Chromosome integration of the Vitreoscilla hemoglobin gene (vgb) mediated by temperature-sensitive plasmid enhances γ-PGA production in Bacillus amyloliquefaciens


  • Wei Zhang,

    1. Key Laboratory of Molecular Microbiology and Technology for Ministry of Education, College of Life Sciences, Nankai University, Tianjin, China
    Search for more papers by this author
  • Hui Xie,

    1. Key Laboratory of Molecular Microbiology and Technology for Ministry of Education, College of Life Sciences, Nankai University, Tianjin, China
    Search for more papers by this author
  • Yulian He,

    1. Key Laboratory of Molecular Microbiology and Technology for Ministry of Education, College of Life Sciences, Nankai University, Tianjin, China
    Search for more papers by this author
  • Jun Feng,

    1. Key Laboratory of Molecular Microbiology and Technology for Ministry of Education, College of Life Sciences, Nankai University, Tianjin, China
    Search for more papers by this author
  • Weixia Gao,

    1. Key Laboratory of Molecular Microbiology and Technology for Ministry of Education, College of Life Sciences, Nankai University, Tianjin, China
    Search for more papers by this author
  • Yanyan Gu,

    1. Key Laboratory of Molecular Microbiology and Technology for Ministry of Education, College of Life Sciences, Nankai University, Tianjin, China
    Search for more papers by this author
  • Shufang Wang,

    1. Key Laboratory of Bioactive Materials for Ministry of Education, Nankai University, Tianjin, China
    Search for more papers by this author
  • Cunjiang Song

    Corresponding author
    1. Key Laboratory of Bioactive Materials for Ministry of Education, Nankai University, Tianjin, China
    • Key Laboratory of Molecular Microbiology and Technology for Ministry of Education, College of Life Sciences, Nankai University, Tianjin, China
    Search for more papers by this author

Correspondence: Cunjiang Song, Department of Microbiology, College of Life Science, Nankai University, Tianjin 300071, China. Tel./fax:+86 22 23503866; e-mail:


Oxygen is a limiting factor in the production of γ-PGA by the glutamic acid-independent strain Bacillus amyloliquefaciens LL3 because of the high viscosity of the culture broth. The vgb gene encoding Vitreoscilla hemoglobin (VHb) was introduced into LL3 to overcome the low concentration of dissolved oxygen (DO). First, recombinant plasmid pWHV was constructed by cloning vgb into the Bacillus expression vector pWH1520 and transformed into LL3. Carbon monoxide difference spectral analysis confirmed the expression of VHb. The γ-PGA yield of LL3 (pWHV) under the optimized fermentation conditions increased by 9.56%. To overcome the instability of pWH1520 and to establish stable expression of VHb, the engineered strain LL3-PVK was constructed by homologous recombination between the integration vector pKSVPVK and the 16S rRNA gene of LL3. The temperature-sensitive plasmid was used to perform the integration, which successfully circumvented the obstacle of the low transformation efficiency of B. amyloliquefaciens LL3. Bacillus amyloliquefaciens LL3-PVK showed an increase of 30% in γ-PGA production, while the biomass was increased by 7.9%. To our knowledge, this is the first report describing enhancement of γ-PGA production in a glutamic acid-independent strain as a result of vgb expression.


Vitreoscilla hemoglobin (VHb) is a soluble homodimeric protein, which is synthesized by the Gram-negative bacterium Vitreoscilla (Wakabayashi et al., 1986). VHb has an average oxygen association rate constant (kon) of 78 Μm−1 s−1, while its oxygen dissociation rate constant (koff) of 5000 μM−1 s−1 is hundreds of times larger than that of other globins (Zhang et al., 2007). To date, several hypotheses have been proposed for the mechanism of VHb action, which remain to be investigated: one hypotheses is that VHb facilitates oxygen diffusion and improves aerobic metabolism in hosts (Khosla et al., 1990); two other reports showed that VHb makes a shift in the electron transport chain and leading to higher efficiency of ATP production (Kallio et al., 1994; Tsai et al., 2002); Dikshit et al. (1992) postulated that VHb itself might act as a terminal oxidase. However, the vgb gene encoding VHb has been successfully expressed in various heterologous microorganisms as well as plant and mammal cells to enhance cell density, improve protein production, elevate chemical production, and improve physical conditions, especially under oxygen-limited conditions (Zhang et al., 2007).

