Expression of PLD by the recombinant B. choshinensis with two different promoters
In this study, two different recombinant B. choshinensis expressing phospholipase D were successfully constructed by using two different plasmid vectors, namely pNY326 (weak P5 promoter) and pNCMO2 (strong P2 promoter).
When the transformed strains were cultured on the plate, the plate culture showed the presence of B. choshinensis/pNY326-PLD, while no any trace of B. choshinensis/PNCMO2-PLD. The promoters of the two expression vectors had significant differences in strength. As depicted in Fig. 1, B. choshinensis/pNY326-PLD containing P5 promoter was consistent with the growth of the control group and expressed a small amount of PLD, which indicated that the low expression of PLD by B. choshinensis under weak constitutive promoter P5 would not damage the host cell. It was speculated that the strength of the P2 promoter in pNCMO2-PLD was much greater than P5 promoter  in pNY326-PLD, inducing strong expression of PLD, but high PLD expression severely inhibited cell growth due to its cytotoxicity. Therefore, for improving PLD expression, two problems need further consider: One is how to weaken the strength of the P2 promoter, and the other is how to reduce the adverse effects of PLD on cell growth.
According to reports, magnesium ions could specifically inhibit the activity of P2 promoter  and alleviate the cytotoxicity of PLD . Therefore, in this study, 60 mM of MgSO4 was added to the transformation plate to culture B. choshinensis/pNCMO2-PLD. As depicted in Fig. 2, B. choshinensis/pNCMO2-PLD showed normal growth, with a significant increase in PLD production. Comparing the growth of the two recombinant cells (B.choshinensis/pNY326-PLD was cultured in the TM medium and B.choshinensis/pNCMO2-PLD was cultured in the TM medium with 60 mM MgSO4) (Fig. 3), the higher extracellular activity of PLD (2.03 U/mL) was obtained by the pNCMO2-PLD expression system, which was 67-folds than pNY326-PLD (0.03 U/mL). These results indicated that the strong promoter could enhance PLD expression under magnesium ion stress, which plays a vital role in PLD expression by B.choshinensis/pNCMO2-PLD. However, its mechanism needs further investigation.
Determination of the culture medium
In the initial experiments, the TM medium was used to culture the cells. However, it found the recombinant cells showed unstable growth, suggesting that the TM medium was not optimal for PLD expression. Therefore, we first compared three culture medium, including 2SY, TM medium, and a fermentation medium from B. choshinensis to efficiently express the pullulan  under 60 mM MgSO4, then chosed the fermentation medium for futher optimization. The results are depicted in Fig. S1 and S2. After optimization, the modified fermentation medium contained 30.0 g/L glucose, 30.0 g/L beef extract, 25.0 g/L yeast extract, and 60 mM MgSO4, which increased PLD production by 7 times.
Effect of metal ions on phospholipase D expression by B. choshinensis
The optimal concentration of Mg2+ was investigated because Mg2+ could increase PLD production in B. choshinensis. As shown in Fig. 4a, the biomass exhibited a slight increase at different Mg2+ concentration, while the PLD activity continued to increase until it reached a peak value of 9.8 U/mL at 100 mM Mg2+.
The result showed that magnesium ion had a positive promoting effect on PLD expression in B. choshinensis/pNCMO2-PLD, speculating similar effects from other metal ions. Therefore, the effects of Na+, K+, Ca2+, and Mg2+ on cell growth and PLD production were investigated in the modified fermentation medium. As depicted in Fig. 4b, only Mg2+ and K+ stimulated cell growth to some extent, while Ca2+ and Na+ inhibited cell growth. For PLD activity, all ions had promoting effect, but the growth was highly stimulated in the presence of Mg2+, reaching the highest activity of 10.09 U/mL. So, the modified fermentation medium with 100 mM MgSO4 (named as medium A) was used as the optimal medium in the following experiments. Also, the anions of Na+, K+, and Mg2+ groups were all SO42-, but the PLD activity was significantly different, which indicated that anion was not the key factor for promoting the PLD activity.
Analysis of secreted PLD production by SDS-PAGE
In order to understand how Mg2+ affects PLD production, SDS-PAGE analysis was performed. As shown in Fig. 5, a band about 64 kDa in protein samples of B. choshinensis/pNCMO2-PLD was detected (lanes 1, 2, 6, 7), while there was no obvious band at 64 kDa in protein samples of B. choshinensis (lanes 3, 5). It indicated that PLD was successfully expressed in recombinant cells. Comparing extracellular protein samples showed in lane 6 and 7, Mg2+ induced more PLD expressed and secreted, which is in good agreement with the PLD activity value (Fig. 4a). Besides, expressing PLD could cause sever protein leakage (lanes 5, 6, 7) due to its cytotoxicity. Comparing intracellular protein samples showed in lane 1 and 2, there were PLD detected. It suggested PLD might accumulate in cells. Furthermore, lane 2 presented less PLD than lane 1, indicating more accumulated intracellular PLD could inhibit its extracellular expression and secretion, and Mg2 might hinder intracellular PLD to accumulate and benefit PLD expression and secretion.
