10 Enzyme Chemistry By A. G. SLITHERLAND Department of Chemistry University of Exeter Exeter EX4 400 1 Introduction Interest continues to increase in the use of isolated enzymes and whole cell systems in organic synthesis.' This has been reflected in an increasing specialization in the subject area of review articles that have appeared in the last year. Thus reviews on the biotransformation reactions of organometallic' and carbohydrate3 systems; ester condensation and hydrolysis reaction^;^ enzyme catalyzed peptide synthesis;' and reactions mediated by bakers yeast6 and -even more specifically -Pseudomonas fluorescens lipase7 have tended to replace the more general review of the recent past.* In this account we shall consider in turn the use of hydrolytic enzymes followed by reduction oxidation then carbon-carbon bond forming reactions which broadly represents a decreasing order of current interest in each area -to some extent reflecting the availability cost and ease of use of the corresponding enzyme systems.2 Hydrolysis and Condensation Reactions Complex Alcohols.-The kinetic resolution of secondary alcohols (and the desym- metrization of meso or prochiral diols) by lipase and esterase catalyzed acylation or hydrolysis of the corresponding esters remains the most heavily studied area in the field of biotransformations. Despite the ease of execution resolutions in the hydrolytic sense are becoming relatively uncommon. However illustrations of the power of this approach still appear for a wide range of substrates ranging in complexity from the carbacyclin precursor (1)9 through the carbocyclic nucleoside intermediate (2)'' to simple 'synthons' such as (3)" -all of which were successfully resolved by Pseudomonas fluorescens lipase (Scheme 1).' N. J. Turner Annu. Rep. Prog. Chem. Sect. B Org. Chem. 1990 333. 'A. D. Ryabov Angew. Chem. Int. Ed. Engl. 1991 30 931. D. G. Drueckhammer W. J. Hennen R. L. Pederson C. F. Barbas 111 C. M. Gautheron T. Krach and C.-H. Wong Synthesis 1991 499. W. Boland C. Frossl and M. Lorenz Synthesis 1991 1049. V. Schellenberger and H. D. Jakubke Angew. Chem. Znt. Ed. EngL 1991 30 1437. R. Csuk and B. I. Glanzer Chem. Rev. 1991 91 49. '2. F. Xie Tetrahedron Asymmetry 1991 2 733. For example J. B. Jones Tetrahedron 1986 42 3351.Z. F. Xie H. Suemune K. Funakoshi T. Oishi H. Akita and K. Sakai J. Chem. SOC. Perkin Trans. 1 1991 3087. S. M. Roberts and K. A. Shoberu J. Chem. Soc. Perkin Trans. 1 1991 2605. 'I U. Goergens and M. P. Schneider J. Chem. SOC. Chem. Commun. 1991 1066. 263 264 A. G. Sutherland G:OzMeOAcI MeOz:B?H 8t,02MeOAc (+)H 4 H+H - - - (1) (51% 96% e.e.) (47%,87% e.e.) (3) R = SiMe,'Bu (-50% >95% ex.) (-5o% >95% e.e.) Reagents i Pseudomonas fluorescens lipase pH 7 buffer Scheme 1 An original application of the lipase mediated desymmetrization of prochiral esters was demonstrated in a synthesis of chiral 5,Sdisubstituted barbiturates.12 Hydrolysis of a bis-N-(acyloxymethyl) substrate (e.g. 4),with subsequent loss of formaldehyde from the N-hydroxymethyl intermediate (Scheme 2) induced chirality at the 5-position allowing conversion to a chiral biologically active N-methyl derivative.A feature of this work was that enantiocomplementary lipases were discovered -a rare occurrence. Thus Candida rugosa lipase hydrolysed (4)with (S)-selectivity (92% e.e.) while Humanicola lanuginosa lipase showed increased selectivity for the (R)-enantiomer (990/ e.e.). This finding proved particularly impor- tant in this instance as the enantiomers of the active derivatives both have significant but different biological profiles. Attempts to predict the enantioselectivity of esterases and lipases have met with some success re~ently.'?'~ Kazlauskas has proposed a rule to predict which enan- tiomer of a secondary alcohol ester will react faster in hydrolysis reactions catalyzed 12 M.Murata and K. Achiwa Tetrahedron Lett. 1991 32 6763. 13 E. J. Toone M. J. Werih and J. B. Jones J. Am. Chem. SOC.,1990 112 4946. Enzyme Chemistry by cholesterol esterase and the lipases from Pseudomonas cepucia and Candidu rug~sa.'