2 Synthetic methods Part (iii) Enzyme chemistry By A. J. CARNELL Department of Chemistry Robert Robinson Laboratories University of Liverpool Liverpool UK L69 7ZD 1 Introduction The use of enzymes in synthetic chemistry continues to capture the imagination of organic chemists interested in using highly selective catalysts under mild conditions. The range and scope of biotransformations is being extended by the increasing availability of the enzymes and in particular the ability of gene technology to provide the catalysts in larger quantities through overexpression of the proteins in recombinant strains. Several general reviews have appeared in the last year1,2 as well as reviews on specific classes of reaction. The microbial hydrolysis of nitriles and amides3 and the use of epoxide hydrolases from mammalian microbial and other sources has been reviewed.4 Holland has summarized recent advances in predictive active site modelling with an emphasis on cubic space models developed for sulfoxidations aromatic and aliphatic hydroxylation and Baeyer–Villiger oxidation.5 Asymmetric epoxidations using enzymes have been extensively reviewed6 and several reviews summarizing recent advances in the use of aldolases and transketolase in the synthesis of polyfunctional oligosaccharides have appeared.7–9 2 Hydrolytic enzymes Several groups have used hydrolytic enzymes for the desymmmetrization or resolution of enol esters.Duhamel used Candida cylindracea lipase to hydrolyse prochiral dienol diacetates 1 in aqueous or biphasic conditions to a§ord the ketone acetates (S)-2 in good yields (62–80%) and enantiomeric purity ([98%ee).10 Replacement of the methyl group with an allyl group in the substrate where R\CH 2 Ph gave diminished selectivity.Carnell has resolved the racemic enol acetate 3 derived from 4-cyano-4- phenylcyclohexanone 4 with Pseudomonas fluorescens lipase in tetrahydrofuran–nbutanol. 11 Chemical recycling of the prochiral ketone the hydrolysis product allowed isolation of enantiomerically pure (S)-enol ester 3 in 55% yield. Again the selectivity obtained appeared to be dependent on a large di§erence in steric requirement of the two groups attached to the quaternary centre (Scheme 1). The cheap and readily available sugar dulcitol was desymmetrized by first making the diacetonide in order to make the substrate rigid and more hydrophobic and then 39 OAc AcO R Me O AcO R Me OAc Ph CN O Ph CN OAc Ph CN R = CH2CH=CH2 CH2CH=CHCH3 ( E) CH2CH=C(Cl)CH3( E) CH2C=CCH3 CH2Ph 1 ( S)-2 (±)-3 + ( S)-3 100% ee 4 P.fluorescens lipase THF–Bu nOH recycle C. cylindracea lipase Scheme 1 O O R O O Me O O Pri OH Pri CO2Bn OH Me CO2Bn CAL-B MTBE BnOH R = Pri + CAL-B MTBE BnOH R = Me + (±)-5 or 6 ( R)-6 ( S)-5 ( R)-7 ( S)-8 Scheme 2 acylating one terminal hydroxy group with a lipase in vinyl acetate–diethyl ether.12 Enantiocomplementary selectivity was obtained with lipase PS a§ording the (])- acetate and Pseudomonas fluorescens lipase the ([)-product. In another example of reversal of selectivity the a-methylene-b-lactones 5 and 6 were resolved using Candida antartica lipase (fraction B) in methyl tert-butyl ether–BunOH.13 The reversal of enantioselectivity was rationalized by substituent size di§erence i.e.for lactone 6 a-methylene is large and methyl is small and for lactone 5 a-methylene is small and isopropyl is large (Scheme 2). An e¶cient synthesis of 1a,25-(OH) 2 -vitamin D 3 A-ring carbamate derivatives 10 was