Download Cloning and expression of the phosphotriesterase

Microbiology (2002), 148, 2687–2695
Printed in Great Britain
Cloning and expression of the
phosphotriesterase gene hocA from
Pseudomonas monteilii C11
Irene Horne, Tara D. Sutherland, John G. Oakeshott and Robyn J. Russell
Author for correspondence : Irene Horne. Tel : j61 2 6246 4110. Fax : j61 2 6246 4173.
e-mail : irene.horne!ento.csiro.au
CSIRO Entomology, GPO Box
1700, Canberra, ACT 2601,
Australia
The cloning of a gene encoding the novel phosphotriesterase from
Pseudomonas monteilii C11, which enabled it to use the organophosphate (OP)
coroxon as its sole phosphorus source, is described. The gene, called hocA
(hydrolysis of coroxon) consists of 501 bp and encodes a protein of 19 kDa. This
protein had no sequence similarity to any proteins in the SWISS-PROT/GenBank
databases. When a spectinomycin-resistance cassette was placed in this gene,
phosphotriesterase activity was abolished and P. monteilii C11 could no longer
grow with coroxon as the sole phosphorus source. Overexpression and
purification of HocA as a maltose-binding protein fusion produced a protein
having a broad substrate specificity across oxon and thion OPs.
Michaelis–Menten kinetics were observed with the oxon OPs, but not with the
thion OPs. End-product inhibition was observed for coroxon-hydrolytic activity.
Increased expression of hocA was observed from an integrative hocA–lacZ
fusion when cultures were grown in the absence of phosphate, suggesting that
it might be part of the Pho regulon, but the phosphate-regulated promoter
was not cloned in this study. This is believed to be the first study in which a
gene required for an organism to grow with OP pesticides as a phosphorus
source has been isolated.
Keywords : organophosphate hydrolase, coroxon
INTRODUCTION
Phosphatases and other phosphoester hydrolases are
produced by bacteria to cope with phosphate-limiting
conditions in the environment. These enzymes collectively hydrolyse a wide range of substrates to obtain
phosphate bound covalently in an organic form. They
are generally not required under phosphate-rich conditions, and the expression of many is repressed by
phosphate in the medium (Wanner, 1983). Furthermore,
phosphate can also inhibit enzymic activity (von Tigerstrom & Stelmaschuk, 1986).
Synthetic organophosphates (OPs) are widely used as
insecticides. OPs contain three phosphoester linkages,
and hydrolysis of any one of the phosphoester bonds
dramatically reduces the toxicity of the pesticides by
virtue of their inability to inactivate acetylcholin.................................................................................................................................................
Abbreviations : OP, organophosphate ; MBP, maltose-binding protein.
The GenBank accession number for the hocA gene is AF469117.
esterase. The OPs can be divided into two groups : those
containing a PUO bond (oxon OPs) and those containing a PUS bond (thion OPs). Several enzymes
capable of detoxifying OPs have been isolated from
micro-organisms capable of using OPs as carbon sources
(Dumas et al., 1989 ; Cheng et al., 1996) and are being
evaluated for application in various pesticide bioremediation processes.
The most widely characterized phosphotriesterase enzyme identified is the OPH protein from Flavobacterium
sp. ATCC 27551 (Mulbry & Karns, 1989). This bacterial
strain is capable of using the OP diazinon (O,O-diethyl
O-[6-methyl-2-(1-methylethyl)-4-pyrimidinyl]
phosphorothioate) as a carbon source (Sethunathan &
Yoshida, 1973). A similar gene was also identified in a
Pseudomonas strain capable of using methyl parathion
(O,O-dimethyl O-p-nitrophenyl phosphorothioate) as a
carbon source (Chaudhry et al., 1988). OPH is a 31 kDa
protein that is metal-requiring, with Zn(II) identified in
the native enzyme and a reaction mechanism that does
not involve a phosphorylated intermediate (Lewis et al.,
0002-5489 # 2002 SGM
2687
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:47:20
I. Horne and others
1988). The OPAA (organophosphate acid anhydrolase)
proteins from Alteromonas spp. are much larger, at
approximately 60 kDa. These proteins are also metalloenzymes [with Mn(II)] and have sequence similarity
with prolidases (Cheng et al., 1996). However, OPs have
not been reported to serve as a nutrient source for these
organisms. Two further OP-hydrolytic enzymes have
been identified in Nocardia-related species isolated for
their ability to use coumaphos (3-chloro-4-methyl-7coumarinyl diethyl phosphorothioate) as a carbon
source (Shelton & Somich, 1988). Mulbry (1992) identified a hydrolytic enzyme (AdpB) from a Nocardia strain
which possessed very different physical properties to
OPH, including the lack of a requirement for metal ions.
Mulbry (2000) subsequently identified a phosphotriesterase in the related Nocardiodes simplex NRRL B24074. This enzyme was suggested to be a metalloenzyme, and had a molecular mass of 45 kDa.
library construction, was performed according to the method
of Gardiner et al. (1996) : a size-fractionated genomic library
of P. monteilii C11 DNA was constructed. DNA fragments
(from a partial EcoRI digestion) in the range 12–20 kb were
cloned into the EcoRI site of plasmid pK18 and transformed
into E. coli DH10β. Tri-parental matings of the library clones
were performed with E. coli JM109 pR751 : : Tn813 as previously described (Horne et al., 2002a) but with P. aeruginosa
BENNY, a Pseudomonas strain that did not have OPhydrolytic activity, as the recipient and matings were performed on a reduced-phosphate medium (Horne et al., 2002b).
Conjugal transfer of plasmids from E. coli S17-1 into
Pseudomonas strains was carried out by the method of Bonnett
et al. (1995). All PCR reactions were performed using buffer
supplied by the manufacturer (Gibco-BRL) and 2 mM MgCl .
