Rational improvement of the engineered isobutanol-producing Bacillus subtilis by elementary mode analysis
© Li et al.; licensee BioMed Central Ltd. 2012
Received: 23 March 2012
Accepted: 13 July 2012
Published: 3 August 2012
Isobutanol is considered as a leading candidate for the replacement of current fossil fuels, and expected to be produced biotechnologically. Owing to the valuable features, Bacillus subtilis has been engineered as an isobutanol producer, whereas it needs to be further optimized for more efficient production. Since elementary mode analysis (EMA) is a powerful tool for systematical analysis of metabolic network structures and cell metabolism, it might be of great importance in the rational strain improvement.
Metabolic network of the isobutanol-producing B. subtilis BSUL03 was first constructed for EMA. Considering the actual cellular physiological state, 239 elementary modes (EMs) were screened from total 11,342 EMs for potential target prediction. On this basis, lactate dehydrogenase (LDH) and pyruvate dehydrogenase complex (PDHC) were predicted as the most promising inactivation candidates according to flux flexibility analysis and intracellular flux distribution simulation. Then, the in silico designed mutants were experimentally constructed. The maximal isobutanol yield of the LDH- and PDHC-deficient strain BSUL05 reached 61% of the theoretical value to 0.36 ± 0.02 C-mol isobutanol/C-mol glucose, which was 2.3-fold of BSUL03. Moreover, this mutant produced approximately 70 % more isobutanol to the maximal titer of 5.5 ± 0.3 g/L in fed-batch fermentations.
EMA was employed as a guiding tool to direct rational improvement of the engineered isobutanol-producing B. subtilis. The consistency between model prediction and experimental results demonstrates the rationality and accuracy of this EMA-based approach for target identification. This network-based rational strain improvement strategy could serve as a promising concept to engineer efficient B. subtilis hosts for isobutanol, as well as other valuable products.
KeywordsRational strain improvement Metabolic network Elementary mode analysis Target prediction Bacillus subtilis Isobutanol
Isobutanol is considered as a leading candidate for the replacement of current fossil fuels [1, 2]. Due to global environmental problems and fuel crises, isobutanol is expected to be produced in biotechnological process, which fulfills the demands of green and sustainable energy production . Atsumi et al.  launched isobutanol bio-production in engineered Escherichia coli by harnessing the power of natural L-valine biosynthetic pathways. At present, isobutanol can be biosynthesized in several engineered microorganisms [2, 4–8].
As the best-characterized Gram-positive microorganism, Bacillus subtilis is regarded as a promising isobutanol producer owing to some valuable features. In addition to high isobutanol toxicity tolerance, B. subtilis has no significant codon usage bias, which facilitates the functional heterologous gene expression and pathway engineering. Besides, it can secrete several enzymes to depolymerize polysaccharides that are presented in large amounts in plant, and further utilize some resulted oligosaccharides and C5 sugar (e.g. L-arabinose) , which benefits isobutanol production from low-value feedstocks. So far, B. subtilis has been engineered for isobutanol production , whereas it still needs to be improved for higher yield.
Pathway modifications that direct metabolic flux towards the desired products play an important role in strain optimization. Several corresponding metabolic strategies, such as pathway reconstruction  and cofactor manipulation , have been well applied for metabolic evolution of the isobutanol producers. Nevertheless, these approaches are always time-consuming and subjected to laborious experiments for target validation. As cells are elaborate systems with highly interconnected metabolic networks, it is challenging to capture the full range of behaviors of a cell and identify the accurate targets for efficient strain improvement by analyzing a set of linear pathways.
To solve this problem, it is necessary to investigate the cell behaviors systematically. Current state-of-the-art omics technologies together with the next-generation sequencing promote the progress of systems biology, which allows the quantitative understanding of pathway operations during cellular metabolism by using the mutually related mathematical modeling and experiment . This kind of framework addresses the questions of traditional metabolic engineering, accelerates the strain improvement process, and opens the door to a new era of network-based strain evolution [13–17]. Presently, several computational tools are developed for systematic cellular metabolism analysis . Among all of them, elementary mode analysis (EMA) is acknowledged as a powerful tool to identify the metabolic network properties. Based on the nullspace and convex analysis, as well as the steady-state, EMA decomposes the complex metabolic network of a cell into a set of unique and indivisible pathways, which link all possible cellular physiological states [19, 20]. Knowledge of these pathways allows the rational in silico design of an ideal host with specialized metabolic functionalities. In addition to previous applications for theoretical yield analysis and cellular phenotype prediction, EMA has drawn more and more attentions to develop efficient bioconversion platforms for the desired chemicals [13, 15, 21, 22].
EMA is an attractive approach for strain improvement, whereas few attempts were performed in the isobutanol-producing microorganisms except for the recent cases implemented by Trinh et al.  and Matsuda et al. . Furthermore, EMA has not been employed to explore the metabolic behaviors of B. subtilis until now. Based on the preceding successes, here we presented a network-based EMA strategy to rationally improve the engineered isobutanol-producing B. subtilis. First, the genome-scale metabolic network model of this strain was reconstructed and refined. Then, potential targets that influence isobutanol biosynthesis were identified, and the strain engineering strategy was proposed. Finally, the in silico designed isobutanol-producing B. subtilis mutants were experimentally constructed and further tested to verify the model prediction.
