Molecular and catalytic properties of monoacetylphloroglucinol acetyltransferase from pseudomonas sp. YGJ3

Article abstract of DOI:10.1271/bbb.110860

Monoacetylphloroglucinol (MAPG) acetyltransferase, catalyzing the conversion of MAPG to 2,4-diacetylphloroglucinol (DAPG), was purified from Pseudomonas sp. YGJ3 grown without Cl^-. Cl^- and pyoluteorin repressed expression of the enzyme. SDS-polyacrylamide gel electrophoresis showed that the purified enzyme (M_r = 330 kDa) was composed of three subunits of 17, 38, and 43 kDa, and protein sequencing identified these as PhlB, PhlA, and PhlC respectively. The enzyme catalyzed the reversible disproportionation of 2 moles of MAPG to phloroglucinol (PG) and DAPG. The equilibrium constant K (=[DAPG][PG]/[MAPG]²) was estimated to be about 1.0 at 25°C. A KpnI 20-kb DNA fragment was cloned from the genomic DNA of strain YGJ3, and a 12,598-bp long DNA region containing the phl gene cluster phlACBDEFGHI was sequenced. PCR cloning and expression of the phl genes in Escherichia coli confirmed that expression of phlACB genes produced MAPG ATase.

Full text of DOI:10.1271/bbb.110860

Biosci. Biotechnol. Biochem., 76 (3), 559–566, 2012
Molecular and Catalytic Properties of Monoacetylphloroglucinol
Acetyltransferase from Pseudomonas sp. YGJ3
Asuka HAYASHI,1 Hiroki SAITOU,1 Tomomi MORI,1 Ikue MATANO,1
y
1;
Hiroyuki SUGISAKI,2 and Kiyofumi MARUYAMA
1Department of Biomolecular Science, Faculty of Engineering, Gifu University, 1-1 Yanagido, Gifu 501-1193, Japan
2Institute for Chemical Research, Kyoto University, Uji, Kyoto 611-0011, Japan
Received November 10, 2011; Accepted December 20, 2011; Online Publication, March 7, 2012
[doi:10.1271/bbb.110860]
Monoacetylphloroglucinol (MAPG) acetyltransfer-
ase, catalyzing the conversion of MAPG to 2,4-diacetyl-
phloroglucinol (DAPG), was purified from Pseudomonas
sp. YGJ3 grown without Clꢀ. Clꢀ and pyoluteorin
repressed expression of the enzyme. SDS-polyacryl-
amide gel electrophoresis showed that the purified
enzyme (Mr ¼ 330 kDa) was composed of three subunits
of 17, 38, and 43 kDa, and protein sequencing identified
these as PhlB, PhlA, and PhlC respectively. The enzyme
catalyzed the reversible disproportionation of 2 moles of
MAPG to phloroglucinol (PG) and DAPG. The equili-
brium constant K (¼[DAPG][PG]/[MAPG]2) was esti-
mated to be about 1.0 at 25 ꢁC. A KpnI 20-kb DNA
fragment was cloned from the genomic DNA of strain
YGJ3, and a 12,598-bp long DNA region containing the
phl gene cluster phlACBDEFGHI was sequenced. PCR
cloning and expression of the phl genes in Escherichia
coli confirmed that expression of phlACB genes pro-
duced MAPG ATase.
cluster phlACBDEFGHI.7,11–14) Among these, genes
phlEFGH are not directly involved in the synthetic
process of DAPG: PhlE and PhlG are a putative
permease and a DAPG hydrolase respectively,13–15)
and PhlF and PhlH are DAPG synthetic pathway-
specific transcriptional regulators.12,14) PhlI is a puta-
tive uncharacterized protein (UniProt accession no.
C0J9E0).7) Bangera and Thomashow cloned the
phlACBDEF genes from Pseudomonas fluorescens Q2-
87, and found that DAPG biosynthesis depended on
the phlACBD genes.11) Based on qualitative analysis of
gene function, they proposed a possible mechanism for
DAPG biosynthesis: PhlA, PhlC, and PhlB jointly
catalyze acyl transfer both at the initial step, the
formation of acetoacetyl-CoA from acetyl-CoA, and at
the final step, the acetylation of monoacetylphloroglu-
cinol (MAPG) to DAPG. PhlD which shows a high level
of homology to chalcone synthase, a type III-polyketide
synthase,16,17) is assumed to catalyze the elongation
through sequential condensation of malonyl-CoA to
acetoacetyl-CoA and the cyclization of the polyketide
chain to MAPG. Shanahan et al. measured the enzy-
matic acetylation of MAPG to DAPG by HPLC with a
cell-free extract of Pseudomonas sp. F113, and named
the enzyme MAPG acetyltransferase (ATase).18) Al-
though PhlA, PhlC, and PhlB are necessary for MAPG
ATase activity, the correct characterization of MAPG
ATase molecule remains unknown.
Achkar et al. found enzymatic evidence for the
formation of a polyketide chain.19) They cloned the
phlACBDE genes from P. fluorescens Pf-5 and found
that recombinant Escherichia coli carrying a plasmid
with a phlD insert markedly produced phloroglucinol
(PG), a condensation product of three molecules of
malonyl-CoA. Subsequently, Zha et al. prepared a PhlD-
His6 (PhlD whose C-terminus is tagged with His6), and
found that PhlD-His6 catalyzed the synthesis of PG from
malonyl-CoA.20) Hence it is expected that MAPG ATase
is involved in the sequential acetylation of PG to give
MAPG and DAPG. To test this possibility, a homoge-
neous preparation of MAPG ATase is required.
Key words: acetyltransferase; diacetylphloroglucinol;
disproportionation; monoacetylphlorogluci-
nol; phloroglucinol
Fluorescent pseudomonads produce a variety of
secondary metabolites that more or less affect microbial
growth in microflora.1) Secondary metabolites showing
strong antibiotic activity are useful in medical and
pharmaceutical applications.2) Polyketide 2,4-diacetyl-
phloroglucinol (DAPG), chlorinated pyrrole pyoluteorin,
and phenazine-1-carboxylate are particularly important
in agriculture due to their potent anti-fungal effects in
the rhizosphere.2–4) Among these, DAPG has a broad
spectrum of antifungal activity, and is a key compound
for the biocontrol activity of fluorescent pseudomo-
nads.5–7) Bacterial strains with a high capacity for DAPG
production effectively suppress the plant diseases caused
by pathogenic fungi such as take-all disease, found in
wheat, black root rot in tobacco, and the damping-off of
sugar-beet seedlings.8–10)
To establish a reliable biocontrol with fluorescent
pseudomonads, it is essential to understand fully the
mechanism and the regulation of DAPG biosynthesis.
