Appl Microbiol Biotechnol (2006) 70: 564–572
DOI 10.1007/s00253-005-0111-x
APPLIED GENE TICS AN D MO LECULA R BIO TECH NOLOGY
Junya Narita . Kenji Okano . Toshihiro Tateno .
Takanori Tanino . Tomomitsu Sewaki .
Moon-Hee Sung . Hideki Fukuda . Akihiko Kondo
Display of active enzymes on the cell surface
of Escherichia coli using PgsA anchor protein
and their application to bioconversion
Received: 26 June 2005 / Revised: 20 July 2005 / Accepted: 20 July 2005 / Published online: 20 August 2005
# Springer-Verlag 2005
Abstract We have developed a novel Escherichia coli cell
surface display system by employing PgsA as an anchoring
motif. In our display system, C-terminal fusion to PgsA
anchor protein from Bacillus subtilis was used. The en-
zymes selected for display were α-amylase (AmyA) from
Streptococcus bovis 148 and lipase B (CALB) from Can-
dida antarctica. The molecular mass values of AmyA and
CALB are approximately 77 and 34 kDa, respectively.
The enzymes were displayed on the surface as a fusion
protein with a FLAG peptide tag at the C terminus. Both
the PgsA-AmyA-FLAG and PgsA-CALB-FLAG fusion
proteins were shown to be displayed by immunofluores-
cence labeling using anti-FLAG antibody. The displayed
enzymes were active forms, and AmyA and CALB ac-
tivities reached 990 U/g (dry cell weight) and 4.6 U/g (dry
cell weight), respectively. AmyA-displaying E. coli cells
grew utilizing cornstarch as the sole carbon source, while
CALB-displaying E. coli cells catalyzed enantioselective
J. Narita . T. Tanino . H. Fukuda
Division of Molecular Science,
Graduate School of Science and
Technology, Kobe University,
Nada-ku, Kobe, 657-8501, Japan
K. Okano . T. Tateno . A. Kondo (*)
Department of Chemical Science and Engineering,
Faculty of Engineering, Kobe University,
1-1 Rokkodaicho, Nada-ku,
Kobe, 657-8501, Japan
e-mail: akondo@kobe-u.ac.jp
Tel.: +81-78-8036196
Fax: +81-78-8036206
T. Sewaki
BioLeaders Japan Corporation,
Saito Asagi,
Ibaraki, 567-0085, Japan
M.-H. Sung
Department of Bio and Nanochemistry,
Kookmin University,
Songbuk-gu,
Seoul, 136-702, South Korea
transesterification, indicating that they are effective whole-
cell biocatalysts. Since a target enzyme with a size of
77 kDa and an industrially useful lipase have been suc-
cessfully displayed on the cell surface of E. coli for the
first time, PgsA protein is probably a useful anchoring motif
to display various enzymes.
Introduction
In recent years, there has been considerable progress in the
development of expression systems for the display of het-
erologous peptides and proteins on the surface of micro-
organisms (Benhar 2001). Microbial cell surface display
has a wide range of biotechnological and industrial ap-
plications in areas such as whole-cell biocatalysis for bio-
conversion, biosensing, biosorption, immobilization, oral
vaccine development, and combinatorial peptide library
screening (Lee et al. 2003). In gram-negative bacteria, sur-
face display systems based on various anchoring strategies
have been reported, with outer membrane proteins (Chang
and Lo 2000; Lee et al. 2004), pili and flagella (Westerlund-
Wikström et al. 1997), modified lipoproteins (Georgiou
et al. 1996), ice nucleation proteins (Jung et al. 1998),
and autotransporters (Veiga et al. 2003) among the an-
chors used. Escherichia coli is an attractive host because
various genetic tools and mutant strains are available, and
its high transformation efficiency makes it ideal for screen-
ing a peptide or protein libraries after surface display. In the
current E. coli cell surface display systems, however, there
remain problems such as the size limitation of the target
protein, misfolding, mislocalization, and destabilization of
the outer membrane (Georgiou et al. 1997; Wan et al. 2002).
The size limitation and misfolding in particular are sig-
nificant problems for broader applications. In most of the
previous studies, peptides and small proteins have been
displayed. The largest target protein reported to have been
displayed so far is a bacterial lipase with a molecular size of
49.9 kDa using OprF as an anchor motif (Lee et al. 2005).
However, for instance, to use E. coli cells for bioconversion
and biosensing, it is necessary to display various enzymes,