Metabolic network analysis of the isobutanol-producing B. subtilis
Potential target identification
EMA calculates the reaction flux and gives an extensive and profound insight into cell behaviors based on the metabolic network. Thus the potential bottlenecks could be investigated according to the flux distribution of EMs. For a precise prediction, EMs were classified and screened reasonably by taking the actual cellular physiological state into consideration. First, 2,216 EMs were generated by eliminating the extreme EMs from the total 11,342 EMs. Then, the unreasonable EMs (see Methods) were further excluded form 2,216 EMs, which finally resulted in 239 qualified EMs for the following analysis.
Potential targets predicted by EMA based on flux correlation
pyruvate dehydrogenase complex
glucose 6-phosphate dehydrogenase
As stated above, inactivation of LDH and PDHC could obviously redirect the pyruvate flux towards isobutanol biosynthetic pathway, and thus improve the theoretical yield. Therefore, LDH and PDHC were finally chosen for in vivo implementation.
Experimental validation—I. Construction and characterization of the LDH-deficient isobutanol-producing B. subtilis
According to the prediction, gene ldh encoding for LDH was disrupted in BSUL03 by integrating plasmid pUCLKm (Additional file 2, Figure S1A) into the chromosome of BSUL03. Recombinants resistant to kanamycin were selected for PCR confirmation by using a pair of primers ldh-F and ldh-R. BSUL03 showed a 0.9 kb PCR band, while the ldh-disrupted strain BSUL04 showed a 2 kb PCR band (Additional file 2, Figure S1B). Assay of enzyme activity showed that LDH activity of BSUL04 was 0.06 ± 0.01 U/mg, while that was 5.20 ± 0.06 U/mg of BSUL03. These results indicated that LDH activity was destroyed in BSUL04.
Comparison of metabolic profiles of different isobutanol-producing B. subtilis under microaerobic conditions
1.95 ± 0.18
0.16 ± 0.01
3.67 ± 022
3.73 ± 0.29
1.09 ± 0.13
2.11 ± 0.15
0.18 ± 0.02
10.65 ± 1.04
1.82 ± 0.27
1.17 ± 0.12
0.36 ± 0.02
2.28 ± 0.17
0.29 ± 0.01
4.46 ± 0.85
Experimental validation—II. Construction and characterization of the LDH- and PDHC-deficient isobutanol-producing B. subtilis
In B. subtilis, the nonpolar PDHC-deficient mutants can be obtained by the interruptions of any gene involved in pdhABCD operon except for the essential gene pdhA. As the core of PDHC, the E2 subunits encoded by pdhC affect PDHC activity most significantly. Therefore, pdhC was reasonably selected for gene knockout in the present work. Recombinants with kanamycin and tetracycline resistance suggested that pdhC were deleted by integrating the plasmid pUCPTet (Additional file 2, Figure S1C) into the chromosome of BSUL04 via double homologous recombinant. By using a pair of primers pdhC 1-F and pdhC 2-R, the band of BSUL04 and the pdhC-disrupted B. subtilis BSUL05 were 2 and 2.75 kb, respectively (Additional file 2, Figure S1D). PDHC activity of BSUL03 and BSUL04 were 0.31 ± 0.01 U/mg and 0.27 ± 0.01 U/mg, respectively, whereas it could not be detected in BSUL05, implying that PDHC was inactivated in BSUL05.
Strain BSUL05 exhibited a longer transition time to exponential phase: 4 h compared to 2 h in BSUL03 and BSUL04. Besides, biomass (0.68 ± 0.02 g/L) and the specific growth rate (0.17 ± 0.02 h-1) were 35 % and 44 % of BSUL03, respectively (Figure 3). Along with the suppressed cell growth, isobutanol production and intracellular ATP concentration of BSUL05 sharply decreased to 1.17 ± 0.12 g/L and 96 ± 7 nM, respectively. Meanwhile, a conspicuous increase of intracellular pyruvate concentration (approximately by 50-fold) was also observed (Figure 4). Satisfactorily, lactate, acetate and ethanol were undetected in fermentation broth (Table 2). In comparison with BSUL04, BSUL05 doubled isobutanol yield to 0.36 ± 0.02 C-mol/C-mol, which was 61% of the predicted value (0.59 C-mol/C-mol). Though cell growth and isobutanol production were inhibited in BSUL05, both of them could be well restored by external acetate addition (Table 2). Simultaneously, the intracellular ATP concentration increased by 180% to 270 ± 18 nM, and the intracellular pyruvate concentration decreased by 95% to 2.6 mM (Figure 4). However, an unexpected net acetate accumulation was noticed during fermentations (Table 2).
Isobutanol biosynthesis profile of B. subtilis BSUL05
Some progress is being made in strain improvement of the engineered isobutanol producers [2, 4, 6]. However, these strategies are always time-consuming and laborious due to the incomplete understanding of the complex cellular behaviors. Network-based EMA is believed to be an attractive approach to handle this problem, while the relevant investigations are rather limited in isobutanol producers. For that reason, the EMA-based design strategy was first employed as a guiding tool to tailor the engineered B. subtilis for better isobutanol-producing performance .
