Metabolic Engineering of Bacillus Amyloliquefaciens as A Novel Cell Factory to Produce Spermidine


 Background: Spermidine is a biologically active polyamine with extensive application potential in foods and pharmaceuticals. However, previously reported spermidine titers by biosynthesis methods are relatively low, which hinders the industrial fermentation of spermidine. To improve the spermidine titer, key genes affecting the spermidine production were mined to engineer the Bacillus amyloliquefaciens.Results: Genes of S-adenosylmethionine decarboxylase (speD) and spermidine synthase (speE) from different microorganisms were expressed and compared in B. amyloliquefaciens. Therein, the speD from Escherichia coli and speE from Saccharomyces cerevisiae were confirmed to be optimal for spermidine synthesis, respectively. Then, these two genes were co-expressed to generate an engineering strain B. amyloliquefaciens HSAM2(PDspeD-SspeE) with a spermidine titer of 91.31 mg/L, improving by 10.90-fold compared with the control (HSAM2). Through further optimization of fermentation medium, the spermidine titer was increased to 227.35 mg/L, which was the highest titer among present reports. Moreover, the consumption of the substrate S-adenosylmethionine was consistent with the accumulation of spermidine, which contributed to understanding the synthetic pattern of spermidine. Conclusions: Two critical genes for spermidine synthesis were obtained, and an B. amyloliquefaciens cell factory was constructed for enhanced spermidine production, which laid the foundation for further industrial production of spermidine.


Background
Spermidine is a multifunctional polyamine with the positive charge and low molecular weight [1,2].
Therefore, spermidine has promising application potential in the elds of foods and pharmaceuticals, and development of spermidine-related products was extremely valuable.
At present, spermidine is mainly produced by chemical synthesis, while it has several disadvantages such as high energy consumption, mass toxic by-product, and heavy environmental pollution. In comparison, the microbial fermentation method is safe, e cient and environmentally friendly [13]. Currently, many microorganisms have been reported to synthesize trace spermidine [14][15][16]. The core metabolic pathway for spermidine synthesis is shown in Fig. 1. Putrescine and S-adenosylmethionine (SAM) are the precursors for spermidine formation. Firstly, the SAM is catalyzed to decarboxylated SAM (dcSAM) by SAM decarboxylase encoded by speD gene. Then, the aminopropyl group of the dcSAM is transfered to putrescine to produce the spermidine, which is mediated by the spermidine synthase encoded by speE gene [17][18][19]. In Saccharomyces cerevisiae, the spermidine production was increased by over-expressing genes of ornithine decarboxylase, SAM decarboxylase and spermidine synthase and deleting genes of anti-ornithine decarboxylase and polyamine transporter [20,21]. In Synechocystis sp, overexpression of arginine decarboxylase genes Adc1 and Adc2 could enhance the intracellular spermidine content [22]. However, previously reported fermentation titers of spermidine are relatively low, and more work is needed to improve the spermidine production, such as explaining the metabolic mechanism, mining key genes and engineering the strain.
B. amyloliquefaciens has been an e cient platform workhorse for production of various bioproducts [23][24][25]. The complete spermidine synthesis pathway exists in B. amyloliquefaciens. Moreover, the precursors of spermidine including putrescine and SAM have been highly produced in B. amyloliquefaciens after metabolic engineering [26,27]. Therefore, the B. amyloliquefaciens has the potential to be developed as an e cient spermidine cell factory. However, the genes responsible for spermidine synthesis from putrescine and SAM have not been investigated in B. amyloliquefaciens. Herein, the downstream genes of spermidine synthesis including SAM decarboxylase gene (speD) and spermidine synthase gene (speE) were selected from different microorganisms, and their effects on spermidine production were evaluated in B. amyloliquefaciens. Two e cient genes affecting the spermidine production were obtained, and the spermidine production was enhanced dramatically by combined expression and fermentation optimization.