receptors and other signal-sensitive components with large
molecular sizes in active form.
In the E. coli cell surface display system for active
enzymes developed in the present study, the anchor protein
used was PgsA protein from Bacillus subtilis (Ashiuchi
and Misono 2002), which is a poly-γ-glutamate (PGA)
synthetase complex. PGA is an unusual anionic polypep-
tide in which glutamate is polymerized via γ-amide link-
ages. According to the previous report, PgsA functions to
stabilize the complex by anchoring in the cell membrane
(Ashiuchi et al. 2001). PgsA is localized in the membrane
and is expected to be used in the future for cell surface
display of heterologous protein.
The heterologous target enzymes selected were α-
amylase from Streptococcus bovis 148 (AmyA) and lipase
B from Candida antarctica CBS6678 (CALB), which have
respective molecular mass sizes of approximately 77 and
34 kDa, respectively (Satoh et al. 1993; Uppenberg et al.
1994). The cell-surface-displayed α-amylase and lipase are
expected to be effective for various bioconversions. We
succeeded in showing that, through fusion to its C termi-
nus, PgsA protein can be used in E. coli to display enzymes
in a conformationally active form on the cell surface. This
is the first report on successful cell surface display of an
enzyme larger than a molecular size of 60 kDa in active
form. In addition, this is the first report on expression of
active CALB in E. coli. We further demonstrated that
E. coli cells displaying enzymes on their surface can be
used as whole-cell biocatalyst.
Materials and methods
Bacterial strains and growth conditions
Escherichia coli NovaBlue {endA1 hsdR17 (rK12−mK12+)
supE44 thi-1 recA1 gyrA96 relA1 lac[F’ proAB+lacIq
ZΔM15::Tn10 (Tetr)]} (Novagen, Inc., Madison, WI) was
used for construction of vectors and expression of heter-
ologous protein on the cell surface. It was cultivated in
Luria–Bertani (LB) medium (10 g of tryptone, 5 g of yeast
extract, and 5 g of NaCl in 1 l of distilled water) or on LB
agar plates containing 100 μg/ml of ampicillin at 37°C.
Construction of plasmids for cell surface display
The basic plasmid for cell surface display, pHLA with a
high constitutive expression (HCE) promoter (Fig. 1a),
was constructed as described below; all PCRs were car-
ried out with KOD-Plus-polymerase (Toyobo Co., Osaka,
Japan). The pgsA gene was PCR-amplified directly from
pHCE1LB-pgsBCA-HB168 (Sung et al. 2003) using the
oligonucleotides pgsA_F and pgsA-MCS_R (Table 1), and
the PCR fragment was inserted into pHCE1LB, which is a
derivative of pHCE (Poo et al. 2002) with EcoRV and
HindIII sites, to give pHLA. The plasmids for cell surface
display of α-amylase (pHLAαAF) and lipase B (pHLA-
CALBF) (Fig. 1b), in which the fusion proteins of PgsA-α-
565
a HCE
Cmr pgsA
pHLA MCS
rrnBT1T2
6.35 kb
b ori
Ampr
pHLAαAF
HCE pgsA amyA rrnBT1T2
terminator
pHLA-CALBF
FLAG tag
rrnBT1T2
HCE pgsA CALB
terminator
FLAG tag
Fig. 1 Construction of vector plasmid for cell surface display.
a Expression plasmid (pHLA) for display on E. coli cell surface.
b Structure of relevant part of cell surface display plasmid. HCE
promoter and rrnBT1T2 terminator were used. MCS multicloning
sites
amylase (or lipase B)-FLAG tag (DYKDDDDK) were
expressed, were constructed as follows: The mature α-
amylase gene was amplified by PCR from pQE31::amyA
(Shigechi et al. 2004) with amyA_F and amyA-FLAG_R
primers, and the amplified fragment was digested with
BamHI and HindIII and introduced into the BamHI and
HindIII site of plasmid pHLA. The resulting plasmid was
designated pHLAαAF. The DNA sequence coding the
mature lipase B containing a Pro sequence and no signal
sequence was amplified by PCR (972 bp) from C. ant-
arctica CBS6678 genomic DNA (Uppenberg et al. 1994)
using ProCALB_F and CALB-FLAG_R primers. The am-
plified fragment was digested with SalI and XhoI and
introduced into the SalI and XhoI site of plasmid pHLA.
The resulting plasmid was designated pHLA-CALBF.
Preparation of membrane fractions of E. coli
Membrane fractions were isolated from a 5-ml culture of
E. coli cells grown at 37°C for 24 h in LB medium using
the following method. The cultures were diluted to an op-
tical density of 0.8 at 600 nm (OD600) with 20 mM Tris-HCl
(pH 8.0) buffer. Samples of 1 ml were centrifuged at
4,000×g for 5 min at 4°C and then resuspended in 1 ml of
20 mM Tris-HCl (pH 8.0) buffer, after which the cells were
disrupted by sonication (UD-201, Tomy Seiko, Tokyo,
Japan) (output, 1; duty, continue; time, 15-s three times on
ice). The undisrupted cells removed by repeated centrifu-
gation at 1,500×g for 5 min at 4°C. The supernatant was
centrifuged for 30 min at 21,800×g at 4°C, and the mem-
brane fraction was recovered as the pellet, which was re-
suspended in equal amounts of 20 mM Tris-HCl buffer and
2× loading buffer of sodium dodecyl sulfate–polyacryla-
mide gel electrophoresis (SDS-PAGE). Prior to loading of
SDS-PAGE, all samples were boiled for 5 min.
566
Table 1 Plasmids and oligonucleotide primers used
Plasmid and oligonucleotide primer Phenotype and sequence Reference

