INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Bacillus strains produce many kinds of bioactive peptides as secondary metabolites. Some of them are synthesized nonribosomally by a large multifunctional enzyme complex. Among them, surfactin (4), tyrocidine (21), gramicidin S (42), and bacitracin (15) are well characterized at the genetic level. Surfactin is a lipopeptide produced by Bacillus subtilis (14). Genetic studies of surfactin biosynthesis were performed extensively following the transfer of a genetic locus responsible for surfactin production to strain JH642, a derivative of B. subtilis 168 (24).

The B. subtilis genome project determined the DNA sequence of strain 168 and revealed that there are two large operons which encode nonribosomal peptide synthetases (17). The surfactin operon is located between 32 and 35°. The other operon, located between 167 and 171° (pps operon), was thought to be the fengycin operon, because significant homology was observed between the fengycin synthetase gene of fengycin-producing B. subtilis F29-3 (2) and the operon from strain 168 (2, 38, 39). Fengycin is a lipopeptide fungicide which consists of almost the same kind of amino acids and -hydroxy fatty acids as plipastatin (36, 45). However, there is no direct evidence which indicates a correlation between the operon of strain 168 and the production of fengycin at the product level.

Strain MI113 is a derivative of strain 168 which was generated by the transformation of strain RM125 (leuB8 arg-15 hsdRM) with BD224 (trpC2 thr-5 recA4) DNA (13, 36a) and does not show any clear fungicidal effect in vitro. In hybridization experiments performed by Chen et al. (2) on MI113 chromosomal DNA with the fengycin synthase gene of F29-3 as a probe, no hybridization signal was detected. From these results, it was determined that there was no fengycin operon in strain MI113 (2).

We isolated B. subtilis YB8, which suppresses the growth of phytopathogenic fungi in vitro (32). The suppressive effect of strain YB8 is mainly due to production of the antifungal lipopeptide antibiotic plipastatin (41, 46). Plipastatin was originally isolated from Bacillus cereus BMG302-fF67 as an inhibitor of phospholipase A₂ (44).

The structure of plipastatin is as follows:

where R is a -hydroxy fatty acid (26-28). We discovered that strain YB8 produces surfactin as well as plipastatin (40) and cloned and characterized the gene lpa-8 (sfp [23]), which is required for the production of both plipastatin and surfactin in strain YB8 (41). The gene lpa-8 encodes 4'-phosphopantetheinyl transferase, which converts inactive apoenzyme peptide synthetases to their active holoenzyme forms by posttranslational transfer of the 4'-phosphopantetheinyl moiety of coenzyme A to the synthetases (19). Strains 168 and MI113 have an inactive sfp allele (sfp⁰), which is why they cannot produce surfactin even though they have the intact surfactin operon. Although lpa-8 is essential for the production of two lipopeptides in YB8, the introduction of lpa-8 into MI113 or 168 induced only surfactin production, and in our previous study, plipastatin was not detectable either by high-performance liquid chromatography (HPLC) or by an assay for antifungal activity in vitro (41).

In this study, MI113 was converted into a coproducer of plipastatin and surfactin by transformation with YB8 chromosomal DNA. We applied transposon mutagenesis to the resultant transformant, strain 406, and determined that the pps operon and lpa-8 are both essential for plipastatin production.

To prove directly that the pps operon in strain 168 encodes plipastatin synthetases, we improved the HPLC system so as to enable the detection of a trace amount of plipastatin production from strain 168 supplied with lpa-8. This is the first report showing that although the pps operon in strain 168 is still active, this strain cannot produce plipastatin because of the sfp⁰ mutation.

