Efficient secretory expression of phospholipase D for the high-yield production of phosphatidylserine and phospholipid derivates from soybean lecithin

Abstract

Phospholipase D (PLD) is an essential biocatalyst for the biological production of phosphatidylserine and phospholipid modification. However, the efficient heterologous expression of PLD is limited by its cell toxicity. In this study, a PLD was secretory expressed efficiently in Bacillus subtilis with an activity around 100 U/mL. A secretory expression system containing the signal peptide SP_EstA and the dual-promoter P_HpaII-SrfA was established, and the extracellular PLD activity further reached 119.22 U/mL through scale-up fermentation, 191.30-fold higher than that of the control. Under optimum reaction conditions, a 61.61% conversion ratio and 21.07 g/L of phosphatidylserine production were achieved. Finally, the synthesis system of PL derivates was established, which could efficiently synthesis novel PL derivates. The results highlight that the secretory expression system constructed in this study provides a promising PLD producing strain in industrial application, and laid the foundation for the biosynthesis of phosphatidylserine and other PL derivates. As far as we know, this work reports the highest level of extracellular PLD expression to date and the enzymatic production of several PL derivates for the first time.

1. Introduction

Phosphatidylserine (PS), a natural acidic phospholipid (PL), mainly exists in brain cells. As one of the active substances of the cell membrane, PS plays a vital role in the process of biological metabolism [1]. Recent studies indicated that PS can effectively improve brain function, optimize memory, relieve brain fatigue, and protect the central nervous system, called "intelligent nutrient" [2].

The majority sources of PS are generally extracted from animals and plants, especially bovine brains, soybeans, and egg yolks. However, the composition of natural PLs is complex, they have similar physical properties and the PS content is sparse [3]. Thus, in order to obtain high purity PS, multiple purification steps and large quantities of solvents need to be performed during the extraction process, which increases the environmental impact. In addition, extraction from animal tissue may also carry a risk of disease transmission, such as bovine spongiform encephalopathy [4].

In this situation, Phospholipase D (PLD, EC 4.1.4.4)-based PS biocatalysis process has the potential to address a number of extraction method challenges [5]. PLD can catalyze transphosphatidylation reaction between the phosphoester bonds of phosphatidylcholine (PC) and l-serine (L-Ser) to produce PS, which can obtain highly pure PS and reduce solvent consumption [6]. In addition to PS, PLD can also facilitate the production of PL derivatives that are difficult to synthesize through chemical catalysis, e.g., phosphatidyl-ascorbic acid [7], phosphatidyl-glucose [8], and phosphatidyl-tyrosol [9]. The functionality of some PL derivates, such as antioxidant and anticancer activity, have been reported [10,11]. Furthermore, the substrate PC of these reactions can easily be sourced from numerous natural sources, such as soybean and egg, which are relatively inexpensive to produce [4], and the most important problem is the limited supply and low expression level of PLD.

To date, a wide variety of PLDs have been successfully heterologously expressed in Pichia pastoris [12], Corynebacterium glutamicum [13], Escherichia coli [14], and Bacillus subtilis [15]. Among these heterologous expression systems, the E. coli expression system accounts for the vast majority (approximately 86%) [16]. However, E. coli is relatively limited in producing food processing enzymes due to its generation of endotoxin. Most reports of heterologous expression of PLD in E. coli are intracellular expression, which are unfavorable for enzyme separation and industrial production. Moreover, it is reported that PLD accumulation in cells will result in plasmid instability, cell lysis and further reduction of PLD activity, the efficient secretory expression of PLD can reduce PLD accumulation in cells and increase PLD production [[5], [17], [18]]. Owing to its excellent fermentation performance, capable of efficiently secreting proteins, food-grade safety, and free of endotoxins and exotoxins, reports of PLD expression in B. subtilis have increased significantly in recent years [15,19]. Nevertheless, most of the reported expression levels of PLD in B. subtilis are less than 100 U/mL, which requires further improvement [17].

In this study, we aimed to construct a food-grade and high-efficient PLD expression strain for the efficient production of PS and PL derivates. To this end, a combination strategy was employed to increase the extracellular PLD activity in B. subtilis. Plasmids with various screened signal peptides and promoters from B. subtilis were engineered to screen the optimal SPs and promoters. Then, the PLD production capacity of the strain was further evaluated by scale-up fermentation. In addition, in order to develop an efficient enzymatic synthesis method for preparing PS and other PL derivates catalyzed by PLD, the effects of conversion conditions were systematically investigated. This strain exhibited obvious improvements in enzyme activity, with PS and other PL derivates synthesized efficiently.