#
The DNA was denatured at 94 mC for 3 min, after which 0n5 U
Taq polymerase (Gibco-BRL) was added. Thirty cycles of
94 mC for 35 s, 48 mC for 70 s and 72 mC for 2 min were
routinely used, followed by a final extension period of 72 mC
for 5 min.
Many studies have identified bacterial strains capable of
using OPs as phosphorus sources (Cook et al.,
1978 ; Rosenberg & Alexander, 1979), but as yet no
genes involved in OP metabolism have been characterized from these organisms. A recent study by Horne
et al. (2002b) described the isolation of a Pseudomonas
monteilii strain (C11) capable of using the oxon OP
coroxon (3-chloro-4-methyl-7-coumarinyl diethyl phosphate) as a phosphorus source (with glucose as a carbon
source). P. monteilii C11 appeared to possess a single
enzyme capable of hydrolysing coumaphos and coroxon, and this enzyme was associated with the soluble
fraction of the cell. OP-hydrolytic activity appeared to
be due to a novel phosphotriesterase enzyme that was
not a metalloenzyme, and this activity was regulated by
the presence of phosphate in the medium. Here, we
describe the isolation of this novel phosphotriesterase
gene\enzyme system and present a preliminary characterization of its properties.
Assays and biochemical techniques. Protein concentrations in
METHODS
Strains and growth conditions. All strains and plasmids used
in this study are listed in Table 1. All cultures were routinely
grown with modified LB medium (Harcourt et al., 2002).
When grown on minimal media, Pseudomonas strains were
grown on the medium of Sutherland et al. (2000) with the
modification of Horne et al. (2002b). For the examination of
various phosphorus sources (which were added to a final
concentration of 100 µM) in the medium, phosphate was
omitted from this minimal medium, which was then buffered
with Tris\HCl (Horne et al., 2002b). Both Escherichia coli
and Pseudomonas strains were grown at 37 mC. When used in
the medium for E. coli, ampicillin, kanamycin, tetracycline
and trimethoprim were used at concentrations of 100, 25, 10
and 25 µg ml−", respectively. When included in the medium for
Pseudomonas, rifampicin, tetracycline, streptomycin and kanamycin were used at 100, 1, 25 and 25 µg ml−", respectively.
DNA manipulations. All standard DNA manipulations were
carried out according to Sambrook et al. (1989). For PCR
detection of insertional mutants, chromosomal DNA was
extracted according to the method of Rainey et al. (1992).
Isolation of chromosomal DNA from P. monteilii C11, for
crude extracts and pure proteins were determined by previously described methods (Bradford, 1976 ; Gill & von
Hippel, 1989). SDS-PAGE (10 % ; acrylamide\bisacrylamide
30 : 1) was performed according to the method of Laemmli
(1970). Hydrolysis of coroxon, coumaphos, O,O-dimethylumbelliferyl phosphate, 4-methylumbelliferyl phosphate, 4methylumbelliferyl acetate and 4-methylumbelliferyl β-galactopyranoside was measured by examining the formation
of fluorescent hydrolysis products (Roth, 1969 ; Harcourt et
al., 2002). Parathion (O,O-diethyl p-nitrophenyl phosphorothioate), paraoxon (O,O-diethyl p-nitrophenyl phosphate),
methyl parathion and bis-p-nitrophenyl phosphate hydrolyses
were determined colorimetrically by measuring the formation
of p-nitrophenol (Dumas et al., 1989). All assays were
performed in 50 mM Tris\HCl (pH 7n5) at 25 mC. The effect of
EDTA as a metal-chelator was tested by dialysis of pure
protein against three changes of EDTA (5 mM) in 50 mM
Tris\HCl (pH 7n5). When included in assays, MnSO , MgSO ,
%
%
ZnCl and CoCl were added to a final concentration
of
#
#
1 mM. All assays were performed at least twice on two
separate samples. Standard errors are displayed throughout
the text. Cell density measurements were performed as
previously described (Sutherland et al., 2000).
Construction of a plasmid for HocA overexpression and
purification. To construct a maltose-binding protein (MBP)–
HocA overexpression plasmid, the hocA gene was amplified
using PCR. The upstream and downstream oligonucleotide
primers, hoc5 (5h-GTCTAAGGATCCATGAAAGAAGAACTAAAAACC-3h) and hoc3 (5h-GTCTAAAAGCTTTTACCAGTTTAGCTTTAG-3h) contained, respectively, a BamHI
restriction site at the hocA start codon and a HindIII
restriction site at the stop codon (underlined bases). The PCR
fragment was subsequently cloned into the BamHI–HindIII
restriction sites of pMAL-c2X (New England Biolabs) to
generate the recombinant plasmid pmalhoc2. The correct
sequence of the insert was confirmed. Optimal production of
MBP–HocA was obtained when mid-exponential-phase cells
(OD 0n6) were induced with 0n1 mM IPTG for 5 h at 37 mC.
&*&
Harvested
cells were disrupted by sonication and the soluble
fraction loaded onto an amylose resin (New England Biolabs)
equilibrated with 50 mM Tris\HCl (pH 7n5). MBP–HocA was
eluted with 10 % maltose in 50 mM Tris\HCl (pH 7n5). The
purity of fractions was assessed by SDS-PAGE, and fractions
that appeared to be homogeneous were pooled. Approximately 720 µg pure protein was obtained from 500 ml culture.
