All the strains and plasmids used in this work are listed in Table 1. Luria-Bertani (LB) medium (10 g/L tryptone, 5 g/L yeast extract, 10 g/L NaCl, pH = 7.0) was used for the growth of Escherichia coli and the construction of B. subtilis strains under a shaking condition at 200 rpm and 37°C, during which appropriate antibiotics (ampicillin 100 μg/mL, kanamycin 50 μg/mL for E. coli and 25 μg/mL for B. subtilis) was added into the medium for strains with antibiotic resistance.

To prepare seed cultures, B. subtilis 168 and its derivative strains were cultivated in LB medium at 200 rpm and 37°C for about 9 h, after which OD₆₀₀ reached 5.5–6.0. To produce fengycin and detect various indexes during the fermentation, the seed cultures were inoculated into the fermentation medium (glucose/xylose 20 g/L, peptone 20 g/L, KH₂PO₄ 2 g/L, K₂HPO₄ 4 g/L, KCl 0.5 g/L, MgSO₄·7H₂O 0.5 g/L, MnSO₄ 5.0 mg/L, CuSO₄ 0.16 mg/L, FeSO₄ 0.15 mg/L, pH = 7.0–7.2) with the initial OD₆₀₀ of 0.4 and the fermentation broth was shaken at 180 rpm and 30°C.

To measure the efficiency of xylose utilization, cultivation was performed in M9 medium with xylose as the sole carbon source (xylose 20 g/L, Na₂HPO₄ 6.8 g/L, KH₂PO₄ 3 g/L, NH₄Cl 1 g/L, NaCl 0.5 g/L, tryptophan 5 g/L, MgSO₄·7H₂O 1 mol/L, CaCl₂ 0.1 mol/L, MnCl₄·4H₂O 1 mg/L, CuCl₂·2H₂O 0.43 mg/L, ZnCl₂ 1.7 mg/L, CoCl₂·2H₂O 0.6 mg/L, NaMoO₄ 0.6 mg/L, FeCl₃·6H₂O 13.5 mg/L) at 180 rpm and 30°C with the initial OD₆₀₀ of 0.4.

Oligonucleotide primers mentioned in this study are listed in Table 2. The general isolation and manipulation of recombinant DNA followed standard techniques (Sambrook, 2001). The plasmids used for gene deletion and integrated overexpression in this work were all derived from the CRISPR-Cas9 plasmid of B. subtilis named PJOE8999 (Altenbuchner and Müller, 2016). The gRNAs were designed in the website¹ and inserted into PJOE8999 to give a series of plasmids named PJOE-sgRNA as described previously (Tan et al., 2019). To obtain the plasmids for knockout, the upstream and downstream fragments were connected to PJOE-sgRNA plasmids by seamless cloning kit. While to construct the plasmids used for integrated overexpression, upstream regions, promoter regions, target genes as well as downstream regions were amplified from the corresponding templates, and then fragments were inserted into corresponding PJOE-sgRNA plasmids by seamless cloning kit or ligase as required. Unless otherwise specified, the genes were amplified from B. subtilis 168. After introducing the correctly constructed plasmids into B. subtilis strains, the selection of engineering strains was carried out as previous description (Altenbuchner and Müller, 2016; Tan et al., 2019). The correct mutants were confirmed by PCR and DNA sequencing.

To determine the concentration of fengycin produced by the B. subtilis 168 and its derivative strains constructed in this study, the cultures were centrifuged at 12,000 rpm and 4°C for 30 min, and the cell-free supernatant was collected and adjusted to pH 2.0 with hydrochloric acid and kept it at 4°C for 6–8 h. Then the precipitate was collected and dissolved with methanol. After adjusting the mixture to a final pH of 7.0, the samples were kept at 4°C for 12–24 h for fengycin extraction. Finally, the supernatant was collected and filtered through a 0.22 μm filter, which can be quantified by high performance liquid chromatography (HPLC) 1200 series equipped with an Agilent Eclipse XDB-C18 column (5 μm, 4.6 × 150 mm, United States) at an operating temperature of 30°C (Tan et al., 2021; Jin et al., 2022). The mobile phase consisting of acetonitrile, water and trifluoroacetate (50:50:0.1, v/v/v) was used at a flow rate of 0.8 mL/min, and the detection wavelength was 210 nm (Li and Wen, 2023).

