Introduction

1,3-Propanediol (1,3-PD) has been considered as an important chemical for a long time. A large-scale application of 1,3-PD is as a monomer in the synthesis of polyesters for use in carpet and textile fibers (Skraly et al. 1998). The conventional chemical method for producing 1,3-PD is rather expensive and generates toxic by-products, and its raw material is derived from nonrenewable petroleum. Microbial production of 1,3-PD is particularly attractive in that it uses renewable resources and reduces pollution to a great extent (Zeng et al. 1993). The 1,3-PD producers are all bacteria including Klebsiella (Huang et al. 2002), Citrobacter (Boenigk et al. 1993), Clostridium (Abbad-Andaloussi et al. 1996) and Lactobacillus (Schutz and Radler 1984). Among these microorganisms, Clostridium and Klebsiella have received more attention because of their relatively high yield and productivity (Zeng and Biebl 1994).

Biosynthesis of 1,3-PD from glycerol is considered as a two-step process in Klebsiella pneumoniae. First, glycerol dehydratase (GDHt) (EC 4.2.1.30), encoded by three genes dhaB, dhaC and dhaE, catalyzes the first reaction from glycerol to 3-hydroxypropionaldehyde (3-HPA). Second, 1,3-propanediol oxidoreductase (PDOR) (EC 1.1.1.202), encoded by the dhaT gene, transfers a reducing equivalent from NADH to 3-HPA, yielding 1,3-PD (Skraly et al. 1998). Both dhaBCE and dhaT were expressed successfully in exogenous strains (Tong et al. 1991; Zheng et al. 2006; Tobimatsu et al. 1996), but the activity of PDOR in exogenous strains was far lower than that of GDHt (Zheng et al. 2006; Sun et al. 2003). In the original K. pneumoniae strain, similar conditions also existed. As a result, the intermediate 3-HPA was accumulated. Also the activities of GDHt and the growth of the microorganisms may be inhibited by excessive toxic 3-HPA (Zeng and Biebl 2002). Researchers pointed out that any attempts for further metabolic engineering of 1,3-PD production in K. pneumoniae should carefully control the accumulation of 3-HPA (Zheng et al. 2006). For this purpose, there are two conceivable approaches including the overexpression of PDOR or weakening the expression of GDHt. However, the exclusion of functional dhaT gene in recombinant Escherichia coli (only harboring gene dhaBCE) resulted in a significantly higher titer of 1,3-PD (Emptage et al. 2003). Therefore, identification of the non-specific enzyme(s) or gene(s) in native E. coli responsible for converting 3-HPA to 1,3-PD may lead to an improved process for the production of 1,3-PD in K. pneumoiae. The yqhD gene encoding the above non-specific enzyme (alcohol dehydrogenase or PDORI) in E. coli was isolated by Emptage et al. (2003). Recently, the completion of the K. pneumoniae entire genome sequence has already enabled many functional genomic studies (http://genome.wustl.edu/). This will undoubtly accelerate the developments in the genetic improvement of K. pneumoniae. Also they will prove useful for K. pneumoniae as an excellent platform host for gene expression in metabolic engineering efforts.

In this study, the yqhD gene of E. coli JM109 were cloned into K. pneumoniae ME-308 to strengthen the catalytic reaction from 3-HPA to 1,3-PD. The inhibitory effect of 3-HPA was reduced. The production and the yield of 1,3-PD were significantly improved. K. pneumoniae is resistant to ampicillin, but sensitive to tetracycline and kanamycin. Consequently, replicating plasmids which encode resistance to these latter two antibiotics are potentially useful as cloning vectors for Klebsiella strains. Thus the tetracycline resistant gene (tetR) from plasmid pHY300PLK was inserted into the expression vector pUC18 for screening positive colonies.

Materials and methods

Strains, plasmids and culture media

The strains and plasmids used in this study are listed in Table 1. LB medium was used as the seed culture of E. coli, K. pneumoniae and their derivatives. When K. pneumoniae was used for electroporation, EDTA was added to the germ medium to a final concentration of 0.7 mM.

Table 1 Bacterial strains and plasmids
Full size table

After the electroporation, the seed culture for activating the recombinant K. pneumoniae was SOC medium (Joseph and David 2001). When necessary, 10 μg/ml tetracycline was added to the medium as the selection marker.