developed using a two-step chemoenzymatic approach involving enzymatic synthesis of C-5 carbonates followed by reaction with amino derivatives.14 Candida antartica lipase (CAL SP435L Novo Nordisk) in toluene was found to be the best enzyme for di§erentiating the C-3 and C-5 positions of compound 9 catalysing the formation of the 5-vinyloxycarbonyl derivatives in quantitative yield. The best alkoxycarbonyl donor was the oxime (R2\Me 2 C––N). Enzyme-catalysed Cbz-protection was acheived in 85% yield under the same conditions a potentially useful mild method of introduction (Scheme 3).Waldemann has reported an extremely mild and e¶cient way to remove phenylacetyl amino protecting groups used in oligodeoxyribonucleotide synthesis using Penicillin G acylase either in solution or on solid support.15 Chemists from Lonza have resolved piperazine 2-carboxamide to give the (S)-acid a key intermediate in the synthesis of the Merk HIV protease inhibitor Crixivan in 41% yield 99.4% ee using Klebsiella DSM 9174. The organism can grow on the substrate with a substrate concentration of 22 g l~1 on a large scale.16 40 A. J. Carnell HO OH O OH NH O R3 1. R1OCO2R2 enzyme solvent R3NH2 = ammonia amines amino alcohols diamines and amino acids 9 10 5 5 R1 = Me CH2 CH Ph; R2 = Me2C N CH2 CH 2.R3NH2 Scheme 3 R C N C N X M M hydration hydration Scheme 4 3 Nitrile and epoxide hydrolysis A series of papers by Meth-Cohn reported a detailed and systematic study of the synthetic potential for selective hydrolysis and possible mechanism of a nitrile hydratase system present in Rhodococcus rhodochrous AJ270.17–19 Aliphatic dinitriles NC(CH 2 )nCN such as succinonitrile and glutaronitrile were hydrolysed selectively giving moderate yields of the monoacids. Generally dinitriles with n\4 gave selective hydrolysis however upon extended reaction times the monoacids are metabolized resulting in low yields. Adiponitrile (n\4) gave the monacid in 40% yield if the reaction was stopped after 3 h. Dinitriles with n[4 generally gave no selectivity a§ording diacids and in good yield (ca.90%) for longer chains (n\7 or 8). A series of a,x-dinitriles NC(CH 2 )nX(CH 2 )nCN (X\O S NR) were examined for regioselective hydrolysis. For X\O the chain length n\4 was optimal giving 72% yield of the monoacid after 2 h. A b or c-oxygen (n\2 or 3) was also e§ective whereas five methylene groups between X and the nitrile gave no selectivity. For X\S a c-sulfur allowed optimal control the b-analogue showed selectivity but not the d-analogue. Good selectivity was found for a b or c-nitrogen although the precise dependence on chain length was not defined. The observed selectivity was explained by invoking bidentate complexation to an active site metal ion (iron or cobalt) by the nitrile and the heteroatom (X or CO 2 ~) which deactivates hydrolysis of the coordinated nitrile by the pyrroloquinone quinone cofactor (Scheme 4).In a parallel study on dintriles separated by an aromatic or aliphatic ring the latter were found to undergo hydrolysis preferentially. cis,cis-Mucononitrile and fumaronitrile were regioselectively hydrolysed to monoacids as were m- and p-diacetonitriles. However o-phenylene-diacetonitrile gave the diamide in 65% yield. In contrast to suberonitrile NC(CH 2 ) 6 CN which gives the diacid trans-cyclohexane-1,4-diyldiacetonitrile in which the CN groups are separated by six carbons a§orded only the 41 Synthetic methods Part (iii) Enzyme chemistry O