Rational design of a transition state analogue with picomolar affinity for pseudomonas aeruginosa PvdQ, a siderophore biosynthetic enzyme

Article abstract of DOI:10.1021/cb400345h

The Pseudomonas aeruginosa enzyme PvdQ can process different substrates involved in quorum-sensing or in siderophore biosynthesis. Substrate selectivity was evaluated using steady-state kinetic constants for hydrolysis of N-acyl-homoserine lactones (HSLs) and p-nitrophenyl fatty acid esters. PvdQ prefers substrates with alkyl chains between 12 and 14 carbons long that do not bear a 3-oxo substitution and is revealed here to have a relatively high specificity constant for selected N-acyl-HSLs (k_cat/K_M = 10⁵ to 10⁶ M^-1 s^-1). However, endogenous P. aeruginosa N-acyl-HSLs are ≥100-fold disfavored, supporting the conclusion that PvdQ was not primarily evolved to regulate endogenous quorum-sensing. PvdQ plays an essential biosynthetic role for the siderophore pyoverdine, on which P. aeruginosa depends for growth in iron-limited environments. A series of alkylboronate inhibitors was found to be reversible, competitive, and extremely potent (K_i ≥ 190 pM). A 1.8 A X-ray structure shows that 1-tridecylboronic acid forms a monocovalent bond with the N-terminal β-chain Ser residue in the PvdQ heterodimer, mimicking a reaction transition state. This boronic acid inhibits growth of P. aeruginosa in iron-limited media, reproducing the phenotype of a genetic pvdQ disruption, although co-administration of an efflux pump inhibitor is required to maintain growth inhibition. These findings support the strategy of designing boron-based inhibitors of siderophore biosynthetic enzymes to control P. aeruginosa infections.

Full text of DOI:10.1021/cb400345h

Articles
pubs.acs.org/acschemicalbiology
Rational Design of a Transition State Analogue with Picomolar
Affinity for Pseudomonas aeruginosa PvdQ, a Siderophore
Biosynthetic Enzyme
Kenneth D. Clevenger, Rui Wu, Joyce A. V. Er, Dali Liu,* and Walter Fast*
†
¶
‡
,¶
,‡
†
‡
Department of Chemistry and Biochemistry and the College of Pharmacy, Medicinal Chemistry Division, University of Texas,
Austin, Texas 78712, United States
¶
Department of Chemistry and Biochemistry, Loyola University Chicago, Chicago Illinois 60660, United States
S Supporting Information
*
ABSTRACT: The Pseudomonas aeruginosa enzyme PvdQ can process
different substrates involved in quorum-sensing or in siderophore biosyn-
thesis. Substrate selectivity was evaluated using steady-state kinetic constants
for hydrolysis of N-acyl-homoserine lactones (HSLs) and p-nitrophenyl fatty
acid esters. PvdQ prefers substrates with alkyl chains between 12 and 14
carbons long that do not bear a 3-oxo substitution and is revealed here to
have a relatively high specificity constant for selected N-acyl-HSLs (k /K =
cat
M
5
6
−1 −1
1
0 to 10 M s ). However, endogenous P. aeruginosa N-acyl-HSLs are
≥
100-fold disfavored, supporting the conclusion that PvdQ was not primarily
evolved to regulate endogenous quorum-sensing. PvdQ plays an essential
biosynthetic role for the siderophore pyoverdine, on which P. aeruginosa depends for growth in iron-limited environments. A
series of alkylboronate inhibitors was found to be reversible, competitive, and extremely potent (K ≥ 190 pM). A 1.8 Å X-ray
i
structure shows that 1-tridecylboronic acid forms a monocovalent bond with the N-terminal β-chain Ser residue in the PvdQ
heterodimer, mimicking a reaction transition state. This boronic acid inhibits growth of P. aeruginosa in iron-limited media,
reproducing the phenotype of a genetic pvdQ disruption, although co-administration of an efflux pump inhibitor is required to
maintain growth inhibition. These findings support the strategy of designing boron-based inhibitors of siderophore biosynthetic
enzymes to control P. aeruginosa infections.
he human opportunistic pathogen Pseudomonas aeruginosa
medium, and has reduced virulence in plant and animal models
12
T
is a worldwide clinical threat associated with diverse
of infection. A small molecule inhibitor of PvdQ can also
1
11
nosocomial infections that are increasingly difficult to treat. To
develop novel therapeutics, new antimicrobial targets have been
proposed, including pathways that facilitate iron acquisition and
inhibit growth in a iron-limited medium.
A second notable feature of PvdQ is its ability to function as
a quorum-quenching enzyme. Artificial constitutive expression
of PvdQ (which is typically only upregulated during iron
starvation) or addition of exogenous purified PvdQ prevents
accumulation of the quorum-sensing signal N-3-oxo-doecanoyl-
2,3
those that regulate quorum-sensing. The protein PvdQ is an
unusual example found at the nexus of these two pathways
because its enzymatic function has been proposed to play roles
both in siderophore production and in degradation of some N-
HSL (3-oxo-C12-HSL) and reduces production of the
4
,5
acyl-homoserine lactone (HSL) signaling molecules.
4,13
virulence factors pyocyanin and elastase.
It has been
3
+
P. aeruginosa requires iron (Fe ), which is only sparingly
soluble and is found at a concentration of 10 M in serum in
its free form. In response, biosynthetic pathways have evolved
suggested that the primary biological function of PvdQ is
−24
siderophore biosynthesis and not regulation of quorum-
6
5,12
sensing.
Nevertheless, its ability to degrade N-acyl-HSLs
to produce two siderophores, pyoverdine and pyochelin, that
are released into the surrounding environment to scavenge iron
and to shuttle it back, with pyoverdine being the predominant
can still be a useful activity. For example, administration of
PvdQ as a therapeutic enzyme has been proposed to limit
quorum-sensing-dependent virulence in pulmonary P. aerugi-
7
iron supplier. Siderophore biosynthetic enzymes have been
14
nosa infections.
suggested as novel antibiotic targets since inhibitors would
block iron acquisition and severely limit bacterial growth in
host tissues. This strategy has been shown to be effective with
Mycobacterium tuberculosis, Yersinia pestis, and recently, P.
Regardless of whether one’s goal is to develop an inhibitor to
block PvdQ action in iron acquisition or to develop a quorum-
quenching enzyme, understanding how PvdQ recognizes and
8
−11
aeruginosa.
As illustration of this concept, a genetic
knockout of P. aeruginosa pvdQ, which encodes a key enzyme
in pyoverdine biosynthesis, results in a bacterial strain that does
not produce pyoverdine, is growth inhibited in an iron-limited
Received: May 15, 2013
Accepted: July 24, 2013
Published: July 24, 2013
©
2013 American Chemical Society
2192
dx.doi.org/10.1021/cb400345h | ACS Chem. Biol. 2013, 8, 2192−2200

ACS Chemical Biology
Articles
Figure 1. Selected PvdQ Ligands. The C14 alkyl substitution (black) of a PvdI precursor (purple) is removed by PvdQ-catalyzed amide hydrolysis.