The fermentation medium for K. pneumoniae was: yeast extract 5 g/l, glycerol 40 g/l, KHPO4 · 3H2O 10 g/l, KH2PO4 2 g/l, NH4Cl 1 g/l, NaCl 0.5 g/l, MgSO4 · 7H2O 0.1 g/l,FeCl3 · 6H2O 30 mg/l, CoCl2 · 6H2O 5 mg/l, VB12 5 mg/l. The fed-batch fermentation was carried out in 3-l bioreactor (NBS, New Brunswick, USA) under the following conditions: medium volume 2 l, inoculation volume 5% (v/v), temperature 31°C, initial pH 7.0, aeration rate 0.4 l/min, and agitation speed 150 rev/min. The glycerol concentration in the medium was maintained at around 25 g/l by manually regulating the feeding rate of the glycerol solution.

Construction of plasmids and strains

The yqhD gene was amplified by PCR using total genome DNA of E. coli JM109 as template DNA and primers A1 and A2, which were designed using the sequence information of yqhD gene (gene ID 947493, GenBank) from E. coli K12.

  • A1: 5′-GAATTCATGAGCTATCGTATGTTTG-3′

  • A2: 5′-GGATCCCACTCAGAATGCCTGGCGG-3′

The PCR product was purified and connected into cloning vector pMD18-simple-T (products of Takara) for sequencing. After the sequencing, the yqhD gene was connected with the expression vector pUC18 using the restriction endonuclease sites EcoRI and BamHI that were introduced by the primers used (italicized), yielding plasmid pUC18-yqhD. The tetracycline-resistant gene (tetR) was amplified by PCR using plasmid pHY300PLK as template DNA and primers P1 and P2.

  • P1: 5′-ACCGGGATCCATGAAATACTGAATTTAAAA-3′

  • P2: 5′-ACCGAAGCTTCTTAACGATTTAGAAATCCC-3′

The purified PCR product of tetR was then inserted into pUC18-yqhD using the restriction endonuclease sites BamHI and HindIII, which were introduced by primers P1 and P2, yielding recombinant plasmid pUC18-yqhD-tetR. Both the PCR programs were performed on thermocycler (Ependdorf) and consisted of: 95°C for 5 min, followed by 30 cycles at 95°C for 30 s, 55°C for 30 s, 72°C for 90 s, as well as final extension step of 72°C for 10 min. The PCR mixture consisted of 1 ng each template, 200 mM each dNTP, 20 mM each of the primers, 2.5 μl Taq PCR buffer and 0.3 μl Ex Taq DNA polymerase (purchased from Takara) in 25 μl. The recombinant plasmid pUC18-yqhD-tetR was purified and transformed into K. pneumoniae ME-308 by standard transformation protocol using the method of electroporation (Emptage et al. 2003; Fournet-Fayard et al. 1995).

Preparation of cell-free extracts and assay of the recombinant enzyme activity

IPTG was added into the culture to a final concentration of 1 mM to induce protein expression when recombinant cells grew until the optical density at 600 nm reached 0.4–0.6. The cells were harvested by centrifugation from the fermentation culture. After washing the cell pellet with 0.035 mol/l potassium phosphate buffer (pH 8.0), the cells were resuspended in the same buffer in 3.0 ml. The cells were disrupted ultrasonically at 4°C for 10 min by a cell sonicator (SCIENTZ JY92-II, China). Cell debris was removed by centrifugation, resulting in a cell-free extract, which was subsequently used for the enzyme activity assay.

According to Johnson and Lin (1987), the activity of the enzyme for the reaction of 3-HPA to 1,3-PD can be measured by the formation speed of NADPH from NADP from the reverse reaction due to the instability of 3-HPA and a conversion factor of 3.95 for the forward reaction to 1,3-PD was used according to Ahrens et al. (1998). The reaction was carried out at 25°C in the presence of 0.6 mM NADP, 30 mM (NH4)2SO4, 100 mM 1,3-PD and an appropriate volume of the enzyme solution in a 2 ml final K2CO3–KHCO3 buffer (pH 9.0). The activity was calculated from the linear slope of increasing absorption of NADPH at 340 nm. One unit of enzyme activity is defined as the amount of protein that produces 1 μmol NADPH per minute. Specific enzyme activity is expressed as units per mg of protein.

