Sources and reaction mechanism of EHs
Sources of EHs
EHs have been found in many prokaryotic and eukaryotic organisms, including bacteria, fungi, yeast, plants, insects, fish and mammals (Figure 3). The physiological role of EHs appears to be manifold; they are involved in the detoxification of potentially harmful, naturally occurring or anthropogenic epoxides (Decker et al., 2009), in lipid metabolism in plants and animals (Newman et al., 2005; Morisseau, 2013), and in the metabolism of juvenile hormones in insects (Newman et al., 2005). Recently, a new role of EHs in the biosynthesis of two antibiotics has been established in two Streptomyces strains (Lin et al., 2006; Lin et al., 2010). The substrates of EHs are structurally very diverse, representing a broad range of metabolites and xenobiotics. The substrate specificity of individual EHs appears to be diverse as well, being in many cases broad, but occasionally limited to a few available epoxidic compounds (Elfström and Widersten, 2005; van Loo et al., 2006; Kotik et al., 2009). Sources of novel EHs are not limited to known (micro-) organisms. Metagenomic or environmental DNA (eDNA), i.e. the total microbial DNA of a microcosm such as a small soil or groundwater sample, can serve as a source of novel EHs without the need to isolate and cultivate the microorganisms. PCR-based amplification of EH gene fragments in conjunction with genome-walking techniques (Kotik, 2009) have been used to retrieve entire genes encoding α/β-hydrolase fold EHs directly from the metagenomic DNA (Kotik et al., 2009; Kotik et al., 2010). Moreover, activity screening of recombinant clones containing fragments of eDNA and hybridization to EH-specific target sequences led to the discovery of novel eDNA-derived EHs with considerable potential for biotransformations (Zhao et al., 2004).
Figure 3. Phylogenetic relationships among protein sequences encoding EHs with confirmed activities. Each sequence is represented by its GenBank accession number; for sequences and further data regarding EHs with codes starting with BD, see Zhao et al., 2004. The data set includes sequences of mammalian EHs (), plant EHs (▲), fish EHs (), insect EHs (), yeast EHs (), fungal EHs (), bacterial EHs (), and eDNA-derived EHs (). A star represents an EH with a determined X-ray protein structure: human EH (NP_001970), murine EH (NP_031966), potato EH (AAA81892), a bacterial EH from Agrobacterium radiobacter AD1 (CAA73331), a fungal EH from Aspergillus niger LCP 521 (CAB59812), and two bacterial EHs which are not members of the α/β-hydrolase fold superfamily (CAA77012 and O33283). The bootstrap consensus tree was inferred from 1000 replicates. The evolutionary distances were computed using the Poisson correction method. The bar represents 0.2 amino acid substitutions per site.
Heterologous expression of EHs
Most EH expression systems were based on Escherichia coli as a host; however, heterologous expression of EHs was also established in mammalian cells (Grant et al., 1993), the baculovirus system with Spodoptera frugiperda and Trichoplusia ni insect cell lines (Kamita et al., 2013), in the yeast strains Pichia pastoris, Saccharomyces cerevisiae and Yarowia lipolytica (Kim et al., 2006; Labuschagne and Albertyn, 2007; Botes et al., 2008), and in Aspergillus niger NW 219 (originally published as A. niger NW171; Naundorf et al., 2009). The latter heterologous host offered the possibility to use a low-cost culture medium with inexpensive corn steep liquor as the main component. Further, E. coli RE3 as the recombinant host enabled EH production in a minimal growth medium with inexpensive sucrose as the sole carbon source (Grulich et al., 2011). Co-expression of molecular chaperones together with the optimization of culture conditions resulted in lower levels of inclusion bodies in the recombinant strain E. coli BL21(DE3) when overexpressing the EH from Rhodotorula glutinis (Visser et al., 2003).
