Synthesis of poly (γ-glutamic acid) and heterologous expression of pgsBCA genes

Article abstract of DOI:10.1016/j.molcatb.2010.07.014

The genes required for synthesis of poly (γ-glutamic acid) (γ-PGA) were cloned from Bacillus licheniformis NK-03, a strain isolated from fermented food, natto. There were three open reading frames pgsB, pgsC, pgsA in the cloned fragment, all of which were greatly similar with those from typical Bacillus subtilis strains. The alignment of deduced amino acid sequences showed that PgsC was the most conservative part in PgsBCA. Recombinant plasmid pXMJ19-PGS was constructed by a shuttle vector pXMJ19, and it was successfully transformed and expressed in the recombinant strains of Escherichia coli JM109 and Corynebacterium glutamicum ATCC13032, respectively. Expression of pgsBCA in C. glutamicum indicated that it could synthesize γ-PGA with a yield of 0.69 g/L and 97% proportion of l-glutamate monomer in the absence of glutamic acid. The results suggest that γ-PGA biosynthesis directly from glucose by genetic engineering is feasible and significant.

Full text of DOI:10.1016/j.molcatb.2010.07.014

Journal of Molecular Catalysis B: Enzymatic 67 (2010) 111–116
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Synthesis of poly (␥-glutamic acid) and heterologous expression of pgsBCA genes
Mingfeng Caoa, Cunjiang Songa,∗, Yinghong Jina,b, Li Liua,b, Jing Liua,b, Hui Xiea, Wenbin Guoa,
Shufang Wangb,∗∗
a Key Laboratory of Molecular Microbiology and Technology for Ministry of Education, College of Life Sciences, Nankai University, 94 Weijin Road, Tianjin 300071, PR China
b Key Laboratory of Bioactive Materials for Ministry of Education, Nankai University, Tianjin 300071, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 22 May 2010
Received in revised form 22 July 2010
Accepted 27 July 2010
Available online 4 August 2010
The genes required for synthesis of poly (␥-glutamic acid) (␥-PGA) were cloned from Bacillus licheniformis
NK-03, a strain isolated from fermented food, natto. There were three open reading frames pgsB, pgsC,
pgsA in the cloned fragment, all of which were greatly similar with those from typical Bacillus subtilis
strains. The alignment of deduced amino acid sequences showed that PgsC was the most conservative
part in PgsBCA. Recombinant plasmid pXMJ19-PGS was constructed by a shuttle vector pXMJ19, and it
was successfully transformed and expressed in the recombinant strains of Escherichia coli JM109 and
Corynebacterium glutamicum ATCC13032, respectively. Expression of pgsBCA in C. glutamicum indicated
that it could synthesize ␥-PGA with a yield of 0.69 g/L and 97% proportion of l-glutamate monomer in the
absence of glutamic acid. The results suggest that ␥-PGA biosynthesis directly from glucose by genetic
engineering is feasible and significant.
Keywords:
Poly (␥-glutamic acid)
Bacillus licheniformis
pgsBCA genes
Shuttle vector pXMJ19
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
According to the nutrient requirements, ␥-PGA producing
strains are divided into two types: one produces PGA depending
Poly (␥-glutamic acid) (␥-PGA) is an unusual macromolecular
anionic polypeptide, which consists of d- and l-glutamic acid units
connected by amide linkages between ␣-amino and ␥-carboxyl
groups [1]. The molecular weight of microbial ␥-PGA varies from
10k to 1000k and the stereo-chemical structure can be divided
into three types: a homopolymer of d-glutamic acid (␥-d-PGA), a
homopolymerof l-glutamicacid (␥-l-PGA), and copolymerlinedup
by random d-/l-glutamic acid (␥-dl-PGA) [2]. Because of the abun-
dant active sites of carboxyls and optically chiral center in gluta-
mate unit, it can be modified by crosslinked, derivative and chelate
reactions. ␥-PGA is water soluble, biodegradable, nontoxic and edi-
ble, which make it useful for various potential applications, such
as hydrogels, flocculants, thickeners, dispersants, cryoprotectants,
drug carriers, cosmetic, and food additives [3]. Since it was first
found that Bacillus anthracis could produce ␥-PGA as capsule com-
ponent and important pathogen causing virulence [4,5], several
Bacillus species have been reported to synthesize ␥-PGA including
B. subtilis, B. licheniformis and B. megaterium. In addition, Hezayen
classified a new archaeobacteria species Natrialba aegyptiaca, and
the strain could survive in extreme environmental stress (>20%
NaCl), attributing to the strong water binding capacity of PGA [6].