2688
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:47:20
Phosphotriesterase from Pseudomonas monteilii C11
Table 1. Strains and plasmids used in this study
Strain or plasmid
E. coli strains
DH10β
S17-1
S17-1 λpir
JM109 λ
Relevant characteristics*
Source or reference
Strain for α-complementation cloning
Mobilizing strain ; carries chromosomally integrated derivative of RP4
S17–1 lysogenized with λpir
Host for α-complementation of cloning vectors (lysogenized with phage λ
carrying the pir gene ; allows for replication of R6K-based suicide vectors)
Gibco-BRL
Simon et al. (1983)
Penfold & Pemberton (1992)
Penfold & Pemberton (1992)
Pseudomonas strains
P. monteilii C11
P. monteilii HOC
P. monteilii C11lac
P. aeruginosa BENNY
Wild-type strain, Cor+
C11 hocA : : ΩSm\SpR, Cor−
C11 hocA : : lacZ
Wild-type strain
Horne et al. (2002b)
This study
This study
This study
Plasmids
pK18
pK16
pK265
pRK415
pRK7
pBSKS+
pBSRK7(1)
pBSRK7(3)
pMAL-c2X
pmalhoc2
pJP5603
pJPhoc
pJPhocSp
pEM32m
pBSlac
pC11lac
pC11lacmob
pR459II
pUI1188Sp
pR751 : : Tn813
pDSK519
pDSK-C11lac
pGEMT Easy
pGhocA5
pGhocA3(3)
pGhocA3j5
Cloning vector, KnR
pK18j12 kb of P. monteilii C11 DNA with hocA
pK18j12 kb P. monteilii C11 DNA in opposite orientation to pK16
Broad-host-range plasmid, TcR
pRK415j1n2 kb BamHI fragment with hocA
Cloning vector, ApR
pBSj1n2 kb BamHI fragment with hocA in same orientation as lacZ
pBSj1n2 kb BamHI fragment with hocA in opposite orientation as lacZ
Overexpression vector
pMAL-c2X with hocA in translational fusion with malE
R6K-based suicide vector
1n2 kb BamHI fragment from pGhocA3j5 in pJP5603
2n3 kb ΩSpR fragment from pUI1188Sp in XhoI site of pJPhoc
Tn5367 : : lacZ
pBSj3n0 kb ClaI–XbaI with lacZYA from pEM32
pBSlacj500 bp of hocA with upstream region
pC11lacj6n9 kb HindIII fragment from pR459II with mob-Tc
pK18j8 kb EcoRI fragment with mob-Tc
pBSKS+ with tandemly repeated double polylinker with ΩSm\Sp
Conjugative, cointegrative broad-host-range plasmid
Broad-host-range plasmid, KnR
3 kb PstI fragment with hocA–lacZ from pC11lac in pDSK519
T-tail vector
500 bp PCR fragment in pGEMT Easy
800 bp PCR fragment in pGEMT Easy
800 bp SpeI fragment from pGhocA3(3) in pGhocA5
W. Klipp
This study
This study
Keen et al. (1988)
This study
Stratagene
This study
This study
New England Biolabs
This study
Penfold & Pemberton (1992)
This study
This study
Machowski et al. (2000)
This study
This study
This study
W. Klipp
Eraso & Kaplan (1994)
Bowen & Pemberton (1985)
Keen et al. (1988)
This study
Promega
This study
This study
This study
* Abbreviations : Ap, ampicillin ; Cor, coroxon-hydrolytic activity ; Kn, kanamycin ; mob, mobilization genes ; ΩSpR, spectinomycinresistant cassette bordered by transcription terminators ; R, resistant ; Tc, tetracycline.
Construction of hocA–lacZ transcriptional fusions. A 3n0 kb
ClaI–XbaI lacZ-containing fragment from pEM32m
(Machowski et al., 2000) was ligated with similarly digested
pBluescript DNA to create plasmid pBSlac. The 5h region of
hocA (containing approximately 162 bp upstream of the hocA
start codon) and the upstream region were amplified by PCR
from pBSRK7(1), using the vector primer T3 (5h-AATTAACCCTCACTAAAGGG-3h) and the RK7T3r primer [5h-GATCCTCGAGTAAGGCTGATTGTTCAAGTTC-3h, containing a XhoI restriction site at the 5h end (underlined)]. This
generated a 500 bp PCR fragment that contained a XhoI site at
either end, the second generated from vector sequence. This
PCR fragment was digested with XhoI and ligated into
similarly digested pBSlac, and the correct orientation con-
firmed by digestion with PstI to produce a 3n2 kb fragment and
a 2n9 kb fragment. This plasmid was called pC11lac. A 6n9 kb
HindIII fragment containing the mobilization genes (mob)and
tetracycline-resistance cassette from pR459II was ligated into
pC11lac to create the integrative lacZ fusion plasmid,
pC11lacmob. A replicative transcriptional hocA–lacZ fusion
was also constructed. A 3n5 kb PstI fragment from pC11lac,
containing the 5h region of hocA with lacZYA, was cloned into
the broad-host-range plasmid pDSK519, to create
pDSKC11lac. Both integrative and replicative lacZ fusion
vectors were transferred into P. monteilii C11 using E. coli
S17-1.
Construction of a defective HocA mutant. The 5h PCR
fragment generated by amplification (see above) using the T3
2689
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:47:20
I. Horne and others
primer and RK7T3r (containing an XhoI restriction site at the
5h end), and using pBSRK7(1) as a template was ligated into
pGEM-T Easy (Promega) to generate the plasmid pGhocA5. A
3h region of hocA was also amplified by PCR with pBSRK7(1)
as a template, using the primers RK7T3f (5h-GAACTTGAACAATCAGCCTTA-3h) and the T7 vector primer (5h-TAATACGACTCACTATAGGGAGA-3h). This generated an 800 bp
fragment that was cloned into pGEM-T Easy to create
pGhoc3(3). Digestion of pGhoc3(3) with SpeI released the
800 bp PCR fragment, which was then ligated with SpeIdigested pGhocA5. Transformants were screened for correct
orientation by a positive PCR amplification using the hoc3 and
hoc5 primers. One was chosen and designated pGhocA5j3.