DAPG is biosynthesized under the direction of phl gene
In a previous study, we isolated the pyoluteorin/
DAPG-producing bacterium Pseudomonas sp. YGJ3 and
found that strain YGJ3 mainly produced pyoluteorin
y
To whom correspondence should be addressed. Fax: +81-58-294-4633; E-mail: maruyama@gifu-u.ac.jp
Abbreviations: ATase, acetyltransferase; DAPG, 2,4-diacetylphloroglucinol; IPTG, isopropyl ꢀ-thiogalactopyranoside; MAPG, monoacetylphlor-
oglucinol; PAGE, polyacrylamide gel electrophoresis; PG, phloroglucinol

560
A. HAYASHI et al.
was treated with (NH4)2SO4. The precipitate obtained at 45–65%
saturation was dissolved in 2 mL of the buffer and placed on a column
(1:7 ꢂ 25 cm) of Superdex 200 equilibrated with the buffer, and 0.55-
mL fractions were collected. The active fractions (28–34) were pooled,
diluted 5-fold with H2O, and placed on a column (1:7 ꢂ 25 cm) of
hydroxyapatite (Bio-Rad, Hercules, CA) equilibrated with 10 mM
potassium phosphate (pH 7.0). After it was washed with 15 mL of
10 mM potassium phosphate (pH 7.0), the enzyme was eluted with a
linear gradient established with 60 mL of 10 mM potassium phosphate
(pH 7.0) and 60 mL of 500 mM potassium phosphate (pH 7.0), and
3-mL fractions were collected. The active fractions (28–31) were
pooled and stored on ice but not frozen, since freezing and thawing
would spoil the enzyme significantly.
Fig. 1. Acetyl Transfer Catalyzed by MAPG ATase.
during growth in the presence of Clꢀ, while it produced
DAPG in the absence of Clꢀ.21) Production of DAPG
was strongly inhibited by Clꢀ or pyoluteorin in the
culture medium, whereas Brꢀ and Iꢀ were almost
ineffective. Consistently with this, significant activity of
MAPG ATase was observed in cell-free extract prepared
from bacterial cells grown without Clꢀ. The activity
decreased markedly when strain YGJ3 was grown in the
presence of Clꢀ or pyoluteorin. In the present study,
based on these findings, Pseudomonas sp. YGJ3 was
cultivated in the absence of Clꢀ to achieve a high degree
of expression of MAPG ATase, and MAPG ATase was
purified from an extract of bacterial cells to examine its
molecular and catalytic properties. We found for the first
time that MAPG ATase is composed of PhlA, PhlC, and
PhlB, and that it catalyzes the disproportionation of 2
moles of MAPG to phloroglucinol (PG) and DAPG
(Fig. 1). Furthermore, a phl gene cluster was cloned
from strain YGJ3, confirming that genes phlACB encode
MAPG ATase.
Enzyme assay. In 50 mM potassium phosphate (pH 7.0), DAPG
showed an absorption peak at 370 nm, while MAPG and PG scarcely
showed any absorption at this wavelength. Hence the enzyme activity
was determined by measuring the increase (forward reaction) and the
decrease (reverse reaction) in A370 ("370 ¼ 6:0 mMꢀ1 cmꢀ1) with a
spectrophotometer UV-300 (Shimadzu, Kyoto, Japan). When the
activity was measured at
a different pH, the molar extinction
coefficient determined at that pH was used. The reaction was carried
out at 25 ꢁC in a cuvette (light path, 1 cm). The standard reaction
mixture (3 mL) contained 150 mmol of potassium phosphate (pH 7.0),
1.2 mmol of MAPG dissolved in 30 mL of ethanol, and the enzyme. The
activity of the reverse reaction was determined with a mixture (3 mL)
containing 150 mmol of potassium phosphate (pH 7.0), 0.3 mmol of
DAPG dissolved in 30 mL of ethanol, 6 mmol of PG, and the enzyme.
Unless otherwise noted, the enzyme activity was determined in the
forward reaction. One unit of enzyme activity was defined as the
amount catalyzing the formation or the disappearance of 1 mmol of
DAPG per min under the assay conditions. Specific activity was
expressed as units per mg of protein. Protein was measured by the
method of Lowry et al., with bovine serum albumin as standard.23)
Materials and Methods
Chemicals. MAPG and DAPG were purchased from Tokyo
Chemical Industry (Tokyo) and Toronto Chemical Research (Toronto,
Canada) respectively. Pyoluteorin was isolated from culture broth of
Pseudomonas sp. YGJ3 as described previously.21) All other chemicals
were of commercially available analytical grade.
Identification and quantification of reaction products. After the
enzyme reaction, the enzyme protein was removed by centrifugal
filtration with an Amicon Ultra 10K (Millipore, Carrigtwohill, Ireland).
The filtrate was applied to TLC with a silica gel 60 F254 (Merck,
Darmstadt, Germany), and CHCl3–MeOH (4:1, v/v) was used as
solvent. The spots of the products detected by UV were extracted with
H2O (for PG) and with ethanol (for MAPG and DAPG). The extracts
were applied again to silica gel TLC with authentic samples: the
Rf-values were 0.47, 0.67, and 0.82 for PG, MAPG, and DAPG
respectively.
Microorganisms. Pseudomonas sp. YGJ3 (Faculty of Engineering,
Gifu University) was aerobically cultivated at 30 ꢁC for 24 h in a
synthetic liquid medium in which Clꢀ was excluded and ethanol was
used as carbon and energy source, as described previously.21) Bacterial
cells were collected by centrifugation at 4 ꢁC, washed with 50 mM
potassium phosphate (pH 7.0), and stored at ꢀ25 ꢁC until needed.
E. coli XL1-Blue MRA as host for Charomid vector and XL1-Blue
MRF0 as host for the plasmid vector were from Stratagene (La Jolla,
CA).22) The former was grown at 37 ꢁC in Luria-Bertani (LB) medium
(polypeptone 20 g, yeast extract 10 g, and NaCl 20 g per 1 L of distilled
H2O, pH 7.0), and the latter in LB medium containing tetracyclin
(10 mg/mL).
