13 Carbohydrates By J. THIEM 0rganisch-Chem isch es Ins titu t Un iversita t Muns ter Orleans -Ring 23 D-4400,Munster Federal Republic of Germany 1 Introduction A previous review in this series discussed the monosaccharide literature of 1976/7 and the present report is intended to focus on the processes currently enjoying interest among carbohydrate chemists with emphasis on the literature of 1983/4. For a comprehensive overview of the intermediate period the reader is referred to more detailed reviews e.g. Specialist Periodical Reports. Carbohydrate chemistry today may be viewed from a variety of standpoints depending on the focus of interest. One very old area fermentation is highlighted today by the biotechnology rush and is expected to become even more important in the near future.Another more technologically oriented branch concerns the transformation of mono- oligo- or polymeric carbohydrate materials into tailor- made polymers using renewable resources. A third main area might be called classical carbohydrate chemistry in which glycosylation modification and esoteric or useful construction of distinct carbohydrates from readily available sugars are dealt with. Finally a vastly expanding fourth branch of carbohydrate chemistry deals with on the one hand de novo syntheses of carbohydrates from non-carbohydrate precursors and on the other hand the transfer of carbohydrates or selectively prepared deriva- tives thereof into other chiral materials -the carbohydrate chiral template approach. These last mentioned areas are both inspired by partial and complete syntheses of complex structures relevant to biochemical processes.This review will concentrate on the last two areas with some emphasis on classical aspects. 2 Monosaccharides G1ycosides.-The stereoselective formation of glycosides continues to be one of the central goals of carbohydrate chemistry. Even though the original general approach by Koenigs and Knorr in 1901 has been improved considerably there remain a number of problems and challenges associated with the glycosylation procedure.' Most of the present glycoside syntheses aim at the construction of biologically important compounds and these frequently demand the preparation of di- tri- and higher oligomers. Consequently in this subchapter only a short survey of the development of glycosylation agents procedures and the formation of simple ' H.Paulsen Angew.Chem. 1982,94 184. 311 3 12 J. Thiern glycosides is discussed. Oligosaccharide synthesis will be treated in more detail below. By use of trimethylsilyl trifluoromethanesulphonate (TMSOTf) the pyranose (1) could be condensed (+)-4-demethoxy-anthracyclinone (2) to give the pure a-glycosylated anthracycline (3) in 99% yield.2 Schmidt and collaborators have pro- vided evidence for the detailed formation of trichloroacetimidates. The primary kinetic product seems to be the P-trichloroacetimidate which anomerizes slowly to its thermodynamically favoured ~y-anomer.~.~ Glycosylation of 2-azido-galactosyl trichloroacetimidates promoted by TMSOTf led exclusively to a-glycosides.A similar treatment of the glum-isomer with both TMSOTf or boron trifluoride etherate however gave alp mixtures.’ pNBzO NHCOCF Acyloxy leaving groups previously used for glycosylations in the presence of Lewis acids could be used successfully for stereoselective glycoside formation. Thus treatment of the p-bromoacetyl gluco-derivative (4) even with sterically demanding alcohols in the presence of trityl perchlorate gave clean inversions and almost exclusively a-glucoside formation (5 ; 75-90% yield). Similarly the P-ribose ,OBn furanose acetate (6) promoted by trityl perchlorate gave only the P-riboside (7) whereas in the presence of lithium perchlorate and molecular sieves a predominant a-riboside synthesis (8; a /3 = 70 30 75-90°h yield) was accomplished.6 In another interesting approach Mukaiyama and co-workers7 demonstrated a stereoselective P-xylofuranoside (10) preparation by reaction of the xylofuranose (9) with the aglycone and caesium fluoride in the presence of methyl fluorosulphonate (65-90% yield almost pure P).The attractive class of 2-0x0-glycosyl bromides [ e.g. (12)] was advantageously prepared by N-bromosuccinimide treatment of the hydroxyglycal esters [ e.g. (1 1)18 Y. Kimura M. Suzuki T. Matsumoto R. Abe and S. Terashima Chem. Lett. 1984 501. R. R. Schmidt and J. Michel Tetrahedron Lett. 1984 25 821. R. R. Schmidt J. Michel and M. Roos Liebigs Ann. Chem 1984 1343. G. Grundler and R. R. Schmidt Liebigs Ann.Chem. 1984 1826. T. Mukaiyama S. Kobayashi and S. Shoda Chem. Lett. 1984 907. M. Murakami and T. Mukaiyama Chem. Lett. 1983 1733. F. W. Lichtenthaler E. Cuny and S. Wepretz Angew. Chem. 1983 95 906. Carbohydrates AMe Bnovo 313 {/>OR ROH ROH {?OR TrCI0,-TrCIO OBn LICIO BnO OBn 0Bn (8) (6) (7) MeOSO,F CsF ROH BnomR OH I or by photochemically induced bromination of 1,s-anhydro-D-keto-hexose deriva-tives [e.g. (13)19in excellent yields. The latter process could also be applied to the oximino compounds [e.g. (14)]and gave an access to the oximino bromides [e.g. (191,derivatives which in turn represent very attractive saccharide units for the sequential construction of aminosugar-containing oligosaccharides. BzO BzO BzO 0Bz Br (1 1) (12) x=o (13) (15) X = NOH (14) Following early work on glycosyl fluorides by Micheel and collaborators," some time ago Mukaiyama et aZ.,ll then Noyori et all2 and a number of other groups have become interested in their preparation properties and glycosylation reactions.Although some of the material is still in the process of being patented a few papers have appeared already. Their increased stabilities render the fluorides so much different from the previously used glycosyl halides that a totally different approach that is by use of Lewis acid catalysis had to be developed for glycosylations. Nicolaou et aL13 prepared glycosyl fluorides from selectively blocked thiophenyl glycosides and claimed a 'practical' synthesis of oligosaccharides (see below).The same group could apply known glycosyl fluorides in the synthesis of 0-,N- S- and C-glycosides (see be lo^).'^.'^ In the very elegant total synthesis of the complex F. W. Lichtenthaler P. Jarglis and W. Hempe Liebigs Ann. Chem. 1983 1959. 10 F. Micheel and A. Werner Ado. Carbohydr. Chem. 1961 16 85. T. Mukaiyama Y. Murai and S. Shoda Chem. Left. 1981 431. 12 S. Hashimoto M. Hayashi and R. Noyori Tetrahedron Lett. 1984 1379. 13 K. C. Nicolaou R. E. Dolle and D. P. Papahatjis J. Am. Chem. SOC.,1984 106 4189. 14 K. C. Nicolaou R. E. Dolle A. Chucholowski and J. L. Randall J. Chem. SOC,Chem Commun. 1984 1153. Is K. C. Nicolaou A. Chucholowski R. E. Dolle and J. L. Randall J. Chem. SOC.,Chem. Commun. 1984 1155.314 J. Thiem aminoglycoside apramycin Tatsuta et all6 condensed the glycosyl fluoride of the terminal sugar unit to the neaminyl octodiose glycoside. A classical approach gave access to the a-and &fluorides of N-acetylneuraminic acid by reaction of the P-chloride with silver fluoride and pyridinium fluoride respectively." By use of the previously developed18 and often advantageously applied" N-iodosuccinimide procedure a successful anthracycline synthesis was nicely accom- plished.20 Treatment of diacetyl-L-rhamnal (16) with the racemic demethoxyadriamycinone (17) gave two easily separable 2-iodoglycosides because of a concommitant resolution of the diastereomers. Only the one with the natural (7S 9s) configuration (18) after deprotection led to a compound which showed activity in a leukaemia assay.0 -HO 0 OSiMe .Bu' OH 0 -HO OH = rac.Q-H AcO AcO NIS I (16) (18) Deoxy Branched and Higher Sugars.-Novel procedures for introduction of a deoxy function (or removal of a hydroxy function) continue to command interest. A reductive displacement of triflates using sodium borohydride was shown to be more effective than Barton-type reactions with primary alcoholic functions ; secondary groups however gave poor results because of concommitant eliminations.21 The stereoselective formation of 2-deoxy-a-~-glycopyranosidesmay be con-sidered a resolved problem2* because efficient methods are available like the N-iodosuccinimide glycosylation18 or the alkoxy ~elenation~~ followed by a reductive process.This approach was displayed in the use of the NIS method for assembling the sequence of the lanatoside oligodeoxy tetrasa~charide.'~ A series of attempts to apply modified glycosylations to 2-deoxy halides did not always meet with convincing results. Glycosylation of the labile 2,6-dideoxy glycosyl bromides most of which are easily accessible in situ by treatment of the corresponding acyl derivatives with trimethylsilyl bromide,24 was studied in the D-ribo case and with complex and simple aglycons led to only little ~tereoselection.~~ A first report on the use of a 2,6-dideoxy glycosyl fluoride for construction of the a,l + 4-linked disaccharide fragment of the antiparasitic agent avermectin B, and the natural compound itself a~peared,'~ 16 K.Tatsuta K. Akimoto H. Takahashi T. Hamatsu M. Annaka and M. Kinoshita Buff. Chem. SOC. Jpn. 1984 57 529. 17 M. N. Sharma and R. Eby Carbohydr. Res. 1984 127 201. J. Thiem H. Karl and J. Schwentner Synthesis 1978 693. l9 J. Thiem and P. Ossowski Liebigs Ann. Chem. 1983 2215. 2o D. Horton W. Priebe and 0.Varela Carbohydr. Res. 1984 130 C1. 21 E.-P. Barette and L. Goodman J. Org. Chem. 1984 49 176. 22 J. Thiem Nachr. Chem. Tech. Lab. 1984 32. 6. 23 G. Jaurand J.-M. Beau and P. Sinay J. Chem. Soc. Chem. Commun. 1981 571. 24 J. Thiem and B. Meyer Chem. Ber. 1980 113 3058 3075. 25 J. Thiem and S. Kopper J. Carbohydr. Chem. 1983 2 75. Carbohydrates 315 which may represent an interesting if more laborious alternative to the existing methodology.In an extension of the N-iodosuccinimide method uronic acid ester glycals could be glycosylated and showed enhanced yields of the 1,2-?runs diequatorial product which led to 2-deoxy-P-glycosides. In addition to a model study26 this concept could by applied in the preparation of a disaccharide part of flarnbamy~in.~~ A novel approach was reported to lead to 2-deoxy-P-glycosides exclusively.28 By addition of 0,O-dimethylphosphorodithioicacid to glycals [e.g. (19)] regio- and stereo-selectively the a-thiophosphate (20) was obtained which on reaction with sodium alcoholates gave 2-deoxy-P-glycosides (21) of simple alcohols (R = Me to Bu') in greater than 85% yield. Extension of this concept to more complex cases will be of relevance.OAc PAC Aco+O) 1. RONa P 2. Ac,O,Py OR AcO SP(S)(OMe)z AcO The ready a~ailability~~ of the C-2 epimeric 2,6-dibromo-2,6-dideoxy-a-~-manno-and gluco-pyranosyl bromides opens possibilities for the preparation of a-and P-2-deoxyarabino glycosides. As was demonstrated with simple alcohols30 the rnunno-isomer gave only a-glycosides in high yield whereas the glucosyl bromide led predominantly to the P-compounds. Reaction of the 2-bromoglucosyl bromide (22) with the selectively blocked 2,6-dideoxy-a- D-arabino compound (23) promoted by silver triflate gave 16% of the a-linked disaccharide and 57% isolated yield of the desired P,1-+4-linked compound (24). The latter could be transformed into the bamflalactone part (25) of flambamy~in.