Establishment of co-production fermentation extraction method
We developed two novel extraction methods, Method A and B, for comparison with traditional methods. The results The results revealed that Method A not only conserved fermentation broth but also minimized the yield reduction of iturin A and Vitamin K2 by 1.2% and 1.8%, respectively, in comparison to the traditional method. Method B was simpler than both Method A and the traditional method, with minimal impact on the yield of iturin A during the operation process, albeit a 4.32% reduction in yield compared to the traditional method. However, the extraction yield of Vitamin K2 decreased by 28.7% compared to Method A and 30.0% compared to the traditional method. Considering these factors, we selected Method A as the preferred extraction approach for subsequent co-production fermentation.
Co-production fermentation of vitamin K2 and iturin A single-factor experiment
As the temperature increased from 25℃ to 37℃, the yields of vitamin K2 and iturin A changed significantly. At 37℃, the maximum yield of vitamin K2 was 30.56 mg/L, while the yield of iturin A was 1.62 g/L. In contrast, at 28℃, the maximum yield of iturin A was 2.75 g/L, and that of vitamin K2 was 20.09 mg/L(Fig. 2A). Therefore, it can be inferred that 37℃ is favorable for vitamin K2 synthesis but not for iturin A production. When pH value increased from 6.0 to 8.0, the yield of iturin A kept increasing. When pH increased from 8.0 to 9.0, due to high pH value affecting bacterial metabolism, resulting in bacterial aging and a corresponding reduction in iturin A yield At pH = 7, the maximum yield of vitamin K2 was 27.31 mg/L; at pH = 8, the maximum yield of iturin A was 2.92 g/L; followed by pH = 7 at 2.65 g/L for iturin A production. Considering comprehensive production of iturin A and vitamin K2, we determined that the optimal temperature was 32℃ and optimal pH was 7.
Appropriate concentrations of carbon and nitrogen sources can improve the yields of vitamin K2 and iturin A individually. The most favorable soybean meal concentrations for vitamin K2 and iturin A production were respectively 14% and 12%, yielding 30.99 mg/L and 3.02 g/L(Fig. 2C). The optimal glycerol concentration for vitamin K2 production was 8%, yielding 32.42 mg/L. At a glycerol concentration of 7%, the maximum yield of iturinA reached 3.32g/L(Fig. 2D). The most suitable yeast extract powder concentration for vitamin K2 and iturinA production were 1.6%, with maximum yields of 33.62 mg/L and 3.59 g/L,respectively. Therefore, the concentrations of 12%-14% soybean meal powder, 7%-8% glycerol, and 1.6% yeast extract powder are beneficial for concurrent production.
The yield of vitamin K2 and iturin A demonstrates a gradual decline as the concentration of KCl increases. Maximum yield of vitamin K2 is achieved at a NaCl concentration of 0.5%, reaching 35.72 mg/L(Fig. 2F).. However, the addition of NaCl does not positively impact the production of iturin A. Consequently, in the subsequent co-production process, the incorporation of KCl and NaCl is kept at 0.00%. Glutamic acid serves as a crucial precursor for synthesizing glutamine, proline, arginine, and lysine. Iturin A, a polypeptide chain composed of multiple amino acids(Yue et al. 2021), is also influenced by α-ketoglutarate(El Asmar et al. 2014) in vitamin K2 production. Thus, adding glutamic acid can promote the co-production of vitamin K2 and iturin A.
Screening of Significant Factors Affecting Co-production of vitamin K2 and iturin A
The maximum yield of vitamin K2 and iturin A reach 38.89 mg/L and 3.97 g/L, respectively, at the medium level. The variance analysis presented in Tables 2 and 3 reveals that glycerol, soybean meal powder, and yeast powder have a significant effect (P < 0.05) on the fermentation process of vitamin K2. This suggests that adjusting the levels of these variables can enhance vitamin K2 yield.
During the fermentation process of iturin A, glycerol, soybean meal powder, and L-glutamic acid sodium also exhibit significant effects (P < 0.05). The determination coefficients (R2) of vitamin K2 and iturin A are 0.9774 and 0.9925, respectively, indicating that the models can explain 97.74% and 99.25% of the response variation. The relatively high values of the adjusted determination coefficient (R2adj = 0.9021; 0.9675) signify that the models possess high significance. The Ct and Pt values for the encoding coefficients of vitamin K2 and iturin A are 0.005 and 0.006 (< 0.05), respectively. This implies that the center point is significant, and additional axial points will be added for response surface analysis to refine the model further.
Optimization of medium composition by response surface experiment
"Table 3" and "Table 4" present the regression models in the form of variance analysis. If "model P" is less than 0.05, it signifies the significance of the model. The determination coefficient values (R²=0.9767 and R²=0.9579) indicate that 97.67% and 95.79% of the variation in vitamin K2 and iturin A production, respectively, can be explained. The high adjusted determination coefficient values (R²adj = 0.9534 and R²adj = 0.9159) indicate a well-fitting condition for the model. A "P" value of (< 0.05) signifies the significant influence of each variable.
Regarding the model terms, soybean meal powder (X1), glycerol (X2), soybean meal powder combined with glycerol (X1X2), soybean meal powder combined with L-glutamic acid sodium (X1X4), glycerol combined with yeast extract powder (X2X3), yeast extract powder combined with L-glutamic acid sodium (X3X4), soybean meal powder squared (X1²), glycerol squared (X2²), yeast extract powder squared (X3²), and L-glutamic acid sodium squared (X4²) all show significant effects on vitamin K2 production.