on the existence of l-glutamic acid in the medium, the other does
not [13]. The former occupies most of the known PGA producers,
such as B. subtilis chungkookjang [7], B. licheniformis ATCC9945A
[8,9], B. subtilis (natto) IFO3335 [10] and B. subtilis RKY3 [11]. The
later includes B. subtilis C1 [12], B. subtilis TAM-4 [13], B. licheni-
formis A35 [14] and B. licheniformis SAB-26 [15]. Although it seems
better to choose l-glutamic acid independent strains for the studies
on the mechanism of ␥-PGA formation and its industrial applica-
tions, little is known about these kinds of bacteria [16]. Since ␥-PGA
producers with high productivity are mostly l-glutamic acid depen-
dent bacteria, the l-glutamate is an important substrate for ␥-PGA
production. People are interested in whether the glutamic acid pro-
ducer such as Coryneform bacteria could synthesize ␥-PGA directly
from glucose or other sugars by expressing ␥-PGA synthetase genes.
Park and Sung et al. developed a method for producing ␥-PGA using
Coryne-bacteria. The result of the surface display by C. glutamicum
E12 harboring vector pMT-HCE-pgsBCA showed that Coryneform
bacteria could express ␥-PGA synthetase genes from Bacillus subtilis
and be served as a considerable host of gene display [17,18].
In this study, a new ␥-PGA producer, B. licheniformis NK-03,
was isolated from Okame natto. It could synthesize ␥-PGA with
high molecular weight Mw and high content of l-glutamate (l-
Glu). The ␥-PGA biosynthesis genes (pgsBCA) were cloned from B.
licheniformis NK-03 and introduced into E. coli JM109 and C. glu-
tamicum ATCC13032, respectively, by shuttle vector pXMJ19 with
lac promoter. The expression of the pgsBCA genes was carried out
in C. glutamicum ATCC13032 without adding l-glutamic acid in
∗
Corresponding author. Tel.: +86 22 23503866; fax: +86 22 23503866.
Corresponding author. Tel.: +86 22 23503753; fax: +86 22 23503753.
E-mail addresses: songcj@nankai.edu.cn (C. Song), wangshufang@nankai.edu.cn
∗∗
(S. Wang).
1381-1177/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcatb.2010.07.014

112
M. Cao et al. / Journal of Molecular Catalysis B: Enzymatic 67 (2010) 111–116
fermentation medium. The fraction ratios of l-isomer and d-isomer
of ␥-PGA monomer from recombinant strains were measured by
reversed phase high performance liquid chromatography (reversed
phase HPLC).
2.3. Sequencing and cloning of ꢀ-PGA biosynthesis genes
For the amplification of ␥-PGA biosynthesis genes (pgsBCA), PCR
was carried out using chromosomal DNA of NK-03 as a template.
A sense primer (5ꢀ-CCCCCAAGCTTCATAGTGATTCTATATACTGATG-
3ꢀ), and an antisense primer (5ꢀ-CCCCGGGATCCTTTGAATATGT
TAAGAGACTTTT-3ꢀ), which contained restriction sites of HindIII
and BamHI, respectively (shown by a solid underline) were
designed based on the sequences of B. subtilis IFO3336 pgsBCA genes
[23]. The amplified DNA fragment was sequenced by commercial
company. The alignments of pgsBCA genes and deduced amino acids
were carried out by DNAMAN Software (Lynnon Co., U.S.) with
other known sequences achieved from NCBI database, such as B.
subtilis (natto), B. subtilis IFO3336, B. subtilis 168 and B. licheniformis
ATCC14580.
Then, the fragment was digested with HindIII and BamHI
(TaKaRa Biotech Co., Ltd., Japan), and inserted into an 6.6 kb
size of Escherichia coli–Corynebacterium glutamicum shuttle vec-
tor pXMJ19 (Gifted from Chinese Academy of Sciences) containing
IPTG inducible lac protomer and chloramphenicol acetyl trans-
ferase gene (cat), so it can be screened by white-blue plaque and Cm
resistant. The resultant plasmid pXMJ19-PGS was introduced into
E. coli JM109 competent cells by the method previously reported
[24]. The positive clones were tested by the methods of colony PCR
and recombinant plasmid enzyme-digestion.
2. Materials and methods
2.1. Isolation and identification of bacterium
Okame natto, a sticky fermented food made from steamed soy-
beans, was bought from local market (Tianjin, China). It has been
popular in Japan for more than 400 years. Almost all the natto-
producing bacteria have the capability to produce PGA and have
been classified as B. subtilis or Bacillus species [19]. The inoculation
liquid was prepared by adding some Okame natto into 50 ml ster-
ilized isolation liquid media containing (g/l): l-glutamic acid, 15;
glucose, 50; KH2PO4, 2.7; Na2HPO4·12H2O, 4.2; MgSO4·7H2O, 0.5;
NaCl, 0.5 and biotin, 5 × 10−4, pH 6.4, which was entirely dispersed
with a magnetic stirrer at 4 ◦C for 30 min. Then, 0.1 ml liquid was
inoculated on a 2.0% (w/v) agar isolation plate and incubated at
35 ◦C until mucous colonies were obtained. The culture was stored
in 15% glycerol solution at −80 ◦C after it was inoculated and puri-
fied in Luria-Bertani media [20].
