Fermentation of a recombinant strains of K. pneumoniae harboring different AHAS isoenzymes
We first prepared plasmids in which genes that functioned in isobutanol synthesis were sequentially linked with different acetolactate synthase isoenzymes (ilvIH, ilvBN, and ilvGM) (Table 1), and introduced them into K. pneumoniae Cu ΔldhAΔbudA. We then examined the effect of different AHAS isoenzymes on isobutanol production in the recombinant strains using batch fermentation in a 5-L jar-fermenter for 24 h, with a pH 6.0 (maintained using ammonium water) and aeration of 0.2 vvm.
Table 1
Strain or plasmid
|
Genotype and description
|
Source
|
Strain
|
|
|
E. coli DH5α
|
Host of plasmid
|
Lab stock
|
K. pneumoniae Cu ΔldhAΔbudA
|
|
(32)
|
E. coli BL21 (DE3)
|
Host of expression vector for acetolactate synthase isoenzymes
|
Novagen
(Madison, WI, USA)
|
Plasmid
|
|
|
pBR-lac-IH-ISO
|
Lac promoter, pBR322 carrying ilvIH-ilvC-ilvD-kivd-adhA, Tet.R
|
This study
|
pBR-lac-BN-ISO
|
Lac promoter, pBR322 carrying ilvBN-ilvC-ilvD-kivd-adhA, Tet.R
|
This study
|
pBR-lac-GM-ISO
|
Lac promoter, pBR322 carrying ilvGM-ilvC-ilvD-kivd-adhA, Tet.R
|
This study
|
pUC-lac-BN-ISO
|
Lac promoter, pUC19 carrying ilvBN-ilvC-ilvD-kivd-adhA, Tet.R
|
This study
|
pSTV-lac-BN-ISO
|
Lac promoter, pSTV28 carrying ilvBN-ilvC-ilvD-kivd-adhA, Tet.R
|
This study
|
pUC-tac-BN-ISO
|
Tac promoter, pUC19 carrying ilvBN-ilvC-ilvD-kivd-adhA, Tet.R
|
This study
|
pUC-bud-BN-ISO
|
Bud promoter, pUC19 carrying ilvBN-ilvC-ilvD-kivd-adhA, Tet.R
|
This study
|
pET28a(+)
|
T7 promoter, pET-28a(+) with N-terminal His-tag, Kan.R
|
Novagen
(Madison, WI, USA)
|
ilvIH/pET28a(+)
|
T7 promoter, pET-28a(+) carrying ilvIH with N-terminal His-tag, Kan.R
|
This study
|
ilvBN/pET28a(+)
|
T7 promoter, pET-28a(+) carrying ilvBN with N-terminal His-tag, Kan.R
|
This study
|
ilvGM/pET28a(+)
|
T7 promoter, pET-28a(+) carrying ilvGM with N-terminal His-tag, Kan.R
|
This study
|
(Tet., tetracycline; Kan., kanamycin) |
Each of the three recombinant strains consumed all the glycerol after 24 h, and K. pneumoniae Cu ΔldhAΔbudA/pBR-lac-ilvBN-ISO had the highest growth (OD600 nm = 4.13; Table 2); however, this strain also had the lowest production of 1,3-PDO (6.65 g/L). Aside from isobutanol, the total production of metabolites derived from the oxidation pathway (ethanol, acetate, and succinate) also differed among the three strains. The maximum isobutanol production was from K. pneumoniae Cu ΔldhAΔbudA/pBR-lac-ilvBN-ISO (0.48 g/L). These results can be explained by a restriction of carbon flux to metabolites of the oxidative pathway due to the higher enzymatic activity of ilvBN relative to ilvIH and ilvGM.
Table 2
Metabolites produced by glycerol fermentation of K. pneumoniae strains that had different recombinant acetohydroxyacid synthase isoenzymes.
Metabolite (g/L)
|
CuΔldhAΔbudA/pBR-lac-ilvIH-ISO
|
CuΔldhAΔbudA/pBR-lac-ilvBN-ISO
|
CuΔldhAΔbudA/pBR-lac-ilvGM-ISO
|
Residual glycerol
|
0.00 ± 0.00
|
0.00 ± 0.00
|
0.00 ± 0.00
|
OD600 nm
|
3.84 ± 0.13
|
4.13 ± 0.12
|
3.91 ± 0.11
|
1,3-Propanediol
|
7.61 ± 0.10
|
6.65 ± 0.08
|
7.13 ± 0.09
|
Isobutanol
|
0.32 ± 0.04
|
0.48 ± 0.09
|
0.36 ± 0.05
|
2,3-Butanediol
|
0.00 ± 0.00
|
0.00 ± 0.00
|
0.00 ± 0.00
|
Ethanol
|
1.61 ± 0.08
|
1.23 ± 0.07
|
1.48 ± 0.09
|
Lactate
|
0.00 ± 0.00
|
0.00 ± 0.00
|
0.00 ± 0.00
|
Acetate
|
0.88 ± 0.09
|
1.03 ± 0.08
|
0.95 ± 0.08
|
Succinate
|
0.24 ± 0.11
|
0.12 ± 0.09
|
0.19 ± 0.01
|
Analysis of AHAS isoenzyme activity
To better understand the role of AHAS on isobutanol production in the three recombinant strains of K. pneumoniae, we measured the in vitro enzymatic activity of the different AHAS isoenzymes. These AHAS isoenzymes convert pyruvate into α-acetolactate. α-acetolactate can be converted to acetoin by decarboxylation, which can be measured using a simple colorimetric assay (17). Thus, we indirectly measured the enzymatic activity of the different AHAS isoenzymes using the colorimetric assay for acetoin.