In this study, glucose concentrations in culture broth were determined enzymatically by a bioanalyzer (SBA-40C, Shandong, China). To measure the concentration of xylose, lactate as well as acetate of the fermentation broth, HPLC (Agilent, USA) equipped with the HPX-87H column (300 mm × 7.8 mm, Cosmosil, United States) and the differential refractive index detector (SFD GmbH, Schambeck, Germany) was used with a mobile phase of 5 mM H₂SO₄ at a flow rate of 0.6 mL/min. The column temperature was 30°C.

In order to compare the fluorescence intensity of green fluorescent protein (GFP) under different promoter expression, samples were collected from the fermentation broth at different times and the precipitate was collected and resuspended in 0.9% (w/v) NaCl solution after washing three times. The fluorescence intensity of GFP was measured at an excitation wavelength of 480 nm and an emission wavelength of 527 nm. Samples were assayed in triplicate.

Due to the frameshift mutation in sfp and the single base substitutional mutations at the −10 position of degQ promoter, B. subtilis 168 lost its ability for fengycin production (Tsuge et al., 2007), which is a major problem that must be solved first. Therefore, the strain BSUY00 was constructed by the mean of replacing mutated sfp gene of B. subtilis 168 with normally-worked sfp from Bacillus amyloliquefaciens FZB42 and degQ from B. subtilis 168 under the control of P43 promoter. As shown in Figure 2A, both B. subtilis 168 and BSUY00 grew well in the fermentation medium, indicating that the expression of sfp and degQ, as well as the synthesis of fengycin, did not have a negative effect on the growth of B. subtilis. Surprisingly, the biomass of BSUY00 was significantly higher than that of B. subtilis 168 during the whole fermentation process. In addition, the strain BSUY00 could consume approximately 20.33 g/L glucose after 24 h of fermentation while the strain B. subtilis 168 required 32 h to consume all of the glucose. And we speculated that the earlier glucose consumption of strain BSUY00 might be due to the significant increase in biomass.

The results of the fengycin fermentation are shown in Figure 2B. As we suspected, the strain B. subtilis 168 was unable to accumulate fengycin throughout the fermentation process. In the strain BSUY00, fengycin accumulation was began at the mid-exponential phase (about 16 h) and the titer of fengycin was increased rapidly in the mid-exponential and stationary phases. After 48 h of fermentation, the titer of fengycin in BSUY00 reached 71.21 mg/L. The fermentation results mean that the expression of sfp and degQ enabled the BSUY00 strain to successfully gain the ability to synthesize fengycin using glucose as a substrate. However, the titer of fengycin of BSUY00 strain was extremely low and it is necessary to carry out further research based on this strain.

By examining the pH of the fermentation broth of the two strains at different times (Figure 2B), it was found that the pH of the fermentation broth of the two strains decreased significantly during both the delayed and pre-exponential phases (0–8 h), which indicated that the strains started to synthesize organic acids with the glucose uptake, and the phosphate buffer system in the fermentation medium was not sufficient for rapid neutralization of the produced organic acids. While during the phase when fengycin began to accumulate rapidly, the pH of the fermentation broth of the strain BSUY00 was higher than that of the fermentation broth of the strain B. subtilis 168. By determining the acetate and lactate contents of fermentation broth of B. subtilis 168 and BSUY00, it was found that the strain BSUY00 accumulated 0.91 g/L lactate and 0.67 g/L acetate after 16 h of fermentation, which were 24.79% and 47.66% lower than those of B. subtilis 168 (1.21 g/L lactate and 1.28 g/L acetate), respectively (Figure 2C). Based on this result, it was speculated that the synthesis of fengycin led to a reduction in the synthesis of organic acids and a redirection of intracellular metabolic fluxes in the BSUY00 strain, which increased the metabolic flux of intracellular precursors for fengycin synthesis as well as made the pH environment in the culture medium more suitable for the growth of B. subtilis.