A 1n2 kb BamHI fragment from pGhocA5j3 that contained
both the 5h and 3h fragments and only one XhoI site (between
the 5h and 3h fragments) was ligated with similarly digested
pJP5603, to create pJPhoc. The streptomycin-resistance cassette from pUI1188Sp was cloned into the unique XhoI site in
pJPhoc to create pJPhocSp, which was transferred into P.
monteilii C11 using E. coli S17-1 λpir ; kanamycin-sensitive,
streptomycin-resistant exconjugants were selected. One exconjugant was chosen and shown by PCR to lack an intact
hocA gene. Rather than the 0n5 kb PCR product produced for
the wild-type strain (C11), a PCR fragment of 2n5 kb was
obtained, demonstrating the insertion of the 2 kb streptomycin-resistance cassette in hocA. The PCR was performed
using the hoc5 and hoc3 primers.
Chemicals. Methyl parathion and parathion were obtained
from Riedel-de Haan. Coumaphos (3-chloro-4-methyl-7-coumarinyl diethyl phosphorothioate) was a gift from Bayer.
Coroxon was purchased from Alltech. Paraoxon, bis-pnitrophenyl phosphate, 4-methylumbelliferyl phosphate, 4methylumbelliferyl acetate and 4-methylumbelliferyl β-galactopyranoside were obtained from Sigma. O,O-Dimethyl
4-methylumbelliferyl phosphate was synthesized by Professor
Alan Devonshire (A. L. Devonshire and others, unpublished).
RESULTS
Construction of a genomic library and subcloning of
the OP-hydrolytic gene
It was not known whether the phosphotriesterase gene
would be expressed adequately in E. coli to detect
enzymic activity in library clones. In a previous study, an
OP-degrading gene was successfully isolated from an
Agrobacterium strain by mobilizing a genomic library
into a related strain lacking activity (Horne et al.,
2002a). A similar approach was taken here, using
Pseudomonas aeruginosa BENNY, a strain previously
isolated in our laboratory and unable to hydrolyse OPs.
The library of P. monteilii C11 genomic DNA, constructed in pK18 in E. coli DH10β, was transferred to P.
aeruginosa BENNY in a tri-parental mating procedure.
Approximately 500 transformants were transferred.
Matings were assayed for coumaphos-hydrolytic activity, yielding two resultant positive clones (pK16 and
pK265), which were shown by restriction enzyme
analysis (using EcoRI, BamHI, HindIII and KpnI) to be
identical. Since neither of these clones conferred detectable coumaphos-hydrolytic activity on E. coli
DH10β, all subcloning was performed in the broad-hostrange plasmid pRK415 and transferred via E. coli S17-1
into P. aeruginosa BENNY. A subclone containing a
E
B
B
H
E
Clone
Activity in
P. aeruginosa
Benny
pRK16
+
pRK9
–
pRK5
–
pRK6
–
pRK8
+
pRK7(1)
+
hocA
1 kb
.................................................................................................................................................
Fig. 1. Restriction-enzyme map of the 12 kb EcoRI DNA
fragment in pK16, which conferred coumaphos-hydrolytic
activity on P. aeruginosa BENNY. B, BamHI ; E, EcoRI ; H, HindIII.
The various subclones that were tested for activity in P.
aeruginosa BENNY are indicated along with the 501 bp hocA
gene. The presence or absence of coumaphos-hydrolytic activity
conferred by these fragments is noted on the right-hand side.
1n2 kb BamHI fragment (Fig. 1) was obtained that
conferred coumaphos-hydrolytic activity on P. aeruginosa BENNY [820p7 pmol min−" (mg protein)−"] and
13-fold less activity on E. coli DH10β when expressed
from the lacZ promoter [62p2 pmol min−" (mg protein)−"].
The sequence of the 1.2 kb EcoRI–BamHI fragment
contains one ORF
Restriction mapping demonstrated that the 1n2 kb
BamHI fragment was contained at one end of the 12 kb
EcoRI fragment. Sequence from the opposite end, and of
other subclones of the EcoRI fragment, had similarity to
a segment of the P. aeruginosa PAO1 genome (83n5 % ;
Stover et al., 2000). The sequence of the 1n2 kb BamHI
fragment did not resemble any part of the P. aeruginosa
genome and contained a single ORF of 504 bp, revealed
using the GenScan program in Bionavigator (Burge &
Karlin, 1997). This ORF was preceded by a ribosomebinding site (AGGAG ; 7 bp upstream of the putative
ATG start codon) and, in constructs that produced
coumaphos-hydrolytic activity in E. coli DH10β, this
gene was in the same orientation as the lacZ promoter.
This ORF was designated hocA (hydrolysis of coroxon).
The GenBank accession number for the hocA gene is
AF469117. HocA is predicted to be 19n6 kDa and
composed of 167 amino acids. The protein has a
predicted pI of 6n50, and a hydropathy plot of this
protein by the method of Kyte & Doolittle (1982)
revealed no obvious membrane-spanning domains, indicating that it is likely to be a soluble protein, probably
contained in the cytoplasm. This is consistent with the
previous finding of Horne et al. (2002b), i.e. that OPhydrolytic activity in P. monteilii C11 is located in the
soluble fraction of the cell.
2690
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:47:20
Phosphotriesterase from Pseudomonas monteilii C11
(a)
Substrate
Km ( íM)
kcat (min–1)
68 ±13
0·12± 0·009
(b)
Coroxon
Substrate
[Substrate] (íM) Activity (min–1)
100
0·039 ±0·005
87
0·074 ±0·015
261
0·041±0·001
Coumaphos
166 ± 50
0·0084±0·0008
Parathion
dMUP
105 ± 36
0·58± 0·08
Methyl parathion
Paraoxon
.................................................................................................................................................................................................................................................................................................................
Fig. 2. (a) Kinetics of activity of HocA against oxon OP substrates. (b) Specific activity of HocA against thion OPs. dMUP,
O,O-dimethylumbelliferyl phosphate.
