Improved enantioselectivity of a lipase by rational protein engineering

A model based on two different binding modes for alcohol enantiomers in the active site of a lipase allowed rational redesign of its enantioselectivity. 1-Halo-2-octanols were poorly resolved by Candida antarctica lipase B. Interactions between the substrates and the lipase were investigated with molecular modeling. Unfavorable interactions were found between the halogen moiety of the fast-reacting S enantiomer and a region situated at the bottom of the active site (stereoselectivity pocket). The lipase was virtually mutated in this region and energy contour maps of some variants displayed better interactions for the target substrates. Four selected variants of the lipase were produced and kinetic resolution experiments were undertaken with these mutants. Single point mutations gave rise to one variant with doubled enantioselectivity as well as one variant with annihilated enantioselectivity towards the target halohydrins. An increased volume of the stereoselectivity pocket caused a decrease in enantioselectivity, while changes in electrostatic potential increased enantioselectivity. The enantioselectivity of these new lipase variants towards other types of alcohols was also investigated. The changes in enantioselectivity caused by the mutations were well in agreement with the proposed model concerning the chiral recognition of alcohol enantiomers by this lipase.

Rational design of enantioselective enzymes requires considerations of entropy

Entropy was shown to play an equally important role as enthalpy for how enantioselectivity changes when redesigning an enzyme. By studying the temperature dependence of the enantiomeric ratio E of an enantioselective enzyme, its differential activation enthalpy (Delta(R-S)DeltaH(++)) and entropy (Delta(R-S)DeltaS(++)) components can be determined. This was done for the resolution of 3-methyl-2-butanol catalyzed by Candida antarctica lipase B and five variants with one or two point mutations. Delta(R-S)DeltaS(++) was in all cases equally significant as Delta(R-S)DeltaH(++) to E. One variant, T103G, displayed an increase in E, the others a decrease. The altered enantioselectivities of the variants were all related to simultaneous changes in Delta(R-S)DeltaH(++) and Delta(R-S)DeltaS(++). Although the changes in Delta(R-S)DeltaH(++) and Delta(R-S)DeltaS(++) were of a compensatory nature the compensation was not perfect, thereby allowing modifications of E. Both the W104H and the T103G variants displayed larger Delta(R-S)DeltaH(++) than wild type but exhibited a decrease or increase, respectively, in E due to their different relative increase in Delta(R-S)DeltaS(++).

Resolution of an iridoid synthon, gastrolactol, by means of dynamic acetylation and lipase-catalyzed alcoholysis

A short synthetic route to asymmetric iridoids was developed. The three key steps were an intramolecular [4 + 2] cycloaddition reaction of an enamine derivative of 8-oxocitral (2), a dynamic acetylation, and an enzymatic resolution of the gastrolactyl acetates 5a and 5b, iridoids with three stereocenters. Some regio- and stereoselective heterogeneous catalytic hydrogenations of double bonds in iridoid aglucones were discussed.

Mass transport limitations reduce the effective stereospecificity in enzyme-catalyzed kinetic resolution

[reaction-see text] The kinetic resolution of seudenol catalyzed by Candida antarctica lipase B in hexane was investigated. Large differences in reaction rate and stereospecificity were observed when different enzyme preparations were used. These differences were ascribed to mass transport limitations which reduced both reaction rate and stereospecificity. Lyophilized enzyme preparations were more apt to give this problem than immobilized preparations. Further, low substrate concentrations enhanced the effect. Thus, high alcohol concentrations and enzyme immobilization can be recommended.

An active-site titration method for lipases

A method for active-site titration of lipases has been developed based on irreversible inhibition by methyl p-nitrophenyl n-hexylphosphonate. This method was applied to five lipases displaying from minor to pronounced interfacial activation. Soluble and immobilized lipases were successfully titrated in aqueous media. A low concentration of sodium dodecyl sulfate was needed for lipases displaying pronounced interfacial activation. The carrier of some of the immobilized preparations adsorbed part of the produced p-nitrophenolate. This problem could be solved by extracting the p-nitrophenolate after inhibition. The method was extended to apolar organic solvents in the case of immobilized lipase preparations.