Thiolactone modulators of quorum sensing revealed through library design and screening

Article abstract of DOI:10.1016/j.bmc.2011.06.071

Quorum sensing (QS) is a process by which bacteria use small molecules or peptidic signals to assess their local population densities. At sufficiently high density, bacteria can alter gene expression levels to regulate group behaviors involved in a range of important and diverse phenotypes, including virulence factor production, biofilm formation, root nodulation, and bioluminescence. Gram-negative bacteria most commonly use N-acylated l-homoserine lactones (AHLs) as their QS signals. The AHL lactone ring is hydrolyzed relatively rapidly at biological pH, and the ring-opened product is QS inactive. We seek to identify AHL analogues with heightened hydrolytic stability, and thereby potentially heightened activity, for use as non-native modulators of bacterial QS. As part of this effort, we probed the utility of thiolactone analogues in the current study as QS agonists and antagonists in Gram-negative bacteria. A focused library of thiolactone analogs was designed and rapidly synthesized in solution. We examined the activity of the library as agonists and antagonists of LuxR-type QS receptors in Pseudomonas aeruginosa (LasR), Vibrio fischeri (LuxR), and Agrobacterium tumefaciens (TraR) using bacterial reporter strains. The thiolactone library contained several highly active compounds, including some of the most active LuxR inhibitors and the most active synthetic TraR agonist reported to date. Analysis of a representative thiolactone analog revealed that its hydrolysis half-life was almost double that of its parent AHL in bacterial growth medium.

Full text of DOI:10.1016/j.bmc.2011.06.071

Bioorganic & Medicinal Chemistry 19 (2011) 4820–4828
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Thiolactone modulators of quorum sensing revealed through library
design and screening
⇑
Christine E. McInnis, Helen E. Blackwell
Department of Chemistry, University of Wisconsin–Madison, 1101 University Avenue, Madison, WI 53706, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Quorum sensing (QS) is a process by which bacteria use small molecules or peptidic signals to assess their
local population densities. At sufficiently high density, bacteria can alter gene expression levels to regu-
late group behaviors involved in a range of important and diverse phenotypes, including virulence factor
production, biofilm formation, root nodulation, and bioluminescence. Gram-negative bacteria most com-
Received 20 May 2011
Revised 22 June 2011
Accepted 26 June 2011
Available online 1 July 2011
monly use N-acylated L-homoserine lactones (AHLs) as their QS signals. The AHL lactone ring is hydro-
lyzed relatively rapidly at biological pH, and the ring-opened product is QS inactive. We seek to
identify AHL analogues with heightened hydrolytic stability, and thereby potentially heightened activity,
for use as non-native modulators of bacterial QS. As part of this effort, we probed the utility of thiolactone
analogues in the current study as QS agonists and antagonists in Gram-negative bacteria. A focused
library of thiolactone analogs was designed and rapidly synthesized in solution. We examined the activity
of the library as agonists and antagonists of LuxR-type QS receptors in Pseudomonas aeruginosa (LasR),
Vibrio fischeri (LuxR), and Agrobacterium tumefaciens (TraR) using bacterial reporter strains. The thiolac-
tone library contained several highly active compounds, including some of the most active LuxR inhibi-
tors and the most active synthetic TraR agonist reported to date. Analysis of a representative thiolactone
analog revealed that its hydrolysis half-life was almost double that of its parent AHL in bacterial growth
medium.
Keywords:
N-Acylated L-homoserine lactones
Gram-negative bacteria
LuxR-type receptors
Pseudomonas aeruginosa
Quorum sensing
Thiolactone AHL analogs
Ó 2011 Elsevier Ltd. All rights reserved.
9
–12
1
. Introduction
decade, the use of abiotic small molecules
probes
and macromolecular
1
3,14
to attenuate QS pathways has emerged as a prominent
While humans and other higher order organisms can use a
research strategy in the bacterial communication field. Most of
these agents work by intercepting or inactivating the native QS
signal.
combination of sound, scent, movement, and sight to communi-
cate, bacteria use a language of chemical signals to express their
messages. This chemical communication process is termed
Gram-negative bacteria use N-acylated
L-homoserine lactones
1
4,11,15
quorum sensing (QS), and allows bacteria to assess their local
(AHLs) as their primary autoinducers for QS (Fig. 1A).
These
population densities using a class of signaling molecules called
low molecular weight, cell permeable signals are produced by
AHL synthases (LuxI-type proteins) and are sensed by their cognate
cytoplasmic receptors (LuxR-type proteins). The LuxR-type
proteins are AHL-regulated transcription factors. Productive
LuxR-type receptor binding occurs once a sufficiently high AHL
concentration is achieved in the cell. The AHL:LuxR-type protein
complex most commonly dimerizes, binds DNA, and activates the
2
–4
autoinducers.
In general, autoinducer concentration increases
with bacterial cell density. Once a threshold population density is
5
achieved in a given environment, bacteria can alter gene expres-
sion levels in order to initiate processes that are largely only pos-
sible as a multicellular community. Such processes can play
critical roles in both mutualistic symbiosis and the pathogenesis
6
16
of bacterial infections. Phenotypes under the control of QS include
transcription of QS controlled genes. Over 100 Gram-negative
biofilm formation, antibiotic production, bioluminescence, root
nodulation, swarming, and virulence factor production. As several
of the most prevalent human pathogens use QS to control viru-
bacteria use LuxI/LuxR-type systems to control QS, and many spe-
cies have multiple LuxI/LuxR-type pairs to regulate different and
1
7
overlapping aspects of their QS regulon. The known, naturally
occurring AHL signals all have a conserved -HL head group and
7
lence, there is significant interest in delineating the mechanisms
L
8
4
of QS and targeting QS as an anti-infective strategy. Over the past
simply vary in their acyl tail. The selectivity of an AHL for its cog-
nate LuxR-type receptor is therefore controlled by acyl chain struc-
ture. These tails are commonly derived from fatty acids and can
have between 4 and 18 carbons and different levels of oxidation
⇑
Corresponding author. Tel.: +1 608 262 1503; fax: +1 608 265 4534.
E-mail address: blackwell@chem.wisc.edu (H.E. Blackwell).
1
8
at the 3-position. Representative naturally occurring AHLs are
0
968-0896/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2011.06.071

C. E. McInnis, H. E. Blackwell / Bioorg. Med. Chem. 19 (2011) 4820–4828
4821
A
B
C
O
O
O
O
O
O
X
S
O
N
N
n
N
n
H
H
H
O
O
O
OHHL: n = 1
OOHL: n = 3
OdDHL: n = 7
5
: X = H
6: X = 3-NO2
: X = 4-OMe
8: X = 4-Br
9: X = 4-Ph
11: n = 1; DL
12: n = 1; L
3: n = 3; DL
4: n = 7; DL
5: n = 7; L
1
1
1
7
O
O
N
n
H
O
O
R
O
BHL: n = 1
O
O
S
N
H
N
H
1
2
3
4
: n = 3
: n = 4
: n = 5
: n = 9
NH
O
O
10
Cl
1
1
6: R = H
7: R = Me
Figure 1. (A) Selected naturally occurring AHLs. OHHL and OOHL are the native AHL signals for V. fischeri and A. tumefaciens, respectively. OdDHL and BHL are native AHLs
used by P. aeruginosa. (B) Representative AHLs with non-native acyl groups studied in our laboratory. (C) Previously reported thiolactone AHL analogues.