Results And Discussion
Effects of different SAM decarboxylase genes on spermidine production SAM can be catalyzed to decarboxylated SAM (dcSAM) by SAM decarboxylase, which is a key step for spermidine synthesis. Therefore, e cient SAM decarboxylase genes are believed to enhance the spermidine production. Previously, we constructed a high SAM-producing strain B. amyloliquefaciens HSAM2 [26], which might provide abundant SAM substrate for spermidine synthesis. Therefore, the HSAM2 was used as the host strain to express SAM decarboxylase genes (speD) from different microorganisms. The speD genes from B. amyloliquefaciens HZ-12, S. cerevisiae CICC 31001 and E. coli DH5α were selected to construct recombinant expression strains, named HSAM2(PHspeD), HSAM2(PSspeD) and HSAM2(PDspeD) respectively. After expressing these genes, the spermidine titers were improved signi cantly in all recombinant strains (Fig.2a). Among them, the maximum spermidine titer of 91.91 mg/L was obtained in HSAM2(PDspeD), increasing by 8.49-fold compared with the control strain HSAM2(pHY300PLK). These results indicated that the speD gene from E. coli DH5α was the optimal gene for enhanced spermidine production.
Subsequently, the speD gene of E. coli was integrated into the genome of HASM2 by homologous recombination, resulting in the integrated expression strain HSAM2 DspeD. After fermentation for 60 h, the spermidine titer reached 33.87 mg/L, which was 2.40-fold higher than that of the control strain HSAM2 (Fig.2b). In comparison, the spermidine titer of the integrated expression strain HSAM2 DspeD was much lower than that of plasmid-based expression strain HSAM2(PDspeD). This phenomenon was probably due to the low gene copy number during integration expression, while the pHY300PLK plasmid had the high copy number [28]. Therefore, recombinant plasmid expression was more suitable for the speD gene.

Effects of different spermidine synthase genes on spermidine production
Spermidine synthase catalyzes the transfer of the aminopropyl group from dcSAM to putrescine to form the spermidine [29], which is considered to be a rate-limiting step in biosynthesis of spermidine [30]. It is particularly essential to exploit e cient spermidine synthase genes (speE). Therefore, different speE genes were evaluated. According to the KEGG database, three spermidine synthase genes from E. coli DH5α, C. glutamicum ATCC13032 and S. cerevisiae CICC31001 were selected and expressed in B. amyloliquefaciens HZMD, resulting in recombinant strains HZMD(PDspeE), HZMD(PGspeE), and HZMD(PSspeE), respectively.
As shown in Fig.3, expressing speE genes from E. coli and S. cerevisiae signi cantly improved the spermidine production, while the gene from C. glutamicum showed no signi cant difference. Therein, the maximum spermidine titer reached 49.58 mg/L in HZMD(PSspeE), with a 23% increase than that of the control strain HZMD(pHY300PLK). Previously, overexpression of the native speE gene was con rmed to be e cient for spermidine synthesis in S. cerevisiae [31]. Herein, our results demonstrated that the speE gene from S. cerevisiae also enhanced the spermidine production in B. amyloliquefaciens. It can be inferred that overexpression of this speE gene presumably increased the enzymatic activity of spermidine synthase to promote the spermidine synthesis.
Effect of co-expressing speD and speE genes on spermidine production Above results showed that genes of speD from E. coli DH5α and speE from S. cerevisiae CICC31001 were e cient to enhance the spermidine production. Therefore, these two genes were ligated into one pHY300PLK plasmid, co-expressed in HSAM2 to generate a recombinant strain HSAM2(PDspeD-SspeE). Then, the control strain HSAM2(pHY300PLK), single gene expression strain HSAM2(PDspeD), and coexpression strain HSAM2(PDspeD-SspeE) were compared after fermentation for 60 h. As shown in Figure  4, the spermidine titer of HSAM2(PDspeD-SspeE) reached 105.24 mg/L at 60 h, further improving by 15% compared with the HSAM2(PDspeD). It indicated that co-expression of speD and speE was effective to increase the spermidine titer, which was probably due to that more upstream substrates of SAM and putrescine were converted to form spermidine.