2. Materials and methods

2.1. Enzymes and chemicals

One Step Cloning Kit was supplied by Vazyme (Nanjing, China). DNA Polymerase, restriction enzymes, and T4 DNA ligase were obtained from Takara Bio (Beijing, China). Lecithin (90% PC, from soybean), 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (99%, DPPC), and 1,2-dipalmitoyl-sn-glycero-3-phospho-l-serine (99%, DPPS) standard were purchased from Macklin (Shanghai, China). PS (97%, from soybean), Choline oxidase (from Alcaligenes sp.), and peroxidase (from horseradish) were purchased from Sigma-Aldrich (St. Louis, MO, USA). The reagents and chemicals, except for the high-performance liquid chromatography (HPLC) phases of chromatographic grade, were of analytical grade.

2.2. Bacterial strains and growth conditions

A Streptomyces sp. PLD encoding gene was deposited in our lab [20]. Plasmid construction and cloning were performed on E. coli JM109, and B. subtilis WB600 was utilized for PLD expression. Table S1 and Table S2 list the bacterial strains, plasmids, and primers used in this study.

Cultures of E. coli were carried out in Luria-Bertani (LB) medium (5 g/L yeast extract, 10 g/L tryptone, and 10 g/L NaCl). For B. subtilis WB600, LB medium and Terrific Broth (TB) medium (72 mM K₂HPO₄, 17 mM KH₂PO₄, 24 g/L yeast extract, 5 g/L glycerol, and 12 g/L tryptone) were used as seed and fermentation media in the flask. The fed medium (25 g/L glycerol, 120 g/L soybean cake powder, 60 g/L peptone (soya), 17 mM KH₂PO₄, 72 mM K₂HPO₄) was utilized in the process of fed-batch fermentation. Transformed bacteria were selected using two antibiotics (0.1 mg/mL ampicillin for E. coli JM109 and 0.05 mg/mL kanamycin for B. subtilis WB600).

2.3. Construction of the expression vector for PLD production

The PLD gene, deposited in our lab, was cloned into the BamHI/NdeI sites of the plasmid pMA5 and transformed into E. coli JM109. The plasmid pMApld was sequenced and transformed into B. subtilis WB600 to produce PLD.

2.4. Selecting optimal expression elements

To achieve a higher extracellular expression level of PLD, we selected different signal peptides and promoters based on the pMApld. The signal peptides of B. subtilis were retrieved from the Signal Sequence Database and analyzed signal peptide properties by web servers (GRAVY CALCULATOR, ExPASy, and SignalP 5.0). We screened signal peptides with different properties and amplified these SPs genes from the chromosome of B. subtilis 168. The SPs genes were fusion to the N-terminus of the PLD gene by overlap extension PCR. The construction of the promoter screening system is based on our previous study [21]. A total of 10 native promoters from B. subtilis were ligated to the pMApld-SP_EstA (Table S1).

2.5. Assay of PLD enzyme activity

One unit (U) of hydrolytic activity of phospholipase D was defined as the amount of enzyme that produces 1 μmol of choline per minute under the assay conditions. According to a modified protocol by Hou et al. [13], enzyme-linked colorimetry was used to measure PLD hydrolysis activity. The reaction mixture (100 μL) contained 60 μL substrate solution (10 mg/mL lecithin, 0.1% (v/v) Triton X-100, 15 mM CaCl₂, and 40 mM Tris-HCl (pH 7.5)) and 40 μL of an enzyme sample. Incubation at 60 °C for 20 min with shaking was followed by boiling for 5 min with 50 mL EDTA solution (50 mM) to terminate the reaction. After cooling down, 500 μL colorimetric solution (40 mM Tris–HCl, 0.75 U of choline oxidase, and 0.5 U of peroxidase) were added and heated at 37 °C for 30 min. An absorbance measurement was performed at 505 nm on the reaction mixture. In place of the enzyme solution, a standard solution of choline chloride was used to obtain the calibration curve. A triplicate of each assay was performed.

2.6. SDS-PAGE analysis

After mixing with SDS-PAGE loading buffer, protein samples were boiled for 10 min. 20 μL of the mixtures were analyzed by 12% (w/v) polyacrylamide gel. Gels were stained with coomassie brilliant blue dye after electrophoresis.

