Introduction

l-Lysine is one of the eight essential amino acids for humans and animals. With its advantages, l-lysine has been applied in more and more fields, for example, feed additives [1], pharmaceuticals, cosmetics, and polymer materials [2]. With the continual enlarging of application range of l-lysine, the demands of l-lysine are rapidly increasing every year. Therefore, the least expensive fermentation substrates and optimal production strains are needed for fermentation in order to reduce the costs of production. Currently, the industrial l-lysine producers are almost auxotroph, regulatory and combined Corynebacterium glutamicum, or the closely related Brevibacterium flavum mutants, which have been obtained by repeated physical and/or chemical mutagenesis [3]. C. glutamicum can use a relatively broad substrate as carbon and energy sources, which include glucose, sucrose, fructose, pentose, ethanol, and organic acids, etc., because it possesses several different uptake systems to metabolize different carbon sources [4–6]. Glucose uptake in C. glutamicum occurs via a PTSGlc and subsequently phosphorylated to glucose-6-phosphate [6], and 50 % of glucose-6-phosphate pass the pentose phosphate pathway to generate anabolic reducing equivalents [7]. Fructose is simultaneously phosphorylated by a PTSFru and a PTSMan in C. glutamicum to gives fructose-1-phosphate and fructose-6-phosphate, respectively [8]. The PTSFru is the predominant fraction, more than four-fifths of fructose via this mechanism to synthesis fructose-1-phosphate, and then synthesis fructose-1,6-bisphosphate [8]. However, C. glutamicum lacks completely a viable fructose-1,6-bisphosphatase (FBPase) for conversion of the fructose-1,6-bisphosphate into fructose-6-phosphate; 80 % of total fructose consumption enters glycolysis at the level of fructose-1,6-phosphate [7]. On the other hand, fructose-1,6-bisphosphate inhibits the key enzymes of the pentose phosphate pathway glucose-6-phosphate dehydrogenase and 6-phosphogluconate dehydrogenase [5]. Therefore, expression of the FBPase gene fbp, which catalyzes the hydrolysis of fructose-1,6-bisphosphate to fructose-6-phosphate and inorganic phosphate, is proposed to increase the pentose phosphate pathway flux as well as lysine production on fructose and sucrose [4, 5, 9, 10]. In addition, FBPase is necessary for C. glutamicum to grow on the carbon sources acetate, citrate, glutamate, and lactate [4]. C. glutamicum was found to take up and phosphorylate sucrose by a phosphoenolpyruvate-dependent sucrose phosphotransferase system (PTSSuc) to yield sucrose-6-phosphate and subsequently hydrolyzed by an invertase to produce glucose-6-phosphate and fructose [8]. However, C. glutamicum does not possess such a fructokinase (ScrK) activity [11]; further metabolism of the intracellular fructose requires efflux of the fructose and reassimilation via the fructose PTS (PTSFru). In order to phosphorylate directly the fructose without any fructose efflux, heterologous expression of the ScrK gene scrK from Clostridium acetobutylicum (C. acetobutylicum) in C. glutamicum is proposed. On the other hand, ScrK not only directs fructose flux to the pentose phosphate pathway for the supply of NADPH but also modifies significantly the rigidity of phosphoenolpyruvate (PEP) note. Therefore, ScrK expression may be a potential strategy for enhancing l-lysine production during growth on sucrose or molasses.