We had previously cloned the degQ gene of YB8 (designated degQ_YB8) as an enhancer of extracellular protease production in strain 168. degQ is a pleiotropic regulatory gene which controls the production of degradative enzymes, an intracellular protease and several secreted enzymes (levansucrase, alkaline proteases and metalloproteases, -amylase, -glucanase, and xylanase) (16, 22). When we ablated the gene degQ_YB8 in strain YB8, plipastatin production in YB8 was severely reduced but no significant change was observed in surfactin production, indicating that degQ_YB8 is necessary for plipastatin production in YB8. Therefore, the role of degQ_YB8 in the production of plipastatin was investigated in strain 168 expressing plasmid-borne lpa-8.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Strains and media. The B. subtilis strains and plasmids used in this study are listed in Table 1. Plasmid pHV1249 was obtained from the Bacillus Genetic Stock Center of Ohio State University, and plasmids pNEXT24, pNEXT44, and pNEXT24A were obtained from M. Itaya. Low-salt Luria-Bertani (LB) medium contained (per liter) 10 g of Polypeptone (Nippon Pharmaceutical Co. Ltd., Tokyo, Japan), 5 g of yeast extract, and 5 g of NaCl and was adjusted to pH 7.2. ACS medium (45) containing (per liter) 100 g of sucrose, 11.7 g of citric acid, 4 g of Na₂SO₄, 5 g of yeast extract, 4.2 g of (NH₄)₂HPO₄, 0.76 g of KCl, 0.420 g of MgCl₂ · 6H₂O, 10.4 mg of ZnCl₂, 24.5 mg of FeCl₃ · 6H₂O, and 18.1 mg of MnCl₂ · 4H₂O was adjusted to pH 6.9 with NH₄OH and was used in the production of plipastatin. When necessary, antibiotics were added at the following concentrations: ampicillin, 50 µg/ml; chloramphenicol, 5 µg/ml; erythromycin, 10 µg/ml; tetracycline, 20 µg/ml; and neomycin, 20 µg/ml.

Purification of plipastatins. The purification of plipastatin was carried out by using a modification of a method developed previously (41). B. subtilis 406 was cultivated in 200 ml of ACS medium in a 500-ml shake flask for 2 days at 30°C. The 60 liters of culture collected from 300 flasks was adjusted to pH 2.0 with HCl, and the acid precipitate was collected and extracted with 3 liters of 95% ethanol. The extract was concentrated under reduced pressure and extracted with a solution of butanol and H₂O (1 liter:1 liter). The butanol phase was recovered, and 3 volumes (3.3 liters) of hexane were added. After separation of the suspension, the aqueous phase was recovered and 300 ml of butanol-hexane (1:1 [vol/vol]) was added. After centrifugation (8,000 × g for 10 min), the supernatant was concentrated to produce a brown crude powder, which was then dissolved in 12 ml of propanol. This solution was loaded on a propanol-filled column of Silica Gel 60 (Merck) (×20 [wt/wt] powder) and was successively eluted with 1 column volume of propanol, 2 column volumes of 90% propanol, and 2.5 column volumes of 80% propanol. The 80% propanol eluate was further purified by reversed-phase preparative HPLC with a Prep-ODS column (GL Sciences, Tokyo, Japan; 2-cm diameter by 25 cm at a flow rate of 10 ml/min with detection at 205 nm) with a mixture of acetonitrile and 1 mM trifluoroacetic acid (1:1 [vol/vol]). The main peak was concentrated and subjected to the same HPLC system with a mixture of acetonitrile and 10 mM ammonium acetate (1:1 [vol/vol]). Two main fractions, called fractions A and B, were collected. Fraction A was further separated into peaks 1 and 2 by using the same HPLC system with a mixture of acetonitrile and acetate buffer (2% potassium acetate plus 6% acetic acid) (1:1 [vol/vol]). Fraction B was separated into peaks 3 and 4 by using the same HPLC system with a mixture of 40 mM ammonium acetate and acetonitrile (1:1 [vol/vol]). All of these fractions were passed through Sephadex LH-20 columns with 80% methanol to remove the salts.

HPLC analysis of plipastatin. Plipastatin was extracted as follows. After 2 days' cultivation of a B. subtilis strain in 40 ml of ACS medium, the culture was acidified to pH 2.0 with 12 N HCl. Then the precipitate was collected by centrifugation and was extracted with ethanol. The extracted solution was filtered through a 0.2-µm-pore-size polytetrafluoroethylene membrane (Advantec, Tokyo, Japan).