2.7. Scale-up fermentation production of PLD in a 7-L fermenter

On an agar plate, a single recombinant colony was selected and inoculated into LB medium for 12 h at 37 °C and 220 rpm. Subsequently, 1 mL culture solution was used to inoculate the seed medium and cultivated for 18 h under the same conditions. The fermentation was performed by inoculating 5% of the seed culture into a 7-L fermentation medium and cultivating it at 37 °C. The dissolved oxygen level was maintained at above 10% through regulation of the rotation speed and ventilation, and a pH value of 7.0 was maintained. The nitrogen sources (tryptone and yeast extract) in the optimized medium were also replaced with peptone (soya) and soybean cake powder to further reduce the cost of the medium used in scale-up fermentation.

2.8. Enzymatic production of PS catalyzed by PLD

The biphasic systems, described by Hagishita et al., were used to synthesize PS from L-Ser and PC [22]. The aqueous phase contained 80 mg/mL L-Ser dissolved in 3 mL of the crude enzyme (40 U/mL), and the organic phase contained 16 mg/mL lecithin dissolved in 6 mL toluene. A reaction period of 12 h was conducted at 40 °C with shaking of the mixed solution.

PLD-catalyzed reaction samples were analyzed by HPLC (UltiMate U-3000, DIONEX) equipped with an evaporative light scattering detector (HPLC-ELSD) and liquid chromatograph-mass spectrometer (LC-MS, MALDI SYNAPT MS). The PS conversion rate (%) was defined as the percentage of PS obtained as compared to the initial concentration of PC.

2.9. Enzymatic production of PL derivates catalyzed by PLD

The procedure for enzymatic production of PL derivates is guided by the optimal reaction conditions of PS synthesis. 342 mg DPPC was dissolved in 9 mL butyl acetate, and 240 mg alcohol was dissolved in 3 mL enzyme solution. It was incubated at 35 °C for 10 h with shaking of the mixed solution.

The PL derivates samples were analyzed by LC-MS. Additionally, four PL derivates (phosphatidyl-glucose, phosphatidyl-mannitol, phosphatidyl-glucosamine, and phosphatidyl-thiamine chloride) were separated from samples by thin-layer chromatography (TLC). I₂ was used as the chromogenic reagent, and the developing solvent was chloroform/methanol/7 M ammonia (65:30:4). The PL derivates were scraped off from the TLC plate and then eluted by methanol/chloroform (10:1, v/v). The extracted PL derivates were analyzed by HPLC-ELSD and LC-MS, and used to determine the conversion rate. The PL derivates conversion rate (%) was defined as the percentage of PL derivates obtained as compared to the initial concentration of PC.

3. Results and discussions

3.1. Signal peptide engineering to facilitate the secretory PLD expression

In B. subtilis, signal peptide (SP) plays an essential role in the secretory system. In B. subtilis, there are three known secretion pathways: the general secretory (Sec) pathway, twin-arginine (Tat) translocation, and ABC transporters [23]. Several reports have shown that choosing an appropriate signal peptide could also enhance the expression of proteins [19,24].

Though the existing studies suggest that parameters including relative hydrophobicity level, molecular weight, and Gibbs free energies of its mRNA might be factors that influence signal peptide-guided secretory protein expression, there is still a lack of comprehensive understanding of how they can lead to high level protein secretion [25]. Moreover, no specific way of predicting the optimal SP for a target enzyme has been developed, and the non-rational screening process is usually necessary for the optimization of the SP. Thus, a total of 244 SPs, derived from B. subtilis, in the Signal Sequence Database (http://www.signalpeptide.de/index.php) were used for analysis. SPs with a D-score below 0.5 were excluded by the SignalP 5.0 web server (https://services.healthtech.dtu.dk/service.php?SignalP-5.0). The N-terminal region, the hydrophobic core H-region, and the C-terminal region of the remaining signal peptides were also analyzed [26]. Then, the grand average of hydropathicity (GRAVY) and molecular weight of signal peptides were calculated by the ProtParam tool (https://web.expasy.org/protparam/), and the Gibbs free energies of folding of mRNA, a 170-nt fragment following the transcription start site of the promoter, were calculated through the FoldRNA web server (http://www.softberry.com/berry.phtml?topic=foldrna&group=programs&subgroup=rnastruct). We finally selected two Tat signal peptides, YwbN and PhoD, and a total of 43 Sec signal peptides with different GRAVY, Mw, and ΔG (Table 1).