Molasses, a waste product of the sugar industry, is wildly used for the chemical synthesis of many organic chemicals by microbial fermentation, such as ethanol, amino acid, and glycerin, etc. [12, 13]. Reuse of molasses not only provides a solution to the environmental issue arising from the disposal of molasses but also helps in stabilizing the agricultural sector in the sugar-producing countries [14]. Molasses is a mixture of sugar, nonsugars, and water, from which no sugar crystallizes under any conceivable physical and technically optimum conditions, with no regard to time [13]. The sugar in beet molasses mainly includes sucrose, glucose, fructose, and raffinase, but the main component is sucrose. On the other hand, molasses contains not only large amounts amino acid (glutamic acid, aspartic acid, leucine, etc.) pantothenic acid, inositol, and biotin, but also other growth substances (nicotinic acid, folic acid, thiamine, etc.). In addition, molasses contain a large variety of salts, for example, potassium and manganese. All in all, most components in molasses are advantageous to the growth of microorganisms. Molasses is wildly used for amino acid fermentation by Corynebacterium or Brevibacterium, but most published work focuses on single glucose, or fructose, or sucrose as the carbon source [4, 8, 9, 15]. Although the molasses is used as a carbon source during the evaluation of the molecular targets for l-lysine production, this is only a fraction of the carbon source used in the fermentation medium [16]. As we have seen, the molasses used as the sole carbon source to l-lysine production has not been reported.

Given that the advantages of molasses, such as resource-rich and advantageous to cell growth, the present study was focus on the development of a high-yielding strain for the production of L-lysine with the beet molasses as sole carbon source. Firstly, the strain which heterologous expression the ScrK gene scrK from C. acetobutylicum in C. glutamicum lysC fbr was constructed, and designated as C. glutamicum lysC fbr pDXW-8-scrK; secondly, the strain which heterologous expression FBPase gene fbp from Escherichia coli in C. glutamicum lysC fbr was constructed, and designated as C. glutamicum lysC fbr pDXW-8-fbp; finally, the strain which heterologous expression fbp and scrK in C. glutamicum lysC fbr was constructed, and designated as C. glutamicum lysC fbr pDXW-8-fbp-scrK. The study shows the first time that heterologous expression of scrK and fbp are beneficial to l-lysine production and cell growth during growth on molasses as sole carbon source.

Materials and methods

Strains, growth media, and culturing conditions

The strains, plasmids, and oligonucleotides used in this study are listed in Table 1. DNA was extracted from E. coli and C. acetobutylicum using the TaKaRa MiniBEST Universal Genomic DNA Extraction Kit according to the protocol supplied by the manufacturer (Dalian, China). The E. coli DNA was used as a template to amplify the fbp operon with fbp-F and fbp-R, and the C. acetobutylicum DNA was used to amplify the scrK operon with scrK-F and scrK-R. The plasmid construction and transformation were performed according to the descriptions of Sambrook and Russel [17] and van der Rest et al. [18] (Fig. 1). Firstly, we constructed the strains which are heterologous expression the ScrK gene scrK from C. acetobutylicum or/and FBPase gene fbp from E. coli in C. glutamicum lysC fbr. Plasmids were extracted from E. coli or C. glutamicum using the Tianpure Mini Plasmid Kit according to the protocol supplied by the manufacturer (Tiangen, Beijing, China), and then confirmed by restriction endonuclease reaction (Fig. 2). Luria–Bertani (LB) was used as the standard medium for E. coli BL21. LBG (LB supplemented with 5 g/l glucose) was used for C. glutamicum [19]. E. coli and C. glutamicum were respectively grown at 37 °C and 30 °C. When appropriate, E. coli and C. glutamicum strains were cultured with kanamycin (50 μg/ml) and 120 rpm, and a reduced concentration of kanamycin (25 μg/ml) was used to obtain recombinant strains of C. glutamicum [5].

