Enzymes: chloramphenicol acetyltransferase (EC 22.214.171.124); pyruvate kinase (EC 126.96.36.199); lactate dehydrogenase (EC 188.8.131.52); purine nucleoside phosphorylase (EC 184.108.40.206); inorganic pyrophosphatase (EC 220.127.116.11); pyrophosphate-dependent fructose-6-phosphate kinase (EC 18.104.22.168); fructose-1,6-bisphosphate aldolase (EC 22.214.171.124), triosephosphate isomerase (EC 126.96.36.199); α-glycerophosphate dehydrogenase (EC 188.8.131.52).
Physiological and biochemical characteristics of poly γ-glutamate synthetase complex of Bacillus subtilis
Article first published online: 20 DEC 2001
European Journal of Biochemistry
Volume 268, Issue 20, pages 5321–5328, October 2001
How to Cite
Ashiuchi, M., Nawa, C., Kamei, T., Song, J.-J., Hong, S.-P., Sung, M.-H., Soda, K., Yagi, T. and Misono, H. (2001), Physiological and biochemical characteristics of poly γ-glutamate synthetase complex of Bacillus subtilis. European Journal of Biochemistry, 268: 5321–5328. doi: 10.1046/j.0014-2956.2001.02475.x
- Issue published online: 20 DEC 2001
- Article first published online: 20 DEC 2001
- (Received 12 June 2001, accepted 20 August 2001)
- nonribosomal polypeptide synthesis;
- poly γ-glutamate synthetase complex;
- membranous amide ligase;
- gene disruption;
- in vitro transcription;
†An enzymatic system for poly γ-glutamate (PGA) synthesis in Bacillus subtilis, the PgsBCA system, was investigated. The gene-disruption experiment showed that the enzymatic system was the sole machinery of PGA synthesis in B. subtilis. We succeeded in achieving the enzymatic synthesis of elongated PGAs with the cell membrane of the Escherichia coli clone producing PgsBCA in the presence of ATP and d-glutamate. The enzyme preparation solubilized from the membrane with 8 mm Chaps catalyzed ADP-forming ATP hydrolysis only in the presence of glutamate; the d-enantiomer was the best cosubstrate, followed by the l-enantiomer. Each component of the system, PgsB, PgsC, and PgsA, was translated in vitro and the glutamate-dependent ATPase reaction was kinetically analyzed. The PGA synthetase complex, PgsBCA, was suggested to be an atypical amide ligase.
chloramphenicol acetyltransferase gene
purine nucleoside phosphorylase
Poly γ-glutamate (PGA) is an unusual anionic polypeptide in that glutamate, mainly the d-enantiomer, is polymerized via a γ-amide linkage . Thus, it should be synthesized in a ribosome-independent manner. PGA is produced by bacteria [2–6] and a few other organisms such as hydra . PGA composed only of d-glutamate, d-PGA, was first isolated from the capsule of Bacillus anthracis. Recently, Hezayen et al.  showed that a hyperhalophilic archaeon, Natrialba aegyptiaca, produced d-PGA on a solid medium but poly γ-dl-glutamate (dl-PGA) in a liquid medium. Our recent study suggested that d-PGA is an extremely enlarged polymer with the molecular mass of over 10 000 kDa (M. Ashiuchi, T. Yamamoto and T. Kamei, unpublished data). Natto, a traditional Japanese fermented food made from soybeans by Bacillus subtilis (formerly Bacillus natto), contains abundant dl-PGA, which consists of 50–80% d- and 20–50% l-glutamate [9–13], as a main component of its extracellular viscous materials. Because the PGA depolymerase accumulates during the PGA production , PGAs with apparently varying molecular masses (10–1000 kDa) are obtained from the culture of B. subtilis (natto). Bacillus licheniformis produces stereochemically various dl-PGAs (d-isomer contents, 10 to 100%) [1,6,15]. Bacillus megaterium also synthesizes dl-PGA (d-isomer content, 50%) . l-PGA is produced by an alkalophilic Bacillus, although it is quite different from d-PGA and dl-PGA in molecular size.
