The enantioselectivity of pig liver esterase catalysed hydrolysis of cis-N-benzyl-2,5-bis(methoxy-carbonyl)pyrrolidine (1) has previously been shown to be very dependent on the reaction conditions.1,2 Hydrolysis performed in media buffered with tris(hydroxymethyl)aminomethane (Tris) afforded a monoester with much higher optical purity than hydrolysis in media without Tris. Detailed product studies in a Tris-buffered medium have been performed using NMR-techniques and a 13C-labelled ester. The NMR-studies revealed the presence of (2S,5R)-N-benzyl-2-methoxycarbonyl-5-[[[2-hydroxy-1,1-bis(hydroxymethyl) (4) as an intermediate, which together with the isolated product (2S,5R)-N-benzyl-2-carboxy-5-[[[2-hydroxy-1,1-bis(hydroxymethyl)ethyl]amino]carbonyl]pyrrolidine (3) suggested Tris as a competitive nucleophile to water. The increased enantioselectivity seen in the produced (2R,5S)-N-benzyl-2-methoxy-carbonyl-5-carboxypyrrolidin (2) was explained by the preference of Tris to react faster with one of the diastereomeric acyl enzymes over the other.

S-Ethyl thiooctanoate was used as acyl donor in the transesterification of 2,3-butanediol (1), 2,4-pentanediol (2), and 2,5-hexanediol (3), catalysed by a lipase from Candida antarctica. Mixtures of all stereoisomers were used as substrates in each case. 2,5-Hexanediol was transesterified with high stereoselectivity and the (2S,5S)-2,5-hexanediol was isolated in good yield with >99 % ee. The diester of (2R,5R)-2,5-hexanediol was formed in good yield and was hydrolysed to yield the (2R,5R)-2,5-hexanediol of high enantiomeric excess (>99% ee). Similar results were obtained for 2,4-pentanediol with >99% ee for both enantiomers. The stereoselectivity for 2,3-butanediol was lower than for 2 and 3, giving 89% ee for the R,R-enantiomer and 34% ee for the SS-enantiomer.

A method has been developed for the calculation of the enantioselectivity of chymotrypsin catalysed hydrolytic reactions using molecular mechanics and molecular dynamics. Nine different ester substrates, which are hydrolysed by the enzyme over a wide range of reaction rates have been studied. Models of the transition state of the ester hydrolysis were built using computer aided molecular modelling. The energies of the transition state models were calculated by molecular mechanics and molecular dynamics methods. The point charges of the substrates were modelled from known force field parameters and by semiempirical methods. The difference in free energy of activation between the enantiomers of each substrate were compared with experimental values. The calculations approximated the experimental results. The calculated structure of the transition state model of the chymotrypsin catalysed hydrolysis of acetyl-phenylalanine ester was virtually the same as the published crystal structure of a chymotrypsin-trifluoromethyl ketone inhibitor complex (Brady et al., Biochemistry 29: 7600-7607, 1990).

A simple method to overcome low equilibrium conversion in lipase catalysed resolution of alcohols by transesterification was developed. Ethyl octanoate was used as acyl donor as well as solvent and the reaction equilibrium was shifted by applying reduced pressure, forcing the co-product ethanol to evaporate during the reaction. Using a lipase fromCandida antarctica 2-octanol, 1-phenyl ethanol, 1-cyclohexyl ethanol andtrans-2-methylcyclohexanol were resolved in good optical and chemical yields.

Pig liver esterase was separated into isoenzyme fractions with known subunit compositions. The fractions showed differences in enantiotopic ester group selectivity in hydrolysis of two substrates of synthetic value, benzylmethylpropanedioic acid dimethyl ester and cis-N-benzyl-2,5-bismethoxy-carbonylpyrrolidine. A difference in aliphatic chain length specificity between the isoenzyme fractions was also observed. The results indicate that pig liver esterase cannot be regarded as homogeneous when used in organic synthesis.