Optimize fermentation medium
To further improve the spermidine titer of the engineered HSAM2(PDspeD-SspeE), the key components of fermentation medium were optimized, including carbon sources, nitrogen sources and antibiotics. Carbon sources were important for cell growth and metabolites synthesis [32,33]. Firstly, effects of carbon sources types on spermidine production were investigated. As shown in Fig. 5a, the maximum titer of spermidine was obtained using xylose as the carbon source. Furthermore, the concentration of xylose was optimized (Fig.5b). The maximum titer of spermidine reached 151.79 mg/L when the xylose concentration was 40 g/L, and no signi cant increase was observed at 60 g/L of xylose.
Corn pulp was a nutrition-rich nitrogen source, and effects of different corn pulp concentrations on spermidine production were investigated. As was indicated in Fig. 5c, the corn pulp concentration signi cantly affected the spermidine titer, and the maximum spermidine titer was obtained at 20 g/L. Several previous studies investigated the impact of antibiotics on the fermentation process [34][35][36].
Herein, different concentrations of tetracycline were added into medium. The increased tetracycline concentration resulted in the improved spermidine titer, and no further increase was noted when the tetracycline concentration reached 4 mg/mL (Fig.5d). These results indicated that adding tetracycline could improve spermidine production, which was probably due to that more plasmids could be maintained under the tetracycline stress.
The spermidine production process under the optimized fermentation midium Under the optimized fermentation medium (40 g/L xylose, 10 g/L peptone, 20 g/L corn pulp, 2 g/L urea, 6.3 g/L (NH4) 2 SO 4 , 2.5 g/L NaCl, 3 g/L KH 2 PO 4 , and 4.2 g/L MgSO 4 ·7H 2 O), the spermidine production process was investigated (Fig. 6). The spermidine was synthesized as the cell grew at the early stage of fermentation. The cell growth entered into the stationary phase at about 48 h, while the spermidine was further synthesized until 84 h, indicating that the spermidine synthesis was a partly growth-coupled process. At 84 h, the maximum spermidine titer reached 227.35 mg/L, which was currently the highest titer reported by microbial fermentation. SAM was the key precursor for spermidine synthesis [5,37], and the SAM concentration was also measured to further understand the fermentation process. In the initial 24 h stage, the SAM was accumulated rapidly to reach the highest point, while the spermidine concentration did not have obvious improvement. When the fermentation time exceeded 24 h, the spermidine production showed a rapid increase accompanied by a sharp drop in the SAM concentration, indicating that the SAM was probably consumed to synthesize spermidine.

Conclusions
This study explored the possibility of synthesizing spermidine via engineering the B. amyloliquefaciens. Effects of different speD and speE genes on spermidine production were investigated in B. amyloliquefaciens. Two e cient genes of speD and speE were mined to enhance the spermidine production. Subsequently, these two genes were co-expressed to construct an engineering strain HSAM2(PDspeD-SspeE) with high spermidine production. After further optimization of the key medium components, the maximum spermidine titer of 227.35 mg/L was obtained, which was the highest titer among the microbial fermentation methods. Moreover, the detected SAM consumption also explained the accumulation of spermidine. This study excavated the key genes for spermidine synthesis and successfully acquired the highest spermidine titer, which provided the reference for breeding the industrial strain for spermidine production in the future. Table 1 listed all the strains and plasmids constructed in present study. All engineering B. amyloliquefaciens strains were modi ed from the wild-type strain B. amyloliquefaciens HZ-12 (M 2015234). Escherichia coli DH5α was used as the platform strain to construct expression and integration vectors based on plasmids of pHY300PLK and T2(2)-ori [38]. All designed primers in this study were showed in supplementary material (Table S1). Cells were picked up from LB solid plates (10 g/L tryptone, 5 g/L yeast extract, 10 g/L NaCl and 15 g/L agar), and transferred into LB liquid medium to culture for 12 h at 37°C and 180 rpm to obtain seed cultures. Then, the inoculum was pipetted into the fermented medium. The initial fermentation medium contained 40 g/L sucrose, 3 g/L aspartate, 10 g/L peptone, 5 g/L corn pulp, 2 g/L urea, 6.3 g/L (NH 4 ) 2 SO 4 , 2.5 g/L NaCl, 3 g/L KH 2 PO 4 , 4.2 g/L MgSO 4 ·7H 2 O, and pH 6.5. The inoculum size was 3% (v/v) and the tetracycline concentration was 8 µg/mL. The optimized fermentation medium was consisted of 40 g/L xylose, 10 g/L peptone, 20 g/L corn pulp, 2 g/L urea, 6.3 g/L (NH4) 2 SO 4 , 2.5 g/L NaCl, 3 g/L KH 2 PO 4 , and 4.2 g/L MgSO 4 ·7H 2 O, 4 µg/mL tetracycline, and pH 6.5. Fermentation was carried out for 60 h at 37°C with shaking at 180 rpm.