~' Use of the halide (22) with other more reactive sugar aglycons could be elaborated into a high yielding P-2-deoxydisac- charide synthesis (92% ;P :a = 7 :1) for oligosaccharides with a terminal arabino unit.32 A similar general stereoselective method for 2-deoxy-P-glycosides having the rib0 configuration did not exist but was highly desirable with respect to many natural compounds having this structural feature e.g.the cardiac glycosides. Wiesner and co-~orkers~~ have developed a method using the 4-p-methoxy-benzoyl-3-methyl-urethane of digitoxose (27) which on condensation with the digitoxigenin precursor (steroid 3-OH) in the presence of p-toluenesulphonic acid gave the glycoside (28) in 83% yield. After reductive cleavage of both ester groups the anomers (29) could be separated and were obtained in the ratio P :a = 7 1.The intermediate formation 26 J. Thiem and P. Ossowski J. Carbohydr. Chem. 1984 3 287. 27 A. Prahst Diss. Univ. Hamburg 1984. 28 M. Michalska and J. Borowiecka J. Carbohydr. Chem. 1983 2 99. 29 K. Bock I. Lundt and C. Pedersen Carbohydr. Res. 1981 90 7. 30 K. Bock I. Lundt and C. Pedersen Carbohydr. Res. 1984 130 125. 31 I. Lundt J. Thiem and A. Prahst 1.Org. Chem. 1984 49 3063. 32 J. Thiem and M. Gerken J. Carbohydr. Chem. 1982/3 1 229. 33 H. Jin T. Y. R. Tsai and K. Wiesner Can. J. Chem.. 1983 61 2442. 316 J Thiem Me HO HO of a charged species (30) by 1,3-participation is believed to account for appreciable stereoselection.In further studies towards the stereoselective synthesis of the com- plete digit~xin~~ the strategy had to be varied because a p-TsOH catalysis is ruled out owing to the acid lability of 2,6-dideoxy oligosaccharides. Along similar lines in a somewhat tedious protection-deprotection sequence the glycoside (3 1) was obtained which could be condensed with the a-thiodigitoxoside (32) in the presence of HgCI2-CdCO3 giving in 60% yield virtually only the &1+4 species. After partial deacylation a 3.5 :1 mixture of the regioisomeric esters (34) and (35) were obtained and separated the former could be used for another similar glycosylation with (33) to give the trisaccharide species in 58% yield. By further cleavage of the ester groups and oxidative refunctionalization of the butenolide portion in the aglycone the total synthesis of digitoxin (36) was finished.Contrary to the previously reported inversion of configuration in the course of HgC1,-promoted glycosylations of thioglyc~sides,~~ here the stereospecific outcome of the reaction was ascribed to the intermediacy of a 1,3-p-methoxybenzoxonium ion (37). Syntheses of methyl-branched amino- and nitro-sugars which represent the sugar moieties of antibiotics attract further efforts. A very efficient sequence consists of the cyanomesylation of a (mostly 3-) dose followed by reductive spiroaziridine formation and final reductive ring-opening. This approach of the groups of Brima~ombe~~ gave the compounds with equatorial methyl and and Y0shimu1-a~~ axial amino-group.By an additional oxidative step the corresponding methyl- branched nitro-sugars are obtained thus a kijanose derivative3* (for a previous preparation see re$ 39) and rubranitrose4’ were prepared. Under the catchword ‘pyranosidic homologation’ Fraser-Reid and his group4’ have studied the extension of the carbohydrate template via the secondary positions 34 T. Y. R. Tsai H. Jin and K. Wiesner Can. J. Chem. 1984,62 1403. 35 R. J. Ferrier R. W. Hay and N. Vethaviyasar Carbohydr. Res. 1973 27 55. 36 J. S. Brimacombe R. Hanna and L. C. N. Tucker J. Chem. Soc. Perkin Trans. I 1983 2277. 37 T. Yasumori K. Sato H. Hashimoto and J. Yoshimura Bull. Chem. SOC.Jpn. 1984 57 2538. 38 J. S. Brimacornbe and K. M. M. Rahman Carbohydr. Res.1983 123 C19. 39 K. Funaki K. Takeda and E. Yoshii Tetrahedron Lerr. 1982 23 3069. J. Yoshimura T. Yasumori T. Kondo and K. Sato Bull. Chem. SOC.Jpn. 1984 57 2535. 41 B. Fraser-Reid L. Magdzinski and B. Moho J. Am. Chem. Soc. 1984 106 731. Carbohydrates 3 17 OH .low NHMe OR' (27) (28) R' = pMeBz; R2= CONHMe (29) R' = R2= H (31) R' = R2= pMeBz OpMeBz SEt (32) R = pNBz +NHMe (33) R =p-MeBz I (30) Rlo++o-steroid OR' Op-MeOBz Rovl o-.+-0 (34) R' = H; R2= pMeBz 1. +(33) 2. LiAH41! (35) R' = pMeBz; R2= H OMe ~o (37) ODigitoxigenin OH OH OH C-2 to C-4 and the primary position at C-6. In this approach complete regio- and stereocontrolled reactions are made possible at 'off -template' positions because these are embedded into additional carbohydrate-like pyran or dihydropyran ring struc- tures.As an example the mannose derivative (38) on treatment with the diethyl- aluminium salt of propargylic alcohol gave as expected a regiospecific attack at the 2-position. Partial hydrogenation by Lindlar catalyst gave the ally1 alcohol (39) and after hydrolysis and reprotection of the primary C-6 position furnished the internal glycoside (40) which may be used for epoxidation and further functi~nalization.~~ A radical carbon-carbon linkage process by treatment of sugar methyl xanthates with tri-n-butyl stannic hydride and acrylonitrile introduced a P-cyanoethyl side- L. Magdzinski B. Cweiber and B. Fraser-Reid Tetrahedron Lett. 1983 24 5823. 318 J.Thiem OTBDMS 1. Et2Al-C=C- CH 2. Bu‘Me,SiCI 2. Lindlar BnO (38) (39) ~hain.4~ In the glucofuranose series a 7 :3 ratio in favour of the 3-branched gluco- compound was obtained and with the corresponding galactofuranose the reaction proceeded with complete diastereoselectivity to the C-branched galucto- derivative. By a biomimetic approach the a-alkylation of uloses could be achieved. Following proton abstraction with lithium di-isopropylamine methylation (or hydroxy methyl- ation) occurs stereo- and regio-specifically with methyl iodide (or formaldehyde) in HMPT at -70 “C in 40-80% yield.44 Routes to gem-dialkylated sugars by application of a highly stereoselective Claisen rearrangement were reported by Fraser-Reid and collaborator^^^ Following Wittig reaction of a 2-ulose and reduction the exocyclic ally1 alcohol was obtained and mildly transformed into the vinyl ether (41).Thermal treatment in refluxing ben- zonitrile gave exclusively the derivative of gluco-configuration (42;axial acetal- dehyde and equatorial vinyl group) in 85% yield. After cleavage of the TBDMS- blocking group lactol formation (43) occurs which is only possible in the depicted configuration. Obviously the presence or absence of the anomeric methoxy group (R = OMe or R = H) seems to be irrelevant which implies neither a steric hindrance nor a stereoelectronic influence for this oxy-Cope rearrangement. The stereoselec- tivity is attributed to an exclusive axial folding which results in an axial attack at C-2 and for similar reasons an equatorial attack at C-3.Another ‘pyranosidic homologation’ described by Fraser-Reid’s group& makes use of a Wittig-extension from a C-6-aldehydo-sugar (44). The facile formation of the olefin sugar (82% 4 1 E/Z-mixture) was followed by methanolysis to give a single isomer (45) in 90% yield termed a ‘satellite’ hexenopyranoside which proved to be a useful carbohydrate template. Part of the anthracycline antibiotic nogalamycin (46) is an unusual optically active 2,6-epoxy-2H-1-benzoxocin moiety. Starting with the xylofuranose (47)addition of a methyl Grignard then Collins oxidation and treatment with a second Grignard reagent obtained from 2-bromo-4-methyl-2-benzyloxybenzene gave the D-gluco 43 B. Giese J. A. Gonzalez-Gomez and T.Witzel Angew. Chem 1984 96,51. 44 A. Klemer and H. Thiemeyer Liebigs Ann Chem. 1984 1094. 45 D. B. Tulshian R. Tsang and B. Fraser-Reid J. Org. Chem. 1984,49 2347. 46 B. F. Moho L. Magdzinski and B. Fraser-Reid Tetrahedron Left. 1983 24 5819. Carbohydrates CHO Me0 7 HOWo\ 2. MeOH/PyH+OTs-BnO Bn OMe derivative (48) exclusively in 80% yield. A similar treatment with inversion of the Grignard additions 1 and 3 led to the L-ido isomer (49). This demonstrates stereo- specific additions of the Grignard reagents to the ketone intermediates according to Cram's rule. Further hydrogenolysis and hydrolysis of the glum compwnd (48) gave only a modest yield of the product (50) contaminated with the ~-ido isomer (51) owing to acid mediated R/S interconversion at the chiral tertiary alcohol function.The similar treatment of (49) gave the more stable pure L-ido-isomer (51) in approximately 60% yield.47 XD \ /\ OH--O--HO 9 (47) AcO Me Me 0 OAc OAc AcO OAc (48) R' = OH;R2 = H (50) (51) (49) R' = H;R2 = OH Halogeno- Phospho- and Thio-sugars.-Fluorinated saccharides seem to be of paramount interest owing to their biological properties. A review which updates the literature appeared recently.48 One prominent target molecule was 2-deoxy-2-(fluorine- lS)-fluoro-D-glucose useful in positron emission tomography. Rapid procedures were developed by SN2 inversion of a 2,3-cyclic sulphated mannose 47 F. M. Hauser and T.C. Adams Jr. J. Org. Chem 1984 49 2296.48 A. A. E. Penglis Adv. Carbohydr. Chem. Biochem 1981,38 195. 320 J. 7’hiem derivative with tetraethyl ammonium “fl~oride,”~ and of a 2-0-trifluoromethylsul-phony1 mannose with CSH’~F,.’’ Addition of hypofluorite and acetyl fluoride to the arabino-glycal,” and reaction of the selectively prepared benzy13,4,6-tri-O-benzyl-P-D-mannopyranoside with DAST and subsequent hydrogenoly~is’~ led to the same compound. Through reaction of the 1-triflate of 2,3;4,5-di- 0-isopropylidene-P- D- fructopyranose with tris(dimethy1amino)sulphonium difluorotrimethylsilicate (TASF) 1 -deoxy- 1 -fluoro-D-fructose was obtained and subsequently enzymatically condensed with UDP-glucose to give 1’-deoxy- l’-fl~oros~crose.~~ There have been only a few reports on the preparation of sugars with sulphur in the hemiacetal ring but an enormous number of papers for the syntheses of sugars with phosphorus in the hemiacetal ring and these have been recently re~iewed.’~.’’ The base-catalysed dialkylphosphite addition to aldose derivatives (free aldehydo- sugars6 or C-1 unblocked compound57) with subsequent internal transesterification led to formation of y-or 8-phostones respectively.Most isosteric phosphonate analogues of naturally occurring phosphates are prepared by the Michaelis-Arbuzov reaction on a halogenomethylene precursor. In this way the phosphono analogues of a-and P-D-mannopyranosyl pho~phate’~ were obtained. A more attractive approach59 to the a-and P-D-ribofuranosyl phosphate analogues [2,5-anhydro-1- deoxy-1 -(phosphono)-D-altritol or -allit011 may be considered the Horner-Emmons reaction of the C-1.unblocked ribofuranose derivative with methylene bis-dimethyl- phosphonate previously introduced by Moff att et aL6’ Although no synthetic reports have appeared yet the isolation and structure e1ucidation6l of the ribofuranoside (52) from the arsenic-accumulating brown kelp Ecklonia rudiata may inspire some. Me O \ I1 II HO OH R = SO,H,OH Amino-sugars.-An attractive cis-oxyamination procedure which starts from allyl- alcohols leads to aminodeoxy sugars of biological relevance was simultaneously developed and applied by several groups. Following comparable pathways improved 49 T. J. Tewson J. Nucl. Med. 1983 24 718 (Chem. Abstr. 