pHCE Vector under HCE promoter control, ampicillin Poo et al. (2002)
and chloramphenicol resistance marker
pHCE1LB Vector under HCE promoter control, ampicillin Sung et al. (2003)
and chloramphenicol resistance marker, derivative of pHCE
pHCE1LB-pgsBCA-HB168 Vector for display of HB168 using PgsBCA anchor motif, Sung et al. (2003)
ampicillin and chloramphenicol resistance marker
pHLA Cell surface display vector containing pgsA gene under Present study
HCE promoter control, ampicillin resistance marker
pHLAαAF Vector for display of α-amylase with FLAG tag, derivative of pHLA Present study
pHLA-CALBF Vector for display of CALB with FLAG tag, derivative of pHLA Present study
pgsA_F 5′-CCCAAGCTTCAGCTGATGAAAAAAGAACTGAGCTTTCA
TGAAAAGC-3′
pgsA-MCS_R 5′-CCCAAGCTTCTCGAGTCTAGAGTCGACCCCGGGAGATCTGGATCC
TTTAGATTTTAGTTTGTCACTATGATCAATATCAAACGTC-3′
amyA_F 5′-CGCGGATCCGATGAACAAGTGTCAATGAAAGATGGTACG-3′
amyA-FLAG_R 5′-CCCAAGCTTACTTGTCATCGTCATCCTTGTAGTCTTTTAGCCCATC
TTTATTATAGTTTCCAGATTTTACAAGG-3′
ProCALB_F 5′-ACGCGTCGACGCCACTCCTTTGGTGAAGCGTCTACCTTCCGG-3′
CALB-FLAG_R 5′-TGCTCTAGACTCGAGCTACTTGTCATCGTCATCCTTGTAGTCGGG
GGTGACGATGCCGGAGCAGG-3′
Restriction enzyme sites are underlined; FLAG sequences are double-underlined

Western blot analysis
Protein samples were analyzed on SDS-PAGE with 8%
gel. As the MW marker, Precision Plus Protein standards,
dual-color prestained (Bio-Rad Laboratories, Inc. Rich-
mond, CA), were used. The proteins were subsequent-
ly electroblotted on polyvinylidene difluoride membrane
(Millipore Co., Boston, MA) and allowed to react with
primary mouse anti-FLAG M2 (Sigma, St. Louis, MO) and
secondary goat anti-mouse IgG alkaline phosphatase-con-
jugated antibody (Promega Co., Madison, WI), after which
the membrane was then stained with nitroblue tetrazoli-
um (NBT; Promega) and 5-bromo-4-chloro-3-indolylphos-
phate (BCIP; Promega) according to the protocol specified
by the supplier. The staining solution was prepared by
adding 95 μl of NBT and 50 μl of BCIP sequentially to
15 ml of alkaline phosphatase buffer (100 mM Tris-HCl
[pH 9.0] containing 150 mM NaCl and 1 mM MgCl2).
Immunofluorescence microscopy
Immunostaining was performed as follows: E. coli cells
were cultivated in LB medium at 37°C for 24 h, collected
by centrifugation at 3,500×g for 5 min at 4°C, and washed
with PBS (pH 7.2). After resuspension in PBS containing
10 g/l bovine serum albumin (PBS+BSA) (OD600=1.0) and
incubation for 30 min at room temperature, the cells and
the primary antibody were incubated in PBS+BSA for
1.5 h at room temperature. As a primary antibody, mouse
anti-FLAG M2 antibody was used at a dilution ratio of
1:500. The cells were then incubated for 1 h at room tem-
perature with the second antibody, a 1:500 dilution ratio
of goat anti-mouse IgG (H+L) antibody conjugated with
Alexa Fluor 488 (Molecular Probes, Inc., Eugene, OR).
After washing with PBS, the cells were observed by mi-
croscopy. Immunofluorescence microscopy analysis was
performed using an Axioskop epifluorescence microscope
equipped with a 12-V, 50-W halogen lamp for transmitted-
light illumination, a 50-W mercury arc lamp for epifluo-
rescence illumination, and an AxioCam HRm camera (Carl
Zeiss, Oberkochen, Germany).
Flow-cytometric analysis
Flow-cytometric analysis was performed on a FACSCali-
bur flow cytometer (Becton Dickinson Immunocytometry
Systems, San Jose, CA) equipped with a 15-mW, 488-nm,
air-cooled argon ion laser. Cells immunostained as de-
scribed above were diluted to approximately 106 cells/ml
and delivered at a low flow rate, corresponding to 100–300
cells/s. A band pass filter of 530 nm (515–545 nm) was
used to collect the green fluorescence (FL1). Data were
analyzed with the CELLQuest program (version 3.3;
Becton Dickinson). In each experiment, 10,000 counts
were analyzed. Counts were made in triplicate for each
procedure.
Measurement of α-amylase activity
The α-amylase activity on the E. coli cell surface was
measured with an α-amylase measurement kit (Kikkoman

