Engineering the epoxide hydrolase from Agromyces mediolanus for enhanced enantioselectivity and activity in the kinetic resolution of racemic epichlorohydrin

Abstract

The biocatalytic production of enantiopure epichlorohydrin (ECH) has been steadily attracting more attention. For industrial applications, it is important to obtain an epoxide hydrolase (EH) that possesses the desired enantioselectivity. Site-saturation and site-directed mutagenesis of the Ser207, Asn240 and Trp182 positions were used to generate variants of EH from Agromyces mediolanus ZJB120203 with enhanced enantioselectivity for the kinetic resolution of racemic ECH. The best variant, VDF (W182F/S207V/N240D), displayed a 7-fold enhanced enantioselectivity toward racemic ECH, with an increase in the enantiomeric ratio value (E value) preferring the (R)-ECH enantiomer from 12.9 of wild-type to 90.0, as well as a 1.7-fold improvement in activity. Furthermore, we successfully applied the created recombinant Escherichia coli whole cells expressing variant VDF in the kinetic resolution of racemic ECH. Enantiopure (S)-ECH could be obtained with an enantiopurity of >99% ee and a yield of 40.5% from 450 mM racemic ECH, which is better than those of other reported EHs. These results demonstrated that the EH obtained in this study could be applied for the efficient resolution of racemic ECH.

---

Introduction

Epoxide hydrolases (EHs, EC 3.3.2.3) are ubiquitous enzymes found in mammals, insects, plants, and microorganisms, and can catalyze the hydrolysis of epoxides to the corresponding vicinal diols through the addition of a water molecule.^1,2 Most EHs for which sequence information is presently available are members of the α/β-hydrolase fold family and share a main domain consisting of a central β sheet surrounded by α helices, with a variable cap domain sitting on top.^2,3 The catalytic triad is formed of a nucleophile (Asp), which attacks the epoxide and forms a covalent ester-intermediate, a catalytic histidine and a carboxylic acid (Asp or Glu) that together subsequently activate a water molecule and hydrolyze the ester bond to release the product.^4,5

Over the past 20 years, EHs have been explored for the preparation of optically active epoxides and diols. The most common application of EHs is the preparation of optically pure epoxides by the enantioselective hydrolysis of a racemic epoxide.⁶ Enantiopure epichlorohydrin (ECH) is a valuable epoxide intermediate for producing optically active pharmaceuticals. ECH has been used for the synthesis of β-blockers, L-carnitine, ferroelectric liquid crystals, and atorvastatin.^7,8 Several enantioselective EHs have been tested in kinetic resolutions to obtain an enantiopure ECH. However, the yield of the remaining ECH and the enantiopurity of the diol are often not very high, due to the low enantioselectivity of the EH.^9–11 Therefore, it is desirable to improve the enantioselectivity of EHs.

It has been reported that site-directed mutagenesis, saturation mutagenesis, the error-prone polymerase chain reaction and DNA shuffling have all been applied to enhance the activity and enantioselectivity of EHs.^12–16 Using a high throughput screening method, EH mutants with improved functions have been obtained, which shows a good potential for using the EHs in biotechnological applications.^{12,13,15–18}

In our previous study, a novel EH from Agromyces mediolanus ZJB120203 (AmEH) was cloned and expressed in E. coli. The kinetic mechanism of AmEH was solved in detail for both enantiomers of ECH; also wild-type (WT) AmEH was found to be moderately enantioselective toward (R)-ECH (Fig. S1†).⁵ The variants constructed in this study were evaluated for altered enantioselectivity and activity toward racemic ECH. Furthermore, saturation mutagenesis was used to engineer AmEH with altered enantioselectivity and enzymatic activity toward racemic ECH. The residues, especially those with a near-reaching effect on the substrate binding that could be reliably predicted by rational design, were identified. The production of optical pure (S)-ECH was also studied in depth and then optimized.

Results and discussion

Construction and screening of mutant libraries of AmEH

In order to choose appropriate randomization sites, we performed the homologous modelling of AmEH using the EH structure (PDB accession no. 4i19) as the template, and further induced the docking of (R)-ECH and (S)-ECH into the obtained model. It is well recognized that residues adjacent to the substrate play a critical role in catalytic activity and enantioselectivity, and thus can act as the target sites for rational design.¹⁹ HotSpot Wizard could give predictions of hot spots (usually the amino acid residues that mediate the substrate binding, the transition-state stabilization, or the product release are selected as the ‘hot spots’) in enzymes based on structural, functional and evolutionary information obtained from different databases.²⁰ In the case of AmEH, HotSpot Wizard was used to list the annotated residues ordered by their estimated mutability. Based on different considerations, Trp182, Ser207, Ser233, Asn240, Arg313, Phe318 and Arg338 were chosen for the mutagenesis and catalytic property analysis (Fig. 1).

