Construction and description of a constitutive plipastatin mono-producing Bacillus subtilis

Background Plipastatin is a potent Bacillus antimicrobial lipopeptide with the prospect to replace conventional antifungal chemicals for controlling plant pathogens. However, the application of this lipopeptide has so far been investigated in a few cases, principally because of the yield in low concentration and unknown regulation of biosynthesis pathways. B. subtilis synthesizes plipastatin by a non-ribosomal peptide synthetase encoded by the ppsABCDE operon. In this study, B. subtilis 3NA (a non-sporulation strain) was engineered to gain more insights about plipastatin mono-production. Results The 4-phosphopantetheinyl transferase Sfp posttranslationally converts non-ribosomal peptide synthetases from inactive apoforms into their active holoforms. In case of 3NA strain, sfp gene is inactive. Accordingly, the first step was an integration of a repaired sfp version in 3NA to construct strain BMV9. Subsequently, plipastatin production was doubled after integration of a fully expressed degQ version from B. subtilis DSM10T strain (strain BMV10), ensuring stimulation of DegU-P regulatory pathway that positively controls the ppsABSDE operon. Moreover, markerless substitution of the comparably weak native plipastatin promoter (Ppps) against the strong constitutive promoter Pveg led to approximately fivefold enhancement of plipastatin production in BMV11 compared to BMV9. Intriguingly, combination of both repaired degQ expression and promoter exchange (Ppps::Pveg) did not increase the plipastatin yield. Afterwards, deletion of surfactin (srfAA-AD) operon by the retaining the regulatory comS which is located within srfAB and is involved in natural competence development, resulted in the loss of plipastatin production in BMV9 and significantly decreased the plipastatin production of BMV11. We also observed that supplementation of ornithine as a precursor for plipastatin formation caused higher production of plipastatin in mono-producer strains, albeit with a modified pattern of plipastatin composition. Conclusions This study provides evidence that degQ stimulates the native plipastatin production. Moreover, a full plipastatin production requires surfactin synthetase or some of its components. Furthermore, as another conclusion of this study, results point towards ornithine provision being an indispensable constituent for a plipastatin mono-producer B. subtilis strain. Therefore, targeting the ornithine metabolic flux might be a promising strategy to further investigate and enhance plipastatin production by B. subtilis plipastatin mono-producer strains.

