Highly potent inhibitors of quorum sensing in staphylococcus aureus revealed through a systematic synthetic study of the group-III autoinducing peptide

Article abstract of DOI:10.1021/ja3112115

Methods to intercept bacterial quorum sensing (QS) have attracted significant attention as potential anti-infective therapies. Staphylococcus aureus is a major human pathogen that utilizes autoinducing peptide (AIP) signals to mediate QS and thereby regulate virulence. S. aureus strains are categorized into four groups (I-IV) according to their AIP signal and cognate extracellular receptor, AgrC. Each group is associated with a certain disease profile, and S. aureus group-III strains are responsible for toxic shock syndrome and have been underestimated in other infections to date. A limited set of non-native AIP analogs have been shown to inhibit AgrC receptors; such compounds represent promising tools to study QS pathways in S. aureus. We seek to expand this set of chemical probes and report herein the first design, synthesis, and biological testing of AIP-III mimetics. A set of non-native peptides was identified that can inhibit all four of the AgrC receptors (I-IV) with picomolar IC₅₀ values in reporter strains. These analogs also blocked hemolysis by wild-type S. aureus group I-IV strains - a virulence trait under the control of QS - at picomolar concentrations. Moreover, four of the lead AgrC inhibitors were capable of attenuating the production of toxic shock syndrome toxin-1 (also under the control of QS) by over 80% at nanomolar concentrations in a wild-type S. aureus group-III strain. These peptides represent, to our knowledge, the most potent synthetic inhibitors of QS in S. aureus known, and constitute new and readily accessible chemical tools for the study of the AgrC system and virulence in this deadly pathogen.

Full text of DOI:10.1021/ja3112115

Article
pubs.acs.org/JACS
Highly Potent Inhibitors of Quorum Sensing in Staphylococcus aureus
Revealed Through a Systematic Synthetic Study of the Group-III
Autoinducing Peptide
Yftah Tal-Gan,†,§ Danielle M. Stacy,†,§ Mary K. Foegen,‡ David W. Koenig,‡ and Helen E. Blackwell*,†
†Department of Chemistry, University of Wisconsin−Madison, 1101 University Avenue, Madison, Wisconsin 53706, United States
‡Kimberly-Clark Corporation, 2100 Winchester Road, Neenah, Wisconsin 54956, United States
S
* Supporting Information
ABSTRACT: Methods to intercept bacterial quorum sensing
(QS) have attracted significant attention as potential anti-
infective therapies. Staphylococcus aureus is a major human
pathogen that utilizes autoinducing peptide (AIP) signals to
mediate QS and thereby regulate virulence. S. aureus strains are
categorized into four groups (I−IV) according to their AIP
signal and cognate extracellular receptor, AgrC. Each group is
associated with a certain disease profile, and S. aureus group-III
strains are responsible for toxic shock syndrome and have been
underestimated in other infections to date. A limited set of
non-native AIP analogs have been shown to inhibit AgrC receptors; such compounds represent promising tools to study QS
pathways in S. aureus. We seek to expand this set of chemical probes and report herein the first design, synthesis, and biological
testing of AIP-III mimetics. A set of non-native peptides was identified that can inhibit all four of the AgrC receptors (I−IV) with
picomolar IC50 values in reporter strains. These analogs also blocked hemolysis by wild-type S. aureus group I−IV strainsa
virulence trait under the control of QSat picomolar concentrations. Moreover, four of the lead AgrC inhibitors were capable of
attenuating the production of toxic shock syndrome toxin-1 (also under the control of QS) by over 80% at nanomolar
concentrations in a wild-type S. aureus group-III strain. These peptides represent, to our knowledge, the most potent synthetic
inhibitors of QS in S. aureus known, and constitute new and readily accessible chemical tools for the study of the AgrC system
and virulence in this deadly pathogen.
considerable recent attention as a novel anti-infective
target.12,13 In turn, the dependence of bacteria on small
molecule and peptidic signals for QS has ignited interest in the
development of non-native ligands capable of blocking QS
pathways.14 In contrast to antibiotics, which target bacterial
pathways that are essential for survival,15,16 QS antagonists
could provide an alternative anti-infective therapy17,18 that does
not place selective pressure on the bacterial population to
develop resistance.19 This is especially important in the case of
S. aureus, which rapidly develops resistance to antibiotics,
including to the once last-resort antibiotic vancomycin.20
S. aureus is a Gram-positive bacterium that uses QS to
establish both acute and chronic infections.21,22 This pathogen
produces an arsenal of virulence factors, including tissue-
degrading enzymes, immune evasion factors, and pore-forming
toxins (hemolysins), all of which are regulated by its accessory
gene regulator (agr) QS system.23−25 The agr system is
comprised of four components, termed AgrA−D (illustrated in
Figure 1A), and is centered on the autoinducing peptide (AIP)
QS signal. AgrB is an integral membrane endopeptidase that
INTRODUCTION
■
Many bacteria utilize chemical signals to assess their local
population densities in a process termed quorum sensing
(QS).1−3 This intercellular signaling process effectively allows
bacteria to “count” themselves and behave as a multicellular
group at high cell number. While the specifics may vary
between species, QS circuits share general organizing
principles: bacteria produce, secrete, and detect signal
molecules referred to as autoinducers. At high population
densities in a given environment, the autoinducers will reach a
sufficiently high concentration to bind and activate their
cognate receptors. Signal:receptor binding then alters the
expression of genes involved in bacterial group behaviors, such
as swarming, sporulation, bioluminescence, conjugation, biofilm
formation, and virulence factor production.4−6 These pheno-
types can have widespread and sometimes devastating effects
on human health, agriculture, and the environment.7,8 For
example, many pathogenic bacteria utilize QS to launch
synchronized attacks on their hosts only after they have
achieved a high cell density, thereby overwhelming the host’s
defense mechanisms.9−11 As several prevalent human patho-
gens (e.g., Staphylococcus aureus, Pseudomonas aeruginosa, and
Vibrio cholerae) use QS to control virulence, QS has received
Received: November 14, 2012
Published: May 6, 2013
© 2013 American Chemical Society
7869
dx.doi.org/10.1021/ja3112115 | J. Am. Chem. Soc. 2013, 135, 7869−7882

Journal of the American Chemical Society
Article
Figure 1. Schematic of the agr quorum sensing circuit in S. aureus. (A) (a) The precursor peptide AgrD is processed by AgrB. (b) The mature AIP
signal is secreted across the cell membrane. (c) AIP binds the extracellular domain of AgrC. (d) The histidine kinase domain of AgrC phosphorylates
AgrA. (e) AgrA binds the P2 and P3 promoters to autoactivate the agr system and upregulate RNAIII transcription. (f) In certain group-III S. aureus
strains, RNAIII promotes the production of TSST-1. (B) Structures and sequences of the four AIP signals (I−IV) corresponding to the four S. aureus
groups (I−IV). Letters represent amino acid codes.
converts the precursor of the AIP signal, AgrD, to the mature
AIP. This conversion involves cyclization of AgrD via a cysteine
sulfhydryl group and its C-terminus to form the AIP as a 16-
atom thiolactone macrocycle with an N-terminal exocyclic tail
(Figure 1B). AgrB is also involved in the secretion of AIP
across the cell membrane. Once a threshold extracellular
concentration of AIP is reached, the AIP ligand binds to its
target receptor AgrC, a transmembrane histidine kinase. The
AIP:AgrC complex acts to phosphorylate the intracellular
response regulator, AgrA. Phosporylated AgrA then binds to
the P2 and P3 promoters to autoinduce the agr system and
upregulate RNAIII transcription, respectively.26 RNAIII thus
represents the main effector of the agr system and regulates the
production of many virulence factors and surface proteins
associated with biofilm production.27
There is a hypervariable region within the S. aureus agr
operon that has led to the classification of four agr specificity
groups of S. aureus (I−IV) with distinct AIP and AgrC
sequences.28−30 The structures of the four AIP signals (I−IV)
are shown in Figure 1B; all have a conserved 16-atom
thiolactone macrocycle, and AIPs-I and -IV share a nearly
identical primary sequence, while AIP-II and AIP-III have more
dissimilar primary sequences. The four different S. aureus agr
groups have been correlated with specific disease types: group-I
and -II are associated with the majority of invasive
infections,31−33 while group-IV is considered rare and limited
to exfoliative toxin-related syndromes.31 At first, group-III S.
aureus was also considered rare. However, recent studies have
revealed that the prevalence of this group has been under-
estimated in infections, and in fact, it is the most abundant
group in nasal carriage cases and is predominately responsible
for toxic shock syndrome (TSS) in humans.31,32 Toxic shock
syndrome toxin-1 (TSST-1) is the causative agent in all cases of
menstrual TSS and most cases of nonmenstrual TSS.31,34
Notably, TSST-1 production is directly regulated by the agr-III
QS system (Figure 1A).23,34,35 As such, methods to inhibit the
agr-III system in S. aureus could provide new insights into and
therapeutic strategies for this deadly disease.
QS is dependent on autoinducer:receptor binding, and the
development of chemical agents capable of blocking this
binding event has been a focus of considerable research over
the past ∼20 years.36−40 Such abiotic agents represent
promising tools to further elucidate the role of QS pathways
in bacterial virulence and other group behaviors.12,14,41−43 Our
laboratory has been actively engaged in the development of
small molecule QS modulators, largely in Gram-negative
bacteria that use LuxR-type receptors for QS.37,44−48 The
development of small molecule tools to probe AgrC signaling in
S. aureus has proceeded more slowly in the research
community, however.49 We note that Janda and co-workers
have reported a complementary, macromolecular strategy based
on antibodies that sequester the AIP ligand away from AgrC
and effectively “quench” QS in group-IV S. aureus.50,51
McCormick and co-workers have also shown that naturally
occurring cyclic dipeptides produced by Lactobacillus reuteri
((cyclo-(Tyr-Pro) and cyclo-(Phe-Pro)) can modulate the agr
system in S. aureus (albeit at mid- to high micromolar
concentrations), suggesting an interesting possibility for
interspecies signaling between Gram-positive bacteria using
small molecules.34 Nevertheless, new strategies are needed to
expand the chemical arsenal active against the agr QS system.
Early studies of the AgrC receptors revealed that each of the
four native AIPs were capable of cross-inhibiting the other
three, noncognate receptors.28,52−54 This activity has been
suggested to provide each group some competitive advantage
when establishing an infection and could explain in part the
predominance of a single S. aureus group in many infection
types (see above).28,55 In terms of the design of peptidic non-
native AgrC modulators, AIP-I and -II have received the most
scrutiny so far.29,52,53,56−59 Studies by Muir, Novick, Williams,
7870
dx.doi.org/10.1021/ja3112115 | J. Am. Chem. Soc. 2013, 135, 7869−7882

