Introduction

Because of the safety and regulation pressure on the chirality of biologically active compounds by US FDA, the commercial need of chiral intermediates for the synthesis of chiral pharmaceuticals is tremendously increasing (Breuer et al. 2004). Enantiopure epoxides and their corresponding vicinal diols (vic-diols) are valuable synthetic building blocks for the synthesis of chiral pharmaceuticals (Kasai et al. 1998; Monterde et al. 2004).

Many biocatalytic conversions have been developed to prepare chiral epoxides and diols (Breuer et al. 2004; Lee and Shuler 2007). One of the promising biocatalytic approaches is the kinetic resolution of racemic epoxides via enantioselective hydrolysis reaction catalyzed by epoxide hydrolase (EH, EC 3.3.2.3) (Orru et al. 1998). Even though kinetic resolution has been proved to be commercially important, it has an intrinsic limitation that the theoretical yield cannot exceed 50%. To overcome the limitation of kinetic resolution, many efforts have been paid to develop so-called enantioconvergent processes leading to a 100% yield and 100% enantiopurity from a racemic mixture as the substrate (Lutje Spelberg et al. 1998; Steinreiber et al. 2001; Strauss et al. 1999). We can set up an enantioconvergent process by the combined use of two different EHs possessing complementary enantio- and regioselectivities (Manoj et al. 2001; Genzel et al. 2002). In some cases, it is possible to enantioconvergently hydrolyze racemic epoxides by using single EH biocatalyst based on ‘stereochemical flexibility’ of some EHs (Fig. 1) (Moussou et al. 1998; Steinreiber et al. 2001).

Fig. 1
figure 1

Reaction schemes of the enantioconvergent hydrolysis of racemic epoxides using the recombinant cell containing C. crescentus EH. Given yields are theoretical values

Full size image

The plant EH from potato was employed to catalyze the enantioconvergent bioconversion of racemic styrene oxide derivatives (Monterde et al. 2004). This is only one report on the enantioconvergent hydrolysis of styrene oxide derivatives using single biocatalyst, and no enantioconvergent process for the preparation of enantiopure diol using heavy metal-based chemocatalysts has been reported so far, to the best of our knowledge. Recently, we have reported enantioconvergent bioconversion of racemic styrene oxide to prepare (R)-phenyl-1,2-ethanediol by employing a bacterial EH from Caulobacter crescentus and a marine fish EH from Mugil cephalus (Kim et al. 2008). The bacterial EH of C. crescentus exhibited interesting activity that some racemic epoxides were converted into the corresponding diols in enantiomerically enriched way. In this paper, the hydrolysis of eight styrene oxide derivatives was investigated to achieve an enantioconvergent conversion by using single cell containing C. crescentus EH (Fig. 2). The enantioconvergent conversion of p-chlorostyrene oxide was optimized and preparative-scale preparation was also carried out.

Fig. 2
figure 2

Various epoxides used as substrates for enantioconvergent hydrolysis by the recombinant cell containing C. crescentus EH (1, styrene oxide; 2, o-chlorostyrene oxide; 3, m-methylstyrene oxide; 4, m-chlorostyrene oxide; 5, m-nitrostyrene oxide; 6, p-methylstyrene oxide; 7, p-chlorostyrene oxide; 8, p-nitrostyrene oxide)

Full size image

Materials and methods

Media and culture conditions

The recombinant E. coil BL21(DE3) harboring the EH gene of C. crescentus was grown in 1 LB broth containing 50 μg kanamycin/ml at 37°C (Hwang et al. 2005). When the OD600 reached 0.6, 1 mM IPTG was added to the culture broth. After induction, the cells were incubated at 20°C overnight with shaking at 250 rpm. The cells were harvested by centrifugation (1,000g, 10 min, 4°C), washed with 50 ml 20 mM phosphate buffer (pH 8.0). The harvested cells were freeze-dried for the bioconversion reactions.

General procedure for the synthesis of racemic epoxides

To a stirred solution of various styrene derivatives (1 equiv.) in dry methylene dichloride (13 ml/mmol), a solution of m-CPBA (1.2 equiv.) in dry methylene dichloride (13 ml/mmol) was added at room temperature. The reaction was monitored by TLC. After the reaction was finished, the reaction mixture was washed with NaHSO3 40% (w/v ) and then saturated with aqueous NaHCO3 solution. The organic layer was dried over Na2SO4, filtered, and then concentrated in vacuum.

Substrate screening

The freeze-dried recombinant whole cells (0.27 U/ml) were preincubated in 20 mM phosphate buffer (pH 8.0) at 25°C. Enantioselective hydrolysis reaction was initiated by adding 10 mM styrene oxide derivatives (see Fig. 2). The reaction was periodically monitored by withdrawing the samples from the reaction mixture. The samples were extracted with the same volume of ethyl acetate containing decane as an internal standard for GC analysis. The styrene oxide derivatives and their corresponding diols (see Fig. 3) in the ethyl acetate phase were analyzed by chiral GC and HPLC (Table 1).