Table 1.

Comparison of the signal peptides used for PLD secretion in this study.

Name	Origin	Theoretical Mw (Da)	GRAVY of H-region	ΔG	Extracellular PLD activity (U/mL)
AmyE	B. subtilis	3520.33	2.26	−40.4	0.41
YwoF	B. subtilis	3147.86	2.65	−46.4	36.81
XynD	B. subtilis	2886.63	3.00	−42.6	1.69
WapA	B. subtilis	3319.21	2.89	−41.70	0.82
BglS	B. subtilis	3027.76	2.57	−34.7	1.05
SacB	B. subtilis	3014.58	1.44	−44.6	4.77
EstA	B. subtilis	3447.33	2.45	−39.50	5.74
Bpr	B. subtilis	3145.8	2.02	−49.40	0.98
YolA	B. subtilis	3044.69	2.49	−44.00	0.65
NprE	B. subtilis	2663.23	1.96	−42.5	0.38
Yxal	B. subtilis	3974.74	2.09	−58.90	0.84
NprB	B. subtilis	2949.54	1.7	−40.1	0.88
YncM	B. subtilis	3654.43	1.81	−50.30	0.14
SpsC	B. subtilis	3494.35	2.07	−33.7	0.61
Csn	B. subtilis	4094.03	2.25	−33.00	3.82
DacB	B. subtilis	3035.75	2.82	−41.6	0.47
WprA	B. subtilis	3328.07	2.61	−51.8	0.52
LytE	B. subtilis	2450.92	1.76	−40.3	1.10
LytF	B. subtilis	2542.07	1.41	−41.1	0.65
Epr	B. subtilis	2955.63	2.42	−39.90	0.52
PelB	B. subtilis	2753.49	2.56	−39.9	0.93
PenP	B. subtilis	3875.66	2.16	−32.90	25.14
OppA	B. subtilis	2350.99	2.64	−44.70	1.93
LytB	B. subtilis	2649.35	3.23	−37	3.87
Pel	B. subtilis	2118.62	1.97	−46.20	1.13
DacA	B. subtilis	3348.29	2.49	−40.70	5.52
YclQ	B. subtilis	2985.59	2.75	−41.5	4.00
PglS	B. subtilis	3592.46	2.95	−41.6	3.74
YdjH	B. subtilis	3198.04	3.04	−49.40	14.22
YvnB	B. subtilis	2530.03	2.10	−46.40	5.65
YqzG	B. subtilis	2350.94	2.84	−43.30	35.17
YoqM	B. subtilis	2559.15	1.98	−45.80	6.98
SsuA	B. subtilis	1785.35	3.29	−43.7	5.64
YybD	B. subtilis	1803.28	2.69	−42.10	3.74
YozM	B. subtilis	2755.55	2.39	−39.50	0.44
YoeB	B. subtilis	2513.14	2.63	−36	0.41
YybP	B. subtilis	1713.15	2.10	−45.00	0.39
YkvT	B. subtilis	2884.51	2.18	−35.6	0.38
YwmC	B. subtilis	2561.20	2.06	−37.60	0.42
AraN	B. subtilis	2356.14	2.87	−36.30	0.25
YxkH	B. subtilis	2067.62	2.18	−47.10	0.34
YczF	B. subtilis	2364.04	2.82	−42.6	0.42
MntA	B. subtilis	1898.31	1.90	−48.2	0.40
PhoD	B. subtilis	6284.07	\	−38	7.99
YwbN	B. subtilis	2700.1	\	−46.40	0.46

B. subtilis WB600, a protease-deficient strain, was selected as the expression host; multicopy plasmid pMA5 was used to construct expression vector pMApld. The extracellular PLD activity produced by this recombinant strain was 0.62 U/mL.