shown in Figure 1A, including N-(3-oxo)-hexanoyl HL (OHHL) from
the marine symbiont Vibrio fischeri, N-(3-oxo)-octanoyl HL (OOHL)
from the plant pathogen Agrobacterium tumefaciens, and N-(3-oxo)-
dodecanoyl HL (OdDHL) and N-butanoyl HL (BHL) from the oppor-
tunistic pathogen Pseudomonas aeruginosa.
ing SdiA at single nanomolar concentrations in their bacterial
reporter strain. Most recently, Bassler and co-workers have
reported a set of thiolactone AHL mimics (e.g., 16 and 17) that
are highly active antagonists of CviR in C. violaceum.3
2,33
In view
of these past studies, we reasoned that chimeric ligands made by
uniting -homocysteine thiolactone with our lead AHL acyl groups
Our research laboratory has focused on the design and synthesis
of non-native AHLs that differ in their acyl chain structures for use as
L
could yield LuxR-type protein modulators with heightened or
novel activity profiles.
9
,19
probes to study LuxI/LuxR-type QS.
5–10) are shown in Figure 1B. In our initial studies, we sought to
determine the structural features of non-native acyl chains that
Several of these AHLs
(
Here, we report the design and parallel synthesis of a focused
library of thiolactone analogs based on native and non-native AHLs
studied in our laboratory. The library was examined for agonistic
and antagonistic activity in the well-characterized QS receptors
LuxR, LasR, and TraR using bacterial reporter strains. For a subset
engendered LuxR-type protein selectivity,2
0–22
and we uncovered
several highly potent and selective modulators of LuxR in V. fischeri,
LasR in P. aeruginosa, and TraR in A. tumefaciens. In more recent work,
we have evaluated the activity of these compounds in an expanded
of these compounds, both the racemic and enantiopure (L) thiolac-
23
set of receptors, including QscR in P. aeruginosa, CviR in Chromo-
tone were evaluated. Several highly potent and receptor selective
thiolactone agonists and antagonists were found for these systems,
further underscoring the potential utility of thiolactone derivatives
as chemical probes to study QS. This study represents the first
comparative analysis of thiolactone analogues across different
LuxR-type receptors. In addition, the results of this study allowed
us to formulate a new hypothesis with regards to how several of
the thiolactone agonists elicit their activity in LuxR-type receptors.
An analysis of the hydrolysis rates for a representative thiolactone
and its AHL analog concludes this study, and suggests that thiolac-
tones can exhibit heightened stability relative to AHLs in standard
bacterial growth media.
2
4
bacterium violaceum, and ExpR1/ExpR2 in Pectobacterium caroto-
2
5
vora. Overall, these past studies have revealed that even subtle
changes to a non-native AHL acyl chain can impart dramatic changes
in ligand activity, ranging from converting a poor agonist into a
strong and broad spectrum agonist, to converting an agonist into a
receptor-selective antagonist.22
An important goal of our current research is to improve the po-
tency of our lead AHL agonists and antagonists. One approach to
address this objective is to increase or tune the hydrolytic stability
of the lactone ring, as an intact lactone ring is required for activ-
2
6,27
ity.
replacing the lactone head group in several of our AHLs with a thio-
lactone (i.e., derived from -homocysteine) on LuxR-type receptor
In the work reported herein, we examined the effects of
L
agonism and antagonism. Previous work has demonstrated that se-
lected thiolactone analogues of native AHLs (Fig. 1C) can behave as
agonists or antagonists of LuxR-type proteins. For example,
Passador et al. reported that thiolactone 15, the analogue of LasR’s
native ligand, OdDHL (Fig. 1A), had comparable agonistic activity
to OdDHL in LasR. Thiolactone 15 was also analyzed in LuxR by
Schaefer et al., who found that both 15 and OdDHL were unable
to agonize LuxR in an E. coli reporter, while 12, the thiolactone
analogue of LuxR’s native ligand, OHHL (Fig. 1A), was a weak LuxR
agonist. A later study by Chhabra et al. directed at the use of AHL
analogues as possible immune modulators revealed thiolactone
2. Results and discussion
2.1. Library design and synthesis
We designed a 21-member library of thiolactone AHL ana-
logues, with all of the compounds containing a conserved thiolac-
tone head group and a variable acyl tail (Fig. 2C). A brief
description of our design process and compound synthesis is pro-
vided here. Thiolactones 12 and 15, previously shown to be ago-
nists in LuxR and LasR (see above) and analogs of OHHL and
OdDHL, respectively, were included in the library to serve as con-
2
8
2
8,29
1
5 to cause approximately 40 times less of an immune response
trol compounds.
To complement these two controls, we also
in mice than the native OdDHL, suggesting that thiolactone-
derived QS agonists and antagonists could be useful in clinical set-
tings.3 Janssens et al. have studied the effects of non-native AHLs
on the (orphan) LuxR homolog from Salmonella enterica, SdiA, and
included the thiolactone analog (18) of TraR’s native ligand, OOHL,
along with five other thiolactone analogs (19–23) of native AHLs
used by a range of bacteria, including P. aeruginosa (BHL, 19),
P. chlororaphis (C6-HL, 20), Pantoea ananatis (C6-HL, 20; C7-HL,
21; C8-HL, 22) Burkholderia cepacia (C8-HL, 22), and Acidithiobacil-
0
uncovered several thiolactone-derived agonists (e.g., racemic 11,
3, and 14, Fig. 1C).3 Thiolactone 11 was markedly potent, activat-
1
34
1
lus ferrooxidans (C12-HL, 23). The remaining thiolactone library

4
822
C. E. McInnis, H. E. Blackwell / Bioorg. Med. Chem. 19 (2011) 4820–4828
A
O
O
EDC, TEA
Cl-+H N
3
S
+
S
R
N
H
HO
R
CH Cl , rt, 14 h
2
2
O
O
B
O
O
O
Cl-+H3N
TEA
S
n
O
DMAP
CH Cl
O
O
O
O
O
O
2
2
S
O
O
+
N
H
Cl
O
n
rt, 12 h
CH CN, rt, 2 h
3
O
then reflux, 5 h
n = 1, 3, 7
C
O
O
O
MeO
O
S
S
S
S
N
H
O
N
H
S
S
4
N
H
O
O
O
1
2
21
2
2
8: DL
9: L
O
O
O
Br
O
S
N
H
O
N
H
7
5
9
N
H
1
1
4: DL
5: L
22
O
30: DL
3
1: L
O
O
O
S
N
H
N
H
O
O
3
O
O
S
N
H
1
1
8
23
O
3
3
2: DL
3: L
O
O
S
S
N
H
O
N
H
O
S
9
N
2
2
4: DL
5: L
H
NH
O
3
3
4: DL
5: L
O
O
S
S
N
H
O
O N
N
H
2
3
26: DL
7: L
O
2
0
2
Figure 2. Thiolactone library. (A) Synthesis of non 3-oxo thiolactones. EDC = 1-ethyl-3-(3-dimethyl aminopropyl) carbodiimide. TEA = triethylamine. (B) Synthesis of 3-oxo
thiolactones. DMAP = dimethyl amino pyridine. (C) Library of thiolactones analyzed in this study.