R O R OH R OH O R O R H H O H OH R OH S R S R d– d+ + S Nocardia EH1 epoxide hydrolase yields 94–98% ee 92–99% 11 12 12 buffer pH 8.0 retention R = (CH2)4CH3 (CH2)3CH CH2 CH2Ph Scheme 5 monoacid suggesting the importance of a more constrained substrate.A variety of aliphatic aromatic and heterocyclic mono nitriles were also hydrolysed to acids with this organism. Substrates bearing an adjacent substituent such as an ortho-substituent on an aromatic nitrile an adjacent heteroatom in a heterocyclic ring or a geminal substituent in an a,b-unsaturated nitrile undergo slow hydrolysis of the intermediate amides allowing them to be isolated in good yield. A substrate size of [7Å and the presence of functional groups near to the nitrile capable of metal complexation inhibit nitrile hydrolysis. E§enburger et al. have used polyurethane immobilized resting cells of Rhodococcus erythropolis MP50 for the resolution of naproxen amide to give (S)-naproxen in [99%ee.20 The free cells were inactive in organic solvents.Optimized conditions involved use of immobilized cells in butyl acetate containing 3 vol% DMSO and residual water. Recent contributions in the area of epoxide hydrolysis focus on the use of microbial sources of hydrolase enzymes. Faber et al. have identified six novel bacterial strains for the enantioselective hydrolysis of 2-methyl-2-alkyl and 2-methyl-2-aryl oxiranes.21 The best results (E[200) were obtained for the resolution of 2-methyl-2-pentyl oxirane with a Nocardia sp. which gave optically pure ([99%ee) (R)-epoxide and (S)-diol at 50% conversion. The microorganisms used in the study were identified by a search for strains known to catalyse the asymmetric epoxidation of alkenes. The author noted that the chance of finding selective epoxide hydrolases from a random screen would be low due to both epoxide enantiomers being equally toxic to the living cell.The same group has used a chemoenzymatic approach for the deracemization of 2,2-disubstituted epoxides via enantioconvergent hydrolysis using Nocardia EH1 epoxide hydrolase (Scheme 5).22 Racemic epoxides 11 were resolved with the enzyme to give (S)-diols 12 and unreacted (R)-epoxides. After slightly beyond 50% conversion treatment of the crude product mixture with concentrated sulfuric acid in dioxane –water gave near quantitative yields of the (S)-diol product 12. The two steps proceed with complementary selectivity. The enzyme hydrolysis occurs by attack at the less substituted carbon leading to retention of configuration whereas the acidcatalysed opening occurs with inversion at the more substituted centre.A lyophilized preparation of Aspergillus niger has been used to enantioselectively (E\41) hydrolyse p-nitrostyrene oxide,23 and a series of para-substituted styrene oxides have been resolved with the fungus Beauvaria densa CMC 3240 with stereoinversion of the hydrolysed enantiomer. However for o- and p-methyl- and chloro-styrene oxides the enantioselectivity is compromised and negligible activity was observed with p-nitrostyrene oxide.24 42 A. J. Carnell R1 R2 OH OH R1 R2 E. coli 13,14 R1 = H R2 = MeO (2.5 g l–1) 15,16 R1 = MeO R2 = MeO (0.8 g l–1) JM 109 (pDTG 601) Scheme 6 X O O O HO OH OH F OH O OH O O F OH X = Cl 8 steps X = Br 4 steps 18 19 17 1 6 4 5 Scheme 7 4 Oxidations Biooxidations continue to attract much interest transformations often being carried out with whole-cell systems due to the instability of the isolated enzymes or