PGA-producing B. anthracis and B. subtilis (natto)  usually harbor plasmids. The plasmid pXO2 of B. anthracis is responsible for encapsulation . capBCA were identified from the plasmid as the genes involved in PGA synthesis, but the gene products, CapBCA have not been characterized as a PGA synthetase complex. It was previously reported that genes related to PGA synthesis were mainly carried on plasmids found only in B. subtilis (natto) ; later, it was proposed that γ-glutamylpeptidase encoded on the plasmids synthesizes PGA . The plasmids, however, were shown not to encode any genes required for the production . We cloned the B. subtilis (natto) genes encoding a PgsBCA system, which synthesizes PGA from both enantiomers of glutamate, into Escherichia coli cells . The E. coli clone produced PGA extracellularly. The sequence analysis of the PgsBCA system revealed that it was an orthologue of the CapBCA complex . Unlike the capBCA genes of B. anthracis, the pgsBCA genes of B. subtilis were present on the chromosome .
In order to know whether PGA synthetic pathways other than the PgsBCA system, such as that by γ-glutamylpeptidase, coexist in B. subtilis, we attempted to construct a PgsBCA system-defective mutant from B. subtilis (natto), but have not succeeded yet because of low or no competence in the strains harboring plasmids. Although B. subtilis (natto), the starters of Japanese fermented food natto, are important as PGA producers, they are substantially unsuitable strains for genetic manipulation, unlike B. subtilis Marburg, a plasmid-free strain. Therefore, we isolated a plasmid-free PGA producer from the Korean fermented soybean paste, chungkookjang, which contains viscous materials and was made mainly by the action of Bacillus. Its physiological characteristics (unpublished data) and the finding that the sequence of 16S ribosomal DNA of the isolated strain shows 99.7% similarity to that of B. subtilis NCDO 1769 allowed this strain to be classified as B. subtilis[25a]. We designated the strain B. subtilis (chungkookjang) and used it throughout the experiments.
Here, we show evidence that the PgsBCA system is the sole machinery of PGA synthesis in B. subtilis (chungkookjang) and also describe some characteristics of the enzymatic system as a PGA synthetase complex.
Materials and methods
An E. coli cloning vector pKF3, all restriction enzymes, proteinase K, and isopropyl thio-β-d-galactoside (IPTG) were from TaKaRa Shuzo, Kyoto, Japan. A plasmid pBEST4F , carrying a chloramphenicol acetyltransferase gene (cat) cassette for Bacillus was a kind gift of M. Itaya, Mitsubishi Kasei Institute of Life Science, Japan. The PRISM DNA sequencing kit was from PerkinElmer. Mini-ProteanII Ready Gel J and a protein assay kit were from Bio-Rad. Vials of distilled HCl (6 m) and a Surfact-Pak detergent sampler: Tween-20, Tween-80, Triton X-100, Triton X-114, Nonidet P-40, Chaps, Brij-35, Brij-58, octyl β-glucoside, and octyl β-thioglucopyranoside were from Pierce. Pyruvate kinase (PK), lactate dehydrogenase (LDH), phosphoenolpyruvate, and RNase A were from Boehringer Mannheim. EnzChek phosphate (Pi) and pyrophosphate (PPi) assay kits were from Molecular Probe, Oregon, USA. Pyrophosphate reagent was from Sigma. A plasmid pET32a(+)  and Single Tube Protein System STP3/T7 were from Novagen. All other chemicals were of analytical grade.