members were chimeric ligands based on acyl groups that we have
the requisite alkyl acid chloride to afford the Meldrum’s acid deriv-
ative, which was then coupled to -homocysteine thiolactone
(Fig. 2B). Racemic thiolactones were made in similar manner from
DL-homocysteine thiolactone (see Section 4).
previously identified in AHL-based LuxR-type receptor agonists
L
and antagonists (Fig. 1B).2
0–22,35
Thiolactones 30/31 and 32/33
were modeled after AHLs 8 and 9, which are strong antagonists
of both LuxR and TraR. Likewise, thiolactones 28/29 were based
on AHL 7, which is a moderate antagonist of LuxR. Phenylacetanoyl
HL 5 was previously shown to be largely inactive in many LuxR-
2.2. Library assay design
2
1
type receptors, and we therefore included thiolactone analogs
4/25 of AHL 5 to test whether this inactivity profile would be
Small molecules are usually screened for LuxR-type agonism or
antagonism using a bacterial strain containing a reporter gene for a
2
9
maintained in thiolactones. The 3-nitro phenylacetanoyl thiolac-
given LuxR-type protein. These strains typically lack a functional
tones 26/27 were based on AHL 6, which is an extremely strong
LuxI-type synthase, yet retain the functional LuxR-type receptor.
Exogenous native AHL therefore must be added to activate the
LuxR system. These strains provide a straightforward way to exam-
ine the agonistic and antagonistic activities of non-native ligands
(by adding only the compound of interest or the compound in
competition with the native AHL ligand (at its EC50 value),
respectively).
2
0
21,22
LuxR agonist and a moderate LasR antagonist.
importance of stereochemistry on ligand activity, thiolactones 14,
5, and 24–35 were synthesized in both racemic (DL) and enantio-
pure ( ) form. The -thiolactone enantiomer was chosen based on
several previous studies that have shown that the active enantio-
To assess the
1
L
L
1,27
mer of native AHL signals is the
L
-form.2
Such an analysis of
the stereochemical requirements for thiolactone modulation for
LuxR-type proteins is yet to be reported.
We utilized four bacterial reporter strains in this study to exam-
ine the LuxR-type modulatory activities of the thiolactone library
in LasR, LuxR, and TraR. Two strains were selected for the LasR
The thiolactone derivatives that lacked 3-oxo functionality were
3
6
synthesized by routine EDC couplings between
L-homocysteine
screens: Escherichia coli DH5 (pJN105L + pSC11) and P. aeruginosa
3
7
thiolactone and various carboxylic acids. (Fig. 2A). The remainder
of the library was synthesized by reacting Meldrum’s acid with
PA01 MW1 (pUM15). E. coli DH5 (pJN105L + pSC11) is a heterol-
ogous reporter strain containing one plasmid for the LasR gene and

C. E. McInnis, H. E. Blackwell / Bioorg. Med. Chem. 19 (2011) 4820–4828
4823
a second plasmid containing the promoter region for LasI fused to
b-galactosidase (b-gal). LasR activity is read-out using a standard
colorimetric assay with ortho-nitrophenyl-b-galactosidase (ONPG)
as the substrate for b-gal. The PA01 MW1 (pUM15) strain is a LasR
reporter in P. aeruginosa that lacks a functional LasI and contains a
plasmid with a LasR responsive promoter for Yellow Fluorescent
Protein (YFP), which facilitates straightforward evaluation of LasR
activity using fluorescence. Examining the thiolactone library in
both of these strains allowed us to study the effects of these com-
pounds on LasR in an isolated system (E. coli) and then in the pres-
ence of P. aeruginosa’s more complex QS network (including RhlR
and QscR, in PAO1). (We note that E. coli and P. aeruginosa have dif-
ferent compound uptake/efflux profiles, and this feature should be
taken into account when comparing small molecule screening data
between the two strains (see below)).
shown in Table 1. Several intriguing trends in ligand activity are
immediately apparent upon analysis of these data. First, all of the
thiolactones exhibited slight to moderate agonistic activities in
the antagonism assay in the P. aeruginosa LasR reporter strain.
Natural AHL thiolactone analogues 12, 14/15, 19, and 20 and
phenylacetanoyl thiolactones 24/25, 29, and 32–34 were capable
of agonizing LasR by greater than 20% in the antagonism assay,
yet only the natural AHL analogs (12, 14/15, 19, and 20) showed
appreciable agonistic activity in the corresponding agonism assays
in the same strain. Compounds capable of agonizing LasR in both
assays can be clearly defined as LasR agonists, but those that fail
to agonize in the agonism assay could operate by a different mech-
anism (see below). Second, while native AHL thiolactone mimics
19 and 20 exhibited an expected correlation between agonism
and antagonism trends in the P. aeruginosa LasR reporter strain,
this pattern is not mimicked in the E. coli LasR reporter strain.
2
1
V. fischeri ESI 114 (-LuxI)38 and A. tumefaciens WCF (pCF372)39
were used to examine the activity of the thiolactone library in LuxR
and TraR, respectively. The V. fischeri mutant strain lacks a func-
tioning LuxI synthase, but retains its native lux operon, allowing
a quantitative luminescent readout based on LuxR activity. Simi-
larly, A. tumefaciens WCF (pCF372)39 lacks a functioning TraI, yet
contains a plasmid with a TraR responsive promoter for the b-gal
gene, thereby allowing for direct quantitation of TraR activity.
We used bacteriological assay protocols for small molecule
screening that were analogous to those reported in our earlier
studies (See Section 4). All synthetic compounds were screened
2
Conversely, 3-NO thiolactones 26/27 were highly active agonists
in the E. coli LasR reporter strain, yet minimally active in the P. aeru-
ginosa reporter strain. This activity profile is the opposite of its par-
ent AHL 6, which is one of the stronger LasR inhibitors known.21
2
We note, however, that 3-NO AHL 6 is also an extremely potent
2
0
agonist of LuxR. Third, the remaining potent LasR agonists iden-
tified in the E. coli reporter strain (22 and 34) were also significant
LasR antagonists in this strain. On the whole, the only LasR agonists
in the thiolactone library capable of agonizing LasR in the agonist
and antagonist assays in both reporter strains were thiolactone
analogs of natural AHL ligands: OdDHL analog 14/15, OHHL analog
12, and N-dodecanoyl thiolactone 23. These results, while unex-
pected, corroborate with those previously reported by Passador
at 10
lM in both agonism and antagonism assays in the bacterial
reporter strains. Notably, these compound concentrations parallel
those used by our laboratory in past studies,2
2,40
allowing for com-
2
8
parisons to be made between the assay data reported here and this
past work. No effects on bacterial growth were observed over the
time course of the reporter gene assays (4–16 h).
and co-workers for 15.