The HocA protein demonstrated no significant sequence
similarity to any of the previously identified OPhydrolytic enzymes, OPH\OpdA (GenBank accession
numbers M20392 and AY043245, respectively), OPAA2 (GenBank accession number U29240) or AdpB (GenBank accession number M91040). Some sequence similarity was observed with a putative cytoplasmic protein
from Salmonella typhimurium LT2 that was identified
in the genome sequencing project (GenBank accession
number AE008708). This putative protein showed 41 %
sequence identity at the amino acid level to HocA over
the entire protein, and 59 % sequence similarity. The
function of this putative Salmonella protein is not
known. No other strong sequence similarity to any other
proteins in the GenBank, PDB, EMBL, SWISS-PROT
and Prosite databases were observed using the 
program. However, some similarity (43 %) was seen
with isocitrate dehydrogenase from Methanococcus
jannaschii (Bult et al., 1996) over the entire protein
(23 % identity), while the central third showed 60 %
similarity (36 % identity) with phospholipase C from
Arabidopsis thaliana (Sato et al., 2000). Both of these
proteins have some involvement with phosphate. Phosphatidyl-inositol-specific phospholipase C catalyses the
cleavage of the membrane lipid, phosphatidylinositol or
its phosphorylated derivatives, to produce diacylglycerol
(Griffith & Ryan, 1999). Isocitrate dehydrogenase is an
enzyme that is regulated by phosphorylation of a serine
residue in its active site (Hurley et al., 1990). Key
residues in isocitrate dehydrogenase and phospholipase
C that bind phosphate are not conserved in HocA, nor
are key catalytic residues. It is likely that these sequence
similarities have no functional significance. No conserved domains or motifs were identified in this protein
using either EMBL or Prosite, and so this appears to be
a novel protein with no obvious homologues.
P. monteilii HOC, which contains an insertional mutation in hocA, showed limited phosphotriesterase
activity with coroxon as a substrate (1n6p0n3 % activity
of the wild-type). Furthermore, P. monteilii HOC was
unable to grow with coroxon as a phosphorus source.
This confirms that HocA is a protein required for P.
monteilii C11 to grow with coroxon as a phosphorus
source.
Purification and substrate specificity of HocA
To confirm that hocA did indeed encode a phosphotriesterase, the enzyme activity of an MBP fusion was
examined. The purified protein possessed phosphotriesterase activity against the substrates coumaphos,
coroxon, O,O-dimethylumbelliferyl phosphate, paraoxon, parathion and methyl parathion. The kinetics of
hydrolysis against the oxon OPs coroxon, paraoxon and
O,O-dimethylumbelliferyl phosphate are shown in Fig.
2a. Reactions with the thion OPs did not appear to
follow Michaelis–Menten kinetics (Fig. 3). This may be
due to the Km of the enzyme being quite low, and the
substrate concentrations tested being greater than the
Km. Specific activities for these substrates are approximately 30–50 % of the activity with oxon OPs at the
same substrate concentration (Fig. 2b). While it is
possible that fusion with MBP interfered with these
activities, the activity of untagged HocA expressed from
the lacZ promoter in E. coli DH10β (pBSRK7(1))
demonstrated the same Km for coroxon as for the fusion
protein, so we find this unlikely. In a previous study, the
2691
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:47:20
I. Horne and others
decreased with increased concentrations of diethyl
phosphate. This demonstrated that diethyl phosphate
produced a non-competitive inhibition with respect to
coroxon, and the Ki was determined to be 67 µM.
Interestingly, this is approximately the same as the Km
for the protein for coroxon, suggesting that HocA has as
much affinity for the hydrolysis product as it does for
coroxon. No inhibition was observed with inorganic
phosphate.
0·050
Activity (min–1)
0·040
0·030
0·020
.................................................................................................................................................
The effect of metal-chelators on the coroxon-hydrolytic
activity of HocA was examined because most of the
previously described phosphotriesterase enzymes appear
to be metalloenzymes. Extensive dialysis of MBP–HocA
against EDTA had no effect on coroxon-hydrolytic
activity. Nor did the metal ions Co(II), Zn(II), Mg(II) or
Mn(II) affect coroxon-hydrolytic activity when added
directly to the assay mixture.
Fig. 3. Reaction rates of HocA against coumaphos, a thion OP,
with increasing substrate concentration, demonstrating that
Michaelis–Menten kinetics are not observed.
Phosphate levels regulate hocA expression
0·010
40
80
120
160
200
[Coumaphos] (íM)
0·0030
1/Fluorescence (min)
0·0025
0·0020
0·0015
0·0010
0·0005
– 40
– 20
0
20
40
1/[Coroxon] (mM–1)
60
.................................................................................................................................................
Fig. 4. Lineweaver–Burk plots for coroxon-hydrolytic activity in
the presence of increasing amounts of diethyl phosphate (DEP).
specific phosphotriesterase activity in P. monteilii C11
was 1n33 nmol min−" (mg protein)−" in crude cell extracts (Horne et al., 2002b). Given that pure HocA has a
specific activity of 0n12 nmol min−" (nmol enzyme)−",
this would suggest that there is approximately 11n1 nmol
HocA (mg protein)−" in cells grown under those
conditions. The HocA protein did not show phosphatase, phosphodiesterase or carboxylesterase activity
against 4-methylumbelliferyl phosphate, bis-p-nitrophenyl phosphate or 4-methylumbelliferyl acetate, respectively.