Reaction mechanisms of EHs
A large fraction of EHs belongs to the α/β-hydrolase fold superfamily, which contains– besides EHs – other structurally related hydrolytic enzymes with a characteristic arrangement of α-helices and β-sheets: esterases, haloalkane dehalogenases, lipases, amidases, and some more (Heikinheimo et al., 1999; Lenfant et al., 2013). It appears that all EHs of the α/β-hydrolase fold superfamily share a common three-step reaction mechanism, which involves the action of the active site-located catalytic triad (Asp-His-Glu/Asp), two tyrosine residues and a water molecule (Figure 4). In a first step, the carboxylic acid of aspartate attacks an oxirane carbon of the bound epoxide substrate, resulting in a transiently formed ester intermediate. Two conserved tyrosines assist in this step of catalysis, polarizing the epoxide ring by hydrogen bonding with the oxirane oxygen. The second step of the reaction mechanism, which is often rate-limiting, is characterized by the hydrolysis of the ester intermediate, catalyzed by an activated water molecule; the amino acid pair His-Glu/Asp of the catalytic triad is responsible for this activation. In the third and final step the formed diol product is released, leaving behind the restored catalytic triad of the enzyme.
Some EHs are not members of the α/β-hydrolase fold superfamily, as shown for the limonene-1,2-epoxide hydrolase from Rhodococcus erythropolis (Arand et al., 2003) and the EH from Mycobacterium tuberculosis (Johansson et al., 2005). These two enzymes, which are dimers, have similar overall structures, each subunit consisting of a curved six-stranded β-sheet and four helices. The active site is located in a deep pocket with an Asp-Arg-Asp catalytic triad at its bottom. In contrast to the above-mentioned catalytic mechanism of α/β-hydrolase fold EHs, a single-step push-pull mechanism has been proposed, which includes the activation of a water molecule by hydrogen bonding, resulting in a nucleophilic attack at the epoxide ring; at the same time, the epoxide is polarized and thereby activated by making available a proton to the oxirane oxygen (acid catalysis) (Figure 5).
Figure 4. Proposed catalytic mechanism of EHs which are members of the α/β-hydrolase fold superfamily.
Figure 5. Proposed catalytic mechanism of the limonene-1,2-epoxide hydrolase from Rhodococcus erythropolis and the EH from Mycobacterium tuberculosis. The catalytic water molecule is activated by the formation of hydrogen bonds involving aspartic acid, asparagine and tyrosine residues. At the same time, polarization and activation of the oxirane ring is achieved by hydrogen bonding with the oxirane oxygen involving another aspartic acid residue.
Mono-functional epoxides as chiral building blocks for the synthesis of biologically active compounds
Mono-functional epoxides are the simplest EH substrates and some of them, such as styrene oxide and derivatives thereof, are routinely used as standard substrates for EH activity measurements. Even within such a simple class of compounds it is possible to distinguish between, for example, mono-, di- and tri-substituted epoxides, aromatic and non-aromatic epoxides, and meso epoxides. The following reflects such a subdivision of epoxides and may assist in finding the right enzyme for as yet not studied epoxides just by structure comparison with already described substrates of EHs.
Mono-substituted aromatic epoxides
Both enantiomers of styrene oxide (SO) and phenyl-1,2-ethanediol are common chiral aromatic molecular building blocks for the synthesis of pharmaceuticals and other specialty chemicals. In the last decades a substantial amount of work has been dedicated to the enantioselective biohydrolysis of SO, the reason probably being the commercial availability of SO and the corresponding diol in both racemic and enantiopure forms. In 1993 Furstoss and co-workers described the first preparative access to both enantiomers of SO by enantioselective hydrolysis of the racemate using cells of two fungal strains that were enantiocomplementary, i.e. enantioselectively hydrolyzed either of the two SO enantiomers (Pedragosa-Moreau et al., 1993; Pedragosa-Moreau et al., 1995). In addition, it was observed that the hydrolysis of racemic SO using cells of Aspergillus niger LCP 521 proceeded with retention of configuration at the chiral center, generating the (R)-diol and leaving behind the (S)-epoxide as the residual compound. On the other hand, hydrolysis of SO using cells of Beauveria sulfurescens ATCC 7159 resulted in the formation of the (R)-diol from the (S)-epoxide by inversion of configuration, leaving behind the unreacted (R)-epoxide. These findings were used for a biohydrolysis reaction in the presence of both molds, resulting in the first enantioconvergent bi-enzymatic process for the production of (R)-phenyl-1,2-ethanediol in high yield (92%) and with an ee as high as 89% (Figure 6).