Polyacrylamide gel electrophoresis and determination of protein concentration

The SDS–PAGE was conducted on a 12% polyacrylamide gel by the method of Joseph and David (2001). Operation was performed on the Mini-ProteanIII Electrophoresis System (Bio-Rad). Protein on the gel was stained with 0.2% (w/v) Coomassie brilliant blue R-250. Marker proteins with molecular weights ranging from 14.3 to 97.2 kDa were used to estimate the molecular weight of the expression products. The concentration of protein was measured according to the Bradford method with bovine serum albumin (BSA) as standard protein.

Purification of PDORI and measurement of its kinetic properties

After induction of IPTG for 12 h in flasks, the recombinant cells carrying pUC18-yqhD-tetR were harvested, washed with potassium phosphate (20 mM, pH 8.0) buffer twice and suspended in the same buffer at 10% of the volume of harvested culture. The cells were disrupted by sonication at 4°C. Then the mixture was centrifuged at 10,000 rev/min for 10 min. The precipitate obtained was discarded and ammonium sulfate was added to the supernatant to raise the saturation to 50%. After centrifugation, the precipitate was collected and the supernatant was 70% saturated with ammonium sulfate. The 50–70% fraction, which appeared colorless and free of all pigments, was subsequently used for further purification.

The 50–70% fraction was desalted by dialysis and loaded onto a DEAE-Sepharose FF column equilibrated with the potassium phosphate buffer (0.005 mM, pH 8.0). Then it was eluted using a linear gradient 0–0.15 M KCl in 0.005 mM potassium phosphate, pH 8.0. The purified enzyme was used for biochemical characterization. The values of apparent K m for the purified enzyme were determined according to Michaelis–Menten equation.

Analysis of biomass and metabolic products

Biomass was determined by measuring the value of optical density at 600 nm with appropriate dilution using a UV–visible spectroscopy system (DU-640, BecKm an, USA). The value of the optical density was converted to cell dry weight (CDW) using a calibration equation \( ( {\text{CDW}} = 4.183 \times {\text{OD}}_{ 6 0 0} - 0.117 ). \)

Glycerol and 1,3-PD were measured by high-performance liquid chromatography (Summit P 680 HPLC, Dionex, USA; Shodex RI-101 Refractive Index Detector, Showa Denko, Japan; Aminex HPX-87 H Ion Exclusion Column 300 × 7.8 mm, Bio-Rad, USA) under the following conditions: sample volume 10 μl; mobile phase 0.005 M H2SO4; flow rate 0.8 ml min−1; column temperature 65°C. 3-HPA was assayed based on the colorimetric method of Circle et al. (1945).

Results and discussion

All values presented in this study are averages of at least three independent trials

Construction of recombinant strains and sequence analysis of yqhD gene

The results of agarose gel electrophoresis of PCR products showed that the two DNA fragments (including the exogenous tetracycline resistance gene) were about 1.16 and 1.56 kb, respectively. The two fragments were inserted into the same plasmid (pUC18), yielding recombinant plasmid pUC18-yqhD-tetR. The recombinant plasmid was transformed into E. coli DH5α and K. pneumoniae ME-308 successfully. The positive colonies were screened. The recombinant plasmid from the transformed strain was purified and analyzed. Its restriction enzyme identification by agarose gel electrophoresis is shown in Fig. 1, and the structure of recombinant plasmid is shown in Fig. 2.

Fig. 1
figure 1

The restriction enzyme identification of the recombinant plasmid pUC18-yqhD-tetR. 1: Recombinant plasmid digested by EcoRI; 2: Circular pUC18-yqhD-tetR plasmid purified from recombinant strain; 3: Recombinant plasmid digested by EcoRI and BamHI; 4: Recombinant plasmid digested by BamHI and HindIII; 5: DNA marker 4500

Full size image
Fig. 2
figure 2

The structure of the recombinant plasmid pUC18-yqhD-tetR

Full size image

The amplified yqhD gene was connected to plasmid pMD18-simple-T for sequencing. DNA sequence analysis showed that the yqhD gene was 1,164 bp, which was in accordance with E. coli yqhD gene sequence announced by NCBI Genbank, encoding 387 amino acid residues with a predicted molecular weight of 43 kDa.