the need for cofactor recycling.Escherichia coli JM109 (pDTG601) is a recombinant organism which overexpresses the enzyme toluene dioxygenase. This strain has been used to convert substituted biphenyls 13 and 15 into the corresponding 3-aryl (1S,2R)-cyclohexadienediols 14 and 16 (Scheme 6).25 The absolute configurations of the diol products were determined by chemical correlation with (1S,2R)-3-iodocyclohexa-3,5- diene-1,2-diol. There have been further elegant examples of the use of bromocyclohexadiene cis-diol 17 (X\Br) obtained from the microbial oxidation of bromotoluene including the synthesis of two fluorinated inositols26 and a multigram synthesis of allo-inositol. 27 A versatile approach to deoxyfluorosugars allowed synthesis of the fluorinated lactol 18 and 2-deoxy-2-fluoroglucose 19 (Scheme 7).28 The key feature of this approach is ozonolysis of the C-1–C-6 double bond after substitution at C-4 and C-5 as a fluorohydrin was established.The naphthalene dioxygenase gene and its regulatory region from Pseudomonas fluorescens N3 has also been cloned in Escherichia coli JM109 giving an e¶cient bacterial system inducible by salicylic acid. The recombinant organism showed fairly broad substrate specificity for a range of naphthalenes with alpha or beta subtituents in the aromatic ring giving dihydrodiols in 50–94% yields.29 Continuing the work of Fonshen and Furstoss on remote hydroxylation of cyclic and bicyclic alkanes by Beauvaria bassiana ATCC 7159 Pietz et al.have proposed a key distance of 5.5Å from the oxygen of an N-phenyl carbamate ‘anchoring group’ and the position which is hydroxylated.30,31 Filamentous fungi such as Absidia coerula 43 Synthetic methods Part (iii) Enzyme chemistry OAc AcO H AcO OAc H OAc OAc AcO R1 AcO OAc H OAc R2 14 1 Absidia coerula 21 R1 = OH R2 = H (39%) 22 R1 = H R2 = OH (26%) 20 Scheme 8 O Cl O Cl O H2N OH O Cl H H HCl. Cunninghamella 4 steps ( R)-baclofen 25 23 ( R)-24 echinulata NRLL 3655 Scheme 9 have been used to hydroxylate the taxane skeleton in compound 20 giving novel C-1 and C-14 hydroxylated derivatives 21 and 22 (Scheme 8).32 The substrate for these biotransformations can be isolated from Taxomyces baccata T. mairei and related strains. Two new luciferase enzymes are able to catalyse the model Baeyer–Villiger reaction of bicyclohept-2-en-7-one.33 The synthetic application of the enzymatic Baeyer–Villiger reaction has been further demonstrated.In the synthesis of the GABA B agonist (R)-([)-baclofen 2534 asymmetric Baeyer–Villiger oxidation of the prochiral cyclobutanone 23 with Cunninghamella echinulata NRLL 3655 gave the enantiomerically pure (R)-chlorobenzyl lactone 24 in 30% yield (Scheme 9). Use of the better known biocatalyst Acinetobacter calcoaceticus NCIMB 9871 a§orded the (S)-lactone 24 with complementary selectivity but in lower enantiopurity (85% ee). The enzyme from the latter organism cyclohexanone monooxygenase (CHMO) is also able to catalyse asymmetric sulfoxidation reactions and a recent review by Willetts35 discusses synthetic applications and predictive active site models for this and related Baeyer–Villiger monooxygenase enzymes.It is interesting to compare a recently discovered enantioselective metal-catalysed Baeyer–Villiger reaction with the enzymatic results.36 Colonna has carried out highly selective asymmetric sulfoxidation of dialkyl sulfides with CHMO and chloroperoxidase (CPO) from Caldariomyces fumago.37 The CPO enzyme