Fig. 3
figure 3

Diols as products for enantioconvergent hydrolysis by the recombinant cell containing C. crescentus EH (1, phenyl-1,2-ethanediol; 2, o-chlorophenyl-1,2-ethanediol; 3, m-methylphenyl-1,2-ethanediol; 4, m-chlorophenyl-1,2-ethanediol; 5, m-nitrophenyl-1,2-ethanediol; 6, p-methylphenyl-1,2-ethanediol; 7, p-chlorophenyl-1,2-ethanediol; 8, p-nitrophenyl-1,2-ethanediol)

Full size image
Table 1 Analysis conditions for determination of enantiopurity and yield of enantiopure epoxides
Full size table

Determination of regioselectivity coefficient

The freeze-dried recombinant whole cells were preincubated in 20 mM phosphate buffer (pH 8.0) at 25°C. The reaction was initiated by adding 10 mM p-chlorostyrene oxide. The enantiopurities of the remaining epoxide and the formed diol were measured by chiral GC and HPLC, respectively. Regioselectivity coefficients were calculated by the computer program made by Faber and Kroutil (2002).

Co-solvent effect on enantioconvergent biohydrolysis

The freeze-dried recombinant whole cells (2 U/ml) were preincubated in 20 mM phosphate buffer (pH 8.0) containing 10% (v/v) various co-solvents at 25°C. The reaction was initiated by adding p-chlorostyrene oxide to 10 mM final concentration. The sample was periodically withdrawn to analyze the enantiopurity of the remaining epoxide and the formed diol.

Preparative-scale enantioconvergent hydrolysis of 4-chlorostyrene oxide

In a 300 ml reactor, 1g rac-7 was dissolved in 60 ml phosphate buffer (pH 8.0). The freeze-dried recombinant whole cells (400 U) were added, and then the mixture was vigorously stirred at 25°C. After the reaction, the diol product was extracted with ethyl acetate three times from the mixture, dried over MgSO4, concentrated in vacuum, and purified with flash chromatography (hexane:ethyl acetate = 2:1, by vol.).

Results and discussion

Substrate screening for enantioconvergent hydrolysis

The enantioselective hydrolysis of racemic styrene oxide derivatives (substrates 1–8) was investigated to evaluate the enantioselectivity of C. crescentus EH (Table 2). Among the eight styrene oxide derivatives, the cell containing C. crescentus EH exhibited enantioselectivity toward rac-1, 2, 7 and 8 substrates. Among these substrates, the hydrolysis of rac-7 occurred at a higher hydrolysis rate. After about 120 min, the hydrolysis of rac-7 (p-chlorostyrene oxide: pCSO) reached a 99% conversion ratio (Fig. 4). The C. crescentus EH hydrolyzed (S)-pCSO in preference to the other enantiomer, which is confirmed by comparison to authentic samples. The enantiomeric ratio (E) value, determined on the basis of the calculation equation suggested by Faber and Kroutil (2002), was about 30. The resulting vicinal diol 7d was shown to have more than 90% enantiopurity analyzed by the chiral HPLC. The formed diol 7d appeared to be also (R)-configuration determined by polarimeter ([α]14 D = −52.16, c = 1.02, CHCl3). These results represent that C. crescentus EH has enantioconvergent hydrolysis activity toward racemic pCSO.

Table 2 Reaction parameters of the biohydrolysis of 1– 8 (10 mM, 25°C) using the C. crescentu s EH
Full size table
Fig. 4
figure 4

Biohydrolysis of 10 mM rac-7 using the recombinant cell containing C. crescentus EH (The figure legends, ee_S and ee_P, represent the enantiomeric excess of substrate and product, respectively)

Full size image

Determination of regioselectivity coefficient

To obtain enantiopure diol with more than 90%ee via a hydrolysis of racemic pCSO, each enantiomer should be hydrolyzed in an enantioconvergent manner. As an epoxide substrate can be attacked at either carbon atom of an epoxide ring, we determined the regioselectivity for the (R)-and (S)-enantiomer of pCSO. When only (S)-enantiomer of rac-7 was supplied, (R)-diol was formed as the main product with the inversion of stereochemistry (Table 3). Hence, we could expect that C. crescentus EH mainly attacked at the benzylic carbon instead of terminal carbon of the epoxide ring of (S)-enantiomer. The (R)-diol was also formed from (R)-enantiomer of rac-7. The (R)-enantiomer of rac-7 was attacked at the less hindered carbon atom, resulting in the formation of (R)-7d with unchanged configuration. Hence, we determined the regioselectivity of C. crescentus EH toward pCSO as shown in Table 3. The retention-inversion ratio (RI ratio) and regioselectivity (α) were determined based on the program proposed by Faber and Kroutil (2002). More than 94% of (S)-7 was attacked at the benzylic position, whereas 99% of (R)-7 was attacked at the terminal carbon atom by C. crescentus EH. In conclusion, the cell biocatalyst containing C. crescentus EH appeared to possess a good to excellent enantioconvergency. The enantioconvergent bioconversion of rac-7 was further investigated to optimize the reaction condition for an enhanced enantiopurity and yield.