Disruption of the pgsBCA genes in B. subtilis
Our strategy for the disruption of the pgsBCA genes in B. subtilis (chungkookjang) is shown in Fig. 1. First, the plasmid pPGS1  harboring the pgsBCA genes (3.0 kb) was digested with a restriction enzyme PstI; then the internal region (nucleotides 404–2050) of the genes was removed. The plasmid was self-ligated and digested with SmaI and HindIII. The resulting DNA fragment (1.4 kb) was ligated into the SmaI–HindIII site of pKF3. We named it pKPSΔ. The pBEST4F DNA was digested with PstI, and the 1.0-kb fragment containing the cat cassette was ligated into the PstI site of pKPSΔ. The constructed plasmid, newly termed pKPSd, was introduced into cells of B. subtilis (chungkookjang) by the competence method . Colonies grown on a plate of Luria–Bertani medium  containing 5 µg·mL−1 chloramphenicol were collected as pgs null mutants. The gene disruption was verified by long-PCR  of the chromosomal DNA of the disruptant with a sense primer PPGS-U (5′-CATAGTGATTCTATATACTGATGAATG-3′) and an antisense primer PPGS-D (5′-TTTGAATATGTTAAGAGACTTTTTAAT-3′), followed by DNA sequencing of the amplified fragment.
PGA production by the wild strain and the pgs null mutant of B. subtilis
Cells of B. subtilis (chungkookjang) were first inoculated into 5 mL of Luria–Bertani medium and cultivated at 30 °C overnight. The culture was transferred to 50 mL of a complete medium (pH 7.0) containing 0.5% polypeptone, 0.25% yeast extract, 0.5% NaCl, and 0.05% MgSO4·7H2O and incubated at 30 °C. When turbidity at 600 nm of the culture reached 2.1 (5 × 109 cells·mL−1; containing mainly cells in the early stationary phase), cells were harvested by centrifugation at 8000 g for 15 min, washed with 50 mL of 0.85% NaCl solution, and then suspended in 5 mL of 0.85% NaCl solution. The cell suspension was transferred into 45 mL of a basal medium (pH 7.0) composed of 5% sucrose, 1% (NH4)2SO4, 0.27% KH2PO4, 0.42% Na2HPO4, 0.05% NaCl, 0.5% MgSO4·7H2O, and an MS vitamin solution (JRH Bioscience, Kansas, USA). When the pgs null mutant was used, chloramphenicol was added to the media. Then, cells were aerobically incubated at 30 °C for 72 h. Isolation and assay of PGA were performed by the method described previously [9,24].
Preparation of cytosolic and extracellular enzymes and cell membranes of the E. coli clones
An E. coli clone harboring the pgsBCA genes of B. subtilis (chungkookjang), JM109/pTPGS1, and each clone harboring the pgsB gene only, the pgsC gene only, the pgsA gene only, the pgsBC genes, the pgsBA genes, or the pgsCA genes were prepared as described previously . Cells were grown in 10 mL of Luria–Bertani medium containing ampicillin at 30 °C. When the turbidity of the culture broth at 600 nm reached 1.2, 1 mm IPTG was added to the culture broth. The cultivation was continued at 30 °C for another 8 h. The culture was centrifuged at 8000 g for 30 min. The supernatant was brought to 70% ammonium sulfate saturation and then centrifuged at 12 000 g for 1 h. The resulting precipitate was dissolved in 1 mL of the standard buffer (0.1 m Tris/HCl buffer containing 10% glycerol, 0.2 m KCl, and 5 mm dithiothreitol, pH 8.0) and dialyzed against 1 L of the same buffer at 4 °C overnight. The dialyzed solution was concentrated to 100 µL with a Centricon-10 concentrator (Amicon) and used as the extracellular enzyme preparation. The harvested cells were suspended in 1 mL of the standard buffer and disrupted by sonication at 4 °C. The supernatant obtained by centrifugation at 12 000 g for 1 h was further centrifuged at 56 000 g for 30 min to separate soluble cytosolic enzymes and cell membranes. The supernatant was used as the cytosolic enzyme preparation. The precipitates were washed with 1 mL of the standard buffer and then suspended in 100 µL of the standard buffer supplemented with 10% glycerol. This was used as the suspension of the cell membrane.