Fig. 1 Sites for the mutant were chosen as described in the text and are color coded with magenta.

Seven libraries were generated using saturation mutagenesis, ideally each containing all possible amino acid substitutions at one of these positions. These mutant libraries were screened for hydrolysis activity towards ECH using 4-(4-nitrobenzyl)pyridine (NBP), which reacts with terminal epoxides to form a blue adduct that can be qualitatively analysed (Fig. 2). 2000 clones were screened, and approximately 50% of them showed appreciable activity, as revealed by the pretests, and subsequently we obtained the respective conversion and ee values through applying automated GC analyses. The cell-free extracts of WT AmEH and its variants were purified by one-step nickel affinity chromatography on a Ni-NTA resin. Using SDS-PAGE, it was found that there was no difference in the molecular mass (about 43 kDa) of the WT compared to the mutant AmEH, which is similar to results in the previous report (Fig. S2†).⁵

Fig. 2 Activity screening of the site-saturation library for ECH in 96-well plates using NBP assay.

For most of the mutants, the yields of (S)-ECH declined, indicating that their enantioselectivities decreased towards racemic ECH. No positive mutant with an improved enantioselectivity was isolated at the Arg313, Arg338, Ser233 and Phe318 positions. A series of variants from the libraries of the other three sites (Ser207, Trp182 and Asn240) showed higher enhancements in enantioselectivity over the wild-type enzyme (data not shown). Among the mutants, three variants with the highest enantioselectivities at each position, namely S207V, W182F and N240D, were selected and subsequently shown to have not only increased total activity, but also increased (R)-enantioselectivity (Table 1). The introduction of Val instead of a Ser at position 207 raised the E-value from 12.9 to 20.3, while the introduction of Asp instead of an Asn at position 240 raised the E-value from 12.9 to 21.4.

Table 1 Specific activities and enantioselectivities of wild-type AmEH and variants toward racemic ECH

Enzyme	Racemic ECH		Enantioselectivity
	Specific activity (U mg^−1)	Fold	E-value^a	Fold
a E values are calculated from the kcat/Km values for the separate enantiomers.
Wild-type	24.6	1.0	12.9	1.0
W182F	26.1	1.1	16.9	1.3
S207V	38.9	1.6	20.3	1.6
N240D	59.7	2.4	21.4	1.7
W182F/S207V	30.1	1.2	28.5	2.2
VD	69.0	2.8	48.1	3.7
VDF	42.6	1.7	90.0	7.0

---

In the EHs, the residue next to the active site nucleophilic amino acid is not strictly conserved, but it is often a tryptophan, such as the Trp182 in AmEH.^21,22 The corresponding residue was Phe108 in the EH from Agrobacterium radiobacter AD1, which was investigated and confirmed by saturation mutagenesis. Mutants of EH from A. radiobacter AD1 at this position have great impact on the activity and enantioselectivity. Mutant F108W of the EH from A. radiobacter AD1 had decreased enantioselectivities and activities for all the test aromatic substrates; whereas, for the two aliphatic epoxides, the activities were lower than for the WT.¹⁷ The mutation of Trp182 of AmEH had improved enantioselectivity and activity for ECH compared to the WT. The changes in both enantioselectivity and activity as a result of mutations at this position provide further evidence that the side chain of this conserved residue is involved in substrate binding. It is proposed that the conservative tyrosine residues stabilize the transition state towards the formation of the intermediate by protonation of the leaving oxygen group. Based on the previous report,²³ Asn240 flanking the Tyr239 was also proposed as a potential hot spot for enhancing the activity and enantioselectivity. The mutagenesis studies revealed that N240D had a higher activity (2.4-fold) than the WT, while its E-value was raised from 12.9 to 21.4. Based on the structure analysis, it was found that another residue, Ser207, was positioned in a loop, and closely located to the catalytic residues, approximately at a distance of 8.3 Å. The mutation of Ser207 therefore ought to have some impact on the catalytic efficiency of AmEH. For ECH, the mutant S207V had a higher enantioselectivity and activity, which showed that Ser207 plays a role in the AmEH-catalyzed hydrolysis of ECH. These studies showed that residues Trp182, Ser207 and Asn240 were enantioselectivity “hot spots”, with several mutants affording product (S)-ECH with a high ee; also, residue Asn240 exerted a greater effect on the enzyme activity than residues Ser207 and Trp182.

