Efficient biotransformation of (S)-1-phenyl-1, 2-ethanediol using xylan as co-substrate by a coupled multi-enzyme Escherichia coli system

Background: ( S )-1-phenyl-1,2-ethanediol is an important chiral intermediate in the synthesis of liquid crystals and chiral biphosphines. ( S )-carbonyl reductase II from Candida parapsilosis catalyzes the conversion of 2-hydroxyacetophenone to ( S )-1-phenyl-1,2-ethanediol with NADPH as a cofactor. Glucose dehydrogenase with a Ala258Phe mutation is able to catalyze the oxidation of xylose with concomitant reduction of NADP + to NADPH, while endo-β-1,4-xylanase 2 catalyzes the conversion of xylan to xylose. In the present work, the Ala258Phe glucose dehydrogenase mutant and endo-β-1,4-xylanase 2 were introduced into the ( S )-carbonyl reductase II-mediated chiral pathway to strengthen cofactor regeneration by using xylan as a naturally abundant co-substrate. Results: We constructed several coupled multi-enzyme systems by introducing ( S )-carbonyl reductase II, the A258F glucose dehydrogenase mutant and endo-β-1,4-xylanase 2 into Escherichia coli . Different strains were produced by altering the location of the encoding genes on the plasmid. Only recombinant E. coli /pET-G-S-2 expressed all three enzymes, and this strain produced ( S )-1-phenyl-1,2-ethanediol from 2-hydroxyacetophenone as a substrate and xylan as a co-substrate. The optical purity was 100% and the yield was 98.3% under optimal conditions of 35°C, pH 6.5 and a 2:1 substrate-co-substrate ratio. The introduction of A258F glucose dehydrogenase and endo-β-1,4-xylanase 2 into the ( S )-carbonyl reductase II-mediated chiral pathway caused a 54.6% increase in yield, and simultaneously reduced the reaction time from 48 h to 28 h. Conclusions: This study demonstrates efficient chiral synthesis

The reaction that is catalyzed by alcohol dehydrogenase is limited by cofactor recycling.
To enhance the regeneration of cofactors, substrate-coupled and enzyme-coupled techniques are always included in chiral synthesis reactions, through the use of multienzyme systems [8,9]. For example, Xu et al. described the preparation of ethyl (S)-4chloro-3-hydroxybutanoate via the coupling of carbonyl reductase with mannitol or sorbitol dehydrogenase [10]. Kosjek et al. carried out asymmetric synthesis of 4,4dimethoxytetrahydro-2H-pyran-3-ol using a ketone reductase and in situ cofactor recycling by glucose dehydrogenase [11]. Yamamoto et al. reported the efficient production of ethyl (S)-4-chloro-3-hydroxybutanoate with over 99% optical purity by co-expressing carbonyl reductase from Kluyveromyces aestuarii and formate dehydrogenase in Escherichia coli [12]. Zhang et al. increased the biotransformation efficiency of (R)-PED by employing an enzyme-coupling system containing SCRII and GDH [13]. These systems enabled scientists to improve the efficiency of the enantioselective reaction with glucose or formate as cosubstrates. However, the use of xylan in chiral synthesis to enhance cofactor recycling would be more interesting since they are naturally abundant materials.
Lignocellulose is well known as the most widely distributed and abundant polysaccharide in nature, which has been recognized as an attractive feedstock for the production of fuels and industrially important metabolites [14,15]. Lignocellulosic biomass is mainly composed of cellulose, hemicellulose, and lignin [16]. Pretreatment with alkaline, acidic, and/or enzymatic hydrolysis causes hemicellulose to decompose to pentoses, mainly xylan and lignin [17,18]. Xylan can be broken down to xylose by endo-β-1,4-xylanases (XYN2), and the expression and characterization of XYN2 from Trichoderma reesei Rut C-30 h has been reported [19].
We have previously isolated a GDH with good solvent-resistant from Bacillus sp.YX-1, and introduced this enzyme for cofactor regeneration in chiral synthesis with glucose as a cosubstrate [20]. The Ala258Phe mutant of GDH exhibited altered cofactor preference and improved xylose binding ability while reducing NADP + to NADPH [21]. An SCRII from C. parapsilosis has been shown to catalyze the conversion of 2-HAP to (S)-PED with NADPH as an electron donor [22]. The present study aimed to improve the efficiency of chiral synthesis by the construction of an NADPH-recycling multi-enzyme system containing SCRII, A258F/GDH and XYN2 from T. reesei Rut C-30 h (Fig. 1). XYN2 and A258F/GDH were introduced into the SCRII-mediated chiral synthesis pathway. In conditions of optimal pH, temperature and substrate-co-substrate ratio, the E. coli/pET-G-S-2 system performed well in terms of (S)-PED production without the addition of external cofactors. This work thus presents a novel strategy for cofactor recycling by introducing xylan into an enantioselective reaction pathway, which offers efficient chiral synthesis using an abundant natural chemical as a co-substrate.