Journal of the American Chemical Society
Article
and co-workers closely examined the SAR of AIPs-I and -II60−62
and provided the first non-native mimetics of these peptides
that were capable of inhibiting both their cognate and
noncognate AgrC receptors in S. aureus. This group delineated
two key components of AIP:AgrC interactions: (1) initial
recognition of the AIP by an AgrC receptor, and (2) the
subsequent induction of allosteric changes within the AgrC
receptor that drives activation. In general, the AIP macrocycle
was found to be responsible for initial receptor recognition/
binding, and the AIP exocyclic tail then engaged in interactions
that resulted in receptor activation.62 For instance, the acyclic
native AIP-II was completely inactive, indicating that an intact
macrocyclic core was essential for activity.52 Removal of the
exocyclic tail converted the native AIPs-I and -II into self-
inhibitors, suggesting that these truncated peptides could bind
yet could not activate their cognate AgrC receptors. Likewise,
alterations to the native AIP exocyclic tails did not significantly
affect their cross-inhibitory activities against noncognate AgrCs,
which implied that these activities also result from macro-
cycle:receptor binding. Within the AIP-I and -II macrocycles,
the hydrophobic residues at the C-termini (residues 6−8 for
AIP-I and residues 8 and 9 for AIP-II; Figure 1B) were found to
be critical for both cognate and noncognate AgrC recognition/
binding.29,52,56 By combining these observations, Muir and co-
workers identified several potent and global inhibitors of all
four AgrC receptors (I−IV).56 Their most active inhibitor was a
truncated version of AIP-I that lacked an exocyclic tail and had
an aspartic acid (D) to alanine (A) mutation in the macrocyclic
core (tAIP-I D2A, shown in Figure 2).
and these prior studies provide a foundation on which to
expand the current set of non-native AgrC agonists and
antagonists.63 In the present study, we focused on AIP-III
(Figure 2) as the SARs that dictate its activity remain largely
unknown, and mimetics thereof are yet to be reported. The
prevalence of group-III S. aureus infections in human TSS (vide
supra) provides additional motivation for the development of
such molecules. Herein, we report the design, synthesis, and
systematic biological testing of a series of first- and second-
generation AIP-III mimetics. Evaluation of these mutant
peptides for their ability to attenuate each of the four AgrC
receptors (I−IV) revealed three key residues (i.e., Ile1, Asn2,
and Asp4) that can be modified in AIP-III to produce potent
AgrC inhibitors. Notably, a set of AIP-III analogs was identified
that can inhibit all four of the AgrC receptors with picomolar
IC50 values in cell-based reporter gene assays. All of these
analogs also blocked hemolysis by wild-type S. aureusa
virulence phenotype under the control of QSat picomolar
levels. Moreover, the lead compounds were capable of reducing
TSST-1 production levels in a wild-type S. aureus group-III
strain by over 80% at low nanomolar concentrations. The
results of this study are significant, as these peptides represent,
to our knowledge, the most potent S. aureus QS inhibitors to be
reported and suggest that, relative to the other AIPs, AIP-III
provides a superior scaffold for the development of peptide-
based AgrC inhibitors. More broadly, these new peptides
represent readily accessible chemical tools for the study of the
agr QS system and virulence in this deadly pathogen and
provide support for the potential development of QS inhibitors
as anti-infective agents.
Many questions remain about the mode of action of native
AIPs and the peptide-based AgrC modulators developed so far,
EXPERIMENTAL SECTION
■
Chemical Reagents and Instrumentation. All chemical reagents
were purchased from commercial sources (Alfa-Aesar, Sigma-Aldrich,
and Acros) and used without further purification. Solvents were
purchased from commercial sources (Sigma-Aldrich and J.T. Baker)
and used as obtained, with the exception of anhydrous dichloro-
methane (CH2Cl2), which was stored over molecular sieves. Water (18
MΩ) was purified using a Millipore Analyzer Feed System. Solid-phase
resin was purchased from Chem-Impex International. Cyclic dipeptide
(cyclo-(Tyr-Pro) and cyclo-(Phe-Pro)) controls were synthesized
according to our previously reported method.64
Reversed-phase high-performance liquid chromatography (RP-
HPLC) was performed using a Shimadzu system equipped with an
SCL-10Avp controller, an LC-10AT pump, an FCV-10ALvp solvent
mixer, and an SPD-10MAvp UV−vis diode array detector. Full details
of the HPLC columns and conditions used in this study are provided
in the Supporting Information. Matrix-assisted laser desorption
ionization time-of-flight mass spectrometry (MALDI-TOF MS) data
were obtained on a Bruker RELEX II spectrometer equipped with a
337 nm laser and a reflectron. In positive ion mode, the acceleration
voltage was 25 kV. Exact mass (EM) data were obtained on a Waters
(Micromass) LCT electrospray ionization TOF spectrometer. The
samples were sprayed with a sample cone voltage of 20 V.
Peptide Synthesis. Table 1 lists the peptides synthesized in this
study. Linear peptides were synthesized on Boc-protected, amino acid
preloaded 4-hydroxymethyl-phenylacetamidomethyl (PAM) resin
(0.6−0.8 mmol/g) using standard solid-phase synthesis protocols;
see Supporting Information for full details of peptide synthesis,
cleavage, HPLC purification, and MS analyses.
Representative Peptide Macrocyclization Protocol. Purified
peptide thioester was dissolved in a 60% guanidinium chloride (6 M
solution in 0.1 M phosphate buffer):40% ACN solution to a final
concentration ranging from 100 μM to 2 mM. The pH of the solution
was then adjusted to 6.8. The peptide was gently agitated using a
multipurpose rotator at 50 °C, and cyclization was monitored by
Figure 2. Structures of a known global AgrC inhibitor tAIP-I D2A
(top) and AIP-III (bottom). The two main structural elements of the
AIP signals (exocyclic tail and macrocycle core) are highlighted.
7871
dx.doi.org/10.1021/ja3112115 | J. Am. Chem. Soc. 2013, 135, 7869−7882

Journal of the American Chemical Society
Article
a
Table 1. Structures of the Peptides Synthesized in This Study
peptide name
sequence
peptide name
sequence
b
b
AIP-I
Y−S-T-(C-D-F-I-M)
G-V-N-A-(C-S-S-L-F)
Ac-(C-A-F-I-M)
AIP-III
I-N-(C-D-F-L-L)
b
b
AIP-II
AIP-IV
Y-S-T-(C-Y-F-I-M)
b
tAIP-I D2A
first generation analogs
second generation analogs
AIP-III D-I1
AIP-III D-N2
AIP-III D-C3
AIP-III D-D4
AIP-III D-F5
AIP-III D-L6
AIP-III D-L7
AIP-III I1A
AIP-III N2A
AIP-III D4A
AIP-III F5A
AIP-III L6A
AIP-III L7A
DI-N-(C-D-F-L-L)
I-DN-(C-D-F-L-L)
I-N-(DC-D-F-L-L)
I-N-(C-DD-F-L-L)
I-N-(C-D-DF-L-L)
I-N-(C-D-F-DL-L)
I-N-(C-D-F-L-DL)
A-N-(C-D-F-L-L)
I-A-(C-D-F-L-L)
I-N-(C-A-F-L-L)
I-N-(C-D-A-L-L)
I-N-(C-D-F-A-L)
I-N-(C-D-F-L-A)
AIP-III I1A/N2A
AIP-III I1A/D4A
AIP-III N2A/D4A
A-A-(C-D-F-L-L)
A-N-(C-A-F-L-L)
I-A-(C-A-F-L-L)
A-A-(C-A-F-L-L)
Ac-(C-D-F-L-L)
Ac-(C-A-F-L-L)
AIP-III I1A/N2A/D4A
tAIP-III
tAIP-III D2A
tAIP-III D2A/F3Y
tAIP-III D2A/F3W
Ac-AIP-III
Ac-(C-A-Y-L-L)
Ac-(C-A-W-L-L)
Ac-I-N-(C-D-F-L-L)
G-I-N-(C-D-F-L-L)
A-I-N-(C-D-F-L-L)
Y-I-N-(C-D-F-L-L)
G-AIP-III
A-AIP-III
Y-AIP-III
a
b
See Table S-1 for MS and HPLC characterization data. Represents control peptides.
analytical RP-HPLC. Upon completion, cyclic peptide was purified by
semipreparative RP-HPLC and lyophilized. The resulting white
powder was dissolved in a small portion of 1 M hydrochloric acid
(∼400 μL) and lyophilized prior to bioanalysis.
Biological Reagents and Strain Information. All standard
biological reagents were purchased from Sigma-Aldrich and used
according to enclosed instructions. Suspended rabbit blood cells (10%,
washed and pooled) were purchased from Lampire Biological
Laboratories and stored at 4 °C until use in the hemolysis assay.
Reagents for the TSST-1 enzyme-linked immunosorbent assay
(ELISA) were purchased from Toxin Technology, Inc. Tryptic soy
broth (TSB) and brain heart infusion (BHI) were prepared as
instructed with pH = 7.35.
Clear polystyrene 96-well EIA/RIA high-binding microtiter plates
(Costar) were used in the TSST-1 ELISA.
Reporter Gene Assay Protocol. Peptide stock solutions were
diluted with DMSO in serial dilutions (either 1:3, 1:5, or 1:10), and 2
μL of the diluted solution was added to each of the wells in a black 96-
well microtiter plate. An overnight culture of S. aureus gfp strain was
diluted 1:50 with fresh TSB (pH 7.35). A 198 μL portion of diluted
culture was added to each well of the microtiter plate containing
peptide. Plates were incubated at 37 °C for 24 h. Fluorescence (EX
500 nm/EM 540 nm) and OD600 of each well were then recorded
using a plate reader, and IC50 values were calculated.
For the competition assays, 2 μL of AIP-III D4A stock solution was
added to wells in a black 96-well microtiter plate to final
concentrations of 2 nM (group-I and -II strains) or 0.3 nM (group-
III and -IV strains). Native AIP (I−IV) stock solutions were diluted
with DMSO in serial dilutions (1:3 dilutions) and added to the wells
containing AIP-III D4A. The fluorescence assay was performed as
described above in the four respective S. aureus reporter strains.
Hemolysis Assay Protocol. Peptide stock solutions were diluted
with DMSO in serial dilutions (either 1:3, 1:5, or 1:10), and 2 μL of
the diluted peptide solution was added to each of the wells in a clear
96-well microtiter plate. An overnight culture of S. aureus wild-type
strain was diluted (1:10, 1:25, or 1:100, assay conditions dependent on
strain; see Table S-2 for additional detail) with fresh TSB. A 198 μL
portion of the diluted culture was added to each well of the microtiter
plate containing peptides. Plates were incubated statically at 37 °C for
6−8 h. The cultures were then assayed for hemolytic activity.
Suspended rabbit red blood cells (1 mL) were centrifuged (2000 rpm,
2 min, 25 °C, 450 g), the supernatant was removed, and the cells were
resuspended in saline phosphate buffer (PBS, pH = 7.35, 1 mL). After
the OD600 of each well of the 96-well microtiter plate was recorded, a
13 μL portion of the suspended red blood cells was added to each well.
After 15−25 min (see Table S-2) of static incubation at 37 °C, the
microtiter plates were centrifuged to pellet the cells (4 min, 25 °C, 450
g). A 150 μL portion of supernatant from each well of culture was
transferred to a fresh 96-well microtiter plate. Absorbance at 420 nm
was measured for each well using a plate reader, and IC50 values were
calculated.
The bacterial strains used in this study are listed in Table 2.
Bacterial cultures were grown in a standard laboratory incubator at 37
Table 2. S. aureus Strains Used in This Study Listed
According to Group
assay type
strain
ref.
Fluorescence
group-I
AH1677
AH430
51
51, 65
51
group-II
group-III
AH1747
AH1872
group-IV
51
Hemolysis and TSST-1
group-I
RN6390B
RN6923
MN8
27
28
66
30
group-II
group-III
group-IV
RN4850
°C with shaking (200 rpm) unless noted otherwise. The bacterial
dilutions and incubation periods were chosen in each assay to provide
the greatest dynamic range between positive and negative controls for
each bacterial strain. Absorbance and fluorescence measurements were
obtained using a Biotek Synergy 2 microplate reader using Gen5 data
analysis software. All biological assays were performed in triplicate.
IC50 values were calculated using GraphPad Prism software (v. 4.0)
using a sigmoidal curve fit.
Compound Handling Protocol. Stock solutions of synthetic AIP
analogs (1 mM) or cyclic dipeptides (10 mM) were prepared in
DMSO and stored at 4 °C in sealed vials. The amount of DMSO used
in biological assays did not exceed 2% (v/v). Black or clear polystyrene
96-well microtiter plates (Costar) were used for bacteriological assays.
TSST-1 ELISA Protocol. The TSST-1 ELISA protocol was based
in part on the procedure supplied by the toxin producer (Toxin
Technologies, Inc.) with some modifications. Peptide stock solutions
(AIP-III analog or cyclic dipeptide control) were diluted with DMSO
to the desired concentration, and 20 μL of the diluted solutions was
added to 15 mL Falcon tubes. An overnight culture of S. aureus MN8
was diluted 1:100 with fresh BHI, and a 1 mL portion of diluted
7872
dx.doi.org/10.1021/ja3112115 | J. Am. Chem. Soc. 2013, 135, 7869−7882