For quantitative determination of MAPG and DAPG, the reaction
mixture (0.7 mL) at equilibrium was extracted 3 times with 0.7 mL of
CHCl3. The extract was evaporated at 25 ꢁC. The resulting residue was
dissolved in 200 mL of 50% methanol-H3PO4 (pH 3.0), and 10 mL of it
was analyzed by HPLC. To determine PG, the reaction mixture was
deproteinized by centrifugal filtration, and the filtrate (2.5 mL) was
analyzed by HPLC. HPLC was performed at 45 ꢁC with a LC-9A
HPLC system (Shimadzu) composed of a Shim-pack C18 reversed-
phase column (0:6 ꢂ 15 cm), a LC-9A pump, a SPD-6AV UV-Vis
monitor, and a C-R6A data processor. Elution was done with 50%-
MeOH–H3PO4 (pH 3.0) at a flow rate of 0.7 mL/min. A310 (for MAPG
and DAPG) and A210 (for PG) were monitored. Under these conditions,
the elution times were 4.7, 7.8, and 47 min for PG, MAPG, and DAPG
respectively.
Purification of MAPG ATase. Unless otherwise specified, all
manipulations were done at 0–4 ꢁC with 50 mM potassium phosphate
buffer (pH 7.0). Bacterial cells (20 g, wet weight) were suspended with
60 mL of the buffer and broken with a Sonifier (Branson, Danbury,
CT). After centrifugation at 30;000 ꢂ g for 15 min, the supernatant
solution (crude extract) was treated with (NH4)2SO4. The precipitate
obtained between 35 and 65% saturation was collected by centrifu-
gation and dissolved in about 18 mL of the buffer. Half of the solution
was put on a column (3 ꢂ 30 cm) of Superdex 200 (GE-Healthcare,
Buckinghamshire, UK) equilibrated with the buffer, and 3-mL
fractions were collected. At this step, most of the PhlG (DAPG
hydrolase) was separated from the MAPG ATase. The active fractions
(fractions 25–28) were pooled and placed on a column (1:2 ꢂ 12 cm)
of Q Sepharose (Pharmacia, Uppsala, Sweden) equilibrated with the
buffer. After it was washed with 200 mL of the buffer, the enzyme was
eluted with a gradient established with 60 mL of the buffer and 60 mL
of the buffer containing 1 M NaCl, and 3-mL fractions were collected.
The active fractions (12–14) were pooled. The other half of the
solution was treated similarly and the pooled solutions were combined.
The solution was diluted with the buffer so as to make A280 0.85, and
Analytical methods. Polyacrylamide gel electrophoresis (PAGE)
was performed with 5–15% Ready Gels J (Bio-Rad) and Davis’s
electrophoresis buffer.24) SDS–PAGE was done with 12.5% polyacryl-
amide gel, as described by Laemmli.25) Protein bands were stained with
Coomassie Brilliant Blue R-250. Densitometric measurement of
protein on the polyacrylamide gel was done with a Flying Spot
Scanner (Shimadzu). Electroblotting of protein from polyacrylamide
gel to polyvinylidene difluoride membrane was performed with semi-
dry blotting apparatus HorizBLOT AE-6675 (Atto, Tokyo), and the N-
terminal amino acid sequence was determined by Edman degradation
with a Procise 492 protein sequencing system (PE Applied Biosystems,
Foster City, CA). The molecular mass of the enzyme was estimated by
gel filtration with Superdex 200 (1:8 ꢂ 22 cm) previously equilibrated

Properties of Monoacetylphloroglucinol Acetyltransferase
561
Table 1. Primers for PCR Cloning of phl Genes
Primer
Sequencea
Target genes
5A
5C
5B
5D
3A
3C
3B
3D
50-TGCGGATCCGCGACCCAGTTTTCTGA-30
50-CCAGGATCCACGAGTTCAAGTACCTG-30
50-CGCGGATCCGCGTGTCGCAGAACCTT-30
50-TCGGGATCCCTTAACTTGTTGGCT-30
50-CGGGGTACCTTGCGTAGACAGGCGTA-30
50-CGGGGTACCCTGGGTACATGGACATG-30
50-ATGGGTACCTTCAAGGTGCCTGATAT-30
50-CGGGGTACCCCGGCAACGTCAGGT-30
phlA, phlAC, phlACB, phlACBD
phlC
phlB
phlD
phlA
phlC, phlAC
phlB, phlACB
phlD, phlACBD
aBamHI and KpnI sites are underlined.
with 50 mM potassium phosphate (pH 7.0) and calibrated with ferritin
(Mr ¼ 450 kDa), bovine liver catalase (230 kDa), yeast alcohol
dehydrogenase (150 kDa), bovine serum albumin (67 kDa), ovalbumin
(45 kDa), and chymotrypsin (23.5 kDa).
electrophoresis showed at least seven DNA fragments, at 6.3, 5.2, 3.9,
3.5, 2.8, 1.6, and 1.3 kb. Among these, the 6.3-kb fragment showed a
hybridization signal. The 6.3, 5.2, and 1.3-kb fragments were
introduced into SphI-digested pUC18 to give plasmids pUSS6.3,
pUSS5.2, and pUSS1.3 (Fig. 7). Partial DNA sequences were briefly
determined by cycle sequencing with M13 forward and reverse primers
DNA preparation and manipulation. The genomic DNA of
Pseudomonas sp. YGJ3 was prepared by the method of Smith,26)
except that the lysate of the bacterial cells was treated at 37 ꢁC for
30 min with protease K (1 mg/mL). A plasmid vector, pUC18, and a
cosmid vector, Charomid 9-28, were obtained from Takara (Ohtsu,
Japan) and Nippon Gene (Tokyo) respectively.27,28) Oligonucleotides
for sequencing primers and PCR primers and a hybridization probe
were obtained from Rikaken (Nagoya, Japan). On the basis of the N-
terminal amino acid sequence of a 38 kDa subunit of the purified
MAPG ATase and nucleotide sequence of phlA of P. fluorescens Pf-5
(GenBank accession no. CP000076, locus tag PFL 5954), an oligonu-
cleotide ONP1 (50-TATGGCGCGGGTATTCCGGTATGCCGCCTG-
AAA-30) was designed. ONP1 was labeled at the 30-terminal with
(Takara).
A comparison of the sequence data with those for
P. fluorescens Pf-5 suggested that these three plasmids completely
covered the phl gene cluster. The full DNA sequences of the three
DNA fragments were determined, and were put together to complete
the whole sequence (12,598-bp long) containing a phl gene cluster.