It has been proved that pH of fermentation medium plays an important role in cycle lipopeptides production (Dang et al., 2019; Gao et al., 2022). Thus, combined with aforementioned findings, ackA, which encodes acetate kinase, and ldh, which encodes lactate dehydrogenase, were subsequently knocked out in the BSUY00 strain, respectively. Contrary to our expectations, the knockout of ackA and ldh resulted in a 49.84% and 25.95% decrease in fengycin production, respectively (Figure 2D). It was speculated that although the production of acetate and lactate may compete with fengycin production for precursors and energy, while affect the pH of the fermentation system, the knockout of ackA and ldh has a negative effect on fengycin production, possibly because it dramatically disrupts the metabolic flux balance in the cell. In particular, it has previously been shown that the lactate dehydrogenase encoded by the ldh gene oxidizes NADPH to NADP⁺, and the knockout of ldh resulted in the overaccumulation of NADPH in the cell, which led to a severe inhibition on the growth of B. subtilis (Romero et al., 2007). Subsequently, the addition of 10 g/L CaCO₃ was attempted in this study to adjust the pH of the fermentation system, however, no significant changes in fengycin titer were detected, which could be attributed to the regulation of three two-component signal-transduction systems ResD/ResE, PhoP/PhoR and DegU/DegS caused by released Ca²⁺ (Chen et al., 2022).

As the most direct and effective way to regulate gene expression levels, promoter optimization has been widely used in the regulation of carbon flux distribution and the expression of exogenous proteins. Hence, in order to improve the transcription level of pps cluster and to achieve the purpose of increasing the production of fengycin, four promoters (including P43 promoter, promoter of veg, promoter of spoVG and promoter of lytR) were selected and their expression strength was analyzed through fluorescence intensity of GFP reporter system, which gave the strains named BSUY00-P43-gfp, BSUY00-P_veg-gfp, BSUY00-P_spoVG-gfp and BSUY00-P_lytR-gfp, respectively. As shown in Figures 3A,B, although the growth of BSUY00-P_veg-gfp was slightly lower than that of the other strains, which probably due to a metabolic stress on the strain caused by the high expression of the GFP fluorescent protein, it showed the highest fluorescence intensity throughout the fermentation process. And the maximum relative fluorescence intensity of BSUY00-P_veg-gfp was about 3.08 times, 8.19 times and 13.25 times that of BSUY00-P43-gfp, BSUY00-P_spoVG-gfp and BSUY00-P_lytR-gfp, respectively (Figure 3C). The results demonstrate that the four promoters selected in this study exhibit different strength of gene expression, which can be applied for the subsequent replacement of the original promoters in the BSUY00 strain to regulate the expression level of the target genes and modify its metabolic pathway. Subsequently, fengycin production of different strains constructed by replacement of native P_ppsA promoter with different promoters, which were named BSUY04, BSUY04-1, BSUY04-2 and BSUY04-3, respectively, were compared with the fengycin titer of BSUY00 strain as a control (Figure 3D). The replacement of native weak P_ppsA promoter with all four promoters selected in this study improved fengycin titer, in which the strain BSUY04-1 constructed with P_veg showed the highest fengycin production, reaching 371.63 mg/L, which was approximately 5.2 times that of BSUY00.

In B. subtilis, xylose transport is in the charge of the AraE protein (Krispin and Allmansberger, 1998). Once xylose is successfully taken up into cells, xylose is first isomerised to xylulose by xylulose kinase encoded by xylA, and then xylulose is subsequently phosphorylated to form xylulose-5-phosphate by xylulokinase encoded by xylB, after which the xylose metabolic pathway is channeled to glycolysis via the pentose phosphate pathway (PPP). However, the expression of AraE was negative regulated by the protein encoded by araR (Park et al., 2012). And the low efficiency of xylose transport in B. subtilis is the problem needed to be addressed first.