EMA decomposes a metabolic network into a set of unique and non-divisible pathways that represent all possible physiological states of the cells. Different from algorithms such as MOMA  and OptGene  that could identify only one optimal solution, EMA calculates all the possible optimal pathway solutions. It offers an opportunity to investigate the flux distribution at different performances, comprehend the underlying cellular behaviors, and draft the blueprint for strain improvement.
For a precise prediction, the qualified EMs should be chosen from the total EMs, as only those conformed to the real cellular physiological state are meaningful to strain optimization. Here, two points should be taken into account. One is that the extreme EMs without the synchronous formation of isobutanol and biomass need to be eliminated. The other one is that the central metabolism is active under the oxygen-limited condition despite the expression of the involved genes are downregulated [11, 29]. Therefore, EMs with non-positive carbon flux through glycolysis and the irreversible reactions of PPP and TCA cycle are unreasonable and also need to be excluded. Previous reports showed that linear relationship existed between fluxes through the target candidates and the objective reaction [22, 30]. Therefore, the statistically significant reactions were picked out for further analysis. Both flux flexibility analysis (Table 1) and intracellular flux simulation (Figure 2) proposed LDH and PDHC as the most promising inactivation targets. This fully coincides with the fact that both LDH (K m = 3.0 mM)  and PDHC (K m = 4.3 mM)  possess much higher pyruvate affinity than ALS (K m = 13.6 mM)  in B. subtilis. In addition, LDH can be activated in an FNR (a transcriptional activator for anaerobically induced genes) independent manner under oxygen-limited conditions . Following ldh and pdh C, alsS ilvC and ilvD involved in KIV biosynthetic reactions were identified as potential amplification targets (Table 1), holding an identical view with the previous findings that overexpression of these genes is beneficial to isobutanol production enhancement [2, 5, 6]. Apart from the above targets, potential candidates also included a transhydrogenase and five PPP-related encoding genes, implying that the intracellular redox state maintains close ties with isobutanol biosynthesis.
As analyzed above, the corresponding mutants were experimentally constructed to validate the EMA prediction. Disruption of ldh eliminated lactate and increased isobutanol yield by 12.5 %, meaning that the carbon flux is indeed redirected towards isobutanol as simulated. Although the observations were consistent with the findings from several experimental studies [4, 5], the yield increment was not so much as predicted. We speculated that this could be primarily put down to the undesirably drastic increase of acetate, which is a well-known inhibitor in fermentation. Romero et al.  demonstrated that, in B. subtilis, LDH plays a pivotal role in maintaining the redox equilibrium (NAD(P)H/NAD(P)+) under fermentative conditions. As ketol-acid reductoisomerase and alcohol dehydrogenase for isobutanol biosynthesis require NADPH and NADH as cofactor, respectively, the discrepancy might also be ascribed to the depressed enzyme activity, which was relevant to the disturbed intracellular redox state caused by ldh disruption. This conjecture agreed with the speculation about the intimate connection between cellular redox state and isobutanol biosynthesis during target identification. Further disruption of pdhC obviously doubled isobutanol yield to 0.36 ± 0.02 C-mol/C-mol, and decreased acetate and ethanol to an undetectable level (Table 2). These data suggested that, on one aspect, PDHC inactivation triggered a significant carbon flux shift, which favors isobutanol biosynthesis and was confirmed by intracellular metabolites analysis (unpublished data). On another aspect, byproduct elimination prevented the broth from overacidification and increased isobutanol yield . Therefore, EMA is accurate enough to predict the targets and guide the rational improvement of the engineered isobutanol-producing B. subtilis.
When taking glucose as the sole carbon source, both cell growth and isobutanol production were suppressed in BSUL05 (Figure 3 and Table 2). As shown in Figure 1, EMA reveals an inverse relationship between the yield of isobutanol and biomass owing to the shared precursor competition [15, 22]. Thus, it was plausible to observe that strain with higher isobutanol biosynthetic efficiency exhibited a lower cell growth. Furthermore, the impaired cell growth of this PDHC-deficient strain might also be accounted for the lack of a pivotal building block, acetyl-coenzyme A (AcCoA). As for the decreased isobutanol production, it might be explained by two reasons. On one hand, an appropriate biomass is necessary for the production of desired products, so the lower isobutanol production of BSUL05 might be attributed to the impaired cell growth. On the other hand, the depressed TCA cycle caused by AcCoA shortage led to ATP scarcity (Figure 4), which influenced cell growth and normal expression of the enzymes for isobutanol biosynthesis. Delightedly, we found that external acetate addition could increase the intracellular ATP concentration and restore the impaired cell growth along with isobutanol production (Figure 4). These results turned out that acetate could not only serve as a carbon source, but also as an energy source in BSUL05. Unexpectedly, a net acetate accumulation was observed, which might be relevant to the putative pyruvate oxidase (POX) in B. subtilis. Originally, the small amount of acetate biosynthesized via POX could be assimilated via acetyl-CoA synthetase, whereas this kind of balance might be destroyed by the accelerated metabolism induced by external acetate addition. Fortunately, acetate could be well controlled at a low concentration in fed-batch fermentations (Figure 5), implying the acetate production-consumption equilibrium could be balanced under appropriate conditions.