16S rRNA gene and BIOLOG MicroLogTM System Release 4.20
(BIOLOG Co. Ltd., U.S.) were employed to identify the highly
mucous strain. 16S rDNA was amplified with universal primers
as follows: 27F, 5ꢀ-AGAGTTTGATCM(M = C,A)TGGCTCAG-3ꢀ; 1492R,
5ꢀ-CGGY(Y = T, C)TACCTTGTTACGACTT-3ꢀ [21]. The PCR sequence
was aligned with published NCBI database sequences by BLAST pro-
gram, and then deposited with GenBank accession number. As a
microorganism begins to use the carbon sources in certain wells of
the MicroPlate, its respiration process reduces a tetrazolium redox
dye and those wells turn purple. The strain of NK-03 was chosen
on GP2 MicroPlate after distinguished by gram and endospore stain
methods, and cultivated for 4–6 h, then the characteristic pattern of
purple wells were read by MicroStation Reader. Furthermore, the
traditional taxonomic characterizations of the strain were investi-
gated towards Bergey’s Manual of Systematic Bacteriology [22].
The identified NK-03 was used to produce ␥-PGA and amplify
␥-PGA synthesis relative genes. Corynebacterium glutamicum
ATCC13032, a typical glutamate producing strain, was supplied by
Center of Industrial Culture Collection (CICC, Beijing, China) and
served as a host to express ␥-PGA synthetase genes.
2.4. Construction of recombinant C. glutamicum
The recombinant plasmid was transformed into C. glutamicum
as follows: the cells were inoculated into LB media and cultured
overnight. Then, 1% amount of suspension was inoculated into a
500 ml conical flask containing 50 ml Complete Media (CM) (g/l:
tryptone, 10; yeast extract, 10; glucose, 5; NaCl, 2.5; pH 7.2) at
30 ◦C. Subsequently, 11 ␮l Penicillin G (2500 U/ml) was added into
the media when the cells’ optical density at 600 nm (OD600) was
up to about 0.15. Cells were collected by centrifugation and sus-
pended in hypertonic TSMC buffer (10 mM MgCl2·6H2O, 30 mM
CaCl2·2H2O, 500 mM sodium succinate, 50 mM Tris–HCl, pH 7.5)
containing 1 mg/ml lysozyme until the OD600 reached 0.5–0.6. The
formation of protoplasts was observed by a microscope. After sev-
eral hours, protoplasts were washed three times with ice-cold 10%
(v/v) glycerol, and finally re-suspended in 1 ml 10% glycerol. The
recombinant plasmid pXMJ19-PGS was extracted from E. coli cells
by the alkaline lysis technique and added to 100 ␮l protoplasts with
5 ␮l volume. Then, the mixture was transferred to an electropo-
ration cuvette. After a single electric pulse at 0.8 kV/cm, the cell
suspension was immediately transferred into 1 ml Recover Media
(RM) (g/l: tryptone, 10; yeast extract, 5; glucose, 10; NaCl, 2.5;
MgCl2·6H2O, 4; sodium succinate, 135; pH 7.2) and incubated stati-
cally for 6 min at 46 ◦C. Subsequently, the cells were incubated and
spread on selective LB agar plates containing chloramphenicol in
order to recover transformant.
2.2. Production and purification of biopolymer in medium
Single colony of B. licheniformis NK-03 was first inoculated
into 5 ml LB media, and incubated at 35 ◦C for 12 h as the seed
culture. Then, 0.5 ml suspension was inoculated into a 500 ml coni-
cal flask containing 100 ml optimal fermentation media obtained
from orthogonal design (g/l: glucose, 50; l-glutamic acid, 40;
KH2PO4, 2.7; Na2HPO4·12H2O, 4.2; MgSO4·7H2O, 0.5; NaCl 0.5,
biotin, 2.5 × 10−4, pH 6.4). The conical flasks were incubated at 35 ◦C
in a rotary shaker at 150 rpm.
␥-PGA was recovered and purified by the methanol precipita-
tion reported previously [10]. The pH of the fermentation liquid was
adjusted to 3.0 with 6 M H2SO4 and then centrifuged at 20,000 × g
at 4 ◦C for 15 min. The supernatant was poured into four volumes
of cold methanol and stayed at −20 ◦C over night. The precipitate
was collected and dissolved in deionized water. The solution was
desalted by ultrafiltration (MWCO 6000, TP-10-20, Motian Mem-
brane Co. Ltd., China) and centrifuged at 25,000 × g at 4 ◦C for
15 min to remove insoluble materials. The solution was lyophilized
(−60 ◦C) to obtain ␥-PGA.