The candidate signal peptides were inserted upstream of the N-terminus of the PLD gene in pMApld, constructing pMApld-SP_x (x is the name of the signal peptide). To evaluate the efficiency of SPs, the extracellular PLD activity was measured (Fig. 1). Extracellular PLD activity has been observed to vary significantly between 0.14 U/mL and 36.81 U/mL. The signal peptide EstA, a Sec pathway signal peptide, was the most optimal candidate for the secretory expression of PLD with an activity almost 59.37-fold of that of the control. As indicated in Fig. S1A, PLD activity was in agreement with the SDS-PAGE result, a significant protein band (lane 3, 4, and 5) with an approximate molecular mass of 55 kDa (Fig. S1A). Western blot analysis and theoretical values were in agreement (Fig. S1B). Additionally, PLD with three other signal peptides (YdjH, PenP, and YoqM) also showed higher extracellular activity (>10 U/mL). Thus, it was concluded that PLD tended to be highly secretory expressed under the guidance of suitable signal peptides. As a result, the following experiments were conducted using the pMApld-SPEstA, which contained EstA as a signal peptide. According to all of the data in this study, high secretion efficiency appears not to correlate with the aforementioned characteristics, in accordance with some reports [24,27].

Fig. 1.

Secretion expression of PLD directed by all candidate signal peptides. The blue bar denotes the Tat pathway signal peptides, and the red bar denotes the control (pMApld).

3.2. Promoter engineering to improve the PLD expression level

In addition to signal peptide screening, optimizing the promoter is also a cost-effective method. A widely adopted strategy for developing efficient promoters is double promoter systems. Double promoter systems in B. subtilis are generally designed as two promoters arranged in tandem, which are located upstream of the gene [28]. It is reported that expression activity may be further increased by double promoters rather than single promoter [[24], [27], [32]]. In B. subtilis, there are 17 sigma factors which associate with RNA polymerase core enzyme for promoter recognition and transcription initiation. Two promoters in double promoter systems are targeted either by the same RNA polymerase core enzyme or different ones, which may result in enhancing the promoter activity within the window and extending the gene expression window, respectively [31]. To achieve higher PLD expression, 12 different promoters from B. subtilis were evaluated. The candidate promoters contained P_sigX, P₄₃, P_veg, P_pst, four promoters of extracellular protease genes (P_nprB, P_bpr, P_epr, P_aprE), and two auto-inducible promoters (P_srfA and P_gsiB). Other characteristics of promoters can be seen in Table S3. To decrease the use of inducer and simplify the PLD fermentation procedures, these promoters are all constitutive or auto-inducible promoters. The promoters were inserted between the upstream of the N-terminal of SP_EstA and downstream of P_HpaII to establish 12 recombinant plasmids.

The strains carrying three promoters (P_srfA, P_nprB, and P₄₃) displayed higher extracellular PLD activities than the control strain (pMApld-SP_EstA) and the other recombinant strains (Fig. 2). Obviously, the extracellular PLD activity of the strains carrying P_srfA and P₄₃ were higher than the anther promoter, and reached 42.77 U/mL and 40.90 U/mL, which was about 1.16-fold and 1.11-fold of that of the control, respectively. As shown in Fig. S2, the supernatant proteins were analyzed by SDS-PAGE. There was a correlation between enzyme activity and protein band thicknesses.

Fig. 2.

Comparison of PLD expression from various promoters.

3.3. Scale-up PLD production in the fermenter

Based on the optimization of SPs and promoters in the shake flask experiment, the recombinant B. subtilis WB600 harboring pMApld4-SP_EstA was further cultivated with scale-up fermentation in a 7-L fermenter to evaluate PLD production capacity.

During the batch culture process, the cells grew faster in the early stage, and the maximum cell density OD₆₀₀ reached 17.18 at 36 h (Fig. 3). Similar to cell growth, the PLD activity remained on an upward trend and reached a peak value of 92.82 U/mL, which was nearly 2.17-fold of that observed in shake flask cultivation.

Fig. 3.

Fermentation curve of batch culture.

The fed-batch culture strategy was further investigated in this study. The feeding was started after 16 h of growth according to the results of batch culture. It could be found that the cell density OD₆₀₀ and enzyme activity remained on an upward trend after 36 h (Fig. 4A). Although cell density OD₆₀₀ steadily increased, the PLD activity reached a maximal activity of 119.22 U/mL at 44 h (Fig. 4A), which was 2.78-fold higher than in shake flask cultivation. SDS-PAGE analysis was also performed to verify the results (Fig. 4B). In comparison to batch culture, the cell density OD600 was 2.59-fold higher, whereas PLD activity increased by only 1.28-fold. This might result from nutrients being used for cell growth rather than protein synthesis; Plasmid loss due to prolonged fermentation time might also explain the insignificant enhancement of secreted PLD. Accordingly, compared to the recent reports on extracellular PLD expression, it is promising to meet industrial requirements (Table 2).

Fig. 4.