The N-acyl substitutions of varying length (black) on N-acyl-L-homoserine lactones (HSL, green) are hydrolyzed by PvdQ, as are the corresponding
p-nitrophenol (PNP, orange) esters. N-Acyl-HSLs with 3-oxo substitutions bear a carbonyl at C3 (not pictured). Alkylboronic acids (C -B(OH) ,
n
2
red) are inhibitors of PvdQ.
processes ligands provides information valuable for improving
the desired outcome. Herein, we quantify substrate selectivity
and use this information to rationally design a potent transition
state analogue inhibitor of PvdQ that is applied in structural
and cell-based studies. The implications of these findings for
understanding the catalytic mechanism, for designing more
effective inhibitors as molecular probes or therapeutics, and for
developing more effective quorum-quenching enzymes are
discussed.
evolutionarily optimized to regulate endogenous P. aeruginosa
signaling, this activity of PvdQ is still useful as a biochemical
tool for degrading certain N-acyl-HSLs. For example, the
specificity constants of PvdQ for selected substrates compares
favorably with quorum-quenching enzymes from different
superfamilies such as AiiA from Bacillus thuringiensis (dizinc
4
−1 −1
metalloform k /K values ≈ 10 M s ; dicobalt metallo-
cat
M
6
−1 −1 15
form k /K ≈ 10 M s ), and an engineered version of
cat
M
4
−1 −1
PLL from Mycobacterium avium (k /K ≈ 10 M s ),
cat
M
indicating that PvdQ might serve equally well to degrade
16
RESULTS AND DISCUSSION
certain N-acyl-HSLs.
Unlike these other quorum-quenching enzymes,
■
1
5,17
Substrate Selectivity. To quantify the selectivity of PvdQ
for different N-acyl-substrates, we determined steady-state
kinetic parameters for a series of substrates (Figure 1, Table
PvdQ
appears to tolerate different ring substitutions yet is more
selective for chain length. So, we defined the selectivity of PvdQ
more precisely by comparing steady-state kinetic constants for
p-nitrophenyl fatty acid ester reporter substrates (Figure 1,
5
−1 −1
1
). C12-HSL is a good substrate (k /K = 10 M s ), but
cat M
Table 1). Notably, the k /K values measured here are
Table 1. Steady-State Rate Constants for PvdQ-Catalyzed
Hydrolysis of Selected Substrates
cat
M
18
signifignatly larger than earlier reports due mostly to lower
KM values, possibly reflecting increased substrate solubility in
our assay conditions (see Methods). PvdQ preferentially
hydrolyses substrates with 12−14 carbon alkyl chains, with
shorter or longer chains showing ≥2 orders of magnitude
decrease in k /K (Table 1). Taken together, these findings
substrate
C8-HSL
KM (μM)
kcat (min−1)
k /K (M−1 s−1
)
cat
M
a
2.2 × 102
b
ND
ND
ND
b
C10-HSL
C12-HSL
ND
11 ± 2
ND
2.2 × 103
2.2 × 105
148 ± 7
ND
cat
M
3
b
3-oxo-C12-HSL
2.3 × 10
suggest that PvdQ has evolved in response to a need for
processing myristoylated substrates and has not been optimized
C10-PNP
C12-PNP
C14-PNP
C16-PNP
61 ± 13
0.8 ± 0.1
0.60 ± 0.06
4 ± 1
49 ± 5
52 ± 3
86 ± 3
1.9 ± 0.1
1.3 × 104
1.1 × 106
for the endogenous quorum-sensing signals of P. aeruginosa
6
19
2.4 × 10
(
although moonlighting roles are possible). This conclusion is
8 × 103
consistent with the proposed primary role of PvdQ in
processing myristoylated precursors of the siderophore
a
ND: not determined because limited substrate solubility prevents
b
5,12
determination of Vmax. Fitting error is ≤8%.
pyoverdine (Figure 1).
Inhibitor Design. Two previous reports of high-throughput
screening for PvdQ inhibitors describe the discovery of
structurally diverse inhibitors. Two compounds of moderate
potency, NS2028 (8-bromo-4H-[1,2,4]oxadiazolo[3,4-c][1,4]-
benzoxazin-1-one; IC = 130 μM) and SMER28 (6-bromo-N-
(prop-2-en-1-yl)quinazolin-4-amine; IC50 = 65 μM) were
identified from a 1280-compound library of bioactive
compounds (LOPAC) and determined by X-ray crystallog-
raphy to bind in the myristoyl binding pocket, predominantly
shorter acyl chains are only very poorly hydrolyzed (k /K ≤
cat
M
3
−1 −1
1
0 M s ) (Figure 1, Table 1). Notably, 3-oxo-C12-HSL is a
poor substrate. Therefore, PvdQ does not appear to be
13
optimized for hydrolysis of either C4-HSL or 3-oxo-C12-
HSL, the two endogenously produced N-acyl-HSLs of P.
aeruginosa. These results are consistent with qualitative studies
showing a preference for substrates with acyl chains of at least 8
50
4
carbons. Although these results suggest that PvdQ was not
2
193
dx.doi.org/10.1021/cb400345h | ACS Chem. Biol. 2013, 8, 2192−2200

ACS Chemical Biology
Articles
1
8
through hydrophobic interactions. A more potent inhibitor,
ML218 (2-(4-fluorophenyl)-2-(6-(trifluoromethyl)pyridine-2-
yl)acetonitrile; IC50 = 6 nM) was recently discovered by
screening the NIH molecular libraries probe production centers
more potent than the two previously reported compounds
known to bind at the alkyl-binding site.
18
Due to the potency of the alkylboronic acid inhibitors and
their ability to form reversible tetrahedral adducts, we suspected
that they act as transition state mimics. In general, the free
energy of binding transition state analogues is expected to
parallel the free energy of transition state stabilization for a
11
network compound library. The mode of inhibition for
ML218 was not reported, but micromolar concentrations were
shown to inhibit growth of P. aeruginosa in an iron-limited
medium.
Taking an alternative rational design approach, we first
started with a boronate moiety since these are known inhibitors
of other Ser/Thr-dependent hydrolases
a series of alkyl substitutions that mimic the observed substrate
preference. Potency varies with chain length, with C12- and
2
2
structurally similar series of substrates. However, the
structural offset of the inhibitors and substrates described
here (Figure 1) precludes traditional linear comparison plots
between the compounds tested. So instead, we sought to
determine whether the ΔΔG (calculated from K values) for
2
0,21
and then proposed
bind
i
⧧
changes in inhibitor chain length correlates with ΔΔG
(calculated from k /K values) for a similar series of
C13-B(OH) having the lowest K values (∼200 pM) (Table
2
i
cat
M
2
). Please note that the C13-B(OH) inhibitor is the closest
2
substrates. When these ΔΔG values are plotted against
normalized chain length (Figure 2), they clearly overlap, a
result that is consistent with transition state mimicry.