1984 100 6969~).so S. Levy E. Livini D. R. Elmaleh D. A. Varnum and G. L. Brownell Znt. J. Appl. Rudiat. hot. 1983 34 1560 (Chem. Abstr. 1984 100 121 462u). ” M. J. Adam B. D. Pate J. R. Nesser and L. D. Hall Curbohydr. Res. 1983 124 215. s2 A. Dessinges A. Olesker G. Lukacs and T. T. Thang Curbohydr. Res. 1984 126 C6. ” P. J. Card and W. D. Hitz J. Am. Chem. Soc. 1984 106 5348. 54 Z. J. Witzak and R. L. Whistler J. Carbohydr. Chem. 1983 2 351. s5 H.Yamamoto and S. Inokawa Adu. Carbohydr. Chem. Biochem. 1984,42 135. 56 A. E. Wroblewski Curbohydr. Res. 1984 125 C1. s7 J. Thiem and M.Giinther Phosphorus Sulfur,1984 20 67. F. Nicotra R. Perego F. Ronchetti G. Russo and L. Toma Curbohydr. Res. 1984 131 180. 59 R. B. Meyer Jr. T. E. Stone and P. K. Jesthi J.Med. Chem. 1984 27 1095. G. H. Jones E. K. Hammura and J. G. Moffatt Tetrahedron Lett. 1968 5731. J. S. Edmunds and K. A. Francesconi J. Chem. SOC. Perkin Trans. 1 1983 2375. Carbohydrates 321 syntheses of L-daunosamine were reported.62d4 Thus L-rhamnose was converted into the glycal(53) which on Ferrier glycosylation C-4 mesylation and SN2inversion using caesium propionate provided the benzyl a-~-threo glycoside (54) in excellent yield. On treatment of its trichloroacetimidate (55) with N-iodosuccinimide an iodonium ion induced cyclization gave the 2-iodo-oxazoline (56).By acid hydrolysis the 3-amino-2,3,6-trideoxy-2-iodo-a-~-lyxo-derivative (57) was obtained which was converted into L-daunosamine using standard procedures.a Another elegant reaction sequence was developed along the idea of an iodonium induced cyclization of allylic amide systems by Fraser-Reid et al.The preparation of a garosaminide (62)65depicts the concept nicely; the well-known keto epoxide (59) is transformed into the methylene derivative (60)via Wittig reaction and then the opening of the epoxide ring by ammonia occurs regio- and stereo-selectively. For the decisive cyclization step to (61)iodo-bis-collidinium perchlorate was used effectively. Further reduction of the iodomethylene group with Bu;SnH then quater- nization with methyl iodide reduction with sodium borohydride and acid methanoly- sis gave the desired garosaminide (62)in an eight-step synthesis with 14% overall yield.65 3BzH WOMe OwoMe N 0Bz __* OMe 0 0Bz (59) MeHN-OH (62) 62 H.W. Pauls and B. Fraser-Reid J. Chem. SOC.,Chem. Commun. 1983 1031. 63 G. Cardillo M. Orena S. Sandri and C. Tomasini J. Org. Chem. 1984 49 3951. 64 D. Springer Diss. University Hamburg 1984. 65 H. W. Pauis and B. Fraser-Reid Can. J. Chem. 1984 62 1532. 322 J. Thiem Shifting of amino-functions opens up novel preparative processes for amino- sugars. Thus the xylopyranoside (63) in apolar medium formed an aziridinium intermediate (64) (proven by n.m.r.) which is opened regioselectively at C-3. This results in formation of the 4-amino-~-hexoside (65); in water a concerted process led to the 3-hydroxy-4-amino-derivative(66).66 Another complex synthesis of daunosamine also made use of such a shift of amino-groups from the 2-to the 3-position via an epimine intermediate.67 OMe (64) OMS-(65) R = Ms (66) R = H Primary and secondary triflate groups could be easily substituted with ammonia to give the inverted amino-sugars in 50-80% yield:' and amino-acid-substituted carbohydrates were similarly ~btained.~' There were two interesting reports on the Pictet-Spengler reaction between a sugar and biogenic amines.For instance dopamine and 2,5-anhydro-~-mannose (67) gave the isoquinoline adduct (68)." Similarly in following a route from methyl a-D-mannoside to heteroyohimbine alkaloids this reaction was used as a key step.71 C'HO Dopamine yo$ OH I LOH HO HO Finally a novel procedure for the synthesis of nitro-sugars may be mentioned here.Following known preparations of the corresponding primary and secondary azides (69a 70a) stirring with triphenyl- or tri-n-butylphosphine gave the phosphine imines (69b 70b) ozonolysis of which at -78 "C with 34 molar equivalents yielded the nitro-compounds (69c 70c) smoothly and in good yields.72 Anhydro-sugars.-An improved method for the preparation of 173-anhydro-sugars with gluco-configuration is reported.73 Using 170 n.m.r. and "0-induced isotopic shifts in 13C n.m.r. evidence is pre~ented'~ that in the formation of the anhydro- 66 D. Picq M. Cottin D. Anker and H. Pacheco Tetrahedron 1983 39 1797. 67 M. K. Gurjar V. J. Patil J. S. Yadav and A. V. R. Rao Carbohydr. Res. 1984 129 267. 68 A. Malik N. Mza M. Roosz and W.Voelter J. Chem. SOC.,Chem. Commun. 1984 1530. 69 A. Malik W. Kowallik P. Scheer N. Mia and W. Voelter J. Chem SOC.,Chem. Commun. 1984 1229. 70 D. B. MacLean W. A. Szarek and I. Kvarnstrom J. Chem SOC. Chem Commun. 1983 601. 71 J. Kervagoret J. Nemlin Q. Khuong-Huu and A. Pancrazi J. Chem SOC., Chem. Commun. 1983,1120. 72 E. J. Corey B. Samuelson and F. A. Luuio J. Am. Chem. Soc. 1984 106 3682. 73 F. Good and C. Schuerch Carbohydr. Res. 1984 125 165. 74 A. Dessinges S. Castillon A. Olesker T. T. Ton and G. Lukacs J. Am. Chem. Soc. 1984 106. 450. 323 Carbohydrates 3 Aco~ocH2cc,3 (b) X = N=PR AcO (a) N3 galactopyranose (72) by treatment of the glucopyranose (71) with sodium azide a P-oxyanion is the intermediate as previously proposed by Brimacombe et aZ.75 D-Mannose on tosylation subsequent treatment with water and then with sodium hydroxide at pH 9 gave an easy access to 1,6-anhydro-P-mannose (mannosan) which was isolated as the 2,3-acetonide in 60% overall yield.76 By epoxidation using the Payne procedure 1,2;5,6-dianhydro-3,4-dideoxy-~-threo-hex-3-enitol gave a mixture of two enantiomeric trianhydro-hexitols with D-rnanno (73) and Dido (74) configur-ation (ratio 1 3).A similar treatment of the corresponding erythro-isomer gave the racemate of the D,L-ghco-derivative (75).77This new class of sugar oligoepoxides awaits further interesting transformations. An important feature of the anhydro-compounds is an attractive transfer into linear glycanes using a Lewis acid ring-opening polymerization.The formation of a number of ( 1 +6)-glycanes from different 1,6-anhydro-precursors was studied and published.78i79Another approach by Schuerch's group described the polymeriz-ation of 1,3-anhydro-D-mannose derivatives into (1 -* 3)-a-D-rnannan.*' A stereoregular (3+6)-a-D-xylofuranan could be made from the oxetane derivative 75 J. S. Brimacombe J. Minshall and L. C. N. Tucker J. Chem. SOC,Perkin Trans. I 1973 2691. 76 M. Georges and B. Fraser-Reid Carbohydr. Res. 1984 127 162. 77 P. KOII M. Olting and J. Kopf Angew Chem. 1984 96 222. ?a T. Uryu Y. Sakamoto K. Hatanaka and K. Matsuzaki Macromolecules 1984 17 1307. 79 K. Kobayashi and H. Sumimoto Polym. 1.(Tokyo) 1984 16 297. 80 F. Kong and C.Schuerch Macromolecules 1984 17 983. 324 J. Thiem 3,5-anhydro-1,2-0-isopropylidene-~-~-xylofuranose.~~ The recent activities in this field were summarized by Uryu et aL8*and reference should be made to Schuerch’s previous review.83 Esters Ethers Acetals and some Protection-group Properties.-A thorough study of the direct mono-benzoylation of pento- and hexo-pyranosides using dibutyl stannic oxide or bis(tributy1 stannic)oxide has been published. Even in the presence of a primary OH-group an equatorial OH-group flanked by a cis-OH (or OMe) group is always selectively a~tivated.~ Another paper on the use of stannylidene com- pounds originally introduced by Moffatt et d.,85 also noted the regioselective activation of secondary OH-groups.It was shown here that different solvents or the presence of Lewis acids changed the co-ordination ability and hence the acylation ratios.86 By a classical approach methyl 4,6- 0-benzylidene-a-D-mannopyranoside was shown to be benzylated at the 3-position predominantly (66’/0).~~ Using the stannyl- idene approach for methyl 4,6- 0-benzylidene-P- D-glucopyranoside benzylation catalysed with tetrabutylammonium bromide gave the 3- and the 2-0-benzyl ethers in the ratio 2 1 (total yield 90’/0).~~ A very useful selective partial benzylation of galactosides was observed:89 methyl a-D-galactopyranoside and benzylchloride in the presence of lithium hydroxide as the base gave the 2,3,6-tri-O-benzyl derivative. With potassium or rubidium hydroxide however the 2,4,6-tri- 0-benzyl compound resulted as the main product.The P-anomer gave a 3,4,6-tri- 0-benzyl-ether irrespec- tive of the nature of the base. The kinetic acetonation in DMF with 2-methoxypropene and p-toluenesulphonic acid catalysisg0 could be used to prepare the 4,6-monoacetal (goo/,) or the 2,3;4,6- diacetal (73%) of a-or P-D-glucopyranosides depending on the molar ratio.” A detailed gas-liquid chromatography study of the preparation of 1,2 ;5,6-di- 0-isopropylidene-D-mannitol,an attractive starting material for the synthesis of other chiral compounds revealed that the classical procedure with acetone and zinc chloride led to the best results (65%) whereas dimethoxypropane in DME with SnC1,-catalysis led to a complex mixture.92 In contrast to previous results93 the use of 2-methoxypropene gave only 44% of that diacetal and in addition another 29% of the 1,2;4,6- and 17% of the 1,2 ;3,6-di- 0-isopropylidene-D-mannitolisomers.94 Three new protecting groups for the regiospecific blocking of a primary OH-group gave promising results.4-( Methylthiomethy1enoxy)butyricacid (76) and 0-(methyl-thiomethy1enoxy)methylbenzoic acid (77) provided the 5’-0-acyl derivatives of 81 T. Uryu Y. Koyama and K. Matsuzaki Macromol. Chem 1984 185 2099. 82 T. Uryu and K. Hatanaka Yuji Gosei Kagaku Kyokai Shi 1984,42 557 (Chem. Absrr. 1984 101 111 275s). 83 C. Schuerch Adu. Carbohydr. Chem. Biochem 1981 39 157. 84 Y. Tsuda M. E. Haque and K. Yoshimoto Chem Pharm. Bull. 1983 31 1612. 85 D. Wagner J. P. H. Verheyden and J.G. Moffatt J. Org. Chem. 1974 39 24. 86 C. W. Holzapfel J. M. Koekemoer and C. F. Marais S. Afr. J. Chem. 1984 37 19. 87 Y. Kondo K. Noumi S. Kitagawa and S. Hirano Carbonydr. Res. 1983 123 157. 88 K. Takeo and K.Shibata Carbohydr. Res. 1984 133 147. 89 N. Morishima S. Koto M. Oshima A. Sugimoto and S. Zen Bull. Chem. SOC.Jpn. 1983 56 2849. 90 J. Gelas Ado. Carbohydr. Chem. Biochem 1981 39 71. 91 J. L. Debost J. Gelas D. Horton and 0. Mols Carbohydr. Res. 1984 125 329. 92 J. Kuszman E. Tomori and I. Meerwald Carbohydr. Res. 1984 128 87. 93 J. L. Debost J. Gelas and D. Horton J. Org. Chem. 1983 48 1381. 94 J. Kuszman E. Tomori and P. Dvortsak Carbohydr. Res. 1984 132 178. Carbohydrates MeS-CH2-0-CH,-CH2-CH2-COOH (76) (77) MeS-CH2-0-CH 9 COOH CH,X CH,X I I H,C=C-O-CH,-Ph /O-C-O-CH,-Ph thymidine in 70% yield.95 Deacylation with conc.aqueous ammonia was slow but treatment with mercury perchlorate/2,4,6-~ollidineand then mild base gave a rapid cleavage. The 4,4',4''-tris-( 4,5-dichlorophthalimido)trityl group (CPTr) introduced via its bromide (78) in DMF with silver nitrate promotion also blocks primary OH-groups specifically in high yield. Cleavage is effected mildly using hydrazine in pyridine-acetic acid.96 A novel open acetal protecting function for primary OH-groups was developed by Mukaiyama et Treatment of the glucopyranoside (79) with 2-(benzy1oxy)- 1-propene and catalytic amounts of palladium dichloride- 1,5-octadiene complex gave the acetal (80; X = H) and no alkylidene derivative (in contrast cf ref 90).The fluoro-acetals (80; X = F) proved to be more resistant to acid hydrolysis?