Poly-γ-glutamic acid (γ-PGA) is an unusual macromolecular anionic polypeptide, in which d- and l-glutamic acid units are polymerized by γ-amide linkages (Ashiuchi et al., 2001). Because γ-PGA is nontoxic, edible, biodegradable, and rich in free carboxyls, it has a wide range of applications such as hydrogels, flocculants, thickeners, dispersants, drug delivery, cosmetic, and feed additives (Shih & Van, 2001). Most γ-PGA producers are bacilli and are divided into two types according to nutrient requirements: glutamic acid-dependent strains, such as B. subtilis (chungkookjang) (Ashiuchi et al., 2001), B. licheniformis NK-03 (Cao et al., 2010) and B. subtilis RKY3 (Jeong et al., 2010), and glutamic acid-independent strains including B. subtilis C1 (Shih et al., 2005), B. subtilis TAM-4 (Ito et al., 1996) and B. licheniformis A35 (Cheng et al., 1989). The glutamic acid-independent γ-PGA-producing strains have great potential in industrial production systems because of lower cost and simplified process. The culture medium for γ-PGA production becomes highly viscous due to the production of the polymer and exhibits non-Newtonian rheology, which limits oxygen transfer to the cell and inhibits cell growth and γ-PGA production (Richard & Margaritis, 2003). The presence of 2,3-butanediol in cultures of B. licheniformis ATCC 9945A indicated the level of oxygen in the medium no longer supported a fully aerobic mode of metabolism (Birrer et al., 1996). Furthermore, the production of γ-PGA increased substantially (from 6.3 to 23 g L−1) by increasing the supply of oxygen in the fermenter (Cromwick et al., 1996).

In Bacillus subtilis, linear DNA-mediated procedures were used to perform gene disruption or integration based on its ability to form a naturally competent physiological state (Fabret et al., 2002). However, some B. amyloliquefaciens strains cannot develop natural competence as found in Bacillus subtilis, which allows automatic incorporation of DNA and subsequent integration into the chromosomes at a high frequency (Zhang et al., 2011). This makes genetic manipulation of B. amyloliquefaciens difficult, and few studies focusing on its genome modification have been reported. Temperature-sensitive plasmids have the advantage of independence of transformation efficiency and, thus, are widely used in strains recalcitrant to exogenous DNA (Okibe et al., 2011). The introduction of a temperature-sensitive plasmid, pKSV7, in this study, made the chromosome integration possible.

In previous studies, a glutamic acid-independent γ-PGA-producing strain, Bacillus amyloliquefaciens LL3, was isolated, and the process parameters of a 200-L fermentation using LL3 showed that dissolved oxygen decreased greatly [to 1.5 vessel volumes per minute (vvm)] at the exponential phase of growth (Cao et al., 2010; Geng et al., 2011). In this study, the vgb gene was first introduced into LL3 under the control of the xylose promoter on plasmid pWH1520 to examine the effects of VHb on γ-PGA production. Then, vgb gene was integrated into the LL3 genome downstream of the strong constitutive P43 promoter from B. subtilis to establish a stable strain suitable for use in industrial production.

Materials and methods

Strains, plasmids, primers, and culture conditions

The strains, plasmids, and primers used in this study are listed in Table 1. Escherichia coli DH5α was used for plasmid construction. The dam, dcm-deficient E. coli GM2163, was used to prepare the unmethylated E. coli–Bacillus shuttle plasmids. Bacillus subtilis 168 and V. stercoraria ATCC 15218 were the source of the P43 promoter and vgb gene, respectively. All strains were grown at 37 °C in Luria Bertani (LB) medium except during plasmid integration/excision experiments. When required, media were supplemented with ampicillin (Ap; 100 μg mL−1), kanamycin (Km; 10 μg mL−1), chloramphenicol (Cm; 5 μg mL−1), or tetracycline (Tc; 10 μg mL−1).