Table 3 Regioselectivity of C. crescentu s EH toward p-chlorostyrene oxide
Full size table

Effect of substrate concentration and co-solvent on enantioconvergent reaction

In order to set up an efficient process for the enantioconvergent biohydrolysis of rac-7, the effects of substrate concentrations on enantiopurity and yield were studied. As shown in Fig. 5, we could obtain more than 90% enantiopurity with up to 100 mM substrate although the reaction rate decreased and conversion ratio could not reach the 100% in the prolonged reaction time for the high substrate concentration. Reaction retardation at high substrate concentrations seemed to be due to product inhibition. To evaluate the product inhibition, hydrolysis reaction (5 mM initial substrate concentration) in the presence of various rac-7d concentrations was carried out. The hydrolysis of p-chlorostyrene oxide using cell containing C. crescentus EH was inhibited in the presence of above 40 mM rac-7d concentration (Fig. 6).

Fig. 5
figure 5

Effect of initial substrate concentration. The cell activity/substrate ratio was kept constant (0.2 U/mM for rac-7). The figure legend, ee_diol, represents the enantiomeric excess of diol product

Full size image
Fig. 6
figure 6

Influence of the initial 7d concentration on the biohydrolysis rate. The initial hydrolysis rate in the absence of the diol was 0.45 μmol/min mg of cell

Full size image

The effect of cosolvent addition on enantioconvergent biohydrolysis of rac-7 was examined since the organic solvent might be required to increase the solubility of rac-7. Various water-miscible organic solvents were tested in hydrolysis of 40 mM rac-7 at 25°C. The values of conversion ratios and 7d enantiopurity were measured. Except DMSO and glycerol, all water-miscible organic solvents tested inhibited the biohydrolysis rates of rac-7, and decreased the enantiopurity of the formed 7d. DMSO and glycerol did not affect on any positive/negative effect on hydrolysis of rac-7 (data not shown).

Preparative scale experiments

At the most appropriate experimental conditions for the biohydrolysis of rac-7, a preparative-scale batch reaction was carried out. In a 300 ml reactor, 400 U of cell biocatalyst containing C. crescentus EH was preincubated with 60 ml phosphate buffer (pH 8.0) at 25°C, and then, 1 g rac-7 was added, which corresponded to 16.8 g/l concentration. Due to low solubility of rac-7 in water, the reaction was done at the biphasic system where the substrate constitutes one phase by itself. The time course of the reaction was monitored by using chiral GC analysis. After 3 h, the reaction reached to the final conversion and then the aqueous solution containing the product 7d was extracted with ethyl acetate in order to isolate the resulting diol 7d. We could obtain 0.87g (R)-7d (overall yield = 78%) with an enantiopurity of 95%ee by using single cell containing EH of C. crescentus. To our best knowledge, it is the first report that the enantioconvergent biohydrolysis of racemic styrene oxide derivatives has been accomplished by using single bacterial EH. Recently, similar study on an enantioconvergent hydrolysis of racemic styrene oxide derivatives was carried out by using an isolated plant EH enzyme (Monterde et al. 2004). The bacterial EH has advantages over the plant EH in an enantioconvergent biohydrolysis since the bacterial EH can be efficiently expressed in E. coli. Hence, we could carry out the enantioconvergent hydrolysis reaction efficiently by using the whole-cell biocatalyst containing C. crescentus EH. The EH gene of C. crescentus was readily expressed in E. coli up to 75% of the total protein, corresponding to about 85% of the soluble protein estimated by densitometric analysis of SDS-PAGE gel (Hwang et al. 2006).

Conclusion

The enantioselective hydrolysis of eight racemic styrene oxide derivatives was investigated by using the recombinant cell containing C. crescentus EH. C. crescentus EH exhibited an enantioconvergent biohydrolysis activity on racemic p-chlorostyrene oxide, thus resulting in the formation of the corresponding (R)-diol with high enantiopurity up to 95%ee and a yield more than 50%. Each enantiomer of racemic p-chlorostyrene oxide was hydrolyzes with complementary enantioselectivity and regioselectivity. Enantiopure (R)-p-chlorophenyl-1,2-ethanediol with 72% preparative yield and enantiopurity as high as 95%ee was obtained via enantioconvergent biohydrolysis of racemic p-chlorostyrene oxide in preparative scale (16.8 g/l) by using single cell biocatalyst containing C. crescentus EH. Since enantiopure styrene oxide derivatives and their corresponding diols are valuable synthetic intermediates for pharmaceuticals, single recombinant cell biocatalyst-catalyzed enantioconvergent biotransformation can have industrial applications.