Preparation of the membranous enzymes
The cell membrane was incubated with 8 mm Chaps at 25 °C for 1 h and centrifuged at 64 000 g for 2 h. The supernatant was used as the membranous enzyme preparation. The preparation showed little glutamate-independent ATPase activity (less than 4 U·mg−1), indicating that few ATPases embedded ubiquitously into bacterial membranes were solubilized under this condition.
Conditions for enzymatic synthesis of PGA
The reaction mixture (100 µL) consisted of 10 µmol of Tris/HCl buffer (pH 8.0), 5 µmol of d-glutamate, 5 µmol of ATP, 0.5 µmol of MgCl2, 20 µmol of KCl, 0.1 µmol of dithiothreitol, various concentrations of Chaps, and the membrane suspension (containing 40 µg of proteins). The reaction was performed at 30 °C for 24 h. Excess amounts of Chaps (final concentration, 0.1 m) were added to the mixture and incubated at 25 °C overnight. Proteinase K (100 µg) was then added to the mixture and incubated at 37 °C for 24 h. The solution was diluted to 1 mL with water and dialyzed twice against 1 L of water at 4 °C overnight. The dialyzed solution was lyophilized and dissolved in 10 µL of water. The PGA production was assessed by the method described previously .
Measurement of glutamate-dependent ATPase activity
Glutamate-dependent ATPase activity was measured by the following four methods.
ADP formed during the reaction was assayed by some modification of the coupled method with PK and LDH . The assay mixture (100 µL) contained 50 µL of solution A (0.2 m Tris/HCl buffer, pH 8.0, 10 mm ATP, 2 mm MgCl2, 0.4 m KCl, 1 mm dithiothreitol, and 0.2 mg·mL−1 BSA), 5 µmol of d- or l-glutamate, 0.4 µmol of phosphoenolpyruvate, 40 nmol of NADH, 5 U of each PK and LDH, and enzyme. The reaction was started by addition of glutamate at 25 °C. Glutamate was replaced with water in a blank. The ATPase activity was determined by following a decrease in the absorbance at 340 nm with a BioSpec-1600 spectrophotometer (Shimadzu, Kyoto, Japan). One unit was defined as the amount of enzyme that catalyzes the hydrolysis of 1 nmol of ATP (i.e. formation of 1 nmol of ADP) per min in the presence of glutamate.
Pi formed during the reaction was assayed with the EnzChek Pi assay kit including purine nucleoside phosphorylase (PNP) and 2-amino-6-mercapto-7-methylpurine ribonucleoside (MESG). The assay mixture (100 µL) contained 50 µL of solution A, 5 µmol of d-glutamate, 40 nmol of MESG, 0.2 U of PNP, and enzyme. The reaction conditions were the same as those of method A. The Pi yields were estimated from an increase in the absorbance at 360 nm with the spectrophotometer.
Production of PPi by the reaction was examined with the EnzChek PPi assay kit including PNP, an inorganic pyrophosphatase (IPP), and MESG. The assay mixture (100 µL) contained 50 µL of solution A, 5 µmol of d-glutamate, 40 nmol of MESG, 0.2 U of PNP, 2 mU of IPP, and enzyme. The reaction and detection conditions were the same as those of method B. Because IPP catalyzes the conversion of PPi into two equivalents of Pi, the PPi yields could be estimated by subtracting the values obtained by method B from those by method C.
The PPi yields were alternatively determined with pyrophosphate reagent composed of 45 mm imidazole/HCl (pH 7.4), 0.5 mm citrate, 0.1 mm EDTA, 2 mm MgCl2, 0.2 mm MnCl2, 0.02 mm CoCl2, 0.8 mm NADH, 12 mm fructose 6-phosphate, 5 mg·mL−1 BSA, 5 mg·mL−1 sugar stabilizer, 0.5 U·mL−1 pyrophosphate-dependent fructose-6-phosphate kinase, 7.5 U mL−1 fructose-1,6-bisphosphate aldolase, 50 U·mL−1 triosephosphate isomerase, and 5 U·mL−1α-glycerophosphate dehydrogenase . The assay mixture (150 µL) contained 50 µL of pyrophosphate reagent, 15 µmol of Tris/HCl buffer (pH 8.0), 7.5 µmol of d-glutamate, 0.75 µmol of ATP, 30 µmol of KCl, 0.15 µmol of dithiothreitol, and enzyme. The assay conditions were the same as those of method A.