Mass spectrometry and ¹H NMR. Collisionally activated dissociation (CAD) mass spectra of protonated molecular ions [M+H]⁺ by fast atom bombardment (FAB) ionization were obtained by the B/E linked scanning method with a JMS-700 MStation (JEOL Ltd., Tokyo, Japan) under the following conditions: 6 keV of xenon for impacting particles, glycerol plus m-nitrobenzylalcohol for the matrix, 0.1% trifluoroacetic acid for the solvent, and helium for the collision gas. ¹H nuclear magnetic resonance (NMR) spectra of purified substances were recorded on a JNM-GX500 NMR spectrometer (JEOL Ltd.) at 500 MHz and 55°C in methanol-d₄.

Transformation and DNA manipulation. The preparation of competent cells, transformation of Escherichia coli and B. subtilis, and routine DNA manipulations were performed by the method of Anagnostopoulos and Spizizen (1) with slight modifications, as described previously (41).

Transposon mutagenesis. Transposon mini-Tn10 mutagenesis was carried out as described by Petit et al. (31). The transposon-carrying plasmid pHV1249 was introduced into B. subtilis 406 by the competent cell method (41), and the resultant chloramphenicol-resistant (Cm^r) colonies were inoculated into LB medium containing chloramphenicol and incubated at 30°C. When the culture reached the logarithmic phase (optical density at 660 nm of 0.6 to 0.8), it was plated on LB agar medium containing chloramphenicol, followed by incubation at 51°C. Thermoresistant Cm^r colonies were regarded as mini-Tn10 insertional mutants.

Southern blot analysis. For NotI restriction analysis, DNA was purified in agarose plugs as described by Itaya and Tanaka (12). DNA fragments were separated by using a hand-made contour-clamped homogeneous electric field pulsed-field gel electrophoresis (PFGE) apparatus (3) with a 1% agarose gel (Sigma; type II) with 0.5× TBE (1× TBE is 89 mM Tris-89 mM boric acid-2 mM EDTA) at 5 V/cm for 20 h at 13°C with a switching ramp of 6 to 15 s, which was designed to optimally separate DNA fragments in the range from 9 to 300 kb. Concatemers of  phage (New England Biolabs, Inc.) were used as standard molecular weight markers. The digoxigenin DNA labeling and detection kit (Boehringer, Mannheim, Germany) was used for hybridization and detection according to the methods described by the manufacturer, and disodium 3-(4-methoxyspiro {1,2-dioxetane-3,2'-(5'-chloro)tricyclo[3.3.1.1^3,7]decan}-4-yl) phenyl phosphate (CSPD; Tropix, Inc., Bedford, Mass.) was used as a chemiluminescent substrate of alkaline phosphatase. Prehybridization was carried out for 1 h at 42°C in a solution containing 5× SSC (1× SSC is 0.15 M sodium chloride plus 0.015 M sodium citrate [pH 7.0]), 50% formamide, 50 mM sodium phosphate buffer (pH 7.0), 7% sodium dodecyl phosphate, 2% skim milk, 0.1% lauroylsarcosine, and 50 µg of fish sperm DNA per ml. Hybridization was performed overnight at 42°C in a prehybridization solution containing a denatured digoxigenin-labeled DNA probe. Filters were washed twice for 15 min at 68°C in 0.1× SSC containing 0.1% sodium dodecyl sulfate.

Construction of degQ-related mutants. The 0.8-kb EcoRV fragment containing degQ of YB8 (degQ_YB8) was cloned into the HincII site of pUC19, creating plasmid pUC19HP1. The BamHI fragment of the neomycin resistance (Nm^r) gene cassette from pUCN1 was inserted into the FbaI site in the degQ_YB8 coding region of pUC19HP1 which was prepared from strain JM110, generating plasmid pUCdegQ_YB8::Nm^r. Plasmid pUCdegQ_YB8::Nm^r was digested with Sse8387I and ligated into the PstI site of pE194, forming pEdegQ_YB8::Nm^r (Fig. 1).

View larger version (17K): [in this window] [in a new window]   FIG. 1.   Disruption of degQYB8 in strain YB8 by insertional plasmid mutagenesis. Details are described in Materials and Methods.