To visualize the effect of the knockout of the araR, the BSUY04-1 and BSUY05 strains were cultured in M9 medium with xylose as the sole carbon source and the results were shown in Figure 4A. As we suspected, the strain BSUY04-1 grew very slowly, with growth rate of only 0.04 h⁻¹, while the BSUY05 strain can grow rapidly at a growth rate of 0.14 h⁻¹. After 48 h of incubation, the BSUY04-1 strain consumed only 2.11 g/L xylose with the consumption rate of 0.04 g/L/h and the BSUY05 strain could consume about 12.71 g/L xylose with xylose consumption rate of 0.26 g/L/h, which is 5.5-fold higher than that of BSUY04-1. These results show that knockout of the araR gene can significantly improve the xylose uptake capacity.

To enhance the expression of the xylose transport protein AraE, and thus further improve the efficiency of xylose transportation into the cell to participate in metabolic process, the araE gene was intensively constitutive expressed in subsequent studies, and the engineered strain BSUY06 was cultured in M9 medium to visualize the effect with the BSUY05 strain as a control. As shown in Figure 4B, the growth rate of BSUY06 was not significantly different from that of BSUY05 during 48 h of incubation period. However, the xylose consumption rate of BSUY06 reached 0.29 g/L/h, which was 11.54% higher than that of BSUY05. These results suggested that the enhancement of araE gene expression on the basis of the knockdown of araR gene still promotes an increase in xylose uptake rate.

Through the previous studies, we had successfully obtained a B. subtilis strains BSUY06, which can efficiently utilize xylose. To give the fermentation characteristics for fengycin production, the strain BSUY06 was cultured in fermentation medium for 64 h and the results were shown in Figure 5A. Under the fermentation conditions, strain BSUY06 was able to consume approximately 20.41 g/L xylose within 40 h, with a xylose consumption rate of 0.51 g/L/h. No fengycin accumulation was observed during both the delayed and pre-exponential phases (0–8 h), and fengycin production in BSUY06 strain was began after about 16 h of incubation, with a final titer of 376.58 mg/L. The aforementioned results suggested that the strain BSUY06 displays a formidable capability to product fengycin with xylose as the substrate.

To ensure a sufficient carbon source and avoid the inhibition of cell growth and waste of residual xylose caused by high concentration substrate, a series fermentation of the BSUY06 strain was performed under different initial xylose concentrations (10, 20, 30, 40, 50, 60 and 70 g/L xylose) in this study. As shown in Figure 5B, at the end of the fermentation, the biomass of BSUY06 strain was increased from 3.06 g/L to 5.23 g/L when the initial xylose concentration was increased from 10 g/L to 40 g/L. In contrast, the decrease in biomass could be observed when the initial xylose concentration continued to increase to 70 g/L. This result indicated that the BSUY06 strain demonstrated favorable growth under low concentration of xylose, while high initial xylose concentration (50–70 g/L) can considerably impede cell growth due to substrate inhibition and increased osmotic pressure. Notably, the fengycin titer increased from 344.35 mg/L to 430.86 mg/L when the initial xylose concentration was increased from 10 g/L to 40 g/L while the opposite trend could be observed when the initial xylose concentration continued to increase to 70 g/L, which showed a similar trend with the biomass. In addition, when the initial concentration of xylose reached 50 g/L, a residual xylose concentration of 6.93 g/L was detected in the broth at the end of fermentation, and as the initial xylose concentration continued to increase to 70 g/L, the level of residual xylose rose gradually from 6.93 g/L to 30.53 g/L, leading to significant wastage of xylose. Based on the aforementioned results, it can be concluded that the most optimal initial concentration of xylose for fengycin fermentation is 40 g/L, as this concentration resulted in the highest biomass and fengycin titer, with no residual waste of xylose at the end of the fermentation.