Journal of the American Chemical Society
Article
culture was added to each Falcon tube containing peptide. The
cultures were incubated at 37 °C for 24 h. Simultaneously, rabbit
polyclonal anti-TSST-1 IgG (100 μL, 10 μg/mL in coating buffer
(0.01 M sodium carbonate, pH 9.6)) was added to a 96-well ELISA
plate and incubated with shaking (200 rpm) in a humid chamber (i.e.,
a sealed plastic box containing several water-soaked paper towels) at
37 °C for 18 h. The ELISA plate was then washed 3× (300 μL each)
with PBS solution containing 0.05% Tween-20 (PBS-Tween). To
block the plate, bovine serum albumin (BSA, 1% in PBS-Tween, 100
μL) was added to each well and incubated for 15 min at rt, after which
it was washed 3× again with PBS-Tween (300 μL each).
Scheme 1. Solid-Phase Synthetic Route to AIPs and
Analogs
a
The bacterial cultures were centrifuged to pellet the cells (10 min,
25 °C, 450 g), and the supernatants were sterile filtered. The
supernatants were diluted 1:10 in normal rabbit serum (NRS, 1% in
PBS-Tween) and incubated for 15 min at rt. The incubated solutions
were further diluted using PBS-Tween to a total dilution of 1:100 to
1:2000. The diluted supernatant samples (100 μL) and TSST-1
standards with known concentrations (ranging from 10 ng/mL to 0.32
ng/mL, 100 μL) were added to the ELISA plate and incubated with
shaking (200 rpm) in a humid chamber at 37 °C for 2 h. After
incubation, the plate was washed 3× with PBS-Tween (300 μL each),
and anti-TSST-1 IgG horseradish peroxidase conjugate (100 μL, 3.33
μg/mL) was added and incubated with shaking (200 rpm) in a humid
chamber at 37 °C for 1 h. The plate was washed 5× with PBS-Tween
(300 μL each), after which 2,2′-azinobis(3-ethylbenzthiazoline-sulfonic
acid) (ABTS solution, 100 μL) was added and incubated at rt for 15
min. Absorbance at 405 nm was measured using a plate reader, and a
TSST-1 standard curve was constructed for each assay plate using the
data for the TSST-1 reference standards (the R2 of linear regression
analysis for these data was ≥0.99). The TSST-1 concentration in the
test samples was then determined from the regression equation and
presented as a percentage of the TSST-1 concentration in the
untreated cultures.
a
See Experimental Section for further details. TFA = trifluoroacetic
acid; SPPS = solid-phase peptide synthesis; GndHCl = guanidinium
chloride.
RESULTS AND DISCUSSION
■
Design of First-Generation AIP-III Analogs. We initiated
our research by independently evaluating each residue of AIP-
III through systematic alanine and D-amino acid point
mutations. This initial set of 13 peptides was designed to
identify key residues and stereocenters for AIP-III:AgrC
interactions. Six of the seven residues could be modified in
the alanine scan (Cys3 had to be maintained for thiolactone
formation), while all seven residues were evaluated in the D-
amino acid scan. The 13 AIP-III analogs were generated by
solid-phase synthesis (Table 1), as detailed below.
method of Hilvert and co-workers gave the linear peptide
thioesters,68,69 which were then purified to homogeneity by
semipreparative RP-HPLC and isolated in acceptable yields
(25−50%).
We next explored intramolecular thiol-thioester exchange
reactions in a range of buffers to effect macrocyclization of the
linear peptides (Scheme 1). However, the previously reported
cyclization buffers and conditions proved ineffective in our
hands, even after 24 h.56,58 We therefore undertook
optimization studies to develop our own macrocyclization
conditions and found that performing the reaction in 6 M
guanidinium chloride in 0.1 M phosphate buffer and ACN
(3:2) at a pH = 6.8 gave quantitative macrocyclization within
24 h at room temperature. Elevating the temperature to 50 °C
reduced reaction times to <2 h (see Figure S-1 for HPLC
analyses of macrocyclization reactions). We therefore used
these conditions to effect macrocyclization of all the AIP
analogs prepared in this study (Table 1; see Supporting
Information for full characterization details). We also used this
synthesis protocol to generate the four native AIPs I−IV and
the known global AgrC inhibitor, tAIP-I D2A (Figure 2), for
use as key controls in our biological experiments (see below).
Overall, this synthetic route to AIPs represents an improvement
over previously reported methods52,56,58 and will facilitate
research in this general area. Further, the thiol-thioester
exchange reaction conditions could also prove useful in other
contexts, for example, in the total synthesis of proteins using
native chemical ligation.70
Synthesis of AIPs and Analogs. Two solid-phase
synthesis approaches have been implemented in the past to
construct native AIPs and their analogs. The first approach
utilizes chemoselective cleavage of the linear, protected peptide
from the solid support with concomitant unmasking of the
cysteine sulfhydryl group. The protected peptide can then be
macrocyclized via a carbodiimide coupling and subsequently
deprotected.29,57,67 This approach, however, is limited in part
by the poor solubilities of the protected peptides and by low
macrocyclization efficiencies due to the steric bulk of the
protecting groups. The second synthetic approach addresses
these challenges by incorporating an initial global deprotection
step prior to a solution phase, chemoselective thiol-thioester
exchange to form the macrocyclic products.52,56,58 This latter
approach is generally more efficient and has the potential for
even further improvement; we therefore selected this strategy
to synthesize our AIP-III analogs. First, we utilized standard
Fmoc/tBu solid-phase peptide synthesis methods to generate
the linear peptides on 4-hydroxymethyl-phenylacetamido-
methyl (PAM) polystyrene resin (Scheme 1; see Experimental
Section). Cleavage and global deprotection according to the
S. aureus Reporter Gene Assays. We next tested the
ability of the 13 alanine and D-amino acid AIP-III mutants to
7873
dx.doi.org/10.1021/ja3112115 | J. Am. Chem. Soc. 2013, 135, 7869−7882