* Cleavage was smoothly effected by hydrogenation on palladium-charcoal. De No00 Synthesis of Carbohydrates.-For the majority of classical carbohydrate chemists the first approaches towards the preparation of n/ L-mixtures of sugars from non-carbohydrate sources were not interesting. Reasons for that were the rather troublesome resolutions necessary to obtain pure enantiomers and also the better understanding of transformation processes starting with abundant carbohydrate materials in enantiomerically pure form. During the last two decades however the new approach has developed into an attractive tool which nowadays in certain cases may compete advantageously with classical syntheses.The topic has been reviewed earliep9 and recently updated,'oo*'o' and another novel review article has 95 J. M. Brown C. Christodoulou C. B. Reese; and G. Sindona J. Chem. SOC.,Perkin Trans. 1 1984 1785. 96 M. Sekine and T. Hata .IAm. Chem. SOC.,1984 106 5763. 97 T. Mukaiyama M. Ohshirna and M. Murakarni Chem. Lett. 1984 165. 98 T. Mukaiyarna M. Ohshima H. Nagaoka and M. Murakami Chem. Lett.. 1984,615. 99 J. K. N. Jones and W. A. Szarek in 'The Total Synthesis of Natural Products' ed. J. W. ApSimon Vol. 1 Wiley-Interscience New York 1973 p. 1. I00 A. Zamojski A. Bannaszek and G. Grynkiewicz Adu. Carbohydr. Chem. Biochem. 1982,40 1. lo' A. Zamojski and G. Grynkiewicz in 'The Total Syntheses of Natural Products' ed.ApSimon Vol. 6 Wiley-Interscience New York 1984 p. 141. 326 J. Thiem covered the acyclic stereoselective synthesis of carbohydrates.lo2 Attractive target carbohydrates for total synthetic procedures are the antibiotic sugars like amino- deoxy- and branched-chain carbohydrates as well as unnatural enantiomers and complex higher carbon sugars. An unusual approach to amino-sugars by Weinreb et aL103makes use of an intramolecular N-sulphinyl dienophile Diels- Alder process. Treatment of the carba- mate of the (E,E)-diene alcohol (81) with thionyl chloride in pyridine gave a single Diels- Alder adduct (83) in 80% yield. Its stereoselective formation could be rational- ized by assuming the cycloaddition of the intermediate N-sulphinyl carbamate to occur through the depicted transition state (82) with the carbonyl group endo and the sulphinyl oxygen em.Further reaction of the dihydrothiazine oxide (83) via an allylic sulphoxide its [2,3]-sigmatropic rearrangement desulphurization and a series of further steps led to 5-epi-desosamine (84). Me Me (82) An asymmetric seven-step synthesis of methyl 3,4-anhydro-2,6-dideoxy-a-~-ribo-hexopyranoside the precursor for L-daunosamine was accomplished starting with cyclopentadiene. Key steps were the asymmetric hydroboration stereoselective epoxidation and Baeyer-Villiger oxidation giving the carbohydrate 1act0ne.l'~ Start- ing with L-glutamic acid the separable D-amiCetOnO and L-rhodinono y-lactones could be obtained which following reductions furnish the corresponding amino- sugars.'05 Ethyl (S)-lactate (85) was transformed into the nitrile (86) which on treatment with the magnesium enolate of t-butyl acetate gave the (Z)-B-amino acrylate derivative (87).This was further processed into N-benzoyl-L-acosamine (88). A corresponding approach using this nitrile-acetate coupling process led to formation of the C-4 epimeric N-benzoyl-L-daunosamine (89).'06 The nitrone cycloaddition represented another useful approach to amino-sugar components. Thus De Shong et aZ.lo7 transformed the ester (90) into a single (2) -benzyl nitrone (91) using standard procedures. This displayed high diastereofacial- and stereo-selectivities for the endo transition-state and its cycloaddi- tion reaction with ethyl vinyl ether gave the isoxazolidine isomer (92) exclusively.By catalytic hydrogenation and simple transformation the D/ L-daunosaminide (93) was obtained. Along corresponding lines an interesting use of the 1,3-dipolar 102 G. J. McGarvey M. Kimura T. Oh and J. M. Williams J. Carbohydr. Chem. 1984 3 125. 103 S. W. Remiszewski R. R. Whittle and S. M. Weinreb J. Org. Chem. 1984 49 3243. 104 G. Grethe J. Sereno T. H. Williams and M. R. Uskokovich J. Org. Chem 1983 48 5315. 105 G. Berti P. Caroti G. Catelani and L. Monti Carbohydr. Rex 1983 124 35. 106 T. Hiyama K. Nishide and K. Kobayashi Tetrahedron Lett. 1984 25 569; Chem. Lett. 1984 361. lo' P. DeShong C. M. Dicken J. M. Leginus and R. R. Whittle J.Am. Chem. Soc. 1984 106 5598. Carbohydrates OBu' I CH,=C-OMgX (88) R' = OH;R2 = H (89) R' = H;RZ= OH Me 0-PhCH2 -N I CH2Ph A~O MfKJoMe NHAc (93) cycloaddition of nitrile oxides was reportedlog to give isoxazolines which could be further transformed into amino-sugars e.g. methyl lividosaminide. A double asym- metric induction was observed in the diastereoselective hydroxyalkylation of the lithium salt of Schollkopf's L-alanine dimer (94) by 2,3-O-isopropylidene-~- gly~erinaldehyde.'~~ After hydrolyses with acetic acid the adduct (95) was obtained (80% yield) and could be further transformed into the pure lactone (96). 3. MeC0,H (94) (95) 108 V. Jager and R. Schohe Tetrahedron 1984,40 2199.109 J. C. Depezay A. Dureault and T. Prange Tetrahedron Lett. 1984. 25 1459 328 J. Thiem Roush et uL"o*lllhave developed very nice diastereoselective syntheses of most of the 2,6-dideoxyhexoses. The racemic (E)-allylic alcohol (97) by kinetic reso- lution-enantioselective epoxidation procedure of Sharpless et ~~2.l'~ gave the (+)-erythro epoxide (98) in addition to the kinetically resolved (-)-E-allylic alcohol (102). The urethane (99) obtained from (98) was transformed into the carbonate (100) with Et,AlCl and this on subsequent treatment with base and ozonolysis gave D-(+)-OliVOSe (101). The resolved (-)-E-allylic alcohol (102) could be epoxidized to give the (-)-erythro-epoxide (103) which on acid hydrolysis gave the D-n'bO-triOl (104) and further ozonolysis led to D-(+)-digitoxose (105).Based on crotonaldehyde the six-step synthesis of olivose gave approximately 17% overall yield; that of digitoxose amounted to 22%. (-)-DIPT Bu'OOH OH OR (97) (98) R = H (99) R = CONHPh Me Ti (0wj (+)-DET Bu'OOH Y -OH OH HO OH (104) 110 W. R. Roush R. J. Brown and M. DiMare J. Org. Chem. 1983,48 5083. 111 W.R. Roush and R. J. Brown J. Org. Chem 1983,48 5093. 112 V.S.Martin S. S. Woodard T. Katsuki Y. Yamada M. Ikeda and K. B. Sharpless J. Am. Chem SOC. 1981 103 6237. Carbohydrates 329 An excellent new approach for the synthesis of 3-deoxy-~- manno-2-octulosonic acid (KDO) used a biomimetic pathway in condensing a C3 unit acrylic acid derivative (mimicking pyruvate) to aldehydo-arabinose as the Cs unit."3 The crystal- line amide (106) was transformed into the dilithium salt (107) which attacked the arabinose (108) in a highly diastereoselective way giving the crystalline manno- compound (109) predominantly (85% yield munno :ghco > 15 1).Further trans- formation gave the enol lactone (110) which was previously reacted to give the KDO ammonium salt. OH R'bo (106) R = H (107) R = Li (R'/CHo) KDO In the cyclocondensation of the chiral butadiene (1 11) with furfural catalysed by Eu(hfc) after work-up with triethylamine-methanol the optically active 3-ketone (1 12) was obtained in 75% ~ie1d.I'~ Stereospecific reduction and ozonolytic cleavage of the furan ring with final reduction and acetylation steps gave the 4-deoxy-~- arabino-glycoside (1 13).This combination of a chiral auxiliary on the diene moiety with chiral catalyst seems to permit syntheses of optically pure saccharides without a resolution step. Zinc chloride catalysed cycloadditions of formaldehyde to sub- stituted dienes like (111) and led to a series of pentopyranose derivatives.115 0-( -)-Menth Eu(hfc) b? 0-(-)-Menth OAc 3 Oligosaccharides Part of the glycosylation procedures as well as synthesis and application of precursors for oligosaccharide synthesis have been discussed above under the chapter 'glyco- sides'. Nowadays an increasing number of researchers head towards larger substruc- tures of carbohydrate-containing biological material and there are some six groups in Canada France Germany Japan and Sweden whose leading activities in these areas should be mentioned.Most of the preparations of higher hetero-R. R. Schmidt and R. Betz Angew. Chern 1984 % 420. M. Bednarski and S. Danishefsky J. Am. Chern. Soc 1983 105,6968. 1IS S. Danishefsky and R. R.Webb 11. J. Org Chern. 1984. 49. 1955. 330 J. Thiem oligosaccharides require multi-step syntheses and a highly reliable sophisticated analysis mostly provided by extended 'H and 13Cn.m.r. spectroscopy is a necessity. The number of carbohydrate units and the complexity of the synthesis need not necessarily exhibit a linear correlation that is a disaccharide synthesis may be more difficult than a block synthesi3 of e.g. a nonameric oligosaccharide.An updating of the general concepts previously discussed' and some specific newer developments in oligosaccharide synthesis have been given recently by H. Paulsen.l16 G1ycoconjugates.-Within the living cells most hetero-oligosaccharide structures are more or less closely associated (or bound) to a variety of other mono- or oligomeric materials of biochemical relevance. The biochemical and structural aspects of these classes of compounds termed glycoconjugates have been reviewed recently e.g. see rej 117. During the recent decade synthetic efforts have increased considerably and there are quite a large number of most interesting glycoconjugates or substructural units thereof available by modern glycosylation approaches. Surveys by some of the leading groups have been given recently on the syntheses of complex oligosac- charide chains of glycoprotein~."~~"~ A particularly intensely- studied question was the approach to complex glycan chains which represent the core region of glycoproteins.One of the first successful reports'20*'21 started from a mannoside which by use of suitable blocking groups was built up to the mannotrisaccharide (115). After selectively deblocking the positions 2 and 4 in the one and 2 and 6 in the other terminal mannopyranoside x-2\a-D-Manp-(1 +3) x-4/ \ 4 x p-D-Galp-(1 -.* LZ)-cy-~-GlcNPhth-Br 1 + x-2 (114) \a-D-Manp-(1 -P 6)P-Man / X-6 (115) X = OH (1 16) X = p-D-Galp-(1 .+ 4)-p-~-GlcNAc unit this compound was glycosylated using a four molar equivalent of the disac- charide bromide of a P-D-Gal-( 1+4)-~-GlcNAcderivative (114) in a silver triflate promoted condensation.By this process the trisaccharide mannose unit was enlarged to an undecamer oligosaccharide (116) in a two-step process in 5% yield by a comparatively simple and efficient method. A similar approach was used in the synthesis of cell surface glycans which in the first series led to the preparation of a P-D-Man-( 1 -+4)-p-~-GlcNAc-( 1-* 4)-~-GlcNAc derivative (118) selectively 116 H. Paulsen in 'Selectivity -a Goal for Synthetic Efficiency' ed. W. Bartmann and B. M. Trost Proceedings 14th Workshop Conferences Hoechst Verlag Chemie Weinheim 1983 p. 169. 117 'The Glycoconjugates' ed. M. Horowitz Vol. 3 and 4 Academic Press New York 1982.H. Fsulsen Chem SOC.Reu. 1984 13 15. 