~ The study of the outcome of hydrolysis of a wide range of substrates has led to the prediction that all three enzymes will react more quickly with the enantiomer represented in Figure 1 where greater selectivity is observed as the difference in relative size of the groups increases. The utility of this rule was demonstrated by increasing the selectivity of some resolutions through rational substrate modifications prompted by the model.M = mediumsizegroup L = large size group Figure 1 An alternative approach to improving the enantioselectivity of a hydrolysis that has met with some success is the use of additives. In the study of a Pseudomonas species lipase mediated hydrolysis of a secondary acetate (5),15 it was found that while some additives had insignificant or even deleterious effects on the reaction the addition of L-methioninol (6) markedly improved the selectivity. Studies on the 0 A jc, individual enantiomers of (5) suggested that the observed enhancement was pre- dominantly a result of a reduction in the rate of hydrolysis of the less favoured (S)-enantiomer. The results were interpreted as a consequence of a remote binding of (6) to the enzyme causing a conformational change at the active site which promoted the selectivity.A similar effect was observed in the diastereoselectivity of the cholesterol esterase hydrolysis of 24 RS)-a-tocopheryl acetate (7),16 where a dramatic increase in selec- tivity was observed in the presence of taurocholate relative to other bile salts. The authors suggested that this result -all the more striking when the remoteness of the position of hydrolysis to the centre being discriminated is considered -might also 14 R. J. Kazlauskas A. N. E. Weissfloch A. T. Rappaport and L. A. Cuccia J. Org. Chem. 1991,56 2656. T. Itoh E. Ohira Y. Takagi S. Nishiyama and K. Nakamura Bull. Chem. SOC.Jpn. 1991 64 624. 16 H. A. Zahalka P. J. Dutton B. O'Doherty T.A. M. Smart J. Phipps D. 0. Foster G. W. Burton and K. U. Ingold J. Am. Chem. Soc. 1991 113 2797. A. G. Sutherlund arise from an interaction between the additive and enzyme but did not discount diastereomeric influences upon the epimeric acetates within the mixed micelles formed by the bile salt. A new protocol that is likely to have wide applicability for improving the selectivity of resolution of a C,-symmetrical diol has been rep~rted.'~ The pig liver esterase (ple) catalyzed hydrolysis of ( f)-trans-1,2-diacetoxycyclohexane(Scheme 3) proceeded with only moderate selectivity. It was observed that the first hydrolysis (*)-rrans Scheme 3 proceeded far more quickly than the second which contrasted with the accepted notion that such sequential resolution processes would give optimum selectivity when the two rates were equal." It was discovered that a marked increase in selectivity could be achieved by the simple expedient of adding a hexane phase to the reaction.This had the effect of reducing the relative concentration of diester to monoester available to the enzyme in the aqueous phase with the consequence that the rates of hydrolysis became similar. The enantiomeric excess of the product diol increased from 58% to 94% at similar conversions as a result of this simple modification. As indicated above there is an increasing trend to approach the kinetic resolution of secondary alcohols in the acylation rather than hydrolysis direction. The com- plexity of structure capable of resolution by this means again appears diverse -from simple structures to the highly complex.For example the calicheamicinone (analogue) precursors (8) and (9) were resolved by acylation at the secondary and primary hydroxyl positions respectively with vinyl acetate catalyzed by Pseudomunus cepuciu lipase." \\\ Ill \\\ IIl The popularity of this approach undoubtedly stems from the observation that changes in solvents in these reactions are often rewarded with improvements in the l7 G. Caron and R. J. Kazlauskas J. Org. Chem. 1991 56 7251. 18 E. L. A. Macfarlane S. M. Roberts and N. J. Turner J. Chem. SOC.,Chem. Commun. 1990 569. 19 (a) D. S. Yamashita V. P. Rocco and S. J. Danishefsky Tetrahedron Lett. 1991 32 6667; (6) V. P. Rocco S. J.Danishefsky and G. K. Schulte Tetrahedron Lett. 1991 32 6671. Enzyme Chemistry 267 observed enantioselectivity. A recent illustration is in the resolution of the mucolytic drug trans-sobrerol (10)” using vinyl acetate and a Pseudomonas species lipase. The selectivity towards acylation of the (IS 5R)-enantiomer was demonstrated to increase from moderate to essentially total through the series tetrahydrofuran acetone dioxane 3-pentanone to t-amyl alcohol. AcO’ OH That the solvent of choice can vary according to substrate and lipase was illustrated in the complete resolutions of ferrocene (11) in toluene;” glycal (12) in dimethoxyethane;22 halohydrin (13) in neat vinyl acetate;23 alkyne (14) in he~ane;’~ and sulfide (15) in t-b~tylrnethylether.’~ It was also demonstrated that selectivity in these processes can also be improved by immobilizing the enzyme on an epoxy resin.26 The range of alternative acylating agents to vinyl acetate continues to grow.Thus while isopropenyl acetatez7 and both trifl~oroethyl~~~~~ esters and trichl~roethyl~~ continue to prove useful other reagents such as methyl pr~pionate,~~ vinyl aler rate,'^ vinyl la~reate,~’ and even diethyl carbonate33 have been advocated. Scheme 4 20 R. Bovara G. Carrea L. Ferrara and S. Riva Tetrahedron Asymmetry 1991 2 931. 21 M.-J. Kim H. Cho and Y. K. Choi J. Chem. Soc. Perkin Trans. 1 1991 2270. 