Ion perturbation in the PLD-imposed cells
Xiong et al. found that the PLD-imposed cells in E.coli underwent ion perturbations, resulting in cell injury or cell death. Na+ stress protects the PLD-imposed cells by improving oxidative phosphorylation through a positive change in the membrane potential via redistribution of Na+/K+ inside and outside the cell . Therefore, B. choshinensis/pNCMO2 was cultured in medium A without Mg2+ (NP group), and B. choshinensis/pNCMO2-PLD was cultured in medium A (EPMg group) and without Mg2+ (EP group) to determine whether Mg2+ has a similar regulatory mechanism for PLD expression. The changes of seven ions in cells (Na+, K+, Mg2+, Ca2+, Fe2+/3+, Zn2+ and Mn2+) were evaluated in the NP, EP and EPMg groups.
We initially focused on ion distribution at 24 h (Fig. 6a and 6b) and found the same change trend in the three groups. For instance, [Na+] and [Fe2+/3+] decreased, while [K+], [Mg2+], [Ca2+], [Mn2+], and [Zn2+] increased, resulting in similar cell growth. Compared with the EP group, higher [Mg2+] and [Mn2+] with an increasing rate of 67.86% and 61.62%, and lower [K+] and [Ca2+] with a decreasing rate of 23.91% and 54.17% were observed in the EPMg group. In 36 h, most of the ions maintained a steady-state, while [Ca2+] and [Zn2+] showed an increase of 101.93% and 140.32% (Fig. 6c and 6d). Meanwhile, the expressed PLD exhibited significant changes in [K+], [Mg2+], and [Ca2+], suggesting that PLD expression might affect the cell growth status by influencing ion distribution. Under Mg2+ stress, [K+] and [Mg2+] increased by 198.55% and 287.83%, while [Na+] and [Ca2+] decreased by 31.08% and 66.31%, respectively. These results demonstrate that Mg2+ stress changes the regulation of some intracellular ions in the PLD-imposed cells, which might change the cell physiology status.
The changes of ion contents in the cells were not consistent at different growth periods under Mg2+ stress. Notably, [Ca2+] exhibited extraordinary change at 24 h and 36 h. Compared to the NP group, [Ca2+] of the EP group significantly increased. When Mg2+ stress was introduced to the medium (EPMg group), [Ca2+] reduced sharply, indicating that [Ca2+] might regulate cell growth. Comparing the changes of ions in three groups at 36 h, [K+] and [Mg2+] were found to have the lowest concentration in the EP group and the highest concentration in the EPMg group. On the contrary, [Ca2+] had the highest value in the EP group and the lowest value in the EPMg group. The combined results of Fig. 6e and 6f suggest that the addition of Mg2+ in the PLD-imposed cells could keep the cells with low [Ca2+], high [K+], and [Mg2+] contents, which is beneficial for cell growth.
The toxic effect of PLD on cells was minute due to the low level of PLD expression in the early growth stage. Therefore, the cell growth of both EP and EPMg groups was slightly lower than the NP group before 24 h (Fig. 6e). From 24 h, PLD expression started to increase to a great extent, which significantly changed the ion distribution of the experimental group, inducing cells showed significant different growth states. Of the three groups, the biomass in the EP group was the lowest at 36 h, which began to decline after 36 h. The biomass in the EPMg group surpassed the NP group at 36 h and continued to rise. As depicted in Fig. 6b and 6e, [Ca2+] of the EP group with the worst growth state was significantly higher than the other two groups at 24 h and 36 h. However, the EPMg group with the best growth state at 36 h was significantly lower than the EP and NP groups, indicating that PLD significantly increased the intracellular [Ca2+] content, which was harmful to the cells. In contrast, the addition of Mg2+ could significantly inhibit the increase of Ca2+ induced by PLD expression to alleviate the toxic effect of PLD on cells, resulting in the continuous increase of cell growth and PLD expression. Meanwhile, the growth of B. choshinensis was inhibited under Ca2+ stress (Fig. 6a), indicating Ca2+ to be an essential factor affecting the growth of B. choshinensis.