The positive linear coefficient for glycerol (0.7308) indicates that vitamin K2 production increases with higher glycerol concentration. However, negative squared coefficients for the four factor variables suggest that there is an optimal concentration beyond which these variables have an inhibitory effect on vitamin K2 production. Based on variance analysis, the following quadratic polynomial equation was derived at the coding level:
Y(vitamin K2) = 41.50–0.7508X1 + 0.7308X2–0.38X3–0.1617X4–1.46X1X2–0.55X1X3–2.87X1X4 + 1.96X2X3 + 0.5875X2X4–1.11X3X4–4.74X1² − 2.07X2² − 8.25X3² − 1.34X4²
The response surface diagram for producing vitamin K2 and iturin A by Bacillus velezensis ND was generated by keeping two variables constant at the center point while keeping the other two variables within the experimental range (Figs. 3 and 4). Figures 3 (a-c) show that there is a significant interaction between soybean meal powder concentration and other variables in terms of vitamin K2 yield parameters. When the soybean meal powder concentration increased from 100 g/L to 130 g/L, the vitamin K2 yield concentration (33.62–41.23 mg/L) increased significantly. However, if the soybean meal powder concentration exceeded 130 g/L, the vitamin K2 yield would decrease. Similar effects were observed for yeast extract powder; increasing yeast extract powder to the optimal point would increase vitamin K2 production to its maximum level, but further increasing yeast extract powder concentration would decrease vitamin K2 production (Figs. 3b,d,f). Figures 3a,e,f indicate that higher vitamin K2 yields can be obtained if glycerol concentration is between 50 and 85 ml/L. Increasing glycerol concentration will reduce vitamin K2 production because high concentrations of glycerol cannot be utilized by strains, resulting in metabolic inhibition effects. With increasing L-glutamic acid sodium concentration, there was no significant change in vitamin K2 production (Figs. 3c,d,e)).
In these 29 experiments shown, “Model P” of this model has high significance (< 0.0001). The coefficient of determination (R2 = 0.9579) and adjusted coefficient of determination (R2adj = 0.9159) indicate that the model has good fit (Tables 3). The significant variable factors affecting are soybean meal powder (X1), interaction term between soybean meal powder and L-glutamic acid sodium (X1X4), square term of soybean meal powder (X12), and square term of yeast extract powder (X32). The negative square coefficient of square terms indicates that there is a maximum value for variable concentration, beyond which there is an inhibitory effect on iturin A production. According to the analysis of variance, the following quadratic polynomial equation is obtained at the coding level:
Y (iturin A) = 4.05-0.1183X1 + 0.0158X2 + 0.0283X3-0.0508X4-0.0775X1X2-0.0550X1 X3-0.2425 X1X4 + 0.1X2X3 + 0.06X2X4 + 0.095X3X4-0.3970X12-0.0532X22-0.7395X32-0.133X42
Figure 4 (a-c) shows the interaction of soybean meal, glycerol, yeast extract and sodium L-glutamate concentration with other parameters on iturin A yield. The fermentation process of Iturin A is similar to that of vitamin K2. When the concentration of soybean meal powder and yeast extract powder increased to a certain concentration, the yield of iturin A increased, but if the concentration of carbon and nitrogen sources exceeded the threshold of effective carbon and nitrogen sources available to strains, the yield of iturin A decreased. The output value of vitamin K2 and iturin A was predicted by using the "numerical optimization" tool of Design-Expert 12 software. Under the conditions of soybean meal powder 124.04 g/L, glycerol 82.78 mL/L, yeast extract powder 17.10 g/L, L-sodium glutamate 1.36 g/L, MgSO4·7H2O 0.5 g/L, KH2PO4 1.0 g/L, trace elements FeSO4·7H2O 0.15 mg/L, MnSO4·H2O 5 mg/L, CuSO4·5H2O 0.16 mg/L, the maximum yield of vitamin K2 and iturin A was predicted to be 41.64 mg/L and 4.07 g/L respectively.
RSM Optimized Co-production Fermentation Medium Composition and Verification
Under the condition of 32 ℃ and pH 7, the optimum fermentation time of vitamin K2 was 7 days, but the yield of iturin A and the number of viable bacteria reached the maximum at the 5th day, which were 4.21 g/L and 5.7 x1012 cfu/ml, respectively. The yield of iturin A decreased by 0.13 g/L and 0.16 g/L on the 7th and 8th day, respectively. Therefore, when nutrients were insufficient, Bacillus valesei ND would use its lipopeptide antibiotic iturin A to meet the growth needs, thus reducing its yield. After 7 days of co-production, the actual yields of vitamin K2 and iturin A were 41.72 mg/L and 4.08 g/L, which were close to the predicted values of 41.64 mg/L and 4.07 g/L, respectively.
Scale-up experiment of biofilm reactor combined with vitamin K2 and iturin A for co-production and fermentation
In the combined fermentation of iturin A and vitamin K2 (Fig. 6), the total output of vitamin K2 and iturin A reached the maximum of 46.88 mg/L and 5.58 g/L respectively on the 7th day of biofilm reactor fermentation, and the output of vitamin K2 and iturin A obtained by suspension cell reactor was 39.72 mg/L (7 days) and 4.26 g/L (8 days) respectively. The total yields were 56.07 mg/L (6 days) and 5.71 g/L (7 days) in biofilm reactor alone, and 43.80 mg/L (8 days) and 4.56 g/L (7 days) in suspension cell reactor alone. It can be seen that the biofilm reactor has the advantages in the joint production of these two products. The yield of iturin A is only reduced by 0.12 g/L, and at the same time, the yield of vitamin K2 produced by suspension cell reactor can be higher than that produced by suspension cell reactor, and a certain amount of fermentation time can be saved.