The rational improved isobutanol-producing B. subtilis BSUL05 displayed much better isobutanol biosynthetic performances than the parental strain BSUL03 both in batch and fed-batch fermentations (Table 2 and Figure 5). However, the maximal isobutanol yield of BSUL05 obtained in the present work (0.36 ± 0.02 C-mol/C-mol) was 39 % lower than the in silico prediction (0.59 C-mol/C-mol) (Figure 1). Such a discrepancy was also noticed by other researchers . It is inferred that this phenomenon might be explained by the following reasons. First, EMA revealed that efficient isobutanol biosynthesis depends on the low oxygen level (data not shown), whereas the precise control of the oxygen-limited condition is really a challenge [4, 23]. Then, as EMA hinted the important role of cellular redox state in isobutanol biosynthesis (Table 1 and Figure 2), this discrepancy might partially ascribe to the disturbed redox equilibrium caused by ldh and pdhC deletion [4, 5]. Next, isobutanol biosynthesis might be inhibited by some negative effects of the cell regulatory system that was not considered by EMA. Besides, some unknown regulation mechanisms could also interpret the fact. Finally, as the ‘just-in-time’ gene transcription (and the associated enzyme expression) shows clear influence on efficient bio-production [37, 38], the lower isobutanol yield of the mutant might be attributed to the uncoordinated isobutanol biosynthetic pipelines, which could not be explored by the metabolic network. Therefore, the aforementioned factors need to be investigated in future studies to construct a more efficient isobutanol-producing B. subtilis. At these points, the improved performance of BSUL05 has confirmed that EMA is a valuable tool for rapid and precise target identification to engineer B. subtilis with stronger isobutanol biosynthetic capability. This metabolic network-based rational strain improvement strategy could be applied to construct valuable industrial B. subtilis hosts for the production of isobutanol and other desired products.
In this work, we presented the first report on rational improvement of the isobutanol-producing B. subtilis by employing an EMA-based strategy. Flux flexibility analysis and in silico simulation predicted LDH and PDHC as the most promising inactivation targets, which were further validated experimentally. The maximal isobutanol yield and titer of the mutant BSUL05 reached 2.3- and 1.7-fold of BSUL03 to 0.36 ± 0.02 C-mol/C-mol and 5.5 ± 0.3 g/L, respectively, showing a stronger isobutanol biosynthetic capability. The consistency between model prediction and experimental results demonstrates that EMA is a reliable approach for target identification and strain optimization. Moreover, EMA dropped a hint of the close relationship between isobutanol biosynthesis and cellular redox state, which provides a valuable insight into further improvement of the isobutanol-producing B. subtilis. Our results demonstrate that the EMA-based prediction could serve as a useful strategy to guide strain engineers towards improved bio-production in B. subtilis, as well as other microorganisms.
Metabolic network reconstruction
The genome-scale metabolic network of isobutanol-producing B. subtilis was constructed by introducing the isobutanol biosynthetic reactions, which allows the conversion from KIV to isobutanol, into the previously described B. subtilis 168 network . For EMA, the model was further refined. Except for the central carbon metabolism, linear reactions involved in other subsystems were lumped when necessary (Additional file 1, Table S1). It takes glucose as the sole carbon source via phosphotransferase system, and contains other substrates ammonium, sulfur and oxygen. Products of isobutanol, lactate, acetate, ethanol, valine, carbon dioxide and biomass, as well as ATP are considered as external metabolites. For ATP production in the respiratory chain, a P/O ratio of 1.33 (for NADH) and 0.89 (for FADH2) was assumed . The biomass composition and cell molecular were taken from Dauner and Sauer . Water, protons and phosphate were assumed to be ubiquitous and unlimited in the cells.
EMA was implemented by METATOOL . The script files and compiled shared library of METATOOL 5.1 were downloaded from the METATOOL website (http://www.biozentrum.uni-wuerzburg.de/bioinformatik/). EMA results were analyzed using Excel Microsoft Corp. for mode sorting and filtering.
Potential target identification based on flux correlation
Here, S stands for the EMs matrix with the dimension of j × i., where i and j are the number of reactions and EMs, respectively. The symbol ξ represents the molar carbon content expressed in C-mol per mol biomass or isobutanol.
Here, and refer to the deviation and the average flux of reaction i, respectively. The variable m refers to the number of reactions in the metabolic network.