Background The preservation of food security is a global concern, especially with regard to increasing population. At the same time, there is a growing demand for organic agriculture products in both developed and developing countries. This high demand emphasizes the need for an effective, environmentally friendly alternative to chemical fertilizers, fungicides, insecticides, etc. Bacillus strains e.g. B. subtilis, B. amyloliquefaciens and B. velezensis are among the beneficial microorganisms known as effective cell factories that can produce many different secondary antimicrobial metabolites, including lipopeptides [1][2][3]. Plipastatin is a bioactive lipopeptide produced by Bacillus subtilis. In general, the main lipopeptides produced by Bacillus strains are classified in the three families of surfactin, iturin and fengycin [4,5], including plipastatin as a member of the fengycin family [6]. Lipopeptides mostly display additional biological activities besides their amphiphilic properties. It is reported that the fengycin group exhibits a broad antagonistic effect on various soil-borne and post-harvest fungal phytopathogens, specifically on filamentous fungi [7,8]. Besides antifungal activity, various authors [6][7][8][9][10] have frequently reported the antibacterial, antiviral and anticancer properties of fengycins. Additionally, these lipopeptides act also as the elicitor of induced systemic resistance in plants [11]. However, in contrast to the very well investigated applicability of surfactin, the application potential of fengycins have so far been investigated only in a few cases, principally based on poor bacterial productivity of fengycins and therefore, laborious and ineffective production approaches.
The structure of lipopeptides comprises a fatty acid connected to a peptide moiety [12]. The composition of amino acids in the circular peptide chain and length of the fatty acid residue provides a unique property for every lipopeptide [13]. Plipastatin consists of a decapeptide chain (L-Glu -D-Orn -L-Tyr -D-Thr -L-Glu -D-Ala -L-Pro -L-Gln -D-Tyr -L-Ile), 8 amino acids of which form a peptide ring and is linked to a 3-hydroxy fatty acid with 14-19 carbon atoms that may be saturated or unsaturated [14,15]. Although, the structure of fengycin is almost identical to plipastatin, the enantiomers of the amino acids in positions 3 and 9 (Tyr3 and Tyr9) are present in the l-and d-form in plipastatin, respectively, while in fengycin, they are in the reverse configuration [6,13,16]. Plipastatin and other lipopeptides are formed by step-by-step reactions of specific non-ribosomal peptide synthetases (NRPSs) [17]. The lipopeptide biosynthesis depends on the 4-phosphopantetheinyl transferase Sfp which converts the inactive apoforms of NRPSs to the active holoforms [18]. In case of B. subtilis 168, the genome has two large operons of srfAA-AD and ppsABCDE, which encode the subunits of NRPSs for surfactin and plipastatin production, respectively. However, due to a single base duplication in sfp gene, B. subtilis 168 is incapable to synthesize these lipopeptides [19][20][21]. Nevertheless, after repairing the sfp mutation, its lipopeptide production is restored [22,23].
Usually, Bacillus spp. encode for more than one lipopeptide synthetase, which are not synthesized simultaneously in the same growth phase showing the involvement of different regulators. More specifically, in the genome of bacteria that have the ppsAB-CDE operon, the srfAA-AD operon is always present as well [24,25]. However, while surfactin is being produced in the late exponential phase, the biosynthesis of plipastatin has been characterized in the stationary phase [25][26][27]. Lipopeptide synthetases are usually regulated by complex regulatory networks in a growthphase dependent manner. In the case of plipastatin, the ppsABCDE operon is repressed by AbrB, the transition state regulator, during exponential growth phase [28]. Moreover, expression of the pleiotropic regulator degQ gene has a positive effect on the synthesis of plipastatin showing the possible involvement of DegSU twocomponent system [20,21,29]. In B. subtilis 168, the -10 promoter region of degQ has a single base mutation that leads to low gene expression. Nevertheless, substitution of this mutation enables overexpression of the degQ gene and results in increased plipastatin biosynthesis [23,30,31].
In this study, we used B. subtilis 3NA as a model strain for genetic engineering of plipastatin production in order to construct a plipastatin mono-producer. This strain was previously described as a hybrid strain encoding genetic features from B. subtilis W23 and 168 [21]. In detail, the surfactin (srfAA-AD) operon was deleted in a way that their competence formation remained intact. Subsequently, the plipastatin was increased by promoter exchange and manipulation of the degQ expression. Finally, the effect of supplementing potentially critical amino acids on plipastatin formation was evaluated.

Plipastatin production in different B. subtilis strains
To construct a plipastatin mono-producer, two B. subtilis strains were compared for their potential in biosynthesis of plipastatin. This included the derivatives of the wellknown laboratory model strain B. subtilis 168 and strain 3NA. Plipastatin operon and its promoter region are identical in these two strains [21]. Nevertheless, strain 3NA is a sporulation-deficient strain caused by a frame shift mutation in spo0A which makes this strain suitable for fermentation [21]. These domesticated B. subtilis strains are known to have a mutation in their sfp gene disabling them to produce lipopeptides, such as surfactin and plipastatin [20,21]. Therefore, the sfp + derivates of strain 168 (JABs24) and strain 3NA (BMV9) were used to ensure lipopeptide production. Both strains were cultivated in mineral salt medium. Figure 1 shows the lipopeptide production, glucose consumption and optical density (OD 600 ) over 72 h of cultivation at 30 °C. The comparison of growth rates showed a faster cell growth of BMV9 strain compared to JABs24. However, optical density of BMV9 was drastically decreased without any stationary phase after glucose depletion. In contrast, JABs24 exhibited an entry into the stationary phase when 20% of glucose were metabolized. Both strains revealed surfactin production from the beginning of cultivation through to the early stage of stationary phase. Thus, surfactin quantities gradually increased until a peak was reached in the late exponential phase. In case of plipastatin, no detectable amounts could be measured up to the middle of exponential phase. The highest amount of plipastatin was produced by BMV9 in the stationary phase and was about 17 mg/L. Conversely, the production of plipastatin by JABs24 was lower than BMV9 (10 mg/L after 72 h). Based on these results strain BMV9 was considered for further strain engineering.