DNA sequencing of the degQ region. The degQ region of each B. subtilis strain was cloned by PCR with primers 5'-CCTATTGAGATTTGCGGTGTCACGCAGGAC-3' and 5'-CCCCCCTCCCATTCCATTTTACTAAATGGGA-3'. PCR was performed with TaKaRa LA PCR kit, version 2 (Takara Shuzo, Kyoto, Japan). PCR cycling conditions were as follows: 4 min at 94°C, 30 cycles of 20 s at 98°C, and 1 min at 68°C in a model TP480 TaKaRa PCR thermal cycler.

Nucleotide sequence accession number. The DNA sequence data from the degQ region of YB8 have been deposited in the DDBJ, EMBL, and GenBank nucleotide sequence databases under accession no. AB010576.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Conversion of strain MI113 into a plipastatin producer. In order to convert strain MI113 into a plipastatin producer, the transformation of MI113 with the chromosomal DNA of strain YB8 was necessary. However, due to the absence of proper markers, the detection of the transformants was not possible. Therefore, congression was done as follows. Cotransformation of MI113 with the chromosomal DNA of YB8 and a replicative plasmid DNA of pTB522 conferring tetracycline resistance (9) was performed and tetracycline-resistant transformants were selected. Colonies exhibiting growth suppression of the fungus F. oxysporum on a solid medium were selected as plipastatin producers. The plasmid was cured by overnight cultivation without selective pressure. A plipastatin-producing clone (designated strain 406) was selected and further characterized.

Transposon mutagenesis of strain 406. Strain 406 was transformed with the transposon-carrying plasmid pHV1249, and Tn10 insertion mutants were isolated by selection for Cm^r. Among 6,000 mini-Tn10 insertion mutants assayed for the loss of growth suppression of F. oxysporum in vitro, 39 mutants showed defective production of plipastatin on the plate assay. All these mutants except one (mutant strain 706) exhibited hemolytic activity on blood-agar plates, which was evidence for surfactin production. The location of the mini-Tn10 insertion in mutant strain 706, which produces neither plipastatin nor surfactin, was confirmed to be in lpa-8 (sfp), which is required for the production of both plipastatin and surfactin (41), because the production of both lipopeptides was recovered by the introduction of an intact lpa-8 gene carried on a plasmid.

Isolation and identification of the antifungal antibiotic produced by strain 406. We compared the HPLC peak pattern of the culture extract of strain 406 with that of a mutant strain with a transposon insertion. Figure 2 shows HPLC peak patterns of strain 703, which is representative of 29 Cm^r and Em^s mutants, and of wild-type strain 406. There were several peaks from the extract of strain 406 which could not be detected from the extract of strain 703. Assuming that these peaks from strain 406 were antifungal substances, we isolated and purified four substances corresponding to peaks 1 to 4 as described in Materials and Methods, and their structures were determined. Amino acid analysis of the hydrolysates of the purified substances, as well as chirality analysis, showed that the hydrolysate contained D-Ala(1) [D-Val(1) in peak 3], D-allo-Thr(1), L-Glx(3), L-Pro(1), L-Ile(1), L-Tyr(1), D-Tyr(1), and D-Orn(1) in molar ratios.

View larger version (19K): [in this window] [in a new window]   FIG. 2.   HPLC analysis of ethanol extracts of the acid precipitate from culture broths of B. subtilis 406 (left) and mutant strain 703 with a transposon inserted (right). Arrows indicate the peaks which were isolated and purified. SF and PL indicate the surfactin group and plipastatin group, respectively. OD, optical density.

View larger version (27K): [in this window] [in a new window]   FIG. 3.   (A) Calculated mass from the fragment ion peaks of plipastatin A1 (left) and its ring form opened between Ala and Pro (right). (B) FAB-linked scanning CAD mass spectrum of the protonated molecular ion peak ([M+H]+; m/z, 1463.8) of the substance purified from peak 1 as shown in Fig. 2.