Co., Tokyo, Japan) with 2-chloro-4-nitrophenyl 65-azido-
65-deoxy-β-maltopentaoside (N3-G5-β-CNP) as substrate.
The assay mixture contained 400 μl of reaction solution
and 40 μl of sample solution. To prepare the sample so-
lution, E. coli cells harvested from the culture medium
were washed twice with distilled water, lyophilized with
FreeZone FZ-1 (Labconco, Kansas, MO) for 48 h. The dry
cell weight was measured. Then cells were resuspended
in distilled water with vigorous agitation, and a specified
amount of the suspension was used for α-amylase activity
assay. The assay mixture was incubated at 37°C for 10 min,
and the enzyme reaction was terminated by addition of
800 μl of reaction stop solution. The activity was assayed
by measuring the absorbance of liberated 2-chloro-4-
nitrophenol (CNP) at 400 nm. One unit (U) of activity
was defined as the amount of enzyme needed to release
1 μmol CNP/min from N3-G5-β-CNP at 37°C.
Measurement of lipase activity
Lipase activity on the E. coli cell surface was measured by
the spectrometric method using p-nitrophenyl butyrate
(PNPB) as substrate. The final concentration of PNPB was
0.5 mM in a substrate solution containing 0.5% ethanol.
The assay mixture, with a total volume of 1 ml, contained
750 μl of substrate solution, 200 μl of 20 mM phosphate
buffer (pH 7.0), and 50 μl of sample solution. To prepare
the sample solution, E. coli cells harvested from the culture
medium were washed twice with distilled water and ly-
ophilized with FreeZone FZ-1 for 48 h. The dry cell weight
was measured. Then cells were resuspended in distilled
water with vigorous agitation, and a specified amount of
the suspension was used for lipase activity assay. The assay
mixture was incubated at 30°C for 10 min with shaking
at 170 strokes/min, and the enzyme reaction was stopped
by addition of 50 μl of 5% (w/v) trichloroacetic acid. The
activity was assayed by measuring the absorbance of lib-
erated p-nitrophenol (PNP) at 400 nm. One unit (U) of
activity was defined as the amount of enzyme needed to
release 1 μmol PNP/min from PNPB at 30°C.
Protease accessibility assay
For whole-cell trypsin treatment, the sample solution for
measurement of activity was adjusted to an OD600 of 1.5
with PBS, and surface-displayed enzyme was cleaved by
incubation of the suspension at 37°C for 3 h with trypsin at
the final concentration of 20 mg/l. After the reaction, the
cells were washed twice with PBS by gentle centrifugation
to remove typsin, and the enzyme activity was measured as
described above.
Growth of cells utilizing cornstarch
The medium used for growth experiment of α-amylase-
displaying cells was consisted of M9 salt solution (6.3 g of
567
Na2HPO4, 3.0 g of KH2PO4, 0.5 g of NaCl, and 1.0 g of
NH4Cl in 1 l of distilled water) plus 1 mM MgSO4, 0.1 mM
CaCl2, 10 μg/ml of thiamine, and 100 μg/ml of ampicillin.
The growth experiments of recombinant E. coli utilizing
cornstarch as a sole carbon source were performed by a 2-l
fermentor. Before inoculation, cornstarch solution (100 g/l;
Wako Pure Chemical Industries, Ltd., Osaka, Japan) was
incubated at 80°C for 5 min for low-temperature cooking,
after which it was mixed with M9 growth medium to a final
cornstarch concentration of 20 g/l. An early stationary
phase seed culture was then used to inoculate the freshly
mixed medium at 5% volume. The growth experiments
were carried out at 37°C. To determine cell growth, count-
ing of colony-forming units (CFU) was performed on
an LB agar plate containing 100 μg/ml of ampicillin by
a spread plate method. Colorimetric methods based on
phenol-sulphuric acid (Dubois et al. 1956) and 3,5-dinitro-
salicylic acid (Robyt and Whelan 1972) were used to
determine the amount of total sugar and reducing sugar,
respectively. The concentration of glucose was determined
using a Glucose CII-Test Wako kit (Wako).
Enantioselective transesterification reaction
To prepare the CALB-displayed transformants as whole-
cell biocatalyst, the cells were collected after 36 h culti-
vation in LB medium, washed with 20 mM HEPES buffer
(pH 7.0), and lyophilized for 48 h. The reaction mixture,
consisting of 50 mg (RS)-1-phenylethanol [(RS)-1-PE],
35.3 mg vinyl acetate (VA), 50 mg lyophilized cells, 0.5 g
molecular sieves 4A beads (Nacalai Tesque, Kyoto, Japan),
and 5 ml n-hexane was incubated at 37°C with shaking at
150 strokes/min. The concentration of (RS)-1-PE and VA
was 80 mM. The progress of enantioselective esterification
was determined with HPLC according to the previously
reported method (Kaieda et al. 2004). The experimental
data divided by the theoretical yield calculated from the
initial substrate quantity was taken as product yield. Enan-
tiomeric excess (%) was calculated as follows: enantio-
meric excess (%)=|R-S|/(R+S)×100, where R and S are the
concentration of (R)- and (S)-1-phenylethyl acetate[(R)-
and (S)-1-PEA], respectively.
Results
Expression of PgsA-AmyAF and PgsA-CALBF
fusion proteins on cell surface
We constructed the plasmids for display of the PgsA-
AmyA-FLAG (PgsA-AmyAF) and PgsA-CALB-FLAG
(PgsA-CALBF) fusion proteins on the cell surface of
E. coli (Fig. 1). Since these fusion proteins have a FLAG
tag at the C terminus of AmyA and CALB, the expressed
fusion proteins were analyzed by Western blotting with
anti-FLAG M2 antibody. Figure 2 shows a Western blot
analysis of membrane fractions of E. coli harboring the
plasmids pHLAαAF and pHLA-CALBF. Since the secre-
568