Table 1 Strains, plasmids, and oligonucleotides used in this study
Full size table
Fig. 1
figure 1

Strategy used for construction of recombinant plasmids pDXW-8-fbp, pDXW-8-scrK, and pDXW-8-fbp-scrK

Full size image
Fig. 2
figure 2

Identification of recombinant plasmids pDXW-8-fbp, pDXW-8-scrK, and pDXW-8-fbp-scrK by enzyme digestion. M DNA marker DL15000 (TAKARA, Dalian, China), Lane 1 pDXW-8 HindIII digested, Lane 2 fbp PCR product, Lane 3 pDXW-8-fbp XhoI digested, Lane 4 pDXW-8-fbp XhoI and NcoI digested, Lane 5 scrK PCR product, Lane 6 pDXW-8-scrK EcoRI digested, Lane 7 pDXW-8-scrK EcoRI and NheI digested, Lane 8 pDXW-8-fbp-scrK EcoRI digested, Lane 9 pDXW-8-fbp-scrK EcoRI and NheI digested

Full size image

Fermentations of molasses

Molasses were supplied by the COFCO Biochemical (Anhui) Co., Ltd., (Anhui, China). They were diluted tenfold in CgXII-media without carbon source and subsequently filtered to remove the suspended components by filter (ϕ 0.45 μm). Diluted molasses were used as fermentation media, and sterilized at 121 °C for 15 min before employed in the fermentations.

For lysine fermentation, a preculture was inoculated from a fresh LBG-plate and cultivated overnight. After washing cells with CgXII-media without a carbon source [5], the main culture with minimal medium was inoculated with OD600 of 1 in 60 ml of fermentation media in 500-ml Erlenmeyer flasks. The cultures were grown at 30 °C and 120 rpm agitation. Heterologous gene expression of recombinant strains was induced by adding isopropyl-β-d-thio-galactoside (IPTG, final concentration 1 mM) at OD600 = 1.6. As a control, molasses was replaced by glucose where the concentration was equivalent to the total sugar concentration of molasses.

Enzyme activity assay

The crude enzyme was prepared according to the methods previously described [4, 19]. After centrifugation at 4 °C for 30 min at 10,000 × g, the cell-free supernatant was immediately used to determine the enzyme activities. Protein concentrations were determined using the Bradford Protein Quantification Kit according to the protocol supplied by the manufacturer (Tiangen, Beijing, China) with bovine serum albumin as standard. The analyses of enzyme activities and protein concentrations were done in triplicate. Specific activity was given as the number of mU/mg of protein. The FBPase assay was based on the protocol of Becker et al. [4]. The fructokinase assay was based on the method of Porter et al. [20].

Analytical methods

Determination of sugar concentration

The sugar concentration was determined according to the descriptions of Periyasamy et al. [21]. The sugar concentrations were determined at initial fermentation and final fermentation.

Determination of glucose, fructose, sucrose, and trehalose concentration

Samples of 200 μl were taken from Erlenmeyer flasks every 4 h. Then, 100 μl of those samples was used to determine the glucose, fructose, sucrose, and trehalose concentration, and the other for measuring the cell growth and l-lysine production. The determinations of glucose, fructose, and trehalose in 1:10-diluted cultivation supernatant were carried out by high-performance liquid chromatography (HPLC, Agilent Technologies, Palo Alto, CA, USA) with a Sugar-Pak 1 6.5 × 300-mm column at 70 °C using isocratic elution with ultrapure H2O at a flow rate of 0.6 ml/min, and subsequent detection via determination of refraction indexes. The determination of sucrose was carried out by HPLC as described previously [4].

Determination of l-lysine, glycerol, and dihydroxyacetone concentration

The l-lysine concentration was determined by an SBA-40E immobilized enzyme biosensor (Shandong, China). The determinations of glycerol and dihydroxyacetone were in 1:10-diluted cultivation supernatant were carried out by HPLC with a Kromasil C-18 4.6 × 250 mm column (AkzoNobel, Amsterdam, The Netherlands) at 30 °C using isocratic elution with 20 mM KH2PO4 and 0.5 % acetonitrile at a flow rate of 0.8 ml/min, and subsequent detection via determination of UV absorption at 210 nm.

Determination of cell growth

One hundred microliters of cultures was used to measure the biomass concentration by a photometer at 600 nm after an appropriate dilution or by gravimetric analysis as described previously [19]. Under these experimental conditions, the correlation factor (g of biomass to OD600 nm) between dry cell weight and OD600 nm was determined as 0.341.