The results indicated that ilvBN had significantly higher acetolactate synthase activity (13.26 U/mg protein) than ilvIH (0.64 U/mg protein) and ilvGM (1.26 U/mg protein) (Fig. 2). Previous research reported that AHAS I (encoded by ilvBN) had much higher affinity for pyruvate than 2-ketobutyrate, relative to AHAS II and AHAS III (encoded by ilvGM and ilvIH) (18). However, our in vitro assay showed that ilvBN had significantly higher activity than ilvIH and ilvGM, but difference in isobutanol production between these AHAS isoenzymes in cells were not as great as in vitro enzymatic activity. The synthesis of the three AHAS isoenzymes (ilvBN, ilvGM, and ilvIH) is strictly controlled by the cellular levels of valine, leucine, and isoleucine, and cells regulate the activities of these AHAS isoenzymes to satisfy the need for the branched amino acids used for cell growth (19). There is also evidence that valine functions as a feedback-inhibitor of AHAS enzyme activity in vitro, and that the extent of this inhibition varies among species. In particular, studies of Corynebacterium glutamicum reported the maximum AHAS activity decreased by 50% in the presence of 10 mM valine (20), studies of E. coli reported the maximum AHAS activity decreased by 80% in the presence of 4.8 µM valine (21), and studies in S. cerevisiae reported the maximum AHAS activity decreased by 74.4% in the presence of 1.0 mM valine (22).
In general, cellular α-acetolactate cell is mainly used for the synthesis of 2,3-butanediol, and only a small portion is used for the synthesis of valine and branched amino acids. A previous study reported that knockout of budA, a gene needed for synthesis of 2,3-butanediol, led to increased production of α-acetolactate, and increased production of valine and isobutanol (14). We therefore suggest that the activity of AHAS was decreased by intracellular feedback inhibition from branched amino acids, such as valine.
Effect of different promoters and plasmid copy numbers on isobutanol production
The expression of recombinant proteins is influenced by physical and transcriptional conditions, including temperature, agitation speed, promoter characteristics, and number of copies of the expression vector. The main factors affecting metabolic burden are promoter strength and origin of replication (ori), which influences the copy number of an expression vector (23). Therefore, we examined the effect of different promoters and plasmid copy numbers on isobutanol production.
Effect of plasmid copy number on isobutanol production
We first determined the effect of plasmid copy number on isobutanol production by performing in batch fermentation is a 5-L jar-fermenter for 24 h at pH 6.0 (maintained using ammonium water) and aeration at 0.2 vvm. Thus, we measured isobutanol production using constructs that had low (pSTV28), medium (pBR322), or high (pUC19) copy numbers of plasmids at the same inducer concentration (Table 1 and Table 3). The results showed that isobutanol production increased as the plasmid copy number increased, in order of pUC19 (0.73 g/L), pBR322 (0.48 g/L), and pSTV28 (0.39 g/L) (Table 3). Multi-copy plasmids are traditionally used for the overexpression of heterologous genes, and changing the copy number of a plasmid is an easy method to adjust gene copy number (24). Furthermore, cells with many plasmids might have greater expression of genes on that plasmid than cells with a small number of plasmids. Thus, if the genes residing on a plasmid are responsible for synthesizing metabolites, greater gene expression from a larger number of plasmids could generate larger amounts of a product. On the other hand, the combination of a high copy number and weak promoter activity could lead to excessive gene expression and a significant metabolic burden (25). Therefore, it is also necessary to select a promoter that is suitable for the high copy number plasmid.
Table 3
Metabolites produced by glycerol fermentation of K. pneumoniae strains that had different recombinant plasmid copy numbers.