Microbial strains and media
Strains and plasmids used in this study
E. coli JM109
recA 1, endA 1, gyrA 96, thi-1, hsdR 17, supE 44, relA 1, Δ(lac-proAB)/F’[traD 36, proAB+, lacIq, lacZ ΔM 15]
ΔamyE::(P43::kivd-adh 2), P43::ilvD-ilvC-alsS; Spcr, Emr
BSUL03 with lactate dehydrogenase inactivation (Δldh); Spcr, Emr, Kmr
BSUL04 with pyruvate dehydrogenase complex E2 subunit inactivation (Δldh ΔpdhC); Spcr, Emr, Kmr, Tetr
E. coli cloning vector; Ampr
B. subtilis integration vector with amyE locus; Ampr, Kmr
E. coli-B. subtilis cloning vector; Ampr, Tetr
pUC18 containing ldh fragment of B. subtilis, Ampr
pUCL containing kanamycin resistance cassettes from pDK; Ampr, Kmr
pUC18 containing homology arm pdhC 1 fragment of B. subtilis; Ampr
pUCP01 containing homology arm pdhC 2 fragment of B. subtilis; Ampr
pUCP02 containing tetracycline resistance cassettes from pHY300PLK; Ampr, Tetr
Gene cloning and plasmid construction
All oligonucleotides used in this work are listed in (Additional file 3 Table S3). Standard techniques for nucleic acid manipulation were used as described by Sambrook et al. . By using genomic DNA of B. subtilis BSUL03 as template, the ldh gene (code for LDH) [GenBank: 938348] was amplified with a pair of primers ldh-F and ldh-R, the homologous arms pdhC 1 and pdhC 2 of pdhC (code for PDHC E2 subunit) [GenBank: 936010] were amplified with two pairs of primers pdhC 1-F and pdhC 1-R, pdhC 2-F and pdhC 2-R, respectively. The Xba I-Xma I digested ldh PCR product was cloned into pUC18 cut with the same enzymes to create pUCL. Then the kanamycin-resistant cassette cut from plasmid pDK with EcoR I and EcoR V was blunted and ligated into the plasmid pUCL cut with EcoR V, creating pUCLKm (Additional file 2, Figure S1A). Plasmid pUCP01 was obtained by cloning the Hind III-Pst I digested pdhC 1 PCR product into pUC18 cut with the same enzymes. Then the BamH I-EcoR I digested pdhC 2 PCR product was cloned into pUCP01 cut with the same enzymes, creating pUCP02. The tetracycline-resistant cassette was amplified by using plasmid pHY300PLK as template with a pair of primers Tet-F and Tet-R. The Pst I-BamH I digested PCR product was then cloned into pUCP02 cut with the same enzymes, creating pUCPTet (Additional file 2, Figure S1C).
Construction of isobutanol-producing B. subtilis mutants
To obtain B. subtilis mutants, the integration vectors pUCLKm and pUCPTet were sequentially transformed into BSUL03 cells by using the competent cell method . B. subtilis recombinants were selected by kanamycin resistance or/and tetracycline resistance, and further confirmed by PCR using two pairs of primers ldh-F/ldh-R and pdhC 1-F/pdhC 2-R, respectively. The LDH mutant and the LDH and PDHC mutant were designated as BSUL04 and BSUL05, respectively.
Pre-cultures were prepared by cultivating one fresh colony in the liquid LB medium at 240 rpm for 8 h. The 1% (v/v) inoculation was adopted in all experiments. Batch fermentations for phenotype growth and metabolic profile assays were carried out in 500 mL screw-cap flasks with LBGSM-I medium under the microaerobic conditions (40% work volume, 200 rpm, 37°C) for 40 h. Fed-batch fermentations were performed in 400 mL LBGSM-III cultures in a fed-batch Pro fermentation system (DASGIP, Germany) under two-stage (aerobic/oxygen-limited) conditions for 60 h. Dissolved oxygen was measured by an oxygen electrode (Mettler Toledo, Germany). For aerobic conditions (0–10 h), it was controlled at 30 ± 1% of saturation in a cascade by stirring from 200 to 700 rpm with 1 volume of air per volume of medium per minute (vvm). For oxygen-limited conditions (10–60 h), it was adjusted to 5 ± 1% by reducing the stirrer and aeration to 50 rpm and 0.5 vvm, respectively. The pH adjusted by 2 M NaOH and 2 M HCl was maintained at 7.0 by a standard pH electrode (Mettler Toledo, Germany). Foam was prohibited by manual injection of appropriate antifoamer (Sigma 204). When glucose concentration fell below 1 g/L, 1.6 mL of glucose feeding solution was added. For PDHC-deficient strains, sodium acetic acid were originally supplemented into the medium at a final concentration of 3 g/L and 3.4 g/L in batch and fed-batch fermentations, respectively. Besides, 0.4 mL of acetate feeding solution was added coupled with glucose feeding solution during feeding period. All the fermentative experiments were performed in triplicate.
Enzyme activity assays
Cells for LDH and PDHC enzyme activity assays were grown at 37°C in 25 mL LB medium, harvested at the exponential phase (6 h) and the beginning of the stationary phase (14 h), respectively, washed twice with cold potassium phosphate buffer (100 mM, pH 7.0), and then suspended in 5 mL of the same buffer. Cell extracts were obtained by sonication (UH-250A, Autoscience instrument Co., Ltd.; 40 k Hz, 10-s pulse and 5-s intervals, total 10 min) and centrifugation (17,900 g for 10 min at 4°C). The supernatant was kept on ice until determination. The standard assay for LDH activity (towards lactate) was spectrophotometrically (340 nm) monitored at 25°C by following the oxidation of NADH . The activity of the PDHC was performed as described by Murarka et al. . Total protein concentrations were measured by Bradford assay .
Measurement of intracellular ATP concentration
The intracellular ATP concentration was measured by the ATP assay kit (Beyotime, China). Cells (1 × 104) harvested at the mid-log phase were mixed with 20 μL lysis buffer and homogenized by vortex, and then centrifuged at 15,300 g for 5 min at 4°C. Other procedures such as lysis buffer preparation, background luminescence correction and ATP measurement were performed as the ATP Assay protocol (Beyotime). The emitted light by luminescent reaction was quantified by a luminometer (Synergy H4 Hybrid Microplate reader, BioTek, USA). ATP concentrations of the samples were calculated from the standard curve using linear regression (0.5-50 nM). Each experiment was carried out in triplicate and each sample was measured five times.