Fig. 2 Western blot analysis of membrane fractions from E. coli

harboring the plasmid pHLA (lane 1), pHLAαAF (lane 2), and
pHLA-CALBF (lane 3). Cells were grown at 37°C for 24 h in LB
medium containing 100 mg/l of ampicillin. The membrane fractions
were separated by SDS-PAGE with 8% gel and stained with primary
mouse anti-FLAG M2, followed by secondary goat anti-mouse IgG
conjugated with alkaline phosphatase
Fig. 3 Immunofluorescence labeling of transformed cells: Nomars-
ki differential interference micrographs (column 1) and immuno-
fluorescence micrographs (column 2). E. coli harboring the plasmid
tion signal cleavage site of PgsA is predicted to be the pHLA, pHLAαAF, and pHLA-CALBF were grown at 37°C for 24 h
position between 38 and 39 amino acids by the method of in LB medium containing 100 mg/l of ampicillin. Cells were labeled
von Heijne (1986), molecular sizes of PgsA and mature with primary mouse anti-FLAG M2, followed by secondary goat
PgsA are approximately 43 and 38 kDa, respectively. As anti-mouse IgG conjugated with Alexa Fluor 488
the molecular sizes of the AmyA, CALB, and FLAG tag
are approximately 77, 34 and 1 kDa, respectively, those
of the mature PgsA-AmyAF and PgsA-CALBF fusion
proteins are estimated to be 116 and 73 kDa, respectively.
Clear bands of the fusion proteins at the estimated mo-
lecular sizes were observed, indicating their successful ex-
pression in the membrane fraction.

Immunofluorescence microscopy
and flow-cytometric analysis

Immunofluorescent labeling of cells was performed using

mouse anti-FLAG M2 antibody and goat anti-mouse IgG
(H+L) conjugated with Alexa Fluor 488. The green fluo-
rescence of the immunostained PgsA-AmyAF and PgsA-
CALBF fusion proteins was observed in E. coli cells
harboring the plasmids pHLAαAF and pHLA-CALBF,
respectively (Fig. 3); cells harboring the control plasmid
pHLA were not immunostained. This confirms that the
PgsA-AmyAF and PgsA-CALBF fusion proteins were
anchored on the E. coli cell surface. We additionally used
flow cytometry to confirm and quantitatively analyze the
cell-surface-displayed enzyme (Fig. 4). The cell-surface-
displayed AmyAF and CALBF were stained with the first
and second antibodies, and E. coli cells harboring plasmid
pHLA were used as the control strain for flow cytometry.
Both AmyAF- and CALBF-displaying cells showed a Fig. 4 Flow cytometric analysis of E. coli harboring the plasmid a
significantly higher intensity of fluorescence signal than pHLAαAF and b pHLA-CALBF. Transformants were grown at
37°C for 24 h in LB medium containing 100 mg/l of ampicillin.
the control cells. This result is consistent with that shown in Cells were labeled with primary mouse anti-FLAG M2, followed by
Fig. 3, and together, the results indicate successful cell secondary goat anti-mouse IgG conjugated with Alexa Fluor 488. In
surface display of the enzymes. each experiment, 10,000 cells were analyzed
 569
1500 2.5
α-amylase activity (U/g dry cell weight)
no detectable enzyme activity was observed in the culture
medium. The activity of cell-surface-displayed enzyme
1200 2.0 was not changed (AmyA) or slightly improved (CALB) by
lyophilization (data not shown). One gram of dry cells per
liter of culture medium corresponds to OD600 of 0.86.

OD600 (-)
900 1.5 In all the experiments described above, the cells reached
stationary phase after 12–16 h cultivation (Figs. 5 and 6).
600 1.0 The enzymes were produced and accumulated on the cell
surface during both growth and stationary phases. The
activity of α-amylase increased significantly after the cell
300 0.5 growth reached stationary phase, while the activity of
CALB gradually increased. The amount of cells in the
0 0.0 stationary phase was slightly reduced by the cell surface
0 20 40 60 80 expression of enzymes.
Time (h) The protease accessibility assays were performed using
α-amylase- and lipase-displaying cells cultivated for 48
Fig. 5 Time course of α-amylase activity on cell surface (closed and 36 h, respectively. By the trypsin treatment, the α-
symbols) and cell growth (optical density at 600 nm) (open symbols).
E. coli harboring the plasmid pHLA (squares) and pHLAαAF amylase and lipase activities of cells were decreased to 49
(circles) were grown at 37°C in LB medium containing 100 mg/l and 21% of untreated cells, respectively. These results
of ampicillin. The data points represent the average of three indepen- indicate that a large part of enzymes was displayed on the
dent experiments cell surface.