Results And Discussion
Construction of the multi-enzyme-coupled-system containing SCRII,

A258F/GDH and XYN2
We successfully constructed four recombinant plasmids pET-S-2-G, pET-S-G-2, pET-G-2-S and pET-G-S-2 ( Fig. 2), in which the three genes were located in different sites of the pET-28a plasmid. For example, SCRII was located nearest to and A258F/GDH farthest from the promoter, while XYN2 was located between SCRII and A258F/GDH in pET-S-2-G.
Transformation into E. coli BL21 (DE3) competent cells resulted in the generation of E.
All three enzymes were successfully expressed in E . Turbidity measurements revealed that E. coli/pET-G-S-2, E. coli/pET-SCRII, E. coli/pET-A258F/GDH, E. coli/pET-XYN2 and the control strain E. coli/pET-28a all had similar growth trends (Fig. 4). This suggests that the expression of A258F/GDH and XYN2 in E. coli does not affect cell growth. Therefore, we used E. coli/pET-G-S-2 expressing SCRII, A258F/GDH and XYN2 for further experiments and chiral synthesis.

The functions of SCRII and A258F/GDH are well balanced
The activities of all three enzymes in the cell-free extracts of recombinant strains are summarized in Table 2. The activities of SCRII, A258F/GDH and XYN2 from E. coli/pET-G-S-2 were found to be 0.85, 1.36 and 55.33 U/mg toward 2-HAP, xylose and xylan, respectively. The activities of SCRII from E. coli/pET-SCRII toward 2-HAP, GDH from E.
coli/pET-A258F/GDH toward xylose, and XYN2 from E. coli/pET-XYN2 toward xylan were found to be 2.73, 6.85 and 267.51 U/mg, respectively. The activities toward 2-HAP, xylose and xylan of E. coli/pET-G-S-2, which expressed all three enzymes, was lower than that of E. coli/pET-SCRII, E. coli/pET-A258F/GDH and E. coli/pET-XYN2, which only expressed ( Table 2). The activity of XYN2 from E. coli/pET-XYN2 toward xylan was much higher than that of SCRII from E. coli/pET-SCRII or A258F/GDH from E. coli/pET-A258F/GDH, although SCRII and A258F/GDH exhibited similar activity toward their corresponding substrates.
These results suggest that the functions of SCRII for chiral catalysis and GDH for cofactor regeneration are balanced. Table 2 Enzyme activities in the cell-free extracts of recombinant strains.

Strains
Specific activities (U/mg) Analysis of the activities of SCRII, A258F/GDH and XYN2 in E. coli/pET-G-S-2 were evaluated under different conditions, which revealed the optimum pH to be 6.5, 6.5 and 5.0 and the optimal temperature to be 35, 50 and 50 °C for SCRII, A258F/GDH and XYN2, respectively.

Optimization of biotransformation of ( S ) -P E D b y E . c o l i/ p E T -G -S -2
Since the optimal temperature and pH conditions were different for each enzyme, the conditions for biotransformation of 2-HAP to (S)-PED by E. coli/pET-G-S-2 required further optimization. When the cell concentration was 10% (w/v), the yield of (S)-PED from 6 g/L binding strength [28]. As depicted in Fig. 5a, the present study revealed that the yield of (S)-PED from E. coli/pET-G-S-2 was increased with the increased pH value between 3.0 and 6.5. At the optimal pH of 6.5, biotransformation of (S)-PED was achieved with an optical purity of 100% and a yield of 90.5% (Fig. 5a).
Temperature also has a significant influence on enzyme activity [29], and can affect product stereoselectivity during enzyme-catalyzed chiral biosynthesis [30]. As shown in Fig. 5b, we found that the optimal temperature for the biotransformation of (S)-PED by E.
coli/pET-G-S-2 was 35 °C, at which the optical purity was 100% and the yield was 93.6%.
The optimal temperature of E. coli/pET-SCRII has been reported to be 35 °C [22], while E.
coli/pET-A258F/GDH exhibited the highest activity toward xylose at 55 °C [21] and the optimal temperature for the activity of XYN2/E. coli toward xylan has been shown to be 50 °C [19]. In the present study, we found the activity of SCRII to decrease rapidly with increasing temperature, while the activity of other two enzymes A258F/GDH and XYN2 remained active at 35 °C.