Identification of NotI fragment with mini-Tn10 inserted. Chromosomal DNAs from strain 406 and its plipastatin-defective mutants were digested with restriction enzyme NotI and separated by PFGE (Fig. 4A). Following electrophoresis, the DNAs were transferred to a nylon membrane and were hybridized with a probe for mini-Tn10 to determine the location of the insertion of mini-Tn10 in the chromosome, as shown in Fig. 4B. Three sizes of NotI fragments, 13 kb (6 of 29 mutants), 24 kb (20 of 29 mutants), and 144 kb (3 of 29 mutants), were detected. The 144-kb fragment was judged to be equivalent to the NotI fragment designated 5N (12), because the value of 144 kb agreed well with the sum of 1 kb from the integrated mini-Tn10 and 143 kb from the 5N NotI fragment (12) of strain 168. In addition, the existence of 12-kb (12,313 bp) and 23-kb (22,859 bp) NotI fragments between 5N and 10N (122 kb) in strain 168 is known (12, 17), and we surmised that they correspond to the detected 13- and 24-kb fragments.

View larger version (53K): [in this window] [in a new window]   FIG. 4.   Southern blot analysis to identify the location of the gene cluster responsible for plipastatin production. Samples were digested with restriction enzyme NotI prior to PFGE. (A) Ethidium bromide (EtBr) staining. (B) Detection of mini-Tn10 insertion-containing NotI fragments by using the 1-kb EcoRV fragment of pHV1249 as a probe. (C) Identification of the pNEXT24 homologous fragment by using pNEXT24 as a probe. (D) Identification of the pNEXT44 homologous fragment by using pNEXT44 as a probe.

View larger version (13K): [in this window] [in a new window]   FIG. 5.   ORFs and NotI restriction map in the region between the two NotI fragments, 5N and 10N. This figure is compiled from data in references 38 and 39. Five ORFs which encode plipastatin synthetase are indicated. The putative amino acid-activating domain, racemase domain (dotted area), and thioesterase domain (black area) are also shown. Note that this amino acid-activating domain arrangement is different from that proposed by Tosato et al. (39). The cloned region in pNEXT24 and pNEXT44 and the insertions in the chromosome of strains 751, 752, and 753 are displayed above. + and  indicate plipastatin production positive and negative, respectively.

Inactivation of plipastatin synthetase gene with pNEXT24 and pNEXT44. The cloned fragments of strain 168 chromosomal DNA in pNEXT24 and pNEXT44 were sequenced. The sequences of the PstI fragment of pNEXT24 and the EcoRI fragment of pNEXT44 matched that previously determined by Tognoni et al. (38). These results implied that the large open reading frames (ORFs), believed to correspond to putative peptide synthetases in strain 168 (38, 39), encode plipastatin synthetases.

Production of plipastatin by strains carrying lpa-8. To determine whether strain 168(pC81AP), which expresses lpa-8 contained on a plasmid, can produce plipastatin, we constructed a ppsB disruption strain of strain 168(pC81AP) by a double crossover recombination of the linearized plasmid pNEXT24A. The newly constructed strain was named strain 790(pC81AP), and we compared the HPLC peak pattern of 168(pC81AP) with that of 790(pC81AP) (Fig. 6). They showed almost the same pattern except for the difference observed in the fractions taken at 18 to 22 min, when the plipastatin group is eluted (×16 area in Fig. 6). Although the concentration of plipastatin B1 in 168(pC81AP) was very low (approximately 1 ppm), it was reproducible.

View larger version (20K): [in this window] [in a new window]   FIG. 6.   HPLC analysis of ethanol extracts of the acid precipitate from the culture broths of B. subtilis 168(pC81AP), 790(pC81AP), and 791(pC81AP). SF and PL indicate the surfactin group and plipastatin group, respectively. ×16, 16-fold magnification of the optical density (OD) at 205 nm.