Table 2. K Values for Alkylboronic and Fatty Acid Inhibitors
i
of PvdQ
(
Comparison of ΔG
for inhibitor with the free energy of
bind
ground state binding for substrate is not presented since the
inhibitor
Ki (nM)
rate-limiting step for PNP ester substrates is not known and KM
C8-B(OH)2
C10-B(OH)2
C12-B(OH)2
C13-B(OH)2
C14-B(OH)2
161 ± 3
1.1 ± 0.4
0.19 ± 0.02
0.20 ± 0.04
1.0 ± 0.3
may not be a useful substitute for K values.)
d
Structural Characterization. To test the proposed mode
of inhibition, we determined a structure of the inhibited
complex. In general, boron-containing inhibitors can bind
noncovalently or bind through reversible mono-, di-, or
C15-B(OH)2
8.3 ± 0.9
21,23−26
tricovalent bonds with N or O nucleophiles.
Changes
C12-COOH
6000 ± 200
in inhibitor structure can even change which active-site residue
27
is covalently modified. These different possible binding
mechanisms make a priori predictions problematic. So, we
structural match to myristoylated (C14) substrates because the
boron takes the position of the initial carbon in the
corresponding substrate (Figure 1). The mode of inhibition
for C12- and C13-B(OH) was determined to be competitive
Figure 2, Supplementary Figure 1). We did not observe any
time-dependent onset of inhibition and some activity was
restored upon dialysis (not shown) consistent with a rapid
equilibrium reversible mode of inhibition. The boronic acid
moiety is clearly important for potency since fatty acids are
several orders of magnitude less potent (Table 2, Supple-
mentary Figure S2). These inhibitors are at least 325,000-fold
grew cocrystals of C13-B(OH) inhibited PvdQ at pH 7.5 and
2
determined the X-ray crystal structure at 1.8 Å resolution
(Table 3, Figure 3, and Supplementary Figures S3 and S4). The
2
(
structure of PvdQ in complex with C13-B(OH)
2
was solved in
space group C222 . The final model was built as a heterodimer
1
and comprises an α subunit (residues 28−192) and a β subunit
(residues 217−907) as a result of autoproteolysis. Ligand atoms
were refined with occupancies of 1 and B-factors ranging from
18.45 to 44.97. The overall structure of the protein resembles a
‘V’-shape (Supplementary Figure S3). The active site, including
Figure 2. Competitive tight-binding inhibition of PvdQ by C12-B(OH) . (A) IC50 values were determined for the inhibitor using varying
2
concentrations of C12-PNP substrate. Main panel shows a representative plot with 5 μM substrate. Inset shows a linear dependence of IC50 on [S]/
K , indicating competitive inhibition. (B) Free energy correlation between K and k /K for a structurally related set of inhibitors and substrates.
M
i
cat
M
Inhibitors are designated using their chain length +1 to account for boron at the initial position (Figure 1). The ΔΔG for inhibitor K values
bind
i
⧧
(
purple ●) relative to that of the normalized C14 length are plotted along with ΔΔG values for k /K of PNP substrates (orange ■) relative to
cat
M
C14-PNP.
2
194
dx.doi.org/10.1021/cb400345h | ACS Chem. Biol. 2013, 8, 2192−2200

ACS Chemical Biology
Articles
Table 3. Crystallographic Statistics for C13-B(OH)2-
Inhibited PvdQ
RMSD of 0.3 Å for all Cα atoms. A simulated annealing
composite omit map to remove phase bias reveals extra electron
density at the active site (Figure 3) that is well fit by a covalent
Data Processing
Ser217−C13-B(OH) adduct. The inhibitor alkyl chain binds
2
space group
C2221
in the myristoyl-binding pocket and a monocovalent bond is
formed between the inhibitor boron and the side chain alcohol
of Ser217 (the N-terminal residue of the heterodimer β-chain)
cell dimensions
α, β, γ (deg)
a, b, c (Å)
resolution (Å)
90.0; 90.0; 90.0
121.0; 166.4; 94.2
41.0−1.8
(
Figure 3 and Supplementary Figure S4). The tight packing of
the alkyl chain into the myristoyl-binding site helps explain the
observed substrate preference (Table 1) because substrates
longer than C14 would be disfavored by steric occlusion. It is
less clear how PvdQ disfavors substrates shorter than C12. The
side chains of several residues, including Phe240, Thr285,
Val286, Val403, and Pro238, are found in close proximity (≤5
Å) to C2 of the inhibitor, which corresponds to the position
bearing the disfavored 3-oxo-substitutent in N-acyl-HSL
substrates. Therefore, these residues likely contribute to the
substrate preference of PvdQ for N-acyl-HSLs bearing acyl
chains unsubstituted at the 3-carbon position.
mosaicity (deg)
0.33
a
b
Rmerge (%)
0.124 (0.603)
I/σ
9.6 (2.1)
99.7 (100)
6.3 (6.3)
488124
completeness (%)
multiplicity
no. Reflections
no. unique reflections
86070
Refinement
c
d
Rwork /Rfree (%)
no. of atoms
protein
16.23/18.78
6304
22
Formation of a dative covalent bond between Ser217 and the
inhibitor (Figure 3 and Supplementary Figure S4) results in an
adduct that mimics an anionic tetrahedral reaction transition
ligand
water
598
2
B-factors (Å )
state, consistent with the observed correlation between
overall average
29.5
⧧
protein average
23.68
ΔΔG
bind
and ΔΔG values (Figure 2) and the contribution
ligand B-factor range
e
18.45/44.97
of the boronic acid moiety to potency (Table 2, Supplementary
Figure S2). Previously, PvdQ crystals at low pH have been used
RMSD
1
8,28
bond length (Å)
bond angle (deg)
0.007
1.126
to trap a covalent ester reaction intermediate.
Here, the use
of a transition state mimic allows a representative structure of a
reaction step to be obtained at pH 7.5, avoiding any structural
perturbations due to altered pH conditions that inhibit catalysis
and provides additional insight into catalysis.
Ramachandran statistics (%)
most favored
allowed
96.07
3.37
0.56
outliers
Mechanistic Implications. The structure of a transition
state analogue helps elucidate the catalytic mechanism.
Previously, a mechanism has been proposed based on
structures of the resting enzyme, a product complex, and a
covalent ester intermediate trapped at low pH (Figure 4, paths
a
b
Rmerge = Σ|I − I |/ΣI
The values for the highest resolution
obs
avg
avg
c
d
bind are in parentheses. R
= Σ|F − F |/ΣF . Five percent of
work
obs
calc
obs
the reflection data were selected at random as a test set, and only these
data were used to calculate Rfree. RMSD, root-mean-square deviation.
e
28
1
b, 3b, 5b). Here, we propose amendments resulting in a
the catalytic N-terminal residue of the β subunit, Ser217,
resides in a deep cleft that is large enough to accommodate the
proposed siderophore precursor as a substrate.
simplified mechanism that does not include product-assisted
catalysis (Figure 4, paths 1a, 3a, 5a): Based on precedence in
other superfamily members, a direct deprotonation of the
Ser217 side chain by the N-terminal amine (path 1a), proceeds
to attack of the bound substrate and formation of the first
The overall protein structure of PvdQ matches those
previously reported except for differences near the active site
1
8,28
29
(
Supplementary Figure S3).