Table 1. Strains, plasmids, and primers used in this study
Strains, plasmids, or primersDescriptionSource or literature
B. amyloliquefaciens LL3glutamic acid-independent poly-γ-glutamic acid (γ-PGA)-producing strain,preserved in our laboratoryCao et al. (2010)
V. stercoraria ATCC 15218The source of vgbATCC
E. coli DH5asupE44 ΔlacU169 (_80 lacZΔM15) recA1 endA1 hsdR17 (rK-mK+) thi-1gyrA relA1 F- Δ(lacZYA-argF)Tiangen
E. coli GM2163F- dam-13::Tn9 (Camr) dcm-6 hsdR2 (rk-mk+) leuB6 hisG4 thi-1 araC14 lacY1 galK2 galT22 xylA5 mtl-1 rpsL136 (Strr) fhuA31 tsx-78 glnV44 mcrA mcrB1Fermentas
pWB980E. coli–Bacillus shuttle vector, source of kanamycin resistance gene (Kanr)Wu and Wong, (1999)
pWH1520E. coli–Bacillus shuttle vector,pBC16 ori, pBR ori, PxylA, xylR, Tcr, AprYang et al. (2010)
pWHVpWH1520 derivative, vgb expression vector coding for VHbThis study
pKSV7E. coli–Bacillus shuttle vector,colE1, pE194ts, Apr, CmrSmith & Youngman (1992)
T-16SPVKpMD19-T simple vector derivative consists of LL3 16S rRNA gene, promoter P43, vgb, and Kanr for selection.This study
pKSPVKpKSV7 derivative consists of LL3 16S rRNA gene, promoter P43, vgb, and Kanr for selectionThis study

Construction of recombinant plasmids

The 0.5-kb vgb gene was first amplified from the V. stercoraria ATCC 15218 using primers P1 and P2, digested with KpnI/SphI, and ligated into similarly digested pWH1520, an Ecoli-Bsubtilis shuttle vector, to generate pWHV.

The DNA fragments of vgb, P43 promoter, 1.5-kb 16S rRNA gene, and 0.9-kb Kan were amplified by PCR from the V. stercoraria ATCC 15218, B. subtilis 168, B. amyloliquefaciens LL3, and pWB980 with primer pairs P3/P4, P5/P6, P7/P8, and P9/P10 respectively. The 16S rRNA gene was first ligated into the pMD19-T simple vector to generate T-16S. The vgb gene and P43 promoter were then digested with NcoI/SpeI and EcoRI/NcoI, respectively, and ligated into T-16S digested with EcoRI/SpeI, and the resulting plasmid was designated T-16SPV. T-16SPVK was obtained by ligating the Kan gene into T-16SPV using the SpeI and PstI restriction sites.

pKSV7, an E. coliBacillus shuttle vector, preserving a temperature-sensitive origin of replication, replicates stably at 30 °C, but not between 37 and 42 °C (Smith & Youngman, 1992). The fusion fragment incorporating P43, vgb, kan, and homologous 16S rRNA gene was PCR-amplified from T-16SPVK with primer 16S, digested with BamHI/HindIII, and ligated into similarly digested pKSV7 to generate pKSPVK. All plasmid construction was verified by DNA sequencing.

Chromosomal integration of Bacillus amyloliquefaciens LL3

pWHV and pKSPVK were transformed into B. amyloliquefaciens LL3 using the high-osmolality electroporation method (Xue et al., 1999) with slight modifications. Because of the high viscosity of γ-PGA, the LL3 culture was harvested by centrifugation at 8000 × g (4 °C) for 10 min to obtain electro-competent cells. Following four washes in ice-cold electroporation medium (0.5 M sorbitol, 0.5 M mannitol, and 10% glycerol), the cells were suspended in the required volume of the electroporation medium to obtain a cell concentration of 2 × 1010 cfu mL−1. DNA (100 ng) was added to competent cells (100 μL), and the mixture was then transferred to an ice-cold electroporation cuvette (2-mm electrode gap). The electrical pulse generated by a Gene-Pulser (Bio-Rad Laboratories, Richmond, CA) was set at 2.1 kV. After recovery and incubation of cells, transformants were selected on LB agar containing tetracycline (10 μg mL−1) or kanamycin (10 μg mL−1).