22 D. R. Berkowitz and S. J. Danishefsky Tetrahedron Lett. 1991 32 5497. 23 J. Sakai H. Sakoda Y. Sugita M. Sato and C.Kaneko Tetrahedron Asymmetry 1991 2 343. 24 K. Burgess and L. D. Jennings J. Am. Chem. Soc, 1991 113 6129. 25 U. Georgens and M. P. Schneider J. Chem. Soc. Chem. Commun.,1991 1064. 26 B. Berger and K. Faber J. Chem. SOC.,Chem. Commun. 1991 1198. 27 C. R. Johnson P. A. PIC and J. P. Adams J. Chem. SOC.,Chem. Commun. 1991 1006. 28 (a) S. Ramaswamy and A. C. Oehlschlager Tetrahedron 1991 47 1157; (b) B. Morgan A. C. Oehlschlager and T. M. Stokes Tetrahedron 1991 47 1611. 29 J. M. Chong and E. K. Mar Tetrahedron Lett. 1991 32 5683. 30 H. J. Bestmann and U. C. Philipp Angew. Chem. Int. Ed. Engl. 1991 30 86. 31 A. J. M. Janssen A. J. H. Klunder and B. Zwanenberg Tetrahedron 1991 47 7409. 32 G. E. Jeromin and A. Scheidt Tetrahedron Lett. 1991 32 7021.33 D. Pioch P. Lozana and J. Graille Biotechnol. Lett. 1991 13 633. A. G. Sutherlund The maximum yield of enantiometrically pure material from a resolution pro- cedure is of course 50%. Oda and co-workers have however developed a procedure for the resolution of aromatic cyanohydrins where theoretical yields of 100% are possible.34 This was achieved by developing conditions which were compatible with both a lipase catalyzed enantioselective acylation and a chemical equilibration between the corresponding aldehyde and cyanohydrin (Scheme 4). Hence unreacted (R)-cyanohydrin was continually racemized and yields of up to 96% of moderately optically pure (S)-cyanohydrin acetates could be obtained in practice -directly from aldehyde starting materials.Complex Acids.-The use of esterases lipases and proteases in the resolution of racemic complex acids has received very little attention in comparison with secondary alcohols. This may well be a reflection of the thinking processes of chemists rather than a reflection of the potential of the technique. The power of this approach was ably illustrated in a Streptornyces griseus protease catalyzed resolution of the threo-serine derivative (16).35Both the product acid and the enantiomeric recoversd ester (each obtained in >45% yield and >%YO enan-tiomeric excess) could be converted in an enantioconvergent fashion to the broad spectrum antibiotic (-)-florfenicol (17). 0 Jones has studied the pig liver esterase catalyzed resolution of a range of monocyc-lic esters (Scheme 5).36 It was found that otherwise similar cyclobutane and cyclo- hexane derivatives were hydrolyzed with essentially total but opposite enantioselec- tivity whilst a corresponding cyclopentyl analogue was hydrolyzed with poor selec- tivity.These results were found to fit well with Jones’ proposed active site m0de1.l~ (18) n = 4,5,8 Intriguing results were observed in a study of the Cundidu cylindrucea lipase catalyzed hydrolysis of the esters (18).37Hydrolysis was found to occur at the methyl ester rather than acetoxy moiety and enantioselectivity was found to increase as the length of the alkyl chain separating the ester from the point of discrimination increased. This apparently anomalous result was rationalized in terms of two hydro-34 M.Inagaki J. Hirate T. Nishioka and J. Oda J. Am. Chem. SOC.,1991 113 9360. 35 J. E. Clark P. A. Fisher and D. P. Schumacher Synthesis 1991 891. 36 E. J. Toone and J. B. Jones Tetrahedron Asymmetry 1991 2 207. 37 U. T. Bhalerao L. Dasaradhi P. Neelakatan and N. W. Fadnavis J. Chem. Soc. Chem Commun. 1991 1197. Enzyme Chemistry C02Me ,C02H <-El,. Me + Me (45% 297% e.e.) (32% 297% e.e.) (*)a12Me -@02H + (y02M' Me Me (28% 22% e.e.) (21% 17% e.e.) phobic binding sites near the active site of the enzyme which might induce folding in the longer chain substrate causing the chiral centre to be close in space to the reactive centre in the active site. This proposal is akin to the Jones' pig liver esterase model which also suggests two hydrophobic binding sites unlike proposed models for other lipases (only one).Resolutions of complex alcohols by condensation have also been shown to be potentially solvent dependant.38 Thus the lipase catalysed condensation of ary- loxypropionic acid (19) with n-butanol was shown to be (R)-selective in cyclohexane n-hexane and toluene but (S)-selective in dichloromethane ethyl acetate and tetrahydro furan. Regioselective Reactions.-The use of hydrolytic enzymes to catalyse regioselective reactions of chiral substrates continues to meet with success. The protease subtilisin has been shown to catalyse the selective acetylation of N-acetylglucosamine (20) at the primary hydroxyl position39 while porcine pancreatic lipase has been shown to selectively butyrylate one (C2) of the three secondary alcohols of the galac- topyranoside (21).40 38 S.H. Wu F. Y. Chu and K. T. Wang Bioorg. Med. Chem. Lett. 1991 1 339. 39 C.