Journal of the American Chemical Society
Article
Table 3. IC50 Values of the Alanine and D-Amino Acid Scan Analogs of AIP-III Against AgrC I−IV Determined Using S. aureus
a
Fluorescence Reporter Strains
b
b
b
b
peptide name
sequence
AgrC-I IC50 (nM)
AgrC-II IC50 (nM)
AgrC-III IC50 (nM)
AgrC-IV IC50 (nM)
AIP-III D-I1
AIP-III D-N2
AIP-III D-C3
AIP-III D-D4
AIP-III D-F5
AIP-III D-L6
AIP-III D-L7
AIP-III I1A
AIP-III N2A
AIP-III D4A
AIP-III F5A
AIP-III L6A
AIP-III L7A
DI-N-(C-D-F-L-L)
I-DN-(C-D-F-L-L)
I-N-(DC-D-F-L-L)
I-N-(C-DD-F-L-L)
I-N-(C-D-DF-L-L)
I-N-(C-D-F-DL-L)
I-N-(C-D-F-L-DL)
A-N-(C-D-F-L-L)
I-A-(C-D-F-L-L)
I-N-(C-A-F-L-L)
I-N-(C-D-A-L-L)
I-N-(C-D-F-A-L)
I-N-(C-D-F-L-A)
Y-S-T-(C-D-F-I-M)
G-V-N-A-(C-S-S-L-F)
I-N-(C-D-F-L-L)
Y-S-T-(C-Y-F-I-M)
Ac-(C-A-F-I-M)
(Y-P)
8.42
2.15
16.4
2.45
>200
24.5
78.3
17.8
77.7
6.23
>200
138
>200
>200
>200
>200
>200
29.2
c,d
>200
>200
12.0
>200
174
e
−
>200
10.5
e
5.36
−
17.9
4.26
194
7.85
c,d
3.60
0.732
0.429
−
3.53
0.485
>200
>200
>200
0.0506
>200
>200
>200
0.522
0.532
0.0349
118
c
>200
>200
>200
8.00
>200
>200
f
e
c,d
AIP-I
−
−
f
e
AIP-II
1.62
−
0.396
f
e
AIP-III
5.05
5.63
0.373
10.1
−
8.53
f
c
,
d
e
AIP-IV
−
0.460
−
f
tAIP-I D2A
3.06
0.260
0.353
f
f
g
g
g
g
Cyclo(Tyr-Pro)
−
−
−
−
−
g
g
g
g
Cyclo(Phe-Pro)
(F-P)
−
−
−
a
See Experimental Section for details of reporter strains and methods. See Supporting Information for plots of antagonism dose response curves. All
b
assays performed in triplicate. IC50 values determined by testing peptides over a range of concentrations (200 fM − 100 μM). See Figure 4 and
Supporting Information for 95% confidence ranges. Dose response curve did not reach 100% inhibition over the concentrations tested. Inhibition
dose response curve upturned at higher concentrations, potentially indicative of partial agonism (see text). Dose response curve revealed agonism
and no antagonism. Control compound. No activity at any concentration tested up to at least 100 μM. Higher concentrations could not be tested
c
d
e
f
g
due to compound insolubility.
modulate the activity of the S. aureus AgrC I−IV receptors
using fluorescence-based reporter gene assays (see Experimen-
tal Section for protocol). Each compound was tested in group
I−IV methicillin-resistant S. aureus strains harboring P3-gfp
reporter plasmids (listed in Table 2).51,65 In these reporter
plasmids, the agr P3 promoter, typically upstream of the main
QS effector RNAIII (Figure 1A), is also upstream of gfp. Thus,
when bacterial cell densities and AIP concentrations are high,
the AIP:AgrC complex will phosphorylate AgrA, which will
then bind P3 and transcribe gfp (in addition to typical
upregulation of RNAIII). GFP fluorescence can then be
quantified to determine the extent of AgrC activation, and in
these wild-type strains (producing native levels of AIP) will be
observable in the absence of an exogenous AgrC modulator.
Compounds capable of reducing fluorescence levels (or
increasing these levels over background), therefore, can be
classified as AgrC inhibitors (or activators). We note that
previous researchers have utilized a similar strategy to identify
AgrC modulators in all four groups of S. aureus but used a
group-I agr-null strain carrying a plasmid containing a P3-blaZ
fusion (to confer β-lactamase activity in response to AgrC
activation) and agrCA from groups-I, -II, -III, or -IV.53,56 The
gfp reporters utilized here allow for the analysis of analogs in all
four wild-type strains producing native levels of AIPs I−IV. In
order to have controls for comparison to our AIP-III analogs,
we also evaluated the activities of the native AIPs I−IV, Muir’s
previously reported global AgrC inhibitor tAIP-I D2A,56 and
McCormick’s previously reported cyclic dipeptide agr modu-
lators ((cyclo-(Tyr-Pro) and cyclo-(Phe-Pro))34 in each S.
aureus gfp-reporter strain. The inhibitory trends for the native
AIPs I−IV and tAIP-I D2A in each of the four AgrC receptors
were comparable to previous reports using alternate S. aureus
AgrC I−IV reporter strains.56 The two cyclic dipeptides were
inactive in the four reporter strains over the concentration
range tested (suggestive of IC50 values much >100 μM), which
was also congruent with the study of McCormick and co-
workers.34 Table 3 summarizes the activities of these control
peptides and the first-generation AIP-III analogs against AgrCs
I−IV in the gfp-reporter strains.
The reporter gene assay data revealed several interesting SAR
trends for AIP-III. Moreover, a number of new, global AgrC
inhibitors were uncovered with either comparable or more
potent activities than the known inhibitor tAIP-I D2A (Table
3). As discussed above, there are two main components to
AIP:AgrC interactions: (1) the initial recognition of the AIP by
an AgrC receptor and (2) the resultant induction of allosteric
changes within AgrC that drives activation. To explore these
components for AIP-III, we examined two different SAR trends
for the AIP-III analogs: (1) cross-inhibition of AgrC-I, -II, and
-IV, and (2) activation of AgrC-III. Each is discussed in turn
below.
Consistent with prior observations of other AIPs,29,52,56 D-
amino acid and alanine replacement of either of the exocyclic
tail residues (Ile1 or Asp2) resulted in AIP-III analogs with
similar cross-inhibitory activities to the parent AIP-III (Table 3,
IC50 values within error or <10-fold change). In addition,
replacement of any one of the three hydrophobic endocyclic
residues in AIP-III (Phe5, Leu6, or Leu7) with alanine resulted
in a significant loss of inhibition in groups-I, -II, and -IV (>10-
fold change relative to AIP-III). Replacement of Phe5 and Leu6
with their D-amino acid counterparts further demonstrated the
stringent requirements of these two residues for inhibitory
activity against AgrC receptors, as these mutants displayed a >
20-fold loss in inhibition in almost every case. However, AIP-III
7874
dx.doi.org/10.1021/ja3112115 | J. Am. Chem. Soc. 2013, 135, 7869−7882

Journal of the American Chemical Society
Article
D-L7 displayed analogous inhibitory activities as the parent AIP-
III in groups-I, -II and-IV (within error), suggesting that Leu7
may not enforce stereodefined interactions of AIP-III with
these AgrC receptors. This observation is congruent with
previous findings for AIP-I by Williams and co-workers.29 We
note two exceptions to these activity trends: replacing the
endocyclic Phe5 or Leu6 with their D-isomers (in AIP-III D-F5
and AIP-III D-L6) neither maintained nor abolished inhibitory
activity against one receptor, AgrC-II. These two analogs
appeared to activate AgrC-II instead. The extent of this
activation relative to native AIP could not be fully explored
using the gfp-reporter assays in this study, however, as the
native AIP signals are produced at normal background levels in
these S. aureus reporter strains (see above). Alternate biological
assays are required to study this phenomenon further and are
ongoing.
The remaining two residues of AIP-III, Cys3 and Asp4, were
found to contribute significantly to cross-receptor inhibition.
The replacement of Cys3 with its D-isomer significantly reduced
inhibition (>20-fold change relative to native AIP-III; Table 3).
This reduction may be due to a conformational change caused
by the reversed stereogenic center of cysteine, forcing a change
in the orientation of the key hydrophobic residues in the
macrocyclic backbone, thereby preventing AIP:AgrC recog-
nition interactions. Additional studies are certainly needed to
confirm this hypothesis. Replacement of Asp4 with its D-isomer
also reduced inhibition against AgrC-I by 28-fold relative to
AIP-III, yet reduced inhibitory activity to a lesser degree against
AgrC-II and AgrC-IV (4- and 3-fold relative to AIP-III,
respectively). This result suggests that the stereochemistry of
Asp4 could reinforce optimal orientations of Phe5 and Leu6 for
interactions with AgrC-I. Perhaps more notable, however, was
that the replacement of Asp4 with alanine (AIP-III D4A)
increased the inhibitory activity of AIP-III by at least 10-fold in
each group, delivering a picomolar global AgrC inhibitor.
Interestingly, AIP-III D4A contains the identical mutation as
the previously reported global inhibitor generated from the
AIP-I scaffold, tAIP-I D2A (residue numbering shifted (4→2)
due to truncated structure), yet is a 5- to 24-fold more active
inhibitor in each group (largest increase in potency against
AgrC-II; see Table 3). AIP-III D4A represented not only the
most active inhibitor identified in this first series of AIP-III
derivatives, but also, to our knowledge, is the most potent AgrC
inhibitor to be reported. The discovery of this lead compound
from such a limited and structurally similar set of AIP-III
derivatives provides strong support for further study of the AIP-
III scaffold for the development of global AgrC inhibitors. A
summary of the key SARs for cross-inhibition of AgrC-I, -II,
and -IV by the AIP-III analogs is provided in Figure 3.
Analysis of the gfp screening data for the AIP-III analogs also
provided valuable SAR trends for the activation of AgrC-III by
AIP-III (Table 3). Although native AIP signals were present in
these wild-type S. aureus reporter strains, we could infer AgrC-
III inhibition as the competitive binding of the non-native AIP
to AgrC-III and either lack of or partial AgrC-III activation.
(We return to the reasoning behind this competitive
mechanism argument below.) As expected, in view of previous
studies of AIP-I and -II analogs,29,52 the exocyclic residue Ile1
of AIP-III appeared to play an important role in AgrC-III
activation. Replacing this residue with either alanine or D-Ile1
converted AIP-III into a weak antagonist (IC50 values = 194 or
78.3 nM, respectively). Converting the exocyclic residue Asn2
to its D-isomer yielded an even stronger AgrC-III antagonist
Figure 3. Summary of key SAR trends for inhibition of AgrC-I, -II, and
-IV by AIP-III analogs as revealed by the gfp-reporter assay data. See
text.
(IC50 value = 17.8 nM). However, replacing Asn2 with alanine
maintained agonistic activity, suggesting that the stereochemical
presentation (and concomitant conformational constraints) of
the peptide backbone may be more essential for AgrC-III
activation relative to the composition of the Asn2 side chain.
Within the AIP-III macrocycle, the two hydrophobic residues
Phe5 and Leu6 proved to be crucial for AgrC-III activation, as
D-isomer or Ala replacements at either of these positions
yielded weak AgrC-III inhibitors (IC50 values > 200 nM; Table
3). The AIP-III D-L7 mutant maintained the agonistic activity
of the native AIP-III, while AIP-III L7A was also a weak
inhibitor, similar to the AIP-III F5A and AIP-III L6A mutants.
This disparate activity trend for the Leu7 mutants suggests that
stereochemistry at this residue may not play as a major of a role
in AgrC-III activation relative to that for Phe5 and Leu6. We
note that the activity profiles for these Phe5, Leu6, and Leu7
mutants in AgrC-III are largely consistent with the observed
7875
dx.doi.org/10.1021/ja3112115 | J. Am. Chem. Soc. 2013, 135, 7869−7882