Figure 6. Biohydrolytic kinetic resolution of racemic styrene oxide (rac-SO) catalyzed by Aspergillus niger LCP 521 (AnEH) and/or Beauveria sulfurescens ATCC 7159 (BsEH). Using both fungi together led to an enantioconvergent production of (R)-phenyl-1,2-ethanediol.
More recently, this bi-enzymatic enantioconvergent strategy was applied for the preparation of the same (R)-diol compound in a higher ee, using different combinations of EHs. For example, by mixing two purified EHs, a wild-type EH from S. tuberosum and an evolved EH from A. radiobacter AD1, SO was rapidly converted to the corresponding (R)-diol in 98% ee and 100% yield (Cao et al., 2006). However, this process had to be carried out at a low substrate concentration of 5 mM due to product inhibition of the S. tuberosum EH. Moreover, a mixture of recombinant whole cells harboring the EH-encoding genes from A. niger LK and C. crescentus was also used in a preparative-scale batch reaction for the enantioconvergent hydrolysis of 1.2 g of racemic SO at a concentration of 33 mM. 1.3 g of (R)-phenyl-1,2-ethanediol with an enantiopurity of 91% was obtained with an overall yield of 95%. Substrate concentrations exceeding 50 mM could not be used due to product inhibition of the bacterial EH (Hwang et al., 2008b). A similar approach was described later in which the A. niger LK EH was replaced by an EH mutant from Mugil cephalus, a marine fish (Min and Lee, 2012). After optimization of the reaction conditions, (R)-phenyl-1,2-ethanediol was obtained in 90% ee and 95% yield from 50 mM racemic styrene oxide. Shen and co-workers isolated two bacterial EHs (SgcF and NcsF2), which are involved in the biosynthesis of enediynes, antitumor antibiotics produced by Streptomyces globisporus and Streptomyces carzinostaticus. Using SO as a substrate mimic, SgcF and NcsF2 were shown to be enantiocomplementary, leading to the formation of (R)-phenyl-1,2-ethanediol in 99% ee and 87% yield from racemic SO (14 mM) when used together in the reaction mixture (Lin et al., 2010). An interesting enantioconvergent process with whole cells of Aspergillus tubingensis TF1 was recently described (Duarah et al., 2013). (R)-phenyl-1,2-ethanediol was isolated from the reaction mixture in 97% ee after 45 min, reaching >99% conversion of racemic SO (8.75 mM). Although a single enzyme was claimed to be responsible for this process, the possibility of two enantiocomplementary EHs with opposite regioselectivity being present in the microorganism cannot be ruled out from the described data.
Although a great number of wild-type EHs was found to kinetically resolve racemic SO, to the best of our knowledge no EH showed a very high enantioselectivity level. Nevertheless, several kinetic resolutions at high SO concentrations have been described using EHs from yeast, bacteria and plants in the last decade. For example, (S)-SO with 98% ee was obtained in 41% yield from racemic SO at a very high concentration of 1.8 M using a Pichia pastoris strain overexpressing the EH from Rhodotorula glutinis. Such a high substrate concentration in the reaction mixture called for optimized reaction conditions, i.e. the reaction was taking place at 4° C in the presence of 40% (v/v) Tween 20 and 5% (v/v) glycerol (Yoo et al., 2008). Biomass of Achromobacter sp. MTCC 5605, which was isolated from a petroleum-contaminated sludge sample, enabled the hydrolytic kinetic resolution of racemic SO at a high concentration of 0.5 M using a biphasic reaction system composed of isooctane and buffer. Under these conditions, an enantiomeric ratio of 64 was determined; the remaining (S)-SO (42% yield) and the formed (R)-phenyl-1,2-ethanediol were isolated in > 99 and 65% ee, respectively (Kamal et al., 2013). It is worth mentioning that an improvement of the enantioselectivity was obtained after covalent immobilization of the multimeric EH from A. niger LCP 521 onto Eupergit C which was partially modified with ethylene diamine (Eupergit C/EDA), resulting in an E-value of 56 instead of 25 (Mateo et al., 2003).