Results

The concentration of total sugar, glucose, fructose, and sucrose

There are differences in components and total sugar concentrations among companies within different counties. In order to have a rough idea of the consumption of sugar, the initial total sugar and residual sugar concentration were detected. The initial total sugar concentration of diluted molasses was 49.57 ± 3.73 g/l (Fig. 3). There were big differences in the residual sugar concentrations of fermentation broth between C. glutamicum lysC fbr and other recombinant C. glutamicum strains (Fig. 3). The residual sugar concentrations of C. glutamicum lysC fbr, C. glutamicum lysC fbr pDXW-8-fbp, C. glutamicum lysC fbr pDXW-8-scrK, and C. glutamicum lysC fbr pDXW-8-fbp-scrK were 6.62 ± 2.26, 3.15 ± 1.53, 2.38 ± 1.04, and 1.71 ± 0.83 g/l, respectively. This phenomenon was also observed by detecting the concentration of glucose, sucrose, and fructose with HPLC in this study (Fig. 4). The concentration of glucose, sucrose, and fructose in diluted molasses was 0.94, 47.33, and 0.79 g/l, respectively. The concept that the sugar in beet molasses consists mainly of sucrose was known to everyone long ago [13]. It should be noted that glucose was firstly consumed, regardless the strains with tested (Fig. 4). The concentrations of sucrose and fructose were changeless during the lag phase of bacterial, regardless the strains with tested. However, the way of sucrose and fructose consumption was different between the tested strains with the passage of fermentation time. Most of the sucrose was consumed at 36 h for C. glutamicum lysC fbr, whereas the concentration of fructose was increased to achieve a peak at 24 h (Fig. 4a). However, for C. glutamicum lysC fbr pDXW-8-fbp, the concentration of sucrose had touched bottom at 44 h (0.55 g/l), whereas the fructose had peaked at 28 h (8.73 g/l). Interestingly, unlike C. glutamicum lysC fbr and C. glutamicum lysC fbr pDXW-8-fbp, the concentration of fructose had been at a low level (0.8 ± 0.05 g/l) for C. glutamicum lysC fbr pDXW-8-scrK and C. glutamicum lysC fbr pDXW-8-fbp-scrK, whereas the sucrose was completely consumed at 36 h (Fig. 4c, d).

Fig. 3
figure 3

Initial total sugar concentration and the residual sugar concentration after fermenting by various strains of C. glutamicum. The gray column indicates the residual sugar concentration and the white column indicates the consumption of sugar concentration

Full size image
Fig. 4
figure 4

Effect of heterogenous expression of fructose-1,6-bisphosphatase gene and fructokinase gene in C. glutamicum on glucose (filled circle), fructose (filled square), and sucrose (filled triangle) catabolism in molasses minimal medium: a C. glutamicum lysC fbr; b C. glutamicum lysC fbr pDXW-8-fbp; c C. glutamicum lysC fbr pDXW-8-scrK; d C. glutamicum lysC fbr pDXW-8-fbp-scrK

Full size image

In vitro activity of FBPase and fructokinase in the genetically defined C. glutamicum strains

FBPase exhibited a low basal in vitro activity in the C. glutamicum lysC fbr, and the activity of FBPase was not affected by additional expression of scrK in C. glutamicum lysC fbr (Table 2). Interestingly, the activity of FBPase from C. glutamicum lysC fbr pDXW-8-fbp was 228 ± 5.46 mU/mg, which was significantly higher than that of C. glutamicum lysC fbr pDXW-8-fbp-scrK (101 ± 4.97 mU/mg). For fructokinase, there was no detectable fructokinase in C. glutamicum lysC fbr and C. glutamicum lysC fbr pDXW-8-fbp. The result is consistent with the precious report [22] because of absence of fructokinase gene in C. glutamicum. For amplification of fructokinase, the recombinant strains C. glutamicum lysC fbr pDXW-8-scrK and C. glutamicum lysC fbr pDXW-8-fbp-scrK showed increased level of fructokinase activity (Table 2). The activity of fructokinase was not affected by additional expression of fructose-1,6-bisphosphatase gene (fbp). Overall, the enzyme data clearly showed that heterologous expression of fbp and scrK in C. glutamicum lysC fbr were successful.