Metabolite (g/L)
|
CuΔldhAΔbudA/pUC-lac-ilvBN-ISO
|
CuΔldhAΔbudA/pBR-lac-ilvBN-ISO
|
CuΔldhAΔbudA/pSTV-lac-ilvBN-ISO
|
Residual glycerol
|
0.00 ± 0.00
|
0.00 ± 0.00
|
0.00 ± 0.00
|
OD600nm
|
4.07 ± 0.14
|
4.13 ± 0.12
|
4.19 ± 0.10
|
1,3-Propanediol
|
6.11 ± 0.11
|
6.65 ± 0.08
|
6.87 ± 0.13
|
Isobutanol
|
0.73 ± 0.04
|
0.48 ± 0.09
|
0.39 ± 0.03
|
2,3-Butanediol
|
0.00 ± 0.00
|
0.00 ± 0.00
|
0.00 ± 0.00
|
Ethanol
|
1.03 ± 0.13
|
1.23 ± 0.07
|
1.52 ± 0.08
|
Lactate
|
0.00 ± 0.00
|
0.00 ± 0.00
|
0.0 ± 0.00
|
Acetate
|
1.08 ± 0.10
|
1.03 ± 0.08
|
1.27 ± 0.12
|
Succinate
|
0.13 ± 0.07
|
0.12 ± 0.09
|
0.19 ± 0.05
|
Effect of different promoters on isobutanol production
We next investigated the effect of different promoters (Plac, Ptac, and Pbud; Table 1) on isobutanol production by using the high copy number plasmid (pUC19) and performing 24-h batch fermentation as above (5-L jar-fermenter, pH 6.0, aeration of 0.2 vvm). The results indicated the highest isobutanol production was obtained using the Ptac promoter (0.82 g/L; Table 4).
Table 4
Metabolites produced by glycerol fermentation of K. pneumoniae strains that had different recombinant promoters.
Metabolite (g/L)
|
CuΔldhAΔbudA/pUC-lac-ilvBN-ISO
|
CuΔldhAΔbudA/pUC-tac-ilvBN-ISO
|
CuΔldhAΔbudA/pUC-bud-ilvBN-ISO
|
Residual glycerol
|
0.00 ± 0.00
|
0.00 ± 0.00
|
0.00 ± 0.00
|
OD600nm
|
4.07 ± 0.14
|
4.12 ± 0.16
|
3.98 ± 0.11
|
1,3-Propanediol
|
6.11 ± 0.11
|
6.02 ± 0.12
|
6.66 ± 0.16
|
Isobutanol
|
0.73 ± 0.04
|
0.82 ± 0.05
|
0.64 ± 0.04
|
2,3-Butanediol
|
0.00 ± 0.00
|
0.0 ± 0.00
|
0.00 ± 0.00
|
Ethanol
|
1.03 ± 0.13
|
0.94 ± 0.16
|
1.10 ± 0.14
|
Lactate
|
0.00 ± 0.00
|
0.0 ± 0.00
|
0.00 ± 0.00
|
Acetate
|
1.08 ± 0.10
|
0.99 ± 0.09
|
1.13 ± 0.13
|
Succinate
|
0.13 ± 0.07
|
0.11 ± 0.08
|
0.21 ± 0.05
|
For most applications in metabolic engineering, it is necessary to replace the native promoter on the heterologous gene with a promoter specific to the host organism or with a promoter that allows more precise control over gene expression. The strength of a promoter can influence on the amount of enzyme that is produced and the flux through a pathway (25). The lac promoter and tac promoter are from E. coli and are strong promoters commonly used as heterogeneous inducible promoters (26). The tac promoter is a synthetic promoter, and is more inducible than the lac promoter. In contrast, the bud promoter is a constitutive promoter that functions as an endogenous promoter of in 2,3-butanediol biosynthesis in K. pneumoniae. Previous studies reported that Ptac had high promoter strength in the presence of an inducer in K. pneumoniae and also had high basal expression, indicating it could be used as a highly efficient constitutive promoter. On the other hand, it was reported that Pbud was a stable constitutive promoter in K. pneumoniae, but was not a strong promoter (27). Generally, a gene expression system that increases DNA copy number by increasing plasmid copy number and increases transcript stability by increasing the stability of mRNA control elements will lead to increased levels of transcript and protein. If these factors are not met, then an increase in plasmid copy number could increase burden on the cell, and lead to reduced transcription and translation (24). Based on these considerations, our results indicated that a high plasmid copy number and a strong promoter were needed to increase metabolic flux into the isobutanol biosynthetic pathway.
Effect of fermentation process parameters on isobutanol production
Each metabolite produced by K. pneumoniae requires appropriate process conditions. Therefore, we examined the effect of culture pH and agitation speed (which is related to oxygen level) on the production of isobutanol.