PCR cloning of phl genes. PCR amplification was carried out using
PrimSTAR GXL DNA polymerase (Takara) following the protocol
recommended by the manufacturer. Forward and reverse primers were
designed to amplify the entire open reading frames (ORFs) including
their ribosome binding sequences (Table 1). BamHI and KpnI sites
were built into the primers for subcloning. pCKK20 was used as
template DNA. The following primers flanking the target phl genes
were used: 5A and 3A for the cloning of phlA; 5C and 3C for phlC; 5B
and 3B for phlB; 5D and 3D for phlD; 5A and 3C for phlAC; 5A and
3B for phlACB; and 5A and 3D for phlACBD (Fig. 7). After digestion
with KpnI and BamHI, the amplified products were ligated into
BamHI/KpnI-digested pUC19 and transferred into E. coli XL1-Blue
MRF0.
digoxigenin (DIG) using
a DIG-tailing label kit (Boehringer
Mannheim, Mannheim, Germany), and used as probe for hybridization.
Southern and colony hybridizations were performed at 56 ꢁC. A
hybridization signal on a nylon membrane was detected with a DIG
luminescent detection kit (Boehringer Mannheim). All the enzymes
used in DNA manipulation were purchased from Takara. Plasmid
isolation and extraction of DNA from the agarose gel were done with a
QIAprep kit (Qiagen, Hilden, Germany) and Suprecꢀ-01 (Takara)
respectively. DNA digestion, ligation, and transformation were done as
described by Sambrook et al.29)
Gene expression in E. coli. A recombinant E. coli strain carrying
the plasmid with phl genes was cultured at 37 ꢁC in 100 mL of a LB
medium containing 50 mg/mL of ampicillin. When A660 reached about
0.15, 1 mM isopropyl ꢀ-thiogalactopyranoside (IPTG) was added, and
culturing was continued for a further 9 h. Bacterial cells were collected
by centrifugation at 4 ꢁC, washed once with 50 mM potassium
phosphate (pH 7.0), and resuspended in about 6 mL of the same
buffer. The cells were disrupted at 4 ꢁC by sonication. Cell debris was
removed by centrifugation at 30;000 ꢂ g for 15 min at 4 ꢁC, and the
supernatant (crude extract) was stored on ice to measure MAPG ATase
activity.
DNA sequencing. Deletion mutants for DNA sequencing were
prepared with a Kilo-sequence deletion kit (Takara) after digestion
using appropriate restriction enzymes. Both strands were sequenced on
denatured double-stranded DNA templates with a BigDye terminator
cycle sequencing kit (Applied Biosystems, Foster City, CA) and an
appropriate oligonucleotide primer (25mer). The nucleotide sequence
was determined with an ABI 3100 DNA sequencer (Perkin-Elmer,
Norwalk, CT). The nucleotide sequence reported in this paper has been
submitted to DDBJ/EMBL/GenBank Nucleotide Sequence Databases
under accession no. AB636682.
Results
Computational tools. The nucleotide sequence and the deduced
amino acid sequence were analyzed using the DNASIS program
(Hitachi Software Engineering, Yokohama, Japan). Alignment was
done using ClustalW (http://www.ddbj.nig.ac.jp/). The domain was
located with the Pfam domain data base (http://pfam.sanger.ac.uk/),
and hydrophobicity was judged with the Sosui WWW server (http://
bp.nuap.nagoya-u.ac.jp/sosui/sosui submit.html).
Purification of MAPG ATase
MAPG ATase was purified from a crude extract
of Pseudomonas sp. YGJ3 grown aerobically in the
absence of Clꢀ. Table 2 summarizes the purification
procedures, showing a 44-fold purification with a yield
of about 14%. The purified enzyme preparation showed
a single protein band on polyacrylamide gel electro-
phoresis (Fig. 2A). The protein band of MAPG ATase
was observable even in the crude extract, suggesting that
a significant amount of the enzyme protein was
synthesized in the bacterial cells grown without Clꢀ.
The addition of 2 mM Clꢀ or 50 mM pyoluteorin to the
culture medium strongly repressed the expression of the
enzyme as judged by decreased enzyme activity and
decreased protein band on PAGE, while Brꢀ (2 mM) or
Iꢀ (2 mM) was not effective (Fig. 2B).
Cloning of the phl gene cluster. The genomic DNA of strain YGJ3
was digested with KpnI. Southern hybridization of the KpnI-digest with
a DIG-labeled oligonucleotide probe ONP1 gave a single hybridization
signal on a 20-kb DNA fragment. The fragment was extracted from
the agarose gel and ligated into KpnI-digested Charomid 9-28. After
in vitro packaging with packaging extract Lambda inn (Nippon Gene),
followed by infection of E. coli XL1-Blue MRA, colony hybridization
gave hybridization-positive colonies. The plasmid DNA, designated
pCKK20, was isolated from one of these following the handling
manual of Charomid vector, and digested with SphI. Agarose gel

562
A. HAYASHI et al.
Table 2. Purification of MAPG ATase
Protein
(mg)
Sp. activity
(mU/mg)
Activity
(U)
Purification
(-fold)
Yield
(%)
Step
Crude extract
1,260
520
136
7.33
13.3
36.4
97.3
106
9.24
6.92
4.95
3.42
2.54
1.85
1.29
1
1.8
5.0
13
14
17
44
100
75
54
37
27
20
14
(NH4)2SO4 (I)
Superdex 200 (I)
Q Sepharose
35.1
24.0
15.2
4.0
(NH4)2SO4 (II)
Superdex 200 (II)
Hydroxyapatite
122
323
A
B
Fig. 3. SDS–PAGE of the Purification of MAPG ATase.
The samples described in the legend to Fig. 2A were applied in
lanes 1–7. Lane M, marker proteins.
Fig. 2. Polyacrylamide Gel Electrophoresis of MAPG ATase.
A, Purification of MAPG ATase. Lane 1, crude extract (35 mg of
protein); lane 2, first ammonium sulfate (31 mg); lane 3, Superdex
200 (14 mg); lane 4, Q Sepharose (10 mg); lane 5, second ammonium
sulfate (1.6 mg); lane 6, hydroxyapatite (2.5 mg); lane 7, hydroxya-
patite (1.3 mg). B, Effects of halide ions and pyoluteorin on the
expression of MAPG ATase. Pseudomonas sp. YGJ3 was grown
aerobically in the absence of halaid ion (lane 1), or in the presence
of 2 mM Iꢀ (lane 2), 2 mM Brꢀ (lane 3), 2 mM Clꢀ (lane 4), or 50 mM
pyoluteorin (lane 5). The crude extracts (50 mg protein), prepared as
described under ‘‘Materials and Methods,’’ except that the cell debris
was removed by brief centrifugation at 10;000 g for 10 min, were
analyzed. The relative enzyme activity of the crude extract is shown
at the bottom of the lane. Arrow shows the position of MAPG
ATase.
