Preparation of enantiopure epoxides by biocatalytic kinetic resolution

Molecular chirality is of great importance for many processes in biological systems. Examples are interactions with enzymes and receptor systems for hormones, sensory recognition and drug metabolism. Activation of biological activity when initiated by interaction with bioactive compounds is highly based on complementary stereochemistry. Synthetic bioactive compounds, like agrochemicals and pharmaceuticals, are therefore now preferably produced as single enantiomers.

Various bioactive compounds can be effectively prepared from enantiopure epoxides. These chiral building blocks can be used to introduce one or two adjacent chirality centers in a target molecule. In the present study, enzymatic kinetic resolution via direct epoxide ring-opening has been studied for the preparation of enantiopure epoxides. For this method, the biocatalytic activities of a bacterial epoxide carboxylase and a yeast epoxide hydrolase have been explored (Scheme 1).

;

Scheme 1. Kinetic resolution of cis - (R ₁ = H, R ₂ = alkyl) and trans - (R ₁ = alkyl, R ₂ = H) 2,3-disubstituted epoxides by enzymatic nucleophilic ring-opening. XH ₂ represents NADPH or a reducing dithiol compound. (Ketone: R= alkyl).

Bacterial epoxide carboxylase

In alkene-utilizing bacteria, epoxides are generated by monooxygenases and subsequently further degraded. The epoxide-degrading enzyme system has been recently identified as an epoxide carboxylase. However, in the absence of CO ₂ , the reaction catalyzed is actually an isomerization of the epoxide. The enzyme can therefore also be regarded as an epoxide isomerase.

Epoxide carboxylase/isomerase from Xanthobacter Py2 was found enantioselective in the conversion of 2,3-disubstituted aliphatic epoxides (Scheme 1). Only (2 S )-enantiomers were converted by propene-grown cells of Xanthobacter Py2 and (2 R )-enantiomers were thus resolved from a racemic mixture with almost maximal feasible yield ( Chapter 2 ). Aliphatic 1,2-epoxides, being intermediates in 1-alkene metabolism, were converted without remarkable enantioselectivity.

In the subsequent study, epoxide bioconversion was studied in more detail ( Chapter 3 ). Epoxide substrates were found to be converted to ketones via an hydroxy intermediate. The enzymatic reaction was dependent on NAD ⁺and a reducing cofactor, which could be replaced by synthetic dithiol compounds. Based on these findings, a four-step reaction mechanism was proposed starting from nucleophilic ring-opening of the epoxide. Follow-up studies by various other research groups concentrated on the metabolism of the physiological substrate 1,2-epoxypropane. By these studies, the enzymatic steps involved in 1,2-epoxypropane metabolism were further elucidated.

Yeast epoxide hydrolase

Cofactor-independent microbial epoxide hydrolases are generally regarded as attractive biocatalytical tools. Epoxide hydrolase catalyzed ring-opening of epoxides can be exploited for the production of enantiopure epoxides and vicinal diols (Scheme 1). The biocatalytical potential of microbial epoxide hydrolases has been first recognized in studies using enzymes from fungal and bacterial origin. Epoxide hydrolase activities in yeasts have been subsequently explored.

Yeast epoxide hydrolase (YEH) activity has been demonstrated for the hydrolysis of various structurally divergent epoxides by Rhodotorula glutinis ATCC 201718 ( Chapter 4 ). Very high enantioselectivities were determined in the hydrolysis of 2,3-disubstituted aryl and aliphatic epoxides (Scheme 1). Asymmetric hydrolysis of meso epoxides has been demonstrated and interestingly this property has been restricted to yeasts in particular.

Typical other substrates for the yeast enzyme are monosubstituted aliphatic epoxides. Enantiomeric discrimination was expected to be difficult for these highly flexible and rather 'slim' molecules. Therefore, kinetic resolution of a homologous range of aliphatic 1,2-epoxides by Rhodotorula glutinis was studied in more detail ( Chapter 5 ). Activities as well as enantioselectivities were found to be strongly influenced by the chain length of the substrate used. Best results were obtained in the resolution of 1,2-epoxyhexane.

Preparative-scale YEH-catalyzed resolution

Preparative-scale kinetic resolutions were investigated with 1,2-epoxyhexane as a model substrate and cells of Rhodotorula glutinis as biocatalyst ( Chapter 6 ). Scaling-up was hampered by inhibition due to substrate toxicity, and to an even higher extend, by product inhibition of the formed diol. A critical inhibitory diol concentration was determined as 50 mM for 1,2-hexanediol. For protection against high epoxide concentrations, aqueous/organic two-phase reaction media were tested. Long-chain aliphatic alkanes were suitable biocompatible solvents and dodecane was selected for further applications. However, dodecane and other biocompatible solvents gave no protection towards the diol.

Preparative-scale resolution of 1,2-epoxyhexane (22 g) was performed successfully in an aqueous/organic two-phase membrane bioreactor. A cascade configuration of two hollow-fiber membrane modules was used ( i ) to separate the biocatalyst from the organic solvent containing feed solution with concentrated epoxide (2 M) and ( ii ) to remove inhibitory amounts of diol.

In a modified process design, the membrane bioreactor was used for continuous extractive kinetic resolution of 1,2-epoxyhexane (1 M in dodecane). Under these conditions, enantiopure ( S )-epoxide (13 g) was obtained in the effluent solvent phase. The process allowed long-term continuous production of enantiopure epoxide without the need for complete resolution of the racemic substrate in the feed reservoir. Optimization of this process will however be necessary for improvement of the productivity.