Journal of the American Chemical Society
Article
Table 4. IC50 Values of the Alanine and D-Amino Acid Scan Analogs of AIP-III for Inhibition of Hemolysis by Group-I−IV S.
a
aureus Strains
b
b
b
b
peptide name
sequence
group-I IC50 (nM)
group-II IC50 (nM)
group-III IC50 (nM)
group-IV IC50 (nM)
c
AIP-III D-I1
AIP-III D-N2
AIP-III D-C3
AIP-III D-D4
AIP-III D-F5
AIP-III D-L6
AIP-III D-L7
AIP-III I1A
AIP-III N2A
AIP-III D4A
AIP-III F5A
AIP-III L6A
AIP-III L7A
DI-N-(C-D-F-L-L)
I-DN-(C-D-F-L-L)
I-N-(DC-D-F-L-L)
I-N-(C-DD-F-L-L)
I-N-(C-D-DF-L-L)
I-N-(C-D-F-DL-L)
I-N-(C-D-F-L-DL)
A-N-(C-D-F-L-L)
I-A-(C-D-F-L-L)
I-N-(C-A-F-L-L)
I-N-(C-D-A-L-L)
I-N-(C-D-F-A-L)
I-N-(C-D-F-L-A)
Y-S-T-(C-D-F-I-M)
G-V-N-A-(C-S-S-L-F)
I-N-(C-D-F-L-L)
Y-S-T-(C-Y-F-I-M)
Ac-(C-A-F-I-M)
18.4
5.31
2.67
0.222
122
>200
>200
23.1
cd
,
−
>200
77.0
>200
>200
>200
>200
>200
d
1.15
>200
>200
>200
29.3
>200
>200
1.96
>200
>200
47.8
12.5
2.64
0.106
>200
>200
>200
189
e
−
f
4.61
1.29
−
c
,
d
1.02
0.137
0.0596
44.7
−
0.0820
>200
>200
>200
--
0.163
>200
>200
>200
6.12
3.59
--
>200
>200
3.34
g
AIP-I
g
AIP-II
0.890
8.07
--
1.19
23.8
--
g
AIP-III
0.456
0.0897
2.50
g
cd
,
AIP-IV
−
1.49
0.853
g
tAIP-I D2A
1.45
0.361
a
See Experimental Section for details of strains and methods. See Supporting Information for plots of antagonism dose response curves. All assays
b
performed in triplicate. -- indicates not tested. IC50 values determined by testing AIPs over a range of concentrations (200 fM to 10 μM). See Figure
4 and Supporting Information for 95% confidence ranges. Dose response curve did not reach 100% inhibition over the concentrations tested. Dose
response curve upturned at higher concentrations. Dose response curve revealed agonism and no antagonism. No activity at any concentration
tested up to at least 10 μM. Control compound.
c
d
e
f
g
inhibition trends against AgrC-I, -II, and -IV described above,
and support Phe5 and Leu6 as the key hydrophobic endocyclic
residues for the modulation of cognate and noncognate AgrC
receptors by AIP-III.
However, such responses, while informative, may not accurately
reflect physiologically relevant QS phenotypes. We therefore
sought to develop an additional assay that would permit us to
screen our AIP-III analogs for modulation of a physiologically
relevant QS phenotype in S. aureus. Previous studies have
established that the production of hemolysins is regulated by
the agr QS system.23−25 Therefore, analogs that inhibit the agr
system should also block the production of hemolysins in S.
aureus, and we reasoned that this outcome could be readily
quantitated by hemolysis assays using red blood cells. To this
end, we modified a standard bacterial hemolysis assay to a 96-
well microtiter plate format to expedite compound screen-
ing72,73 and tested the AIP-III analogs for their ability to inhibit
hemolysis by S. aureus. This work represents, to our knowledge,
the first application of hemolysis assays for the study of non-
native AIPs as QS modulators.
Replacement of the AIP-III Cys3 with its D-isomer gave full
AgrC-III antagonism only at high concentrations (IC50 value >
200 nM; Table 3), suggesting that the stereochemistry of this
residue is important for AgrC-III activation. Substitution of
Asp4 with D-Asp yielded a similarly weak inhibitor. As discussed
above, however, replacing this same residue with alanine (i.e.,
AIP-III D4A) produced the most potent AgrC-III inhibitor in
this series (IC50 value = 0.0506 nM). Together, these SAR
trends indicate that the side-chain of Asp4 may play a major
role in AgrC-III activation but not in initial AIP-III:AgrC-III
binding, as AIP-III D4A could strongly inhibit AgrC-III but was
incapable of activation.
Throughout these SAR analyses, we made the assumption
that the AIP-III analogs were eliciting their activity through
directly binding AgrC receptors and outcompeting the native
AIP signals. Such a hypothesis is reasonable in view of the close
structural similarity of these mutants to the native AIP signals
and has been made in prior studies of AIP-I and -II analogs.71
However, to provide further support for this hypothesis, we
performed a competition assay between the native AIPs and
our most potent AgrC-III inhibitor, AIP-III D4A, using the gfp-
reporter strains (see Experimental Section). We observed that
the native AIP signals could be added in a dose-dependent
manner to completely eliminate AIP-III D4A inhibition and
recover gfp production in each of the four S. aureus groups (see
Figure S-2). These data serve to support a competitive
mechanism by which AIP-III D4A and the related analogs in
this study modulate AgrC activity.
Briefly, we treated wild-type strains of groups-I−IV S. aureus
(listed in Table 2) with the AIP-III analogs and evaluated the
cultures for hemolysin activity (see Experimental Section for
details). Rabbit red blood cells were incubated with peptide-
treated bacterial cultures (∼15 min) in a microtiter plate, after
which the samples were pelleted by centrifugation. The culture
supernatant was then transferred to new plates, and the
concentration of free heme (directly correlated with red blood
cell lysis) was quantified by measuring absorbance at 420 nm.
Similar to the gfp-reporter assays outlined above, we used the
native AIPs (I−IV) and the previously reported global inhibitor
tAIP-I D2A as key controls in the hemolysis assays (the cyclic
dipeptides were omitted as controls due to their low activities
in the gfp-reporter assays). Table 4 summarizes the hemolysis
assay data for the control peptides and the 13 AIP-III analogs.
With a few minor exceptions, the relative IC50 value trends
for the controls and AIP analogs in the hemolysis assay were
largely identical to those in the gfp-reporter assays, validating
our hemolysis assay protocol as a straightforward method for
S. aureus Hemolysis Assays. Bacterial reporter strains,
such as the S. aureus gfp-reporters above, can certainly facilitate
the rapid screening of compounds for potential QS modulators.
7876
dx.doi.org/10.1021/ja3112115 | J. Am. Chem. Soc. 2013, 135, 7869−7882

Journal of the American Chemical Society
Article
Figure 4. IC50 values for first-generation AIP-III analogs as determined by the gfp-reporter assay (dark gray) and hemolysis assay (light gray) against:
(A) AgrC-I/group-I; (B) AgrC-II/group-II; (C) AgrC-III/group-III; and (D) AgrC-IV/group-IV. Only calculated values (below the 200 nM
threshold) are presented. 95% confidence intervals indicated. For a complete listing of data, see Tables 3 and 4.
Table 5. IC50 Values of the Second-Generation AIP-III Analogs Against AgrC I−IV Determined Using S. aureus Fluorescence
a
Reporter Strains
b
b
b
b
peptide name
sequence
AgrC-I IC50 (nM)
AgrC-II IC50 (nM)
AgrC-III IC50 (nM)
AgrC-IV IC50 (nM)
AIP-III I1A/N2A
AIP-III I1A/D4A
AIP-III N2A/D4A
AIP-III I1A/N2A/D4A
tAIP-III
A-A-(C-D-F-L-L)
A-N-(C-A-F-L-L)
I-A-(C-A-F-L-L)
A-A-(C-A-F-L-L)
Ac-(C-D-F-L-L)
Ac-(C-A-F-L-L)
7.40
0.328
0.331
0.304
26.7
4.38
2.35
0.204
0.604
1.53
0.900
1.15
1.90
44.3
13.7
6.40
3.75
0.429
10.1
2.60
0.280
0.0657
0.0734
>200
5.41
0.101
0.0221
0.0161
25.5
tAIP-III D2A
0.257
0.279
0.909
>200
29.9
0.329
0.387
0.509
>200
0.0957
0.0306
0.0363
>200
tAIP-III D2A/F3Y
tAIP-III D2A/F3W
Ac-AIP-III
Ac-(C-A-Y-L-L)
Ac-(C-A-W-L-L)
Ac-I-N-(C-D-F-L-L)
G-I-N-(C-D-F-L-L)
A-I-N-(C-D-F-L-L)
Y-I-N-(C-D-F-L-L)
I-N-(C-A-F-L-L)
Ac-(C-A-F-I-M)
c
G-AIP-III
>200
104
d
d
A-AIP-III
26.1
27.5
28.5
c
Y-AIP-III
8.92
39.2
0.0506
0.260
78.2
e
AIP-III D4A
0.485
3.06
0.0349
0.353
e
tAIP-I D2A
a
See Experimental Section for details of reporter strains and methods. See Supporting Information for plots of antagonism dose response curves. All
b
assays performed in triplicate. IC50 values determined by testing AIPs over a range of concentrations (200 fM to 10 μM). See Figure 5 and
Supporting Information for 95% confidence ranges. Dose response curve did not reach 100% inhibition over the concentrations tested. Inhibition
dose response curve upturned at higher concentrations, potentially indicative of partial agonism (see text). Data included for comparison.
c
d
e
the detection and quantification of AgrC modulators in S.
aureus (Tables 3 and 4; see Figures 4 and S-3−S-6 for a data
overlay).74 The most potent AgrC inhibitors identified in the
gfp-reporter assays were capable of completely inhibiting
hemolysis in the S. aureus strains at nanomolar concentrations
or lower. For example, the global S. aureus QS inhibitor
identified in the gfp-reporter assay, AIP-III D4A, inhibited
hemolysis in all four groups at subnanomolar concentrations
(IC50 value < 0.2 nM). This peptide was 3-fold more active
than the previously reported global inhibitor, tAIP-I D2A, in
the hemolysis assay in the group-IV strain; more strikingly, it
was 18-fold and >40-fold more active in the group-I and -II
strains, respectively (Table 4). These data are significant, as
they indicate that AIP-III D4A and related analogs are capable
7877
dx.doi.org/10.1021/ja3112115 | J. Am. Chem. Soc. 2013, 135, 7869−7882