2.5. ꢀ-PGA biosynthesis in the recombinant strains
␥-PGA was biosynthesized by recombinant E. coli as follows:
recombinant cells were inoculated into 5 ml LB medium containing
chloramphenicol (50 ␮g/ml). After overnight growth at 37 ◦C with
shaking, a 1% (v/v) inoculum was transferred into a 500 ml conical
flask containing 200 ml LB medium plus 2% l-glutamic acid, 0.5%
MgCl2 and 50 ␮g/ml chloramphenicol. 1 mM IPTG was added into
the culture broth when the OD600 reached 1.8. Then, it was further
incubated for 48 h at 37 ◦C.
The protocol of ␥-PGA production by recombinant C. glutam-
icum was similar with that of recombinant E. coli. However, a seed
medium (g/l: glucose 25, tryptone 25, yeast extract 25, KH2PO4 1,

M. Cao et al. / Journal of Molecular Catalysis B: Enzymatic 67 (2010) 111–116
113
The biopolymer of the fermentation was determined by 1H NMR
spectrum. The number-average molecular weight (Mn), weight-
average molecular weight (Mw) and polydispersity index (PDI)
of PGA were measured by gel permeation chromatography (GPC)
using Waters 410 refractive-index detector equipped with Waters
Ultrahydrogel column (Waters, Co., U.S.). The eluant was 0.25N
NaNO3 and at a flow rate of 0.6 ml/min. The l-glutamate in the
fermentation liquid and the ratios of l- and d-isomer of ␥-PGA
hydrolysate were determined by reversed phase high performance
liquid chromatography (reversed phase HPLC) using Alltech sys-
tem controller (Alltech Associates, Inc., U.S.) after pre-column
derivation (Waters ␮BondapakTM C18 Column) by FDAA (Marfey’s
Reagent, Fig. 1A) [25], while acetonitrile was used as gradient elu-
tion phase, and the derivatives were detected by UV at 340 nm. The
two derivatives named d-glutamate-DNPA and l-glutamate-DNPA
presented their peaks at certain retention time and area. Then,
the metrical calibration curve and the equation of d-glutamate-
DNPA about mass percentage content Xm% and peak area ratio Yarea%
was drawn out by different concentrations of d-glutamate standard
derivatives (Fig. 1B): 0%, 5%, 10%, 20%, 30%, 40%, and 50%.
3. Results and discussion
Fig. 1. The stereo-chemical composition of ␥-PGA measured by the reversed phase
HPLC method. Note: (A) The chiral separation reaction of d/l-glutamic acid by FDAA.
(B) The calibration curve about mass percentage content (m%) and peak area ratio
(area%) of d-glutamate-DNPA.
3.1. Identification and taxonomic status of the isolated strain
A highly mucous strain of NK-03 was isolated from Okame natto
as described in Part 2.1. 16S ribosomal DNA sequence (GenBank
accession no. DQ020262) of the strain was compared with that
of 22 strains from NCBI database. As shown in Fig. 2, NK-03 had
the greatest similarity for about 98% and the nearest phylogenetic
relationships with Bacillus subtilis var. chungkookjang, B. mojavensis
JF-2 and B. licheniformis HK-1, respectively. The result of BIOLOG-
System, which presented the best match of similarity (SIM) 0.954,
distance (DIST) 0.68 and probability (PROB) 99%, showed that the
species was B. licheniformis (data of MicroPlate purple wells not
shown). The morphological and physiological characteristics of NK-
03 are summarized in Table 1. It was rod-shaped, capsule shielded,
gram-positive, spore-forming, catalase and amylase active. Accord-
ing to Bergey’s Manual of Systematic Bacteriology, the strain was
classified as Bacillus licheniformis NK-03.
MgSO4·7H2O 0.4, biotin 3.5 × 10−3, pH 7.0), was prepared for better
state of incubation. Then, 6% (v/v) inoculum was transferred into a
500 ml conical flask containing 100 ml fermentation medium (g/l:
glucose 50, tryptone 5, yeast extract 2, KH2PO4 1, MgSO4·7H2O 0.4,
MnSO4 0.01·FeSO4 0.01, biotin 0.5 × 10−3, pH 7.6) plus 50 ␮g/ml
chloramphenicol, and it was cultivated at 33 ◦C after the addition
of IPTG for 48 h at 33 ◦C.
2.6. Analytical procedures
Cell growth of the strains was monitored by optical density at
600 nm (OD600). The yield of ␥-PGA was determined by weight on
electronic analytical balance after lyophilized.
3.2. Characteristics of ꢀ-PGA produced by B. licheniformis NK-03
␥-PGA synthesized by NK-03 was determined by 1H NMR.
