Metabolic engineering of Escherichia coli for production of (2S,3S)-butane-2,3-diol from glucose

Background Butane-2,3-diol (2,3-BD) is a fuel and platform biochemical with various industrial applications. 2,3-BD exists in three stereoisomeric forms: (2R,3R)-2,3-BD, meso-2,3-BD and (2S,3S)-2,3-BD. Microbial fermentative processes have been reported for (2R,3R)-2,3-BD and meso-2,3-BD production. Results The production of (2S,3S)-2,3-BD from glucose was acquired by whole cells of recombinant Escherichia coli coexpressing the α-acetolactate synthase and meso-butane-2,3-diol dehydrogenase of Enterobacter cloacae subsp. dissolvens strain SDM. An optimal biocatalyst for (2S,3S)-2,3-BD production, E. coli BL21 (pETDuet–PT7–budB–PT7–budC), was constructed and the bioconversion conditions were optimized. With the addition of 10 mM FeCl3 in the bioconversion system, (2S,3S)-2,3-BD at a concentration of 2.2 g/L was obtained with a stereoisomeric purity of 95.0 % using the metabolically engineered strain from glucose. Conclusions The engineered E. coli strain is the first one that can be used in the direct production of (2S,3S)-2,3-BD from glucose. The results demonstrated that the method developed here would be a promising process for efficient (2S,3S)-2,3-BD production. Electronic supplementary material The online version of this article (doi:10.1186/s13068-015-0324-x) contains supplementary material, which is available to authorized users.

As shown in Fig. 2, the genes encoding ALDC (accession number: 392323960), ALS (accession number: 392323961) and meso-BDH (accession number: 392323962) are sequentially clustered in one operon and under control of transcriptional regulation protein AlsR in E. cloacae. Two separate plasmids, pETDuet-P T7 -budB-P T7 -budC and pET28a-lysR-P abc -budB-budC, were used for expression of the budB and budC genes of E. cloacae subsp. dissolvens SDM in recombinant E. coli (Fig. 2). In a previous report, the 2,3-BD pathway genes of strain SDM were expressed under the control of different types of promoters. The recombinant E. coli strain with the native promoter (P abc ) of 2,3-BD synthesis gene cluster of strain SDM had the best ability to produce 2,3-BD [17]. Thus, in pET28a-lysR-P abc -budB-budC, the genes lysR, budB, budC and P abc of E. cloacae subsp. dissolvens SDM were ligated through gene splicing by  Construction of pET28a-lysR-P abc -budB-budC. lysR, the gene encoding the transcriptional regulator; budB, the gene encoding ALS; budA, the gene encoding ALDC; budC, the gene encoding meso-BDH, P abc , the predicted promoter of the 2,3-BD pathway gene cluster from E. cloacae subsp. dissolvens strain SDM overlap extension and cloned into the multiple clone site of pET28a. The expression of both budB and budC was also under the control of transcriptional regulation protein AlsR and the promoter P abc of the 2,3-BD pathway gene cluster of strain SDM (Fig. 2b). In pETDuet-P T7 -budB-P T7 -budC, budB and budC were cloned into the two multiple clone sites of pETDuet-1 and under the control of the promoter P T7 (Fig. 2a).
The budB and budC genes were successfully cloned from E. cloacae subsp. dissolvens SDM and then inserted into pETDuet-1 to get pETDuet-P T7 -budB-P T7 -budC (Additional file 1: Figure S1). The fragment budB-budC amplified from the genomic DNA of SDM was about 2500 bp; while the fragment lysR-P abc amplified from the genomic DNA of SDM was about 1000 bp (Additioanl file 1: Figure S1). These two fragments were ligated through recombinant PCR to get fragment lysR-P abc -budB-budC. This fragment was inserted into pET28a and resulted in plasmid pET28a-lysR-P abc -budB-budC.