Protein was determined using a protein assay kit with BSA as a standard.
In vitro synthesis of PgsB, PgsC, and PgsA
Template DNAs for in vitro synthesis of each component of PgsBCA system were constructed as follows. The pgsB gene was amplified by long-PCR with a sense primer PPGSB-NF2 (5′-GCGACTAGTCCATGGCCCGGGAAAAGCAATGTGGTTACTC ATTATAGCCTG-3′) and the antisense primer PPGSB-CR . The amplified DNA fragment (1.2 kb) was digested with SpeI and KpnI and then ligated into the XbaI–KpnI site of the pET32a(+) DNA. The plasmid was designated pEPGSb and used for the synthesis of PgsB. The 0.45-kb fragment of the pgsC gene was obtained by digestion of pPGC1  with KpnI and BamHI and ligated into the KpnI–BamHI site of pET32a(+). The plasmid was designated pEPGSc and used for the synthesis of PgsC. The pgsA gene was amplified by long-PCR with the sense primer PPGSA-NF  and an antisense primer PPGSA-CR2 (5′-GCGGCATGCGGCCGCTTATTTATTGGCGTTTACCGGTTCT TCCTG-3′). The amplified fragment (1.4 kb) was digested with BamHI and NotI and ligated into the BamHI–NotI site of pET32a(+). The plasmid was designated pEPGSa and used for the synthesis of PgsA. Each component of the PgsBCA system was synthesized by using STP3/T7. The in vitro transcription mixture (10 µL) consisted of 8 µL of the STP3 transcription mixture and 2 µL of a solution containing 1 µg of template DNA(s): 1 µg of pET32a(+), 0.3 µg of pEPGSb plus 0.7 µg of pET32a(+), 0.3 µg of pEPGSc plus 0.7 µg of pET32a(+), 0.3 µg of pEPGSa plus 0.7 µg of pET32a(+), 0.3 µg of each pEPGSb and pEPGSc plus 0.4 µg of pET32a(+), 0.3 µg of each pEPGSb and pEPGSa plus 0.4 µg of pET32a(+), 0.3 µg of each pEPGSc and pEPGSa plus 0.4 µg of pET32a(+), or 0.3 µg of each pEPGSb, pEPGSc, and pEPGSa plus 0.1 µg of pET32a(+) for the synthesis of the control (only TrxA, thioredoxin of E. coli), PgsB, PgsC, PgsA, PgsBC, PgsBA, PgsCA, or PgsBCA. The reaction was performed at 30 °C for 20 min. The translation mixture (50 µL) was composed of 10 µL of the resulting transcription mixture, 2 µL of 625 µm l-methionine, 30 µL of the STP3 translation mixture, and 24 µL of nuclease-free water. The reaction was carried out at 30 °C for 2 h. The STP3 control DNA (1 µg) carrying the E. coliβ-galactosidase gene was used as a positive control to verify whether the protein synthesis was successful  and to estimate the amount of newly synthesized proteins by the application of the β-galactosidase assay and kinetic parameters of the E. coli enzyme . About 1.7 ± 0.3 µg of proteins was newly synthesized in a tube under the conditions used. These reaction mixtures (50 µL) were immediately added to 50 µL of a dilution buffer composed of 0.2 m Tris/HCl buffer (pH 8.0), 1% BSA, 0.4 m KCl, and 10 mm dithiothreitol.