In contrast to the P. aeruginosa LasR reporter strain in which no
antagonists were found, several highly potent LasR antagonists were
found using the E. coli reporter strain. The non-3-oxo aliphatic thio-
lactones 20–22 and phenylacetanoyl thiolactones 24/25 and 28-34
were all found to be ꢀ50–80% inhibitors of LasR this strain. Many
of these compounds show complex trends in activity. For example,
both the racemic and enantiopure phenylacetanoyl thiolactones
2
.3. Antagonism and agonism assays in LasR
The primary antagonism and agonism data for the thiolactone
library in the LasR E. coli and P. aeruginosa reporter strains are
Table 1
LasR primary antagonism and agonism assay data for the thiolactone library in P. aeruginosa and E. coli reporter strainsa
Compound
P. aeruginosa LasR
Antagonismb (%)
E. coli LasR
Agonismc (%)
Antagonismb (%)
Agonismc (%)
12
14
15
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
À23
À49
À55
À8
76
127
88
3
22
20
0
À119
À61
À40
À30
17
59
78
64
96
102
94
0
4
1
À24
À23
À8
8
À7
0
85
94
3
À11
À24
À25
À15
À7
42
17
2
3
5
4
7
5
8
1
3
3
10
À93
61
65
2
À3
À13
48
51
68
80
45
56
54
81
82
0
1
8
2
0
9
72
4
À18
À23
À18
À17
À23
À43
À22
À7
16
a
All synthetic compounds were screened at 10 lM. All assays were preformed in triplicate; error did not exceed ±10%. Positive controls were
OdDHL at its EC50 value for the strain in antagonism assays and 100 times its EC50 value for the strain in agonism assays. Negative controls
contained neither thiolactone nor natural AHL, and were subtracted from each sample to account for background. Negative antagonism values
indicate that the compound activates at the tested concentration. See text for details of strains.
b
Antagonism assays were preformed against the EC50 value for OdDHL in each strain: P. aeruginosa PA01 MW1 (pUM15) = 1 M; E. coli DH5
(
pJN105L + pSC11) = 10 nM.
Agonism assays were normalized to the positive control (OdDHL) in each strain.
c

4
824
C. E. McInnis, H. E. Blackwell / Bioorg. Med. Chem. 19 (2011) 4820–4828
2
8–31 are good to excellent inhibitors in the E. coli LasR reporter
strain, yet are mild agonists in the P. aeruginosa reporter strain antag-
onism assays. This is particularly intriguing since neither the 4-
methoxy nor the 4-Br phenylacetanoyl parent AHLs (7 and 8) were
very active in LasR; rather, they are both active antagonists in
2
0
LuxR. Thiolactones 24/25 and 32/33, which contain either an
unsubstituted phenyl or a biphenyl tail, exhibit even more exagger-
ated differences in activity; they are simultaneously good antago-
nists in the E. coli strain and good agonists in the antagonism assay
in the P. aeruginosa strain. The parent AHLs (control phenyl AHL 5
and biphenyl AHL 9) display minimal activity in LasR as well.2 These
data trends for thiolactone analogs versus their parent AHLs clearly
suggest that minor perturbations (i.e., O to S) to the AHL head group
can have dramatic effects on overall compound activity and receptor
selectivity.
1
We note again here that the P. aeruginosa reporter strain retains
the two other LuxR-type proteins that govern QS in this organism
(
QscR and RhlR), while the E. coli strain lacks these receptors. The
lack of these receptors in E. coli could be one reason for the compli-
cated thiolactone activity trends for LasR outlined above when
comparing the two strains. Differences in membrane permeability
between the two strains could be another contributing factor (see
above). Both RhlR and QscR can regulate LasR to some degree, with
QscR directly repressing LasR using its identical cognate ligand,
1
9
OdDHL. In previous work, we have shown that many of our
non-native AHLs can simultaneously inhibit LasR and QscR. In this
context, thiolactones 24, 25, and 28–33 could also be inhibiting
QscR in addition to LasR. A QscR inhibitor would yield LasR agonis-
tic activity in the P. aeruginosa strain, yet would exhibit LasR antag-
onistic activity in the QscR null E. coli strain. This is the trend
observed above for these compounds. Additional experiments with
QscR are necessary to evaluate this hypothesis.
Since it was difficult to determine a definitive acyl group SAR
for the thiolactone LasR agonists and antagonists, we scrutinized
the thiolactone stereochemistry to determine what role it played,
if any, in dictating compound activity. If stereochemistry played
a large role in receptor binding, we hypothesized that the enantio-
pure
racemic thiolactones (screened at the same concentration). Our
reasoning proved incorrect, however, as the enantiopure -thiolac-
L-thiolactones would be approximately twice as active as the
L
tones and their racemates largely exhibited approximately the
same activity (Table 1). (We note that no appreciable racemization
was observed in our synthesis procedure for the L-thiolactones).
This result suggests that however these ligands modulate LasR,
the interaction is less specific than that of native AHL ligands.
Thiolactones 35 and 34, however, are a special case. Although they
are LasR antagonists in the E. coli strain and LasR agonists in the
P. aeruginosa strain, 34 (the racemate) has higher activity in both
cases than 35 (the enantiopure
L-form). For these compounds,
the -enantiomer may be the more active form of the compound.
D
2
.4. Possible agonism model in LasR for thiolactones
The interesting trends in LasR agonism for the thiolactones in
the presence of OdDHL within the individual strains caused us to
consider the mechanism of LasR activation in more detail (Fig. 3).
Many of the thiolactone library members have similar architec-
tures as OdDHL (most notably the aliphatic derivatives), allowing
us to make the reasonable assumption that they can potentially
bind in the OdDHL binding pocket in LasR. A simple schematic of
LasR’s mechanism of action is shown in Figure 3A. OdDHL binds
to LasR, which then homodimerizes and binds to the promoter re-
gion of one of many genes under LasR control including LasI, RhlI,
RhlR, and RsaL, acting as a transcriptional activator.16 Similarly,
Figure 3. A schematic diagram of possible mechanisms for LasR agonism and
antagonism by OdDHL (native AHL) or a non-native AHL analog. OdDHL is shown as
the purple trapezoid, and the non-native ligand is shown as the blue pentagon. (A)
The accepted mechanism for native ligand (OdDHL) activation of LasR. (B) Synthetic
ligand binding to LasR. (C) Proposed binding mode for cooperative agonists. (D)
Proposed binding mode for ‘bimodal binders.’ See text for further detail.

C. E. McInnis, H. E. Blackwell / Bioorg. Med. Chem. 19 (2011) 4820–4828
4825
when a non-native LasR agonist replaces OdDHL in reporter gene
assays, LasR:non-native ligand homodimers are believed to form
Fig. 3B). In competitive antagonism assays, it is generally assumed
2.5. Antagonism and agonism assays in LuxR and TraR
(
The antagonism and agonism primary screening data for the
thiolactone library in LuxR and TraR are shown in Table 2. Several
exquisitely active compounds were identified, indicating the abil-
ity of acylated thiolactones to strongly modulate LuxR-type recep-
tors beyond LasR. Sixteen thiolactones (including racemic
compounds) were found to inhibit LuxR by over 50%. Notably, thio-
lactones 15, 18, 20, 27/28, 30/31, and 32/33, all of which are over
90% inhibitors, represent some of the most potent LuxR antagonists
reported to date. The strong antagonistic activity of 15, the thiolac-
tone mimic of OdDHL, is analogous to OdDHL itself in LuxR.22 Thio-
lactone 15 was also incapable of activating LuxR, corroborating the
that the non-native ligand is able to outcompete OdDHL for the
binding pocket, either by creating inactive homodimers or by pre-
venting dimerization from occurring. Based on the work of Bassler
3
2
and co workers, it is also possible that the inactive homodimers
are able to bind DNA, but the homodimer:DNA complex is then
unable to successfully interact with RNA polymerase. Two
alternate scenarios are possible, however, if one considers LasR
heterodimer formation. Such heterodimers would consist of one
LasR bound to OdDHL and one LasR bound to a non-native ligand
(Fig. 3C and D).