Diethyl phosphate, the hydrolysis product of coroxon,
inhibited coroxon-hydrolytic activity (Fig. 4). While the
Km values showed little difference, the Vmax values
Previously, coumaphos-hydrolytic activity in P. monteilii C11 appeared to be regulated by phosphate levels
(Horne et al., 2002b). In many organisms, including the
pseudomonads, genes of the Pho regulon are either
induced or repressed by the PhoP-R sensor–regulator
pair upon phosphate starvation. The PhoR protein
binds to a distinctive segment of DNA (5h-TTGCAGTCTCGCTGTCACAA-3h). The region contained in the
1n2 kb EcoRI–BamHI fragment did not possess any
obvious promoter region or regulatory DNA sequence
such as the Pho DNA-binding sequence. To examine the
regulation of hocA expression, both an integrative and a
replicative lacZ fusion were constructed. With the
integrated hocA–lacZ fusion, transcription from upstream regions not contained in the 1n2 kb EcoRI–
BamHI fragment can contribute to hocA expression
(and therefore β-galactosidase expression). The β-galactosidase activity from this fusion was examined in media
containing either excess phosphate or 4-methylumbelliferyl phosphate (limited phosphate). The β-galactosidase activity in cells grown without phosphate was
twice that of cells grown with phosphate (109n4p
1n3 nmol min−" mg−"
compared
with
51n7p
0n2 nmol min−" mg−"). This suggests that expression of
the hocA gene is regulated by phosphate levels in the
medium.
A replicative lacZ fusion (pDSKC11lac) was used to
determine whether the 1n2 kb EcoRI–BamHI fragment
contained immediate upstream sequences involved in
phosphate regulation. Greater β-galactosidase activity
was observed from this construct in P. monteilii C11
compared with the integrative hocA–lacZ fusion, presumably because of a copy-number effect. Therefore, no
overall comparison can be made between the integrative
and replicative lacZ fusions with respect to the extent to
which upstream promoters contribute to overall hocA
expression. However, it is clear that the replicative
plasmid conveyed identical β-galactosidase activity in
cells grown with either phosphate (141n9p
2692
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:47:20
Phosphotriesterase from Pseudomonas monteilii C11
7n7 nmol min−" mg−") or 4-methylumbelliferyl phosphate (142n2p3n8 nmol min−" mg−"). This suggests that
regions involved in phosphate regulation are contained
on DNA sequences much further upstream of hocA than
those cloned in this study. The hocA gene may be
contained in a much larger operon that is regulated by
phosphate. Alternatively, the regulatory region is much
further upstream. For example, the ‘ Pho box ’ is almost
600 bp upstream of the phosphate-regulated phoA gene
of P. aeruginosa (Stover et al., 2000). It was not possible
to identify such upstream regulatory regions in this
study, as the cloned fragment containing the hocA gene
contained only 162 bp of sequence upstream of the start
codon.
DISCUSSION
A gene (hocA) was isolated from P. monteilii C11 ; its
product is required for this strain to grow with coroxon
as a phosphorus source. The hocA gene appears to be
contained on the chromosome of P. monteilii C11, as
sequences of subclones of the 12 kb EcoRI fragment
containing hocA had sequence similarity to sequences
on the P. aeruginosa genome. A hocA homologue was
not identified on the P. aeruginosa genome, and the P.
aeruginosa strain BENNY did not demonstrate phosphotriesterase activity. This gene is either unique to P.
monteilii, and possibly some other Pseudomonas strains,
or was acquired via lateral gene transfer. The 1n2 kb
BamHI fragment containing hocA has a 46n9 mol %
GjC content, in contrast to the 66n6 mol % GjC
content of the P. aeruginosa genome, which is typical of
pseudomonads (Stover et al., 2000). Differences in GjC
contents have been suggested to imply lateral gene
transfers (Folkesson et al., 1999). Pseudomonas strains
have been frequently reported to have acquired genes via
lateral gene transfer (Sinclair et al., 1986 ; Williams &
Murray, 1974) among soil microbial communities. This
may be the case here.
hydrolysis by P. monteilii C11 observed in our previous
study (Horne et al., 2002b). Furthermore, the activity is
quite low for these substrates and may not be seen until
the protein is in sufficient concentrations, as seen here
with purified enzyme. The unusual kinetics observed for
HocA with the thion OPs may be a result of end-product
inhibition by the phosphorothioates, diethyl thiophosphate and dimethyl thiophosphate, the hydrolysis products of coumaphos and methyl parathion, respectively.
Alternatively, the enzyme has a low Km for the thion
OPs, and substrates were tested at concentrations far
exceeding the Km of the enzyme. We find this unlikely,
as this would be the first report of a phosphotriesterase
having an extremely low Km for thion OPs relative to the
oxon versions. In general, the converse is true and the
enzymes have a higher affinity and catalytic activity for
the naturally occurring PUO-containing OPs rather
than the bigger, unnatural PUS-containing OPs.
The expression of hocA in P. monteilii C11 was
regulated by phosphate levels. While this suggests a role
for HocA in phosphate metabolism, it is highly unlikely
that the native role is to hydrolyse phosphotriesters,
because OPs have only been in existence for the last 50
years. HocA may have evolved from a pre-existing
phosphatase or phosphodiesterase as it has been shown
that a phosphotriesterase (OPH) could be altered by
only one amino acid to possess phosphodiesterase
activity (Shim et al., 1998), demonstrating how a minor
change can allow a protein to acquire a new activity.
Likewise, HocA may have been a pre-existing phosphodiesterase that has acquired phosphotriesterase activity.