Results and discussion
Physiological significance of PgsBCA system in Bacillus PGA producers
Recently, pgsBCA of B. subtilis (natto) IFO 3336 were identified as the genes encoding the enzymatic system to synthesize PGA from glutamate, i.e. the PgsBCA system . To know if the system occurs ubiquitously in PGA producers, we first carried out the gene amplification by long-PCR using primers designed on the nucleotide sequence of the pgsBCA genes of B. subtilis IFO 3336 and each chromosomal DNA of B. subtilis IFO 3013, IFO 3335, IFO 13169, B. subtilis (natto) A, B, C, and D isolated from commercially available natto foods, B. licheniformis IFO 12107 (ATCC9945a), and B. subtilis (chungkookjang). The DNA fragments (about 3.0 kb) were found in the PCR products of all the strains tested.
The nucleotide sequence of the DNA fragment amplified from B. subtilis (chungkookjang) was the same as that of the pgsBCA genes of B. subtilis IFO 3336. To determine whether a PGA synthetic system other than the PgsBCA system operates in B. subtilis, the pgsBCA gene-disruptant of B. subtilis (chungkookjang) was constructed and the PGA productivity of this mutant was examined. The PGA production by the wild-type strain was strongly induced by glutamate or amino acids belonging to the glutamate family, whereas the pgs null mutant could not produce PGA under any conditions used (Table 1). Other phenotypes of the mutant, such as sporulation, were apparently the same as those of the wild-type strain. The data indicate that the PgsBCA system is the sole machinery of PGA synthesis in B. subtilis.
|Conditions||Concentrations (%)||Volumetric PGA yields (µg·mL−1)|
Biochemical characteristics of the B. subtilis PgsBCA system
The E. coli clone harboring the pgsBCA genes of B. subtilis (chungkookjang), E. coli JM109/pTPGS1, produced highly elongated PGA (about 1000 kDa), like the E. coli clone harboring those of B. subtilis IFO 3336 . We tried to synthesize PGA enzymatically by products of the E. coli clones. No PGA was synthesized by these cytosolic and extracellular enzyme preparations (data not shown). Then, we attempted the enzymatic synthesis of PGA by the cell membranes of the clones in the presence of 0.08, 0.8, 2, 4, 6, 8 or 24 mm Chaps (an amphiprotic cholate-based detergent). In the presence of 2 mm Chaps, PGAs with apparently high molecular masses, which were expected to be over 100 kDa on the SDS/PAGE analysis , were obtained by using the cell membrane of the E. coli JM109/pTPGS1 clone, ATP, and d-glutamate. The yield and the stereochemical composition of the PGA were determined by HPLC with the chiral carrier [9,24] after hydrolysis of PGA with 6 m HCl at 105 °C for 8 h in vacua. Under the conditions used, 13 ± 5 µg of PGA composed only of d-glutamate was produced, suggesting that the membrane contains no activity for the glutamate isomerization [6,34]. In contrast, the membranes of other E. coli clones harboring the pgsB gene only, the pgsC gene only, the pgsA gene only, the pgsBC genes, the pgsBA genes, and the pgsCA genes could not synthesize such elongated PGA.
Because PGA synthesis is considered to be a ligase reaction for glutamate, the coincident ATPase activity should be detected in the presence of glutamate. The cell membrane of the E. coli clone was incubated with various detergents and glutamate-dependent ATPase activities in these preparations were assayed (Fig. 2). As a result, the solubilization with Tween, Triton/Nonidet, and Brij-based nonionic detergents was unsuccessful. d-Glutamate-dependent ATPase activity (28 ± 12 U·mg−1) was found in the preparation solubilized with 8 mm Chaps. When the membrane was incubated with 2, 4, 6, 8, and 24 mm Chaps, yields of the activity were 0, 10, 40, 100, and 0%, respectively. The ATPase activity could be recovered with octyl β-glucoside and octyl β-thioglucopyranoside (nonionic sugar-based detergents), although the yields with them were only about 10 and 12.5% of that with 8 mm Chaps, respectively. Thus, Chaps was the best among detergents tested.