(R)-para- and (R)-meta-chlorostyrene oxides are important building blocks for the synthesis of various biologically active molecules. Indeed, these compounds are, for example, essential chiral intermediates for the production of Eliprodil (Pabel et al., 2000), an effective NMDA receptor antagonist, and various β-3-adrenergic receptor agonists such as SR 58611A or AJ-9677 (Harada et al., 2003). Numerous biohydrolytic kinetic resolutions have been described in the literature with the purpose of preparing these chiral synthons in enantiopure form. However, it is well known that one of the general drawbacks of including a resolution step in a chemical synthesis is its intrinsic 50% yield limitation. This is the reason why various enantioconvergent processes with the aim of approaching the ideal situation of “100% yield and 100% ee” have been elaborated.
As far as para-chlorostyrene oxide (p-ClSO) is concerned, enzymatic extracts of overexpressed EHs from A. niger LCP 521 (AnEH) and S. tuberosum (StEH) were shown to efficiently resolve this racemic epoxide (E-values of 100 at 0 °C). The enantiocomplementarity of these two enantioselective EHs enabled the preparation of both enantiomers of p-ClSO, the (R)-pClSO and (S)-pClSO being preferentially hydrolyzed by the AnEH and the StEH, respectively. The absolute configuration of the formed diol was determined to be R in both cases. Preparative-scale resolutions were performed at very high substrate concentrations of 306 g/L and 30.6 g/L using respectively AnEH and StEH as the biocatalysts (Manoj et al., 2001). Similar results were also obtained in a repeated batch reaction with both enzymes immobilized onto DEAE-cellulose by ionic adsorption (Karboune et al., 2005b).
Based on the complementary enantio- and regioselectivities of these two EHs, an enantioconvergent production of (R)-para-chlorophenyl-1,2-ethanediol from rac-p-ClSO was established using a sequential bi-enzymatic strategy. Thus, a preparative-scale experiment was carried out at a substrate concentration of 30.6 g/L using first the StEH followed by the AnEH. The (R)-diol was obtained with an overall yield as high as 93% and 96% ee (Figure 7) (Manoj et al., 2001).
Figure 7. Enantioconvergent biohydrolytic transformation of para-chlorostyrene oxide using a bi-enzymatic process. The sequential use of Solanum tuberosum and Aspergillus niger EHs as biocatalysts led to the formation of enantiopure (R)-para-chlorophenyl-1,2-diol, a chiral building block for the synthesis of (R)-Eliprodil.
Unfortunately, the substrate concentration had to be decreased by a factor of 10 in this bi-enzymatic process, compared to a single-enzyme resolution process with AnEH, to diminish the inhibitory effect of the formed diol on the StEH activity. Later, a repeated batch experiment using these two EHs, separately immobilized onto DEAE-cellulose, was also described (Karboune et al., 2005a). More recently, other teams reported a similar strategy using two EHs. A sequential bi-enzymatic hydrolysis of p-ClSO was described by Lee et al. (Min and Lee, 2012), using a heterologously expressed EH from Caulobacter crescentus and an EH mutant from Mugil cephalus. The combined use of whole cells overexpressing these two EHs enabled the production of (R)-para-chlorophenyl-1,2-ethanediol with 92% enantiopurity and 71% yield, starting from 17 g/L of rac-p-ClSO. In a previous paper, the same authors described an enantioconvergent process which was based on a mono-enzymatic approach using the EH from C. crescentus (Hwang et al., 2008a). This EH was shown to have opposite enantioselectivity and regioselectivity toward either enantiomer of rac-p-ClSO, which led to the almost exclusive formation of the (R)-diol. With the enantioselectivity being not too high (E-value = 30), a preparative-scale biohydrolysis at a substrate concentration of 16.8 g/L resulted in the formation of (R)-para-chlorophenyl-1,2-ethanediol with 98% ee and 78% overall yield. Kotik et al. have described the first example of regioselectivity engineering in EHs by directed evolution starting from a non-enantioconvergent enzyme (Kotik et al., 2011). The substrate binding cavity of the EH from A. niger M200 was redesigned to generate an enantioconvergent biocatalyst by guiding the point of nucleophilic attack to the benzylic oxirane position of the bound (S)-enantiomer. After nine amino acid exchanges, the final enzyme variant transformed racemic p-ClSO to the (R)-diol with an ee of 70.5%. These authors reported in the same article a sequential bi-enzymatic reaction using the wild-type EH from A. niger M200 and its evolved variant, resulting in the formation of the (R)-diol with an ee-value of 88%. More recently, Kotik and co-workers reported that an EH (named Kau2), whose gene was isolated from a biofilter-derived metagenome, exhibited an opposite regioselectivity for the two enantiomers of p-ClSO, which enabled them to obtain (R)-para-chlorophenyl-1,2-ethanediol in 84% ee at 100% conversion (Kotik et al., 2010).