Structural comparison with
tetrahedral adduct. The N-terminus may have a depressed
28
one previously solved structure (PDB ID: 2WYE) shows a
pK due to nearby electropositive side chains of His239 (2.9 Å)
a
Figure 3. A 1.8 Å structure of PvdQ inhibited by C13-B(OH) . (A) Cutaway surface view showing the alkyl chain of the inhibitor (orange) bound to
2
the myristyl-binding site of PvdQ, with the α subunit in green and the β subunit in purple. See Supplementary Figure S3 for a wider perspective. (B)
A simulated annealing composite omit map at 1.1 σ showing a monocovalent adduct between the inhibitor’s boron (pink) and the side chain of
Ser217 (the N-terminal residue of the β-chain), with H-bonds and water molecules included.
2
195
dx.doi.org/10.1021/cb400345h | ACS Chem. Biol. 2013, 8, 2192−2200

ACS Chemical Biology
Articles
Figure 4. Proposed chemical mechanism for PvdQ. The C13-B(OH) −PvdQ complex most closely mimics the second tetrahedral intermediate
2
(
boxed). R′ is the PvdI precursor or HSL (Figure 1), which is released as an amine as the first product (P ). R is an acyl chain that is released as a
1
fatty acid as the second product (P ). Results presented here favor the more parsimonious mechanism (paths a; see Results and Discussion).
2
Figure 5. P. aeruginosa growth studies with C13-B(OH) . (A) In iron-limited minimum medium, the growth of wild-type P. aeruginosa in the absence
2
of inhibitor (black ●) or after initial treatment with 1 μM inhibitor (orange ◆) is shown in comparison to the untreated tranposon pvdQ insertion
mutant P. aeruginosa strain (purple ■). O.D. is optical density. Inset shows 20−40 h in more detail. (B) Same experiment as in panel A, except all
incubations also contain 40 μM PAβN. (C) In minimal media containing trace iron, the growth of the transposon pvdQ insertion mutant P.
aeruginosa strain is shown in the presence of 0 (purple ●), 0.1 (light blue ■), 10 (dark blue ◆), and 100 (gray ▲) μM inhibitor. (D) Same
experiment as in panel C, with an additional concentration of inhibitor (1 μM, ), except each incubation also contains 40 μM PAβN, and one
incubation (purple ○, dashed line) contains neither inhibitor nor PAβN.
and Arg513 (3.5 Å). The N-terminal amine can again act
directly as a general acid to protonate the leaving group and as a
general base to generate a hydroxide ion required to attack the
ester intermediate (path 3a). In contrast to amides, PNP esters
2
196
dx.doi.org/10.1021/cb400345h | ACS Chem. Biol. 2013, 8, 2192−2200

ACS Chemical Biology
Articles
have leaving groups that do not require protonation and would
therefore not supply the general base needed to generate
hydroxide in path 3b. Yet, the PNP esters described here are
excellent substrates, and so favor a product-independent
pathway for intermediate decay (path 3a). Attack of hydroxide
ion leads to formation of the second tetrahedral adduct (Figure
compound in cell culture is likely limited by poor access to the
target due to outer membrane impermeability or a multidrug
efflux pump. It has been reported that compounds structurally
similar to C13-B(OH) , including N-acyl-HSLs and sodium
2
dodecyl sulfate, are exported by the MexAB-OprM efflux pump,
3
1,32
but other efflux pumps might also play a role.
4
, boxed). The C13-B(OH) −PvdQ complex most closely
As predicted, when iron is not limiting, control experiments
using the same concentrations of inhibitors to treat the
transposon-disrupted pvdQ strain do not result in a similar
growth inhibition (Figure 5C and D). This result is consistent
2
resembles this catalytic intermediate and provides some insight
into its formation and stabilization. Based on comparison of the
transition-state analogue to a trapped ester intermediate,
the incoming hydroxide ion in step 4 appears to attack the re
1
8,28
with our proposal that C13-B(OH) targets iron acquisition.
2
28
face of the ester. As proposed earlier, an “oxyanion” hole,
consisting of the backbone nitrogen of Val268 and the side
chain nitrogen of Asn485, provides H-bonds to one of the
oxygens in the tetrahedral adduct. Additionally, the oxygen
corresponding to the incoming nucleophilic hydroxide is
stabilized by H-bonds to both the backbone nitrogen and
carbonyl of His239 and is also close (2.7 Å) to the N-terminus
of Ser217. Collapse of the tetrahedral adduct leads to C−O
bond cleavage and protonation of the Ser217 side chain. The
Higher concentrations of C13-B(OH) do result in minor
2
growth inhibition (Figure 5C), which suggests that this
particular inhibitor may have deleterious off-target effects
when used at elevated concentrations. This result is not entirely
unexpected due to the hydrophobicity and structural simplicity
of the unsubstituted n-alkyl substituent. However, to the best of
our knowledge, this is the first rationally designed boron-
containing inhibitor of siderophore biosynthesis shown to block
the growth of P. aeruginosa in iron-limited conditions and
serves as a proof of principle.
C13-B(OH) −PvdQ complex shows a 2.8 Å distance from this
2
oxygen to the N-terminal amine, supporting a direct proton
transfer (path 5a) rather than transfer through an intervening
water (path 5b). The closest ordered water molecule observed
is 5.8 Å distant and overlaps with the position likely occupied
by the leaving group amine in the first tetrahedral intermediate,
In summary, PvdQ is revealed to have high specificity
constants for myristoylated substrates, yet it discriminates
against N-acyl-HSLs produced endogenously by P. aeruginosa.
This finding supports the conclusion that PvdQ has not
primarily evolved to regulate quorum-sensing endogenous to P.
aeruginosa. However, the quorum-quenching ability of PvdQ
remains useful as a biochemical tool to target a narrow set of
quorum-sensing signals. To counteract the primary role of
PvdQ in siderophore production, we identified n-alkylboronic
acids as extremely potent PvdQ inhibitors, due in part to the
similarity of their monocovalent tetrahedral adducts to the
reaction transition state. One of the most potent compounds,
disfavoring paths 3b and 5b. The fatty acid product (P ) can
2
then dissociate to regenerate resting enzyme (path 6).
Cultured Cell Growth Assays. Previously, different
enzymes in siderophore biosynthesis have been effectively
targeted by inhibitor design and high-throughput screening.
Inhibitors of aryl acid adenylating enzymes potently inhibit
siderophore production and/or iron-limited growth of M.
8
−10
1
-tridecylboronic acid (C13-B(OH) ), inhibits growth of P.
tuberculosis, Y. pestis, and other microorganisms.
These
2
aeruginosa in iron-limited media and serves as a proof of
principle. The efficacy of this particular compound is limited by
poor target access and likely by limited selectivity due to its
unsubstituted alkyl moiety. However, since boron-based
compounds might also have potential applications with P.
aeruginosa due to similarities of their target enzymes with the
adenylating enzyme PchD used in pyochelin production. To
the best of our knowledge, only one report describes a
siderophore biosynthetic inhibitor that was applied to P.
aeruginosa growth. Micromolar concentrations of the PvdQ
inhibitor ML318 inhibits growth of P. aeruginosa in iron-
20
inhibitors are used clinically and have proven useful in
33
cultured P. aeruginosa, we suggest that this moiety may be a
productive chemotype to retain in the design of more
biologically effective PvdQ inhibitors.