P. monteilii C11 was isolated by virtue of its ability to
utilize OPs as a source of phosphorus. Phosphotriesterase activity is the first step in obtaining phosphate
from coroxon. The doubling time of P. monteilii C11
was not increased when inorganic phosphate, rather
than coroxon, was provided as the sole source of
phosphorus (data not shown), suggesting that phosphotriesterase activity is not a rate-limiting step in providing
inorganic phosphate for P. monteilii C11. This implies
that there is no requirement for a kinetically better
phosphotriesterase in this organism. HocA (the only
enzyme in this organism responsible for phosphotriesterase activity) is less efficient at hydrolysing OPs than
are other phosphotriesterases isolated from organisms
capable of using OPs as a carbon source. In general,
phosphotriesterase activity would be the first step in the
use of OPs as a carbon source. The OPH enzyme of
Flavobacterium sp. ATCC 27551 can hydrolyse paraoxon at rates close to the limits of diffusion. We find it
unlikely that the kinetics of HocA would be able to
support growth of an organism with OPs as a carbon
source. Bacteria require much higher levels of carbon
than phosphorus for growth. Stanier et al. (1986)
suggested that a Pseudomonas putida strain required
7 mM carbon, whereas 100-fold less phosphorus is
needed for growth ; therefore an enzyme that is providing the only source of phosphorus can be 100-fold
less active than an enzyme providing the only source of
carbon. In accordance with this, the kcat values of OPH
for both coroxon and paraoxon appear to be 100-fold
higher than those of HocA (based on the results of
Horne et al., 2002a). We propose that the isolation of
enzymes from organisms using OPs as carbon sources
rather than phosphorus sources potentially yields more
efficient enzymes.
It is common for phosphatases and phosphodiesterases
to be inhibited by their phosphorus-containing hydrolysis product (von Tigerstrom & Stelmaschuk, 1986).
End-product inhibition of HocA activity was observed
by the phosphorus-containing moeity. The severe endproduct inhibition of HocA by diethyl phosphate
probably contributed to the lack of parathion\paraoxon
The kinetics of the OPH\OpdA proteins suggest that
they are far better suited to bioremediation than HocA.
However, HocA does not require a metal ion. It might,
therefore, be better suited than OpdA for bioremediation in certain environments such as heavily polluted
waters containing inhibitory metals or metal-chelating
agents. It might be possible to improve HocA by in vitro
2693
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:47:20
I. Horne and others
evolution (MacBeath et al., 1998) for better kinetics and
reduced end-product inhibition.
ACKNOWLEDGEMENTS
Harcourt, R. L., Horne, I., Sutherland, T. D., Hammock, B. D.,
Russell, R. J. & Oakeshott, J. G. (2002). Development of a simple
and sensitive fluorimetric method for isolation of coumaphoshydrolysing bacteria. Lett Appl Microbiol 34, 263–268.
Horne, I., Sutherland, T. D., Harcourt, R. L., Russell, R. J. &
Oakeshott, J. G. (2002a). Identification of an opd (organophos-
We would like to thank Michelle Williams for technical
assistance in this study. This research was supported by Orica
Australia Ltd and Horticulture Australia Ltd (HG97034).
phate degradation) gene in an Agrobacterium isolate. Appl
Environ Microbiol (in press).
REFERENCES
Horne, I., Harcourt, R. L., Sutherland, T. D., Russell, R. J. &
Oakeshott, J. G. (2002b). Isolation of a Pseudomonas monteilli
Bonnett, T. C., Cobine, P., Sockett, R. E. & McEwan, A. G. (1995).
Phenotypic characterization and genetic complementation of
dimethylsulfoxide respiratory mutants of Rhodobacter sphaeroides and Rhodobacter capsulatus. FEMS Microbiol Lett 133,
163–168.
Bowen, A. R. St G. & Pemberton, J. M. (1985). Mercury resistance
transposon Tn813 mediates chromosome transfer in Rhodopseudomonas sphaeroides and intergeneric transfer in pBR322. In
Plasmids in Bacteria, pp. 105–115. Edited by D. R. Helsinki, S. N.
Cohen, D. B. Clewell, D. A. Jackson & A. Hollaender. New
York : Plenum.
Bradford, M. (1976). A rapid and sensitive method for the
quantitation of microgram quantities of protein utilising the
principle of protein-dye binding. Anal Biochem 72, 248–254.
Bult, C. J., White, O., Olsen, G. J. & 20 other authors (1996).
Complete genome sequence of the methanogenic archaeon,
Methanococcus jannaschii. Science 273, 1058–1073.
Burge, C. & Karlin, S. (1997). Prediction of complete gene
structures in human genomic DNA. J Mol Biol 268, 78–94.
Chaudhry, G. R., Ali, A. N. & Wheeler, W. B. (1988). Isolation of a
methyl parathion degrading Pseudomonas sp. that possesses
DNA homologous to the opd gene from a Flavobacterium sp.
Appl Environ Microbiol 54, 288–293.
Cheng, T.-C., Harvey, S. P. & Chen, G. L. (1996). Cloning and
expression of a gene encoding a bacterial enzyme for decontamination of organophosphorus nerve agents and nucleotide
sequence of the enzyme. Appl Environ Microbiol 62, 1636–1641.
Cook, A. M., Daughton, C. G. & Alexander, M. (1978). Phosphorus-containing pesticide breakdown products : quantitative
utilization as phosphorus sources by bacteria. Appl Environ
Microbiol 36, 668–672.
Dumas, D. P., Caldwell, S. R., Wild, J. R. & Raushel, F. M. (1989).
Purification and properties of the phosphotriesterase from
Pseudomonas diminuta. J Biol Chem 264, 19659–19665.
Eraso, J. M. & Kaplan, S. (1994). prrA, a putative response
regulator involved in oxygen regulation of photosynthesis gene
expression in Rhodobacter sphaeroides. J Bacteriol 176, 32–43.
Folkesson, A., Advani, A., Sukupolvi, S., Pfeifer, J. D., Normark, S.
& Lofdahl, S. (1999). Multiple insertions of fimbrial operons
correlate with the evolution of Salmonella serovars responsible
for human disease. Mol Microbiol 33, 612–622.
Gardiner, A. T., MacKenzie, R. C., Barrett, S. J., Kaiser, K. &
Cogdell, R. G. (1996). The purple photosynthetic bacterium
Rhodopseudomonas acidophila contains multiple puc peripheral
antenna complex (LH2) genes : cloning and initial characterization of four α\β pairs. Photosynth Res 49, 223–235.
Gill, S. C. & von Hippel, P. H. (1989). Calculation of protein
extinction coefficients from amino acid sequence data. Anal
Biochem 182, 319–326.