Efficient transformation of ( S ) -P E D b y E . c o l i/ p E T -G -S -2
We investigated the reaction duration of the enzyme-coupled system E. coli/pET-G-S-2 by targeting the highest optical purity and yield of (S)-PED under optimal conditions of pH, temperature and substrate-co-substrate ratio. The yield of (S)-PED was highest (98.3%) with an optical purity of 100% at 28 h (Fig. 7). Compared with /E. coli/pET-SCRII, the introduction of XYN2 and A258F/GDH into the SCRII-mediated chiral synthesis pathway in E. coli/pET-G-S-2 caused a 54.6% increase in yield and the time required for the reaction to complete was significantly reduced from 48 h to 28 h [22]. The results of the present study suggest that the introduction of XYN2 and A258F/GDH accelerates the bioconversion dimethoxytetrahydro-2H-pyran-3-ol by co-expressing a ketone reductase and glucose dehydrogenase [11]. These approaches both achieved increased cofactor regeneration, which could overcome the limitation of coenzyme restriction for asymmetric synthesis and accelerate the initial reaction rate, resulting in improved catalytic efficiency. The present work describes a simple approach to achieve cofactor regeneration in chiral synthesis using a naturally abundant co-substrate.

Conclusions
To Escherichia coli JM109 (Invitrogen Co., Shanghai, China) was used as a host for plasmid propagation.
E. coli BL21 (DE3) (Invitrogen Co., Shanghai, China) was used for protein expression. E. coli JM109 and E. coli BL21 were cultured at 37 o C in Luria-Bertani (LB) medium supplemented with kanamycin (50 µg/mL) as the selective marker. The strains and plasmids used in this work were summarized in Table 1.   Gene cloning of SCRII, A258F/GDH and XYN2 The oligonucleotide primers (Table 1) were designed based on the gene sequences. The genes SCRII, A258F/GDH and XYN2 were amplified using the plasmids pET-SCRII, pET-A258F/GDH and pET-XYN2 as the DNA template, respectively. The PCR-amplified products were ligated to pMD19-T (Takara-Bio, Kyoto, Japan) to obtain T-SCRII, T-A258F/GDH and T-XYN2 plasmids, which were transformed in E.
coli JM109 cells and the recombinant strains were verified by DNA sequencing in Takara Co.
Construction of co-expression system of SCRII, A258F/GDH and XYN2 Several multi-enzymes coupled systems containing the SCRII, XYN2 and A258F/GDH were constructed using a Shine-Dalgarno (SD) and aligned spacing (AS) sequence (GAAGGAGATATACC) linker between them. Either SCRII or A258F/GDH was nearest to the promoter. The fusion genes SCRII-SD-AS- Their enzyme activities were analyzed using 2-HAP, xylose and xylan as substrates, respectively.
The enzymatic activities of SCRII for oxidation of 2-HAP were measured at 35 o C and pH 6.5 mixture by spectrophotometrically recording the rate of change of NADPH absorbance at 340 nm. One unit of enzyme activity is defined as the amount of enzyme catalyzing the oxidation of 1 μmol of NADPH per minute under the measurement conditions. The standard assays were performed as described by Zhang et al [22].
The oxidation activities of A258F/GDH were measured at 35 o C and pH 7.0 by spectrophotometrically recording the rate of change of NADPH absorbance at 340 nm. One unit of enzyme activity is defined as the amount of enzyme catalyzing the reduction of 1 μmol of NADP + per minute under the measurement conditions. Xylanase activity was assayed by the method described by Bailey et al. with 1% oat-spelt xylan (Sigma) as substrate at 50 o C [23]. Appropriate dilutions of the protein solution in 0.1 M sodium citrate buffer (pH 6.0) were used as the enzyme source. The amount of released sugar was determined by the dinitrosalicylic acid method described by Miller et al. [24]. The protein concentration was determined by the Bradford method with bovine serum albumin as the standard.
The temperature optimum of enzymes activity was determined at various temperatures (20−80 o C).
The pH optimum of enzyme activity was determined at the optimal temperature over a pH range of 3.0-10.0. The buffers used were 0.1 M sodium citrate buffer (pH 3.0 to 6.5), 0.1 M potassium phosphate buffer (pH 6.5 to 7.5), and 0.1 M Tris-HCl buffer (pH 8.0 to 10.0), respectively.

Biotransformation and analytical methods
The recombinant E. coli/pET-G-S-2 cells were used for (S)-PED biotransformation. The biotransformation was carried out as described previously with minor modifications [25]. The reaction mixture (2 mL) consisted of 0.1 M sodium citrate buffer (pH 6.5), 6 g/L 2-HAP, 6 g/L xylan,  The biosynthetic pathway of (S)-PED through a multi-enzyme coupled system containing SCRII, A258F/GDH, and XYN2.

Figure 2
The location of SCRII, A258F/GDH, and XYN2 on plasmid pET-28a.  Cell-growth curve of the recombinant strains. The recombinant strains were cultured at 37°C in 500 mL flask bottles, with an initial working volume of 150 mL, and then 0.1 mM IPTG was added at 3 h to induce protein expression. The induction temperature is 25 °C. The effects of temperature (a) and pH (b) on E. coli BL21/pET-G-S-2 catalyzing (S)-PED biotransformation. The temperature optimum of (S)-PED transformation was determined at