Production of (2S,3S)-2,3-BD by different recombinant E. coli strains
The constructed expression vectors were transformed into E. coli BL21(DE3) and the 2,3-BD synthesis abilities of the whole cells of recombinant strains were assayed. The 20 mL reaction mixtures were incubated at 30 °C and 180 rpm in a 50-mL flask.
The activities of ALDC, ALS, and BDH in the recombinant strains were also assayed ( Table 2). Consistent with the result of 2,3-BD production, E. coli BL21 (pET-Duet-1) exhibited low ALS and BDH activities. E. coli BL21 (pETDuet-P T7 -budB-P T7 -budC) showed the highest ALS and meso-BDH activities. No ALDC activity could be detected in all of the E. coli strains. Since the concentration of 2,3-BD obtained and the ALS and meso-BDH activities of E. coli BL21 (pETDuet-P T7 -budB-P T7 -budC) were higher than that of other recombinant strains, E. coli BL21 (pETDuet-P T7 -budB-P T7 -budC) was chosen for further investigation.

Table 1 Glucose consumption, product and yield analyses of E. coli strains harboring different vectors in 24 h flask cultures
Data are the mean ± standard deviations (SDs) from three parallel experiments

Strain
Glucose consumed (g/L) 2,3-BD (g/L) 2,3-BD yield (g/g) 13.67 ± 0.58 1.14 ± 0.01 0.08 expectation. Thus, we analyzed the bioconversion system and drew the time course of the process. As shown in Fig. 3a, during the bioconversion process, pH was decreased from 7.0 to 5.0 due to production of organic acids. Thus, we expected that the decrease in pH might be the reason of the low stereoisomeric purity of the (2S,3S)-2,3-BD.

Effect of temperature on production of (2S,3S)-2,3-BD by recombinant E. coli
Efficiency of the bioconversion processes and non-enzymatic reaction is temperature dependent. Thus, in this study, the effects of temperature (16,25,30,37,45 and 55 °C) on (2S,3S)-2,3-BD production were also examined. As shown in Fig. 5, the highest (2S,3S)-2,3-BD concentration was obtained when the temperature was maintained at 37 °C. However, the stereoisomeric purity of (2S,3S)-2,3-BD was much lower than that of 25 and 30 °C. The stereoisomeric purity of (2S,3S)-2,3-BD is rather important for its utilization as the building block in asymmetric synthesis. Since both high product concentration and stereoisomeric purity could be obtained at 30 °C, this temperature was chosen for subsequent bioconversions.

Batch bioconversion under optimal conditions
Combining the results mentioned above, an optimal system for the production of (2S,3S)-2,3-BD from glucose was developed. Bioconversion was conducted at 30 °C in 50-mL shake flasks containing 20 mL medium. The medium consisted of 40 g/L glucose, 10 mM FeCl 3 and 5 g DCW/L whole cells of E. coli BL21 (pETDuet-P T7 -budB-P T7 -budC). The pH was maintained at 7.0 through periodical addition of 10 M NaOH.

Condition
Glucose consumed (g/L) 2,3-BD (g/L) 2,3-BD yield (g/g) Purity of (2S,3S)-2,3-BD It has been reported that the cost of substrates accounted for more than about 30 % of total cost of 2,3-BD production. In this work, we constructed a recombinant E. coli coexpressing ALS and meso-BDH for the production of (2S,3S)-2,3-BD from a cheap substrate, glucose. The product concentration of our system was lower than that of processes using other substrates (Table 4). However, the cost analyses indicated that our process could produce (2S,3S)-2,3-BD at a rather low substrate cost (Additional file 1: Table S1). Thus, the method presented in this work would be a promising process for (2S,3S)-2,3-BD production.

Bacterial strains and plasmids
The bacterial strains and plasmids used in this study are listed in Table 5. E. coli DH5α and BL21 (DE3) were used as cloning and expression host, respectively. The pEASY-Blunt cloning vector (TransGen Biotech, China) was used for gene cloning, pETDuet-1 and pET28a were used for gene expression. E. cloacae strain SDM was cultured in a medium containing the following (g/L) at pH 7.0: glucose, 15; peptone, 10; yeast extract, 5; KCl, 5. Luria-Bertani (LB) medium was used for E. coli cultivations.
Ampicillin was used at a concentration of 100 μg/mL and kanamycin was used at a concentration of 50 μg/mL.