2
9
In one scenario (Fig. 3C), non-native ligands may act as what we
work of Schaefer et al. Surprisingly, 3-NO
2
phenylacetanoyl thio-
term cooperative agonists if they are able to form viable heterodi-
mers with LasR, yet are either unable to form homodimers or form
inactive homodimers. This could lead to an observed high agonistic
activity in a competitive antagonism assay against OdDHL There-
fore, if the homodimers of the non-native ligand are either inactive
or unable to form, negligible agonistic activity would be observed
for the non-native ligand alone in an agonism assay. For example,
thiolactones 24/25, 29, and 32–34 exhibit this trend in the
P. aeruginosa strain. Conversely, in a second scenario (Fig. 3D), if
the synthetic ligand is an agonist when OdDHL is absent and is
an antagonist when OdDHL is present, both the non-native and
OdDHL homodimers could likely agonize LasR, while the heterodi-
mers are inactive. An antagonistic effect would therefore be
observed when the natural ligand is present (Fig. 3D). We have
named compounds that could bind in this latter manner ‘bimodal
binders’ because of their dual antagonistic and agonistic proper-
ties. We uncovered thiolactones displaying this activity profile in
both LuxR and TraR (see below). Additional biochemical experi-
ments are of course required to validate these models for LasR
and related LuxR-type proteins, and are currently underway in
our laboratory. Nonetheless, we find these models helpful for
understanding the screening data obtained above for the thiolac-
tones in LasR, and below in LuxR and TraR.
lactones 26/27 were only weak agonists of LuxR, even though their
parent 3-NO AHL (6) is a ‘super activator’ of LuxR. This change in
2
2
0
activity may be a result of the possible bimodal binder activity pro-
file for 26/27 in LuxR (see above), as it is also behaves as a weak
LuxR antagonist. However, several thiolactone displayed activities
that tracked well with their parent AHLs. Thiolactones 28/29 were
found to have analogous LuxR antagonistic activities as their meth-
oxy AHL analog (7), which is a moderate inhibitor of LuxR. Simi-
larly, thiolactones 30–34, modeled after our lactone-based LuxR
inhibitors, were all strong LuxR antagonists, while 24/25 (based
on inert AHL 5) were only weakly inhibitory.
Turning to TraR, we only found one thiolactone, the OOHL mi-
mic 18, which displayed appreciable agonistic activity in TraR
(50%; Table 2). Thiolactone 18 is notable as very few non-native
2
1
TraR agonists have been reported to date. Several thiolactones,
however, appeared to be either capable of what we now propose
to be cooperative agonism (see above), or were extremely strong
TraR antagonists. Compounds 19, 20, 26, and 27 behaved as coop-
erative agonists, with 19 and 20 agonizing TraR in the presence of
OOHL to over 70% more than OOHL alone. The N-hexanoyl thiolac-
tone 20 is the non-3-oxo analog of OHHL, suggesting that close
similarities in ligand structure may permit active heterodimer
formation. In turn, thiolactones 21 and 22 were very strong TraR
Table 2
a
LuxR and TraR primary antagonism and agonism assay data for the thiolactone library in V. fischeri and A. tumefaciens reporter strains
Compound V. fischeri (LuxR)
A. tumefaciens (TraR)
Antagonismb (%)
Antagonismb (%)
Agonismc (%)
Agonismc (%)
12
14
15
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
62
91
99
99
42
93
78
80
57
30
34
30
19
59
70
91
97
85
98
75
68
0
0
0
0
0
0
0
0
0
0
0
26
21
0
1
0
0
0
93
51
92
À89
À70
À78
78
99
6
13
1
1
1
50
0
0
0
0
0
0
3
0
1
1
0
1
0
11
0
À22
30
33
À24
À107
25
À60
26
0
0
0
9
À26
6
0
À80
a
See footnote a in Table 1.
b
Antagonism assays were preformed against the EC50 value for the native AHL each strain: V. fischeri ESI 114 (-LuxI) = 2 M OHHL; A. tumefaciens WCF
(
pCF372) = 200 nM OOHL.
Agonism assays were normalized to the positive control (native ligand) in each strain.
c

4
826
C. E. McInnis, H. E. Blackwell / Bioorg. Med. Chem. 19 (2011) 4820–4828
inhibitors, with 22 surpassing the activity of its parent N-octanoyl
HL (3) and inhibiting TraR by 99%.22 Compound 21 has a one car-
bon shorter acyl chain than 22 and is 20% less active, correlating
well with acyl chain length activity trends seen for similar AHL
tive LasR antagonists in the E. coli strain (i.e., the N-heptanoyl (21)
and biphenyl (33) -thiolactones).
L
The dose response data for four thiolactone agonists are wor-
thy of attention (12, 15, 18, and 23; Table 3). Interestingly, all
four were native AHL analogs. The thiolactone analog of OOHL
(18) is especially notable, as to our knowledge it is the most ac-
tive non-native TraR agonist reported to date (albeit 100-fold less
active then OOHL). In turn, OdDHL thiolactone analog 15 had EC50
values approximately equal to OdDHL in both the E. coli and P.
aeruginosa LasR reporter strains (ꢀ10 nM and 1 mM, respectively).
OHHL thiolactone analog 12 was also a strong agonist in both
LasR reporter strains, but with EC50 values approximately 10-fold
higher than OdDHL in each strain. Interestingly, unlike the thio-
lactone analogs of OOHL (18) in TraR and OdDHL (15) in LasR,
OHHL thiolactone analog (12) was a mild LuxR antagonist instead.
Finally, thiolactone 23, the non-3-oxo thiolactone analog of OdD-
HL, was found to be ꢀ100-fold less active then OdDHL in the
E. coli LasR reporter strain, yet was only ꢀ10-fold less active than
OdDHL in the P. aeruginosa reporter strain. This difference in
activity is significant because many of the thiolactones examined
in the primary assays displayed the opposite trend in the two
strains, that is, heightened activity in the E. coli versus the P. aeru-
ginosa LasR reporter strain.
2
2
analogs in TraR.
Analogous to the LasR primary screening data, thiolactone
stereochemistry did not appear critical for LuxR agonism or antag-
onism. However, it did influence TraR agonism and antagonism
trends in some cases. For example, the racemic thiolactone 30
and
antagonist, while 31 is a 60% TraR agonist in the antagonism assay.
These data suggest that the -stereoisomer may have higher TraR
inhibitory activity, and the -stereoisomer then has higher agonis-
tic activity. In another example, -thiolactone 29 exhibited very
L-thiolactone 31 showed a reversal in activity; 30 is a 25% TraR
D
L
L
strong agonistic activity in the TraR antagonism assay (2x the activ-
ity of OOHL), while the racemic thiolactone 28 is only a 24% TraR
agonist. For both 28 and 29, agonistic activity is only observed
when OOHL is present. These data suggest that 28 and 29 could
function as cooperative agonists of TraR (see above), with the L
enantiomer imparting the majority of the activity. Indole thiolac-
tones 34 and 35 showed a similar agonistic activity profile as 28
and 29, although not quite as exaggerated.