119 T. Ogawa H. Yamamoto T. Nukada T. Kitajima and M. Sugimoto Pure AppL Chem. 1984 56 779. I2O J. harp H. Baumann H. Loenn J. Loenngren H. Nyman and H. Ottosson Acta Chem. Scand. Ser. R 1983 37 329. 121 H. Loenn and J. Loenngren Carbohydr. Res. 1983 120 17. Curboh ydru tes 331 2 x p-D-Galp-(1 +4)-p-~-GlcNAc-( 1 +a)-cu-~-Glcp-Br(1 17) I X-6 'P-D-Manp-( 1 -+ 4)-p-~-GlcNAc-( 1 +4)-p-~-GlcNAc 1 x-3' (118) X = OH (119) X = p-D-Galp-(1 -P 4)-p-~-GlcNAc-(l-+ 2)-a-~-Manp o-w I I Me Me Me unprotected at the positions 4 and 6 in the terminal non-reducing end. Its condensa- tion with two mols of the glycosyl bromide of p-D-Gal-(l+ 4)-&~-GlcNAc- (1 -+ 2)-~-Glc (1 17) led to formation of the nonahexosyl unit (119) in 60% yield.'22 Alternatively three disaccharide blocks could be linked successively and gave another cell surface glycan hexa~accharide.'~~ A stepwise strategy yielded a pentasac- charide part of the exocellular P-D-(1-+ 2)-glucan of Agrobucteriurn turnef~ciens.'~~ Block syntheses have been also employed to achieve the preparation of octa- and several penta-saccharides which represent the fundamental structures of the core regions of glycoproteins of the lactosamine type.12' Glycolipids represent another group of natural compounds which stimulated a number of synthetic efforts.Lipid A the carbohydrate part of which is a p-D-GlcNAcyl-( 1 -+ 6)-a-~-GlcNAcyl disaccharide phosphorylated in 1 and 4' carries long aliphatic chains on the amino-functions and the 3,3-hydroxy-groups are esterified similarly (120).A survey of compounds of the type found in the hydro- phobic region of many endotoxins has been given by Szabo et who also described one of the complex multi-step syntheses of a compound with additional hydroxylated aliphatic ~ide-chains.'~~ Lipid A analogues were prepared by Anderson et and Warren et synthesized similar tri- and tetra-saccharide compounds. Other approaches to lipid A precursors were disclosed by Japane~e'~' and Dutch 122 T. Ogawa T. Kitajima and T. Nukada Carbohydr. Rex 1983 123 C8. 123 T. Ogawa T. Nukada and T. Kitajima Carbohydr. Rex 1983 123 C12. 124 T. Ogawa and Y.Takanashi Curbohydr. Res. 1983 123 C16. 125 H.Paulsen and R. Lebuhn Carbohydr. Res. 1984 125,21; 130 85. 126 D. Charon C. Diolez M. Mondange S. R. Sarfati L. Szabo P. Szabo and F. Trigalo ACSSymp. Ser. 1983 231 301. 127 C. Diolez M. Mondange 8. R. Sarfati L. Szabo and P. Szabo J. Chem. SOC.,Perkin Trans. 1,1984,275. 128 L. Anderson and M. A. Nashed ACS Symp. Ser. 1983 231 255. 129 C. D. Warren M. L. Milat C. Auge and R. W. Jeanloz Carbohydr. Rds. 1984 126 61. 130 M. Imoto H. Yoshimura M. Yamamoto T. Shimamoto S. Kusumoto and T. Shiba Tetrahedron Lett. 1984,25 2667. 332 J. 7hiem group^,'^' and recently the corresponding structures and syntheses of the new lipids X and Y from E. coli mutants were de~cribed.'~~ The synthesis of fragments of bacterial polysaccharides and their application for the preparation of synthetic antigens has been reviewed.'33 Novel di- tri- and tetra-saccharides which represent the repeating units of the 0-specific side-chains of lipopolysaccharides from certain serotypes of Shigellu flexneri and Salmonella fyphimurium were published by a number of groups.134-137 Similarly the synthesis of blood-group antigens remains an area of high activity and tri- and tetra-saccharide haptens related to the asialo-form of gangliosides GM2 and GM1 have been In a block synthesis a hexasaccharide the p(1+ 3)-linked trimer of N-acetyl lactosamine useful as a potent inhibitor of anti-i-antibodies was pre- pared.'40 Other classical approaches were employed to synthesize a trisaccharide which represents the human blood-group PI-antigenic determir~ant'~' and a tetrasac- charide determinant of ABH type 1 and Leb blood-group antigens.'42 A preliminary report about the total synthesis of the pentasaccharide fragment of the polyanionic polysaccharide heparin appeared.'43 Starting with a glucofuranose derivative a number of steps led to the selectively blocked L-ido-orthoester (121) which by orthoester glycosylation with the 4-OH unprotected glucosamine derivative (122) led to the a-L-ido-uronic acid ester-( 1 -+ 4)-glucosaminide (123).In a second sequence the glucuronic acid ester bromide (124) could be linked to the 1,6-anhydro- 2-azido-P- ~-glucopyranose derivative (125) and gave the corresponding disac- charide which was further transformed into the glycosyl bromide (126).Glycosyla- tion of the protected a-L-IdoUA-( 1 -+ ~)-D-G~cNH, block (123) with this bromide of the protected P-D-G~cUA-( 1-+ ~)-D-G~cNH, precursor (126) gave the tetrasac- charide species (127). In the final step this was glycosylated in the 4""-position with the 2-azidoglucopyranosyl bromide (128) and gave a pentasaccharide. Following a series of deprotection and selective sulphation steps the derivative of ~-D-G~cNH,- (1 -+ 4)-P-D-GlcUA-( 1 -+ 4)-a-D-GlcNHZ-( 1 -+ 4)-a-~-IdoUA-( 1--* 4)-D-GlcNHZ (129) which represents the central structural element supposed to bind to antithrom- bin I11 was obtained. In fact this showed a high affinity to antithrombin I11 comparable to that of heparin itself whereas in contrast a trisaccharide precursor'4 exhibited none of these properties.13' C. A. A. van Boeckel J. P. G. Hermans P. Westerduin J. J. Oltvoort G. A. van der Marel and J. H. van Boom J.R. Neth. Chem. SOC.,1983 102 438. 13* S. Kusumoto M. Yamamoto and T. Shiba Tetrahedron Lett. 1984 25 3727. 133 N.K. Kochetkov Atre Appl. Chem. 1984 56 923. 134 K. Bock and M. Meldal Acta Chem. Scand. Ser. B. 1983 37 629; 775. 135 P. J. Garegg T. Norberg P. Konradsson and S. C. T. Svensson Carbohydr. Res. 1983 122 165. 136 H. Paulsen and W. Kutschker Carbohydr. Rex 1983 120 25. 137 D. R. Bundle M. A. J. Gidney S. Josephson and H. P. Wessel ACSSymp. Ser. 1983 231 49. 138 S. Sabesan and R. U. Lemieux Can. J. Chem. 1984,62 644. 139 H. P. Wessel T. Iversen and D. R. Bundle Carbohydr. Res. 1984 130 5.140 J. Alain and A. Veyrieres Tetrahedron Lett. 1983 24 5223. 141 P. H. A. Zollo J. C. Jaquinet and P. Sinay Carbohydr. Res. 1983 122 201. 142 N. V. Bovin and A. Ya. Khorlin Bioorg. Khim 1984 10 853. (Chem. Abstr. 1984 101 211 588q). 143 P. Sinay J. C. Jaquinet M. Petitou P. Duchaussoy I. Lederman Y. Choay and G. Tom Carbohydr. Rex 1984 132 C5. 144 J. C. Jaquinet M. Petitou P. Duchaussoy I. Lederman Y. Choay G. Tom and P. Sinay Carbohydr. Res. 1984 130 221. 333 Carbohydrates -==% 040~~1 Meooc&o Me + Ho&oBn I II CICH,C-0 BnO 0 (121) (1 22) ,OAc MeOOCw* 4' BnO HO OAc OBn COOMe ClCH2-!O* + BnO Bn Br HO N3 \O Ac MmcJyo* NH OBn OAc HO=:Fo COOBn BnO + OBn w*+ 0SO;Na' OH (129) N HSOiNa' 334 J.Thiem Much interest centred on the very ambitious syntheses of oligosaccharides involv- ing higher carbon sugars like the C,(KDO) or the C,(NANA) compounds. The basic chemistry and biological background of 3-deoxy-~-manno-2-octulosonic acid (KDO) has been discussed by U~~ger'~' These previously and enriched re~ent1y.l~~ groups and the one of Paulsen partly in collaboration have accomplished very impressive synthetic achievements involving the KDO molecule. The former have succeeded in stereoselectively condensing a ribofuranose via its bromide and pro- moting it by silver triflate to the KDO molecule.147 Similarly a ribo-ribo-KDO trisaccharide was prepared which constitutes the repeating unit of the E. coli capsular polysaccharide.The KDO glycosyl bromide could be used for the glycosylation of a 3-OH free glucosamine precursor and depending on the solvent by mercury ion promotion either the pure a-or the alpmixture of the KDO-(2 -+ 3)-~-GlcNAc derivative were obtained.14* Along corresponding lines the KDO bromide could be condensed to the 3'-OH unprotected P,1-+ 6-linked glucosamine dimer thus furnish- ing a-KDO-(2 +3)-p-~-GlcNAc-( 1.+6)-~-GlcNAc.'~* In addition to these conventional glycosylation procedures another stereo-controlled approach for a KDO disaccharide has been published'49 which makes use of the intramolecular mercury-cyclization previously applied in a number of other projects150 and by Mukaiyama et aLIS1in carbohydrate synthesis. Here 2,3-di-0- benzyl-D-mannose ( 130) was converted into the aldehydo-D-mannose derivative (131) which constitutes the precursor of the KDO unit.The diastereomeric mixture of the 6-0-phosphonogluco-compound (132) obtained by an insertion reaction of methyl diazo( dimethy1phosphono)acetate into the 6-OH unprotected saccharide was converted into the anion. By Horner-Emmons condensation with (131) the (E/Z) mixture of the enol ethers (133) and (134) was obtained (85% ; E/Z ratio 3 :2) and separated chromatographically. The mercury-cyclization with Hg( OCOCF3)* of the (E)-isomer afforded regio- and stereo-specifically the 3-chloromercury glycoside (85% ) which after demercuration hydrogenolysis and saponification afforded the pure P-linked disaccharide P-KDO-(2 -.+ 6)-D-Glc (135).Similarly the correspond- ing treatment of the (2)-isomer gave the a-KDO-(2 .+6)-D-Glc anomer (136) exclusively. Along a similar approach P-KDO(2 +3)-~-GalNAc which represents the repeating unit of the K antigen of Neisseria meningitis 29e was synthesized from the (E)-enol ether precursor which surprisingly was the only product obtained in that Horner-Emmons c~ndensation.'~~ Further use of this approach should demon- strate its compatibility with the classical procedure with respect to anomeric purity and the overall preparative value. The other highly complex monosaccharide N-acetylneuraminic acid (NANA) the best known member of the sialic acid family and a frequent component of glycoconjugates in the terminal positions of glycan chains has been subject to a 145 F.M. Unger Adu. Curbohydr. Chem. Biochem. 1981,38 323. 146 P. Waldstaetten R. Christian G. Schulz F. M. Unger P. Kosma C. Kratky and H. Paulsen ACS Symp. Ser. 1983 231 121. 147 P. Kosma G. Schulz and F. M. Unger Curbohydr. Res. 1984 132 261. 148 H. Paulsen Y. Hayauchi and F. M. Unger Liehigs Ann. Chem. 1984 1270; 1268. 149 F. Paquet and P. Sinay J. Am. Chem. SOC.,1984 106 8313. 150 P. A. Bartlett in 'Asymmetric Synthesis' ed. J. D. Morrison Vol. 3 Academic Press New York 1984 Chapter 6. 151 K. Suzuki and T. Mukaiyama Chem. Lett. 1982 683. Carbohydrates 335 COOMe OMe \ 11 / CH-P / \ B&$,:C OMe + Bnoh /OH -9 BnO BnO CHO OBn OMe * (133) (15):R' = Q BnO*R R2 (134) (2):R' = COOMe (E) + (2) R2 = Q 4 H"H;q'OH HTy+..HO HO CO;N a+ HO (135) (136) Ho OMe number of glycosylation reactions. The syntheses start with the NANA 2-chloride which e.g. by promotion with silver triflate can be glycosylated with the 6-OH free methyl 2,3,4-tri-O-benzyl-~-~-glucopyranoside to give a mixture of a-and p-NANA-(2 +6)-~-Glcin 50% yield (a:p = 1:4).152 The NANA trisaccharide a//?-NANA-(2 -+ 6)-/?-~-Gal-( 1 -* 4)-~-GlcNAc was correspondingly obtained from the same chloride and the lactosamine precursor in 22% a-and 23% p-yield.lS3 Ogawa et al.ls4 published again only short communications on the preparation of the glycoprotein component a/P-NANA-(2 4 6)-P-D-Gal-(1+4)-p-~-GlcNAc-(1 -+ 2)-D-Glc and that of a cell surface glycan a-NANA-(2 -* 8)-a-NANA-(2-* 1)-gly~erine."