Enzyme activity
Growth of α-amylase-displaying E. coli cells
Figure 5 shows the time course of α-amylase activity on utilizing cornstarch
the cell surface of E. coli harboring pHLAαAF. The α-
amylase-displaying E. coli cells were cultivated in LB In order to confirm the starch-degrading ability of α-
medium containing 20 g/l soluble starch in a 2-l fermentor. amylase-displaying E. coli cells, their growth utilizing low-
To retain enzyme activity on the cell surface, the pH was temperature-cooked cornstarch was examined. M9 growth
adjusted to 6.0 by automatic addition of 5.0 N NaOH. The medium containing 20 g/l cornstarch as the sole carbon
α-amylase activity reached the maximum value of 990 U/g source was used to examine the cell growth. The time
dry cells at 48 h. The expression experiment with the PgsA- course of the viable-cell count and reducing-sugar con-
CALBF fusion protein was performed by flask cultivation centration are shown in Fig. 7. The E. coli harboring
of E. coli recombinants. Figure 6 shows the time course of pHLAαAF was able to grow on cornstarch. Oligosaccha-
lipase activity in whole cells of E. coli harboring pHLA- ride production was observed during the growth of the
CALBF. Lipase activity reached the maximum value of
4.6 U/g dry cells at 36 h. In both enzyme-display systems,
108 10

Reducing sugar concentration (g/l)

7 2.5
Lipase activity (U/g dry cell weight)

Viable cell count (CFU/ml)

8
6 107
2.0
5
6
OD600 (-)

4 1.5 106

3 4
1.0
2 105
2
0.5
1
104 0
0 0.0
0 10 20 30 40 50 0 10 20 30 40
Time (h) Time (h)
Fig. 6 Time course of lipase activity on cell surface (closed symbols) Fig. 7 Time course of viable-cell count (open symbols) and re-
and cell growth (optical density at 600 nm) (open symbols). E. coli ducing-sugar concentration (closed symbols). E. coli harboring the
harboring the plasmid pHLA (squares) and pHLA-CALBF (circles) plasmid pHLA (squares) and pHLAαAF (circles) were grown at
were grown at 37°C in LB medium containing 100 mg/l of am- 37°C in M9 growth medium containing 20 g/l low-temperature-
picillin. The data points represent the average of three independent cooked cornstarch. The data points represent the average of three
experiments independent experiments
570
Fig. 8 a Scheme of enantio- a
selective esterification from
(RS)-1-RE to (R)-1-PEA using OH O
CALB-displaying cells and with CALB
VA as acyl donor. b Time course
of concentration of reaction
+ O

products and enantiomeric ex- (RS)-1-phenylethanol vinyl acetate

cess. Reaction was carried out at
37°C by adding 50 mg lyophi-
lized cells into substrate solution OCOCH3 OH
consisting of 50 mg (RS)-1-PE,
35.3 mg VA, 0.5 g molecular
sieves 4A beads, and 5 ml
+ + CH3CHO

n-hexane. Symbols: squares, (R)-1-phenylethyl acetate (S)-1-phenylethanol acetaldehyde

(S)-1-PEA; circles, (R)-1-PEA;
triangles, enantiomeric excess
(%). The data points represent
the average of three independent b
experiments 100 100

Enantiomeric excess (%ee)

80 98

60
Yield (%)

96

40 94

20 92

0 90
0 5 10 15 20 25 30
Reaction time (h)

α-amylase-displaying cells and reducing-sugar concentra-
tion reached 8.5 g/l after 12 h.
Bioconversion using lipase-displaying E. coli cells
as whole-cell biocatalysts
In order to confirm the performance of the CALB-dis-
playing E. coli whole-cell biocatalyst, optical resolution of
(RS)-1-PE was carried out. (R)-1-PEA was selectively syn-
thesized by reaction of VA as an acyl donor with (RS)-1-PE
(Fig. 8a). Figure 8b shows the time course of (R)-1-PEA
production from 1% (RS)-1-PE using CALB-displaying
cells as whole-cell biocatalyst. After 25 h of reaction, the
yield of (R)-1-PEA reached 76.3% with an enantiomeric
excess of 98.2%.
Discussion
We constructed a novel E. coli cell surface display system
for large proteins. There have been many research projects
aimed at the development of an E. coli cell surface display
system for biotechnological and industrial applications
(Jose et al. 2002; Jose and von Schwichow 2004; Jung et al.
1998; Lee et al. 2004; Richins et al. 1997), but the systems
developed so far cannot display large target proteins of
over 60 kDa and have serious folding problem. Therefore,
they have limited uses. In the present study, we success-
fully developed a novel cell surface display system using
PgsA anchor protein which is suitable for displaying
proteins of large molecular size in an active form. This
anchor protein is a synthetase of poly-γ-glutamic acid
(PGA) with a molecular mass of over 1,000 kDa derived
from Bacillus subtilis chungkookjang (Ashiuchi and
Misono 2002). It is thought to be located between the
cell surface and the membrane of B. subtilis and to be a
transporter of PGA (Ashiuchi et al. 2001). We attempted
the display on the E. coli cell surface of AmyA, namely, a
protein of unprecedentedly large size in the active form.
We also attempted the display on an industrially useful en-
zyme, CALB, which has never been successfully expressed
in E.coli.
In our new surface display system using PgsA protein,
the PgsA anchor can be fused to the N terminus of the
AmyA and CALB proteins (Figs. 1 and 2). Moreover, both
the PgsA-AmyAF and PgsA-CALBF fusion proteins were
demonstrated by immunofluorescence labeling to be suc-
cessfully displayed on the cell surface (Figs. 3 and 4). In
addition, protease accessibility assay confirmed that a large
part of enzymes was displayed on the cell surface. In pre-
vious reports, the largest target protein displayed (49.9 kDa)