Molecular properties of MAPG ATase
Fig. 4. Identification of the Subunits of MAPG ATase Based on
N-Terminal Amino Acid Sequences.
The molecular mass of the enzyme was estimated to
be about 330 ꢃ 30 kDa by gel filtration with Superdex
200 (data not shown). On SDS–PAGE, the purified
enzyme showed three protein bands at 17, 38, and
43 kDa (Fig. 3). The N-terminal amino acid sequences
of these protein bands were determined by Edman
degradation, and were compared with those of PhlA,
PhlB, and PhlC of P. fluorescens (Fig. 4). The amino
acid sequences of the 17, 38, and 43 kDa proteins
matched PhlB, PhlA, and PhlC respectively of P. fluo-
rescens Pf-5 and P. fluorescens Q2-87. As described
below, the phl gene cluster of Pseudomonas sp. YGJ3
was cloned and sequenced. The N-terminal amino acid
sequences as determined by Edman degradation were
consistent with those deduced from the nucleotide
sequences of strain YGJ3, except that no N-terminal
residues or Cys residues were observed. These results
suggest that MAPG ATase is composed of PhlA, PhlC,
and PhlB. The PhlA, PhlC, and PhlB proteins obtained
from triple independent experiments were measured
densitometrically, and the molar ratio PhlA:PhlC:PhlB
was calculated to be 0:8 ꢃ 0:1:1:0 ꢃ 0:1:1:2 ꢃ 0:3 based
on the Mr of these proteins. Thus MAPG ATase appears
to be composed of equimolar amounts of PhlA, PhlC,
The N-terminal amino acid sequences of the subunits of MAPG
ATase were compared with those of PhlA, PhlB, and PhlC as
deduced from the nucleotide sequences of Pseudomonas sp. YGJ3
(GenBank accession no. AB636682), P. fluorescens Pf-5 (GenBank
accession no. CP000076, locus tag PFL 5954-5656), and P. fluo-
rescens Q2-87 (GenBank accession no. PFU41818). The sequences,
determined by Edman degradation of the purified MAPG ATase, are
indicated by bars above the amino acid sequences. Identical amino
acid residues are shown by stars.
and PhlB, and its subunit structure might be (PhlACB)3.
The purified enzyme showed an absorption maximum at
279 nm. The A279 of a 0.1% solution was estimated to
be 0.65, and the "279 was calculated to be 2:1 ꢂ
105 Mꢀ1 cmꢀ1. The shoulder at 290 nm due to the Trp
residue was not significant. The A280=A260 ratio was 2.0.
No significant absorption was detected in the visible
region.
Catalytic properties of MAPG ATase
Analysis of the reaction products by the silica gel
TLC and the HPLC on a C18 reversed-phase column
indicated that the enzyme yielded DAPG and PG from

Properties of Monoacetylphloroglucinol Acetyltransferase
563
Table 3. Estimation of the K Value of the MAPG ATase Reaction
A
C
B
D
Conc. at equilibrium (mM)
Substrate (conc. (mM))
Ka
MAPG
DAPG
PG
MAPG (0.05)
MAPG (0.1)
MAPG (0.2)
MAPG (0.4)
MAPG (0.6)
0.0132
0.0249
0.0576
0.107
0.0169
0.0283
0.0593
0.114
0.0091 0.88
0.0208 0.95
0.0564 1.01
0.118
0.172
1.97
1.17
1.02
1.27
1.12
0.93
0.177
0.186
0.004
DAPG (0.0333) þ PG (2) 0.0466
DAPG (0.1) þ PG (2)
0.168
0.0996
0.0161
0.0579
1.97
0.160
DAPG (0.1) þ PG (0.2)
2
aK ¼ ½DAPGꢅ½PGꢅ=½MAPGꢅ
A
B
Fig. 5. Spectral Changes in MAPG ATase Reactions.
In A (forward reaction), the enzyme (95 mg) was incubated with
0.1 mM MAPG. In B (reverse reaction), it (32 mg) was incubated with
0.1 mM DAPG and 2 mM PG. All other conditions are described
under ‘‘Materials and Methods.’’ Curves were taken at the indicated
times (min). C and D show the time courses of A370 in A and B
respectively.
MAPG in the forward reaction. Similarly, MAPG was
formed from DAPG plus PG in the reverse reaction.
When the enzyme reaction was monitored spectrophoto-
metrically, the absorption around 370 nm due to DAPG
gradually increased in the forward reaction and decreased
in the reverse reaction (Fig. 5). These results show that
MAPG ATase catalyzes the disproportionation of 2
moles of MAPG to DAPG and PG. The forward and
reverse reactions ceased before depletion of the sub-
strates, indicating the establishment of equilibrium. The
concentrations of MAPG, DAPG, and PG at equilibrium
were measured by HPLC (Table 3). The equilibrium
constant K (¼[DAPG][PG]/[MAPG]2) in the forward
reaction at 25 ꢁC with various concentrations (0.05–
0.6 mM) of MAPG was estimated to be about 1.0. The K
value (0.93–1.27) for the equilibrium established in the
reverse reaction with various concentrations of DAPG
and PG appeared to be consistent with this. DAPG was
the donor of an acetyl group, but not an acceptor, since
no decrease in DAPG was observed in incubation with
the enzyme. Resorcinol, but not pyrocatechol, served as
acetyl acceptor in place of PG in the reverse reaction.
At a fixed concentration (0.4 mM) of MAPG, activity
was proportional to the amount of the enzyme added
(0.024–0.19 mg). At a fixed amount of the enzyme
(0.14 mg), activity increased against the MAPG concen-
tration (0.1–1.2 mM) according to Michaelis-Menten
kinetics. The Km for MAPG and the Vmax were estimated
to be 0:26 ꢃ 0:01 mM and 0:66 ꢃ 0:01 mmol/min/mg
respectively by double reciprocal plot. The kcat was
calculated to be about 1.2 sꢀ1, assuming that the enzyme
(Mr ¼ 330 kDa) has three equivalent catalytic sites. In
the reverse reaction at various concentrations (0.005 mM–
0.10 mM) of DAPG and at a fixed concentration (2 mM)
of PG, the apparent Km (Kmapp) for DAPG and the
apparent Vmax (Vmaxapp) were estimated to be 0:019 ꢃ
0:001 mM and 1:55 ꢃ 0:12 mmol/min/mg respectively.