Combinations of mutations can have additive effects in the case of enantioselectivity (Table 1). For the AmEH mutants, two and three pronounced additive effects on enantioselectivity were found, except the variant W182F/N240D, which had a decreased enantioselectivity and activity for ECH (data not shown). Several variants – W182F/S207V, S207V/N240D and W182F/S207V/N240D – isolated from the second and third-generation library, VD and VDF, are particularly efficient in the kinetic resolution of ECH. VD contains two active-site substitutions: S207V and N240D, while VDF also contains S207V and N240D, but in combination with W182F. The variants VD increased the enantioselectivity from 12.9 to 48.1 towards racemic ECH. However, VDF was the most distinct, with the E value increasing from 12.9 for the wild-type enzyme to 90.0, a 7-fold enhancement, which is the highest enantioselectivity described for ECH so far. In addition, VDF also demonstrated a 1.7-fold improvement, compared with the WT AmEH, in specific activity. The significant improvement in the enantiomeric purity and the enhanced catalytic rate make VDF an AmEH variant of considerable interest for preparing enantiomerically pure (S)-ECH.

Kinetics parameters analysis of selected mutants

Enhancing the enantioselectivity of a kinetic resolution can be carried out by increasing the reaction rate of one enantiomer, or by decreasing the rate of the mirror-image compound.¹² To understand the basis of the observed changes in enantioselectivity, the steady-state kinetics parameters of the mutants were determined for ECH using the enantiopure form as the substrate. Kinetics parameters (K_m, V_max, k_cat and k_cat/K_m) of AmEH and variants towards (R)-ECH and (S)-ECH were determined, and are summarized in Table 2. The results showed that the overall catalytic efficiencies (k_cat/K_m) of the mutant AmEHs were lower than for the WT enzymes for (S)-ECH, while for mutant AmEH, the k_cat/K_m values were higher than for the WT enzyme for (R)-ECH. VDF gave a k_cat/K_m value for the (R)-enantiomer that was 2.0-fold greater than that of the WT. However, the k_cat/K_m of VDF for (S)-ECH was reduced to about one-quarter of the WT. The changes in the kinetics parameters for VDF led to a significant enhancement in the E-value. For mutant VD, there was also a significant increase in k_cat/K_m on (R)-ECH and a decrease in k_cat/K_m on (S)-ECH. For the mutant N240D, the increase in enantioselectivity was mainly due to the increased k_cat for its preferred substrate, i.e. the (R)-enantiomer.

Table 2 Kinetic analysis of the selected mutants of AmEH towards enantiopure ECH

Enzyme	S-ECH				R-ECH
	Km (mM)	Vmax (μmol min^−1 mg^−1)	kcat (s^−1)	kcat/Km (mM^−1 s^−1)	Km (mM)	Vmax (μmol min^−1 mg^−1)	kcat (s^−1)	kcat/Km (mM^−1 s^−1)
WT	161.4	7.9	5.7	3.5 × 10^−2	56.6	35.6	25.7	4.5 × 10^−1
W182F	130.7	5.9	4.2	3.2 × 10^−2	48.3	36.2	26.1	5.4 × 10^−1
S207V	108.8	4.8	3.5	3.2 × 10^−2	62.6	56.4	40.7	6.5 × 10^−1
N240D	141.8	6.8	4.9	3.5 × 10^−2	87.3	90.3	65.2	7.5 × 10^−1
W182F/S207V	137.1	3.8	2.7	2.0 × 10^−2	52.1	41.1	29.7	5.7 × 10^−1
VD	134.0	2.9	2.1	1.6 × 10^−2	98.7	105.3	76.0	7.7 × 10^−1
VDF	105.1	1.5	1.1	1.0 × 10^−2	48.8	60.5	43.7	9.0 × 10^−1

---

Moreover, the mutations also caused changes in the V_max values. The N240D mutant had a V_max value for (R)-ECH that was approximately 2.5-fold that of the WT; whereas, it had a lower V_max value for (S)-ECH. These results clearly support the hypothesis concerning the two roles of Asp240, which are the enhancement of reactivity in (R)-ECH and suppression of the reactivity in (S)-ECH. Furthermore, the VD double mutant had a V_max value for (R)-ECH that was approximately 1.2-fold that of the N240D mutant.