The enzymatic system acted exclusively on glutamate but not other amino acids including glutamine and aspartate. The d-glutamate-dependent ATPase activity (Fig. 2, black bars) was about fourfold higher than the l-glutamate-dependent ATPase activity (Fig. 2, white bars).
Gardner & Troy  suggested the occurrence of an extremely unstable PGA synthetase complex in B. licheniformis ATCC9945a. Their proposed reaction mechanism  was similar to those of multienzyme systems , such as the gramicidin S synthetase complex, which were accompanied by the cleavage of ATP into AMP. It is important for the elucidation of the mechanism of PGA synthesis by the PgsBCA system to know whether ADP or AMP is formed by the ATPase reaction. We measured the ATPase activity in the presence of d-glutamate by several methods. During the reaction, ADP and Pi formed stoichiometrically, whereas the production of PPi (corresponding to that of AMP) was not observed (Table 2).
|Reaction time (min)||Reaction product (nmol)|
|ADP||Pi||PPi Method C||PPi Method D|
In B. licheniformis[6,34], l-glutamate is first activated by γ-adenylation and then l-glutamyl-γ-adenylate is isomerized into the d-isomer as the precursor for PGA synthesis. In the presence of hydroxylamine, the glutamyl-γ-adenylate intermediate will be readily converted into glutamyl-γ-hydroxamate. Glutamate-γ-hydroxamate thus formed can be easily determined by the usual method . On the other hand, amide ligases catalyze formation of the amide linkage between two amino acids as substrates, coinciding with the hydrolysis of ATP into ADP. In these reactions, the phosphoryl group of ATP is transferred to a terminal carboxyl group of an acceptor (mainly oligopeptides) through substrate-dependent ATP hydrolysis. Then, an amide linkage is formed by nucleophilic attack of an amino group of a donor (mainly amino acids) into the phosphorylated carboxyl group [29,37–39]. Therefore, glutamyl-γ-hydroxamate cannot be obtained. Our experiments using hydroxylamine showed that no hydroxamate was produced during the reaction. These results suggest that the mechanism of PGA synthesis of B. subtilis is different from that of B. licheniformis. We propose a reaction mechanism of the PGA synthetase complex of B. subtilis, the PgsBCA system, as follows.
Kinetic analysis of each component of the PgsBCA system on glutamate-dependent ATPase reaction
To understand the enzymological features of the PgsBCA system, we synthesized each of the components, PgsB, PgsC, and PgsA, by application of the in vitro transcription/translation system. Glutamate-dependent ATPase activities of each preparation of control, PgsB only, PgsC only, PgsA only, PgsB plus PgsC (PgsBC), PgsB plus PgsA (PgsBA), PgsC plus PgsA (PgsCA), and PgsB plus PgsC plus PgsA (PgsBCA) were determined. The control was prepared with the vector pET32a(+) only. ATPase activity was found only in preparations of PgsBC and PgsBCA (Table 3). The cosubstrate specificity of both the in vitro translated PgsBC and PgsBCA was the same as that of the enzyme preparation solubilized from the membrane of the E. coli JM109/pTPGS1 clone: d-glutamate was the best, followed by l-glutamate. When 50 mm d-glutamate was used, the ATPase activity of PgsBCA was 3.3-fold higher than that of PgsBC; when 50 mm l-glutamate was used, that of PgsBCA was 2.1-fold higher than that of PgsBC. Apparent Km values of PgsBCA for d- and l-glutamate were estimated to be 4.2 and 15 mm, respectively, and those of PgsBC for d- and l-glutamate were 3.6 and 17 mm, respectively. There was little difference in affinity for glutamate between PgsBC and PgsBCA, and the turnover number was apparently increased in the presence of PgsA (Table 3).
|Proteins||Cosubstrate specific act. (U·mg−1)|
|Negative control (TrxA)||0||0|
|PgsBC||148 ± 28 (134)||54 ± 19|
|PgsBCA||495 ± 95 (171)||113 ± 48|
The ATPase activity of these preparations stored for 24 h at 4 °C was measured again by using 50 mm d-glutamate in order to examine the stability of PgsBC and PgsBCA. Five independent experiments showed that, during storage of PgsBCA, its activity was reduced to the level of PgsBC. This indicates that PgsBC is stable but PgsBCA is clearly unstable and suggests that PgsB and PgsC associate tightly to each other but the interaction between PgsBC and PgsA is comparatively loose.