~An approach similar to the one outlined above for KDO-disac- ~harides'~~ was applied successfully to the stereocontrolled preparation of a-and &NANA-(2 +6)-D-Glc which determines the steric outcome of the reaction by a Horner- Emmons process rather than a terminal glycosylation step.'56 152 H.H. Brandstaetter and E. Zbiral Monatsh. Chem 1983 114 1247. 153 H. Paulsen and H. Tietz Carbohydr. Res. 1984 125 47. 154 T. Kitajima M. Sugimoto T. Nukada and T. Ogawa Carbohydr. Res. 1984 127 C1. 155 T. Ogawa and M. Sugimoto Carbohydr. Res. 1984 128 C1. 156 F. Paquet and P. Sinay Tetrahedron Left. 1984 25 3071. 336 J. Thiem General Oligosaccharide Synthesis and Further Topics.-Modification reactions of abundantly available oligosaccharides or their lower fragments have been previously used and worked out predominantly by Hough et al.e.g. with sucrose trehalose and also cellobiose and lactose. Some newer results also make use of the intact interglycosidic linkage and modify the cellobiose molecule starting with the glycal or the 1,6-anhydro-derivative into 2-azido-2-deoxy derivatives. 157 Chitin represents another natural product which may be used as cheap starting material for large scale syntheses. Its selective degradation by acetolysis led after chromatographic separation to chitotetraose peracetate the glycosyl chloride of which was conven- tionally glycosylated with 7-hydroxy-4-methyl-coumarin sodium salt to provide a tetrasaccharide glycoside useful as fluorometric assay for ly~ozyrne.'~~ Oligodeoxyoligosaccharide syntheses have progressed by application of the N-iodosuccinimide procedure employing glycals which e.g.in a block condensatim type approach yielded a lanatoside tetrasaccharide. l9 This approach a number of other methods the general problem of stereospecific 2-deoxy-glycoside syntheses and some further specific application in the aureolic acid synthesis were concisely reviewed.22 Oligomannosides di-to penta-saccharides e.g. a-D-Man-(1-* 3)-a-D-Man-(1-P 2)-a-D-Man-( 1 -+ 2)-a-D-Man were synthesized following the classical approach which as expected gave the a-glycosides excl~sively.'~~ Silver triflate promoted sequential condensation of glycosyl chloride blocked at C-2 with a non-participatihg residue led to formation of the heptasaccharide (137) which glucan fragment was said to be useful as a neoplasm inhibitor.16' In approaching a rather simple a,1-* 6-linked glucan target molecule Nicolaou et al.claimed the development of a 'practical oligosaccharide ~ynthesis'.'~ In their main concept the repetitive and block-based operation needed a variation of OH- a-~-Glcp-(l+ 3)-a-D-Glcp-(l+ 6)-a-~-Glcp-(l+ 6) \ D-Glc / 3)-c~-~-Glcp-(l+ c~-~-Gl~p-(l+ 6)-a-~-Glp-(l+3) (137) OTBDMS BnO BnO BnO R X n = 0,2,4 (138) (139) TBDMS TBDMS a-SPh F (141) (140) H a-SPh 157 T. K. M. Shing and A. S. Perlin Curbohydr. Res. 1984 130 65. 158 T. Inaba T. Ohgushi Y. Iga and E.Hasegawa Chem. Phurm. Bull. 1984,32 1597. 159 Institute of Physical and Chemical Research Sapporo Breweries Ltd. Jpn. Kokai Tokkyo Koho JP 59 36 691 and 59 36 690 (Chem Abstr. 1984 102 7028y and 70292). 160 Institute of Physical and Chemical Research Jpn. Kokai Tokkyo Koho JP 59 33 301 (Chem. Abstr. 1984 101 19 2397h). Carbohydra tes protecting groups in the reducing or aglyconic carbohydrate moiety. More important was the activation at the anomeric centre using phenylthioglycosides as stable precursors and glycosyl fluorides easily obtained therefrom by treatment with NBS and DAST as the reactive compound to be glycosylated. For instance the glucopyranoside (138) could be transformed (NBS DAST) into the a/P-glucosyl fluoride mixture (139).Simultaneously the TBDMS group could be removed and gave the aglyconic glucose derivative (140). Condensation of (139) and (140) using Mukaiyama's conditions (silver perchlorate stannic dichloride in dry ether) fur- nished the a,1+ 6-linked disaccharide (141;n = 0) in 75% yield. Similar treatment of this disaccharide as above gave the two fragments which could be condensed to a tetrasaccharide species (141; n = 2 70% yield) and the hexasaccharide (141; n = 4 66%) was obtained after further attachment of another disaccharide unit. Based on glucose the preparation of the starting phenylthio-glycoside (138) required eight (more or less) conventional steps (no yield mentioned) and the hexasaccharide synthesis gave an overall yield of approximately 26% for the outlined further eight steps.The practicality of this concept is not convincing and other well-established approaches may be favoured particularly if syntheses of more complex hetero- oligosaccharide targets than the ones mentioned above are desired. There was a previous report of an enzymically promoted condensation of a trisaccharide species to carminomycinone forming a class I1 anthracycline deriva- tive.16' A novel patent now describes the uncatalysed condensation of 2-hydroxy- aclavinone (142) and the trisaccharide (143;QH) in the presence of acetic anhydride in THF to give the class I1 anthracycline (144) which shows marked effects in the treatment of tumour growth.'62 Following classical procedures the first leaf-move- ment factor from Mimosa pudica L.consisting of a a,1+ 2-linked apiofuranosyl- glucopyranose P-glucosylated with gentisic acid (145) has been ~ynthesized.'~~ (142) Q = H (144) Q = (143) /OH COOH I I (145) H0 OH 161 A. Yoshimoto Y. Matsuzawa Y. Matsushita T. Oki T. Takeuchi and H. Umezawa J. Antibiotics 1981,34 1492. 162 Sanraku-Ocean Co. Ltd. Jpn. Kokai Tokkyo Koho JP 59 36 641 (Chem Abstr. 1984 101 38 780w). 163 P. Hettinger and H. Schildknecht Liebigs Ann. Chem. 1984 1230. 338 J. Thiem In the aminoglycoside field the groups of Szarek et uZ.164*165 and of Tatsuta et uL16*166 have accomplished syntheses of the very complex bicyclic amino-octodial- dose which together with 2-deoxy-streptamine represent the basic structures of the aminoglycoside antibiotics apramycin (155) and saccharocin (156).Tatsuta et uZ.16 started their total synthesis with the aminoglycoside antibiotic neamine (146) by preferential N-benzo yloxycarbonylation further N-tosylation and 0-cyclohexyl-idenation. The 6’-amino-group was N-methylated and the N-oxide (147) formed. On treatment with benzoyl chloride and Hunig’s base the 6’-aldehyde (148) was obtained which by reaction with allylmagnesium chloride gave a mixture (149) of the (6’s) and (6’R) alcohols in 37 and 41% yield respectively. Cleavage of the olefin-function with osmium tetraoxide and sodium metaperiodate resulted after acid removal of the alkylidene blocking groups in formation of the octodialdose derivative (150). 0-Tosylation and base treatment gave after acetylation the glycal (151).This diastereomer was also obtained from the (6’R) adduct (149) viu a mesylation-inversion sequence and further treatment as above. Azidonitration of the glycal gave both anomers at C-8’ but the 7’-azido groups was introduced stereospecifically (7’sonly). After a number of steps the 3’,6’-dimesylate was used to generate the 3’-deoxy function via a 3’-chloride and tributyltin hydride reduction and following epimerization at C-6’ the subsequent carbamate formation led to the derivative ( 152). By standard deblocking procedures this compound gave directly methyl P-aprosaminide. The unblocked N-carbobenzoxylated amino-octodialdose derivative could be glycosylated employing Mukaiyama’s conditions and the per- benzylated 4-azido-4-deoxy-glucopyranosyl( 153) or the normal glucopyranosyl fluorides (154).Further steps completed the syntheses of apramycin (155) and saccharocin (156). There are again some reports on the large-scale enzymic synthesis of oligosac- charides an area which may supplement and also compete favourably with the existing synthetic methodology. Employing sucrose synthetase 1-deoxy-1 -fluoro-D- fructose was accepted as a substrate and with UDP-glucose transformed into 1’- deoxy-1’-fluorosucrose in 60-80% yield.53 Only 20% yield was obtained in the preparation of P-D-Gal-( 1 + 6)-~-GalNAc using immobilized P-galact~sidase.’~~ Buckwheat a-glucosidase used maltose and sucrose to form one tetra- and two tri-saccharides named erlose esculose and theandrose (maltosylsucrose and two isomeric glucosylsucroses).An enzymic synthesis on an a-chymotrypsin-sensitive polymer was published by Zehavi et uZ.169Thus the ~-N-benzoyloxycarbonyl-2-phenylalanine glycoside of peracetylated cellobiose was elaborated into a polymer and the cellobiose residue represented the acceptor for a D-galactosyltransferase reaction using UDP-galactose. The resulting polymeric trisaccharide was digested with a-chymotrypsin and released the desired trisaccharide species p-D-Gal-( 1 -+ 4)-p-~-Glc-( 1 -+ 4)-~-Glc. Employing a system of five immobilized enzymes and 164 0.Martin and W. A. Szarek J. Chem. SOC.,Chem. Commun. 1983 926. 165 0. Martin and W. A. Szarek Carbohydr. Res. 1984 130 195. K. Tatsuta K. Akimoto H. Takahashi T.Hamatsu M. Annaka and M. Kinoshita Tetrahedron Lett. 1983 24 4861. 167 L. Hedbys P. 0.Larsson K. Mosbach and S. Svensson Biochem. Biophys. Res. Commun. 1984,123,s. 168 S. Chiba Y. Asada-Komatsu A. Kimura and K. Kawashima Agric. BioL Chem. 1984,48 1173. 169 U. Zehavi and M. Herchman Carbohydr. Rex 1984 133 339. Carbohydrates HO HO CX-0 OH 0 CY-0 MeO% BnO NHTs R *&F BnO (153) R’= N (154) R’= OBn (155) R = NH HO (156) R = OH OH 340 J. 7'hiern partially based on previous reports for the synthesis of N-acetyl-lactosamine by Whitesides et ~1.'~'the group of David et published quite simple preparations of the trisaccharides p-D-Gal-( 1 + 4)-p-~-GlcNAc-( 1+ 6)-~-Gal and P-D-Gal- (1 -P 4)-p-~-GlcNAc-( 1-.3)-D-Gal in very good yields. Some polymeric compounds with a carbohydrate base structure should be of interest. Acrylamide was radically copolymerized with a trisaccharide allyl glycoside which led to a linear polyacrylamide copolymer with carbohydrate branches (approximately 30% sugar content). This treatment of the allyl glycoside of p-D-Man-(1 -+ 4)-a-~-Rha-( 1+ 3)-~-Gal led to a polymer with a molecular weight larger than 100kD which functioned as a synthetic antigen and showed a group specificity E of salmonella species. 172 Another interesting project centred on the modification of natural polysaccharides ; e.g. chitosan was reductively alkylated with an aldose and sodium cyanoborohydride and gave aminoalkylated polysaccharides of comb- or tree-like structure.The derivative of chitosan with N-(sorbitol-1-yl) branches (stemming from glucose units) showed novel interesting properties for instance water ~olubility.'~~ 4 Carbohydrates as Chiral Templates The synthesis of challenging chiral structures starting from readily available or modifiable chiral precursors frequently referred to as 'chiral template approach' enjoys ever increasing interest in carbohydrate chemistry. Not surprisingly this area has attracted scientists both from the original carbohydrate field as well as from other areas of organic chemistry. The selection presented here has been deliberately divided into a subchapter on C-glycosides and another one on the use of carbohy- drate precursors or selectively modified derivatives thereof as chiral building units for the construction of various chiral entities.C-G1ycosides.