Effect of pH on isobutanol production
We cultured the K. pneumoniae Cu ΔldhAΔbudA, pUC-tac-BN-ISO strain using fed-batch fermentation in a 5-L jar-fermenter for 48 h, with aeration of 0.2 vvm and an agitation speed of 200 rpm, and adjusted the pH to 5.0, 6.0, or 7.0 using ammonium water (Fig. 3). After culture for 48 h, the results showed that cell growth and glycerol consumption increased as pH increased, but isobutanol production was highest at pH 6.0 (1.02 ± 0.03 g/L). In addition, as pH increased there was increased production of 1,3-propanediol, succinate, acetic acid, and ethanol. These results indicated that acidic conditions favor isobutanol production, but strongly acidic conditions inhibit cell growth and productivity. Therefore, the pH condition was set to pH 6.0 to investigate the subsequent effect of the agitation speed as fermentation process parameters.
Previous studies found increased formation of 2,3-dihydroxyisovalerate (2,3-DHIV), a precursor of isobutanol and valine (Fig. 1), under weak acidic conditions (pH 6.5) (28). This may explain why a low pH had a positive effect on isobutanol synthesis. We also found that a higher pH increased the synthesis of acetic acid and succinate, consistent with the results of previous studies (28). These results might be explained by the increased conversion of pyruvate into acetyl-CoA at high pH (29). Acetyl-CoA can be converted to acetaldehyde, which is then converted to acetic acid. Other research suggested that the reason for the increased production of succinate at high pH was the increased amount of acetyl-CoA entering the TCA cycle (28).
Effect of agitation speed on isobutanol production
We then cultured the K. pneumoniae Cu ΔldhAΔbudA, pUC-tac-BN-ISO strain in fed-batch fermentation using a 5-L jar-fermenter for 48 h at pH 6.0, aeration of 0.2 vvm, and adjusted the agitation speed to 100, 200, 300, or 400 rpm (Fig. 4).
The results indicated that as agitation speed increased, cell growth increased but glycerol consumption decreased. In addition, an increased agitation speed also decreased the production of 1,3-propanediol, ethanol, and succinate. Thus, agitation speed altered the consumption of glycerol and the carbon flux of consumed glycerol. Maximum succinate production also decreased as agitation speed increased. These results are in agreement with other research which reported that highly aerobic conditions inhibited succinic acid production (30). Our results therefore suggest that the carbon balance decreased as the conditions became more aerobic as the agitation speed increased, which is due to dispersion of carbon flux as the TCA activity of the oxidation pathway that generates CO2 increased (28).
We obtained the highest isobutanol production of 1.02 ± 0.03 g/L at an agitation rate of 200 rpm, followed by 100 rpm (0.75 ± 0.05 g/L), 300 rpm (0.56 ± 0.04 g/L), and 400 rpm (0.35 ± 0.06 g/L). A likely explanation is that a high agitation rate led to increased carbon flux into metabolites derived from the oxidative pathway, such as succinate and ethanol, and that a low agitation rate led to greater flux of metabolites into anaerobic metabolism and the synthesis of isobutanol. Therefore, we selected 200 rpm as an appropriate agitation speed for isobutanol production (Fig. 4).
Interestingly, we also found that isobutanol production decreased as the agitation speed increased to more than 200 rpm. However, another study reported that a moderate agitation speed (400 rpm) in K. pneumoniae ΔldhAΔbudAΔilvD efficiently produced 2,3-dihydroxyisovalerate (2,3-DHIV), an intermediate in the isobutanol biosynthesis pathway (28). These results may be explained by the need for NADH regeneration to maintain glycolytic flux (30). Although increased production of 2,3-DHIV may be expected increase isobutanol production, we did not observe this effect at 400 rpm.
The results of a previous study that examined valine production from glucose in K. pneumoniae are consistent with our results about the effect of pH on isobutanol production. This previous study, which also used a pH of 6.0 to improve valine production, reported that valine production increased as the agitation speed increased to 600 rpm (31). However, this study also reported that isobutanol production gradually decreased when the agitation speed increased to 400 rpm. In this case, it could be assumed that an agitation speed of 600 rpm increased flux into valine synthesis and decreased flux into isobutanol synthesis (31).
In conclusion, we evaluated the effect culture factors (agitation speed and pH) on isobutanol production and found that a pH of 6.0 and an agitation speed of 200 rpm provided the greatest isobutanol production. However, the fermentation process conditions and our development of the K. pneumoniae ΔldhAΔbudA, pUC-tac-BN-ISO strain only led to a small increase in the amount of isobutanol produced.
Based on the above overall results, blocking the valine biosynthetic pathway may be an effective method to further increase isobutanol production. This additional step could increase the conversion of 2-ketoisovalerate to isobutyraldehyde (a precursor of isobutanol) by suppressing valine production and reducing the valine-mediated feedback inhibition of AHAS. Therefore, we suggest that future studies should investigate the effect of blocking valine synthesis on isobutanol production, and also increase cell growth by adjustment of fermentation process conditions.