Table 2 In vitro specific activity of FBPase and ScrK in recombinant C. glutamicum strains during growth on molasses
Full size table

Growth performance of genetically defined C. glutamicum strains

To test the influences of the heterologous expressions of fbp and scrK for cell growth during growth on molasses, we analyzed the DCW of these recombinant strains as compared to growth on glucose. There were differences in the DCW and maximal specific growth rate among the recombinant strains during growth on molasses. The DCW of recombinant strains were higher than the starting strain C. glutamicum lysC fbr during growth on molasses. The DCW of C. glutamicum lysC fbr, C. glutamicum lysC fbr pDXW-8-fbp, C. glutamicum lysC fbr pDXW-8-scrK, and C. glutamicum lysC fbr pDXW-8-fbp-scrK were 19.22, 22.13, 22.71, and 20.25 g/l, respectively, during growth on molasses. However, compared with the molasses, the DCW of these strains were different during growth on glucose, 22.3, 21.07, 22.1, and 21.07 g/l, respectively. The maximal specific growth rate was decreased by heterologous expression of fbp or scrK as compared to C. glutamicum lysC fbr. The maximal specific growth rate of C. glutamicum lysC fbr, C. glutamicum lysC fbr pDXW-8-fbp, C. glutamicum lysC fbr pDXW-8-scrK and C. glutamicum lysC fbr pDXW-8-fbp-scrK was 0.295, 0.243, 0.253, and 0.232/h, respectively, during growth on molasses. Interestingly, the maximal specific growth rate of C. glutamicum lysC fbr pDXW-8-scrK was lower than C. glutamicum lysC fbr, whereas the DCW of it was highest among the four tested strains (Figs. 5, 6). It should be noted that heterologous expression of scrK has no effect on cell growth during growth on glucose (Fig. 6). It is generally known that no fructose was synthesized during growth on glucose, so expression of scrK is ineffective.

Fig. 5
figure 5

Effect of heterogenous expression of fructose-1,6-bisphosphatase gene and fructokinase gene in C. glutamicum on DCW and l-lysine production. Symbols are as follows: filled circle DCW of C. glutamicum lysC fbr, filled diamond DCW of C. glutamicum lysC fbr pDXW-8-fbp, filled triangle DCW of C. glutamicum lysC fbr pDXW-8-scrK, filled square DCW of C. glutamicum lysC fbr pDXW-8-fbp-scrK, open circle l-lysine production of C. glutamicum lysC fbr, open diamond l-lysine production of C. glutamicum lysC fbr pDXW-8-fbp, open triangle l-lysine production of C. glutamicum lysC fbr pDXW-8-scrK, open square l-lysine production of C. glutamicum lysC fbr pDXW-8-fbp-scrK. Each value represents mean with standard error of three replicative experiments. The standard errors are shown as bars

Full size image
Fig. 6
figure 6

Comparison of DCW and l-lysine production of recombinant C. glutamicum strains during growth on glucose with growth on molasses minimal medium. The white column indicates the l-lysine production of recombinant strains during growth on glucose and the gray column indicates the l-lysine production of recombinant strains during growth on molasses. Symbols are as follows: filled triangle the DCW of recombinant strains during growth on glucose, filled circle the DCW of recombinant strains during growth on molasses. Each value represents mean with standard error of three replicative experiments. The standard errors are shown as bars