-The detection of C-nucleosides has promoted quite intensive studies and their synthesis was reviewed re~ent1y.l~~ A number of procedures have been developed in this area and later adopted in the carbohydrate field. In recent years however many new approaches have also appeared some of which will be discussed. A general method was described for the conversion of aldono-lactones into the methyl aldulosonates a group of compounds which are represented by such impor- tant members as KDO or NANA (see above).'75 Treatment of e.g.the lactone (157) with tris(methy1thio)methyl-lithium as acyl equivalent at -78 "C and subsequent work-up with HgO-HgCI2 in methanol gave the corresponding methyl heptulosonate (158) in 40% yield.This method with a relatively good yield should be of interest for multi-step syntheses the more so because it was compatible with a number of base-stable protecting group^."^ In another short approach aldono-lactones could be transformed into methylene compounds by treatment with the titanium carbene 170 C. H. Wong S. L. Haynie and G. M. Whitesides J. Org. Chem. 1982 47 5416. 171 C. AugC S. David C. Mathieu and C. Gautheron Tetrahedron Lett. 1984 25 1467. 172 A. Ya. Chernyak A. B. Levinskii B. A. Dmitriev and N. K. Kochetkov Curbohydr. Res, 1984,128,269. 173 M. Yalpani and L. D. Hall Mucromolecules 1984 17 272. 174 J. G. Buchanan Prog. Chem. Org. Nut.Prod. 1983 44 243. 175 J. E. Hengeveld K. Grief J. Tadanier C. M. Lee D. Riley and P. A. Lartey Tetrahedron Lett. 1984 25 4075. Carbohydrates BnO /OBn (MeS),CH Hg2+ -__* BuLi BnO (157) BnO Bnoko OH (158) complex (159) in good yields. These exocyclic glycals could be further elaborated in a number of ways and also into C-glyc~sides.'~~ Baldwin et demonstrated that trapping of the C-1 radical formed by treatment of phenyl 2,3,4,6-tetra-0-acetyl-l-thio-~-~-glucopyranoside and triphenylstannane in refluxing toluene with methyl acrylate gave 40% of the a-(2-methoxycar-bony1)ethyl-glucoside derivative and 10% of a double substituted product a-(2,4-bis-methoxycarbony1)butyl-glucoside.Keck et showed that treatment of thio-phenyl glycosides with methallyl-tri-n-butylstannaneusing a photochemical initi- ation gave the C-allylated derivative in approximately 90% yield with an alp-ratio of 92 :8.However the same reaction performed in the presence of tri-n-butylstannanyl triflate with heating gave in 95% yield the p-anomer virtually exclusively. In contrast to other versions of the Claisen rearrangement the Eschenmoser variation (N,N-dimethylacetamide dimethyl acetal reflux) applied to the alcohol (160) gave the a-glycosylated N,N-dimethylacetamide (161) in 85% yield.'79 On reduction with Li[(OEt),AlH] only minor amounts of the a-aldehyde (162) resulted because of prevailing anomerization to the &derivative (163). In another approach \ R (161) R =NMe (162) R =H Wittig extension of the 4,6- 0-ethylidene-D-glucose (1 64) gave the trans-oct-2-enoate (165) which on cyclization with dilute base after one hour yielded a 1 :1 mixture of the a-and &derivatives (166).By further base treatment the p-anomer of (166) was obtained exclusively which by Cohen-Tipson reductive elimination gave the olefinic ester derivative (167).18' 176 C. S. Wilcox G. W. Long and H. Suh Tetrahedron Lett. 1984 25 395. 177 R. M. Adlington J. E. Baldwin A. Basak and R. P. Kozyrod J. Chem. SOC.,Chem Commun 1983,944. 178 G. E. Keck E. J. Enholm and D. F. Kachensky Tetrahedron Lett. 1984 25 1867. 179 D. B. Tulshian and B. Fraser-Reid J. Org. Chem 1984,49 518. 180 R. D. Dawe and B. Fraser-Reid J. Org. Chem. 1984 49 522. 342 J.Thiem I A most interesting paper of Sinay et reported the synthesis of the first C-disaccharide an interglycosidic oxygen isostere of a cellobioside. The straightfor- ward preparation started with a transformation of the 6-aldehydo-glucose derivative (168) into the dibromo-olefin (169). This on treatment with BuLi at -50 "C in situ provided the acetylenic anion which added to the perbenzylated lactone (170) and gave the hemiacetal (171). Its stereospecific reduction with Et,SiH BF3-Et20 led to the P-glycoside (1 72) exclusively and by a final reduction deblocking and saturation occurred to yield the product (173) which may be of interest in studies of sugar metabolism or as an enzyme inhibitor. In course of the preparation of the vineomycin B2 aglycone which features a P-C-glycosidic bond of a 2,6-dideoxy-arabino sugar to an unsymmetrical anthraquinone Danishefsky et aZ.'82reacted a 2-triethylsilyloxy- 173-pentadiene with + Bnoh BnO BnO OMe BnO 0 (168) R = CHO (170) (169) R = CH=CBr BnO&e!&J/o);n -H\&H2Cd&j OH BnO Ho OMe 0 (1 73) Bn X (171) X = OH (172) X = H 181 D.Rouzaud and P. Sinay 1. Chem Soe Chem. Commun. 1983 1353. 182 S. Danishefsky B. J. Uang and G. Quallich J. Am. Chem. Soc. 1984 106 2453. Carbohydrates 343 an aromatic aldehyde in the presence of Eu(fod)3. This hetero Diels-Alder reaction catalysed by the lanthanide complex gave the correct &bond of the (*) C-aryl glycoside in 92% yield. Reaction of 6-aldehydo-1,2;3,4-di-O-isopropylidene-c-D-galactopyranose with the diene mixture of 1- 0-benzoyl-2- O-trimethylsilyl-4-0- methylbutadiene in the presence of boron trifluoride at -78°C in ether provided 62% of a single cyclocondensation product (174) with high diastereofacial selectivity which may be considered a carbon-carbon linked disaccharide prec~rsor.'~~ 0 The cycloaddition adducts of the isoquinolinium salt (175) with the pyranose or furanose glycals (176) or (177) were opened to the amino-aldehyde derivatives which by abstraction of 2,4-dinitroaniline aromatized to the same sugar-substituted naph- thaldehyde (178) in approximately 60% yield.lS4A number of C-glycosides could be obtained by employing glycosyl fluorides as substrates Lewis acids as catalysts and e.g.ally1 silanes trialkyl aluminium or trimethylsilyl cyanide as nucleophiles ; in generally high yields the predominant formation of the a-compounds was repor- ted.14 Reetz et aZ.'*' published a mild and regiospecific method which could be used (175) CHO 183 S. Danishefsky C. J. Maring M. R. Barbachyn and B. E. Segmuller J. Org. Chem. 1984,49,4564. 184 R. W. Franck and R. B. Gupta J. Chem. SOC Chem Commun. 1984 761. 185 M. T. Reek and H. Muller-Starke Liebigs Ann. Chem 1983 1726. 344 J. Thiem for the synthesis of C-glycosides. Glycosyl halides or even acetates on treatment with silyl enolether 0-silyl ketene acetals or bis-silylacyloins in the presence of zinc halides gave a-alkyl-oxyalkylated carbonyl compounds. Similarly 1-0-acetyl- 2,3,5-tri-0-benzoyl-P-~-ribofuranosewith silyl enol ether and SnCl catalysis led to the formation of C-ribosides.'86 By use of 2,3,5-tri-O-benzyl-P-~-ribofuranosyl fluoride and the silyl enol ether of acetone under BF,.Et,O catalysis the alpmixture of the corresponding octulose was ~btained.'~~ Another grouplS8 demonstrated the direct C-allylation of a L-Z~XO acetate with allyltrimethylsilane in acetonitrile using BF3*Et20 catalysis in high yield and almost exclusive a-anomer formation.A large number of sugar substrates like glycosyl chlorides but also the more stable methyl glycosides could be C-allylated with various allylsilanes as nucleophiles. Catalysed by trimethylsilyl trifluoromethanesul- phonate or iodotrimethylsilane the reaction gave mainly a-anomer and in good ~ie1ds.l'~ Similarly Mukaiyama et uZ.'~' showed that the a-C-riboside was formed in excellent yields on treatment of ribopyranose acetate with silylenolethers or allylsilanes or trimethylsilyl cyanide ;the reaction was catalysed by trityl perchlorate.2-Acyloxy-3-keto glycals undergo stereospecific reaction with silyl enol-ethers with titanium chloride as catalyst to give P-C-glycosides of 4-deoxy-hex-3-en-2-ulose derivatives by way of a Michael-type addition.'" A very promising finding was disclosed by Sinay et ~2Z.l~~ who on treatment of the glucopyranoside (179) with lithium naphthalenide in THF at -78 to +5 "C generated by reductive lithiation and subsequent elimination (according to Ireland's procedure) the tri-0-benzyl glucal (180).After addition of gaseous HC1 the labile intermediate glycosyl .chloride was again treated with LiC10H7 in THF at -78 "C and in another reductive lithiation gave the first glycopyranosyl lithium derivative (181). By quenching with anisaldehyde the a-C-glycoside (182) was obtained as a 4:1 diastereomeric mixture. In a mechanistically somewhat peculiar reaction a-acetobromoglucose and diacetylphloroglucinol with sodium in methanol gave after chromatographic purifi- cation a 9% yield of the p-C-aryl-substituted glucose.'93 Even the classical appro ache^'^^ for the preparation of C-glycosides have been employed for the synthesis of P-C-ben~yl'~~ P-C-butyl xylopyrano~ides,'~~ or compounds which showed remarkable activities against certain tumours.Reaction of 4,6-0-benzyl-idene-3- 0-mesyl-D-allal with several alkyl Grignard reagents gave stereospecifically the P-C-allylated hex-2-enopyranosides in good ~ie1ds.l~~ By mild reaction of la6 Y.S. Yokayama T. Inoue and I. Kuwajima Bull. SOC.Chem Jpn. 1984,57 553. 187 Y. Araki K. Watanabe F. Kuan K. Itoh N. Kobayashi and Y. Ishido Carbohydr. Res. 1984,127 C5. 188 A. P.Kozikowski and K. L. Sorgi Tetrahedron Lett. 1984,25 2085. 189 A. Hosomi Y. Sakata and H. Sakurai Tetrahedron Lett. 1984,25 2383. 190 T.Mukaiyama S. Kobayashi and S. Shoda Chem. Lett. 1984 1529. 191 H.Kunz J. Weismuller and B. Muller Tetrahedron Lett. 1984,25 3571. 192 J. M. Lancelin L. Morin-Allory and P. Sinay J. Chem. SOC.,Chem. Commun. 1984 355. 193 H.Obara M.Hattori,and Y. Matsui Chem. Lett. 1984 1039. 194 W. A. Bonner Adv. Carbohydr. Chem 1951,6 251. 195 R. Noyori S. Suzuki M. Okayama K. Sakurai S. Komohara and Y. Ueno (Seikagaku Kogyo Co. Ltd.) USP 4 454 123 (Chem. Abstr. 1984,101 152261~). 196 R. Noyori S. Suzuki and M. Orayama (Seikagaku Kogyo Co.,Ltd.) USP4446312 (Chem. Abstr. 1984,101 91 397c). T.Ogihara and 0. Mitsunobu Tetrahedron Lett. 1983,24 3505. 19' 345 Carbohydrates pen / OBn /OBn hleO CHO BnO & BnO acylated glycals with diethylaluminium cyanide at room temperature the l-cyano- hex-2-enopyranosides were obtained in good yield and an alp-ratio of 2 3.’98 Under reflux the a-derivative predominated as demonstrated before.’79 Synthesis of Chiral Structures from Carbohydrate Precursors.-Comprehensive reviews covering this topic became available some time ago’99-201 and several more have appeared quite re~ently.~~~~~~~ This subchapter is meant to update and discuss recent developments.The highly specific cationic ionophore calcimycin (187) was synthesized from D-glucose. First the two selectively C-methylated and protected units (chirons) (183) and (184) were elaborated and their dithiane condensation gave the open-chain precursor (185). Mercury-mediated dithiane hydrolysis and acid treatment led to the spiroketal(l86) exclusively. In a number of further steps the two heteroaromatic residues could be introduced to give calcimycin (187).204 Another report has described the transformation of D-glucose uia 3,5-dideoxy-1,2-0-isopropylidene-a-D-erythro-hexofuranose as the key intermediate in only four steps into the enan- tiomerically pure 1,7-dioxaspiro[5,5 Jundecane (1 88) which represents the pheromone component of the olive fly (Dacus ole~e).”