which requires hydrogen peroxide as the oxidant gave (R)-configured sulfoxides with generally high ee ([98%) and conversion (75–98%) for alkyl (cyclopentyl allyl pentyl isopropyl) methyl sulfides. Increasing the ring size (cyclohexyl) chain length ([C5) increased branching or replacement of methyl with ethyl gave lower selectivities. CHMO also exhibited high selectivity for alkyl methyl sulfides for substrates with limited steric requirements giving in most cases (R)-configured sulfoxides.Octyl and pentyl methyl 44 A. J. Carnell sulfoxides were (S)-configured in lower ee (50 and 60%). For the non-specialist the chloroperoxidase is a more convenient enzyme since it uses hydrogen peroxide and does not require regeneration of a redox cofactor. CPO can also be used in ButOH–water (1 1) and enhancements in enantioselectivity have been observed.38 Rabbit lung flavin monooxygenase (FMO2) has been used for the sulfoxidation of 2-naphthyl- and p-tolyl alkyl sulfides.39 For the p-tolyl series the enantioselectivity switched from giving (R)- to (S)-sulfoxides on changing the chain length from methyl to heptyl. Only short alkyl chains were tolerated for the 2-naphthyl series giving (R)- sulfoxides.Whole cells of the fungus Mortierella isabellina ATCC 42613 also gave (R)-configured sulfoxides from methyl aryl sulfides.40 Gallagher et al. have isolated a multicomponent alkene monooxygenase from Nocardia corallina B276 which catalyses the epoxidation of propene to (R)-propene oxide in 83%ee. Use of the whole cell system gave lower ee (69%) due to the presence of an (R)-selective epoxide hydrolase.41 5 Reductions Bakers’ yeast reductions continue to be widely used with the range of substrate types undergoing selective transformation expanding rapidly. Reduction of aromatic azides to amines has been reported in the chemoenzymatic synthesis of a benzodiazepine42 and a 4-aminopodophyllotoxin.43 Reports on the e§ects of organic solvents44 additives such as methyl vinyl ketone or chloroacetone45 and heat treatment46 enable selection of appropriate reaction conditions for a given b-keto ester or b-diketone.Reduction of 3-substituted cyclohexanones 28 with Baker’s yeast gave Prelog selectivity for both enantiomers of the substrate (38–46% yields,[90%ee) where the C-3 side chain possesses a sulfone or nitro function (Scheme 10).47 Evidently axial versus equatorial hydride delivery by NADH does not determine the selectivity as would be assumed for a reduction by a metal hydride (L-selectride gives predominantly the trans-isomer for R\Bun). 1-Methylsulfonylalkan-2-ones have been reduced to the corresponding b-hydroxy sulfones with Baker’s yeast under ‘dilute’ (A) and ‘concentrated’ (B) conditions in terms of the amount of yeast and sucrose used in the biotransformation.48 The reduction proceeded with good enantioselectivity (up to 87%ee) using conditions B (though isolated yields were moderate).These results are an improvement on those previously obtained with phenylsulfonyl derivatives. trans-2-Phenylsulfonyl-3-ethylcyclopentanone underwent a near perfect resolution to give the corresponding (1S,2R,3S)-alcohol product which could be oxidized to the (2R,3S)-ketone after separation. Treatment of either ketone with aqueous sodium hydroxide a§orded the corresponding enantiomer of 4-(phenylsulfonylmethyl)hexanoic acid through a retro-Claisen type process.49 Adam and co-workers have extended the scope of the horseradish peroxidase (HRP)-catalysed asymmetric reduction to that of erythro or threo hydroperoxyhomoallylic alcohols resulting in the (R,R)- or (R,S)-enantiomers being enantioselectively reduced to the allylic diols leaving unreacted peroxides in high ee.50 45 Synthetic methods Part (iii) Enzyme chemistry OH R R OH R O + Bakers' yeast H2O 30 °C (±)-26 (1 S 3 S)-27 (1 S 3 R)-28 R = CH2NO2 SPh SO2Ph CH2SO2Ph Scheme 10 6 Carbon–carbon bond formation The synthesis of N-acetyl-D-neuraminic acid (Neu5Ac) 30 (R\CH 3 ) using the corresponding aldolase to catalyse the reaction between N-acetyl-D-mannosamine (MannNAc) 29 (R\CH 3 ) and pyruvic acid has been scaled up by the Glaxo Wellcome group.51 Base-catalysed epimerization of N-acetylglucosamine (GlcNAc) gave a GlcNAc–MannNAc mixture (4 1) which could be used directly for the aldolase reaction.However GlcNAc is an inhibitor of the enzyme and therefore a high concentration of pyruvate was required to drive the equilibrium towards Neu5Ac.The excess pyruvate could then be removed as a bisulfate adduct. A second approach was developed to enrich the GlcNAc–MannNAc mixtures for MannNAc. This mixture could then be used in much higher concentration obviating the need to use a large excess of pyruvate. Neu5Ac aldolase is specific for pyruvic acid but is known to tolerate substitutions at C-4 C-5 or C-6 of ManNAc and configurations at C-4 and C-5 can be di§erent. The configuration at C-2 is essential but Wong has recently shown that a range of C-2 mannosamines 29 are accepted by the enzyme to produce the corresponding C-5 modified sialic acids 30 in good yields (55–72%) (Scheme 11).52 This will be a useful alternative to the use of the 5-deoxy-5-azide derivative in cases where synthetic elaboration proves di¶cult.Rabbit muscle aldolase (RAMA) catalyses the stereoselectiveC–C coupling between dihydroxyacetone phosphate (DHAP) and a broad range of aldehydes. An improved route for the preparation of DHAP on a large scale has been reported starting from 1,3-dibromoacetone.53 Wong has recently shown that remote dialdehydes with an aliphatic linkage lead to formation of multifunctional monoaldehydes.54 RAMA appeared to have no diastereopreference (R,R S,S and meso all reacted). The same group have used recombinant D- and L-threonine aldolases for the enzymatic synthesis of a range of b-hydroxy-a-amino acids (Scheme 12).55 L-Threonine aldolase (LTA) from E. coli and D-threonine aldolase (DTA) from Xanthomonas oryzae were cloned and overexpressed in E.coli. The enzymes are pyridoxal-5-phosphate dependent and LTA requires Mg2` as a cofactor. Both enzymes tolerate up to 40% DMSO as cosolvent and DMSO-induced rate enhancement of the LTA reactions was observed. LTA gave erythro b-hydroxy-a-L-amino acids with aliphatic aldehydes and the threo isomer with aromatic aldehydes as kinetically controlled products. DTA formed threo b-hydroxy-a-D-amino acids with aliphatic and aromatic aldehydes but the diastereoselectivity was lower than that of LTA. Although yields in many cases were low several b-hydroxy-a-amino acids such as hydroxyleucines c-benzyloxythreonines c-benzyloxymethyl threonines and polyoxamic acids were synthesized stereoselectively on a preparative scale. 