-H. Wong Y. Ichikawa T. Krach C. Gautheron-Le Narvor D. P. Dumas and G. C. Look J. Am. Chern. Soc. 1991 113 8137. D. Colombo F. Ronchetti and L. Toma Tetrahedron 1991 47 103. A. G. Sutherland Ho& H O h CH2C0,EtI HO OH NHAc Ho OMeHO CHC0,EtI CH2CO2Et (20) (21) (22) Regioselective acylation of peptides has also been shown to be possible.41 Both L-Phe-a-L-Lys-0-tBu and L-Ala-a-L-Lys-0-tBu were selectively acetylated at the &-position of the lysine residue in preference to the free a-amino group by trifluoroethyl acetate in Pseudornonas species lipoprotein lipase catalysed reactions.The regioselective hydrolysis of achiral substrates has also met with some study. The ‘retro-fat’ (22) is hydrolyzed non-selectively with pig liver esterase but subtilisin catalyzes only hydrolysis of the 2-e~ter.~* Glycosidation Reactions.-Biotransformations are being employed with increasing frequency in the synthesis of complex carbohydrates. The utility of this approach lies in the inherent selectivity of the enzymes -negating the need for protecting groups. HO OUDP OUDP jOH& OH HOk0& HO HO OH (20-40%) Reagents i UDP-glucose epimerase; ii P-1,CgalactosyI transferase Scheme 6 Uridine disphosphate galactose (UDP-galactose) generated in situ by the action of UDP-glucose epimerase on the less expensive UPD-glucose has been shown to react with a variety of sugar ‘acceptors’ such as 1-deoxynojirimycin (Scheme 6),39,43 under galactosyl transferase catalysis to give the corresponding disaccharide.The modest yields obtained are quite acceptable given the directness of the approach. The discovery that the UPD-glucose required can also be generated in situ from catalytic amount of uridine diphosphate and stoichiometric quantities of glucose-6- 41 L. Gardossi D. Bianchi and A. M. Klibanov J. Am. Chem. Soc. 1991 113 6328. 42 L. Kvittingen V. Partali J. U. Braenden and T. Anthonsen Biotechnol. Lett. 1991 13 13. 43 C. Gautheron-Le Narvor and C.-H. Wong J. Chem. Soc. Chem. Commun. 1991 1130. Enzyme Chemistry 27 1 phosphate by an enzyme catalysed sequence,& should only increase the attractiveness of this procedure.Even more complex targets would seem approachable by this route -for example the tetrasaccharide sialyl Le" (23) has been readily synthesized by a similar fucosyl transferase catalyzed pr0cedu1-e.~~ HO HO OH (23) Other Hydro1yses.-The significance of enzyme systems capable of converting nitriles to carboxylic acids under neutral conditions seems to be becoming more widely appreciated. Although nitrilase activity (catalyzing a direct conversion) is sometimes available evidence seems to suggest that most conversions proceed by a two step process where a nitrile hydratase catalyzes a conversion of the nitrile to an amide which is in turn hydrolyzed to the corresponding carboxylic acid by an amidase.(295% e.e.1 (89%e.e.1 Scheme 7 The micro-organism Brevibacferium imperiale has been shown to convert a range of racemic aryloxypropionitriles -e.g. (24) -into the corresponding chiral (and antipodal) acids and amides (Scheme 7).47The authors demonstrated that the nitrile hydratase was non-enantioselective and that the amidase performed the kinetic resolution. Similar results were observed in the conversion of a range of arylpropionitriles (25) by Rhodococcus butanicaM where in all instances reported amide and/or acid J. Theim and T. Wiemann Angew. Chem Znt. Ed. Engl. 1991 30 1163. 45 D. P. Dumas Y. Ichikawa C.-H. Wong J. B. Lowe and R. P. Nair Bioorg. Med. Chem. Lett. 1991 1 425.46 H. Kakeya N. Sakai T. Sugai and H. Ohta Tetrahedron Lett. 1991 32 1343. 47 D. Bianchi A. Basetti P. Cesti G. Franzosi and S. Spezia Biotechnol. Lett. 1991 13 241. A. G.Sutherland could be obtained in near optical purity. Time-course studies revealed that both amidase and (to a lesser extent) nitrile hydratase systems showed some discrimination between enantiomers in this organism. The mildness of reaction condition required for these processes was illustrated by the conversion of a wide range of acid and base sensitive nitriles -e.g. (26) -to the corresponding acids by an isolated enzyme system derived from an unidentified Rhodococcus species.48 Rhodococcus equi was shown to hydrolyse the racemic p-lactam (27) with kinetic res~lution.~~ The recovered unreacted lactam (40% >99% enantiomeric excess) was converted to the low-toxicity anti-Cundida ulbicans agent (-)-Cispentacin (28).The investigation of the kinetic resolution of a series of terminal epoxides by the action of an epoxide hydrolase from a rabbit liver preparation met with mixed success.so The selectivity varied dramatically according to the alkyl substituent from nil (29 R = n-C,H,,) to apparently complete (29 R = Bu‘). It may be that the epoxide hydrolase activity currently being unearthed in fungal sourcess1 may prove more useful. A curious and possibly related report involves the selective microbial degradation of (S)-and (R)-3-chloro-l,2-propanediol by Pseudomonas and Alcaligenes bacteria respe~tively.