Journal of the American Chemical Society
Article
Table 6. IC50 Values of Selected Second-Generation AIP-III Analogs for Inhibition of Hemolysis by Group-I−IV S. aureus
a
Strains
b
b
b
b
peptide name
sequence
group-I IC50 (nM)
group-II IC50 (nM)
group-III IC50 (nM)
group-IVIC50 (nM)
AIP-III I1A/D4A
AIP-III N2A/D4A
AIP-III I1A/N2A/D4A
tAIP-III D2A
A-N-(C-A-F-L-L)
I-A-(C-A-F-L-L)
A-A-(C-A-F-L-L)
Ac-(C-A-F-L-L)
Ac-(C-A-Y-L-L)
Ac-(C-A-W-L-L)
0.0103
0.0362
0.0411
0.332
0.793
0.0661
0.0606
0.711
0.204
0.126
0.551
0.216
0.243
0.197
0.265
1.08
0.284
0.122
0.140
0.306
0.134
0.194
tAIP-III D2A/F3Y
tAIP-III D2A/F3W
0.279
0.468
a
See Experimental Section for details of strains and methods. See Supporting Information for plots of antagonism dose response curves. All assays
b
performed in triplicate. IC50 values determined by testing AIPs over a range of concentrations (200 fM to 10 μM). See Figure 5 and Supporting
Information for 95% confidence ranges.
of blocking an important QS phenotype directly linked to
virulence in wild-type S. aureus strains. We later examined their
activity against a second virulence phenotype in S. aureus,
TSST-1 production, which is of particular relevance in the
context of AIP-III (Figure 1A). We return to the results of these
experiments below.
We next examined four truncated AIP-III derivatives lacking
exocyclic tails to explore whether this modification, as shown
for AIP-I, -II and -IV,56 could affect their inhibitory activities
against cognate and noncognate AgrC receptors (Table 5, rows
5−8). Such an analysis of the native AIP-III is yet to be
reported. In agreement with past data for truncated AIP-I and
-IV but in contrast to that for truncated AIP-II,56 we found that
the truncated AIP-III (tAIP-III) was only a weak self-inhibitor.
Moreover, we observed no significant activity change between
tAIP-III and native AIP-III against AgrC-II and -IV and slightly
diminished activity for tAIP-III relative to native AIP-III against
AgrC-I (26.7 vs 5.05 nM, respectively). These data suggest that
the exocyclic tail of AIP-III does not play a major role in cross-
group inhibition. We note, however, that inclusion of the D4A
mutation in the truncated AIP-III produced a potent global
AgrC inhibitor (termed tAIP-III D2A; IC50 value < 1 nM for all
groups; note, residue numbering shifted due to truncation),
with activity within error of its full-length analog (AIP-III D4A)
in AgrC-I, -II, and -IV and only somewhat reduced for group-III
(IC50 value = 0.329 vs 0.0506 nM, respectively). This truncated
AIP-III analog represents a more structurally streamlined,
peptide-based AgrC inhibitor and provides an excellent scaffold
for future structural optimization.
To explore SARs for the lead inhibitor tAIP-III D2A, we
replaced Phe3 with other aromatic amino acids (Tyr or Trp) to
determine the role of this residue in inhibitory activity (Table
5). These two analogs (tAIP-III D2A/F3Y and tAIP-III D2A/
F3W) displayed similar inhibitory activities in the gfp-reporter
assays as the parent analog. These data, along with the alanine
scan data above, suggest that bulky, aromatic residues are
crucial at the Phe3 (or Phe5) position for AgrC recognition by
AIP-III analogs but that the residue identity is insignificant.
We next evaluated four elongated AIP-III analogs with either
an acetyl, Gly, Ala, or Tyr extension at the N-terminus of AIP-
III (Table 5, rows 9−12). Again, elongation of the exocyclic tail
of AIP-III with Tyr was previously reported to yield a potent
AgrC-III antagonist.56 As expected, the tyrosine-extended AIP-
III (Y-AIP-III) was a moderate antagonist of AgrC-III in the
gfp-reporter assay. In contrast, the acetylated and Gly variants
(Ac-AIP-III and G-AIP-III) were largely inactive in AgrC-III,
and the alanine-extended AIP (A-AIP-III) was actually a partial
agonist of AgrC-III instead. When evaluated for inhibitory
activity against the noncognate AgrC receptors, Ac-AIP-III was
the least active (>200 nM for group-I and -IV and 44.3 nM for
group-II), implying that a free N-terminus is important for
recognition. None of the elongated AIP-III analogs showed
improved inhibition compared to the parent AIP-III, and G-
AIP-III was the least potent elongated analog overall.
Interestingly, A-AIP-III, as opposed to Y-AIP-III, displayed
Second-Generation AIP-III Analogs. We next designed
and synthesized a second set of AIP-III analogs to further
explore the SARs delineated above for the first-generation
analogs and to integrate additional results from prior studies of
other synthetic AIP analogs. This set included peptides with
double and triple alanine mutations, alternate aromatic residues
in place of Phe5, and truncated and elongated exocyclic tails
(Table 1). As introduced above, previous reports have
demonstrated that truncated AIP-I, -II and -IV analogs lacking
the exocyclic tail are strong cross-group inhibitors and weak to
strong self-inhibitors of AgrC receptors.56 In addition, an earlier
study revealed that elongation of the exocyclic tail of native
AIP-III with a Tyr residue yielded a potent AgrC-III inhibitor.56
Interestingly, this phenomenon was observed only for AIP-III.
While the origins of this effect in AIP-III are unknown, we
sought to explore such elongated AIP-III analogs further in the
context of the present study. The second set of 12 AIP-III
analogs was generated according to the solid-phase synthesis
methods introduced above and evaluated in the four S. aureus
gfp-reporter strains for inhibitory activity against AgrCs-I−IV.
The results of these assays are shown in Table 5; data for our
lead AgrC inhibitor, AIP-III D4A, and the previously reporter
inhibitor, tAIP-I D2A, are included for comparison.
The analogs with double and triple alanine mutations (Table
5, rows 1−4) were designed to examine whether simulta-
neously replacing multiple amino acid residues with alanine
would result in an additive effect on compound activity and
built on our studies of the first-generation analogs described
above. Replacing both exocyclic residues (Ile1 and Asn2) in
AIP-III with alanine yielded an analog (AIP-III I1A/N2A) with
an IC50 value between those of the parent single alanine
mutants for AgrC-I, -II, and -IV (see Tables 3 and 5). In
contrast, AIP-III I1A/N2A was a much more potent inhibitor
against AgrC-III (IC50 = 2.60 nM) relative to the single alanine
mutants (I1A, IC50 = 194 nM; N2A = weak agonist).
Introducing a D4A mutation along with these exocyclic alanine
mutations (Table 5, rows 2−4) yielded analogs with
antagonistic activities analogous to the AIP-III D4A parent
mutant, regardless of other mutations (within error for all
except AIP-III I1A/D4A against AgrC-II, with a < 5-fold
change), suggesting that the inclusion of D4A may convert
most AIP-III mutants into cross-group AgrC inhibitors.
7878
dx.doi.org/10.1021/ja3112115 | J. Am. Chem. Soc. 2013, 135, 7869−7882

Journal of the American Chemical Society
Article
Figure 5. IC50 values for second-generation AIP-III analogs as determined by the gfp-reporter assay (dark gray) and hemolysis assay (light gray)
against: (A) AgrC-I/group-I; (B) AgrC-II/group-II; (C) AgrC-III/group-III; and (D) AgrC-IV/group-IV. Only calculated values (below the 200 nM
threshold) are presented. In this second generation, only lead compounds (IC50 values < 1 nM) were tested in the hemolysis assay. 95% confidence
intervals indicated. For a complete listing of data, see Tables 5 and 6.
partial agonism in AgrC-IV, even though Tyr is the first amino
acid in this receptor’s native AIP sequence (AIP-IV; Figure 1B).
We evaluated the most potent AgrC inhibitors from this
second set of AIP-III analogs (IC50 values <1 nM) in the S.
aureus hemolysis assay (Table 6). Analogous to the results for
the first-generation analogs, the activity trends for the second-
generation analogs in the hemolysis and gfp-reporter assays are
largely consistent (see Figure 5 and Figures S-3−S-6 in
Supporting Information for a data overlay). However, the
hemolysis assay did unmask some variances between certain
potent analogs. For example, AIP-III I1A/D4A inhibited
hemolysis in group-I at 8-fold lower concentrations than the
original AIP-III D4A analog (0.0103 vs 0.0820 nM,
respectively), a discrepancy that was not observed in the gfp
reporter assays (0.328 vs 0.485 nM, respectively), suggesting
that double mutants may enhance the inhibitory activity of AIP-
III analogs. We also note that AIP-III D-D4 was a moderate
AgrC-IV antagonist in the gfp assay (IC50 value = 29.2 nM) but
instead displayed partial agonism in the hemolysis assay (IC50
value > 200 nM). The cause for the latter discrepancy is unclear
at this time. Nevertheless, the overall congruence of the
hemolysis and gfp-reporter assay data for both the first- and
second-generation AIP analogs in this study provides strong
support for the hemolysis assay as a method to characterize
synthetic AgrC modulators.
Attenuation of TSST-1 Production. As highlighted above,
this study was motivated in part by our interest in AIP signaling
in group-III S. aureus, and the production of TSST-1 is a
hallmark QS phenotype in this group. Thus, we sought to
determine whether the four most potent AgrC inhibitors
identified above (AIP-III D4A, tAIP-III D2A, AIP-III N2A/
D4A, and AIP-III I1A/N2A/D4A) were capable of reducing the
production of TSST-1 in group-III S. aureus. We incubated a
wild-type group-III S. aureus strain (MN8, Table 2) known to
produce TSST-1 in the presence of each inhibitor (at 1 and 10
nM) and quantitated toxin production using a standard
sandwich-type ELISA assay with a commercially available
anti-TSST-1 antibody (see Experimental Section for details).
The results of these assays are shown in Figure 6 and
demonstrate that all four AIP-III analogs are capable of strongly
inhibiting TSST-1 production (by >80%) in this group-III S.
aureus strain at 10 nM. Moreover, two analogs, AIP-III D4A
and tAIP-III D2A, maintained 80% reduction of TSST-1 levels
at 1 nM.
These ELISA data are significant, as we are aware of only one
other report of small molecules capable of inhibiting TSST-1
production in S. aureusthe naturally occurring cyclic
7879
dx.doi.org/10.1021/ja3112115 | J. Am. Chem. Soc. 2013, 135, 7869−7882