Intracellular pyruvate analysis
To analyze the intracellular pyruvate, 5 mL mid-log phase cell suspension was harvested using vacuum filtration (AP-01P Vacuum Pump, Tianjin Auto Science Co., Ltd.; cellulose nitrate, 0.22 μm pore size) and washed three times each with 20 mL 0.9 % cold NaCl solution (the whole filtration procedure including the washing was finished in less than 30 s) . Metabolite extraction and sample preparation were carried out as described . Sampling was carried out five times in parallel for each experiment.
Intracellular pyruvate analysis was implemented by liquid chromatography-tandem quadrupole mass spectrometry (Bruker MicroTOF-Q II, Germany) equipped with a 250 × 4.6 mm aminopropyl column (Luna NH2, 5 μm particle size, Phenomenex). The detailed separation and elution conditions, as well as data analysis approaches were carried out as described previously .
Cell concentration was determined by measuring the optical density of culture broth at 600 nm (OD600). Biomass was calculated by multiplying OD600 by a conversion factor of 0.325. The quantitative analysis of glucose and fermentation metabolites (alcohols, organic acids and other compounds) were also performed as before .
elementary mode analysis
pyruvate dehydrogenase complex
pentose phosphate pathway
- TCA cycle:
tricarboxylic acid cycle
The authors appreciate the kind donation of plasmid pDK from Dr. Danier R. Zeigler and the Bacillus Genetic Stock Center (BGSC), The Ohio State University. This research was financially supported by the National 973 Project of China (No. 2007CB714302), the Key Program of National Natural Science Foundation of China (Grant No. 20936002), National Natural Science Foundation of China (No. 20976124), Specialized Research Fund for the Doctoral Program of Higher Education (20110032130005) and the Programme of Introducing Talents of Discipline to Universities (No. B06006).
- Dellomonaco C, Fava F, Gonzalez R: The path to next generation biofuels: successes and challenges in the era of synthetic biology. Microb Cell Fact. 2010, 9: 3- 10.1186/1475-2859-9-3View ArticleGoogle Scholar
- Atsumi S, Hanai T, Liao JC: Non-fermentative pathways for synthesis of branched-chain higher alcohols as biofuels. Nature. 2008, 451: 86-U13. 10.1038/nature06450View ArticleGoogle Scholar
- Fava F, Ohtake H, Pesaresi P: Biotechnology for a more sustainable environment decontamination and energy production. J Biotechnol. 2012, 157: 443-445. 10.1016/j.jbiotec.2012.01.029View ArticleGoogle Scholar
- Blombach B, Riester T, Wieschalka S, Ziert C, Youn JW, Wendisch VF, Eikmanns BJ: Corynebacterium glutamicum tailored for efficient isobutanol production. Appl Environ Microbiol. 2011, 77: 3300-3310. 10.1128/AEM.02972-10View ArticleGoogle Scholar
- Smith KM, Cho KM, Liao JC: Engineering Corynebacterium glutamicum for isobutanol production. Appl Microbiol Biotechnol. 2010, 87. 10.5-1055.View ArticleGoogle Scholar
- Li SS, Wen JP, Jia XQ: Engineering Bacillus subtilis for isobutanol production by heterologous Ehrlich pathway construction and the biosynthetic 2-ketoisovalerate precursor pathway overexpression. Appl Microbiol Biotechnol. 2011, 91: 577-589. 10.1007/s00253-011-3280-9View ArticleGoogle Scholar
- Higashide W, Li YC, Yang YF, Liao JC: Metabolic engineering of Clostridium cellulolyticum for production of isobutanol from cellulose. Appl Environ Microbiol. 2011, 77: 2727-2733. 10.1128/AEM.02454-10View ArticleGoogle Scholar
- Chen X, Nielsen KF, Borodina I, Kielland-Brandt MC, Karhumaa K: Increased isobutanol production in Saccharomyces cerevisiae by overexpression of genes in valine metabolism. Biotechnol Biofuels. 2011, 4: 21- 10.1186/1754-6834-4-21View ArticleGoogle Scholar
- Stülke J, Hillen W: Regulation of carbon catabolism in Bacillus species. Annu Rev Microbiol. 2000, 54: 849-880. 10.1146/annurev.micro.54.1.849View ArticleGoogle Scholar
- Atsumi S, Wu TY, Eckl EM, Hawkins SD, Buelter T, Liao JC: Engineering the isobutanol biosynthetic pathway in Escherichia coli by comparison of three aldehyde reductase/alcohol dehydrogenase genes. Appl Microbiol Biotechnol. 2010, 85: 651-657. 10.1007/s00253-009-2085-6View ArticleGoogle Scholar
- Bastian S, Liu X, Meyerowitz JT, Snow CD, Chen MMY, Arnold FH: Engineered ketol-acid reductoisomerase and alcohol dehydrogenase enable anaerobic 2-methylpropan-1-ol production at theoretical yield in Escherichia coli. Metab Eng. 2011, 13: 345-352. 10.1016/j.ymben.2011.02.004View ArticleGoogle Scholar