An enantioconvergent preparative-scale production of (R)-meta-chlorophenyl-1,2-ethanediol was described by Furstoss and co-workers using the EH from Solanum tuberosum (Monterde et al., 2004). The enzyme exhibited a low enantioselectivity (E-value of 6) in conjunction with an opposite regioselectivity for each enantiomer of m-ClSO, which are ideal conditions for an enantioconvergent process. Starting from rac-m-ClSO, nine cycles of a repeated batch experiment in a stirred reactor at 10 g/L of substrate concentration furnished 20 the (R)-diol with an ee-value of 97% and an 88% overall yield. Very recently, Li and co-workers have studied the biohydrolysis of numerous epoxides (including meso compounds) using the EH from Sphingomonas sp. HXN-200 overexpressed in E. coli (SpEH) (Wu et al., 2013). This EH with an E-value of 41 was shown to be more enantioselective than any other known EH for the hydrolysis of rac-m-ClSO. Interestingly, SpEH reacted preferentially with the (R)-epoxide, forming the (R)-diol and leaving behind the (S)-m-ClSO, which is a useful chiral building block for the preparation of an IGF-1R kinase inhibitor (Witmann et al., 2005). A gram-scale kinetic resolution of rac-m-ClSO was performed in a two-phase system (buffer/n-hexane) at 15 g/L of substrate with resting SpEH-containing cells. Enantiopure (S)-m-ClSO was obtained in 37.9% yield.
(R)-para-nitrostyrene oxide (p-NSO) is the key chiral synthon for the synthesis of Nifenalol, a compound showing β-blocking activity and used in the treatment of hypertensive diseases (Murmann et al., 1967). This epoxide, which is mostly insoluble in water, has been resolved with good enantioselectivity by fungal, bacterial and yeast EHs (Table 1).
A 330 mM (54 g/L) solution of p-NSO was hydrolysed within 6 h to furnish the (S)-epoxide in 49% yield and 99% ee. Then, the controlled acid hydrolysis of the reaction mixture yielded the (R)-diol (80% ee) as a result of steric inversion upon acid hydrolysis of the unreacted (S)-epoxide. After recrystallisation, the (R)-diol could be easily recycled to the epoxide and transformed into (R)-Nifenalol. Interestingly, Botes et al. (Pienaar et al., 2008; Botes et al., 2007c) described the enantioconvergent production of the (R)-diol in one pot at a high substrate concentration using the combined action the Oxy-4 and Oxy-10 biocatalysts (bi-enzymatic-process).