Homology structural modelling and substrate docking

The three-dimensional coordinates of the AmEH variants with the most significant changes in ECH hydrolysis, N240D, VD and VDF, were constructed, and molecular docking was further carried out to try to deeply understand the interactions between the enantiomers of ECH and the amino acid residues in the AmEH active cavity (Fig. 3, Table S2†). Docking of the ECH into the active sites of the various mutants and the WT suggest the reshaping of the binding pocket. The catalytic efficacy of the enzyme is dependent on how often the nucleophile and electrophile are present in near attack conformations (NACs). The through-space distance, d, between the attacking O-atom of Asp181 and the epoxide C-atom was thought to be of particular importance. Also, the angle from the Asp181 oxygen via the attacked epoxide carbon to the epoxide oxygen (α₁) and the Asp181 oxygen via the attacked epoxide carbon to the other epoxide carbon (α₂) were considered.²⁴

Fig. 3 Docking of (R)-ECH and (S)-ECH to the binding pocket of mutant AmEHs. (a) Variant N240D docked with (R)-ECH and (S)-ECH; (b) variant VD docked with (R)-ECH and (S)-ECH; (c) variant VDF docked with (R)-ECH and (S)-ECH.

As Table S2† shows, the favoured enantiomers have a relatively shorter distance (d value) and larger angles (α₁ and α₂ value), corresponding to NACs, or more generally to the productive positions, as expected. The differences in the modelled distance, Δd, for the (R)-ECH and (S)-ECH were well consistent with the measured E values.^7,25 In the WT AmEH, the Δd value was expected to be only 0.3 Å, with the preferred (R)-ECH slightly closer to the attacking Asp181.⁵ In the mutant VD, the Δd value increased to 0.9 Å. For the preferred (R)-ECH, the value for d reduced to 3.2 Å, which may be responsible for the about 2.8-fold increase in the activity for this variant. However, for the disfavoured (S)-ECH, the value of d increased to 4.1 Å. In the case of the highly enantioselective VDF, the result was remarkably different; and Δd increased to 1.1 Å. For the preferred (R)-ECH, the value for d did not clearly change (d = 3.3 Å), but the activated (S)-ECH was positioned much further away (d = 4.4 Å), which disfavours a ring opening nucleophilic attack. Additionally, the calculation results show that there was a bigger difference for WT AmEH in the angle values between (R)-ECH (142.0° and 136.8°) and (S)-ECH (55.2° and 64.7°), which leads to the best mutant VDF being more active on (R)-ECH. These results suggest that structural changes in the binding pocket imposed by the evolutionary process are predicted to make it much more difficult for the disfavoured (S)-ECH to be activated by Tyr308 and at the same time to be positioned close enough to Asp181 for rapid nucleophilic attack.¹²

Biocatalytic synthesis of chiral ECH using recombinant cells of variant VDF

To test the potential applications of the mutant in the enzymatic hydrolysis of ECH, the recombinant E. coli (4.5 g dcw L⁻¹) used as a biocatalyst was incubated with racemic ECH at substrate concentrations of 75 and 150 mM. The time-profile of the kinetic resolution of racemic ECH is shown in Fig. S3.† As a result, enantiopure (S)-ECH with more than a 99% ee was obtained from the 75 mM and 150 mM racemic ECH, with yields from 45.8% to 44.1% (theoretical yield = 50%). Considering industrial applicability, the low concentration of substrate may be a major restriction for the enzymatic reaction. Thus, we verified if the reaction could go beyond 150 mM ECH while still keeping the optical yield and enantioselectivity the same as the amount of enzyme increased. As shown in Fig. S3,† 8.5 and 13.5 g dcw L⁻¹ recombinant E. coli were incubated with racemic ECH ranging from 150 to 300 mM at 30 °C and pH 8.0. It was found that enantiopure (S)-ECH with a high optical purity (ee > 99%) was readily obtained from 150 to 300 mM racemic ECH within 80 min, and also it was found that more AmEH in the reaction could help to shorten the reaction time without significantly affecting the optical yield or enantioselectivity. When the kinetic resolution was conducted by the recombinant E. coli at an ECH concentration of 450 mM, chiral (S)-ECH with an ee higher than 99% was obtained at a 40.5% yield (theoretical yield = 50%) at 90 min (Fig. 4). In the best known examples, EH from Novosphingobium aromaticivorans has been cloned and applied to the resolution of (R,S)-ECH (250 mM) to provide (S)-ECH with only a 17.3% yield and a greater than 99% ee (Table 3).¹¹ In the last few decades, the use of EHs in the kinetic resolution of racemic ECH for the synthesis of chiral ECH has been hindered by the insufficient enantioselectivity of existing EHs,^5,7,9,26 and as ECH is a very unstable epoxide in aqueous buffer.¹⁰ These two disadvantages lead to a poor yield of enantiopure ECH from kinetic resolution in the aqueous phase. The variant VDF was used to catalyze the enantioselective hydrolysis of racemic ECH to give (S)-ECH with an enantiomeric ratio (E) of 90.0 in aqueous buffer, better than any reported EHs, and which makes it very competitive and promising for practical applications in the production of (S)-ECH. However, when the concentration reached 500 mM, the optical purity of (S)-ECH only rose to 90.1%, even when the reaction time was prolonged to 200 min. When the variant VD (S207V/N240D) was used in the production of (S)-ECH by the kinetic resolution of racemic ECH, the ee of (S)-ECH was able to reach 99% at high concentration of 750 mM racemic ECH, although the yield was not excellent (only about 21.3%). This result may be due to the mutant VDF having a higher enantioselectivity, but lower activity, than mutant VD. To the best of our knowledge, the substrate and product inhibition is the limiting factor preventing large-scale applications in many biotransformation reactions.^27,28 This may be ascribed to the enzyme inactivation caused by a high concentration of the substrate and product. This type of inhibition has been reported in the asymmetric reduction of ECH and can be overcome by utilizing a substrate fed-batch strategy and an aqueous–organic solvent two-phase system.^29,30 EHs should also be tuned for better operational stability for application in non-natural reaction conditions. It is of great interest to improve the stability of EHs in organic solvents and to prevent the interfacial deactivation of EHs in aqueous–organic two phase systems by molecular engineering of EHs.