- Cassman M, Arkin A, Doyle F, Katagiri F, Lauffenburger D, Stokes C: Systems Biology. 2007, Springer Verlag: International Research and Development,View ArticleGoogle Scholar
- Becker J, Zelder O, Häfner S, Schröder H, Wittmann C: From zero to hero—Design-based systems metabolic engineering of Corynebacterium glutamicum for L-lysine production. Metab Eng. 2011, 13: 159-168. 10.1016/j.ymben.2011.01.003View ArticleGoogle Scholar
- Park JH, Lee KH, Kim TY, Lee SY: Metabolic engineering of Escherichia coli for the production of L-valine based on transcriptome analysis and in silico gene knockout simulation. P Natl Acad Sci USA. 2007. 10.: 7797-7802.View ArticleGoogle Scholar
- Unrean P: Trinh CT. Srienc F: Rational design and construction of an efficient E. coli for production of diapolycopendioic acid. Metab Eng. 2010, 12: 112-122.Google Scholar
- Brochado AR, Matos C, Møller BL, Hansen J, Mortensen UH, Patil KR: Improved vanillin production in baker's yeast through in silico design. Microb Cell Fact. 2010, 9: 84- 10.1186/1475-2859-9-84View ArticleGoogle Scholar
- Asadollahi MA, Maury J, Patil KR, Schalk M, Clark A, Nielsen J: Enhancing sesquiterpene production in Saccharomyces cerevisiae through in silico driven metabolic engineering. Metab Eng. 2009, 11: 328-334. 10.1016/j.ymben.2009.07.001View ArticleGoogle Scholar
- Ruppin E, Papin JA, de Figueiredo LF, Schuster S: Metabolic reconstruction, constraint-based analysis and game theory to probe genome-scale metabolic networks. Curr Opin Biotechnol. 2010, 21: 502-510. 10.1016/j.copbio.2010.07.002View ArticleGoogle Scholar
- Schuster S, Fell DA, Dandekar T: A general definition of metabolic pathways useful for systematic organization and analysis of complex metabolic networks. Nat Biotechnol. 2000, 18: 326-332. 10.1038/73786View ArticleGoogle Scholar
- Trinh CT, Wlaschin A, Srienc F: Elementary mode analysis: a useful metabolic pathway analysis tool for characterizing cellular metabolism. Appl Microbiol Biotechnol. 2009, 81: 813-826. 10.1007/s00253-008-1770-1View ArticleGoogle Scholar
- Driouch H, Melzer G, Wittmann C: Integration of in vivo and in silico metabolic fluxes for improvement of recombinant protein production. Metab Eng. 2012, 14: 47-58. 10.1016/j.ymben.2011.11.002View ArticleGoogle Scholar
- Melzer G, Esfandabadi ME, Franco-Lara E, Wittmann C: Flux Design: In silico design of cell factories based on correlation of pathway fluxes to desired properties. BMC Syst Biol. 2009, 3: 120- 10.1186/1752-0509-3-120View ArticleGoogle Scholar
- Trinh CT, Li J, Blanch HW, Clark DS: Redesigning Escherichia coli metabolism for anaerobic production of isobutanol. Appl Environ Microbiol. 2011, 77: 4894-4904. 10.1128/AEM.00382-11View ArticleGoogle Scholar
- Matsuda F, Furusawa C, Kondo T, Ishii J, Shimizu H, Kondo A: Engineering strategy of yeast metabolism for higher alcohol production. Microb Cell Fact. 2011. 10. 70.View ArticleGoogle Scholar
- Sauer U, Hatzimanikatis V, Hohmann HPGM, Manneberg M, van Loon AP, Bailey JE: Physiology and metabolic fluxes of wild-type and riboflavin-producing Bacillus subtilis. Appl Environ Microbiol. 1996, 62: 3687-3696.Google Scholar
- Gao HC, Jiang X, Pogliano K, Aronson AI: The E1 beta and E2 subunits of the Bacillus subtilis pyruvate dehydrogenase complex are involved in regulation of sporulation. J Bacteriol. 2002, 184: 2780-2788. 10.1128/JB.184.10.2780-2788.2002View ArticleGoogle Scholar
- Segrè D, Vitkup D, Church GM: Analysis of optimality in natural and perturbed metabolic networks. P Natl Acad Sci USA. 2002, 99: 15112-15117. 10.1073/pnas.232349399.View ArticleGoogle Scholar
- Patil KR, Rocha I, Förster J, Nielsen J: Evolutionary programming as a platform for in silico metabolic engineering. BMC Bioinformatics. 2005, 6: 308- 10.1186/1471-2105-6-308View ArticleGoogle Scholar
- Nakano MM, Zuber P: Anaerobic growth of a "strict aerobe" (Bacillus subtilis). Annu Rev Microbiol. 1998, 52: 165-190. 10.1146/annurev.micro.52.1.165View ArticleGoogle Scholar
- Choi HS, Lee SY, Kim TY, Woo HM: In Silico Identification of Gene Amplification Targets for Improvement of Lycopene Production. Appl Environ Microbiol. 2010, 76: 3097-3105. 10.1128/AEM.00115-10View ArticleGoogle Scholar