Fig. 6. pH Dependence of MAPG ATase.
The forward (A) and the reverse (B) reactions were measured in
the following buffers (each 50 mM): 2-morpholinoethanesulfonic
acid (MES) ( ); potassium phosphate ( ); HEPES ( ). The
activity obtained with potassium phosphate, pH 7.0, was taken to be
100%.
min/mg respectively at various concentrations (0.2–
3 mM) of PG and at a fixed concentration (0.1 mM) of
DAPG. Resorcinol was an active acetyl acceptor. The
app
app
Km
and a Vmax
were estimated to be 0:77 ꢃ
0:06 mM and 0:81 ꢃ 0:08 mmol/min/mg respectively.
DAPG inhibited the forward reaction competitively
(Ki ¼ 0:018 ꢃ 0:001 mM), while PG showed no signifi-
cant inhibition (Ki ꢄ 13 mM). Acetyl-CoA (1 mM) and
malonyl-CoA (1 mM) had no significant effects on
enzyme activity. The kinetic mechanism of the MAPG
ATase reaction could not be determined, because an
accurate assay in the reverse reaction with a low
concentration of DAPG was difficult.
Effects of temperature, pH, and inhibitors on MAPG
ATase
When the enzyme activity was measured at various
temperatures, it increased as the temperature increased
up to 65 ꢁC (data not shown). The enzyme was
preincubated in 50 mM potassium phosphate (pH 7.0)
for 20 min at various temperatures and then residual
activity was measured. The enzyme retained activity
below 40 ꢁC, but was inactivated at above 65 ꢁC,
suggesting that it was moderately thermostable.
The forward and the reverse reactions showed
optimum pH at about 6.3–7.2 (Fig. 6). The highest
activity was obtained with 2-morpholinoethanesulfonic
app
app
Similarly, the Km
for PG and the Vmax
were
determined to be 1:31 ꢃ 0:30 mM and 2:27 ꢃ 0:45 mmol/

564
A. HAYASHI et al.
Table 4. Genes Involved in DAPG Biosynthesis
Location of DNA
sequence
No. of amino
acid residues
Deduced Mr
(kDa)
Gene
Function
phlH
phlG
phlF
phlA
phlC
phlB
phlD
phlE
phlI
2,259–2,930, Ca
3,072–3,956
223
294
200
362
398
146
349
421
310
24.2
33.7
22.9
38.5
42.5
16.6
38.6
44.9
35.3
Regulatory protein14)
DAPG hydrolase14,15)
4,008–4,610, C
5,072–6,160
6,190–7,386
Regulatory protein12)
MAPG ATase 39 kDa subunit11)
MAPG ATase 43 kDa subunit11)
MAPG ATase 17 kDa subunit11)
Polyketide synthase11,19,20)
Permease11,13)
7,399–7,839
8,048–9,097
9,207–10,472
10,479–11,411
Putative uncharacterized protein7)
aComplement
acid (MES) and HEPES buffers. These buffers showed
some absorption at 210 nm, and interfered with the
measurement of PG by HPLC. Hence potassium
phosphate buffer was used throughout this study.
To examine the effects of various reagents on MAPG
ATase activity, the enzyme was preincubated with
various reagents at 25 ꢁC for 5 min and then residual
activity was measured. HgCl2 (0.1 mM) and p-chloro-
mercuribenzoate (1 mM) completely inactivated the
enzyme. It lost 88% of its activity after it was treated
with diethylpyrocarbonate (1 mM). Various metal com-
pounds (1 mM each), including NaCl, MgCl2, CaCl2,
FeCl3, ZnCl2, CuCl2, and NiCl2, and various metal
chelators (0.5 mM each), including 1,10-phenanthroline,
8-hydroxyquinoline, 2,20-dipyridine, and tiron (catechol-
3,5-disulfonic acid disodium salt), had almost no effect
on enzyme activity. Neither activation nor inhibition
was found with thiols (1 mM each) such as L-cysteine,
2-mercaptoethanol, and dithiothreitol.
Fig. 7. Organization of the phl Gene Cluster from Pseudomonas sp.
YGJ3.
The above horizontal lines show the insert DNA of plasmids
pUSS6.3, pUSS1.3, and pUSS5.2, which were used in sequencing a
12,598-bp long DNA region containing the phl gene cluster. phl
gene cluster phlHGFACBDEI is shown by bold arrows on the below
horizontal line, which shows a KpnI 20-kb DNA fragment cloned
into Charomid 9-28. ORFs indifferent as to phl genes are not shown.
The hybridization sites of forward primers (5A–5D) and reverse
primers (3A–3D) for PCR cloning are shown. The solid triangle
shows the position of a possible promoter. Several restriction sites
are marked: K, KpnI; S, SphI; X, XbaI.
Cloning and expression of phl genes
A KpnI 20-kb DNA fragment was cloned with
Charomid 9-28 from genomic DNA of Psedomonas sp.
YGJ3, as described above under ‘‘Materials and
Methods.’’ The resulting plasmid, pCKK20, was digested
with SphI. Of the DNA fragments formed, 6.3, 5.2, and
1.3-kb DNA fragments were subcloned into pUC18 for
DNA sequencing. Finally, the sequence of a 12,598-bp
long DNA region was determined to define the organ-
ization of DAPG biosynthetic locus phlHGFACBDEI
(Fig. 7). Table 4 summarizes the features of these genes.
A possible promoter sequence (TTAAGA for the ꢀ35
sequence and TATTAC for the ꢀ10 sequence) was
present in the region upstream of phlA (nucleotide
numbers 4,825–4,854). Although no ꢁ-independent
terminator sequence was found, another terminator must
have been present in the region downstream of phlI. The
deduced amino acid sequences of all the phl genes
agreed almost completely with those of P. fluorescens
Pf-5 (more than 99% identities), but were less homol-
ogous (80–93% identities) to those of P. fluorescens Q2-
87. PhlA had a 3-oxoacyl-acylcarrier protein synthase
(ACP synthase) III domain, as described by Bangera and
Thomashow,11) although the overall sequence homology
between Pseudomonas sp. YGJ3 PhlA and E. coli K12/
DH10B ACP synthase III (UniProt accession no.