~ Treatment of the tetrabenzyl gluconolactone (170)with the lithium salt of 1-trimethylsilyloxy-3-butyneand mild acid work-up gave the C-1 alkylated alp-lactol(l89) in 92% yield and on hydrogen- ation over Lindlar catalyst this yielded the cis-olefin (190).This could be cyclized 198 D. S. Grierson M.Bonin H. P. Husson C. Monneret and J. C. Florent Tetrahedron Lett. 1984,25,4645. 199 S. Hanessian Acc. Chem. Res. 1979 8 192. 200 B. Fraser-Reid and R. C. Anderson Fortschr. Chem. Organ. Naturst.1980 39 1. 201 A. Vasella in ‘Modern Synthetic Methods’ ed. R. S. Scheffold Salle and Sauerlander Frankfurt 1980. 202 S. Hannessian ‘Total Synthesis of Natural Products’. The Chiron Approach’ Pergamon Oxford 1983. 203 T. D. Inch Tetrahedron 1984,40 3161. Y. Nakahara A. Fujita and T. Ogawa J. Carbohydr. Chem. 1984 3 487. *05 H. Redlich and W. Francke Angew. Chem. 1984 95 506. 346 J. Thiem TBDMSO f-t-D-Glucose OEE (183) (184) TBDMSO (186) R' = CH20H,R2 = CH20Bz YOOH (187) R' = NHMe R2 0 R' R2= fi with camphosulphonic acid to give the unsaturated model spiroacetal. The alp-ratio was 4:3and acid treatment anomerized the thermodynamically unstable @-derivative to the a-anomer (191).By debenzylation (Na in liquid ammonia) the unblocked 1,7-dioxaspiro[5,5]undecene subunit (192)of avermectin B1,was obtained.206 Irradi- ation of an a-(3-keto-3-pheny1propyl)glycoside led in a Norrish-type I1 cyclization to both diastereomers of the 4-hydroxy-4-phenyl-l,6-dioxaspiro[ 5,6]decane system.'07 There is considerable interest in syntheses of variously substituted anthracyc- hones and reports were given on an application of the Marschalk reaction employ- ing sugars.Leucoquinizarin (193)and the arabinose (194)could be condensed in alkaline solution and after oxidation by air gave the adduct (195).This following selective deblocking and oxidative cleavage could be elaborated into the highly functionalized A-ring anthracyclinone derivative (196).208 The same group published 206 S.Hanessian and A. Ugolini Carbohydr. Rex 1984 130 261. 207 P. Bron L. Cottier and G. Descotes J. Heterocycl. Chem. 1984 21 21. D. J. Mincher G. Shaw and E. DeClerq J. Chem. SOC.,Perkin Trans. 1 1983 613. Carbohydrates @&+ (191) R = Bn (192) R = H(189) R = -=*OH (190) R = moH CHO / \ 0 \ I I / o 0 HO OH (193) 0 i another improved anthracyclinone synthesis employing the xylofuranose as chiral template.209 o-Xylylenes ( 197) derived from 1,2-bis( bromomethy1)benzene by treat- ment with zinc powder and ultrasound irradiation gave rise to cycloaddition with the hex-2-enopyranoside-4-ulose ( 198) forming the carbohydrate adduct in 30-70% yield210 (for comparable cycloadditions see rej 21 1). Further transformation of the exocyclic enol ether function in (199) by solvolysis employing mercury salts gave the highly functionalized hexahydro-naphthacenes or -anthracenes (200).In the difficult stereoselective glycosylation steps of anthracyclines the P-glucoside of the functionalized butadiene (202) was employed for the cycloaddition to the tricyclic oxirane derivative (201).2'2 The cycloadduct was transformed into the 3-ketone (203) and following reduction of the oxirane introduction of the side-chain and Hg"- mediated work-up furnished the intact C-7 p-D-glucosylated anthracycline derivative (204). 209 D. J. Mincher and G. Shaw J. Chem. SOC.,Perkin Trans. 1 1984 1279. 210 S. Chew and R. J. Ferrier J. Chem. SOC.,Chem. Commun.,1984 911. 21 1 R.W. Franck and T. V. John J. Org. Chem 1983,48 3269. 212 R. C. Gupta P. A. Harland and R. J. Stoodley J. Chem. Soc. Chem. Commun.,1983 754. 348 J. Thiem I OBz D-Ribose as well as 1-deoxy-D-ribose were used for the preparations of dioxapros- tacyclin analogues.213 Horton et uL2149215 transformed aldehydo-sugars into a$-unsaturated esters using the Wittig method. Their Diels- Alder condensation with cyclopentadiene at low temperature and Lewis acid catalysis furnished sugar- substituted norbornene derivatives. Following hydroxylation glycol cleavage and reduction optically pure cyclopentane derivatives were obtained which owing to their five chiral centres similar to those in PG F1 may be of interest as prostaglandin synthons. Q= OAc AcO OAc 213 P.Heath J. Mann E. B. Walsh and A. H. Wadsworth J. Chem. SOC.,Perkin Trans. 1 1983 2675. 214 D. Horton T. Machinami Y. Takagi C. W. Bergmann and G. C. Christoph J. Chem. SOC Chem. Commun. 1983 1164. 215 D. Horton T. Machinami and Y. Takagi Curbohydr. Res. 1983 121 135. Carbohydrates 349 A key step in an efficient synthesis of methyl shikimate (207) was the intramolecular Horner- Emmons reaction. Starting with D-mannose the 5-O-triflyl-~-lyxofuranose (205) was obtained and alkylated with sodium trimethylphosphono acetate to give (206) as a 1 :1 diasteromeric mixture. Following hydrogenolysis the lactol anomers on base treatment gave the methyl shikimate (207) obtained after mild acid acetal cleavage.216 OH HOA / (205) R = OTf PO(OMe) (206) R = -CHEk -COOMe The C-3-C-8 and C-9-(2-13 fragments of the 14-membered macrelide antibiotic ~leandomycin~” and the C- 19-C-29 aliphatic sequence of rifamycin S218were both synthesized from D-glUCOSe following the chiron approach.Levoglucosenon (208) furnished an attractive starting material for preparations of (-)-S-multistriatin (209) and also of the (+)-Prelog-Djerassi lactonic acid (210).219 ,,COOH D-Glucose can be transformed into 6-epithienamycin (213) in a fourteen-step synthesis. The conversion centred around the base-mediated azetidinone formation (45%) of (212) from the D-lyxo-compound (211). Further elaboration of the carbon framework by Wittig reaction and along previously published lines gave the desired compound (213).220 A corresponding cyclization with Hunig’s base of the carbamate ester (214) derived from D-glucosamine gave the pyrrolidine (215) in 88% yield.216 G. W. J. Fleet and T. K. M. Shing J. Chem. SOC.,Chem. Commun. 1983 849. 217 S. S. Costa A. Olesker T. T. Ton and G. Lukacs J. Org. Chem. 1984 49 2338. 218 S. Hanessian J. R. Pougny and 1. K. Boessenkool Tetrahedron 1984 40,1289. 219 M. Mori T. Chuman and K. Kato Curbohydr. Res. 1984 129 73. A. Knierzinger and A. Vasella J. Chem. Soc. Chem. Commun. 1984,9. 220 350 J. Thiem -+ 0q COOR Its further transformation led to a synthesis of the carbapenem antibiotic (216; R = p-nitrophenyl) .221 Similarly treatment of the D-glucosamine-derived 2-amino- O-isopropylidene-5-O-tosyl-~-sorbitol 4,6-0-benzylidene-2-deoxy-l,3-with base gave the functionalized chiral pyrrolidine (217).222 Hanessian et uL223have published a stereocontrolled synthesis of the azetidinone (218) a monobactam precursor from D-glucosamine.This nine-step process gave the desired product in 16% overall yield and remains quite attractive even though the C4-C6 fragment of the starting sugar and thus two chiral centres had to be discarded. H3Nb-Me O 'so; (218) -+ HHoH N 2,3-0-Isopropylidene-D-erythroseserved as the precursor for a 10-step enantio- specific preparation of the lactone (219),224 the conversion of which into the pyr- rolizidine alkaloid (+)-retronecine (220)had been previously reported. Considerable interest has been centred on synthetic pathways for the hydroxylated indolizidine alkaloids castanospermine (221) deoxynojirimycin (222) and swainsonine (223) which are potent inhibitors of several carbohydrate processing enzymes.A straight-22 1 M. Miyashita N. Chida and A. Yoshikoshi L Chem. SOC,Chem Commun. 1984 195. 222 C. Morin Tetrahedron Lett. 1984 25 3205. 223 S. Hanessian and S. P. Sahoo Can. J. Chem 1984 62 1400. 224 J. G. Buchanan G. Singh and R. H. Wightman J. Chem. SOC.,Chem. Commun. 1984 1299. Carbohydrates 351 forward D-glucose-based preparation of (+)-deoxynojirimycin (222) and of (+)-castanospermine (221) was published.225 1,5-Dideoxy-l,5-imino-D-mannitol (1-deoxy-mannojirimycin) was synthesized either from D-mannose uia hydrogenation of its 5-azido-derivative or from D-glucose with the key step being nucleophilic 2-0-triflate inversion at C-2.226 Swainsonin (223) could be synthesized from D-mannose2279228 and from ~-glucose.~~~ The control of chirality at off-template sites in carbohydrate-derived syntheses is of concern in some recent contributions.Thus the platinum-catalysed hydrogenation of the a,P-unsaturated ester (224) derived from diacetone glucose in six straightfor- ward steps occurred from the si-si face exclusively most likely owing to the bulky substituent at C-3. The ester (225) or lactone (226) obtained could be further functionalized into the ketolactone (227) which is potentially useful as a sesquiter- pene precursor.230 Another 16-step stereocontrolled synthesis from diacetone glucose furnished the synthon (228) for the hemiacetal moiety (C13-C20) of the polyene macrolide antibiotic amphotericin B in an amazing 20% overall yield.231 Other approaches of the same group to achieve the stereocontrolled functionalization at off-template sites have been discussed (224) R = /=ioH Et0,C (226) R =OQ,H (225) R = Et02CxoH OH /OH Me 0 225 R.C. Bernotas and B. Ganem Tetrahedron Lett. 1984.25 165. 226 G. W. J. Fleet M. J. Gough and T. K. M. Shing Tetrahedron Lett. 1984 25 4029. 227 N. Yasuda H. Tsutsumi and T. Takaya Chem. Lett. 1984 1201. H. A. Mezher L. Hough and A. C. Richardson J. Chem. SOC.,Chem. Commun. 1984 447. G. W. J. Fleet M. J. Gough and P. W. Smith Tetrahedron Lett.1984 25 1853. M. Georges T. F. Tam and B. Fraser-Reid J. Chem. Soc. Chem Commun. 1984 1122. D. Liang H. W. Pauls and B. Fraser-Reid J. Chem. SOC.,Chem Commun.,1984 1123. 228 229 230 231 352 J. Thiem An unusual tetra( tripheny1)phosphine palladium-mediated rearrangement of the 5,6-anhydro- 1,2- O-isopropylidene-a-~-erythro-hex-3-enofuranose (229) led to a E/Z-mixture of the 4,5-unsaturated aldehyde (230). Reduction benzylation and hydrolysis gave the free-sugar intermediate which by aldol condensation yielded a precursor of (-)-pentenomycin I (231).232 A novel preparation of alkyl-branched cyclitols has been achieved by Klemer et aZ.233With strong bases for instance methyl lithium 1,6-anhydro-sugars like the P-D-galacto-derivative (232) undergo abstraction of H-5and in consequence a ring-opening reaction which gives the 5-enolate intermediate (233).This undergoes an intramolecular aldol addition and the ketone (234) adds another nucleophile to give both the diastereomers in the case ofmethyl lithium but a stereoselective addition with Bu"Li. This reaction resembles the procedure previously developed by Ferrier et aL234in which 6-deoxy-hex-5-enopyranosides are transformed into 2-deoxy- inososes induced by Hg" salts in aqueous media. A number of reports have appeared of the latter procedure being successfully used for the synthesis ofvarious cyclitols and aminocyclitols from carbohydrate precursor^.^^^,^^^ / u4 X OH OH (233) (234) X = 0 (235) X = OH,Me 232 S.Achab J. P. Cosson and B. C. Das J. Chem SOC.,Chem Commun. 1984 1040. 233 A. Klemer and M. Kohla Liebigs Ann. Chem 1984 1662. 234 R. J. Femer J. Chem SOC.,Perkin Trans. 1 1979 1455. 235 D. Semeria M. Philippe J. M. Delaumeny A. M. Sepulchre and S. D. Gero Synthesis 1983 710. 236 I. Pelyvas E. Sztariskai and R. Bognar J. Chem. SOC.,Chem Commun. 1984,104.