- Garvie EI: Bacterial lactate dehydrogenases. Microbiol Mol Biol Rev. 1980, 44: 106-139.Google Scholar
- Hodgson JA, Lowe PN, Perham RN: Wild-type and mutant forms of the pyruvate dehydrogenase multienzyme complex from Bacillus subtilis. Biochem J. 1983, 211: 463-472.View ArticleGoogle Scholar
- Atsumi S, Li Z, Liao JC: Acetolactate synthase from Bacillus subtilis serves as a 2-ketoisovalerate decarboxylase for isobutanol biosynthesis in Escherichia coli. Appl Environ Microbiol. 2009, 75: 6306-6311. 10.1128/AEM.01160-09View ArticleGoogle Scholar
- Romero S, Merino E, Bolívar F, Gosset G, Martinez A: Metabolic engineering of Bacillus subtilis for ethanol production: lactate dehydrogenase plays a key role in fermentative metabolism. Appl Environ Microbiol. 2007, 73: 5190-5198. 10.1128/AEM.00625-07View ArticleGoogle Scholar
- Jian J, Zhang SQ, Shi ZY, Wang W, Chen GQ, Wu QO: Production of polyhydroxyalkanoates by Escherichia coli mutants with defected mixed acid fermentation pathways. Appl Microbiol Biotechnol. 2010, 87: 2247-2256. 10.1007/s00253-010-2706-0View ArticleGoogle Scholar
- Petersohn A, Antelmann H, Gerth U, Hecker M: Identification and transcriptional analysis of new members of the σB regulon in Bacillus subtilis. Microbiology. 1999, 145: 869-880. 10.1099/13500872-145-4-869View ArticleGoogle Scholar
- Medema MH, Breitling R, Bovenberg R, Takano E: Exploiting plug-and-play synthetic biology for drug discovery and production in microorganisms. Nat Rev Microbiol. 2011, 9: 131-137. 10.1038/nrmicro2478View ArticleGoogle Scholar
- Liu S, Zhang DX, Wang M, Cui WJ, Chen KK, Du GC, Chen J, Zhou ZM: The order of expression is a key factor in the production of active transglutaminase in Escherichia coli by co-expression with its pro-peptide. Microb Cell Fact. 2011. 10. 112.View ArticleGoogle Scholar
- Oh YK, Palsson BO, Park SM, Schilling CH, Mahadevan R: Genome-scale reconstruction of metabolic network in Bacillus subtilis based on high-throughput phenotyping and gene essentiality data. J Biol Chem. 2007, 282: 28791-28799. 10.1074/jbc.M703759200View ArticleGoogle Scholar
- Sauer U, Bailey JE: Estimation of P-to-O ratio in Bacillus subtilis and its influence on maximum riboflavin yield. Biotechnol Bioeng. 1999, 64: 750-754. 10.1002/(SICI)1097-0290(19990920)64:6<750::AID-BIT15>3.0.CO;2-SView ArticleGoogle Scholar
- Dauner M, Sauer U: Stoichiometric growth model for riboflavin-producing Bacillus subtilis. Biotechnol Bioeng. 2001, 76: 132-143. 10.1002/bit.1153View ArticleGoogle Scholar
- Pfeiffer T, Sánchez-Valdenebro I, Nuño JC, Montero F, Schuster S: METATOOL: for studying metabolic networks. Bioinformatics. 1999, 15: 251-257. 10.1093/bioinformatics/15.3.251View ArticleGoogle Scholar
- Sambrook J: Russel DW: Molecular cloning: a laboratory manual. 2001, Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press, 3,Google Scholar
- Anagnostopoulos C, Spizizen J: Requirements for transformation in Bacillus subtilis. J Bacteriol. 1961, 81: 741-746.Google Scholar
- Savijoki K, Palva A: Molecular genetic characterization of the L-lactate dehydrogenase gene (ldhL) of Lactobacillus helveticus and biochemical characterization of the enzyme. Appl Environ Microbiol. 1997, 63: 2850-2856.Google Scholar
- Murarka A, Clomburg JM, Moran S, Shanks JV, Gonzalez R: Metabolic analysis of wild-type Escherichia coli and a pyruvate dehydrogenase complex (PDHC)-deficient derivative reveals the role of PDHC in the fermentative metabolism of glucose. J Biol Chem. 2010, 285: 31548-31558. 10.1074/jbc.M110.121095View ArticleGoogle Scholar
- Bradford MM: A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Chem. 1976, 72: 248-254.Google Scholar
- Bolten CJ, Kiefer P, Letisse F, Portais JC, Wittmann C: Sampling for metabolome analysis of microorganisms. Anal Chem. 2007, 79: 3843-3849. 10.1021/ac0623888View ArticleGoogle Scholar
- Wang ZW, Chen T, Ma XH, Shen Z, Zhao XM: Enhancement of riboflavin production with Bacillus subtilis by expression and site-directed mutagenesis of zwf and gnd gene from Corynebacterium glutamicum. Bioresour Technol. 2011. 10.: 3934-3940.View ArticleGoogle Scholar
- Yuan G, Wong SL: Regulation of groE expression in Bacillus subtilis the involvement of the σA-like promoter and the roles of the inverted repeat sequence (CIRCE). J Bacteriol. 1995, 177: 5427-5433.Google Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.