46 A. J. Carnell HN O HO HO OH O R HO O OH CO2H HO N HO OH HO R O CO2H O H Neu5Mann 29 30 R = amino acyl peptidyl dansyl biotin aldolase Scheme 11 R OH O OH NH2 R OH O OH NH2 R OH O OH NH2 R OH O OH NH2 R H O OH O NH2 + DTA LTA D- threo D- erythro L- threo L- erythro + + Scheme 12 Some novel sources of (R)-oxynitrilases from apple apricot cherry and plum meal have been compared with the commercially available almond meal for the synthesis of aliphatic and aromatic cyanohydrins.Apple meal gave best results accepting sterically hindered substrates such as pivaldehyde to give (R)-cyanohydrins with high enantiopurity. 56 Optically active (S)-cyanohydrins have also been obtained by enantioselective cleavage of racemic cyanohydrins with (R)-hydroxynitrile lyase. Optimum conditions used a biphasic system (citrate bu§er–Pr*OH and capture of the aldehyde as a semicarbazone.57 7 C–O and C–N bond formation Modified sialic acids are present at the termini of many biologically important oligosaccharides and are among the most important residues for interactions with receptors.Halcomb and Chappell have demonstrated that a-2,3-sialyl transferase will accept variation in the 5-substituent on the activated CMP sialyl donor for glycosylation reactions with allyl b-lactoside (Scheme 13).58 The R\NHC(O)CH 2 OH and NH 2 (easily formed from NHCbz) features are found in a number of gangliosides. The sialyl transferase reaction can be used in vivo and therefore has the potential for modification of antigenic properties of intact cells. 3-Methyl aspartase isolated from a recombinant E. coli strain has been used to catalyse the enantioselective conjugate addition of a range of N-nucleophiles to the si-face of substituted fumaric acids in good yields (Scheme 14).59 The size of the substituents (R2) on theN-nucleophile tolerated by the enzyme displayed a profound dependence on the 47 Synthetic methods Part (iii) Enzyme chemistry O OCMP CO2 – HO R HO OH HO O CO2H O HO R HO OH HO O OH OH OH O O HO OH O OH allyl b-lactoside a-2,3-sialyl transfrase R = NHAc (Sialic acid) OH NHC(O)CH2OH NHCbz 5 Scheme 13 HO2C R1 H CO2 H R2 N R3 H HO2C R1 H CO2H H R2 N R3 2 3 + Mg2+ K+ 3-methyl aspartase R1 = H Me hal Et Pr Pri R2 = NH2 Me Et OH OMe R3 = H Me 12–61% yield (2 S 3 S) Scheme 14 size of R1 on the Michael acceptor in a mutually exclusive manner indicating that R1 and R2 are able to access similar regions of space in the enzyme active site.References 1 S.M. Roberts and N.M. Williamson Curr. Org. Chem. 1997 1 1. 2 N. J. Turner Curr. Org. Chem. 1997 1 21. 3 T. Sugai T. Yamazaki M. Yokoyama and H. Ohta Biosci. Biotechnol. Biochem. 1997 61 1419. 4 I.V. J. Archer Tetrahedron 1997 53 15 617. 5 H.L. Holland Adv. App. Microbiol. 1997 44 125. 6 A. Archelas and R. Furstoss Ann. Rev. Microbiol. 1997 51 491. 7 W.D. Fessner and C. Walter Top. Curr. Chem. 1997 184 97. 8 M. Petersen M.T. Zaetti and W. D. Fessner Top. Curr. Chem. 1997 186 87. 9 M. J. Kim I. T. Lim H. J. Kim and C. H. Wong Tetrahedron Asymmetry 1997 8 1507. 10 P. Renouf J.-M. Poirere and P. Duhamel J. Chem. Soc. Perkin Trans 1 1997 1739. 11 A. J. Carnell J. Barkley and A. Singh Tetrahedron Lett. 1997 38 7781.12 C. Bonini R. Ragoppi and L. Viggiani Tetrahedron Asymmetry 1997 8 353. 13 W. Adam P. Groer and G. R. Saha-Mo� ller Tetrahedron Asymmetry 1997 8 833. 14 M. Ferrero S. Fernandez and V. Gotor J. Org. Chem. 1997 62 4358. 15 H. Waldemann and A. Reidel Angew. Chem. Int. Ed. Engl. 1997 36 647. 16 F. E§enberger B. W. Graef and S. Oßwald Tetrahedron Asymmetry 1997 8 27. 17 O. Meth-Cohn and M.-X. Wang Chem. Commun. 1997 1041. 18 O. Meth-Cohn and M.-X. Wang J. Chem. Soc. Perkin Trans. 1 1997 3197. 19 O. Meth-Cohn and M.