B1XA01) was low (19% identical). PhlA lacked the
active site Cys and His residues, which are well
conserved in ACP synthase III of various species.30)
PhlC had a thiolase-like domain, including active site
His and Cys residues,31) although the overall sequence
homology between PhlC and E. coli K12/DH10B 3-
ketoacyl-CoA thiolase (UniProt accession no. B1X9L5)
was low (16% identity). PhlB of Pseudomonas sp. YGJ3
had a Cys-rich region: there were 5 Cys residues in the
region of residue no. 34–61). It had no known enzyme
domain, but a DUF (domain of unknown function)
35 OB-fold domain, which is frequently found in
DNA binding proteins.32) Analysis of hydrophobicity
indicated that PhlA, PhlC, PhlB, PhlD, PhlF, PhlG,
PhlH, and PhlI are soluble proteins, while PhlE is a
membrane protein.11)
To express phl genes in E. coli, PCR cloning with
pCKK20 as template was performed using appropriate
forward and reverse primers (Table 1 and Fig. 7). The
PCR products were digested with BamHI and KpnI and
introduced into pUC19 to construct plasmids pUphlA,
pUphlB, pUphlC, pUphlD, pUphlAC, pUphlACB, and
pUphlACBD. The nucleotide sequences of the DNA
inserts of these plasmids were confirmed by cycle
sequencing with appropriate sequencing primers. Ex-
pression of the phl genes in E. coli XL1-Blue MRF0 was
induced by the addition of 1 mM IPTG in a culture
medium. When the culture broth of the recombinant
E. coli was extracted with CHCl3 and analyzed by
HPLC, neither MAPG nor DAPG was detected. How-
ever, significant activity of MAPG ATase was found in

Properties of Monoacetylphloroglucinol Acetyltransferase
565
no significant activity hydrolyzing MAPG and DAPG
(data not shown), this was partly due to a lack of PhlE
(permease), but chiefly due to the low DAPG production
of the recombinant E. coli. Achkar et al. showed that
P. fluorescens Pf-5 transformed with a plasmid contain-
ing a phlACBDE insert produced large amounts of PG,
MAPG, and DAPG, while recombinant E. coli carrying
phlACBDE synthesized only a little MAPG, and
scarcely produced DAPG.19) Thus, in addition to
phlACBD, some other factors inherent in P. fluorescens
cells but limited in E. coli cells appear to be required to
facilitate DAPG biosynthesis. In connection with this, it
is probable that phlI plays a role in DAPG biosynthesis.
PhlA, PhlC, and PhlB are integral subunits of MAPG
ATase (Fig. 4). PhlA and PhlC have enzyme domains
related to an acetyl transfer, while the intrinsic function
of PhlB in enzymatic catalysis remains uncharacterized.
Since phlACB and phlACBD were expressed at higher
degrees in E. coli than phlA, phlC, phlB, phlD, and
phlAC (Fig. 8), phlB appeared to facilitate high-level
gene expression. The DUF 35 OB-fold in PhlB suggests
that PhlB somewhat interacts with DNA, although no
direct evidence for this has been obtained. MAPG
ATase activity was intricately affected by phlD. phlD
increased MAPG ATase activity by coexpression with
phlACB, while independent expression of phlD led to
the production of the inhibitors of MAPG ATase.
MAPG ATase catalyzes the transfer of acetyl groups
among DAPG, MAPG, and PG. This explains the
observation that recombinant E. coli carrying phlACB
produced no DAPG, MAPG, or PG, but stimulated their
interconversion when they were added exogenously into
the culture medium.19) If MAPG is biosynthesized by
PhlD, MAPG ATase can convert it to DAPG, but PhlD-
His6 synthesizes PG but not MAPG, as reported by Zha
et al.20) To synthesize MAPG and DAPG from PG,
MAPG ATase must be provided with a sufficient amount
of appropriate acetyl donor. Contrary to our expecta-
tions, acetyl-CoA appeared to be inactive as a donor
substrate for purified MAPG ATase. Achkar et al. have
reported that recombinant E. coli carrying phlACBD
produced a small amount of MAPG under culture
conditions different from those used in this study.18)
Hence there must be an acetyl donor in E. coli cells.
PG and its derivatives are found in various natural
products. PG is known to be formed by enzymatic
hydrolysis of phloretin,34) enzymatic trans-hydroxyla-
tion between pyrogallol and 1,2,3,5-trihydroxyben-
zene,35) and PhlD-catalyzed Claisen condensation of
malonyl-CoA.19,20) The MAPG ATase reaction described
in this study is a new biological source of PG.
Fig. 8. Expression of phl Genes in E. coli.
SDS–PAGE was done with a large gel (13:8 ꢂ 10 ꢂ 0:1 cm) to
analyze crude extracts (50 mg protein) of the following recombinant
E. coli: E. coli XL1-Blue MRF0 carrying pUC19 (lane 1), pUphlA
(lane 2), pUphlAC (lane 3), pUphlACB (lane 4), pUphlACBD
(lane 5), pUphlB (lane 6), pUphlC (lane 7), or pUphlD (lane 8).
Lane M, marker proteins; lane E, MAPG ATase (about 1 mg)
purified from Pseudomonas sp. YGJ3.
crude extracts from E. coli XL1-Blue MRF0 harboring
pUphlACB (specific activity, 7:6 ꢃ 0:4 mU/mg) and
pUphlACBD (16:6 ꢃ 1:1 mU/mg). The product of the
enzyme reaction was extracted, and was confirmed to be
DAPG by HPLC. No other recombinant E. coli showed
MAPG ATase activity. SDS–PAGE of the crude extract
from E. coli carrying pUphlA, pUphlC, pUphlB,
pUphlD, or pUphlAC showed faint bands of the
corresponding gene products. The crude extract from
E. coli carrying pUphlACB or pUphlACBD gave protein
bands characteristic of MAPG ATase, although the
bands of PhlC and PhlB were somewhat disturbed by
proteins intrinsic to E. coli (Fig. 8). Expression of
MAPG ATase was also confirmed by native PAGE
(data not shown). When crude extracts of E. coli
harboring pUphlA, pUphlC, pUphlB, and pUphlD were
mixed and incubated at 4 ꢁC for 5 min or for 12 h in the
presence and the absence of 1 mM 2-mercaptoethanol, no
MAPG ATase activity was reconstructed. The crude
extract of E. coli carrying pUphlD activated neither the
enzyme in the crude extract of E. coli carrying pU-
phlACB nor the enzyme purified from Pseudomonas sp.
YGJ3, but rather somewhat inhibited them, possibly due
to PhlD-derived metabolites (data not shown).