Fig. 4 Enantioselective hydrolysis of racemic ECH by the recombinant E. coli variant VDF. The reaction was performed at 30 °C in 200 mM sodium phosphate buffer (pH 8.0), 450 mM ECH and 13.5 g dcw L⁻¹ recombinant E. coli. Samples were removed at set time intervals, and the ECH concentration and the optical purity of the (S)-ECH were determined by chiral GC.

Table 3 The comparison of the kinetic resolution of known EHs toward racemic ECH

EH source	ECH conc. (mM)	Temperature (°C)	pH	ee (%)	Reaction medium	Final yield^a (%)	References
a The final yield of (S)-ECH was determined by GC.
Aspergillus niger	60	27	7.5	100/(S)	Organic solvents	20.0	10
Novosphingobium aromaticivorans	50	30	8.0	>99/(S)	Aqueous buffers	20.7	11
Novosphingobium aromaticivorans	250	30	8.0	>99/(S)	Aqueous buffers	17.3	11
Aspergillus niger	64	30	8.0	98/(S)	Organic solvents	17.5	27
Agromyces mediolanus (AmEH)	64	30	8.0	>99/(S)	Aqueous buffers	21.5	5
Mutant AmEH (W182F/S207V/N240D)	75	30	8.0	>99/(S)	Aqueous buffers	45.8	This study
	150	30	8.0	>99/(S)	Aqueous buffers	44.1	This study
	300	30	8.0	>99/(S)	Aqueous buffers	43.7	This study
	450	30	8.0	>99/(S)	Aqueous buffers	40.5	This study

---

Experimental section

Materials, bacterial strains and plasmids

(R,S)-ECH, (R)-ECH, (S)-ECH and NBP were purchased from Sigma-Aldrich (St. Louis, MO, USA). All other chemicals used were of analytical grade and are commercially available.

PrimeSTAR HS DNA polymerase, restriction endonucleases, PCR reagents and the genomic extraction kit were purchased from Takara Biotechnology (Dalian, China) Co., Ltd. DNA sequencing was performed by Sangon Biotechnology (Shanghai, China) Co., Ltd. The DNA gel extraction, plasmid extraction and PCR product purification kits were purchased from Axygen Biotechnology (Hangzhou, China) Co., Ltd. E. coli BL21(DE3) and pET-28a(+) were used as the host strain and vector for the expression experiments. The plasmid pET28a-AmEH with the gene encoding EH from A. mediolanus ZJB120203 was used as the template for the saturation mutagenesis. The E. coli strains were grown at 37 °C in lysogeny broth (LB) medium (0.5% yeast extract, 1% tryptone and 1% NaCl), and supplemented with kanamycin (50 μg ml⁻¹).

Construction of the saturation mutagenesis libraries

Gene libraries encoding all the possible amino acids at positions 182, 207, 233, 240, 313, 318 and 338 of the AmEH gene in the pET28a-ameh (pET-28a hosting the gene encoding AmEH) were constructed by replacing the target codon with NNN (N = A/T/C/G in a 1 : 1:1 : 1 ratio) via a polymerase chain reaction (PCR). The primers used for the saturation mutagenesis at each site are listed in Table S1.† The mutagenesis PCR was performed with PrimeSTAR HS DNA polymerase using a program of 3 min at 98 °C, followed by 27 cycles of 98 °C for 10 s, 55 °C for 15 s, 72 °C for 7 min and a final extension at 72 °C for 10 min. The restriction enzyme Dpn I was directly added to the PCR reaction tube to degrade the methylated template for 3 h at 37 °C. The mixture was transformed into E. coli BL21(DE3) and cultivated on an LB agar plate containing 50 μg ml⁻¹ kanamycin. The transformants were confirmed by DNA sequencing using the T7 forward and reverse primer. From three positive mutants, a second round of site-directed mutagenesis was performed at the two other respective amino acid residues. A mutant with improved enzyme activity and enantioselectivity was obtained by site-directed mutagenesis and screening.