Role of degQ of YB8 in plipastatin production. We cloned a 0.8-kb EcoRV fragment from YB8 by shotgun cloning, which had enhanced extracellular protease activity when expressed in strain 168 in a previous experiment. The sequence of this fragment completely matched that of degQ of strain 168, but the upstream sequence of degQ has several base substitutions compared with that of strain 168 (Fig. 7). Especially, degQ of strain YB8 has a single base substitution of C for T at the promoter position 10 like degQ36, which has been known to lead to the hyperexpression of the degQ gene (47) (Fig. 7). For convenience, we will refer to degQ of YB8 as degQ_YB8 and to degQ of 168 as degQ⁰. To study the effect of degQ_YB8 on plipastatin production in YB8, we constructed a degQ_YB8 disruption mutant, YB8degQ_YB8::Nm^r. The plipastatin B1 productivity of YB8degQ_YB8::Nm^r was 5 ppm, which was about 10 times lower than that of YB8 (54 ppm). Introduction of a plasmid containing degQ_YB8 into YB8degQ_YB8::Nm^r restored plipastatin B1 production (48 ppm) to the level of YB8 (Fig. 8). These results indicate that the degQ_YB8 gene was responsible for the hyperproduction of plipastatin in YB8.

View larger version (53K): [in this window] [in a new window]   FIG. 7.   DNA sequence of the 0.8-kb EcoRV fragment containing degQYB8. The nucleotides of the strain 168 sequence that differ from that of strain YB8 are shown above. Boxes indicate 10 and 35 regions. , termination codon; >> <<, inverted repeat sequence.

View larger version (16K): [in this window] [in a new window]   FIG. 8.   HPLC analysis of ethanol extracts of the acid precipitate from the culture broths of B. subtilis YB8, YB8degQYB8, and YB8degQYB8(pCRV). SF and PL indicate the surfactin group and plipastatin group, respectively. OD, optical density.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Tognoni et al. (38) and Tosato et al. (39) determined the DNA sequence of this region of the chromosome of strain 168 and suggested the existence of large proteins which contain structural motifs associated with the subunits of previously characterized peptide synthetases. From these reports, there are five ORFs (from ppsA to ppsE) organized into 10 amino acid-activating domains, and the second, fourth, sixth, and ninth domains have putative racemase activity (Fig. 5). Furthermore, at some positions the first and fifth domains showed homology to the glutamic acid-activating domain of surfactin synthetase, and the second domain was remarkably similar to the ornithine-activating domain of gramicidin S synthetase. From these sequencing results, Tosato et al. proposed that this region is the fengycin operon (39). Fengycin is a lipopeptide antibiotic produced by B. subtilis F29-3, and it has the same amino acid composition as plipastatin except for one amino acid. One glutamic acid in fengycin is a glutamine in plipastatin, and it was concluded that fengycin contains three glutamic acid residues and no glutamine from elemental analysis (45). However, ambiguity still exists regarding the structure of fengycin. The fengycin operon had been cloned and sequenced from B. subtilis F29-3 (2, 20) and has significant homology with the putative peptide synthetase operon in strain 168 (20, 39).

In this study, we demonstrated that the antibiotic produced by strain 406 had the same sequence as plipastatin. Considering the biosynthesis of surfactin (4), we think that plipastatin is first synthesized in linear form, according to the amino acid-activating domain order on the pps operon, and then is cyclized by the formation of a lactone bond between the hydroxyl group of L-Tyr and the carboxyl group of L-Ile. The amino acid sequence of the linear form of the plipastatin precursor is thought to be as follows: -hydroxy fatty acid-L-Glu-D-Orn-L-Tyr-D-allo-Thr-L-Glu-D-Ala(Val)-L-Pro-L-Gln-D-Tyr-L-Ile. In the putative peptide synthetase as well, homology is observed among the first, fifth, and eighth amino acid-activating domains. The amino acid sequence of the first domain is 98% identical to that of the fifth. However, the amino acid sequence of the eighth domain is 49% identical to that of the first and fifth, indicating that the eighth domain activates glutamine but not glutamic acid. This peptide sequence and chiral pattern agree well with the domain arrangement of the putative peptide synthetase of strain 168. We propose the amino acid-activating domain assignment of the pps operon that is shown in Fig. 5. This assignment is not the same as that for the fengycin operon proposed by Tosato et al. (39). Recently, fenB, which is a fengycin synthetase gene and a homolog of ppsE, was cloned and sequenced. FenB overexpression revealed that FenB is responsible for isoleucine activation (20). This result agrees with our prediction of activation of isoleucine by PpsE. Indeed, since the ¹³C NMR spectrum of plipastatin resembles that of fengycin (data not shown), it is probable that fengycin has almost the same structure as plipastatin. From these results, we conclude that the gene cluster located between 167 and 171° is the plipastatin synthetase operon and propose that ppsA, ppsB, ppsC, ppsD, and ppsE be renamed pliA, pliB, pliC, pliD, and pliE, respectively, as shown in Fig. 5.