11
limiting conditions. Here, we extend the study of iron-limited
P. aeruginosa growth to alkylboronate inhibitors of PvdQ.
METHODS
The PvdQ inhibitor C13-B(OH) was tested for its effects in
2
■
cell culture (Figure 5). Wild-type P. aeruginosa can grow in
iron-limited minimal media, but the transposon-disrupted pvdQ
mutant strain shows only very limited growth. Wild-type P.
aeruginosa cultures given an initial dose of 1 μM inhibitor (5000
Materials. Unless otherwise noted, all chemicals were obtained
from Sigma-Aldrich Chemical Co. (St. Louis, MO, USA), and all
restriction enzymes were from New England BioLabs (Beverly, MA,
USA). 1-Octylboronic acid (C8-B(OH) ), 1-decylboronic acid (C10-
2
B(OH) ), 1-dodecylboronic acid (C12-B(OH) ), and 1-tetradecylbor-
2
2
×
Ki) are growth inhibited. However, after 30 h, growth of the
onic acid (C14-B(OH) ) were from Alfa Aesar (Ward Hill, MA, USA).
p-Nitrophenyl esters of capric (C10-PNP), lauric (C12-PNP), myristic
C14-PNP), and palmitic (C16-PNP) acids were from Research
Organics, Cleveland, OH. Syringe filters are MillexGP filter units (0.22
μm, PES membrane, Millipore). PAβN was from Bachem (Torrance,
California).
3-Oxo-dodecanoyl-homoserine lactone (3-oxo-C12-HSL) is known
to undergo non-enzymatic rearrangement. To verify the stock
solutions of this compound had not rearranged from the lactone
which is susceptible to AiiA mediated hydrolysis) to the tetramic acid
which is not), LC-ESI-MS (+ ionization) was used to show an 18 Da
increase between untreated (298.2 Da) and AiiA -treated (316.2)
samples, consistent with maintaining an intact lactone, 3-oxo-C12-
HSL, in the stock solution.
1-Tridecylboronic acid (C13-B(OH) ) was synthesized in two steps
using a published route for butylboronic acid, but substituting
2
inhibitor-treated culture recovers and then parallels that of the
untreated wild-type strain (Figure 5A). Recovery cannot be
avoided by repeated dosing of the inhibitor (not shown).
These experiments were then repeated in the presence of
phenylalainine-arginine-β-naphthylamide (PAβN; CAS 100929-
(
99-5), a wide-spectrum competitive inhibitor of multidrug-
3
4
resistance efflux pumps that can also increase outer membrane
30
permeability through other mechanisms. Co-administration
(
(
of both C13-B(OH) and PAβN severely inhibits growth of
2
wild-type P. aeruginosa under iron-limited conditions, and
growth does not recover like it did in the absence of PAβN
35
(
Figure 5B). These studies indicate that the PvdQ inhibitor
C13-B(OH) can inhibit the growth of P. aeruginosa in iron-
2
2
3
6
limited conditions, but that the efficacy of this particular
2
197
dx.doi.org/10.1021/cb400345h | ACS Chem. Biol. 2013, 8, 2192−2200

ACS Chemical Biology
Articles
tridecylic acid for butyric acid. The product was recrystallized from
hexanes as white needles (57% yield). H NMR (DMSO): δ 7.28 (s,
(0.05 M Hepes at pH 7.5); 80 mM RbCl; 9% PEG 4000 and 20%
1
glycerol, which is slightly modified from a previously reported
1
8
2
H), 1.26 (m, 22H), 0.84 (t, J = 6.5 Hz, 3H), 0.55 (t, J = 7.5 Hz, 2H).
C NMR (DMSO): δ 32.1, 31.3, 29.1, 29.0, 28.7, 24.2, 22.1, 15.4,
3.9. 11B NMR (DMSO): δ 32.4. HRMS: [M − H] calcd for
condition ) at a ratio of 1:0.5:1 μL, respectively, at 20 °C. Crystals
appeared in 1 h and grew to maximum size in 5 days. Those with good
morphology and typical dimensions of 0.5 mm × 0.3 mm × 0.2 mm
were picked and flash cooled in N2(l). X-ray diffraction data sets were
collected at the Advanced Photon Source, beamline 23-ID, Argonne
National Laboratory. The wavelength used in monochromatic data
collection was 1.0332 Å. All collected data sets were indexed and
1
3
−
1
C H BO 227.2190, found 227.2190.
13
28
2
Cloning, Expression, and Purification. P. aeruignosa PvdQ was
obtained by heterologous expression from E. coli using standard
methodology described in Supporting Information.
3
8
39
Activity and Inhibition Assays. Steady-state kinetic constants
integrated using iMosflm and scaled in program Scala in CCP4.
were determined for hydrolysis of two substrate classes. PNP-ester
The best data set was processed to a resolution of 1.8 Å.
18
hydrolysis was continuously monitored using UV−vis absorbance
spectroscopy (Varian Cary 50) to detect formation of p-nitro-
The PvdQ−C13-B(OH)
2
structure was solved by molecular
replacement using the program PHASER with a starting model of a
2
8
−1
−1
phenolate; ε402nm = 13,000 M cm . The ε
was determined
reported structure of PvdQ (PDB ID: 2WYE ). The structural
solution of the protein was obtained, and an extended region of
402nm
experimentally using the assay buffer. Substrate stock solutions were
prepared in methanol, which was determined to be an acceptable assay
cosolvent (Supplementary Figure S5). Assays were completed using
positive difference electron density (F
− F map) was identified (not
o
c
shown; see Figure 3B for a simulated annealing composite omit map).
This observed extra density corresponds to the omitted inhibitor
moiety at the active site. Once a structural solution was obtained,
disposable polystyrene cuvettes (Thermo Fisher), substrate, Na HPO4
2
(40 mM) at pH 8.0, 25 °C with 20% methanol as a cosolvent and were
4
0
initiated with addition of PvdQ (10 nM). Fresh working dilutions of
PvdQ (1 μM) were made from more concentrated stocks after each
hour of use. Assay buffer was filtered prior to use to eliminate dust
particles. Initial hydrolysis rates (<10% substrate consumed) were
linear. KaleidaGraph 3.6 (Synergy Software, Reading, PA, USA) was
model building was conducted in COOT. A structural file of the
C13-B(OH) −Ser217 adduct was created in Chemdraw (Perkin-
Elmer, Cambridge, MA, USA); the generated structure of the adduct
2
4
1
was used to generate chemical restraints using JLigand. The covalent
adduct was fitted into the observed extra electron density using
COOT. Rigid body refinement and restrained refinement were
conducted using program REFMAC5 in CCP4. Manual adjustments
and rebuilding of the structural model were conducted using COOT;
used to determine kcat and K values and associated error by nonlinear
M
fitting to the Michaelis−Menten equation. Values determined here are
signifigantly different than previous reported values (e.g., C12-PNP,
4
2
−1
18
further refinements were carried out using PHENIX, including TLS
KM = 1.4 mM, kcat 44 min ). In our hands, the substrates were not
sufficiently soluble to use the previously reported assay conditions.