Full size image

l-Lysine production of genetically defined C. glutamicum strains

As compared to growth on glucose, C. glutamicum lysC fbr showed a lower l-lysine production during growth on molasses, 34.5 and 25 mM, respectively. This is due to the main carbon source of molasses, which contains mainly sucrose as carbon source. The additional heterologous expression of fbp and/or scrK in C. glutamicum lysC fbr led to a significant increase in the l-lysine production on molasses (Fig. 5). The highest increase (88.4 %) was observed for C. glutamicum lysC fbr pDXW-fbp-scrK, but the increase was also significant for C. glutamicum lysC fbr pDXW-8-fbp (47.2 %) and C. glutamicum lysC fbr pDXW-8-scrK (36.8 %). Compared with heterologous expression of fbp, heterologous expression of scrK led to little increase l-lysine production on molasses, but the l-lysine production was similar to growth on glucose, 34.5 and 34.2 mM, respectively (Fig. 5). Although the l-lysine production of C. glutamicum lysC fbr pDXW-8-scrK was not higher, the time of reaching the peak was shortest among these strains (reached peak at 36 h). In addition, C. glutamicum lysC fbr pDXW-8-fbp-scrK showed the biggest the maximal specific l-lysine production rate among these strains (data no shown). It is worth pointing out that heterologous expression of fbp in C. glutamicum lysC fbr leads to a significant increase (32.2 %), whereas heterologous expression of scrK in C. glutamicum lysC fbr was invalid for increasing l-lysine production during growth on glucose (Fig. 6). This is, however, not a consequence of an unexpressed scrK. In fact, the expression of scrK was successfully expressed in C. glutamicum (Table 2).

By-product production of genetically defined C. glutamicum strains

Glycerol and dihydroxyacetone were the leading by-products for molasses-grown cells of C. glutamicum lysC fbr, which is also accompanied by little trehalose (Table 3). The comparative analysis for the other recombinant strains revealed that heterologous expression of fbp or/and scrK led to strong decrease in glycerol and dihydroxyacetone yields (Table 3). Interestingly, single expression of scrK in C. glutamicum lysC fbr led to a larger decrease in glycerol and dihydroxyacetone production as compared to single expression of fbp (Table 3). In addition, the concentrations of glycerol and dihydroxyacetone were lowest during co-expression of fbp and scrK, 0.5 ± 0.07 and 1.7 ± 0.09 mM, respectively. However, the trehalose formation was increased by heterologous expression of fbp or/and scrK as compared to the C. glutamicum lysC fbr (Table 3). Among them, heterologous expression of scrK played a major role in the increase of trehalose. These results showed that heterologous expression of fbp and scrK was best for minimizing glycerol and dihydroxyacetone, especially co-expression of fbp and scrK, during growth on molasses.

Table 3 Yield of glycerol, dihydroxyacetone, and trehalose in the stages of l-lysine production by recombinant C. glutamicum strains during growth on molasses
Full size table

Discussion

Molasses is wildly used for product many organic chemicals by microbial fermentation, such as ethanol, amino acid, and glycerin, etc. [12–14], and is beneficial to cell growth [13]. In the present study, we focus on the development of a high-yielding strain for the production of l-lysine with beet molasses as the sole carbon source. Although molasses is wildly used for l-lysine fermentation by Corynebacterium or Brevibacterium [16], the molasses is firstly used as the sole carbon source to l-lysine production. In order to increase the l-lysine production and cell growth during growth on molasses, the heterologous fbp from E. coli and scrK from C. acetobutylicum are firstly expressed in C. glutamicum lysC fbr.