As shown in Fig. 3, the chemical shifts of ␣-H (4.311 ppm), ␤-H
(1.961 ppm and 2.199 ppm), ␥-H (2.414 ppm) and N–H (8.294 ppm)
of the sample are almost accordant with that of the standard ␥-PGA.
In addition, ␥-PGA was further characterized by Fourier Trans-
form Infrared (FT-IR) spectroscopy, amino acid analysis by HPLC
Table 1
Morphological and physiological characterizations of the newly isolated stain NK-
03.
Characterizations
Results
Shape
Endospores
Optimal growth temperature and pH
Gram stained
Short rod, 1.0 × 2.0 ␮m
Cylindrical, central
35 ◦C, 6.2
a
+
Oxidase and catalase formation
Anaerobic growth
Voges-Proskauer test
Starch and casein hydrolysis
Sporangia swollen
+
+
+
+
b
–
Ammonium salts utilization
Methyl-red test
–
–
Fig. 2. Neighbour-joining tree showing the phylogenetic relationships of the strain
NK-03. Note: The units at the bottom of the tree indicate the numbers of nucleotide
substitutions.
a
Positive.
b
Negative.

114
M. Cao et al. / Journal of Molecular Catalysis B: Enzymatic 67 (2010) 111–116
reduced within the range of 1000–20k and the PDI also decreased as
a function of hydrolysis time. Compared with physical and chemical
methods of polymer degradation, the enzymatic degradation pro-
vided a mild and controllable pathway of obtaining the expected
molecular weight and appropriate polydispersity of ␥-PGA [28]. Up
to now, the ywtD gene of B. licheniformis NK-03 has been cloned and
registered (GenBank accession no. HM067838), further biochemi-
cal and enzymatic catalysis analysis will be carried out on the basis
of culture compositions and genetic manipulation.
3.3. pgsBCA genes of B. licheniformis NK-03
A 3089 bp fragment was amplified and sequenced with the rel-
evant primers from B. licheniformis NK-03 genomic DNA. It was
composed of three open reading frames, pgsB 1182 bp, pgsC 450
bp, pgsA 1143 bp, and registered in GenBank (accession no. pgsB,
EF066513; pgsC, EF071858; pgsA, EF071859). The result of align-
ments of these open reading frames and the deduced amino acids
with those from typical Bacillus subtilis strains, including B. sub-
tilis (natto), B. subtilis IFO3336 and B. subtilis 168, suggested that
PgsC was the most conserved part in the ␥-PGA synthetase com-
plex (100% homologous), while PgsB was the most unconserved
expressive protein (about 99% consistent). However, if the PgsBCA
amino acids of NK-03 were aligned with those of B. licheniformis
ATCC14580, the different result appeared: PgsB and PgsC possessed
much lower similarity of 90%. Therefore, the strain of B. licheni-
formis NK-03 may be an atypical mosaic or transitional strain of
B. licheniformis and B. subtilis. According to the former reports,
it could be easily revealed that PgsC was the most hydrophobic
component to interact with cell membranes; PgsB, which has a
structure of amido-ligase probably acts as the active site of PgsBCA
complex and represents great differences of molecular weight and
monomer ratio in various strains [29,30]. When referring to Can-
dela’s schematics hypothesis, the PGA synthesis may be divided
into two steps, PgsB/CapB and PgsC/CapC are involved in PGA syn-
thesis, whereas PgsA/CapA serves as transporter [31].
Fig. 3. 1H NMR spectrum of ␥-PGA from Sigma and product of NK-03. Note: (A) 1
H
NMR spectrum of ␥-PGA from Sigma in D2O. The chemical shifts of ␣-H (4.269 ppm),
␤-H (1.979, 1.870 ppm), ␥-H (2.240 ppm)and N–H (8.575 ppm) were labeled in
the peaks station; (B) 1H NMR spectrum of ␥-PGA in D2O from NK-03 fermenta-
tion product. The chemical shifts of ␣-H (4.311 ppm), ␤-H (2.199, 1.961 ppm), ␥-H
(2.199 ppm) and N–H (8.5294 ppm) were labeled in the peaks station.
and thin layer chromatography (TLC) (data not shown). The results
suggested that ␥-PGA was the primary product of B. licheniformis
NK-03 and it scarcely synthesized other polypeptides or extracel-
lular polysaccharide by-products.