Enantiocontrolled synthesis of fluorinated organic compounds has gained tremendous impetus over the recent years because it is well known that the presence of fluorine atoms in a molecule can have dramatic effects on its biological activity (Soloshonok, 1999). In this context, the kinetic resolution of a specific trifluoro-methyl-substituted aromatic epoxide family was studied using a recombinant EH from A. niger LCP 521, and the productivity of the biotransformation process was evaluated (Deregnaucourt et al., 2007). A two liquid-liquid phase methodology with an appropriate co-solvent (isooctane, 10 to 35% (v/v)) and optimized operational conditions led to a very efficient and cost-effective resolution process. The best results (high E-value, high TON and TOF) were obtained in the case of para-substituted CF3-, OCF3- and SCF3-derivatives. For example, resolution of (4-trifluoromethoxyphenyl)-oxirane could be performed at 250 g/L, resulting in the residual (S)-epoxide and the formed (R)-diol in good yields and very high enantiomeric excess (Figure 9).
Enantiopure 2-, 3- and 4-pyridyloxirane are key-building blocks for the synthesis of several biologically active compounds, such as β-adrenergic receptor agonists or antiobesity drugs (Mathvink et al., 1999; Devries et al., 1998; Fisher et al., 1996). Up to now, none of these products could be obtained in a satisfactory enantiopure form using the most effective heavy metal-containing catalysts (Jacobsen epoxidation, Jacobsen HKR or Sharpless dihydroxylation). Interestingly, the recombinant EH from A. niger LCP 521 exhibited a rather high enantioselectivity toward all three substrates with E-values of 96, 27 and 47, respectively, hydrolyzing preferentially the (R)-enantiomer and thus enabling the recovery of the slowy reacting (S)-epoxide. Unfortunately, it was shown that the E-value decreased with increasing substrate concentration. Nevertheless, the preparative-scale synthesis of each pyridyloxirane could be performed at about 10 g/L substrate concentration, which enabled these three target compounds to be obtained in nearly enantiopure form (Genzel et al., 2001a). In the same year, it was shown that a Tyr215Phe mutation in the EH from Agrobacterium radiobacter AD1 resulted in an enzyme variant that could efficiently resolve 2-pyridyloxirane (E-value = 55) at a substrate concentration as high as 127 mM (15.5 g/L) (Figure 10) (Genzel et al., 2001b).
Di-substituted aromatic epoxides
p-Bromo-, p-isobutyl- and p-trifluoromethyl-alpha-methyl styrene oxide
To overcome the problem of low solubility of aromatic epoxides in the water phase, Furstoss and collaborators studied in 1998 enzymatic resolutions at high substrate concentration without adding organic solvents (Cleij et al., 1998). They showed that at a high concentration of 80 g/L of para-bromo- α-methyl styrene oxide a biphasic system was formed with the epoxide constituting one phase by itself. This enabled a good kinetic resolution of the aromatic epoxide using an EH-containing extract from A. niger LCP 521 as a biocatalyst. Under these experimental conditions, the residual epoxide was found to be of (S) configuration, whereas the formed product was the corresponding (R)-diol. Surprisingly, the use of this procedure led to a dramatic enhancement in the enantioselectivity with the E-value increasing from 20 at low substrate concentration (1.7 g/L) to 260 at 80 g/L (Figure 11).
One year later, the same authors studied the biohydrolysis of seven differently substituted-methylstyrene oxide derivatives, including the para-bromo-methyl styrene oxide, using 10 different EHs (Cleij et al., 1999). The best results were obtained with the EH from A. niger LCP 521; however, the E-values were relatively moderate. A four-step synthesis of (S)-Ibuprofen, a non-steroidal anti-inflammatory drug, was performed to illustrate the synthetic potential of EHs. The strategy was to achieve the enantioselective hydrolysis of rac-4-isobutyl-α-methylstyrene oxide using the EH from A. niger, which has been shown to specifically hydrolyse the undesired (R)-enantiomer, and to further transform the enantiopure residual (S)-epoxide into (S)-Ibuprofen using classical chemical synthesis. As in the case of the para-bromo derivative, the biohydrolysis was performed at a high substrate concentration of 50 g/L, leading to a biphasic process, and at a low reaction temperature of 4 °C to enhance enzyme stability and decrease the spontaneous hydrolysis of the substrate. Following this strategy, the overall yield of (S)-Ibuprofen was only 27%. Recycling of the formed diol via chemical racemization substantially improved the process yield. Indeed, treatment of the formed diol with HBr/AcOH and subsequent cyclization of the bromhydrin intermediate under basic conditions afforded racemic 4-isobutyl-methylstyrene oxide in 80% yield, which could thus be resubmitted to the enzymatic resolution step. Under these conditions, the overall yield increased from 27 to 47% (Figure 12).