Recombinant plasmid expression
Recombinant plasmid expression was carried out based on the procedures reported previously [39,40]. The speD gene of B. amyloliquefaciens HZ-12 (named as BspeD) was ampli ed using primers BspeD-F and BspeD-R. By Splicing with Overlapping Extension PCR (SOE-PCR), the BspeD fragment was ligated with the P43 promoter from Bacillus subtilis 168 and the TamyL terminator from Bacillus licheniformis WX-02 to obtain gene expression module. Then, this module was inserted into the Bacillus-E. coli shuttle plasmid (pHY300PLK) at restriction sites of BamHI/XbaI to obtain the expression vector PBspeD1, which was subsequently electro-transformed into B. amyloliquefaciens HSAM2 competent cells, resulting in the recombination strain named as HSAM2(PBspeD). Other genes were expressed in pHY300PLK plasmids following the same method.
Homologous recombination T2(2)-ori mediated homologous recombination was employed in gene integration expression [39,40]. The upstream and downstream homology arms were ampli ed using primers A-F/A-R and B-F/B-R, respectively, which were further fused with the expression module of EspeD gene by SOE-PCR. Then, the fused fragment was ligated into the T2(2)-ori plasmid at BamHI/XbaI, resulting in the integrated vector T2(2)-EspeD. Subsequently, this plasmid was electro-transformed into B. amyloliquefaciens HSAM2 competent cells, which were cultured in LB plates containing 20 μg/mL of kanamycin. After veri cation by PCR, positive clones were inoculated into kanamycin-containing LB liquid medium (20 μg/mL), and cultured at 45°C for 8 h at 180 rpm. Single-crossover strains were selected by kanamycin resistance screening and PCR identi cation, subcultured in LB liquid medium for several times at 37°C (8h). The nal cultures were diluted and incubated on LB plates to obtain individual colonies, which were transferred into LB and kanamycin-containing LB plates to screen kanamycin-sensitive colonies. The double-crossover strain was obtained by PCR veri cation.  [41].

Determination of SAM
The SAM was detected by previously reported HPLC method [26] . For sample pretreatment, 1.5 mL of 0.4 M HClO 4 was added into 500 μL fermentation broth to extract the total SAM for 1 h, vortexing for 10 s every 15 min. After centrifugation for 10 min at 12,000×g, 800 μL supernatant was collected and mixed with 95 μL 2 M NaOH and 15 μL saturated NaHCO 3 . Then, the mixture was centrifuged for 5 min at Declarations Ethics approval and consent to participate Not applicable.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used during the current study are available from the corresponding author on reasonable request. Figure 1 The  The effect of expressing different speE genes on spermidine production. Asterisks show the signifcant diference (p < 0.05) compared with the control. Time pro les of spermidine production in HSAM2(PDspeD-SspeE).

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.