Construction of plasmid pETDuet-P T7 -budB-P T7 -budC
E. cloacae strain SDM genomic DNAs were extracted with the Wizard Genomic DNA Purification Kit (Promega, Madison, WI, USA). The budB gene was amplified by PCR using forward primer budB-F with a BglII restriction site insertion and reverse primer budB-R with a XhoI restriction site insertion ( Table 5). The PCR product was firstly ligated to the pEASY-Blunt vector, and the resulting plasmid was designated pEASY-Blunt-budB. Next, pEASY-Blunt-budB was digested with BglII and XhoI, and the gel-purified budB fragment was ligated to the pETDuet-1 vector digested with the same restriction enzymes. The resulting plasmid was designated pET-Duet-P T7 -budB. Using the same process as described above, the budC gene fragment was obtained from the genome of E. cloacae strain SDM using primers budC-F (with the EcoRI restriction site) and budC-R (with the HindIII restriction site) (Table 5), and the pETDuet-P T7 -budB-P T7 -budC was constructed based on the pETDuet-P T7 -budB.

Biocatalyst preparation and bioconversion conditions
LB medium supplemented with 100 μg/mL ampicillin or 50 μg/mL kanamycin was used to cultivate E. coli BL21 (pETDuet-1) and E. coli BL21 (pET28a-lysR-P abc -budB-budC). (2S,3S)-2,3-BD production was carried out using 40.0 g/L of glucose as the substrate and 5.0 g DCW/L whole cells of the recombinant as the biocatalysts. 20 mL of mixture with 10 mM Fe 3+ was reacted at 30 °C and 180 rpm in 50 mL flasks. pH was controlled at 7.0 by adding 10 M NaOH.

Enzyme activity assays
To measure enzyme activity, the cells of the different E. coli strains were resuspended in 67 mM phosphate buffer (pH 7.4) and disrupted with an ultrasonic cell-breaking apparatus (Xinzhi, Ningbo, China). Cell debris was removed through centrifugation at 13,000×g for 15 min. The resulting supernatants were used in the successive enzyme activity assays. All enzyme assays were performed at 30 °C with the proper enzyme in 67 mM phosphate buffer (pH 7.4).
Activity of ALS was measured by monitoring the conversion of pyruvate to α-acetolactate [36]. One unit of ALS activity was defined as the amount of enzyme that produced 1 μmol of α-acetolactate per minute.
ALDC activity was assayed by detecting the production of acetoin from α-acetolactate [37]. α-Acetolactate was prepared immediately before use of ethyl 2-acetoxy-2-methylacetoacetate, according to the protocol supplied by the manufacturer. One unit of ALDC activity was defined as the amount of protein that formed 1 μmol of acetoin per min.
meso-BDH activity was assayed by measuring the change in absorbance at 340 nm corresponding to the oxidation of NADH or reduction of NAD when diacetyl or acetoin was used as the substrate [38]. For the reduction reaction, 5 mM acetoin or diacetyl and 0.2 mM NADH were used for the enzyme assay, and 10 mM meso-2,3-BD and 1 mM NAD were used for the oxidation reactions. One unit of enzyme activity was defined as the amount of enzyme that consumed 1 μmol of NADH or produced 1 μmol of NADH per minute.
The protein concentration was determined by the Lowry procedure using bovine serum albumin as the standard [39].

Analytical methods
Samples were withdrawn periodically and centrifuged at 12,000×g for 10 min. The concentration of glucose was measured enzymatically by a bio-analyzer (SBA-40D, Shandong Academy of Sciences, China) after diluting to an appropriate concentration. The concentrations of 2,3-BD were analyzed by GC (Varian 3800) as described previously [17]. The GC system was equipped with a capillary GC column (AT. SE-54, inside diameter, 0.32 mm; length, 30 m, Chromatographic Technology Center, Lanzhou Institute of Chemical Physics, China). The ratio of the three stereoisomers of 2,3-BD was analyzed by GC (Agilent GC6820) using a fused silica capillary column (Supelco Beta DEXTM 120, inside diameter 0.25 mm; length 30 m) [40].