The transformants into which pWHV and pKSPVK had been introduced were designated LL3 (pWHV) and LL3 (pKSPVK), respectively. Integration of the vgb-cassette into the LL3 was conducted as shown in Fig. 1. Briefly, LL3 (pKSPVK) was cultured at 42 °C supplemented with chloramphenicol and kanamycin for the selection of single-crossover integration. Resolution of cointegration was obtained by passage of the culture in LB medium with kanamycin only and followed by screening for chloramphenicol-sensitive colonies. Correct integration was verified by PCR amplification and DNA sequencing.

Figure 1.

Scheme of recombination events leading to gene insertion. The single-crossover recombinants were obtained by culturing the transformants at unpermissive condition for replication of the temperature-sensitive plasmid (42 °C for pKSV7). Then, the cointegrant was cultured without antibiotic pressure to facilitate intramolecular recombination so as to obtain desired mutants. Chromosome structures of final mutants were confirmed by PCR. Note that the initial recombination can occur either in the upstream portion of the target gene as shown here or at the downstream portion. In both cases, the final structure of the chromosome is the same.

CO-difference spectral analysis

Some hemoproteins are known to react with CO, hemoglobin for example. The complex consisting of carbon monoxide and reduced hemoglobin has a characteristic peak at 420 nm (Choi et al., 2003). Bacillus amyloliquefaciens LL3, LL3 (pWHV), and LL3-PVK were harvested from culture medium by centrifugation at 12 000 g (4 °C) for 10 min and resuspended in 0.25 volumes of 0.1 M potassium phosphate buffer (pH 7.2). Cells were totally disrupted by sonication on ice. The crude extract was centrifuged at 12 000 g (4 °C) for 10 min to remove cell debris. The cytosolic fractions were diluted to a final protein concentration of 20 mg mL−1 and used directly for carbon monoxide (CO)-difference spectral analysis.

Production of γ-PGA by flask culture

The strains, that is, B. amyloliquefaciens LL3, LL3 (pWHV), and LL3-PVK, were used to inoculate LB medium and aerobically incubated at 37 °C for 12 h with shaking at 120 rpm. A 1% (v/v) inoculum was added to a 500-mL flask containing 100 mL of fermentation medium with shaking at 200 rpm. The fermentation medium was as follows: sucrose 50 g L−1 (NH4)2SO4 2 g L−1, MgSO4 0.6 g L−1, KH2PO4 6 g L−1, K2HPO4 14 g L−1, 2 mL mineral elements including 1 mM FeSO4•4H2O, CaCl2•2H2O, MnSO4•4H2O, and ZnCl2 (pH 7.2). The flask cultures were incubated at 37 °C for 60 h. Then, the fermentation broth was centrifuged at 8000 g (4 °C) for 20 min. The cell pellet was washed three times with dH2O then dried and weighed. γ-PGA was recovered and purified according to the method described previously (Goto & Kunioka, 1992). Fourfold volumes of cold anhydrous ethanol were added to the supernatant followed by incubation at 4 °C overnight. The precipitate was centrifuged at 4000 g (4 °C) again and lyophilized to obtain γ-PGA.

The optimal inducing conditions for LL3 (pWHV) were determined by an orthogonal experimental design (Supporting Information Tables S2 and S3).

Results and discussion

Chromosome integration strategy for Bacillus amyloliquefaciens strains

The Bam HI restriction–modification system of B. amyloliquefaciens prevents its own genome from incorporating exogenous DNA, severely reducing transformation efficiency. The successfully constructed plasmids, pWHV and pKSPVK, were first transformed into the adenine and cytosine methylationdeficient strain, E. coli GM2163 and then treated with BamHI methyltransferase (New England Biolabs, Ipswich, MA) according to the manufacturer's instructions.

Bacillus amyloliquefaciens LL3 is among many of the naturally nontransformable B. amyloliquefaciens strains. Furthermore, we were able to achieve a transformation efficiency of only 7.6 × 102 through electroporation under the optimized conditions (Table S1). These two drawbacks render the commonly used methods for chromosomal integration by nonreplicating plasmids or PCR fragments useless for our purposes. Temperature-sensitive plasmids have the advantage of independent of transformation efficiency, because a single transformant is sufficient for the selection of chromosome integrants. Therefore, we used the temperature-sensitive plasmid pKSV7 to perform the homologous recombination. LL3 (pKSPVK) was first cultured overnight at 42 °C in LB broth and then plated on LB agar plate, both with chloramphenicol and with kanamycin. More than 90% of the colonies were accurate single-crossover transformants (19 of 20). In practice, there is no need to screen single-crossover transformants by PCR, for four to six colonies are enough to perform subsequent selection.