3.4. Recombinant strains harboring pXMJ19-PGS
The weight-average molecular weight (Mw), the number-
average molecular weight (Mn) and the polydispersity index (PDI)
were analyzed by Waters system (Table 2). Under the culture con-
ditions in this study, the Mw of ␥-PGA produced by B. licheniformis
NK-03 was about 1.36 × 106. To our knowledge, ␥-PGA with Mw
over 106 was rarely synthesized by B. licheniformis. Although differ-
ent species or strains could produce ␥-PGA with different molecular
weights, the cultivated conditions and medium components, such
as rotate speed, medium ionic concentration, and fermentation
time [26,27], had also been indicated to affect the molecular weight
of ␥-PGA by the way of activation or deactivation of the synthetase
and depolymerase. Taking ␥-PGA depolymerase YwtD of B. subtilis
NX-2 for example, it was produced in cells of the E. coli DE3 clone
harboring pET15b-ywtD and was purified by metal-chelating affin-
ity chromatography. The strain’s YwtD protein was proved to be an
endo-hydrolase and exhibited an obvious activity at wide ranges of
temperature (30–40 ◦C) and pH (5.0–8.0). The Mw of ␥-PGA could be
The pgsBCA genes of B. licheniformis NK-03 were fused between
BamHI and HindIII sites of pXMJ19, constructing recombinant plas-
mid which was named as pXMJ19-PGS (see Fig. 4). E. coli JM109
clones that harbored pXMJ19-PGS were selected by the methods
of colony PCR and recombinant plasmid enzyme-digestion. Judg-
ing from the bands migration distance compared to Marker III in
0.8% agarose gel electrophoresis shown in Fig. 5, it is concluded
that pXMJ19-PGS with a 6.6 kb pXMJ19 and 3.0 kb pgsBCA genes
was successfully introduced into E. coli. To our knowledge, the syn-
thetase genes in recombinant strains that could synthesize PGA
under the addition of inducer (IPTG) and substrate (glutamic acid),
were mostly from B. subtilis. Therefore, the relative work about pgs-
BCA genes cloned from B. licheniformis into E. coli has hardly been
reported.
In addition, considering glutamic acid is a very important sub-
strate for ␥-PGA production, we developed another producer of
␥-PGA by using Corynebacterium glutamicum that was able to
Table 2
Characterizations of ␥-PGA produced by the wild NK-03 and two recombinant strains.
Strains
Weight-average
molecular weight (Mw
Number-average
molecular weight (Mn)
Polydispersity
index (PDI)
l-Glutamate
monomer ratio (%)
Yield of
PGA (g/L)
)
B. licheniformis NK-03
E. coli pXMJ19-PGS
C. glutamicum pXMJ19-PGS
1,362,828
42,433
NDa
850,736
23,964
ND
1.60
1.77
ND
98
97
97
10.5
0.51
0.69
a
Not determined.

M. Cao et al. / Journal of Molecular Catalysis B: Enzymatic 67 (2010) 111–116
115
Fig. 6. The reversed phase HPLC chromatogram profile of ␥-PGA hydrolysates
derivatives synthesized by wild strain B. licheniformis NK-03.
hot-shock could obviously improve the efficiency of gram-positive
bacteria transformation (Fig. 5).
3.5. ꢀ-PGA produced by recombinant strains
The Mw of ␥-PGA produced by the recombinant of E. coli was
42,433, which was much lower than that of wild strain. It was
hypothesized by our results that E. coli was gram-negative bacteria
while B. licheniformis was gram-positive, the membrane-associated
enzyme complex PgsBCA could not be localized very well in cell
membranes of the recombinant E. coli. However, Coryneform bac-
teria are gram-positive bacteria like Bacillus strains, and the rigid
structure of cell walls seems to be more suitable for the display and
expression of pgsBCA than those of E. coli.
Fig. 4. Construction of recombinant plasmid pXMJ19-PGS from shuttle vector
pXMJ19 and ␥-PGA synthetase pgsBCA genes.
express ␥-PGA synthetase genes pgsBCA. As a bacteria of high
glutamate productivity, C. glutamicum was chosen to construct a
recombinant strain that could make use of its own production of
glutamic acid and synthesize ␥-PGA directly, which may enor-
mously decrease the cost of ␥-PGA production. Nevertheless, C.