~~ In both cases the unreacted enantiomer is recovered in high yield and enantiomeric excess (Scheme 8).The metabolized enantiomer is converted to glycidol then glycerol which infers the presence of an enzyme system capable of enantioselective halohydrin ring closure in both organisms. ‘OH OH Y Y Cl-H ClyOH I OH OH (>40% 99.5%e.e.) (>40°% 99.4% e.e.) Reagents i Pseudomonas sp.; ii Alcaligenes sp. Scheme 8 48 H. Klempier A. de Raadt K. Faber and H. Griengl Tetrahedron Lett. 1991 32 341. 49 C. Evans R. McCague S. M. Roberts A. G. Sutherland and R. Wisdom J. Chem. SOC. Perkin Trans. 1 1991 2276. 50 G. Belluci C. Chiappe F. Marioni and M. Benetti J. Chem. SOC. Perkin. Trans. 1 1991 361. 51 X. M. Zhang A. Archelas and R.Furstoss J. Org. Chem. 1991 56 3814. ’* T. Suzuki and N. Kasai Bioorg. Med. Chem. Lett. 1991 1 343. Enzyme Chemistry 3 Reduction Reactions The emphasis of research in this area continues to be on the reduction of ketones to chiral alcohols. The use of whole cell systems predominates although the utiliz- ation of isolated enzyme systems -of greater expense but often more selective and higher yielding -remains an area of note. Whole Cell Systems.-The bakers yeast (Saccharumyces cerevisiae)6 reduction of P-ketoesters remains a popular area of study. Thus the reduction of the viny- lacetoacetate (30) provides access to enantiomerically enriched compac-tin/mevinolin analogues,53 while optically pure diethyl (S)-malate was provided by conversion of sodium diethyl oxalacetate (3 1).54 An interesting variation of this theme was demonstrated by Simpson et aL55who found that when the reduction of P-ketoesters was performed in D20 almost complete di-deuteration at C-2 occured (with some additional incorporation at C-3 also observed).The reactions which could also be performed in conventional mode in water were also high yielding and highly enantioselective (Scheme 9). (75% 85% e.e.) (95% >95% e.e.) Reagents i Saccharomyces cereuisiae glucose D,O 30 "C Scheme 9 Bakers' yeast reductions are often marred by low or poorly reproducible yields and stereoselectivities -often a consequence of competing dehydrogenase activity or metabolism of the substrate or product. A dramatic approach to avoiding this problem has been proposed in the use of benzene as reaction solvent.56 Under these 53 F.D. Bennett D. W. Knight and G. Fenton J. Chem. Soc. Perkin Trans. 1 1991 133 and 519. 54 E. Santaniello P. Ferraboschi P. Grisenti F. Aragozzini and E. Maconi J. Chem. SOC.,Perkin Trans. I 1991 601. 55 M. P. Dillon M. A. Hayes T. J. Simpson and J. B. Sweeney Bioorg. Med. Chem. Lett. 1991 1 223. 56 K. Nakamura S. Kondo Y. Kawai and A. Ohno Tetrahedron Lett. 1991 32 7075. A. G. Sutherland conditions in the presence of a small amount of buffer to allow enzyme activity reductions of a range of a-ketoesters (32) were able to proceed with enhanced (R)-selectivity. This was interpreted in terms of the reaction conditions suppressing the metabolic and (S)-selective dehydrogenase activity.x CS,$CH3 N R CO2Et (32) R = CH3,(CH,),CH,n-C5H, (33) An ample demonstration that bakers’ yeast is not the only organism capable of synthetically useful reductions was given in a study of the reduction of 2-acetylthiazole (33).57No less than thirty eight yeast and mould strains capable of a highly selective reduction were identified. While some -though not all -strains of bakers’ yeast gave the (S)-alcohol in high yield (>70%) and optical purity (>95% e.e.) similar transformations were also accomplished with strains of Candida Rhizopus and Pichiu species. Furthermore two strains of Yarrowia lypolytica capable of enantioselective production of the (R)-alcohol were also found. Traditionally the regeneration of the cofactor (NADH or NADPH) required for the reduction is accomplished in whole cell systems by the metabolism of a carbohy- drate -typically glucose.The micro-organism Profeus vulgaris however has viologen-dependent molybdenum containing dehydrogenase systems which mitigate for cofactor recycling protocols more akin to palladium catalysed hydrogenation reactions than biotransformations. Thus resting cells of P. vulgaris catalyze the conversion of a range of a-ketoacids to the corresponding alcohols in the presence of formate or even hydrogen gas.58 The yields and selectivity are excellent as illustrated by the conversion of (34). P. oulgaris F &yo2 H or HCO; I 0 OH (34) (90%’>97%.e.e.) Isolated Enzyme Systems.