Verification of different putative bottlenecks concerning plipastatin production
As mentioned before, B. subtilis 3NA is a hybrid strain that exhibits genetic features from both strains W23 and 168 [21]. Likewise, the degQ gene and upstream promoter region are identical to that of strain 168. Therefore, a single base mutation in the degQ promoter region drastically decreases the corresponding gene expression [32]. Previous studies demonstrated that this circumstance has a negative effect on plipastatin production [31]. We hypothesized that a repair of degQ expression combined with the deletion of the competitive surfactin operon as well as an exchange of a weak plipastatin promoter against a constitutively active promoter lead to a highly efficient plipastatin mono-producer strain. However, before constructing the mutant strain that exhibits these three characteristics, BMV9 strain (3NA sfp +) was used to construct BMV10, BMV11 and BMV12, showing repaired degQ expression, plipastatin promoter exchange and surfactin elimination, respectively. Based on the observation that plipastatin level reached a plateau after 40 h of cultivation ( Fig. 1), the production outcomes of all following mutant strains were measured after 48 h of cultivation (Fig. 2).
As shown in Fig. 2, BMV10 which holds a repaired degQ gene, showed a twofold higher plipastatin production but a decrease in surfactin formation compared to the control strain (BMV9). Furthermore, the promoter exchange of native plipastatin promoter against the constitutive promoter region of the veg gene resulting in BMV11 strain, increased the plipastatin titer from 15 mg/L in control strain BMV9 to 70 mg/L. Interestingly, the surfactin formation was unaffected by the enhanced plipastatin production. In order to eliminate surfactin synthesis, the entire srfAA-AD operon was deleted and the regulatory comS gene which is co-expressed with srfAA-AD operon restored back to the genome with the native P srfA promoter. Consequently, the strains BMV12 (ΔsrfAA-srfAD with native P pps promoter) and BMV13 (ΔsrfAA-srfAD with P pps ::P veg promoter exchange) were constructed. It was observed that even though the deletion of the surfactin operon caused enhanced cell growth in both BMV12 and BMV13 strains, plipastatin production reduced (Fig. 2a). Plipastatin production was significantly decreased not only under native expression of plipastatin operon (BMV12) but also when P pps was exchanged against constitutive P veg promoter (BMV13).
In order to see the effect of combining the features of promoter exchange and repairing degQ gene expression, the strain BMV14 was constructed. Comparably to strain BMV9, the expression of degQ had a negative effect on surfactin production (Fig. 2b). Interestingly, no additive effect was observed on plipastatin production by BMV14 (Fig. 2a). In summary, we concluded that under the same conditions, BMV11 was able to produce the highest amounts of plipastatin compared to all other mutant strains constructed.

Impact of amino acid supplementation on plipastatin production
To verify the impact of amino acid precursors on plipastatin and surfactin formation (Additional file 1), seven different amino acids including glutamic acid, glutamine, isoleucine, alanine, threonine, proline and ornithine with a concentration of 30 mM were additionally supplemented in mineral salt medium. BMV9 (3NA sfp +) and BMV12 (3NA sfp + plipastatin mono-producer) were selected to evaluate the produced plipastatin under control of native P pps promoter, in the presence and absence of the surfactin operon.
As it is shown in Table 1, except for ornithine, a decrease in plipastatin titer was observed when supplementing BMV9 strain with the other six amino acids. Interestingly, in the plipastatin mono-producer BMV12 strain, supplementation of the ornithine led to a detectable plipastatin titer. Neither in control cultures (without amino acid supplementation) nor in the presence of other amino acids, no detectable production was observed. Subsequently, it was observed that cultivation of BMV13 (constitutive plipastatin mono-producer) in supplemented medium with ornithine let to enhance in plipastatin titer (about 10%). Furthermore, it is noteworthy to mention that the observed plipastatin chromatogram on silica HPTLC plate of BMV12 (plipastatin mono-producer) in the presence of ornithine exhibited a modified pattern compared to the plipastatin standard (B. subtilis plipastatin/fengycin standard, Lipofabrik france) and parental BMV9 strain (Fig. 3A). Although the detected signals in the chromatogram showed similar Rf values compared to the standard, not all standard peaks could be observed. Therefore, HPLC-MS analysis was performed for identification of plipastatin variants or homologs.