Fermentation curve of recombinant B. subtilis WB600 harboring pMApld4-SP_EstA in a 7-L fermenter. (A) Fed-batch culture strategies. (B) SDS–PAGE analysis of PLD expression. Lanes: M, low molecular-weight protein marker; 1, 12 h; 2, 24 h; 3, 36 h; 4, 48 h; 5, 60 h.

Table 2.

The recent reports of extracellular PLD expression.

Recombinant strain	Enzyme activity (U/mL)	PS conversion (%)	PS yield (g/L)	Reference
B. subtilis WB600	24.2	36	1.2	[19]
E. coliRosetta	16	ND	ND	[29]
Streptomyces lividans	69.12	ND	ND	[18]
C. glutamicum	1.9	48.6	1.94	[13]
B. subtilis WB600	2.84	65	0.64	[15]
B. choshinensis	10.09	ND	ND	[17]
B. subtilis WB600	119	61.6	21.1	This study

3.4. PLD-catalyzed production of PS

The PLD-catalyzed PS synthesis was carried out in a biphasic system. In this system, organic solvents can enhance the solubility of nonpolar substrates (lecithin), reduce hydrolysis side reactions, and inhibit microbial pollution. So, we first optimized the system by selecting safe solvents (ethyl butyrate, cyclopentyl methyl ether, butyl acetate, and hexane) allowed in food processing, and compared them with the toxic toluene that was often used [13]. The PS concentration was calculated from the standard curve. The HPLC–ELSD analyses suggested that the PC can be successfully converted into PS when toluene, cyclopentyl methyl ether, butyl acetate, and ethyl butyrate were used as organic solvents (Fig. 5A and S4A). To further determine the successful synthesis of PS, the LC-MS was applied to determine the molecular weight and fragment ions of products and DPPS standard. The m/z 734.49 is consistent with a relative molecular mass of DPPS (Fig. 5B). The fragment ions of the product and DPPS standard are also consistent, and the fragment ions were identified to be m/z 734.50 (C₃₈H₇₃NO₁₀P⁻), m/z 255.23 (C₁₆H₃₁O₂⁻), and m/z 647.47 (C₃₅H₆₈O₈P⁻) (Fig. S3).

Fig. 5.

PLD-catalyzed PS synthesis. (A) HPLC−ELSD results of standard PS and PLD-catalyzed reaction. (B) LC-MS results of PLD-catalyzed DPPS synthesis.

As an enzyme-catalytic reaction, the PS yield is highly influenced by the conversion conditions. To maximize the PS conversion rate, the conversion conditions, such as organic solvents, volume ratios between organic and aqueous phases, substrate molar ratio, reaction time, reaction temperature, and enzyme concentration, were systematically studied (Fig. S4).

An appropriate temperature could improve enzyme activity and substrate solubility without resulting in enzyme inactivation. According to previously reported data, the PLD-catalyzed reaction was conducted at different temperatures (30–45 °C) [13]. As shown in Fig. S4B, the optimum reaction temperature is 35 °C, and used for the subsequent experiments.

It has been widely reported that an appropriate volume ratio and substrate molar ratio may ensure a high PS conversion rate, and decrease the unwanted hydrolysis of PC [12]. Therefore, the conversion reactions were performed at volume ratios of 1:2, 1:1, 2:1, 3:1, 4:1, and 5:1 (organic phase: aqueous phase) and substrate molar ratios of 15:1, 10:1, 5:1, 2.5:1 and 1:1 (L-Ser: PC). As shown in Fig. S4C and Fig. S4D, the conversion rate increased with the increased volume ratios between the organic and aqueous phase, and the conversion rate of 46.06% and PS concentration of 7.37 g/L was obtained at a volume ratio of 3:1. The conversion rate decreased with as organic phase further increase in the solution. On the other hand, an increment in the substrate molar ratio from 1 to 10 led to an increase in conversion rate to 59.92%. There is no further increment in the conversion rate beyond that. Therefore, the appropriate volume ratio and substrate molar ratio could increase the PS conversion rate.

Consideration of the phenomenon of interfacial saturation by the PLD and enzyme cost, enzyme concentration may not have an important effect on the reaction [30]. In order to study this effect, the reaction was carried out with PLD concentration varying from 40 to 5 U/mL (Fig. S4F). It is worth noting that decreased the enzyme concentration (40 U/mL-10 U/mL) almost did not influence the conversion rate, the conversion rate remained above 59.35%, indicating that the minimum enzyme concentration was 10 U/mL.