While there are many genetic tools that can be used in genome modification of Bacillus subtilis, few methods are reported for Bacillus amyloliquefaciens, because of its recalcitrant restriction–modification system and low transformation efficiency. The application of temperature-sensitive plasmid pKSV7 successfully solved these problems, because one transformant was enough for subsequent selection. The desired mutations, such as deletion and integration, can be obtained in a temperature-shift manner. This method was used to successfully overcome the obstacle of low transformation efficiency and is applicable in other B. amyloliquefaciens strains recalcitrant to exogenous DNA to allow improved microbial production.

Confirmation of vgb insertional transformants

The introduction of plasmids was verified by plasmid extraction and restriction enzyme digestion according to manufacturers' instructions. PCR was performed with primers designed to confirm the insertion events in B. amyloliquefaciens LL3. The procedure for selection of strains in which vgb was integrated into the genome is schematically shown in Fig. 1. First, single-crossover cells were obtained by culturing LL3 (pKSPVK) at 42 °C. Strains in which integration did not occur were not able to survive in LB with chloramphenicol due to loss of the plasmid. Subsequently, a KanR Cm S strain was obtained after more than six passages at 42 °C in the absence of chloramphenicol. The double-crossover occurred between the homogenous sequences (the up- or downstream of 16S rRNA gene) in the chromosomal and plasmid DNA. Primers designed according to the sequences flanking the homologous arms and the up- and downstream sequences of the kanamycin resistance gene were used to verify the integration. The fragment of approximately 2500 bp containing the region upstream of 16S rRNA gene, vgb, and the kanamycin resistance gene was PCR-amplified with P11/P12. The fragment of approximately 1500 bp containing region downstream of 16S rRNA gene and kanamycin resistance gene was PCR-amplified with P13/P14 (Fig. 2). Sequence analysis of the two fragments showed that the exogenous gene was correctly integrated into the chromosome of the host strain by homologous recombination.

Figure 2.

Confirmation of insertion-carrying recombinant strains by PCR. Primes P11/P12 were used to amplify the region upstream of 16S rRNA gene, vgb, and the kanamycin resistance gene (approximately 2500 bp) in lanes 1–3; and primers P13/P14 were used to amplify the region downstream of 16S rRNA gene and kanamycin resistance gene (approximately 1500 bp) in lanes 4–6 (the genetic structure of genes is shown in Fig. 1). Both fragments exist in insertion-carrying strains but do not exist in wild-type strains. The templates used in lanes 1 and 4 are the chromosome of LL3 and in lanes 2, 3, 5, and 6 are genome of LL3-PVK. Generation of the 2500-bp (lanes 2 and 3) and 1500-bp fragments (lane 5 and 6) indicated that the vgb was successfully inserted into the chromosome of LL3.

vgb was successfully expressed in B. amyloliquefaciens LL3

The biochemical activity of the VHb was tested using the CO-difference spectrum assay, based on the observation of a characteristic 420-nm absorption peak due to CO binding to VHb (Liu & Webster, 1974). This peak was present in B. amyloliquefaciens LL3 (pWHV) induced by xylose and LL3-PVK, indicating the successful expression of vgb in LL3 in both forms of the plasmid or integration into the genome (Fig. 3).

Figure 3.

CO-spectral analysis of LL3 and engineered bacteria. LL3 (pWHV) and LL3-PVK showed characteristic peaks due to the binding of CO to VHb, indicating the expression of vgb in these two recombinants. A little peak was also formed in wild-type strain LL3 for unknown reasons. Data are expressed as one of triplicate experiments, which was similar with the other two.

Previous studies have confirmed that vgb can be expressed in B. subtilis (Su et al., 2010), a close relative of B. amyloliquefaciens. Therefore, we investigated whether vgb could also be expressed in the glutamic acid-independent γ-PGA producer B. amyloliquefaciens LL3 to enhance bacterial growth and γ-PGA production.