Mass spectrometry analyses of plipastatin variants produced by BMV12 with supplementation of ornithine
Plipastatin produced by B. subtilis is a mixture of several homologs based on the length of fatty acid chain and variants within peptide moiety [33]. The patterns of the HPTLC chromatograms of produced plipastatin are comparable between the B. subtilis plipastatin standard and the sample extracts of the parental BMV9 strain. In both cases, plipastatin splits into several peaks. This might indicate that several plipastatin variants were produced ( Fig. 3A [a, b]). Subsequently, plipastatin pattern of mono-producer strain BMV12 was analyzed by HPTLC and it appeared in one peak in the same Rf values (Fig. 3A [c]). Thus, it was assumed that deletion of srfA operon resulted in an accumulation of specific plipastatin variants whereas other subtypes were diminished. To get a perception about the different plipastatin variants present in B. subtilis, the high-purified plipastatin standard was analyzed by mass spectrometry (HPLC-ESI-MS). The results showed that a large variety of plipastatin subtypes with saturated and unsaturated fatty acid chains is present in the plipastatin standard (Additional file 2A). Mainly, A and B type plipastatin were detected in the LC-ESI-MS analysis, but also other plipastatin variants were observed. Some of them could be assigned to plipastatin C, D or S types [34] based on their diagnostic marker ions. In addition, other plipastatin variants were detected that were not described in previous studies in B. subtilis so far. Therefore, they were named in this study as W, U and Y. To get an overview about the differences in plipastatin variants produced by BMV12, shake flask cultivation was performed in mineral salt medium supplemented with ornithine. Samples were then harvested from the exponential and stationary phase (plipastatin HPTLC chromatograms were comparable in both growth phases) and also analyzed by LC-ESI-MS. Afterwards, the peak areas of identified plipastatin variants were used to determine ratios in between BMV12 samples and were compared to that of the plipastatin standard (Fig. 3B). Variants of the subtypes plipastatin A and plipastatin B were detected in both BMV12 samples and the standard, even though especially plipastatin B C 14 u (unsaturated fatty acid chain with 14 carbon atoms) and plipastatin B C 15 s (saturated fatty acid chain with 15 carbon atoms) showed higher relative abundance in the standard in contrast to BMV12 samples. In contrast, slightly higher abundances of plipastatin B C 16 u and plipastatin B C 18 s were detected in both samples of BMV12. Taken together, no major differences were detected between standard and BMV12 plipastatin extracts. Hence, the reason for the variation of the plipastatin pattern based on the deletion of surfactin operon could not be explained. Another interesting observation was the accumulation of different plipastatin variants (named in this study plipastatin Y, W and U) during the stationary phase. Due to the cultivation with mineral salt medium, a limitation of the amino acids during the stationary phase is very likely. Insofar, the accumulation of plipastatin Y, W and U could be due to the incorporation of different amino acids into the peptide residue of plipastatin.