Overall, the thiolactone analog of OdDHL (15) was highly active
in all strains tested—either as a LasR agonist (94% in E. coli and 88%
in P. aeruginosa) or an antagonist in LuxR (92%) and TraR (65%).
These trends match those observed for OdDHL in each strain, and
suggest that, at least for 15, sulfur replacement in the lactone ring
does not significantly affect its interactions with a LuxR-type
receptor.
Further comparison of the thiolactone dose response data to
that for our previously reported AHLs (1–9) revealed that the
oxygen?sulfur replacement can both augment and diminish
relative compound activities in AHL analogs (Table 3).2
0,21
For
example, thiolactones 21, 22 and 33 largely exhibit similar ago-
nistic or antagonistic potencies as their parent AHLs (2, 3, and 9)
in LasR (E. coli), LuxR and TraR. The N-dodecanoyl thiolactone
(23), however, displayed reduced agonistic activity relative to
its AHL analog (4), suggesting that the sulfur replacement can
impact ligand activity in some cases. Likewise, the phenylaceta-
noyl thiolactone 25 was a moderate LasR inhibitor in E. coli,
while it parent AHL (5) was largely inactive in most LuxR-type
2
.6. Dose response analyses for selected L-thiolactones
Dose response analyses for selected active,
L-thiolactones (12,
1
4, 15, 20–23, 25, 27, 29, 31, 33, and 35) were conducted to deter-
mine their IC50 and EC50 values in the bacterial reporter strains.
These data are listed in Table 3, and revealed several agonists
and antagonists with sub-micromolar activities (15, 18, 20–22,
2
receptors. Notably, the 3-NO thiolactone 27 and indole thiolac-
tone 35 were single digit micromolar agonists of LasR in E. coli,
while their corresponding AHL analogs (6 and 10) were sub-
3
1, and 33). The majority of these compounds were LuxR antago-
nists. The most active compound identified overall was N-octanoyl
-thiolactone (22), which could antagonize LuxR by 50% at a
2
1
micromolar LasR antagonists.
Clearly, additional thiolactones
L
will need to be studied to gain more insight into the subtle SARs
behind these activity switches. Structural studies of these com-
pounds complexed to various LuxR-type proteins could be par-
ticularly illuminating, as indicated by the recent X-ray
crystallographic work of Hughson and co-workers on related
1
0-fold lower concentration than OHHL. Interestingly, thiolactone
2
2 was also the strongest LasR antagonist identified in the E. coli
reporter strain, capable of antagonizing LasR by 50% at a 10-fold
higher concentration than OdDHL. Following this trend, the second
and third most active LuxR antagonists were also the next most ac-
3
3
thiolactones bound to CviR.
Table 3
IC50 and EC50 values for the most active L-thiolactone antagonist and agonistsa
Compound
E. coli LasR
P. aeruginosa LasR
V. fischeri LuxR
A. tumefaciens TraR
IC50b
(l
M)
EC50
0.13
(lM)
EC50
(l
M)
IC50b
(
lM)
EC50
(lM)
IC50b
(lM)
EC50
20
(lM)
c
12
15
18
20
21
22
23
25
27
29
31
33
35
À
13
3.2
3.2
1.8
0.092
0.45
0.35
0.84
0.31
0.13
1.1
0.79
0.14
10
2.8
1.9
4.1
21
2.5
11
7.2
0.40
2.9
0.77
0.35
1.8
a
b
c
See footnotes for Table 1 and Table 2 for details.
The IC50 values were determined in the presence of the native AHL for each strain at its EC50 value.
Vacant cells were not determined.

C. E. McInnis, H. E. Blackwell / Bioorg. Med. Chem. 19 (2011) 4820–4828
4827
2
.7. Thiolactone hydrolysis study
oxygen analogs,42 we hypothesize that 15 is either able to remain
closed for longer than OdDHL or it is able to ring close at higher
pHs, shifting the half-life to longer times. We note that the LasR,
LuxR, and TraR reporter gene assays above were conducted for
no longer than 16 h, so the majority of both OdDHL and 15 is likely
in the closed form and ligand stability (or free thiol reactivity) did
not have a significant impact on the observed activities. Neverthe-
less, these hydrolysis data suggest that thiolactone AHL analogs,
such as 15, could have value for biological experiments conducted
over longer time periods.
We sought to examine the hydrolytic stability of the thiolactone
analogs described above, as generating probe compounds with
heightened hydrolytic activity relative to AHLs was one motivation
for this study. Like lactones, thiolactones are also hydrolyzable.
However, direct comparison of the hydrolysis rates of AHLs versus
their thiolactone analogs has, to our knowledge, yet to be reported.
We provide such an analysis here.
We chose to analyze 15 as it was the most active thiolactone
modulator identified in this study. A biological assay was devel-
oped to determine the functional half-life of 15 and its correspond-
ing AHL analog, OdDHL. Our assay did not directly measure lactone
hydrolysis, but rather the ability of the remaining non-hydrolyzed
ligand to activate LasR. Previous experiments have shown that the
hydrolysis half-life for OdDHL is approximately two days, while
racemization of the chiral center was found to be less than 5% over
3
. Conclusion
We have designed and synthesized a library of thiolactone
analogs of both naturally occurring and non-native AHLs. These
thiolactones were evaluated for both antagonistic and agonistic
activity in three relevant LuxR-type receptors (LasR, LuxR, and
TraR) using cell-based reporter gene assays. The screening data
provided a complex set of activity profiles for these compounds
2
7
the course of a week. These data suggest that most ligand degra-
dation is due to hydrolysis and not epimerization. Previous hydro-
lysis experiments have been monitored by NMR spectroscopy and
(
most notably in LasR). However, several new and highly active
2
7,41
performed in deuterated buffers containing 50% DMSO.
The
QS modulators were discovered, many with multi-receptor activ-
ity. Compounds 15, 18–22, and 30–33 all had nanomolar IC50 val-
ues in V. fischeri and are some of the most potent LuxR antagonists
to be reported. OOHL thiolactone analog 18 is also a strong LuxR
antagonist, and simultaneously is the strongest non-native TraR
agonist to be reported. In terms of LasR, thiolactones 21, 22, and
use of high levels of DMSO reduces the biological relevance of
the assay. In order to assess ligand degradation under more biolog-
ically relevant conditions, we monitored ligand activity levels
using a LasR reporter strain (PA01 MW1 (pUM15); introduced
3
6
above).
Our hydrolysis assay protocol utilized the P. aeruginosa LasR re-
porter to detect ligand activity after incubation in growth media
for specified times using fluorescence (YFP; see Section 4). Fluores-
cence values corresponded to the degree of ligand degradation prior
to contact with the reporter strain, and were normalized to that of a
freshly prepared ligand sample (i.e., the positive control). Since we
predict that the ligand degradation is due to hydrolysis, we assumed
a pseudo first-order rate and plotted the natural log of the ligand
activity as a percent of the positive control versus time. The slope
of the graph therefore could be used to determine the half-life of
the ligand according to the formula t1/2 = ln (2)/-slope (Fig. 4).