Journal of the American Chemical Society
Article
current study, we report the design and synthesis of a series of
30 non-native macrocyclic peptides based on the AIP signal
used by group-III S. aureus for QS, AIP-III. We evaluated the
activities of these peptides against the group I−IV AgrC
receptors using both fluorescence-based reporter assays and QS
phenotypic assays in wild-type S. aureus strains. This work
revealed a series of new and highly potent, pan-group AgrC
inhibitors and represents the first systematic analysis of the
SARs defining AIP-III activity in the group I−IV AgrC
receptors.
There are several important outcomes of this study. First,
SAR analyses of the AIP-III analogs revealed a series of
interesting activity trends for AgrC receptor modulation across
the four S. aureus groups. Our findings both corroborate and
extend previous observations regarding the role of the AIP
exocyclic residues in AgrC receptor activation as well as the
importance of the endocyclic hydrophobic residues for AgrC
receptor recognition.29,52,56 We also demonstrate that, while
combining certain mutations in the AIP-III structure led only to
modest additive inhibitory effects, inclusion of the D4A
mutation consistently converts most AIP-III sequences into
potent and global AgrC inhibitors. Specifically, the D4A
mutation had this effect in the native AIP-III signal, AIP-III
analogs with multiple alanine mutations, truncated AIP-III
lacking the exocyclic tail, and truncated sequences with
hydrophobic residue mutations. Second, we introduce the use
of hemolysis assays as a straightforward, rapid-throughput
method to assess a QS phenotype in wild-type S. aureus linked
to virulence. The similar activity trends observed for the AIP-III
analogs in this hemolysis assay compared to those in the gfp-
reporter assay serve to validate the utility of hemolysis assays
for the study of QS modulators in S. aureus.
Third, we extend our studies to examine, for the first time,
the effects of AIP analogs on attenuating TSST-1 production in
S. aureus, a QS phenotype that is particularly pertinent in
group-III S. aureus infections. Lastly, and perhaps most notably,
this study uncovered several global inhibitors of the S. aureus
agr QS circuit that are active at picomolar concentrations and
represent, to our knowledge, the most potent peptide-based
AgrC inhibitors reported to date. These results suggest that,
relative to the other AIPs, AIP-III provides a superior scaffold
for the development of peptide-based AgrC inhibitors. For
example, AIP-III D4A surpassed the activity of the known
inhibitor, tAIP-I D2A, by 5- to >20-fold in the gfp-reporter
assays and by 3- to >40-fold in the hemolysis assays in all four
S. aureus groups and blocked TSST-1 production in group-III S.
aureus by over 80% at 1 nM. As such, AIP-III D4A represents a
powerful new chemical tool for the study of QS in S. aureus in a
range of fundamental and applied contexts; its activity in all
four groups of S. aureus serves to significantly expand its utility.
Preliminary assays indicate that the lead AIP-III analogs elicit
their inhibitory activity via a competitive mechanism by
blocking native AIP-III:AgrC-III binding. Additional biochem-
ical and structural studies are required to characterize their
mechanisms of action and will serve to illuminate new avenues
for further optimization of their structures to improve their
potency and stability. Such studies are ongoing in our
laboratory, along with the examination of methods and
materials for the controlled release of these agents into
environments of relevance for infection control and will be
reported in due course.
Figure 6. Attenuation of TSST-1 production in wild-type group-III S.
aureus (MN8) using non-native peptides. Bacteria were either
untreated (black bar) or treated with select peptides at 1 nM (light
gray bars), 10 nM (white bars), or 100 μM (dark gray bars). TSST-1
concentrations presented as percentage of the TSST-1 concentration
in untreated cultures. ** indicates control compound. See text.
dipeptides (cyclo-(Tyr-Pro) and cyclo-(Phe-Pro)) reported by
McCormick and co-workers.34 Bacterial supernatants contain-
ing these compounds were shown to reduce TSST-1
production levels by ∼80% using an analogous ELISA. The
purified cyclic dipeptides, however, were not tested in the
ELISA. As we observed these compounds to be largely inactive
in the gfp-reporter assays (over the concentration range tested;
see above), we reasoned that the pure compounds would be
weak inhibitors of TSST-1 production, if at all. We subjected
pure samples of cyclo-(Tyr-Pro) and cyclo-(Phe-Pro) to the
ELISA protocol and found that 100 μM concentrations were
required to reduce TSST-1 production by ∼50% or ∼15%,
respectively (see Figure 6). These data indicate that our AIP-III
analogs are at least 1000-fold more potent than the cyclic
dipeptides in the ELISA. For an additional comparison, we also
evaluated Muir and co-workers’ tAIP-I D2A derivative in TSST-
1 ELISA. In contrast to the cyclic dipeptides, tAIP-I D2A is a
strong inhibitor of the AgrC-III receptor (albeit weaker than
the AIP-III analogs; see above). As expected, tAIP-I D2A was
capable of inhibiting TSST-1 production by ∼80% at 10 nM,
yet was less active than AIP-III D4A and tAIP-III D2A at 1 nM
(∼65% vs ∼80%; Figure 6). These AIP-III analogs therefore
represent, to our knowledge, the most potent inhibitors of
TSST-1 production in S. aureus to be reported. Collectively, the
results of these ELISA assays, in concert with the hemolysis
assay data described above, provide strong support for the use
of non-native AIP-III analogs to attenuate clinically relevant QS
phenotypes in S. aureus.
SUMMARY AND CONCLUSIONS
■
S. aureus utilizes QS to control myriad phenotypes linked to
virulence in human infections. This often lethal pathogen uses
AIP signals and their corresponding membrane-associated
AgrC receptors to mediate QS. Non-native ligands that
attenuate AIP signaling in S. aureus could be applied as
chemical tools to study the role and timing of QS in infection
processes. To date, such probe molecules remain limited. In the
7880
dx.doi.org/10.1021/ja3112115 | J. Am. Chem. Soc. 2013, 135, 7869−7882