Protein expression and purification

The recombinant E. coli was cultivated at 37 °C and in a 200 rpm orbital shaking in 100 ml of LB broth containing 50 μg ml⁻¹ kanamycin. After the cells reached an OD₆₀₀ of 0.6, the heterologous expression of the protein was induced fully with 0.1 mM isopropyl-β-d-thiogalactoside (IPTG) and the culture was continuously incubated at 28 °C for an additional 10 h. The cells were harvested, disrupted by sonication using an ultrasonic processor UP200S (Hielscher, Teltow, Germany) and centrifuged for 15 min at 4 °C and at 10 000×g. The WT AmEH and mutant variants were purified by the AKTA™ explorer system (Amersham Biosciences Corp., Uppsala, Sweden) with a 16 mmD/100 mmL POROS MC 20 μm column (Applied Biosystems Co., USA). His-tagged enzyme was bound to the resin in equilibrating buffer (20 mM sodium phosphate buffer, pH 8.0, containing 0.5 M sodium chloride, and 20 mM imidazole). Unbound and weakly bound proteins were washed out. His-tagged enzyme was eluted by a buffer containing 500 mM imidazole.³¹ The eluted protein was pooled and dialyzed overnight against 20 mM sodium phosphate buffer (pH 8.0), and then stored at 4 °C.³² The protein concentration was quantified via the Bradford method using bovine serum albumin (Sigma-Aldrich, St. Louis, MO, USA) as the standard.³³ The molecular mass of the denatured protein was determined by 12% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), as described by Laemmli.³⁴ The protein was stained with Coomassie brilliant blue G-250 (Aladdin, Shanghai, China).

Determination of the kinetics properties

The kinetic analysis of the WT and AmEH variants for the kinetic resolution of racemic ECH was performed by measuring their initial rates towards varied concentration of (R)-ECH and (S)-ECH, respectively. The maximum reaction rate V_max and the Michaelis constant K_m were determined using the Lineweaver–Burk plotting method based on the Michaelis–Menten equation: 1/v = K_m/V_max[S] + 1/V_max, where v is the initial velocity and [S] is the substrate concentration. The catalytic constant k_cat was deduced using a molecular mass of 43 kDa. All the parameters were calculated using the mean values from three independent experiments. The E-value, corresponding to the enantioselectivity of the enzyme, was derived from the ratio of k_cat/K_m of the separate enantiomers.

Activity screening

The constructed mutant libraries were screened for EH activity using a blue assay in a 96-well plate developed by Cedrone.³⁵ All the colonies from the agar plate prepared in the earlier step were picked up using a 10 μl micropipette tip and then directly transferred into individual wells in a 96-well plate containing 1 ml of LB medium with kanamycin. The plates were incubated at 37 °C for 2 h, and induced with 0.1 mM IPTG for the heterologous expression of EH. The expression plates were incubated overnight at 28 °C on a shaker. After centrifuging for 20 min at 3000×g, the cell pellets were resuspended in 200 μl 50 mM phosphate buffer (pH 8.0). Aliquots (25 μl) of the supernatant were transferred to the new microtiter plates containing 25 μl of epoxide substrate solution (5 mM racemic ECH dissolved in 50 mM phosphate buffer). After incubation at 30 °C for 20 min, the reaction was stopped by the addition of 25 μl of 100 mM 4-(4-nitrobenzyl) pyridine in a 80/20 v/v ethylene glycol/ethanol mixture. The reaction of NBP with ECH was run for 20 min at 80 °C, and 50 μl ethanol and 25 μl 1 M K₂CO₃ were added. The K₂CO₃ interacts with the remaining ECH and a blue colour develops. The darker the blue colour of the solution, the more ECH is left unconverted in the well, indicating either inefficient mutant variants or the absence of enzyme.