In order to enhance the l-lysine production, it is necessary to increase the consumption of sugar. There are big differences of sugar consumption between C. glutamicum lysC fbr and recombinant C. glutamicum lysC fbr strains (Fig. 3). The variation of sugar consumption may be related to components of molasses and gene expression. Molasses contain multiple sugars, for example, glucose, sucrose, fructose, and raffinase [13]. In addition, the differences of various sugar uptake capacity were found in C. glutamicum [5]. Glucose was preferred by C. glutamicum, and next being sucrose and fructose [5, 23], and this is exactly why glucose is firstly consumed, regardless of the strain tested (Fig. 4). Interestingly, the way of sucrose and fructose consumption was different between the tested strains with the passage of fermentation time (Fig. 4). This may be related to the metabolic methods of fructose and sucrose in C. glutamicum and recombined C. glutamicum strains. Sucrose is ingested by C. glutamicum via PTSSuc to yield intracellular sucrose-6-phosphate and subsequently hydrolyzed to produce glucose-6-phosphate and liberated fructose [8], but in the absence of fructokinase, further metabolism of the liberated fructose requires efflux of the fructose [23]. In this way, the concentration of sucrose is decreased, whereas the concentration of fructose is increased during the initial stage of fermentation. Finally, the fructose in culture was reassimilated by PTSFru so that the concentration of fructose is decreased with the fermentation progress. Heterologous expression of scrK in C. glutamicum lysC fbr and C. glutamicum lysC fbr pDXW-8-fbp not only increases the activity of fructokinase but also prevents the efflux of the fructose (Table 2; Fig. 4c, d). However, the FBPase activity is affected by additional expression of scrK in C. glutamicum lysC fbr pDXW-8-fbp (Table 2). Expression of scrK modifies the transportation of sucrose. The intercellular free fructose derived from sucrose was directly phosphorylated to fructose-6-phosphate and thus without fructose efflux [22]. In this way, the concentration of fructose-6-phosphate is increased, whereas the concentration of fructose-1,6-bisphosphate is deceased during expression of scrK. Hereby, the activity of FBPase from C. glutamicum lysC fbr pDXW-8-fbp-scrK is decreased.

The DCW of recombinant C. glutamicum lysC fbr strains is increased by heterologous expression of fbp or scrK (Fig. 5), however, the specific growth rate is decreased by expression of fbp or scrK. This same phenomenon was found in previous works [4, 22]. This is because heterologous expression of fbp and/or scrK leads to a redirection of carbon from glycolysis toward the PPP and an increased NADPH supply, thus meeting the requirement of NADPH for biomass synthesis [4]. On the other hand, heterologous expression of fbp and scrK lead to a decrease in the flux of TCA (see Figure 1 of Georgi et al., [5]). The scrK-expression strains grow faster than the fbp-expression strains (Fig. 4). This could be because of the difference of characteristics between ScrK and FBPase. The intercellular fructose is catalyzed to synthesize the fructose-6-phosphate by ScrK, thus preventing the effusion of fructose [24]. The synthetic fructose-6-phosphate has two metabolic pathways: (1) entering glycolysis via phosphofructokinase and (2) entering glyconeogenesis via glucose-6-phosphate isomerase, thus utilizing by strains. FBPase catalyzes the hydrolysis of fructose-1,6-bisphosphate to fructose-6-phosphate and inorganic phosphate [25]. The intracellular fructose must be transferred to extracellular, subsequently ingested by PTSFru to synthesize fructose-1,6-bisphosphate. Therefore, heterologous expression of fbp should not directly utilize intracellular fructose by strains. Interestingly, Moon et al. [22] reported that growth of the C. glutamicum ATCC13032 harboring the fructokinase gene was noticeably retarded during the initial stage of growth, whereas this phenomenon does not appear in our experiment. This may be due to the difference of starting strains or even media. As compared to growth on glucose, the DCW of C. glutamicum lysC fbr pDXW-8-fbp and C. glutamicum lysC fbr pDXW-8-scrK were higher during growth on molasses (Fig. 6). Apart from the fact that the growth substances are abundant in molasses [13], heterologous expression of fbp and scrK plays a major role in the increase of biomass. Interestingly, co-expression of fbp and scrK leads to a small decrease in biomass (Figs. 5, 6). This may be related to the fact that heterologous co-expression of fbp and scrK influence the intracellular distribution of the carbon source, with a shift from biomass towards amino acid production [4, 22].