Here, inclusion of well-tolerated cosolvent (Supplemetary Figure S5)
4
3
refinement and simulated annealing refinement. Structural figures of
4
4
the protein were created using UCSF Chimera.
P. aeruginosa Growth Assays. The effect of C13-B(OH) on
led to improved kinetic parameters (e.g., C12-PNP, K = 0.6 μM, kcat
2
M
−1
cultured cells was monitored using wild-type P. aeruginosa PA14 or a
transposon-disrupted pvdQ mutant, P. aeruginosa PA14 ID27758 from
8
6 min ) (Table 1).
N-Acyl-HSL amide hydrolysis was monitored with an end point
4
5
the PA14 mutant library. This mutant strain has been used
assay by o-phthaldialdehyde (OPA) derivitization of the product amine
46
3
7
previously to study phenotypes resulting from loss of PvdQ function.
followed by HPLC and fluorescence detection compared to a
Briefly, a 1 mL culture in LB, started from glycerol stocks, is grown
with shaking (220 rpm) for approximately 7 h until A600nm is
approximately 0.5. Cells were pelleted (5000 × g), washed three times
with sterile phosphate-buffered saline (PBS), and resuspended in 250
μL of sterile iron-limited MOPS glucose medium (25 mM
standard curve prepared each day using (S)-(−)-α-amino-γ-butyr-
olactone (linear from 30 nM to 100 μM). Assay details are supplied in
Supporting Information.
Initial screening for PvdQ inhibition by fatty acids and boronic acids
was accomplished using a method similar to the PNP-ester assay
2
morpholinepropanesulfonic acid at pH 7.2, 93 mM NH Cl, 43 mM
described above except using a Wallac Victor plate reader
4
NaCl, 3.7 mM KH PO , 1 mM MgSO , 45 μM diethylene triamine
(
PerkinElmer, Waltham MA), one concentration of C12-PNP (5
2
4
4
4
7
pentaacetic acid (DTPA), 20 mM glucose). DTPA was added to
chelate any trace iron, as described elsewhere. Triplicate cultures (1
μM) and varying concentrations of fatty acids and boronic acids,
diluted from DMSO stock solutions, in triplicate. Final assays do not
exceed 1% DMSO (v/v), which does not inhibit PvdQ.
4
8
mL) in the same medium were then inoculated to a starting OD600 of
0
.01, treated with either vehicle alone (DMSO, 0.1% v/v) or C13-
Concentration−response curves for C10-, C14-, and C15-B(OH)2
were determined in the same fashion and fit to determine IC50 values.
B(OH) (1 μM), and incubated with shaking at 37 °C for up to 70 h.
2
Tests for off-target toxicity were completed in the same manner using
the pvdQ mutant P. aeruginosa strain and the same culture medium,
The IC50 for C8-B(OH) was also measured this way, but using a Cary
2
5
0 spectrophotometer instead of a plate reader. These IC50 values were
then converted to K values for a tight-binding competitive inhibition
but omitting DTPA and instead supplementing with FeSO (3.5
4
i
4
7
2
2
μM) and treating with various inhibitor concentrations (1 nM to 100
μM). When indicated, stock solutions of PAβN (200 mM in DMSO)
were diluted into cultures to a final concentration of 40 μM.
model described elsewhere using enzyme concentration as
determined by A280nm (Supporting Information). Because these
compounds have IC50 values greater than the enzyme concentration,
small variations in enzyme concentration have less of an impact on Ki
than with the more potent inhibitors below.
ASSOCIATED CONTENT
* Supporting Information
This material is available free of charge via the Internet at
http://pubs.acs.org.
■
The K values and the mode of inhibition for C12- and C13-
S
i
B(OH) were determined using a slightly modified method. Briefly,
2
PvdQ activity was measured in singlicate assays using disposable
polystyrene cuvettes and a Cary-50 UV−vis spectrophotometer as
above to determine IC50 values at different concentrations of C12-
PNP. The IC50 values were plotted against [C12-PNP]/KM values to
Accession Codes
Coordinates of the PvdQ−C13-B(OH) complex have been
2
determine mode of inhibition and K values, as described elsewhere
deposited in the Protein Data Bank with the accession number
4M1J.
i
22
(Supplementary Figure S1). Due to the potency of these tight-
binding inhibitors, small changes in inhibitor or enzyme concen-
trations can impact IC50 values, so in contrast to the inhibitors above,
AUTHOR INFORMATION
Corresponding Author
E- mail: walt.fast@austin.utexas.edu; dliu@luc.edu.
■
*
all kinetic experiments for C12- and C13-B(OH) were performed on
2
the same day using the same enzyme and inhibitor stock solutions.
Crystallography. Cocrystals of PvdQ and C13-B(OH)2 were
grown by hanging drop vapor diffusion by mixing PvdQ (10 mg
Notes
−1
mL ), C13-B(OH) (50 mM in DMSO), and crystallization solution
The authors declare no competing financial interest.
2
2
198
dx.doi.org/10.1021/cb400345h | ACS Chem. Biol. 2013, 8, 2192−2200

ACS Chemical Biology
Articles
(
13) Sio, C. F., Otten, L. G., Cool, R. H., Diggle, S. P., Braun, P. G.,
ACKNOWLEDGMENTS
We thank the following for generous gifts: M. Whitely (Univ. of
Texas, Austin) for P. aeruginosa PA14 ID27758, J. Leadbetter
CA Inst Tech) for a PvdQ expression plasmid, and W.
Bachovchin & D. O’Connell (Tufts University) for 1-
pentadecylboronic acid (C15-B(OH) ). We thank J. Golihar
■
Bos, R., Daykin, M., Camara, M., Williams, P., and Quax, W. J. (2006)
Quorum quenching by an N-acyl-homoserine lactone acylase from
Pseudomonas aeruginosa PAO1. Infect. Immun. 74, 1673−1682.
(
(14) Wahjudi, M., Murugappan, S., van Merkerk, R., Eissens, A. C.,
Visser, M. R., Hinrichs, W. L., and Quax, W. J. (2012) Development of
a dry, stable and inhalable acyl-homoserine-lactone-acylase powder
formulation for the treatment of pulmonary Pseudomonas aeruginosa
infections. Eur. J. Pharm. Sci. 48, 637−643.
2
(
Univ. of Texas, Austin) for help designing the growth assays.
This work was supported in part by the Robert A. Welch
Foundation (Grant F-1572 to W.F.) and by Loyola University
Chicago (to D.L.). We thank Ruslan Sanishvili (GM/CA-CAT,
sector 23) at Advanced Photon Source, Argonne National
Laboratory (Argonne, IL) for help with data collection; GM/
CA at APS has been funded in whole or in part with Federal
funds from the National Cancer Institute (Y1-CO-1020) and
the National Institute of General Medical Sciences (Y1-GM-
(15) Thomas, P. W., Stone, E. M., Costello, A. L., Tierney, D. L., and
Fast, W. (2005) The quorum-quenching lactonase from Bacillus
thuringiensis is a metalloprotein. Biochemistry 44, 7559−7569.