Journal of the American Chemical Society
Article
(21) Chambers, H. F.; Deleo, F. R. Nat. Rev. Microbiol. 2009, 7, 629−
ASSOCIATED CONTENT
■
641.
S
* Supporting Information
(22) Deleo, F. R.; Chambers, H. F. J. Clin. Invest. 2009, 119, 2464−
2474.
Full details of peptide synthesis and characterization, additional
bacteriological methods, dose response curves for AIP-III
analogs, and overlays of the fluorescence and hemolysis assay
data. This material is available free of charge via the Internet at
http://pubs.acs.org.
(23) Recsei, P.; Kreiswirth, B.; O’Reilly, M.; Schlievert, P.; Gruss, A.;
Novick, R. P. Mol. Gen. Genet. 1986, 202, 58−61.
(24) Pragman, A. A.; Schlievert, P. M. FEMS Immunol. Med.
Microbiol. 2004, 42, 147−154.
(25) George, E. A.; Muir, T. W. ChemBioChem 2007, 8, 847−855.
(26) Lina, G.; Jarraud, S.; Ji, G. Y.; Greenland, T.; Pedraza, A.;
Etienne, J.; Novick, R. P.; Vandenesch, F. Mol. Microbiol. 1998, 28,
655−662.
(27) Novick, R. P.; Ross, H. F.; Projan, S. J.; Kornblum, J.;
Kreiswirth, B.; Moghazeh, S. EMBO J. 1993, 12, 3967−3975.
(28) Ji, G.; Beavis, R.; Novick, R. P. Science 1997, 276, 2027−2030.
(29) McDowell, P.; Affas, Z.; Reynolds, C.; Holden, M. T.; Wood, S.
J.; Saint, S.; Cockayne, A.; Hill, P. J.; Dodd, C. E.; Bycroft, B. W.;
Chan, W. C.; Williams, P. Mol. Microbiol. 2001, 41, 503−512.
(30) Jarraud, S.; Lyon, G. J.; Figueiredo, A. M.; Gerard, L.;
Vandenesch, F.; Etienne, J.; Muir, T. W.; Novick, R. P. J. Bacteriol.
2000, 182, 6517−6522.
(31) Jarraud, S.; Mougel, C.; Thioulouse, J.; Lina, G.; Meugnier, H.;
Forey, F.; Nesme, X.; Etienne, J.; Vandenesch, F. Infect. Immun. 2002,
70, 631−641.
(32) Holtfreter, S.; Grumann, D.; Schmudde, M.; Nguyen, H. T. T.;
Eichler, P.; Strommenger, B.; Kopron, K.; Kolata, J.; Giedrys-Kalemba,
S.; Steinmetz, I.; Witte, W.; Broker, B. M. J. Clin. Microbiol. 2007, 45,
2669−2680.
(33) Limbago, B.; Fosheim, G. E.; Schoonover, V.; Crane, C. E.;
Nadle, J.; Petit, S.; Heltzel, D.; Ray, S. M.; Harrison, L. H.; Lynfield,
R.; Dumyati, G.; Townes, J. M.; Schaffner, W.; Mu, Y.; Fridkin, S. K. J.
Clin. Microbiol. 2009, 47, 1344−1351.
AUTHOR INFORMATION
■
Corresponding Author
blackwell@chem.wisc.edu
Author Contributions
§These authors contributed equally.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by the Office of Naval Research
(N00014-07-1-0255), Burroughs Welcome Fund, and Kim-
berly-Clark Corporation. D.M.S. was supported in part by a
National Science Foundation Graduate Research Fellowship
(DGE-0718123). We thank Professors Alexander Horswill and
Richard Novick for providing S. aureus strains and Professor
Tom Muir for valuable discussions.
REFERENCES
■
(1) Ng, W. L.; Bassler, B. L. Annu. Rev. Genet. 2009, 43, 197−222.
(2) Bassler, B. L.; Losick, R. Cell 2006, 125, 237−246.
(3) Camilli, A.; Bassler, B. L. Science 2006, 311, 1113−1116.
(4) Eberhard, A.; Burlingame, A. L.; Eberhard, C.; Kenyon, G. L.;
Nealson, K. H.; Oppenheimer, N. J. Biochemistry 1981, 20, 2444−
2449.
(34) Li, J. R.; Wang, W. L.; Xu, S. X.; Magarvey, N. A.; McCormick, J.
K. Proc. Natl. Acad. Sci. U.S.A. 2011, 108, 3360−3365.
(35) Musser, J. M.; Schlievert, P. M.; Chow, A. W.; Ewan, P.;
Kreiswirth, B. N.; Rosdahl, V. T.; Naidu, A. S.; Witte, W.; Selander, R.
K. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 225−229.
(36) Galloway, W. R. J. D.; Hodgkinson, J. T.; Bowden, S. D.; Welch,
M.; Spring, D. R. Chem. Rev. 2011, 111, 28−67.
(37) Geske, G. D.; O’Neill, J. C.; Blackwell, H. E. Chem. Soc. Rev.
2008, 37, 1432−1447.
(5) Marketon, M. M.; Gronquist, M. R.; Eberhard, A.; Gonzalez, J. E.
J. Bacteriol. 2002, 184, 5686−5695.
(6) Joint, I.; Downie, J. A.; Williams, P. Philos. Trans. R. Soc., B 2007,
362, 1115−1117.
(38) Persson, T.; Givskov, M.; Nielsen, J. Curr. Med. Chem. 2005, 12,
3103−3115.
(7) von Bodman, S. B.; Willey, J. M.; Diggle, S. P. J. Bacteriol. 2008,
190, 4377−4391.
(39) Stevens, A. M.; Queneau, Y.; Soulere, L.; von Bodman, S.;
Doutheau, A. Chem. Rev. 2011, 111, 4−27.
(8) Boyer, M.; Wisniewski-Dye, F. FEMS Microbiol. Ecol. 2009, 70,
1−19.
(40) Amara, N.; Krom, B. P.; Kaufmann, G. F.; Meijler, M. M. Chem.
Rev. 2011, 111, 195−208.
(9) De Kievit, T. R.; Iglewski, B. H. Infect. Immun. 2000, 68, 4839−
4849.
(41) Tavassoli, A.; Hamilton, A. D.; Spring, D. R. Chem. Soc. Rev.
2011, 40, 4269−4270.
(10) Bjarnsholt, T.; Givskov, M. Anal. Bioanal. Chem. 2007, 387,
409−414.
(42) O’ Connor, C. J.; Laraia, L.; Spring, D. R. Chem. Soc. Rev. 2011,
40, 4332−4345.
(11) Hibbing, M. E.; Fuqua, C.; Parsek, M. R.; Peterson, S. B. Nat.
Rev. Microbiol. 2010, 8, 15−25.
(43) Kaufmann, G. F.; Park, J.; Janda, K. D. Expert Opin. Biol. Ther.
2008, 8, 719−724.
(12) Rasko, D. A.; Sperandio, V. Nat. Rev. Drug Discovery 2010, 9,
117−128.
(44) Geske, G. D.; O’Neill, J. C.; Miller, D. M.; Mattmann, M. E.;
Blackwell, H. E. J. Am. Chem. Soc. 2007, 129, 13613−13625.
(45) Geske, G. D.; ONeill, J. C.; Miller, D. M.; Wezeman, R. J.;
Mattmann, M. E.; Lin, Q.; Blackwell, H. E. ChemBioChem 2008, 9,
389−400.
(13) Njoroge, J.; Sperandio, V. EMBO Mol. Med. 2009, 1, 201−210.
(14) Sintim, H. O.; Smith, J. A.; Wang, J. X.; Nakayama, S.; Yan, L.
Future Med. Chem. 2010, 2, 1005−1035.
(15) Wright, G. D. Curr. Opin. Chem. Biol. 2003, 7, 563−569.
(16) Walsh, C. Nature 2000, 406, 775−781.
(17) Clatworthy, A. E.; Pierson, E.; Hung, D. T. Nat. Chem. Biol.
2007, 3, 541−548.
(46) Mattmann, M. E.; Blackwell, H. E. J. Org. Chem. 2010, 75,
6737−6746.
(47) Palmer, A. G.; Streng, E.; Blackwell, H. E. ACS Chem. Biol. 2011,
6, 1348−1356.
(18) Barczak, A. K.; Hung, D. T. Curr. Opin. Microbiol. 2009, 12,
490−496.
(19) For an excellent argument against the development of resistance
to QS antagonists, see: Swem, L. R.; Swem, D. L.; O’Loughlin, C. T.;
Gatmaitan, R.; Zhao, B.; Ulrich, S. M.; Bassler, B. L. Mol. Cell 2009, 35,
143−153.
(20) Rotun, S. S.; McMath, V.; Schoonmaker, D. J.; Maupin, P. S.;
Tenover, F. C.; Hill, B. C.; Ackman, D. M. Emerging Infect. Dis. 1999,
5, 147−149.
(48) Stacy, D. M.; Welsh, M. A.; Rather, P. N.; Blackwell, H. E. ACS
Chem. Biol. 2012, 7, 1719−1728.
(49) Gordon, C. P.; Williams, P.; Chan, W. C. J. Med. Chem. 2013,
56, 1389−1404.
(50) Park, J.; Jagasia, R.; Kaufmann, G. F.; Mathison, J. C.; Ruiz, D. I.;
Moss, J. A.; Meijler, M. M.; Ulevitch, R. J.; Janda, K. D. Chem. Biol.
2007, 14, 1119−1127.
7881
dx.doi.org/10.1021/ja3112115 | J. Am. Chem. Soc. 2013, 135, 7869−7882

Journal of the American Chemical Society
Article
(51) Kirchdoerfer, R. N.; Garner, A. L.; Flack, C. E.; Mee, J. M.;
Horswill, A. R.; Janda, K. D.; Kaufmann, G. F.; Wilson, I. A. J. Biol.
Chem. 2011, 286, 17351−17358.
(52) Mayville, P.; Ji, G.; Beavis, R.; Yang, H.; Goger, M.; Novick, R.
P.; Muir, T. W. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 1218−1223.
(53) Lyon, G. J.; Mayville, P.; Muir, T. W.; Novick, R. P. Proc. Natl.
Acad. Sci. U.S.A. 2000, 97, 13330−13335.
(54) Wright, J. S.; Jin, R.; Novick, R. P. Proc. Natl. Acad. Sci. U.S.A.
2005, 102, 1691−1696.
(55) Fleming, V.; Feil, E.; Sewell, A. K.; Day, N.; Buckling, A.;
Massey, R. C. J. Bacteriol. 2006, 188, 7686−7688.
(56) Lyon, G. J.; Wright, J. S.; Muir, T. W.; Novick, R. P. Biochemistry
2002, 41, 10095−10104.
(57) Scott, R. J.; Lian, L. Y.; Muharram, S. H.; Cockayne, A.; Wood,
S. J.; Bycroft, B. W.; Williams, P.; Chan, W. C. Bioorg. Med. Chem. Lett.
2003, 13, 2449−2453.
(58) George, E. A.; Novick, R. P.; Muir, T. W. J. Am. Chem. Soc.
2008, 130, 4914−4924.
(59) Fowler, S. A.; Stacy, D. M.; Blackwell, H. E. Org. Lett. 2008, 10,
2329−2332.
(60) Chan, W. C.; Coyle, B. J.; Williams, P. J. Med. Chem. 2004, 47,
4633−4641.
(61) Lyon, G. J.; Novick, R. P. Peptides 2004, 25, 1389−1403.
(62) Thoendel, M.; Kavanaugh, J. S.; Flack, C. E.; Horswill, A. R.
Chem. Rev. 2011, 111, 117−151.
(63) Gorske, B. C.; Blackwell, H. E. Org. Biomol. Chem. 2006, 4,
1441−1445.
(64) Campbell, J.; Lin, Q.; Geske, G. D.; Blackwell, H. E. ACS Chem.
Biol. 2009, 4, 1051−1059.
(65) Malone, C. L.; Boles, B. R.; Horswill, A. R. Appl. Environ.
Microbiol. 2007, 73, 6036−6044.
(66) Schlievert, P. M.; Blomster, D. A. J. Infect. Dis. 1983, 147, 236−
242.
(67) Otto, M.; Echner, H.; Voelter, W.; Gotz, F. Infect. Immun. 2001,
69, 1957−1960.
(68) Sewing, A.; Hilvert, D. Angew. Chem., Int. Ed. 2001, 40, 3395−
3396.
(69) Swinnen, D.; Hilvert, D. Org. Lett. 2000, 2, 2439−2442.
(70) Dawson, P. E.; Muir, T. W.; Clark-Lewis, I.; Kent, S. B. H.
Science 1994, 266, 776−779.
(71) Lyon, G. J.; Wright, J. S.; Christopoulos, A.; Novick, R. P.; Muir,
T. W. J. Biol. Chem. 2002, 277, 6247−6253.
(72) Blevins, J. S.; Beenken, K. E.; Elasri, M. O.; Hurlburt, B. K.;
Smeltzer, M. S. Infect. Immun. 2002, 70, 470−480.
(73) Lopes, M. D.; Teixeira, L. M. Rev. Microbiol. 1992, 23, 59−65.
(74) We note that the two sets of assay data came from different
bacterial strains (Table 2), and thus strain-to-strain variations could
account for the slight deviations observed.
7882
dx.doi.org/10.1021/ja3112115 | J. Am. Chem. Soc. 2013, 135, 7869−7882