Discussion
DAPG and MAPG show similar UV spectra at
wavelengths longer than 300 nm in an organic solvent.33)
In neutral aqueous solution, DAPG but not MAPG
exhibits a significant absorption peak at 370 nm. Based
on this, we established a convenient spectrophotometric
assay for MAPG ATase and purified it from an extract of
Pseudomonas sp. YGJ3 to elucidate its properties. In
concurrence with the findings of Bangera and Thoma-
show, that phlACB is essential for the conversion of
MAPG to DAPG,11) the MAPG ATase molecule was
indeed composed of PhlA, PhlC, and PhlB (Fig. 4). PCR
cloning of the phlACB genes followed by expression in
E. coli further confirmed the molecular identity of
PhlACB with MAPG ATase.
In conclusion, we purified and characterized Pseudo-
monas sp. YGJ3 MAPG ATase for the first time. We
also cloned phl genes to confirm that phlACB is
necessary and sufficient for the synthesis of MAPG
ATase. In contrast to most acetyl transferases using CoA
derivatives as acyl donors or acceptors, MAPG ATase
catalyzes the transfer of acetyl groups without interme-
diation of CoA derivatives. It is an interesting problem
how this rare transferase catalyzes the cleavage and
formation of C–C bonds. It is probable that in vivo
MAPG ATase interacts with PhlD facilitating the
efficient transfer of intermediates between these func-
tionally related enzymes.
E. coli XL1-Blue MRF0 carrying pUphlACBD scarcely
produced MAPG or DAPG in the culture broth in spite
of high-level expression. Since recombinant E. coli had

566
A. HAYASHI et al.
16) Ferrer J-L, Jez JM, Bowman ME, Dixon RA, and Noel JP, Nat.
Struct. Biol., 6, 775–784 (1999).
The present study should prove useful in future
investigation of the interaction between MAPG ATase
and PhlD and of the structure and catalytic mechanism
of MAPG ATase.
17) Austin MB and Noel JP, Nat. Prod. Rep., 20, 79–110 (2003).
18) Shanahan P, Glennon JD, Crowley JJ, Donnelly DF, and O’Gara
F, Anal. Chim. Acta, 272, 271–277 (1993).
19) Achkar J, Xian M, Zhao H, and Frost JW, J. Am. Chem. Soc.,
127, 5332–5333 (2005).
References
20) Zha W, Rubin-Pitel SB, and Zhao H, J. Biol. Chem., 281,
32036–32047 (2006).
1) Leisinger T and Margraff R, Microbiol. Rev., 43, 422–442
(1979).
´
2) van Pee K-H and Ligon JM, Nat. Prod. Rep., 17, 157–164 (2000).
21) Matano I, Tsunekawa M, Shimizu S, Tanaka I, Mitsukura K,
and Maruyama K, Biosci. Biotechnol. Biochem., 74, 427–429
(2010).
3) Dowling DN and O’Gara F, Trends Biotechnol., 12, 133–141
(1994).
22) Jerpseth B, Greener A, Short JM, Viola J, and Kretz PL, Strat.
Mol. Biol., 5, 81–83 (1992).
4) Loper JE, Kobayashi DY, and Paulsen IT, Phytopathology, 97,
233–238 (2007).
23) Lowry OH, Rosebrough NJ, Farr AL, and Randall RJ, J. Biol.
Chem., 193, 265–275 (1951).
5) Nowak-Thompson B, Gould SJ, Kraus J, and Loper JE, Can. J.
Microbiol., 40, 1064–1066 (1994).
´
6) Keel C, Weller DM, Natsch A, Defago G, Cook RJ, and
24) Davis BJ, Ann. NY Acad. Sci., 121, 404–427 (1964).
25) Laemmli UK, Nature (London), 227, 680–685 (1970).
26) Smith MG, Methods Enzymol., 12A, 545–550 (1967).
27) Yanish-Perron C, Vieira J, and Messing J, Gene, 33, 103–119
(1985).
Thomashow LS, Appl. Environ. Microbiol., 62, 552–563 (1996).
7) Moynihan JA, Morrissey JP, Coppoolse ER, Stiekema WJ,
O’Gara F, and Boyd EF, Appl. Environ. Microbiol., 75, 2122–
2131 (2009).
28) Saito I and Stark GR, Proc. Natl. Acad. Sci. USA, 83, 8664–
8668 (1986).
8) Keel C, Schnider U, Maurhofer M, Voisard C, Laville J, Burger
´
U, Wirthner P, Haas D, and Defago G, Mol. Plant Microbe
29) Sambrook J, Fritsch EF, and Maniatis T, ‘‘Molecular Cloning, a
Laboratory Manual’’ 2nd ed., Cold Spring Harbor Laboratory,
Cold Spring Habor (1989).
Interact., 5, 4–13 (1992).
9) Laville J, Voisard C, Keel C, Maurhofer M, Defago G, and Haas
´
D, Proc. Natl. Acad. Sci. USA, 89, 1562–1566 (1992).
10) Fenton AM, Stephens PM, Crowley J, O’Callaghan M, and
O’Gara F, Appl. Environ. Microbiol., 58, 3873–3878 (1992).
11) Bangera MG and Thomashow LS, J. Bacteriol., 181, 3155–3163
(1999).
30) Khandekar SS, Gentry DR, van Aller GS, Warren P, Xiang H,
Silverman C, Doyle ML, Chambers PA, Konstantinidis AK,
Brandt M, Daines RA, and Lonsdale JT, J. Biol. Chem., 276,
30024–30030 (2001).
31) Jiang C, Kim SY, and Suh D-Y, Mol. Phylogenet. Evol., 49,
691–701 (2008).
12) Delany I, Sheehan MM, Fenton A, Bardin S, Aarons S, and
O’Gara F, Microbiology, 146, 537–546 (2000).
32) Muzin AG, EMBO J., 12, 861–867 (1993).
33) Marchand PA, Weller DM, and Bonsall RF, J. Agric. Food
Chem., 48, 1882–1887 (2000).
13) Abbas A, McGuire JE, Crowley D, Baysse C, Dow M, and
O’Gara F, Microbiology, 150, 2443–2450 (2004).
14) Bottiglieri M and Keel C, Appl. Environ. Microbiol., 72, 418–
427 (2006).
34) Chatterjee AK and Gibbins LN, J. Bacteriol., 100, 594–600
(1969).
35) Brune A and Schink B, J. Bacteriol., 172, 1070–1076 (1990).
15) He Y-X, Huang L, Xue Y, Fei X, Teng Y-B, Rubin-Pitel SB,
Zhao H, and Zhou C-Z, J. Biol. Chem., 285, 4603–4611 (2010).