With the optimal reaction conditions in hand (butyl acetate/enzyme solution 3:1, 40 U/mL PLD, 40 °C, 38 mg/mL lecithin, 80 mg/mL L-Ser, and a reaction time of 10 h), the conversion rate reached up to 61.61%, and the PS concentration was 21.07 g/L, reflecting space-time productivity of 2.11 g/L/h. Compared to previous reports of extracellular PLD expression, the results indicated that the PLD-catalyzed process showed strong potential for industrial applications (Table 2).

3.5. PLD-catalyzed production of PL derivates

Another current focus in PLD is to modify PLs to produce PL derivates [8]. In the presence of appropriate alcohols, the PLD-catalyzed transphosphatidylation reaction is capable of modifying PLs by exchanging the hydrophilic head groups. To investigate the ability of enzymatic modification of PLs, we attempted reactions using several alcohols and DPPC (Fig. 6). The candidate alcohols contained two carbohydrates (glucose and mannitol), two carbohydrate derivates (glucosamine and N-Acetyl-d-glucosamine), and four vitamins (ascorbic acid, pyridoxine hydrochloride, pantothenic acid calcium salt, and thiamine chloride).

Fig. 6.

PLD-catalyzed PL derivates synthesis. Degree of synthesis of PL derivates was roughly measured by LC-MS analysis. High (yield >40%), low (10–40%), poor (<10%) and none (not detected).

After the PLD-catalyzed reaction, a new peak was detected in the HPLC analysis for all samples except trehalose and cellobiose (Fig. S6A). Next, MS was employed to further investigate the product, and the results are coincided with the predicted molecular formula of PL derivates (Fig. S6B and Table S5). The conversion rate of the PL derivates was estimated by HPLC analysis (Fig. 6). The present results indicated that PLD could catalyze the phosphatidylation of these monosaccharides, monosaccharide derivatives, and vitamins. When mannitol and thiamine chloride were used as substrate alcohols, the conversion rate of the PL derivates were 66.73% and 72.56%, which were higher than the conversion rate of PS. The conversion rate of phosphatidyl-glucose decreased to 45.91%. For the two carbohydrate derivates, the conversion rate of PLs were significantly lower than that of phosphatidyl-glucose. In conclusion, several carbohydrates and vitamins can be phosphatidylated by PLD. The phosphatidyl-mannitol, phosphatidyl-glucosamine, phosphatidyl-pyridoxine, phosphatidyl-pantothenic acid, and phosphatidyl-thiamine chloride were prepared by PLD for the first time. The present results provide beneficial data for the enzymatic modification of PLs.

4. Conclusions

In this study, a PLD gene was successfully secretory expressed in B. subtilis WB600. In order to increase the extracellular expression of PLD, versatile strategies, including signal peptide screening, promoter engineering, and fed-batch culture were combined to improve the extracellular activity of PLD to 119.22 U/mL which was 191.30 times more than its original extracellular activity. For the industrial production of PLD, this secretory expression system held a great deal of promise. Additionally, the PLD exhibited a suitable level of transphosphatidylation activity during PS and PL derivates synthesis. It could convert 61.61% of PC to PS within 10 h, yielding 21.07 g/L, and produce novel PL derivates, such as phosphatidyl-mannitol and phosphatidyl-thiamine chloride, with a higher conversion rate of 66.73% and 72.56%. This study would lay the foundation for the secretion expression of similar enzyme proteins in B. subtilis and pave the way for the enzymatic production of PS and enzymatic modification of PLs.

CRediT authorship contribution statement

Peng Zhang: Formal analysis, Methodology, Investigation, Conceptualization, Writing – original draft, Writing – review & editing. Jin-Song Gong: Data curation, Investigation, Conceptualization, Funding acquisition. Zhi-Hao Xie: Writing – original draft, Software, Formal analysis. Chang Su: Writing – review & editing, Investigation. Xiao-Mei Zhang: Software, Validation, Formal analysis. Zhi-Ming Rao: Investigation, Data curation, Formal analysis. Zheng-Hong Xu: Conceptualization, Project administration. Jin-Song Shi: Conceptualization, Supervision, Project administration, Funding acquisition, Writing – review & editing.

Declaration of competing interest

No potential conflict of interest was reported by the author(s).

Appendix A. Supplementary data

The following is the Supplementary data to this article.