Using this reporter assay procedure, we determined the half-life
for OdDHL to be 48.2 h. This value corresponds closely with the
hydrolysis half-life for OdDHL previously reported by Spring and
3
1 displayed nanomolar IC50 values in the E. coli LasR reporter.
The thiolactone analogs of both OdDHL and its non-3-oxo HL
analog (15 and 23) are strong LasR agonists in the E. coli system,
supporting previous studies that showed that acyl chain length is
important for receptor selectivity. However, this trend does not
always hold true, as we found that the OHHL thiolactone analog
with a six-carbon shorter acyl tail (12) had similar LasR agonistic
activity as 15.
The complexity of the thiolactone activity profiles observed in
LasR motivated us to propose models for what we term coopera-
tive agonism and bimodal binding in LuxR-type receptors. Several
of the active thiolactones reported herein could behave via these
pathways. While thiolactones that had the highest structural sim-
ilarity to their parent AHL appeared to retain the highest degree of
activity, the relatively minor change from a lactone to thiolactone
could either reduce activity or flip activity from antagonism to
agonism for many of the other thiolactone library members. Unex-
pectedly, thiolactone stereochemistry only appeared critical for li-
gand activity in TraR from A. tumefaciens, which is known to have
2
7
co-workers. A similar analysis of thiolactone 15 revealed its
half-life to be 82.3 h, almost twice that of OdDHL. Yates and co-
workers have shown that once ring-opened, the lactone does not
4
1
reclose in appreciable quantities until pH values of less than 2.
Since sulfur compounds are stronger nucleophiles than their
4
3,44
fairly rigid binding pocket.
In the course of these studies, a new assay was developed to as-
sess the rates of thiolactone and lactone hydrolysis. Using this as-
say, we found that the half-life of the thiolactone derivative of
OdDHL (15) is almost double that of OdDHL itself. As such, the
new thiolactone agonists and antagonists reported here may offer
a set of compounds with enhanced hydrolytic stabilities relative to
AHLs, and could find utility as new chemical probes to study QS in
Gram-negative bacteria.
4
4
. Experimental section
.1. Acylated thiolactone synthesis
Thiolactones lacking 3-oxo functionalities in their acyl groups
were synthesized in solution using previously reported EDC
coupling protocols and purified using acid-base extraction
procedures.
Thiolactones containing 3-oxo acyl groups were prepared from
acylated Meldrum’s acid derivatives. The Meldrum’s acid
Figure 4. Hydrolysis of OdDHL and its thiolactone analog (15) plotted as a function
45,46
of time versus the natural log of the percent of the positive control. The slope is
2
À0.01399 ± 0.0006 for OdDHL and À0.01025 ± 0.0004 for 15. The R value for both
slopes is 0.96. Error is less than ±2% for each time point.

4
828
C. E. McInnis, H. E. Blackwell / Bioorg. Med. Chem. 19 (2011) 4820–4828
derivatives were synthesized by dissolving Meldrum’s acid (0.5 g,
.5 mmol) in CH Cl (50 mL) under nitrogen at 0 °C. DMAP
0.86 g, 7 mmol) was added and allowed to dissolve completely.
References and notes
3
(
2
2
1
2
3
.
.
.
Platt, T. G.; Fuqua, C. Trends Microbiol. 2010, 18, 383.
Bassler, B. L.; Losick, R. Cell 2006, 125, 237.
Camilli, A.; Bassler, B. L. Science 2006, 311, 1113.
Thereafter, the alkyl acid chloride (3.5 mmol) was added to the
reaction mixture slowly over 1 h at 0 °C. The reaction mixture
was allowed to stir for an additional hour at 0 °C, and then over-
night (ꢀ14 h) at rt. The reaction mixture was washed 2Â with
4. Fuqua, C.; Greenberg, E. P. Nat. Rev. Mol. Cell Biol. 2002, 3, 685.
5
.
Boedicker, J. Q.; Vincent, M. E.; Ismagilov, R. F. Angew. Chem., Int. Ed. 2009, 48,
908.
Pacheco, A. R.; Sperandio, V. Curr. Opin. Microbiol. 2009, 12, 192.
5
6.
2
M HCl and 1Â with brine, and the organic layer was dried over
7. Smith, R. S.; Iglewski, B. H. Curr. Opin. Microbiol. 2003, 6, 56.
8. Rasko, D. A.; Sperandio, V. Nat. Rev. Drug Disc. 2010, 9, 117.
9. Geske, G. D.; O’Neill, J. C.; Blackwell, H. E. Chem. Soc. Rev. 2008, 37, 1432.
4
MgSO , filtered, and concentrated to yield the crude acylated
Meldrum’s acid derivative.
10. Amara, N.; Mashiach, R.; Amar, D.; Krief, P.; Spieser, S. A. H.; Bottomley, M. J.;
The 3-oxo thiolactones were synthesized by stirring the homo-
cysteine thiolactone (25 mg, 0.16 mmol) with the Meldrum’s acid
derivative (0.16 mmol) in acetonitrile (15 mL) at rt under nitrogen
for 2 h, followed by refluxing for 5 h. The acetonitrile was then
removed, and the reaction residue was redissolved in ethyl acetate
and washed 1Â each with saturated sodium bicarbonate, 1 M
Aharoni, A.; Meijler, M. M. J. Am. Chem. Soc. 2009, 131, 10610.
11. Galloway, W. R. J. D.; Hodgkinson, J. T.; Bowden, S. D.; Welch, M.; Spring, D. R.
Chem. Rev. 2011, 111, 28.
2. Stevens, A. M.; Queneau, Y.; Soulere, L.; von Bodman, S.; Doutheau, A. Chem.
1
Rev. 2011, 111, 4.
13. Kaufmann, G. F.; Sartorio, R.; Lee, S. H.; Mee, J. M.; Altobell, L. J.; Kujawa, D. P.;
Jeffries, E.; Clapham, B.; Meijler, M. M.; Janda, K. D. J. Am. Chem. Soc. 2006, 128,
2802.
4 4
NaHSO , and brine. The organic layer was dried over MgSO ,
1
1
1
4. Amara, N.; Krom, B. P.; Kaufmann, G. F.; Meijler, M. M. Chem. Rev. 2011, 111,
filtered, and concentrated to yield the 3-oxo thiolactone products.
Further purification by flash silica gel column chromatography
195.
5. Welch, M.; Mikkelsen, H.; Swatton, J. E.; Smith, D.; Thomas, G. L.; Glansdorp, F.
G.; Spring, D. R. Mol. Biosyst. 2005, 1, 196.
(
25% EtOAc/hexanes) was necessary for 14 and 18.
6. Schuster, M.; Greenberg, E. P. In Chemical Communication Among Bacteria;
Winans, S. C., Bassler, B. L., Eds.; ASM Press: Washington, DC, 2008; p 133.
7. Patankar, A. V.; Gonzalez, J. E. FEMS Microbiol. Rev. 2009, 33, 739.
8. Schaefer, A. L.; Greenberg, E. P.; Oliver, C. M.; Oda, Y.; Huang, J. J.; Bittan-Banin,
G.; Peres, C. M.; Schmidt, S.; Juhaszova, K.; Sufrin, J. R.; Harwood, C. S. Nature
Purities and isolated yields for these compounds were 91–99%
and 35–60%, respectively. See Supplementary data for full com-
pound characterization details.