Discussion
In this study, B. subtilis 3NA was engineered to construct a plipastatin mono-producer strain and to gain more insights about plipastatin production. Previously, it was shown that DegQ positively regulates plipastatin production [30]. In fact, DegQ stimulates autophosphorylation of DegS sensor kinase resulting in enhanced phosphotransfer to DegU response regulator [29,35]. As a result, the phosphorylated and activated DegU-P version causes higher expression of ppsABCDE operon and increases the plipastatin production [36]. In this study, repair of degQ expression (strain BMV10) ensured DegU-P activation, which doubled the production of plipastatin compared to parental BMV9 strain (3NA sfp +). In this context, previous results from Tsuge et al. [30] showed a tenfold higher plipastatin production when ppsABCDE, degQ and sfp were combined in a B. subtilis plasmid expression system. Furthermore, Wang et al [31]. described a decrease of plipastatin after in-frame mutagenesis of degQ in B. subtilis NCD-2. Afterwards, with respect to the relatively low expression level of ppsABCDE operon, an approximately fivefold higher plipastatin formation was achieved by the exchange of the native P pps promoter against constitutive P veg promoter (strain BMV11). Previous to this study, the effect of promoter exchange on the amount of plipastatin BMV12 staƟonary phase B MV12 exponenƟal phase Standard produced was reported only on a few cases. For instance, promoter exchange of native P pps promoter against native P fen from B. subtilis 21332 caused no plipastatin overproduction. In contrast, a tenfold higher plipastatin production was obtained when P fen from strain BBG21 (a spontaneous mutant of B. subtilis ATCC 21332) was integrated [37]. In the following step, it was observed that the combination of both repair of degQ expression and the P pps promoter exchange (BMV14), had no additional effect on the plipastatin titer compared to BMV11 (constitutive plipastatin producer). This is in contrast to the comparison of parental BMV9 and BMV10 when functional degQ expression increases plipastatin production about twofold. An explanation could be addressed by the DegQ mediated activation of DegU-P regulon, which causes in general the increase of secretory proteases [38,39]. These proteases could target plipastatin for degradation. In sum, this negative effect could be more noticeable in a constitutive pps operon expression (P veg ) which make a higher amount of DegQ not benefical. Another aim of this study was to construct a strain that ensures constitutive plipastatin mono-production. In a recent work, it was shown that deletion of srfAA significantly reduced the plipastatin production and on the other hand deletion of srfAC showed no effect [25]. The authors argued that probably srfAA has a regulatory effect on plipastatin production. Furthermore, they also observed another significant decrease in plipastatin production when pnp gene, which is responsible for biosynthesis of multifunctional polynucleotide phosphorylase (PNPase) [40] was deleted. The regulation of PNPase on plipastatin biosynthesis can happen through effect on comS expression. Subsequently, it was hypothesized that comS expression in an unknown complex pathway positively regulates plipastatin formation. In this study, we have deleted the whole srfAA-AD operon and retained comS with the native P srfA promoter (BMV12). In BMV12, even though comS was retained, plipastatin titer decreased to a non-detectable concentration confirming previously described study [25] that srfAA or in general, the expression of surfactin synthetase has a positive effect on plipastatin production. This observation was consistent after exchange of native P pps against P veg (BMV11 compared to BMV13). Likewise, in another study P pps promoter was exchanged against a strong P amyQ promoter from Bacillus amyloliquefaciens resulting in an increase in plipastatin production. However, after elimination of surfactin synthetase by deletion of srfAB, srfAC and srfAD genes, plipastatin production did not change [11]. Hence, in respect to our results and previously described studies [11,25] we conclude that, subunits of the surfactin synthetase have different impacts on plipastatin production.
Another rational factor in increasing the plipastatin production is the presence of a sufficient quantity of precursors in the cultivation medium. Therefore, we have attempted to improve plipastatin production by supply of seven amino acids of the plipastatin peptide chain in the mineral salt medium. Previously, these seven amino acids as the only nitrogen source were used in the mineral salt medium and compared with other nitrogen sources such as urea and ammonium carbonate [41]. Accordingly, urea was introduced as the best nitrogen source and alanine, followed by glutamic acid, were the best sources of nitrogen among the other amino acids. In this study, approximately the same concentration of amino acids were added to the medium besides urea as the main nitrogen source. Since, surfactin synthetase, assumably has a regulatory effect on plipastatin production, the supplementation of amino acids was examined in two strains of BMV9 (3NA sfp +) and BMV12 (3NA sfp + plipastatin mono-producer). Among all the amino acids used, ornithine was the only amino acid induced a detectable plipastatin production in mono-producer (BMV12), while it has not been measured in the other cultivations of BMV12. Interestingly, in the strain BMV9, the supplementation of different amino acids had different effects on final plipastatin titer. As it was shown in Fig. 1, the optical density of BMV9 directly decreased after glucose consumption whereas the plipastatin titer remained stable over time. Therefore, a comparison of the plipastatin production per biomass after 48 h cultivation is not reasonable. Accordingly, the results show that expect for ornithine, the additional supplementation of amino acids reduced in general the plipastatin titer in BMV9. This can be explained due to (de-) activation of stringent response in B. subtilis which occurs in amino acid limitation. Activation of stringent response results in enhanced provision of branched-chain amino acids [42,43]. Conversely, by addition of amino acids in the cultivation medium, the positive side effect of stringent response will be reduced. However, supplementation of ornithine in the medium had no negative effect and it was the only amino acid that enhanced the plipastatin produced by BMV12 (plipastatin mono-producer) to a detectable level. Therefore, it is concluded that ornithine is an indispensable constituent for a plipastatin mono-producer B. subtilis strain.