-The commercially available D-laCtate dehydrogenase from Staphylococcus epiderrnidis has been shown to reduce a range of a-ketoacids to the corresponding (R)-2-hydroxyacids in high yield with essentially complete stereoselectivity (Scheme Given that enantiocomplementary activity has already been demonstrated using L-lactate dehydrogenase from rabbit it would appear that the synthesis of either enantiomer of many a-hydroxy acids and their derivatives -notably terminal epoxides -is possible by this approach with apparently complete and predictable enantioselectivity.57 G. Fantin M. Fogagnolo A. Medici P. Pedrini S. Poli and M. E. Guerzoni Tetrahedron Asymmetry 1991 2 243. 58 A. Schummer H. Yu,and H. Simon Tetrahedron 1991 47 9019. 59 M.-J.Kim and J. Y. Kim J. Chem. SOC.,Chem. Commun. 1991 326. 60 M.-J. Kim and G. M. Whitesides J. Am. Chem. SOC.,1988 110 2959. A-1 275 Enzyme Chemistry Staphylococcus epidermis R CO2H D-lactate R CO2H dehydrogenase R Yield YO e.e.,YO Et 86 >99 C-Pr 86 >99 n-Pr 92 >99 PhCH 80 >99 Scheme 10 4 Oxidation Reactions The use of cis-cyclohexadiene-l,2-diols(35) obtained by the microbial oxidation of arenes by mutant strains of Pseudomonas putida in organic synthesis is now well established. To some extent it might be argued tht these compounds are now full members of the ‘chiral pool’ and hence fall outside the remit of this discussion. However recent important developments in the field suggests their inclusion here.OMTPA OMTPA One major problem was that an accurate assessment of the optical purity of the diols was not predicated because of the non-availability of the racemic material. Boyd6’ has solved this problem by investigating the ‘H and I9F NMR spectra of the di-MTPA (a-methoxy-a-trifl,uoromethyl-phenylacetic) esters of the 4-phenyl- 1,2,4-triazoline-3,5-dione cycloadducts (36). The spectra of the ‘racemic’ adducts could be mimicked by making diesters with both (R)-and (S)-MTPA allowing the enantiomeric excess to be determined. In this way the diols from toluene ethylben- zene chlorobenzene trifluoromethylbenzene and benzyl acetate were all shown to be >98% e.e. while that from fluorobenzene was 60% e.e.. The diesters were all crystalline allowing confirmation of absolute configuration by X-ray crystallogra- phy.The absolute configuration of the diols from chlorobenzene toluene and anisole has also been confirmed by correlation of the CD spectra of their iron tricarbonyl complexes.62 The current major class of synthetic targets from these systems are the conduritols (1,2,3,4-tetrahydroxycyclohex-5-enes), the various stereoisomers of which display a wide range of biological activity. The strategies employed generally commence from the diol derived from fluoro chloro or bromobenzene when a suitable sequence of regioselective dihydroxylation or epoxidation and hydrolysis with expeditious D. R. Boyd M. R. J. Dorrity M. V. Hand J. F. Malone N. D. Sharma H. Dalton D. J. Gray and G. N. Sheldrake J.Am. Chem. SOC.,1991 113 666. 62 G. R. Stephenson P. W. Howard and S. C. Taylor J. Chem. SOC., Chem. Commun. 1991 126. A. G. Sutherland I HO .OH Conduritol E iv Conduritol F 0- Reagents i oso, 0-N' ; ii Bu3SnH AIBN; iii AcOH THF H,O; iv MCPBA /-J 'Me v KOH DMSO H,O Scheme 11 removal of the halogen lead to a wide selection of the series. Representative sequences are depicted in Scheme 11.63 It remains to be seen just how wide the synthetic potential of these diols is. With the current level of understanding of the chemistry of the system it would seem that any polyhydroxy/ amino functionalized system of the inositol/ pinitol/ con- duritol/ conduramine class should be readily preparable from this starting point.Whether other less 'tailor made' target systems will be readily accessible by this approach is not yet clear. Perhaps in response to this a number of groups have examined the possibility of functionalizing the ring system by carbon-carbon bond forming reactions. Thus the bromine in the acetonide derivative (37) can be successfully replaced by vinyl alkenyl and aryl groups via palladium mediated transmetalation approaches (Scheme 12);64 while similar chemistry on the unprotected bromo or iodo diols additionally yields alkyl allyl and nitrile derivative^.^^ These transformations have the effect of formally increasing the range of arenes that can be converted to the cyclohexadienediols without the need to develop new and potentially difficult biotransformations.A number of cycloadditions to the diene system have also been The acetonide derivatives were found to react conventionally in [4 + 21 manner with dimethyl acetylenedicarboxylate to give addition from the least hindered face (Scheme 13). However the reaction with diphenylketene proved less predictable 63 T. Hudlicky H. I.una H. F. Olivo C. Andersen T. Nugent and J. D. Price J. Chem. SOC.,Perkin Trans. 1 1991 2907. 64 S. V. Ley A. J. Redgrave S. C. Taylor S. Ahmed and D. W. Ribbons SYNLEm 1991 741. 