Strain 168 also has a peptide synthetase for surfactin biosynthesis (4) but could not produce surfactin because of a mutation in sfp, which encodes 4'-phosphopantetheinyl transferase (19). We have previously cloned lpa-8, which is identical to sfp from strain YB8, and determined that this gene was essential for the production of both surfactin and plipastatin in YB8 (41). The introduction of a plasmid containing lpa-8 into strain 168 resulted in both plipastatin and surfactin production, but plipastatin production was very poor (Fig. 6), suggesting that there was some other gene responsible for plipastatin production.

In our previous investigation, we had cloned the degQ_YB8 gene, which enhanced the protease activity in strain 168 (data not shown), and we think that degQ_YB8 is a hyperexpression type of degQ. When we introduced degQ_YB8 into strain 168(pC81AP), which has lpa-8 on the plasmid, the resultant strain 168 derivative 791(pC81AP), which possessed both lpa-8 and degQ_YB8, produced a detectable amount of plipastatin when analyzed by HPLC. This shows that degQ_YB8 is responsible for the enhancement of plipastatin production. This gene is not essential, because a degQ_YB8 disruption in YB8 reduced the plipastatin production by 1/10. A similar observation has been reported for degradative enzyme production (18, 30).

We also sequenced the degQ region of strains 406 and MI113 and confirmed that plipastatin-producing 406 has the degQ_YB8 promoter while the host strain MI113 has a degQ⁰ promoter like strain 168. This result indicates that the degQ⁰ region of strain 406 was exchanged with degQ_YB8 of YB8 by transformation. However, strain 786, which is a degQ_YB8-disrupted mutant derived from strain 406, did not exhibit a change in its plipastatin production (Table 2). To clarify whether degQ_YB8 of strain 406 is inactive or not, plasmid pCRV, which carries a copy of degQ_YB8, was introduced into strain 406, but no significant change in strain 406(pCRV) plipastatin production was observed. Although we cannot explain why, it would seem that there is another pathway which bypasses degQ in chimera strain 406. 

degQ expression is known to be controlled by DegS-DegU and ComP-ComA modulator-effector pairs (22). The srfA operon, which comprises the surfactin synthetase gene and comS, is also regulated by ComP-ComA (25, 33). However, competence development requires nonphosphorylated DegU, whereas degradative enzyme production needs phosphorylated DegU (5). In our previous study, plipastatin was produced by YB8 from late logarithmic phase to early stationary phase, whereas surfactin was produced from early to late logarithmic phase (40). This delay in plipastatin production may be due to the requirement for the phosphorylated form of DegU. This is the first report that relates degQ to the production of a metabolite other than degradative enzymes. We summarize the collation of genotypes and plipastatin-producing phenotypes in Table 2 and show the pedigree tree of the key strains used in this study in Fig. 9.

View larger version (15K): [in this window] [in a new window]   FIG. 9.   Pedigree tree of the key strains used in this study. degQ and lpa-8 genotypes are also indicated in boxes. Arrows show the direction of evolution. The construction of strain MI113 from strain 168 (dotted arrow) was done by several steps (13, 36a).

Recently, researchers have developed targeted replacements of amino acid-activating domains within the srfA operon with a variety of other amino acid-activating domains cloned from bacteria and fungi by PCR (35, 43) and succeeded in the production of peptides with modified amino acid sequences (34, 35). On the other hand, de Ferra et al. replaced the integral thioesterase type I domain within surfactin synthetase and demonstrated that the synthesis truncated lipopeptides (6). Because of its high specificity and efficiency of genetic transformation (7), strain 168 is a good candidate as a recipient for recombinant peptide synthetase genes. Therefore, the identification of a new peptide synthetase is significant because it activates unique amino acids in strain 168. In this respect, the presence of plipastatin synthetase will increase the production of a variety of modified peptides.