Conclusions
This study provides evidence that degQ stimulates the native plipastatin production. A significant decrease in plipastatin productivity after deletion of the surfactin operon in a constitutive plipastatin producer strain suggested that full plipastatin production requires the surfactin synthetase or some of its components. Nevertheless, the impact of surfactin synthetase existence on plipastatin formation is still unknown. In order to construct a plipastatin mono-producer strain suitable for cultivation in large quantities in a bioreactor, understanding the mutual impact between surfactin and plipastatin syntheses might help to increase the final plipastatin production. Furthermore, as another conclusion of this study, results point towards ornithine provision being an indispensable constituent for a plipastatin mono-producer B. subtilis strain. Therefore, targeting the ornithine metabolic flux might be a promising strategy to further investigate and enhance plipastatin production by B. subtilis plipastatin mono-producer strains.

Chemicals, materials and standard procedures
All chemicals were acquired from Carl Roth GmbH & Co. KG (Karlsruhe, Germany) if not, mentioned otherwise. Standard molecular techniques were carried out as described by Sambrook and Russell [44]. The desired DNA fragments were amplified in polymerase chain reactions using DNA Polymerase (Phusion High-Fidelity #M0530S, New England BioLabs, Frankfurt am Main, Germany). The PCRs were carried out on a PCR thermal cycler (prqSTAR 96X VWR GmbH, Darmstadt, Germany). Chromosomal DNA was purified with a ready to use kit (innuPREP Bacteria DNA Kit) and plasmid DNA was extracted with innuPREP Plasmid Mini Kit (Analytik Jena AG, Jena, Germany). After PCR reactions, amplified DNA fragments were extracted after agarose-based gel electrophoresis with QIAquick PCR & Gel Cleanup Kit, according to the manufactures' instruction. Restriction enzymes and alkaline phosphatase (#M0290) was purchased from New England BioLabs (Frankfurt am Main, Germany) and T4 DNA ligase were purchased from Thermo Fisher Scientific (Karlsruhe, Germany). All ligation reactions were performed overnight at 4 °C. For better efficiency of ligation, a PEG 8000 solution was added. Oligonucleotides were synthesized by Eurofins MWG (Ebersberg, Germany).

Strains, plasmids and transformation method
All strains and plasmids used in this study are shown in Table 2. Oligonucleotides used for construction of strains and plasmids are listed in Table 3. Escherichia coli JM109 was used for plasmid propagation and cloning. Transformation of E. coli strains were carried out according to the standard heat-shock method [45]. B. subtilis JABs32 strain, a sfp + derivate of B. subtilis 3NA, was used for mannose counterselection. Therefore, erythromycin resistance gene (erm) for manPA deletion was removed by the use of plasmid pJOE7644.2 resulting in BMV9 [46]. Strain BMV9 was used as parental strain for construction of further mutant strains. Transformation of natural competent B. subtilis strains was performed according to the "Paris method" [47]. Depending on the selection marker, the transformants were selected on LB agar supplemented with ampicillin (100 µg/mL), spectinomycin (100 µg/mL) or erythromycin (10 µg/mL for E. coli and 5 µg/mL for B. subtilis). All plates were incubated at 37 °C.

Construction of plasmids for strain engineering
For markerless promoter exchange, LFH-PCR method was used [53]. Accordingly, upstream and downstream wild-type sequence of pps promoter region was fused with veg promoter. After ligation into Sma I digested pJOE4786.1 resulting in pMAV3, target sequence was isolated by Hind III digestion and was subsequently integrated into pJOE6743.1 (results in pMAV4). Afterwards, plasmid pMAV4 was transformed into strain BMV9 followed by the protocol described before [51]. For the integration of the degQ gene including promoter region (+ 510 bp) and terminator structure from B. subtilis DSM10 T , the primer s1410 and s1409, containing Nde I and EcoR I restriction sites, were used. After restriction digestion, the degQ fragment was ligated into pKAM312 [46] resulting in pMAV5. After transformation of pMAV5 into parental strain BMV9 and other mutant strains, transformants were selected on LB agar plates containing spectinomycin. To ensure the correctness of plasmids and mutant strains, all constructs were confirmed by sequencing (Eurofins Genomics Germany GmbH, Ebersberg, Germany).