(
16) Chow, J. Y., Wu, L., and Yew, W. S. (2009) Directed evolution
of a quorum-quenching lactonase from Mycobacterium avium subsp.
paratuberculosis K-10 in the amidohydrolase superfamily. Biochemistry
4
(
8, 4344−4353.
17) Momb, J., Wang, C., Liu, D., Thomas, P. W., Petsko, G. A., Guo,
1
104). Use of the Advanced Photon Source was supported by
the U.S. Department of Energy, Basic Energy Sciences, Office
of Science, under contract No. DE-AC02-06CH11357.
H., Ringe, D., and Fast, W. (2008) Mechanism of the quorum-
quenching lactonase (AiiA) from Bacillus thuringiensis. 2. Substrate
modeling and active site mutations. Biochemistry 47, 7715−7725.
(
18) Drake, E. J., and Gulick, A. M. (2011) Structural characterization
REFERENCES
and high-throughput screening of inhibitors of PvdQ, an NTN
hydrolase involved in pyoverdine synthesis. ACS Chem. Biol. 6, 1277−
1286.
■
(
1) Navon-Venezia, S., Ben-Ami, R., and Carmeli, Y. (2005) Update
on Pseudomonas aeruginosa and Acinetobacter baumannii infections in
the healthcare setting. Curr. Opin. Infect. Dis. 18, 306−313.
(19) Copley, S. D. (2012) Moonlighting is mainstream: paradigm
adjustment required. Bioessays 34, 578−588.
(
2) Frederick, R. E., Mayfield, J. A., and DuBois, J. L. (2009) Iron
trafficking as an antimicrobial target. Biometals 22, 583−593.
(20) Baker, S. J., Ding, C. Z., Akama, T., Zhang, Y. K., Hernandez, V.,
and Xia, Y. (2009) Therapeutic potential of boron-containing
compounds. Future Med. Chem. 1, 1275−1288.
(
3) Romero, M., Acuna, L., and Otero, A. (2012) Patents on quorum
quenching: interfering with bacterial communication as a strategy to
fight infections. Recent Pat. Biotechnol. 6, 2−12.
(21) Smoum, R., Rubinstein, A., Dembitsky, V. M., and Srebnik, M.
(2012) Boron containing compounds as protease inhibitors. Chem.
Rev. 112, 4156−4220.
(
4) Huang, J. J., Han, J. I., Zhang, L. H., and Leadbetter, J. R. (2003)
Utilization of acyl-homoserine lactone quorum signals for growth by a
soil pseudomonad and Pseudomonas aeruginosa PAO1. Appl. Environ.
Microbiol. 69, 5941−5949.
(22) Copeland, R. A. (2005) Evaluation of Enzyme Inhibitors in Drug
Discovery: A Guide for Medicinal Chemists and Pharmacologists, pp 178−
(
5) Hannauer, M., Schafer, M., Hoegy, F., Gizzi, P., Wehrung, P.,
2
13, Wiley-Interscience, Hoboken, NJ.
Mislin, G. L., Budzikiewicz, H., and Schalk, I. J. (2012) Biosynthesis of
the pyoverdine siderophore of Pseudomonas aeruginosa involves
precursors with a myristic or a myristoleic acid chain. FEBS Lett.
(
23) Philipp, M., and Bender, M. L. (1971) Inhibition of serine
proteases by arylboronic acids. Proc. Natl. Acad. Sci. U.S.A. 68, 478−
80.
24) Zervosen, A., Herman, R., Kerff, F., Herman, A., Bouillez, A.,
4
(
5
(
(
86, 96−101.
6) Otto, B. R., Verweij-van Vught, A. M., and MacLaren, D. M.
1992) Transferrins and heme-compounds as iron sources for
pathogenic bacteria. Crit. Rev. Microbiol. 18, 217−233.
7) Schalk, I. J., and Guillon, L. (2012) Pyoverdine biosynthesis and
Prati, F., Pratt, R. F., Frere, J. M., Joris, B., Luxen, A., Charlier, P., and
Sauvage, E. (2011) Unexpected tricovalent binding mode of boronic
acids within the active site of a penicillin-binding protein. J. Am. Chem.
Soc. 133, 10839−10848.
(
secretion in Pseudomonas aeruginosa: implications for metal homeo-
stasis. Environ. Microbiol. 15, 1661−1673.
(25) Transue, T. R., Krahn, J. M., Gabel, S. A., DeRose, E. F., and
London, R. E. (2004) X-ray and NMR characterization of covalent
(
8) Ferreras, J. A., Ryu, J. S., Di Lello, F., Tan, D. S., and Quadri, L. E.
complexes of trypsin, borate, and alcohols. Biochemistry 43, 2829−
(
2005) Small-molecule inhibition of siderophore biosynthesis in
Mycobacterium tuberculosis and Yersinia pestis. Nat. Chem. Biol. 1,
9−32.
9) Neres, J., Wilson, D. J., Celia, L., Beck, B. J., and Aldrich, C. C.
2008) Aryl acid adenylating enzymes involved in siderophore
2
839.
(26) Katz, B. A., Finer-Moore, J., Mortezaei, R., Rich, D. H., and
2
Stroud, R. M. (1995) Episelection: novel Ki ∼ nanomolar inhibitors of
serine proteases selected by binding or chemistry on an enzyme
surface. Biochemistry 34, 8264−8280.
(
(
biosynthesis: fluorescence polarization assay, ligand specificity, and
discovery of non-nucleoside inhibitors via high-throughput screening.
Biochemistry 47, 11735−11749.
(27) Bachovchin, W. W., Wong, W. Y., Farr-Jones, S., Shenvi, A. B.,
and Kettner, C. A. (1988) Nitrogen-15 NMR spectroscopy of the
catalytic-triad histidine of a serine protease in peptide boronic acid
inhibitor complexes. Biochemistry 27, 7689−7697.
(
10) Somu, R. V., Boshoff, H., Qiao, C., Bennett, E. M., Barry, C. E.,
3rd, and Aldrich, C. C. (2006) Rationally designed nucleoside
(28) Bokhove, M., Nadal Jimenez, P., Quax, W. J., and Dijkstra, B. W.
antibiotics that inhibit siderophore biosynthesis of Mycobacterium
tuberculosis. J. Med. Chem. 49, 31−34.
(
2010) The quorum-quenching N-acyl homoserine lactone acylase
PvdQ is an Ntn-hydrolase with an unusual substrate-binding pocket.
(
11) Theriault, J. R., Wurst, J., Jewett, I., Verplank, L., Perez, J. R.,
Proc. Natl. Acad. Sci. U.S.A. 107, 686−691.
Gulick, A. M., Drake, E. J., Palmer, M., Moskowitz, S., Dasgupta, N.,
Brannon, M. K., Dandapani, S., Munoz, B., and Schreiber, S. (2013)
Identification of a small molecule inhibitor of Pseudomonas aeruginosa
PvdQ acylase, an enzyme involved in siderophore pyoverdine
synthesis, Probe Reports from NIH Molecular Libraries Program,
National Center for Biotechnology Information, Bethesda, MD.