First vgb was cloned into the Bacillus expression vector pWH1520 under the control of a xylose-induced promoter. CO-difference spectral analysis and fermentation experiments indicated that vgb was expressed in B. amyloliquefaciens LL3 in an active form and stimulated γ-PGA production (Fig. 3 and Table S2). However, stable expression of the vgb gene required the use of antibiotics in the culture medium, thus adding to the cost of this process, which may therefore, be unsuitable for industrial production. Therefore, vgb was integrated into the chromosome of B. amyloliquefaciens LL3 and expressed stably without a requirement for antibiotic selection. The resulting recombinant LL3-PVK showed slightly higher CO binding activity than LL3 (pWHV) under xylose induction (Fig. 3). Based on the capacity to express VHb constitutively and stably without the need for antibiotics and inducers, LL3-PVK has great potential for large-scale production.

Unexpectedly, a small peak was detected in the CO-difference spectral analysis of wild-type LL3, although no peak was present in the negative control E. coli DH5α (data not shown). No protein sequence was found to be encoded in the genome of LL3 that may function like VHb. Therefore, it is possible that LL3 synthesizes a similar substance with the capacity to bind CO, thus forming the peak. However, a marked difference in CO absorbance between LL3- and vgb-containing recombinants was observed, thus showing that vgb was successfully expressed in LL3.

Comparison between cell densities and γ-PGA yield among flask cultures of B. amyloliquefaciens LL3, LL3 (pWHV), and LL3-PVK

The highest production of γ-PGA for LL3 (pWHV) was gained when 0.7% xylose was added into the medium at OD600 of 0.3 (Table S2). Under the optimized conditions, γ-PGA production of LL3 (pWHV) was 9.56% higher than that of LL3, and the biomass was only slightly higher than that of LL3 (Fig. 4a and b).

Figure 4.

Comparison between γ-PGA yields and biomass of B. amyloliquefaciens LL3, LL3 (pWHV), and LL3-PVK. (a) Time curves of cell growth for LL3 (♦), LL3-PVK (■), and LL3 (pWHV) (▲). Data are expressed as one of triplicate experiments, which was similar with the other two. (b) Time curves of γ-PGA production for LL3 (♦), LL3-PVK (■), and LL3 (pWHV) (▲). Data are expressed as the means of three experiments. (c) γ-PGA production (60 h) and DCW (48 h) of three strains. Data are expressed as the means of five experiments. Standard deviations are shown by error bars.

Stable expression of VHb by LL3-PVK was achieved due to integration of the vgb gene into the chromosome. Under the same fermentation conditions as LL3, as shown in Fig. 4c, LL3-PVK showed slight growth in dry cell weight (DCW) (7.9% higher), while the γ-PGA production increased obviously (30% higher).

The logarithmic phase of both LL3-PVK and LL3 (pWHV) was about 2 h longer than that of LL3. And in the stationary phase, the cell density of LL3-PVK and LL3 (pWHV) was a little higher than that of LL3 (Fig. 4b), which was consistent with the final DCW of the three strains (Fig. 4c).

The expression of vgb enhanced the γ-PGA production and the biomass, although not to the same extent as that previously reported (Su et al., 2010). This may be accounted for by two factors. Firstly as shown in Fig. 3, LL3-PVK showed a moderately increased absorbance of CO compared with wild-type LL3, indicating that the expression level of vgb may be not high enough to facilitate the production of γ-PGA and cell growth. We are now conducting a multicopy insertion of vgb into the chromosome to confirm this hypothesis. On the other hand, the production of γ-PGA by glutamic acid-independent producers is usually lower than that of glutamic acid-dependent strains. Therefore, the limitation of oxygen transfer caused by high viscosity of the medium in LL3 may be less severe than that in strains with higher γ-PGA production. Great efforts are being made to increase γ-PGA production from LL3 by optimization of medium components as well as by genetic modification of its genome. It can be concluded that as the production of γ-PGA by LL3 increases, the enhancement of γ-PGA production by vgb will become more obvious.


This study was financially supported by National key 296 Basic Research Program of China (‘973’-Program) 2012CB725204, National High Technology Research and Development Program of China (‘863’-Program) 2012AA021505, Natural Science Foundation of China Grant Nos. 31070039, 31170030, and 51073081, Project of Tianjin, China (11JCYBJC09500). No author has any conflict of interests to declare.

Authors' contribution

W.Z. and H.X. contributed equally to this work.