SDS–PAGE analysis of protein expression in whole cells

The results of SDS–PAGE analysis of proteins in whole cells are shown in Fig. 3. One protein band about 43 kDa appeared in the recombinant cells, which corresponded with the prediction from gene sequence. SDS–PAGE also revealed that the target protein of recombinant cells accounted for about 20% of the whole cell proteins under the optimum culture conditions.

Fig. 3
figure 3

SDS–PAGE analysis of proteins in whole cells. The left is protein marker. 1: Klebsiella pneumoniae ME-308 (pUC18-yqhD-tetR)-with IPTG induction; 2: Klebsiella pneumoniae ME-308 (pUC18-yqhD-tetR)-without IPTG induction; 3: Klebsiella pneumoniae ME-308- with the addition of IPTG; 4: Klebsiella pneumoniae ME-308-without the addition of IPTG

Full size image

Characterization of PDORI

The enzyme was purified after precipitation by ammonium sulfate, desalination by dialysis and loading onto DEAE-Sepharose FF column. The purified recombinant enzyme exhibited Michaelis–Menten kinetics, and the apparent K m values for 1,3-PD and NADP+ were 12.1 and 0.15 mM, respectively. The low value of K m (0.15 mM) of PDORI showed a high affinity for NADP+ to be linked with the enzyme during the reaction between 3-HPA and 1,3-PD, while the value of K m of native PDOR encoded by gene dhaT from K. pneumoniae was 0.31 mM (Johnson and Lin 1987) for NAD+ and the value of K m of recombinant PDOR expressed in E. coli was 0.21 mM for NAD+ (Wang et al. 2005).

Assay of the reaction activity for 3-HPA to 1,3-PD during the fermentation

The high efficiency of the reduction of 3-HPA to 1,3-PD is important during the fermentation of glycerol to 1,3-PD. The expression of PDORI, which catalyzes the non-specific reduction reaction of aldehydes, may strengthen the reaction from 3-HPA to 1,3-PD. As depicted in Fig. 4, the non-specific enzyme activity increased remarkably compared to the control during the fermentation. The maximum enzyme activity (110 IU/mg total protein) for the conversion of 3-HPA to 1,3-PD was tenfold higher than that of the wild type strain. It was supposed that the high activity was contributed to the original expression of dhaT in K. pneumoniae and exogenous expression of yqhD from E. coli.

Fig. 4
figure 4

Time course of the non-specific enzyme activity for the reaction of 3-HPA to 1,3-PD. Enzyme was extracted and assayed from the 3-l bioreactor at different times during the fermentation

Full size image

Effect of expression of yqhD on cell growth

Growth of the original strain and the recombinant strain in fed-batch cultures was monitored. As shown in Fig. 5, the growth rate (0.16 g/l/h) of the recombinant strain was obviously lower than that of the original strain (0.18 g/l/h) after the IPTG was added. This clearly indicated that the cell growth of the recombinant strain was impaired by the expression of the exogenous gene. The over-expression of plasmid-encoded protein can be a factor to impose a heavy metabolic burden on the host cell because a large amount of energy is consumed during protein synthesis (Bentley et al. 1990).

Fig. 5
figure 5

Time course of cell growth of recombinant Klebsiella pneumoniae ME-308 (pUC18-yqhD-tetR) and its parent strain Klebsiella pneumoniae ME-308