(29) Chilov, G. G., Sidorova, A. V., and Svedas, V. K. (2007)
Quantum chemical studies of the catalytic mechanism of N-terminal
nucleophile hydrolase. Biochemistry (Mosc) 72, 495−500.
(30) Lomovskaya, O., Warren, M. S., Lee, A., Galazzo, J., Fronko, R.,
Lee, M., Blais, J., Cho, D., Chamberland, S., Renau, T., Leger, R.,
Hecker, S., Watkins, W., Hoshino, K., Ishida, H., and Lee, V. J. (2001)
Identification and characterization of inhibitors of multidrug resistance
efflux pumps in Pseudomonas aeruginosa: novel agents for combination
therapy. Antimicrob. Agents Chemother. 45, 105−116.
(
12) Nadal Jimenez, P., Koch, G., Papaioannou, E., Wahjudi, M.,
Krzeslak, J., Coenye, T., Cool, R. H., and Quax, W. J. (2010) Role of
PvdQ in Pseudomonas aeruginosa virulence under iron-limiting
conditions. Microbiology 156, 49−59.
2
199
dx.doi.org/10.1021/cb400345h | ACS Chem. Biol. 2013, 8, 2192−2200

ACS Chemical Biology
Articles
(
31) Pearson, J. P., Van Delden, C., and Iglewski, B. H. (1999) Active
(48) Fung, C., Naughton, S., Turnbull, L., Tingpej, P., Rose, B.,
Arthur, J., Hu, H., Harmer, C., Harbour, C., Hassett, D. J., Whitchurch,
C. B., and Manos, J. (2010) Gene expression of Pseudomonas
aeruginosa in a mucin-containing synthetic growth medium mimicking
cystic fibrosis lung sputum. J. Med. Microbiol. 59, 1089−1100.
efflux and diffusion are involved in transport of Pseudomonas aeruginosa
cell-to-cell signals. J. Bacteriol. 181, 1203−1210.
(
32) Srikumar, R., Kon, T., Gotoh, N., and Poole, K. (1998)
Expression of Pseudomonas aeruginosa multidrug efflux pumps MexA-
MexB-OprM and MexC-MexD-OprJ in a multidrug-sensitive Escher-
ichia coli strain. Antimicrob. Agents Chemother. 42, 65−71.
(
33) Hernandez, V., Crepin, T., Palencia, A., Cusack, S., Akama, T.,
Baker, S. J., Bu, W., Feng, L., Freund, Y. R., Liu, L., Meewan, M.,
Mohan, M., Mao, W., Rock, F. L., Sexton, H., Sheoran, A., Zhang, Y.,
Zhang, Y. K., Zhou, Y., Nieman, J. A., Anugula, M. R., Keramane el, M.,
Savariraj, K., Reddy, D. S., Sharma, R., Subedi, R., Singh, R., O’Leary,
A., Simon, N. L., De Marsh, P. L., Mushtaq, S., Warner, M., Livermore,
D. M., Alley, M. R., and Plattner, J. J. (2013) Discovery of a novel class
of boron-based antibacterials with activity against gram-negative
bacteria. Antimicrob. Agents Chemother. 57, 1394−1403.
(
34) Kaufmann, G. F., Sartorio, R., Lee, S. H., Rogers, C. J., Meijler,
M. M., Moss, J. A., Clapham, B., Brogan, A. P., Dickerson, T. J., and
Janda, K. D. (2005) Revisiting quorum sensing: Discovery of
additional chemical and biological functions for 3-oxo-N-acylhomo-
serine lactones. Proc. Natl. Acad. Sci. U.S.A. 102, 309−314.
(
35) Thomas, P. W., and Fast, W. (2011) Heterologous over-
expression, purification, and in vitro characterization of AHL
lactonases. Methods Mol. Biol. 692, 275−290.
(
36) Charette, A. B., and Lebel, H. (1999) (2S,3S)-(+)-(3-
Phenylcyclopropyl)methanol. Org. Synth. 76, 86.
37) Xu, F., Byun, T., Deussen, H. J., and Duke, K. R. (2003)
(
Degradation of N-acylhomoserine lactones, the bacterial quorum-
sensing molecules, by acylase. J. Biotechnol. 101, 89−96.
(
38) Battye, T. G., Kontogiannis, L., Johnson, O., Powell, H. R., and
AG, L. (2011) iMOSFLM: a new graphical interface for diffraction-
image processing with MOSFLM. Acta Crystallogr., Sect. D: Biol.
Crystallogr. 67, 27−81.
(
39) Collaborative Computation Project (1994) The CCP4 suite:
programs for protein crystallography. Acta Crystallogr., Sect. D: Biol.
Cyrstallogr. 50, 760−763.
(
40) Emsley, P., and Cowtan, K. (2004) Coot: model-building tools
for molecular graphics. Acta Crystallogr., Sect. D: Biol. Crystallogr. 60,
126−2132.
41) Lebedev, A. A., Young, P., Isupov, M. N., Moroz, O. V., Vagin,
2
(
A. A., and Murshudov, G. N. (2012) JLigand: a graphical tool for the
CCP4 template-restraint library. Acta Crystallogr., Sect. D: Biol.
Crystallogr. 68, 431−440.
(
42) Adams, P. D., Grosse-Kunstleve, R. W., Hung, L.-W., Ioerger, T.
R., McCoy, A. J., Moriarty, N. W., Read, R. J., Sacchettini, J. C., Sauter,
N. K., and Terwilliger, T. C. (2002) PHENIX: building new software
for automated crystallographic structure determination. Acta Crystal-
logr., Sect. D: Biol. Crystallogr. 58, 1948−1954.
(
43) Winn, M. D., Isupov, M. N., and Murshudov, G. N. (2001) Use
of TLS parameters to model anisotropic displacements in
marcromolecular refinement. Acta Crystallogr., Sect. D: Biol. Crystallogr.
5
(
7, 122−133.
44) Pettersen, E. F., Goddard, T. D., Huang, C. C., Couch, G. S.,
Greenblatt, D. M., Meng, E. C., and Ferrin, T. E. (2004) UCSF
Chimera–a visualization system for exploratory research and analysis. J.
Comput. Chem. 25, 1605−1612.
(
45) Liberati, N. T., Urbach, J. M., Miyata, S., Lee, D. G., Drenkard,
E., Wu, G., Villanueva, J., Wei, T., and Ausubel, F. M. (2006) An
ordered, nonredundant library of Pseudomonas aeruginosa strain PA14
transposon insertion mutants. Proc. Natl. Acad. Sci. U.S.A. 103, 2833−
2
(
(
838.
46) Overhage, J., Bains, M., Brazas, M. D., and Hancock, R. E.
2008) Swarming of Pseudomonas aeruginosa is a complex adaptation
leading to increased production of virulence factors and antibiotic
resistance. J. Bacteriol. 190, 2671−2679.
(
47) Korgaonkar, A. K., and Whiteley, M. (2011) Pseudomonas
aeruginosa enhances production of an antimicrobial in response to N-
acetylglucosamine and peptidoglycan. J. Bacteriol. 193, 909−917.
2
200
dx.doi.org/10.1021/cb400345h | ACS Chem. Biol. 2013, 8, 2192−2200