was anchored using OprF protein (Lee et al. 2005). In the
present study, we succeeded in displaying for the first time
a target protein (AmyA-FLAG) of approximately 78 kDa
in size on the E. coli cell surface. Moreover, the target
proteins (AmyA-FLAG and CALB-FLAG) showed enzy-
matic activity. This system is expected to be effective for
N-terminal immobilization of target proteins which have
a binding domain or catalytic site near the C terminus.
The fusion position is important for retaining enzyme ac-
tivity on the surface. In the yeast cell surface display, fu-
sion of the C terminus of α-amylase with the N terminus of
the α-agglutinin anchor protein results in low α-amylase
activity, but fusion of the N terminus of α-amylase with
the C terminus of the Flo1 anchor protein results in high
α-amylase activity (Shigechi et al. 2004). In the cell sur-
face display of lipase, the fusion style is similarly impor-
tant (Matsumoto et al. 2004). In the PgsA anchor display
system, AmyA and CALB activities reached 990 U/g (dry
cell weight) and 4.6 U/g (dry cell weight), respectively
(Figs. 5 and 6). It should be noted that we succeeded in
functional expression of CALB using PgsA anchor protein
in E. coli for the first time. We also tried to express free
CALB in E. coli, but its lipase activity was not detected
(data not shown). PgsA anchor protein probably facilitates
the successful folding of CALB. The PgsA anchor display
system is thus effective in fixing the α-amylase and lipase
on the cell surface in the active form.
Since α-amylase from S. bovis 148 is able to hydrolyze
cornstarch (Narita et al. 2004), the growth of α-amylase-
displaying E. coli on cornstarch was examined. As shown
in Fig. 7, α-amylase-displaying E. coli is able to grow
utilizing cornstarch. During growth, the reducing-sugar
concentration reached 8.5 g/l, and glucose was not detected
(data not shown). The glucose produced from the reducing
sugar was probably consumed quickly. This finding dem-
onstrates that E. coli cells induced to display AmyA by
fusion of PgsA to the N terminus have cornstarch-de-
grading ability. In the future, useful substances could be
produced from biomass by displaying enzymes such as
α-amylase on the surface of E. coli.
E. coli cells displaying lipase on their surface are used
as whole-cell biocatalyst for enantioselective transesterifi-
cation reaction. This reaction has been studied using re-
combinant fungi, yeast, and E. coli, which express lipase
such as Aspergillus oryzae, C. antarctica, Bacillus sp., and
Rhizopus oryzae (Kaieda et al. 2004; Kano et al. 1997; Lee
et al. 2004; Matsumoto et al. 2004). With cell-surface-
displayed CALB, the substrate molecules were able to
access CALB readily, and no treatment was needed to
catalyze the enantioselective transesterification reaction.
As shown in Fig. 8b, the yield and enantiomeric excess
were similar to those reported by Matsumoto et al. (2004)
(25 h, 63% yield, and 93% ee) using lipase-displayed yeast
cells. This result demonstrates that the lipase-displaying
E. coli cells can catalyze their enantioselective transesteri-
fication potential as whole-cell biocatalyst for the produc-
tion of various chemicals.
In our E. coli cell surface display system based on PgsA,
a large fusion protein of approximately 78 kDa in size and a
571
protein which is difficult to express in an active form using
the conventional methods were successfully displayed in
active form. Therefore, the system is applicable to a wide
range of biotechnological fields, such as bioconversions
using whole-cell biocatalysts and development of biosen-
sors. In addition, this display system would be effective for
protein library screening to supply novel proteins with new
functions.
Acknowledgements The present study was financed by the 2003
regional innovative consortium project of the Ministry of Economy,
Trade and Industry, Japan.
References
Ashiuchi M, Misono H (2002) Biochemistry and molecular genetics
of poly-γ-glutamate synthesis. Appl Microbiol Biotechnol 59:
9–14
Ashiuchi M, Nawa C, Kamei T, Song JJ, Hong SP, Sung MH, Soda
K, Yagi T, Misono H (2001) Physiological and biochemical
characteristics of poly γ-glutamate synthetase complex of
Bacillus subtilis. Eur J Biochem 268:5321–5328
Benhar I (2001) Biotechnological applications of phage and cell
display. Biotechnol Adv 19:1–33
Chang HH, Lo SJ (2000) Modification with a phosphorylation tag of
PKA in the TraT-based display vector of Escherichia coli. J
Biotechnol 78:115–122
Dubois M, Gilles KA, Hamilton JK, Rebers PA, Smith F (1956)
Colorimetric method for determination of sugars and related