Functions of the components PgsB, PgsC and PgsA
Experiment of PGA synthesis by the cell membrane of the E. coli clone producing the PGA synthetase complex, PgsBCA, and assays of its glutamate-dependent ATPase activity indicate that the enzyme complex is naturally associated with bacterial cell membranes. The enzyme complex is likely to be maintained at an active form through the interaction (association) with the membranes, because the PGA productivity of the enzyme complex solubilized with detergent was immediately lost. Its extreme instability, which resulted in the disappearance of PGA productivity, is probably due to the weak association between PgsBC and PgsA. PGA synthesis by the cell membrane of the E. coli clone producing the PgsBC proteins (therefore lacking the PgsA protein) was unsuccessful. These results, the data shown in Table 3, and the fact that PgsA consists mainly of hydrophobic (Fig. 3) and positively charged amino acid residues suggest that PgsA effectively removes the reaction products charged highly negatively (eventually PGA) from an active site of the PGA synthetase complex and that the enzyme complex is stably anchored into cell membranes by the action of PgsA. PgsA may function as a PGA transporter. It seems likely that this function of PgsA is important for the enlargement (elongation) of PGA. A structural feature commonly seen in amide ligases  was found in PgsB; the consensus sequence of the ATP-binding motif  lies on the residues 37–42 (GIRGKS) of the protein. As all the ligases identified so far have been cytosolic enzymes [29,37,38], this is the first example of membranous amide ligase. As shown in Fig. 3, a highly hydrophobic cluster is present at the N-terminal region of PgsB. This region may be important for the PgsB protein to interact with cell membranes and with other components, PgsA and PgsC. Orthologues of PgsC, the most hydrophobic component (Fig. 3;), have not been found from organisms other than the PGA producers. PgsC was indeed essential for PGA synthesis, but the function of PgsC remains unknown. To our knowledge, active sites of amide ligases are made with only a single peptide; thus, it was presumed that the essential ATP hydrolysis was catalyzed by the action of PgsB only. The results of kinetic analysis on each component, however, gave an unexpected conclusion about the role PgsC: ATP hydrolysis was catalyzed by PgsBC or PgsBCA but not by PgsB alone. This suggests that the active site of the complex is constituted mainly from PgsB and PgsC. We further observed that PgsBC and PgsBCA acted on both enantiomers of glutamate as the substrate. On the analyses of the substrate-dependent ATP hydrolysis for amide ligation, the ligases have been believed to show the strict stereospecificity for amino-acid substrates [29,37,38]. Thus, the PGA synthetase complex of B. subtilis is probably comformationally unique and atypical in enzymes of the ligase superfamily.
The elucidation of the ternary structure of the PGA synthetase complex and subsequently the detailed analyses on its catalysis at molecular levels will provide a deep insight not only into the specific properties of PGA synthesis and the general principles of peptide synthesis, but also into establishing a mass-production system for PGA, an environmentally and industrially useful biopolymer [13,41,42]. Moreover, since B. anthracis, the anthrax agent, produces PGA to escape from the attack by macrophages , inhibitors for the PGA synthetase complex possibly serve as pharmaceuticals against this important pathogen. We are currently investigating the structure of the complex.
The work was supported in part by a Grant-in-Aid for scientific research (no. 12760062) from the Japan Society for the Promotion of Science.
- 21942) The formation of extracellular d-glutamic acid polypeptide by Bacillus subtilis. J. Biol. Chem. 145, 415–424(
- 161956) Optical isomers of glutamic acid comprising bacterial glutamyl polypeptides. Med. J. Osaka University 6, 1043–1046(
- 291989) Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, , (