65 D. R. Boyd M. V. Hand N. D. Sharma J. Chima H. Dalton and G. N. Sheldrake J. Chem. Soc. Chem. Commun. 1991 1630. 66 M. F. Mahon K. Molloy C. A. Pittol R. J. Pryce S. M. Roberts G. Ryback V. Sik J. 0. Williams and J. Winders J. Chem. Soc.Perkin Trans. 1 1991 1255. Enzyme Chemistry Scheme 12 R Scheme 13 R Ph R Scheme 14 with the expected [2 + 21 cycloaddition occuring in competition with a Diels-Alder reaction involving the ketene carbonyl (Scheme 14). Interestingly this latter path-way entirely absent in the reaction with 1,3-~yclohexadiene became predominant in the case of the fluoro derivative. The use of the enzymatic Baeyer-Villiger reaction continues to receive much attention. Whole cells of Acinetobacter calocaceticus were used to oxidize the meso- substrate (38) selectively and in high yield allowing rapid access to a key subunit for the synthesis of the polyether antibiotic ionomycin (Scheme 15).67 Two other groups have examined the use of this organism in the context of the oxidation of racemic bicyclic ketones.Roberts and co-workers68 observed two distinct types of kinetic resolution in action dependant on the substrate under examination. The bromohydrin (39) was resolved in the more typical sense in that only one enantiomer reacted so that both product lactone and recovered ketone were obtained in high optical purity (Scheme 16). Both enantiomers of the simpler 67 M. J. Taschner and Q. Z. Chen Bioorg. Med. Chem. Lett. 1991 1 535. 68 A. J. Carnell S. M. Roberts V. Sik and A. J. Willetts J. Chem. Soc. Perkin Trans. 1 1991 2385. A. G. Sutheriand Me**Qd*Me __*i ii Me0,CVOH Me Me 0 (38) (82%)1iii iv Ho-OTBDPS , I Me Me Reagents i Acinetobacter; ii CH,N,; iii Bu‘Ph,SiCI base; iv LiAlH Scheme 15 1 A I I I OH OH OH (*)(39) Scheme 16 ketone (40) reacted but resolution still occurred as the regioselectivity of the oxygen insertion was different for each enantiomer (Scheme 16).The authors interpreted these and other results in terms of only one enzyme being involved where steric interactions affected the stereochemical mode of attack of the enzyme bound hydro- peroxide moiety resulting in selective migrations in each enantiomer. Furstoss et aZ?9 observed a third distinct mode of resolution in the transformation of the r2.2.11 ketone (41) (Scheme 17). In this case only one enantiomer underwent Baeyer-Villiger oxidation (with an unexpected rearrangement to yield the [3.3.0] system (42)) while the other enantiomer was actually reduced to both the correspond- ing endo- and exo-alcohols all three products being formed in high optical purity.Leaving the rearrangement aside as an artefact of this particular substrate this distribution of products may represent some form of dynamic balance of cofactors during the course of the reaction. The relationship of the cofactors required for the interconversion between alcohol and ketone and ketone and lactone has been elegantly exploited by Roberts et aZ.” 69 K. Konigsberger V. Alphand R. Furstoss and H. Griengl Tetrahedron Lett. 1991 32 499. 70 A. J. Willetts C. J. Knowles M. S. Levitt S. M. Roberts H. Sandey and N. F. Shipton J. Chem. Soc. Perkin Trans. 1 1991 1608. Enzyme Chemistry h ,OCH2Ph Scheme 17 The realization that the cofactor generated by an alcohol dehydrogenase mediated oxidation of an alcohol to a ketone namely NADPH was that required for the monooxygenase catalyzed oxidation to the lactone prompted the investigation of a coupled enzyme system.Thus the use of the isolated enzymes Thermoanaerobium brockii alcohol dehydrogenase and A. caloaceticus monooxygenase in tandem con- verted the alcohol (43)to the corresponding lactones (44and 45) in the presence of only catalytic amounts of NADP+ (Scheme 18). HO Therrnoanaerohium brockii dehydrogenase (43) 0 Acinetobacter calcoaceticus monooxygenase (44) (45) Scheme 18 The monooxygenase catalyzed epoxidation of alkenes is likely to be an important and rewarding area of study in the near future.That high optical purities are obtainable by these processes was illustrated by Wong and co-workers in a study of the oxidation of allylic alcohol derivative^.^' Although the parent alcohol did not react usefully a number of other derivatives (e.g. 46) could be oxidized with high enantiomeric excess by growing cells of Pseudomonas oleovoruns. H.Fu M.Newcomb and C.