1
1
2008, 454, 595.
4
.2. Bacterial reporter gene assays
19. Mattmann, M. E.; Blackwell, H. E. J. Org. Chem. 2010, 75, 6737.
2
2
0. Geske, G. D.; O’Neill, J. C.; Blackwell, H. E. ACS Chem. Biol. 2007, 2, 315.
1. Geske, G. D.; O’Neill, J. C.; Miller, D. M.; Mattmann, M. E.; Blackwell, H. E. J. Am.
Chem. Soc. 2007, 129, 13613.
Reporter gene assays for LasR (in E. coli), LuxR, and TraR were
2
1
performed according to our previously reported methods. The
LasR assay in the P. aeruginosa PA01 MW1 (pUM15) strain was
modified from a literature procedure.37 See Supplementary data
for full details of bacteriological protocols.
22. Geske, G. D.; O’Neill, J. C.; Miller, D. M.; Wezeman, R. J.; Mattmann, M. E.; Lin,
Q.; Blackwell, H. E. ChemBioChem 2008, 9, 389.
2
3. Mattmann, M. E.; Shipway, P. M.; Heth, N. J.; Blackwell, H. E. ChemBioChem
011, 12, 942.
24. Praneenararat, T.; Geske, G. D.; Blackwell, H. E. Org. Lett. 2009, 11, 4600.
5. Palmer, A. G.; Streng, E.; Jewell, K. A.; Blackwell, H. E. ChemBioChem 2011, 12,
38.
2
2
1
4
.3. Thiolactone hydrolysis study
26. Breitbach, A. S.; Broderick, A. H.; Jewell, C. M.; Gunasekaran, S.; Lin, Q.; Lynn, D.
M.; Blackwell, H. E. Chem. Commun. 2010, 47, 370.
2
7. Glansdorp, F. G.; Thomas, G. L.; Lee, J. J. K.; Dutton, J. M.; Salmond, G. P. C.;
Thiolactone ligand was incubated in Luria-Bertani medium at
Welch, M.; Spring, D. R. Org. Biomol. Chem. 2004, 2, 3329.
3
7 °C for predetermined times, and then added to a 96-well plate.
28. Passador, L.; Tucker, K. D.; Guertin, K. R.; Journet, M. P.; Kende, A. S.; Iglewski, B.
H. J. Bacteriol. 1996, 178, 5995.
The P. aeruginosa PA01 MW1 (pUM15) reporter strain was cultured
overnight, pelleted, washed, and resuspended in a minimal amount
of media. These cells were then added to the 96-well plate contain-
ing ligand, incubated for 8 h, and the wells were analyzed for opti-
cal density and YFP fluorescence. Fluorescence was normalized to
cell density per well, and each time point was referenced to the
cell-density-normalized readout of ligand prepared at time 0 (i.e.,
the positive control).
29. Schaefer, A. L.; Hanzelka, B. L.; Eberhard, A.; Greenberg, E. P. J. Bacteriol. 1996,
1
78, 2897.
30. Chhabra, S. R.; Harty, C.; Hooi, D. S. W.; Daykin, M.; Williams, P.; Telford, G.;
Pritchard, D. I.; Bycroft, B. W. J. Med. Chem. 2003, 46, 97.
31. Janssens, J. C.; Metzger, K.; Daniels, R.; Ptacek, D.; Verhoeven, T.; Habel, L. W.;
Vanderleyden, J.; De Vos, D. E.; De Keersmaecker, S. C. Appl. Environ. Microbiol.
2007, 73, 535.
32. 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.
33. Chen, G.; Swem, L. R.; Swem, D. L.; Stauff, D. L.; O’Loughlin, C. T.; Jeffrey, P. D.;
Bassler, B. L.; Hughson, F. M. Mol. Cell 2011, 42, 199.
3
3
4. Dickschat, J. S. Nat. Prod. Rep. 2010, 27, 343.
5. Geske, G. D.; Mattmann, M. E.; Blackwell, H. E. Bioorg. Med. Chem. Lett. 2008, 18,
Acknowledgments
5978.
Financial support for this work was provided by the NIH
36. Lee, J. H.; Lequette, Y.; Greenberg, E. P. Mol. Microbiol. 2006, 59, 602.
37. Muh, U.; Schuster, M.; Heim, R.; Singh, A.; Olson, E. R.; Greenberg, E. P.
Antimicrob. Agents Chemother. 2006, 50, 3674.
(
AI063326) ONR (N00014-07-1-0255), Greater Milwaukee Founda-
tion Shaw Scientist Program, Burroughs Welcome Fund, Camille &
Henry Dreyfus Foundation, Research Corporation, and Johnson &
Johnson. C.E.M. was supported in part through a DOD (Air Force Of-
fice of Scientific Research) National Defense Science and Engineer-
ing Graduate (NDSEG) Fellowship (32 CFR 168a). We gratefully
acknowledge Professors Peter Greenberg (University of Washing-
ton), Stephen Winans (Cornell University), and Edward Ruby
38. Lupp, C.; Urbanowski, M.; Greenberg, E. P.; Ruby, E. G. Mol. Microbiol. 2003, 50,
319.
39. Zhu, J.; Beaber, J. W.; More, M. I.; Fuqua, C.; Eberhard, A.; Winans, S. C. J.
Bacteriol. 1998, 180, 5398.
40. Mattmann, M. E.; Geske, G. D.; Worzalla, G. A.; Chandler, J. R.; Sappington, K. J.;
Greenberg, E. P.; Blackwell, H. E. Bioorg. Med. Chem. Lett. 2008, 18, 3072.
41. Yates, E. A.; Philipp, B.; Buckley, C.; Atkinson, S.; Chhabra, S. R.; Sockett, R. E.;
Goldner, M.; Dessaux, Y.; Camara, M.; Smith, H.; Williams, P. Infect. Immun.
2002, 70, 5635.
(
UW–Madison) for donations of bacterial strains and advice on
42. Smith, M.; March, J. Advanced Organic Chemistry: Reactions, Mechanisms, and
Structure; New York: Wiley, 2001.
their manipulation.
43. Vannini, A.; Volpari, C.; Gargioli, C.; Muraglia, E.; Cortese, R.; De Francesco, R.;
Neddermann, P.; Di Marco, S. EMBO J. 2002, 21, 4393.
Supplementary data
44. Zhang, R. G.; Pappas, T.; Brace, J. L.; Miller, P. C.; Oulmassov, T.; Molyneauz, J.
M.; Anderson, J. C.; Bashkin, J. K.; Winans, S. C.; Joachimiak, A. Nature 2002, 417,
9
71.
Supplementary data (compound characterization data, biologi-
cal assay protocols, and additional screening data) associated with
this article can be found, in the online version, at doi:10.1016/
j.bmc.2011.06.071.
4
4
5. Geske, G. D.; Wezeman, R. J.; Siegel, A. P.; Blackwell, H. E. J. Am. Chem. Soc. 2005,
127, 12762.
6. Smith, K. M.; Bu, Y.; Suga, H. Chem. Biol. 2003, 10, 563.