Activity assay and analytical methods

The standard assay was performed with racemic ECH and the recombinant EH mixed in 10 ml phosphate buffer (pH 8.0). The reaction mixture reaction was incubated at 30 °C at 150 rpm. After 5 min, a 1.0 ml biotransformation sample was taken and centrifuged at 12 000×g for 5 min, and then 400 μl of biotransformation sample was extracted with 1 ml of ethyl acetate and centrifuged at 10 000×g for 5 min. The organic layer was separated and dried with anhydrous sodium sulfate. The reaction mixture was analyzed by GC and chiral GC to determine the enzyme activity and ee value, respectively. The concentrations of the ECH was analyzed using an Agilent 7890A gas chromatograph (Agilent, Santa Clara, CA) with a capillary HP-5 column (0.35 mm × 30 m × 0.25 μl film thickness). Temperature program: 80 °C for 4 min, and then 100 °C at 10 °C min⁻¹. Retention times: 4.7 min for racemic ECH, 5.8 min for 1-chlorohexane (internal standard). The enantioselectivity was determined by GC equipped with a chiral capillary BGB-175 column (0.25 mm × 30 m × 0.25 μl film thickness). The injector, detector and column temperatures were set as 220, 220 and 90 °C, respectively. The retention times of (S)-ECH and (R)-ECH were 4.7 and 4.9 min, respectively.⁵ The ee was derived from the remaining epoxide of the two enantiomers [ee (%) = (S − R)/(S + R) × 100]. One unit of EH activity was defined as the amount of enzyme required to convert 1 μmol ECH at 30 °C. The specific enzyme activities were defined as units per mg enzyme per min.

Biocatalytic synthesis of chiral ECH by the whole-cell biocatalyst

The enantioselective hydrolysis of racemic ECH was conducted in 50 ml screw-capped vials with a working volume of 10 ml. The cultured cells were suspended in a 200 mM sodium phosphate buffer (pH 8.0), and the kinetic resolution reactions were initiated with the addition of various concentrations of racemic ECH at 30 °C in a shaking incubator (200 rpm). The reaction was then stopped by extraction twice with an equal volume of ethyl acetate. The progression of the enantioselective hydrolysis reaction was analyzed through the analysis of samples withdrawn periodically from the reaction mixtures.

Molecular modelling and analysis

The crystal structure of an EH from Streptomyces carzinostaticus (PDB accession no. 4i19), which was 36% identical to AmEH, was chosen as the template. The three-dimensional homology model of the WT and AmEH mutants were generated by Modeller 9.12 in Discovery studio (DS) 2.1 (Accelrys Software, San Diego, USA) with loops refined sufficiently.^36,37 Several models were obtained and validated by Procheck and Profile-3D. Finally, the best quality model was chosen for further calculations, molecular modelling, and docking studies by Autodock 4.0.³⁸ The visualization was performed with PyMOL program version 1.2r1 (http://www.pymol.org).

Conclusions

In summary, the enantioselectivity and activity of AmEH toward (R, S)-ECH were improved by a structure-based rational design approach. Successive rounds of saturation mutagenesis resulted in an increase in enantioselectivity from E = 12.9 for the WT enzyme to E = 90.0 for the best variant (W182F/S207V/N240D). Enantiopure (S)-ECH could be readily prepared with a high enantiopurity more than 99% ee and a yield of 40.5% using the three-point mutant with 450 mM (R,S)-ECH as the substrate. These results indicated that the engineered EH is a promising biocatalyst for the kinetic resolution of racemic ECH and other substrates.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (no. 21176224), the National High Technology Research and Development Program of China (no. 2012AA022201B), and 973 Program (no. 2011CB710806).

