 2423. 94 Y. Araki J. Nagasawa and Y. Ishido Carbohydrate Res. 1977,58 C4. 9s G. R. Woolard E. B. Rathbone W. A. Szarek and J. K. N. Jones J.C.S. Perkin I 1976,950. 96 S. M. David and J.-C. Fischer Carbohydrate Res. 1976 50 239. Biological Chemistry-Part (i) Monosaccharides PhCHZHN A reaction which could be of value for selectively reducing an azido-group in an unsaturated nucleoside has been rep~rted.~' The unsaturated azide is treated with hydrogen sulphide in aqueous pyridine at room temperature for 3 h. The sulphur is filtered off and the unsaturated amino-nucleoside isolated. A method of synthesis of amino-sugar derivatives which has attracted the atten- tion of three is the addition of the Sharpless reagent'" to unsaturated sugars.Thus chloramine T (TsNClNa) in the presence of osmium tetroxide was added99 to the diacetylated derivative of the pseudo-glucal(33) to give in moderate yield a mixture of the 2-and 3-tosylamino-mannosides (1 14) and (1 15). A higher yield of the products (116) and (117) was from the corresponding methyl hexenopyranoside after acetylation of the hydroxy-groups in the initially formed adducts. Attempts have been made to improve the potency and antibacterial spectra particularly against resistant organisms of aminoglycoside antibiotics by chemically modifying their structures. For example gentamicin X2(1 18) which is produced only as a minor component by Micromonospora purpurea has been prepared"' semi-synthetically. The key step in the synthesis is the a-glycosylation of the 1,3,3'-tris-N-benzoyloxycarbonylderivatives of the semi-synthetic pseudodisac- charide garamine (1 19)'** by the Lemieux-Nagabhushan reaction using the nitrosyl chloride adduct of triacetyl glucal.The total syntheses of streptomycin and its 3"-deoxydihydro analogue have been reported by Umezawa's gro~p.'~~.'~~ Labelling studies that have revealed the biosynthetic pathways by which D-glucose is converted into the amino-cyclitol components of streptomycin and spec- tinomycin have been reviewed.'05 97 I. Adachi Y. Yamada and I. Inoue Synthesis 1977,45. 98 K. Heyns and J. Feldmann Tetrahedron Letters 1977 2789. 99 I. Dyong Q. Lam-Chi G. Schulte B. Fraser-Reid and J. L. Primeau Angew. Chem.Internat. Edn. 1977,16 553. 100 K. B. Sharpless A. 0.Chong and K. Oshima J. Org. Chem. 1976,41,177. 101 M. Kugelman A. K. Mallams H. F. Vernay D. F. Crowe G. Detre M. Tanabe and D. M. Yasuda J.C.S. Perkin I 1976 1097. 102 M. Kugelman A. K. Mallmas H. F. Vernay D. F. Crowe and M. Tanabe J.C.S.Perkin I 1976 1088. 103 S. Umezawa Y. Takahashi T. Usui and T. Tsuchiya J. Antibiotics 1974,27,997. 104 H. Sano T. Tsuchiya S. Kobayashi M. Hamada S. Umezawa and H. Umezawa J. Antibiotics 1976 29 978. 10s K. L. Rinehart jun. and R. M. Stroshane J. Antibiotics 1976 319. 366 P.M. Collins AcO [:iR, (1 14) R' = Et R2 = NHTs R3 = OH (115) R'=Et R2=OH R3=NHTs (116) R' =Me R2 =NHTs R3 = OH (117) R' = Me R2 =OH R3 =NHTs (118) R= HO-p-+ Meoq>R CH,OH HO NH R2 (119) R=H (120) R'=H R2=N02 (121) R' =Me R2= AcNH The branched-chain nitro-sugar evernitrose has been shown to possess the L-ambino-configuration (1 20) from an X-ray crystallographic analysis of the 3-acetamidopyranoside (121) which was prepared from it in three steps.lo6 Thus the complete structure of the octasaccharide antibiotic everninomicin D with its uncommon orthoester linkages is now known.'" 10 Synthesis of Non-carbohydrate Compounds There is increasing use of carbohydrate derivatives as precursors for the synthesis of optically active non-carbohydrate compounds.Some examples of this approach that have been applied to the synthesis of natural products this year include the preparation of the antifungal metabolite (-)-isoavenaciolide from hiacetone-D-glucose'os and the conversion of 1,6-anhydro-P-D-glucose into the growth factor (+)-biotin'" in a biomimetic synthesis which involved a stereospecific ring closure to the tetrahydrothiophen.Prostaglandin and related substances have also been prepared from carbohydrate sources. Thus prostaglandin PGEl has been prepared from D-glyceraldehyde"' and thromboxane B2 has been synthesized by two from methyl a-D-glucopyranoside. A. K. Ganguly 0.Z. Sarre A. T. McPhail and K. D. Onan J.C.S. Chem. Comm. 1977,313. lo' A. K. Ganguly 0.Z. Sarre D. Greeves and J. Morton J. Amer. Chem. SOC.,1975 97,1982. lo* R. C. Anderson and B. Fraser-Reid Tetrahedron Letters 1977 2865. lo9 T. Ogawa T. kawano and M. Matsui Carbohydrate Pes.1977,57 C31. 'lo G. Stork and T. Takahashi J. Amer. Chem. Soc. 1977,99 1275. S. Hanessian and P. Lavallee Canad. J. Chem. 1977.55 562. E. J. Corey M. Shibasaki and J. Knolle Tetrahedron Letters 1977 1625.