glutamicum is gram-positive bacteria with thick cell wall and strong
restriction modification system, so it was very intractable to intro-
duce pgsBCA genes into the competent cells of C. glutamicum by
normal electro-transformation. We tried to prepare C. glutamicum
protoplast by lysozyme for transforming, then the regenerate cells
were treated by hot-shock (46 ◦C) for about 6 min. The result of
successful construction of recombinant C. glutamicum carried pgs-
BCA genes suggested that methods of protoplast preparation and
The recombinant strain of C. glutamicum could produce ␥-PGA
successfully without adding glutamic acid. However, the yields
of ␥-PGA in the recombinant strains of E. coli and C. glutam-
icum were merely 0.51 g/L and 0.69 g/L respectively, which were
much lower than the yield in wild strain of B. licheniformis (about
10.5 g/L). As shown in Fig. 6, the reversed phase HPLC profile
of ␥-PGA hydrolysates derivatives from NK-03, which presented
the retention time of 19.0 min (l-glutamate-DNPA) and 20.3 min
(d-glutamate-DNPA), possessed the peak area proportion of 3.7%
(d-glutamate). Furthermore, the associated HPLC results of ␥-PGA
synthesized by recombinant strains E. coli JM109 and C. glutam-
icum ATCC13032 were also measured (data not shown). It was
clearly indicated that the monomers derivatives from recombi-
nant strains held the same retention time as that of NK-03, and
the peak area ratio of d-glutamate was about 4.8%. The equation of
d-glutamic acid about mass percentage content and peak area ratio
was acquired from the curve in Fig. 1B: Yarea% = 1.1759Xm% + 0.0132
(R2 = 0.9973, R is the linear correlation coefficient). Through the
peak area ratio of 4.8%, the mass percentage content of ␥-PGA iso-
mers in the recombinants was determined: 3% d-isomer and 97%
l-isomer, which was similar to that of wild strain NK-03 (l-Glu,
98%). The PgsBCA synthetase could use either l- or d-isomer as
substrates, but the activity of glutamate racemase Glr in strain NK-
03, which could isomerize l-Glu to d-Glu, is lower than in other
␥-PGA accumulating strains. On the other hand, the two kinds of
recombinant strains did not have the racemase Glr or Glr’s activ-
ity is negligible. Therefore, the content of l-glutamate monomer of
␥-PGA synthesized by wild strain NK-03 and recombinant strains
was much higher than d-glutamate.
Fig. 5. Colony PCR and enzyme-digestion profile of recombinant plasmid in E. coli.
JM109 and C. glutamicum ATCC13032. Note: Lane M, DNA marker III (TIANGEN);
Lane 1, pXMJ19/BamHI in E. coli; Lane 2, Colony PCR product in E. coli; Lane 3,
pXMJ19-PGS/BamHI in E. coli; Lane 4, pXMJ19-PGS/HindIII in E. coli; Lane 5, pXMJ19-
PGS/HindIII + BamHI in E. coli; Lane 6, Colony PCR product in C. glutamicum. Lane 7,
pXMJ19-PGS/BamHI in C. glutamicum; Lane 8, pXMJ19-PGS/HindIII in C. glutamicum;
Lane 9, pXMJ19-PGS/HindIII + BamHI in C. glutamicum.
In past works, Ashiuchi et al. had constructed an E. coli strain
harboring the plasmid carrying both pgsBCA genes and glutamate
racemase gene of B. subtilis IFO3336 [29], which could produce

116
M. Cao et al. / Journal of Molecular Catalysis B: Enzymatic 67 (2010) 111–116
a larger amount of ␥-PGA than the clone consisting of only pgs-
BCA. It was indicated that an increase supply of d-glutamate
by co-expression of such two genes probably resulted in higher
polymer production. At present, the authors are looking for the
factors affecting synthesis of ␥-PGA and applying co-expression
of synthetase and racemase methods to improve the production
efficiency in the recombinant strains.
References
[1] M. Ashiuchi, H. Misono, Appl. Microbiol. Biotechnol. 59 (2002) 9–14.
[2] M. Ashiuchi, T. Kamei, H. Misono, J. Mol. Catal. B 23 (2003) 101–106.
[3] M. Ashiuchi, K. Shimanouchi, H. Nakamura, T. Kamei, K. Soda, C. Park, M.H.
Sung, H. Misono, Appl. Environ. Microbiol. 70 (2004) 4249–4255.
[4] G. Ivanovics, L. Erdos, Z. Immunitatsforsch. 90 (1937) 5–19.
[5] S. Makino, I. Uchida, N. Terakado, C. Sasakawa, M. Yoshikawa, J. Bacteriol. 171
(1989) 722–730.
[6] F.F. Hezayen, B.H.A. Rehm, R. Eberhardt, A. Steinbüchel, Appl. Microbiol.
Biotechnol. 54 (2000) 319–325.
[7] M. Ashiuchi, T. Kamei, D.H. Baek, S.Y. Shin, M.H. Sung, K. Soda,
T. Yagi, H. Misono, Appl. Microbiol. Biotechnol. 57 (2001) 764–
769.
4. Conclusion
In this study, a ␥-PGA producing strain was isolated from natto
and identified as Bacillus licheniformis NK-03. The pgsBCA genes
encoding the ␥-PGA sythetase were cloned from NK-03 and trans-
formed to E. coli and C. glutamicum. Enzyme digestion tests of
resultant vector and analysis of fermented production suggested
that the genes were expressed in both strains. The deduced PgsBCA
amide acids from NK-03 were greatly homologous to other Bacillus
species, and PgsC was the most conserved part. ␥-PGA produced
by the recombinant strains E. coli and C. glutamicum consisted of
about 3% d-isomer and 97% l-isomer, which was similar to the wild
strain. Although the yield of the recombinant C. glutamicum was
not high, it had already carried out the ␥-PGA production without
adding any glutamic acid into medium and offered us a outstanding
protocol of ␥-PGA synthesis by molecular modification. The rea-
sons inducing the low production may be that the recombinant
plasmid was not expressed adequately or the fermentation con-
ditions were not optimized. Therefore, selecting better expression
plasmids and effective promoters will be introduced to perform
anticipated production.