-X. Wang J. Chem. Soc. Perkin Trans. 1 1997 1099. 20 F. E§enberger B. W. Graef and S. Oßwald Tetrahedron Asymmetry 1997 8 2749. 48 A. J. Carnell 21 I. Osprian W. Kroutil M. Mischitz and K. Faber Tetrahedron Asymmetry 1997 8 65. 22 R. V. A. Orru W. Kroutil and K.Faber Tetrahedron Lett. 1997 38 1753; see also I. V. J. Archer D. J. Leak and D. A. Widdowson Tetrahedron Lett. 1996 37 8819. 23 C. Morisseau H. Nellaiah A. Archelas R. Furstoss and J. C. Baratti Enzyme Microb. Technol. 1997 20 446. 24 G. Grogan C. Rippe and A. Willetts J. Mol. Catal. B-Enzymatic 1997 3 253. 25 D. Gonzalez V. Schapiro G. Seoane and T. Hudlicky Tetrahedron Asymmetry 1997 8 975. 26 B. V. Nguyen C. York and T. Hudlicky Tetrahedron 1997 53 8807. 27 M. Desjardins L. E. Brammer and T. Hudlicky Carbohydr. Res. 1997 304 39. 28 F. Y. Yan B. V. Nguyen C. York and T. Hudlicky Tetrahedron 1997 53 11 541. 29 P. DiGennaro E. Galli G. Albini F. Pelizzoni G. Sello and G. Bestetti Res. Microbiol. 1997 148 355. 30 S. Pietz D. Wolker and G. Haufe Tetrahedron 1997 53 17 067. 31 S. Pietz R.Frohlich and G. Haufe Tetrahedron 1997 53 17 055. 32 S. H. Hu D. A. Sun X. F. Tian and Q. C. Fan Tetrahedron Lett. 1997 38 2721. 33 R. Villa and A. Willetts J. Mol. Catal. B-Enzymatic 1997 2 193. 34 C. Mazzini J. Lebreton V. Alphand and R. Furstoss Tetrahedron Lett. 1997 38 1195. 35 A. Willetts TIBTECH 1997 15 55. 36 C. Bolm T. K. K. Luong and G. Schlingho§ Synlett 1997 1151 and references therein. 37 S. Colonna N. Gaggero G. Carrea and P. Pasta Chem. Commun. 1997 439. 38 M.P. J. vanDeurzen I. J. Remkes F. vanRantwijk and R. A. Sheldon J. Mol. Catal. A-Chemical 1997 117 329. 39 M.B. Fisher and A. E. Rettie Tetrahedron Asymmetry 1997 8 613. 40 H. L. Holland L. J. Allen M.J. Chernishenko M. Diez A. Kohl J. Ozog and J. X. Gu J. Mol. Catal. B-Enzymatic 1997 3 311. 41 S. C. Gallagher R.Cammack and H. Dalton Eur. J. Biochem. 1997 247 635. 42 A. Kamal Y. Damayanthi B. S. Narayan Reddy B. Lakminarayana and B. S. Praveen Reddy Chem. Commun. 1997 1015. 43 A. Kamal B. Laxminarayana N. L. Gayatri Tetrahedron Lett. 1997 38 6871. 44 O. Rotthaus D. Kruger M. Demuth and K. Scha§ner Tetrahedron 1997 52 935. 45 Y. Kawai K. Takanobe and A. Ohno Bull. Chem. Soc. Jpn. 1997 70 1683. 46 J.-N. Cui R. Teraoka T. Ema S. Takashi and U. Masanori Tetrahedron Lett. 1997 38 3021. 47 R. Tanikaga Y. Obata and K.-i. Kawamoto Tetrahedron Asymmetry 1997 8 3101. 48 A. R. Maguire and D. G. Lowney J. Chem. Soc. Perkin Trans. 1 1997 235. 49 A. R. Maguire and L. L. Kelleher Tetrahedron Lett. 1997 38 7459. 50 W. Adam U. Hoch H. U. Humpf C. R. SahaMoller and P. Schreier Chem. Commun. 1997 2701.51 M. Mahmoudian D. Noble C. S. Drake R. F. Middleton D. S. Montgomery J. E. Piercey D. Ramlakhan M. Todd and M.J. Dawson Enzyme Microb. Technol. 1997 20 393. 52 C. C. Lin C. H. Lin and C. H. Wong Tetrahedron Lett. 1997 38 2649. 53 T. Geßaut M. Lemaire M. L. Valentin and J. Bolte J. Org. Chem. 1997 62 5920. 54 M.J. Kim I. T. Lim H. J. Kim and C. H. Wong Tetrahedron Asymmetry 1997 8 1507. 55 T. Kimura V. P. Vassilev G. J. Shen and C. H. Wong J. Am. Chem. Soc. 1997 119 11 734. 56 E. Kiljanen and L. T. Kanera Tetrahedron Asymmetry 1997 8 1225. 57 F. E§enberger and A. Schwaemmle Biocatal. Biotransform. 1997 14 167. 58 M.D. Chappell and R. L. Halcomb J. Am. Chem. Soc. 1997 119 3393. 59 M.S. Gulzar M. Akhtar and D. Gani J. Chem. Soc. Perkin Trans. 1 1997 649. 49 Synthetic methods Part (iii) Enzyme chemistry mmmm