Griffith, O. H. & Ryan, M. (1999). Bacterial phosphatidyl inositolspecific phospholipase C : structure, function and interaction with
lipids. Biochim Biophys Acta 1441, 237–254.
strain with a novel phosphotriesterase. FEMS Microbiol Lett 206,
51–55.
Hurley, J. H., Dean, A. M., Sohl, J. L., Koshland, D. E., Jr & Stroud,
R. M. (1990). Regulation of an enzyme by phosphorylation at the
active site. Science 249, 1012–1016.
Keen, N. T., Tamaki, S., Kobayashi, D. & Trollinger, D. (1988).
Improved broad-host range plasmids for DNA cloning in Gram
negative bacteria. Gene 70, 191–197.
Kyte, J. & Doolittle, R. F. (1982). A simple method for displaying
the hydropathic character of a protein. J Mol Biol 157, 105–132.
Laemmli, U. K. (1970). Cleavage of structural protein during the
assembly of the head of bacteriophage T4. Nature 227, 680–685.
Lewis, V. E., Donarski, W. J., Wild, J. R. & Raushel, F. M. (1988).
Mechanism and stereochemical course at phosphorus of the
reaction catalysed by a bacterial phosphotriesterase. Biochemistry
27, 1591–1597.
MacBeath, G., Kast, P. & Hilvert, D. (1998). Redesigning enzyme
topology by directed evolution. Science 279, 1958–1961.
Machowski, E. E., McAdam, R. A., Derbyshire, K. M. & Mizrahi, V.
(2000). Construction and application of mycobacterial reporter
transposons. Gene 253, 67–75.
Mulbry, W. W. (1992). The aryldialkyl phosphatase-encoding
gene adpB from Nocardia sp. strain B-1 : cloning, sequencing and
expression in Escherichia coli. Gene 121, 149–153.
Mulbry, W. (2000). Characterization of a novel organophosphorus
hydrolase from Nocardiodes simplex NRRL B-24074. Microbiol
Res 154, 285–288.
Mulbry, W. W. & Karns, J. S. (1989). Parathion hydrolase specified
by the Flavobacterium opd gene : relationship between the gene
and the protein. J Bacteriol 171, 6740–6746.
Penfold, R. J. & Pemberton, J. M. (1992). An improved suicide
vector for construction of chromosomal insertion mutations in
bacteria. Gene 118, 145–146.
Rainey, F. A., Dorsch, M., Morgan, H. W. & Stackebrandt, E.
(1992). 16S rDNA analysis of Spirochaeta thermophila in
phylogenetic position and implications for the systematics of the
order Spirochaetales. Syst Appl Microbiol 15, 197–202.
Rosenberg, A. & Alexander, M. (1979). Microbial cleavage of
various organophosphorus insecticides. Appl Environ Microbiol
37, 886–891.
Roth, M. (1969). Fluorimetric assay of enzymes. Methods Biochem
Anal 17, 189–285.
Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular
Cloning : a Laboratory Manual, 2nd edn. Cold Spring Harbor,
NY : Cold Spring Harbor Laboratory.
Sato, S., Nakamura, Y., Kaneko, T., Katoh, T., Asamizu, E. &
Kotani, H. (2000). Structural analysis of Arabidopsis thaliana
chromosome 5. DNA Res 7, 31–63.
Sethunathan, N. & Yoshida, T. (1973). A Flavobacterium that
degrades diazinon and parathion. Can J Microbiol 19, 873–875.
Shelton, D. R. & Somich, C. J. (1988). Isolation and characteriza-
2694
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:47:20
Phosphotriesterase from Pseudomonas monteilii C11
tion of coumaphos-metabolising bacteria from cattle-dip. Appl
Environ Microbiol 54, 2566–2571.
Shim, H., Hong, S.-B. & Raushel, F. M. (1998). Hydrolysis of
phosphodiesters through transformation of the bacterial phosphotriesterase. J Biol Chem 273, 17445–17450.
Simon, R., Priefer, U. & Pu$ hler, A. (1983). A broad host range
mobilization system for in vivo genetic engineering : transposon
mutagenesis in Gram-negative bacteria. Bio\Technology 1, 784–
791.
Sutherland, T. D., Horne, I., Lacey, M. J., Harcourt, R. L., Russell,
R. J. & Oakeshott, J. G. (2000). Enrichment of an endosulfan-
Prentice Hall.
degrading mixed bacterial culture. Appl Environ Microbiol 66,
2822–2828.
von Tigerstrom, R. G. & Stelmaschuk, S. (1986). Purification and
characterization of the outer membrane associated alkaline
phosphatase of Lysobacter enzymogenes. J Gen Microbiol 132,
1379–1387.
Wanner, B. L. (1983). Overlapping and separate controls of the
phosphate regulon in Escherichia coli K-12. J Mol Biol 166,
283–308.
Williams, P. A. & Murray, K. A. (1974). Metabolism of benzoate
and the methylbenzoates by Pseudomonas putida (arvilla) mt-2 :
evidence for the existence of a TOL plasmid. J Bacteriol 120,
416–423.
Stover, C. K., Pham, X.-Q. T., Erwin, A. L. & 28 other authors
(2000). Complete genome sequence of Pseudomonas aeruginosa
.................................................................................................................................................
PAO1, an opportunistic pathogen. Nature 406, 959–964.
Received 28 January 2002 ; revised 3 May 2002 ; accepted 21 May 2002.
Sinclair, M. I., Maxwell, P. C., Lyon, B. R. & Holloway, B. W.
(1986). Chromosomal location of TOL plasmid in Pseudomonas
putida. J Bacteriol 168, 1302–1308.
Stanier, R. Y., Ingraham, J. L., Wheelis, M. L. & Painter, P. R.
(1986). General Microbiology, 5th edn. Englewood Cliffs, NJ :
2695
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 12:47:20