Ten years later, Furstoss’s group confirmed the high kinetic resolving power of the EH from A. niger LCP 521 by showing that the enzyme could also be used for the efficient resolution of para-trifluoromethyl-α-methyl styrene oxide (Deregnaucourt et al., 2007). Indeed, the preparative-scale biohydrolysis at 100 g/L could be performed in a short reaction time in a biphasic reaction medium containing water-organic solvent. Isooctane (25% v/v) was added to the reactor to obtain a good substrate emulsion, leading to an optimal transfer of the substrate from the organic phase to the water phase. Under these experimental conditions, a loading of 25 g of racemic epoxide resulted in the formation of 10.5 g (42% yield) of (S)-epoxide (99.7% ee) and 13.4 g (78.5% yield) of (R)-diol (78.5% ee) (Figure 13).
Figure 13. Preparative-scale synthesis of enantiopure para-trifluoromethyl-α-methyl styrene oxide using a partially purified recombinant EH from Aspergillus niger LCP 521 as biocatalyst.
cis- and trans- -methylstyrene oxide
Enantiopure form of trans-β-methylstyrene oxide has been used as a building block for the synthesis of a potential cocaine abuse therapeutic agent and an anti-obesity drug (Hsin et al., 2003; Lin et al., 2006). Very satisfactory resolutions of trans-β-methylstyrene oxide were achieved with a metagenome-derived EH (termed Kau2) (Kotik et al., 2010) and EHs from fungi, e.g. Beauveria sulfurescens (Pedragosa et al., 1996b), and yeasts Rhodotorula glutinis CIMW147 (Weijers, 1997), Rhodotorula glutinis UOFS Y-0123 (Lotter et al., 2004), and Rotoruloides mucoides UOFSY-0471 (Botes et al., 2007a). A comparison of all the described results leads to the conclusion that the hydrolysis of trans-β-methylstyrene oxide proceeded with a similar enantioselectivity and stereochemistry. Indeed, in all cases the (1S,2S)-epoxide was preferentially hydrolysed to the (1R,2S)-erythro-diol, indicating that the enzymatic attack occurred at the benzylic position. The best result was achieved in a preparative-scale reaction at 80 g/L of substrate concentration, using freeze-dried E. coli RE3 cells harbouring the plasmid pSEKau2 for expression of the EH Kau2. Both (1R,2R)-epoxide and the corresponding (1R,2S)-diol were isolated in high enantiomeric excess (>99%) and good yield (>45%), corresponding to a very high enantioselectivity (E-value >200) (Figure 14).