Notes and references

1. M. Arand, A. Cronin, M. Adamska and F. Oesch, Phase Ii Conjugation Enzymes and Transport Systems, 2005, vol. 400, pp. 569–588.
2. Z. Q. Liu, Y. Li, Y. Y. Xu, L. F. Ping and Y. G. Zheng, Appl. Microbiol. Biotechnol., 2007, 74, 99–106.
3. M. Widersten, A. Gurell and D. Lindberg, Biochim. Biophys. Acta, Gen. Subj., 2010, 1800, 316–326.
4. N. Bala and S. S. Chimni, Tetrahedron: Asymmetry, 2010, 21, 2879–2898.
5. F. Xue, Z. Q. Liu, S. P. Zou, N. W. Wan, W. Y. Zhu, Q. Zhu and Y. G. Zheng, Process Biochem., 2014, 49, 409–417.
6. M. Kotik, A. Archelas and R. Wohlgemuth, Curr. Org. Chem., 2012, 16, 451–482.
7. Z. Q. Liu, L. P. Zhang, F. Cheng, L. T. Ruan, Z. C. Hu, Y. G. Zheng and Y. C. Shen, Catal. Commun., 2011, 16, 133–139.
8. N. W. Wan, Z. Q. Liu, K. Huang, Z. Y. Shen, F. Xue, Y. G. Zheng and Y. C. Shen, RSC Adv., 2014, 4, 64027–64031.
9. H. S. Kim, J. H. Lee, S. Park and E. Y. Lee, Biotechnol. Bioprocess Eng., 2004, 9, 62–64.
10. W. J. Choi, E. Y. Lee, S. J. Yoon, S. T. Yang and C. Y. Choi, J. Biosci. Bioeng., 1999, 88, 339–341.
11. J. H. Woo, Y. O. Hwang, J. H. Kang, H. S. Lee, S. J. Kim and S. G. Kang, J. Biosci. Bioeng., 2010, 110, 295–297.
12. M. T. Reetz, M. Bocola, L. W. Wang, J. Sanchis, A. Cronin, M. Arand, J. Y. Zou, A. Archelas, A. L. Bottalla, A. Naworyta and S. L. Mowbray, J. Am. Chem. Soc., 2009, 131, 7334–7343.
13. M. Kotik, W. Zhao, G. Lacazio and A. Archelas, J. Mol. Catal. B: Enzym., 2013, 91, 44–51.
14. M. H. Woo, H. S. Kim and E. Y. Lee, J. Ind. Eng. Chem., 2012, 18, 384–391.
15. M. T. Reetz, C. Torre, A. Eipper, R. Lohmer, M. Hermes, B. Brunner, A. Maichele, M. Bocola, M. Arand, A. Cronin, Y. Genzel, A. Archelas and R. Furstoss, Org. Lett., 2004, 6, 177–180.
16. J. H. L. Spelberg, R. Rink, A. Archelas, R. Furstoss and D. B. Janssen, Adv. Synth. Catal., 2002, 344, 980–985.
17. B. van Loo, J. Kingma, G. Heyman, A. Wittenaar, J. H. L. Spelberg, T. Sonke and D. B. Janssen, Enzyme Microb. Technol., 2009, 44, 145–153.
18. M. T. Reetz and H. B. Zheng, ChemBioChem, 2011, 12, 1529–1535.
19. K. L. Morley and R. J. Kazlauskas, Trends Biotechnol., 2005, 23, 231–237.
20. A. Pavelka, E. Chovancova and J. Damborsky, Nucleic Acids Res., 2009, 37, W376–W383.
21. J. H. Woo, J. H. Kang, S. Kang, Y. O. Hwang and S. J. Kim, Appl. Microbiol. Biotechnol., 2009, 82, 873–881.
22. H. Visser, S. Vreugdenhil, J. A. M. de Bont and J. C. Verdoes, Appl. Microbiol. Biotechnol., 2000, 53, 415–419.
23. X. D. Kong, Q. Ma, J. H. Zhou, B. B. Zeng and J. H. Xu, Angew. Chem., Int. Ed., 2014, 53, 6641–6644.
24. B. Schiott and T. C. Bruice, J. Am. Chem. Soc., 2002, 124, 14558–14570.
25. J. Zhao, Y. Y. Chu, A. T. Li, X. Ju, X. D. Kong, J. Pan, Y. Tang and J. H. Xu, Adv. Synth. Catal., 2011, 353, 1510–1518.
26. E. Y. Lee, J. Ind. Eng. Chem., 2007, 13, 159–162.
27. H. X. Jin, Z. C. Hu and Y. G. Zheng, J. Biosci., 2012, 37, 695–702.
28. H. X. Jin, Z. Q. Liu, Z. C. Hu and Y. G. Zheng, Eng. Life Sci., 2013, 13, 385–392.
29. P. F. Gong and J. H. Xu, Enzyme Microb. Technol., 2005, 36, 252–257.
30. H. Baldascini, K. J. Ganzeveld, D. B. Janssen and A. A. C. M. Beenackers, Biotechnol. Bioeng., 2001, 73, 44–54.
31. Z. Q. Liu, Y. Gosser, P. J. Baker, Y. Ravee, Z. Y. Lu, G. Alemu, H. G. Li, G. L. Butterfoss, X. P. Kong, R. Gross and J. K. Montclare, J. Am. Chem. Soc., 2009, 131, 15711–15716.
32. F. Xue, Z. Q. Liu, N. W. Wan and Y. G. Zheng, Appl. Biochem. Biotechnol., 2014, 174, 352–364.
33. M. Bradford, Anal. Biochem., 1976, 72, 248–254.
34. U. K. Laemmli, Nature, 1970, 227, 680–685.
35. F. Cedrone, T. Bhatnagar and J. C. Baratti, Biotechnol. Lett., 2005, 27, 1921–1927.
36. A. Fiser and A. Sali, Methods Enzymol., 2003, 374, 461–491.
37. C. Zhang, J. Li, B. Yang, F. He, S. Y. Yang, X. Q. Yu and Q. Wang, RSC Adv., 2014, 4, 27526–27531.
38. F. Osterberg, G. M. Morris, M. F. Sanner, A. J. Olson and D. S. Goodsell, Proteins, 2002, 46, 34–40.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra02492g

---