Apart from the benefit for cell growth, Becker et al. [4] and Georgi et al. [5] have reported that overexpression of fbp in C. glutamicum lysC fbr led to a strong increase in l-lysine production during growth on glucose, fructose, and sucrose. In addition, Becker et al. [4] thinks that overexpression of FBPase was beneficial to l-lysine production by C. glutamicum with starch or molasses as the carbon source, and Georgi et al. [5] have also pointed out that FBPase was a limiting factor for l-lysine production by C. glutamicum with sucrose as the carbon source. However, the FBPase from C. glutamicum is negatively regulated at the metabolic level by phosphoenolpyruvate or AMP [26] and inhibited by fructose 2,6-bisphosphate [27]. Fortunately, the FBPase from E. coli is not inactivated by catabolite and is insensitive to fructose-2,6-bisphosphate ([28]; Jianzhong Xu et al. unpublished results). Therefore, expression of E. coli fbp is beneficial in increasing l-lysine production and decreasing by-products (Figs. 5, 6; Table 3). It is worth pointing out that the improvements of l-lysine production by C. glutamicum lysC fbr with heterologous fbp expression are superior to the previous results, which are obtained via homologous fbp expression in C. glutamicum lysC T311I during growth on fructose and sucrose [4]. It may be that E. coli FBPase has removed the inactivation or inhibitions of fructose-2,6-diphosphate, AMP and catabolite, thus catalyzing effectively the formation of fructose-6-phosphate from fructose-2,6-diphosphate. Expression of the scrK in C. glutamicum can directly phosphorylate intracellular fructose into fructose-6-phosphate instead of fructose-1-phosphate, thus changing the route of fructose entry into metabolism [22]. In addition, fructokinase uses ATP as a phosphoryl donor rather than phosphoenolpyruvate. Therefore, fructokinase expression would seem to be a potential strategy for enhancing amino acid production. As compared with heterologous expression of fbp, however, heterologous expression of scrK leads to a smaller increase in l-lysine production (Fig. 5). We find the DCW was increased more by heterologous expression of scrK as compared with expression of fbp. So, we deduce that heterologous expression of scrK led to more C-source enter to synthesize biomass than to produce the l-lysine.

The differences in carbon-metabolite accumulation are indicative of a different pattern of carbon-flux distribution through the central metabolic pathways [7]. Heterologous expression of fbp and scrK changes the carbon-metabolite accumulation as compared to staring strains C. glutamicum lysC fbr [3, 4, 22]. Therefore, the l-lysine production is strongly increased, whereas the by-products are greatly decreased by expression of fbp and scrK ([4] and Table 3). However, fbp and/or scrK expression lead to a slight increase in trehalose. Glucose-6-phosphate is used as a precursor to synthesis trehalose [15], which is increased by expression of fbp and scrK. Therefore, reducing the concentration of glucose-6-phosphate is beneficial to reducing the concentration of trehalose. Enhancing the PPP flux may be a promising method for reducing the concentration of glucose-6-phosphate. In this regard, expression of the glucose-6-phosphate dehydrogenase gene zwf or introduction of the modified variants of 6-phosphogluconate dehydrogenase or glucose-6-phosphate dehydrogenase has been suggested [15, 29]. In addition, enhancing the PPP flux is also of critical importance for l-lysine production [30].

In conclusion, we have studied the development of a high-yielding strain for the production of l-lysine, and demonstrated that heterologous expression of fbp and scrK are beneficial to increasing the biomass and l-lysine production with beet molasses as the sole carbon source. On the basis of these and previous results [4, 15, 22], we regard that strategies which lead to an increased PPP flux could increase l-lysine production or other NADPH-dependent compounds. In addition, expression of C. acetobutylicum scrK can effectively stop intracellular fructose effusion and enhance the l-lysine production during growth on molasses.