Table of contents :
CHAPTER I. INTRODUCTION
CHAPTER II. EPOXIDE HYDROLASES AND THEIR APPLICATION IN ORGANIC SYNTHESIS
II.2 SOURCES AND REACTION MECHANISM OF EHS
II.2.1 Sources of EHs
II.2.2 Heterologous expression of EHs
II.2.3 Reaction mechanisms of EHs
II.3 MONO-FUNCTIONAL EPOXIDES AS CHIRAL BUILDING BLOCKS FOR THE SYNTHESIS OF BIOLOGICALLY ACTIVE COMPOUNDS
II.3.1 Mono-substituted aromatic epoxides
II.3.2 Di-substituted aromatic epoxides
II.3.3 Non aromatic epoxides
II.4 PREPARATION OF VALUABLE CHIRAL BUILDING BLOCKS FOR THE SYNTHESIS OF BIOLOGICALLY ACTIVE COMPOUNDS STARTING FROM BI-FUNCTIONAL EPOXIDES
II.4.1 Halogenated epoxides
II.4.3 Protected epoxy-alcohols
CHAPTER III. RESULTS
III.1 INTRODUCTION OF KAU2-EH
III.1.1 Cloning of Kau2-EH from metagenomic DNA and properties
III.1.2 Biocatalytic properties of Kau2-EH
III.1.3 Purposes of this PhD work
III.2 MODELISATION STUDY OF KAU2-EH
III.2.2 Preparation of CDU and CIU
III.2.3 Determination of kinetic parameters of Kau2-EH
III.2.4 Determination of Ki and CDU & CIU type of inhibition with Kau2-EH
III.2.5 Determination of Ki and CDU & CIU type of inhibition of with potato EH
III.3 BIOCONVERSION STUDIES
III.3.2 Chemical Synthesis
III.3.3 Kau2-EH production in fermentor
III.3.4 Bioconversion using Kau2-EH
III.3.5 Overall conclusion
III.4 KINETIC STUDIES
III.4.2 Construction of Kau2-HisTag and mutants
III.4.3 Heterologous expression of Kau2-His6 and the corresponding mutants
III.4.4 Purification of Kau2-His6 and the corresponding mutants
III.4.5 Preparation of (SS)- and (RR)-TSO
III.4.6 Kinetic study
CHAPTER IV. CONCLUSION
CHAPTER V. MATERIAL AND METHODS
V.1.1 Reagents, solvents and chromatography
V.1.2 Buffer and culture media
V.2.1 Nuclear Magnetic Resonance (NMR)
V.2.2 Gas chromatography (GC)
V.2.3 High Pressure Liquid Chromatography (HPLC)
V.2.4 5 L Fermentor
V.3 CHEMICAL SYNTHESIS
V.3.1 Synthesis of N-cyclohexyl-N′-decylurea (CDU)-2
V.3.2 Synthesis of N-cyclohexyl-N’-(4-iodophenyl) urea (CIU)-3
V.3.3 Synthesis of rac-trans-methyl phenylglycidate 4
V.3.4 Synthesis of rac-trans-ethyl-3-phenylglycidate-5
V.3.5 Synthesis of cis and trans-2, 3-epoxy-3-phenylpropanenitrile 7 and 8
V.3.6 Synthesis of rac-trans-2,3-epoxy-3-phenyl-1-bromo propane-9 and rac-trans-2,3-epoxy-3 phenyl-1-chloro propane-10
V.3.7 Synthesis of rac-cis-methyl phenylglycidate-13
V.3.8 Chemical hydrolysis of rac-trans-methyl phenylglycidate-4
V.4 KAU2-EH PRODUCTION
V.4.1 Cells expressing Kau2-EH production in flask
V.4.2 Cells production in fermentor
V.4.3 Determination of EH activity
V.5 MODELISATION STUDY OF KAU2-EH
V.5.1 Determination of EH activities and kinetic parameters (Km, Vmax)
V.5.2 Determination of Ki and type of inhibition using CDU and Kau2-EH
V.5.3 Determination of Ki and type of inhibition using CIU and Kau2-EH
V.5.4 Determination of Ki and type of inhibition using CDU and Potato-EH
V.5.5 Determination of Ki and the type of inhibitor of CIU with Potato EHs
V.6.1 Bioconversion of rac-trans-methyl-phenylglycidate-4
V.6.2 Bioconversion of rac-trans-ethyl-3-phenylglycidate-5
V.6.3 Bioconversion of rac-methyl-trans-3-(4-methoxyphenyl)-glycidate-6
V.6.4 Bioconversion of rac-trans-2,3-epoxy-3-phenylpropanenitrile-7
V.6.5 Bioconversion of rac-cis-2,3-epoxy-3-phenylpropanenitrile-8
V.6.6 Bioconversion of rac-trans-2,3-epoxy-3-phenylpropyl bromide 9
V.6.7 Bioconversion of rac-trans-2,3-epoxy-3-phenylpropyl-chloride 10
V.6.8 Bioconversion of rac-trans-stilbene oxide-11
V.6.9 Bioconversion of cis-stilbene oxide-12
V.7 KINETIC STUDIES
V.7.1 Protein purification