-H. Wong J. Am. Chern. Soc. 1991,113 5878. A. G. Sutherland 5 Carbon-Carbon Bond Forming Reactions The use of fructose 1,6-diphosphate aldolase (FruA) from rabbit muscle72 or E. ~oli~~ as a catalyst for the aldol reaction of dihydroxyacetone phosphate (DHAP) with a wide range of aldehydes is well documented. The reaction reliably gives products with (3S 4R)stereochemistry (Scheme 19).0 Scheme 19 The use of two new aldolases from E. coli namely L-rhamnulose and L-fuculose 1-phosphate aldolases (RhuA and FucA) has now been reported.74 These enzymes are important in that they catalyze the formation of diastereomerically complemen- tary products. Thus RhuA and FucA produce (3R 4s)and (3R 4R)products respectively (Scheme 19). Both these enzymes were shown to accept a wide range of aldehyde substrates including aliphatic and aromatic species. The development of a fourth enzyme D-tagatose 1,6-bisphosphate aldolase (TagA) capable of produc- ing (3S 4s)products is understandably noted as an area of current investigation by this group. The increased interest in aza-sugars and glycoproteins has prompted the investiga- tion of the use of aldolases in synthetic routes to these materials.This approach had been hampered by the fact that many aminoaldehyde equivalents had transpired to be poor substrates for these enzymes. However this obstruction has now been circumvented by the discovery that azidoaldehydes participate readily in the process as illustrated in the high yielding synthesis of the azido-xylulose (47).75 CHO 1) DHAP FruA W O H N3-2) Phosphatase N3 OH (47) (78%) 72 (a) M. D. Bednarski E. S. Simon N. Bischofberger W. D. Fessner M.-J. Kim W. Lees T. Saito H. Waldmann and G. M. Whitesides J. Am. Chem. Soc. 1989 111 62; (b) E. J. Toone E. S. Simon M. D. Bednarski and G. M. Whitesides Tetrahedron 1989 45 5365. 73 C. H. van der Osten A.J. Sinskey C. F. Barbas 111 R. L. Pederson Y. F. Wang and C.-H. Wong J. Am. Chem. SOC.,1989 111 3924. 74 W. D. Fessner G. Sinerius A. Schneider M. Dreyer G. E. Schultz J. Badia and J. Aguilar Angew. Chem. Int. Ed. Engl. 1991 30 555. 75 R. R. Hung J. A. Straub and G. M. Whitesides J. Org. Chem. 1991 56 3851. Enzyme Chemistry 281 This discovery together with the increased realization that in many instances aldolases only utilize one enantiomer of a racemic aldehyde has led to the rapid synthesis of a wider range of aza-sugars for evaluation as glycosidase inhibitors through the use of all three available aldolases (Scheme 20).76*77 i DHAP FruA N3JCH0 ii Phosphatase i DHAP RhuA N3+OH & H6 HO OH *OH *How ii Phosphatase N3 HO HO OH Scheme 20 The use of the aldolase-like enzyme transketolase (TK) is likely to attract much attention in coming years.This commercially available thiamin pyrophosphate dependant enzyme readily catalyzes the irreversible decarboxylative coupling of hydroxypyruvate with aldehydes (Scheme 21). The range of potential 'acceptor' aldehydes has been shown to be very wide,78 including aliphatic aromatic heterocyc- lic and a,P-unsaturated systems. 0 OH R1 H + HOzC &OHZ/-e_OH R Scheme 21 The enzyme has also been shown to accept only one enantiomer of racemic 2-hydroxyaldehyde substrates to give a resultant vicinal diol with D-threo configur- ation (and recovered L-hydroxyaldehyde in high enantiomeric excess) -as utilized by Whitesides in a synthesis of (+)-exo-brevicomin (Scheme 22).79 A rather more distinct carbon-carbon bond forming enzyme is oxynitrilase.This enzyme (also known as mandelonitrile lyase) catalyses the stereospecific reversible addition of HCN to aldehydes (whilst some ketones have also been shown to be substrates"). The process is not without its problems apart from the notional 76 T. Kajimoto K. K. C. Liu R. L. Pederson Z. Zhong Y. Ichikawa J. A. Porco Jr. and C.-H. Wong J. Am. Chem. SOC.,1991 113 6187. 77 K. K. C. Liu T. Kajimoto L. Chen. Z. Zhong Y. Ichikawa and C.-H. Wong J. Org. Chem.,1991,56,6280. 78 C. Demuynck J. Bolte L. Hecquet and V. Dalmas Tetrahedron Lett. 1991 32 5085. 79 D. C. Myles P. J. Andrulis 111 and G. M. Whitesides Tetrahedron Lett.1991 32 4835. 80 F.Effenberger B. Horsch F. Weingart T. Ziegler and S. Kuhner Tetrahedron Lett. 1991 32 2605. A. G. Sutherland Me Scheme 22 hazards involved as the potential for 'background' non-enzymic addition of cyanide together with some racemization of the product cyanohydrin under reaction condi- tions has meant that the consistent achievment of high optical purity has proved difficult. HO CN A ENZ [ENZ . HCN] R "OxHCN R H Scheme 23 An elegant solution to these problems has been revealed.** The use of acetone cyanohydrin as a donor of HCN in a transcyanation process (Scheme 23) means that the level of HCN is kept low (possibly in enzyme bound form) and hence background reaction is kept to a minimum.Product racemization is also reduced by performing the reaction in an ether-buffer two phase system. A stark illustration of the power of this procedure is in the formation of phenylacetaldehyde cyanohydrin (48) where the enantiomeric excess of the product is more than doubled (from 40% to 88%) by adopting this new procedure. It should be noted that this technique is enantiocomplementary to the lipase-mediated route to (S)-cyanohydrin acetates delineated earlier. (48) (83'10 88% e.e.) V. I. Ognyanov V. K. Datcheva and K. S. Kyler J. Am. Cfiem.Soc. 1991 113 6992.