Deletion of surfactin operon and retain of comS gene
The principle of LFH-PCR was utilized to design a DNA fragment for deletion of srfA operon and simultaneously retain of comS gene. A fusion of upstream region of srfA operon with comS gene ensured a wild-type expression. For a simple strain selection, comS gene was additionally linked with erythromycin resistance cassette (erm) of pKAM312. An uncoupled erm gene expression was ensured by maintaining the natural P erm promoter region from pKAM312. Figure 4 illustrates the described strategy.

Cultivation in mineral salt medium
The mineral salt medium used was based on the fermentation medium of Willenbacher et al. [54] with slight changes. The composition of the final medium was: 20 g/L glucose For the first preculture, 10 mL LB medium (10 g/L tryptone, 5 g/L NaCl, 5 g/L yeast extract) was inoculated with 10 μL glycerol stock solution in a 100 mL  Additionally, the influence of potentially critical amino acids including glutamic acid, glutamine, isoleucine, alanine, proline and ornithine on plipastatin production was tested using mineral salt medium complemented with 30 mM of each amino acids.

Extraction of lipopeptides and HPTLC analysis
The cell-free supernatants were obtained by centrifugation at 4700 rpm and 4 °C and were used for extraction of lipopeptides following the method described before with slight changes [55]. In detail, 2 mL of cell-free supernatant was mixed 3 times with 1 mL 1-butanol 95% (v/v) by vortexing for 1 min, followed by 5 min centrifugation at 3000 rpm to separate organic phase. After complete evaporation of butanol phases (RVC2-25 Cdplus, Martin Christ Gefriertrocknungsanlagen GmbH, Osterode am Harz, Germany) at 10 mbar and 60 °C, the remaining residues were dissolved in 2 mL methanol. To quantify surfactin and plipastatin production, these methanolic fractions were separated by HPTLC (CAMAG, Muttenz, Switzerland) as described previously [56].

Structural analysis of plipastatin variants by Mass spectrometry
LC-MS analysis of plipastatin was performed on a 1290 UHPLC system (Agilent, Waldbronn, Germany) coupled to a Q-Exactive Plus Orbitrap mass spectrometer equipped with a heated electrospray ionization (HESI) source (Thermo Fisher Scientific, Bremen, Germany). Analyte separation was achieved by a Waters ACQUITY CSH C18 column (1.7 μm, 2.1 μm × 150 mm). The column temperature was maintained at 40 °C. Samples were dissolved in methanol and 5 µl of each sample was injected. Mobile phase A was 0.1% formic acid in water (v/v), and mobile phase B 0.1% formic acid in acetonitrile (v/v). A constant flow rate of 0.3 mL/min was used and the gradient elution was performed as follows: 0 -15% B from 0 to 15 min, 15-75% B from 15 to 29 min, 75-100% B from 29 to 32 min, isocratic at 100% B from 32 to 36 min, the system was returned to initial conditions from 100% B to 0% B from 36 to 37 min. The HESI source was operated both in positive and negative mode, with a capillary voltage of 3.90 kV and an ion transfer capillary temperature of 350 °C. The sweep Fig. 4 Schematic illustration of the construction of PCR fragment in order to delete surfactin operon and retain back comS with the native P srfA promoter using LFH-PCR method gas and auxiliary pressure rates were set to 35 and 10, respectively. The S-Lens RF level was 50%, and the auxiliary gas heater temperature was 150 °C. The temperature of ion transfer capillary, spray voltage, sheath gas flow rate, auxiliary gas flow rate and S-lens RF level were set to 325 °C, 3.5 kV, 60, 30 and 55, respectively. The Q-Exactive Plus mass spectrometer was calibrated externally in positive and negative ion mode using the manufacturer's calibration solutions (Pierce/Thermo Fisher, Germany). Mass spectra were acquired in MS mode within the mass range of 600-1800 m/z at a resolution of 70,000 FWHM using an Automatic Gain Control (AGC) target of 1.0 × 10 6 of and 100 ms maximum ion injection time. Data dependent MS/MS spectra in a mass range of 200-2000 m/z were generated for the five most abundant precursor ions with a resolution of 17,500 FWHM using an Automatic Gain Control (AGC) target of 5.0 × 10 4 of and 64 ms maximum ion injection time and a stepped collision energy of 20, 60 and 150. Xcalibur ™ software version 4.0.27 (Thermo Fisher Scientific, San Jose, USA) was used for data acquisition and data analysis.