[8] G.A. Birrer, A.M. Cromwick, R.A. Gross, Int. J. Bio. Macromol. 16 (1994) 765–769.
[9] Y.H. Ko, R.A. Gross, Biotechnol. Bioeng. 57 (1998) 430–437.
[10] A. Goto, M. Kunioka, Biosci. Biotech. Biochem. 63 (1992) 110–115.
[11] J.H. Jeong, J.N. Kim, Y.J. Wee, H.W. Ryu, Bioresour. Technol. 101 (2010)
4533–4539.
[12] I.L. Shih, Y.T. Van, C.J. Shieh, Process. Biochem. 40 (2005) 2827–2832.
[13] Y. Ito, T. Tanaka, T. Ohmachi, Y. Asada, Biosci. Biotech. Biochem. 60 (1996)
1239–1242.
[14] C. Cheng, Y. Asada, T. Aida, Agric. Biol. Chem. 53 (1989) 2369–2375.
[15] N.A. Soliman, M.M. Berekaa, Y.R. Abdel-Fattah, Appl. Microbiol. Biotechnol. 69
(2005) 259–267.
[16] I.L. Shih, Y.T. Van, Bioresour. Technol. 79 (2001) 207–225.
[17] M.H. Sung, C. Park, K.S. Kim, J.H. Park, Y.H. Choi, H.B. Kim, S.P. Hong, I.H. Lee,
C.M. Chung, KR/10-0062761, 08, 09, 2003.
[18] C. Park, J.C. Choi, Y.H. Choi, H. Nakamura, K. Shimanouchi, T. Horiuchi, H. Mis-
ono, T. Sewaki, K. Soda, M. Ashiuchi, M.H. Sung, J. Mol. Catal. B 35 (2005) 128–
133.
[19] J. Meerak, H. Iida, Y. Watanabe, M. Miyashita, H. Sato, Y. Nakagawa, Y. Tahara,
J. Gen. Appl. Microbiol. 53 (2007) 315–323.
[20] J. Sambrook, T. Maniatis, Molecular Cloning: A Laboratory Manual, 2nd ed., Cold
Spring Harbor Laboratory, New York, 1989.
[21] A. Felske, A. Wolterink, Lis.R. van, W.M. de Vos, A.D. Akkermans, FEMS Micro-
biol. Ecol. 30 (1999) 137–145.
[22] J.G. Holt, N.R. Krieg, P.H.A. Sneath, J.T. Staley, J. Stanley, Bergey’s Manual of
Systematic Bacteriology, 9th ed., Williams & Wilkins, Baltimore, 1984, pp.
559–562.
[23] M. Ashiuchi, K. Shimanouchi, T. Horiuchi, T. Kamei, H. Misono, Biosci. Biotech-
nol. Biochem. 70 (2006) 1794–1797.
[24] H. Inoue, H. Nojima, H. Okayama, Gene 96 (1990) 23–28.
[25] J.G. Adamson, T. Hoang, A. Crivici, G.A. Lajoie, Anal. Biochem. 202 (1992)
210–214.
[26] I.L. Shih, Y.T. Van, Y.N. Chang, Enzyme Microb. Technol. 31 (2002) 213–
220.
In recent work, we isolated a glutamic acid independent ␥-PGA
producing strain Bacillus amyloliquefaciens LL3, and cloned its pgs-
BCA and ywtC genes successfully. Compared with B. subtilis and B.
licheniformis, the nucleotide sequences of pgsBCA from LL3 were
much less homologous. The future in-depth study on this pgsBCA
operon may give us some revelations about the mechanism of ␥-
PGA synthesis.
[27] I.B. Bajaj, S.S. Lele, R.S. Singhal, Bioresour. Technol. 100 (2009) 826–
832.
[28] J. Yao, J. Jing, H. Xu, J.F. Liang, Q. Wu, X.H. Feng, P.K. Ouyang, J. Mol. Catal. B 56
(2009) 158–164.
Acknowledgements
This work was supported by the National 863 High-Tech
R&D Program of China (2007AA06Z323), Tianjin Scientific project
(China) (09JGZDJC18400, 09ZCKFSH00800).
[29] M. Ashiuchi, K. Soda, H. Misono, Biochem. Biophys. Res. Commun. 263 (1999)
6–12.
[30] M. Ashiuchi, C. Nawa, T. Kamei, J.J. Song, S.P. Hong, M.H. Sung, K. Soda, H. Misono,
Eur. J. Biochem. 268 (2001) 5321–5328.
[31] T. Candela, A. Fouet, Mol. Microbiol. 60 (2006) 1091–1098.