The intracellular concentrations of four ions (Na+, K+, Mg2+, Ca2+) were investigated from 12 h to 60 h to further analyze the effect of ion changes. Compared with the NP group, [Na+] showed a similar decreasing trend at the first 24 h, while [K+] had a similar increase in the EP and EPMg groups (Fig. 7a and 7b). However, after 24 h, [K+] began to decrease accompanied by an increase in [Na+] content in the EP group, while in the EPMg group, [K+] continued to increase accompanied by a steady low [Na+]. In 24 h, the continued PLD expression in the cell caused K+ efflux and Na+ influx in the EP group, which was harmful to the cells . However, the addition of Mg2+ could regulate the change in ions. As depicted in Fig. 7c, [Mg2+] in the EPMg group showed a higher growth rate than the other two groups. It is reported that Mg2+ could enhance the activity of Na+-K+-ATPase , thereby triggering K+ influx and Na+ efflux in the EPMg group, keeping [Na+] lower and [K+] higher inside the cells to promote cell growth. As depicted in Fig. 7d, [Ca2+] in the EP group with the lowest biomass had a high concentration. In contrast, the EPMg group with the highest biomass had low [Ca2+] during the whole process. Although low [Ca2+] was observed in the NP group before 24 h, it began to increase after 36 h, showing a sharp increase. [Ca2+] of the EP group simultaneously exhibited similar increasing trend after 36 h. As depicted in Fig. 6e, the EPMg group showed increased biomass, while the other two groups showed a decreased biomass after 36 h. The results suggested that the growth of B. choshinensis was coupled with the intracellular [Ca2+] and low intracellular [Ca2+] was beneficial for cell growth.
Xiong et al. expressed PLD in E. coli and found that it could cause a continuous influx of Na+, the outflow of K+, and increase of intracellular Ca2+, while cations stress (especially Na+) alleviated cell growth inhibition and profoundly increased PLD production by redistributing the ions inside and outside the cell. The effect of salt stress on PLD expression was consistent with our study results. However, PLD expression in B. choshinensis showed a different phenomenon without any continuous influx of Na+. Therefore, it was speculated that the effect of Mg2+-induced protection on cell growth and PLD expression might be related to the P2 promoter expression.
Transcription level of PLD and HWP genes in B. choshinensis/pNCMO2-PLD under Mg2+ stress
HPD31 Wall Protein (HWP) is the cell wall protein secreted by B. choshinensis, whose expression could also be regulated by the P2 promoter [34, 35]. Therefore, the analysis of HWP expression level could further reflect the level of P2 promoter strength.
As depicted in Fig. 8a and 8b, the transcriptional levels of PLD and HWP genes in the EP group were consistent with the EPMg group at 12 h, but both increased sharply in the EP group at 24 h, which were 5.14 times and 4.52 times higher than at 12 h, respectively. Subsequently, the transcription level of PLD and HWP decreased significantly at 36 h. It might be attributed to PLD accumulation which induced a toxic effect on cell growth (Fig. 6e), resulting in the death of the recombinant strain and decreased gene expression. As for the EPMg group, the transcription levels of PLD and HWP exhibited minute changes at 24 h and 36 h compared to 12 h. The results indicated that Mg2+ reduced the strength of the P2 promoter reflected by the decreased transcriptional level of the HWP gene, further maintaining the PLD gene at a relatively low transcription level during cell growth. Therefore, the toxic effect caused by highly expressed PLD was inhibited, which was conducive to the continuous cell growth and expression of PLD in B. choshinensis (Fig. 7a and 7b).
As depicted in Fig. 6e, cell growth in three groups was consistent before 24 h, but it changed after 24 h. The biomass of the NP and EP groups gradually declined after 36 h, showing the worst growth state in the EP group. In contrast, the biomass of the EPMg group significantly increased from 24 h to 36 h, and continued to increase after 36 h until it reached a higher biomass state than the NP and EP groups after 36 h. The result demonstrated that the decreased growth of B. choshinensis by PLD expression was due to its toxicity, while the addition of Mg2+ attenuated the toxic effect of PLD on cells, exhibiting a positive effect on the growth of B. choshinensis. As depicted in Fig. 6f, the PLD activity of the EPMg group was higher than the EP group during the entire growth process, indicating that Mg2+ effectively promoted the expression of PLD. Besides, PLD activity in the EP group decreased after 36 h, showing the same change in its biomass. Furthermore, the analyzed cells began to die after 36 h due to the high toxicity of PLD expression, which might cause the intracellular protease leakage leading to the degradation of extracellular PLD.