substances. Anal Chem 28:350–356
Georgiou G, Stephens DL, Stathopoulos C, Poetschke HL,
Mendenhall J, Earhart CF (1996) Display of β-lactamase on
the Escherichia coli surface: outer membrane phenotypes con-
ferred by Lpp'-OmpA'-β-lactamase fusions. Protein Eng 9:
239–247
Georgiou G, Stathopoulos C, Daugherty PS, Nayak AR, Iverson BL,
Curtiss R III (1997) Display of heterologous proteins on the
surface of microorganisms: from the screening of combinatorial
libraries to live recombinant vaccines. Nat Biotechnol 15:29–34
Jose J, von Schwichow S (2004) Autodisplay of active sorbitol
dehydrogenase (SDH) yields a whole cell biocatalyst for the
synthesis of rare sugars. Chembiochem 5:491–499
Jose J, Bernhardt R, Hannemann F (2002) Cellular surface display
of dimeric Adx and whole cell P450-mediated steroid synthesis
on E. coli. J Biotechnol 95:257–268
Jung HC, Lebeault JM, Pan JG (1998) Surface display of
Zymomonas mobilis levansucrase by using the ice-nucleation
protein of Pseudomonas syringae. Nat Biotechnol 16:576–580
Kaieda M, Nagayoshi M, Hama S, Kondo A, Fukuda H (2004)
Enantioselective transesterification using immobilized Asper-
gillus oryzae overexpressing lipase. Appl Microbiol Biotechnol
65:301–305
Kano K, Negi S, Kawashima A, Nakamura K (1997) Optical
resolution of 1-arylethanols using transesterification catalyzed
by lipases. Enantiomer 2:261–266
Lee SY, Choi JH, Xu Z (2003) Microbial cell-surface display.
Trends Biotechnol 21:45–52
Lee SH, Choi JI, Park SJ, Lee SY, Park BC (2004) Display of
bacterial lipase on the Escherichia coli cell surface by using
FadL as an anchoring motif and use of the enzyme in enan-
tioselective biocatalysis. Appl Environ Microbiol 70:5074–
5080
Lee SH, Choi JI, Han MJ, Choi JH, Lee SY (2005) Display of lipase
on the cell surface of Escherichia coli using OprF as an anchor
and its application to enantioselective resolution in organic
solvent. Biotechnol Bioeng 90:223–230
572
Matsumoto T, Ito M, Fukuda H, Kondo A (2004) Enantioselective
transesterification using lipase-displaying yeast whole-cell bio-
catalyst. Appl Microbiol Biotechnol 64:481–485
Narita J, Nakahara S, Fukuda H, Kondo A (2004) Efficient
production of L-(+)-lactic acid from raw starch by Streptococ-
cus bovis 148. J Biosci Bioeng 97:423–425
Poo H, Song JJ, Hong SP, Choi YH, Yun SW, Kim JH, Lee SC, Lee
SG, Sung MH (2002) Novel high-level constitutive expression
system, pHCE vector, for a convenient and cost-effective sol-
uble production of human tumor necrosis factor-α. Biotechnol
Lett 24:1185–1189
Richins RD, Kaneva I, Mulchandani A, Chen W (1997) Biodeg-
radation of organophosphorus pesticides by surface-expressed
organophosphorus hydrolase. Nat Biotechnol 15:984–987
Robyt JF, Whelan WJ (1972) Reducing value methods for malto-
dextrins. I. Chain-length dependence of alkaline 3,5-dinitrosa-
licylate and chain-length independence of alkaline copper. Anal
Biochem 45:510–516
Satoh E, Niimura Y, Uchimura T, Kozaki M, Komagata K (1993)
Molecular cloning and expression of two α-amylase genes from
Streptococcus bovis 148 in Escherichia coli. Appl Environ
Microbiol 59:3669–3673
Shigechi H, Koh J, Fujita Y, Matsumoto T, Bito Y, Ueda M, Satoh E,
Fukuda H, Kondo A (2004) Direct production of ethanol from
raw corn starch via fermentation by use of a novel surface-en-
gineered yeast strain codisplaying glucoamylase and α-amylase.
Appl Environ Microbiol 70:5037–5040
Sung MH, Hong SP, Lee JS, Jung CM, Kim CJ, Soda K, Ashiuchi
M (2003) Surface expression vectors having pgsBCA, the gene
coding poly-gamma-glutamate synthetase, and a method for
expression of target protein at the surface of microorganism
using the vector. International Patent WO 03/014360
Uppenberg J, Hansen MT, Patkar S, Jones TA (1994) The sequence,
crystal structure determination and refinement of two crystal
forms of lipase B from Candida antarctica. Structure 2:293–
308
Veiga E, de Lorenzo V, Fernandez LA (2003) Autotransporters as
scaffolds for novel bacterial adhesins: surface properties of
Escherichia coli cells displaying Jun/Fos dimerization domains.
J Bacteriol 185:5585–5590
von Heijne G (1986) A new method for predicting signal sequence
cleavage sites. Nucleic Acids Res 14:4683–4690
Wan HM, Chang BY, Lin SC (2002) Anchorage of cyclodextrin
glucanotransferase on the outer membrane of Escherichia coli.
Biotechnol Bioeng 79:457–464
Westerlund-Wikström B, Tanskanen J, Virkola R, Hacker J,
Lindberg M, Skurnik M, Korhonen TK (1997) Functional ex-
pression of adhesive peptides as fusions to Escherichia coli
flagellin. Protein Eng 10:1319–1326