Full size image

Glycerol consumption, 3-HPA formation and 1,3-PD production

Glycerol consumption and 1,3-PD production are described in Fig. 6. Expression of yqhD significantly improved 1,3-PD production. The final titer of 1,3-propanediol achieved 67.6 g/l, which was 125.33% of that of the original strain. Glycerol consumption was 130 g/l, which was also higher than that of the control. The amount for the recombinant strain assimilating glycerol was increased by 8.33%. The conversions of glycerol to 1,3-PD were 0.513 and 0.442 g/g, respectively. These demonstrated that the recombinant strain with the expression of PDORI acquired an improved 1,3-PD-producing capacity. Metabolic flux of glycerol in recombinant cells distributed to the 1,3-PD pathway was increased with the expression of PDORI. The specific 1,3-PD-producing capability (25.98 g/g, 1,3-PD produced per gram of cells) of the recombinant strain was 1.7-fold increased over the original strain (15.28 g/g). The production of 1,3-PD by recombinant E. coli only harboring gene dhaB and yqhD was 43.1 g/l (Zhang et al. 2006)under the optimal fermentation conditions, which was far higher than that of the recombinant E. coli carrying gene dhaB and dhaT constructed before. DuPont had increased the 1,3-PD production to about 135 g/l by recombinant E. coli with the yqhD gene substituting for the dhaT gene, including other genetic modifications (Emptage et al. 2003).

Fig. 6
figure 6

Time course of 1,3-PD fermentation in 3-l fermentor under micro-aerobic conditions. RS recombinant strain; WT wild type strain

Full size image

The final concentration of 1,3-PD in the fermentation was about 40–60 g/l by the wild-type K. pneumoniae without any optimization of the fermentation conditions. Considerable difference about the final product titer exists for different derived Klebsiella strains. For example, the production was 70 g/l by K. pneumoniae AC 15 isolated from the soil samples with the mathematical optimization method (Zheng et al. 2008), 46.1 g/l by K. pneumoniae ATCC 25955, obtained from American type culture collection (Huang et al. 2002) and 72 g/l by K. pneumoniae DSM2026, obtained from the German Collection of Microorganisms and Cell Cultures. These are the higher production reported from the papers by the original K. pneumoniae strains. In this study, the production of 1,3-PD was improved obviously by recombinants transformed with exogenous gene compared with the original strain. This could encourage the further development of genetic modification of K. pneumoniae to produce 1,3-PD efficiently.

Figure 6 also showed the time courses of 3-HPA concentration during the fermentation of glycerol by recombinant K. pneumoniae and its derivative strain K. pneumoniae ME-308. The 3-HPA concentration of the recombinant strain accumulated to a maximum of around 1.8 g/l after 14 h, which was lower than that of the original strain (2.32 g/l) under the same conditions. Considering the activity of the non-specific enzyme was increased by tenfold, the toxic intermediate 3-HPA accumulation decreased by 22.4%, this also being due to the expression of yqhD. As a result, the inhibition on cell growth and on the key enzymes of 1,3-PD metabolic pathway was decreased.

The reason for this phenomenon was that the PDORI encoded by yqhD from E. coli, which is insensitive to oxygen, can use both NADPH and NADH as its coenzyme (Sulzenbacher et al. 2004). However, the PDOR encoded by gene dhaT in native K. pneumoniae can only use NADH as its coenzyme. NADH is used for oxidative phosphorylation and as donor of hydrogen for reduction reaction in organisms, while NADPH is mainly used for biosynthesis reduction reaction. Under the aerobic or anaerobic conditions, PDORI could always use both the coenzymes NADPH and NADH to catalyze the conversion 3-HPA to 1,3-PD efficiently. So the amount of toxic intermediate 3-HPA was reduced, also the production and the yield of 1,3-PD was improved significantly.

Conclusion

The yqhD gene from E. coli was expressed successfully in K. pneumoniae ME-308, and the activity of PDORI in the recombinant strain was far higher than that of the original strain.

Compared with the original K. pneumoniae, the recombinant strain converts glycerol to 1,3-propanediol with relatively higher production concentration (67.6 g/l) and higher conversion ratio (0.62 mol/mol). And the time achieving the maximum 1,3-PD concentration advanced about 2 h. However, during the first 12 h of the fermentation, the 1,3-propanediol concentration in recombinant K. pneumoniae fermentation broth was a slightly lower than that of the original strain. When fermentation time arrived at 8 h, IPTG inducing the expression of gene yqhD was added into the fermentation broth, which imposed a metabolic burden and resulted in slowing down the growth rate. After about 20 h, the biomass synthesis of the microorganisms and the expression of yqhD achieved the maximum and reached stabilization. Eventually, the 1,3-PD production of recombinant strain also achieved the maximum. The results confirmed the expression of yqhD as a positive factor in the metabolic pathway for 1,3-PD production.