Solid-phase synthesis and biological evaluation of N-dipeptido L -homoserine lactones as quorum sensing activators

Article abstract of DOI:10.1002/cbic.201300533

Bacteria use small signaling molecules to communicate in a process termed quorum sensing (QS), which enables the coordination of survival strategies, such as production of virulence factors and biofilm formation. In Gram-negative bacteria, these signaling molecules are a series of N-acylated L-homoserine lactones. With the goal of identifying non-native compounds capable of modulating bacterial QS, a virtual library of N-dipeptido L-homoserine lactones was screened in silico with two different crystal structures of LasR. The 30 most promising hits were synthesized on HMBA-functionalized PEGA resin and released through an efficient acid-mediated cyclative release mechanism. Subsequent screening for modulation of QS in Pseudomonas aeruginosa and E. coli identified six moderately strong activators. A follow-up library designed from the preliminary derived structure-activity relationships was synthesized and evaluated for their ability to activate the QS system in this bacterium. This resulted in the identification of another six QS activators (two with low micromolar activity) thus illuminating structural features required for QS modulation. Behind the screen: A virtual screening approach identified a number of N-dipeptido L-homo-serine lactones as potential quorum sensing modulators. Selected compounds were synthesized and identified as activators of LasR, a protein regulating quorum sensing in Pseudomonas aeruginosa; EC₅₀ values were in the low micromolar range. Copyright

Full text of DOI:10.1002/cbic.201300533

CHEMBIOCHEM
FULL PAPERS
DOI: 10.1002/cbic.201300533
Solid-Phase Synthesis and Biological Evaluation of N-
Dipeptido l-Homoserine Lactones as Quorum Sensing
Activators
Mette R. Hansen,[a] Sebastian T. Le Quement,[a] Tim H. Jakobsen,[b] Søren Skovstrup,[c]
Olivier Taboureau,[c, d] Tim Tolker-Nielsen,[b] Michael Givskov,[b, e] and Thomas E. Nielsen*[a, e]
Bacteria use small signaling molecules to communicate in
a process termed “quorum sensing” (QS), which enables the
coordination of survival strategies, such as production of viru-
lence factors and biofilm formation. In Gram-negative bacteria,
these signaling molecules are a series of N-acylated l-homoser-
ine lactones. With the goal of identifying non-native com-
pounds capable of modulating bacterial QS, a virtual library of
N-dipeptido l-homoserine lactones was screened in silico with
two different crystal structures of LasR. The 30 most promising
hits were synthesized on HMBA-functionalized PEGA resin and
released through an efficient acid-mediated cyclative release
mechanism. Subsequent screening for modulation of QS in
Pseudomonas aeruginosa and E. coli identified six moderately
strong activators. A follow-up library designed from the pre-
liminary derived structure–activity relationships was synthe-
sized and evaluated for their ability to activate the QS system
in this bacterium. This resulted in the identification of another
six QS activators (two with low micromolar activity) thus illumi-
nating structural features required for QS modulation.
Introduction
Pseudomonas aeruginosa is an opportunistic pathogen associ-
ated with a major part of hospital-acquired infections and lung
infections in cystic fibrosis patients.[1,2] Unfortunately, this bac-
terium shows growing resistance toward classical antibiotics.[3]
Thus, new strategies towards the treatment of infections of
P. aeruginosa are urgently needed.
polymeric matrix material that embeds the bacterial cells and
makes it difficult for the immune system (as well as antibiotics)
to eradicate the bacteria.[9,10] By modulating the QS system,
pathogenic bacteria can be sensitized to immune responses
and antibiotics.[11] Furthermore, as QS is not directly involved in
biological processes that are essential for bacterial survival,
modulation of QS does not impose selective pressure for the
evolution of resistant mutants.
P. aeruginosa uses a quorum sensing (QS) system to control
the formation of drug-resistant biofilm and to generate the vir-
ulence factors that are responsible for extensive tissue damage
in mammals.[2,4–8] The biofilm consists primarily of extracellular
QS is an intercellular communication process used in Gram-
positive and Gram-negative bacteria. The process is driven by
small signaling molecules (“autoinducers”) that allow the bac-
teria to monitor the size of their population and control behav-
iors that are beneficial for the group.[12,13] The most common
class of autoinducers in Gram-negative bacteria is N-acylated
l-homoserine lactones (AHLs). These signaling molecules are
continuously synthesized in low concentrations by a synthase
enzyme (LuxI-type synthase) and diffuse into the extracellular
fluid. The extracellular concentration of AHL increases as the
bacterial population grows, and once an intracellular concen-
tration threshold is reached, AHL binds to an intracellular re-
ceptor (LuxR-type transcriptional regulator), which functions as
a transcription factor. The AHL/LuxR complex dimerizes and
activates transcription of the lux gene cluster, including the
luxI-type gene, thus generating a positive feedback loop.[14]
The QS system of P. aeruginosa is complex; it comprises at
least two AHL molecules and three LuxR-type receptors (LasR,
RhlR, and QscR).[15] The LasR system upregulates the RhlR
system, whereas the QscR system represses the LasR and the
RhlR systems. The signaling molecule used in the LasR system
is N-(3-oxododecanoyl)-l-homoserine lactone (OdDHL); the
RhlR system uses N-butanoyl-l-homoserine lactone (BHL). Each
[a] Dr. M. R. Hansen, Dr. S. T. Le Quement, Prof. T. E. Nielsen
Department of Chemistry, Technical University of Denmark
Kemitorvet 201, 2800 Kgs. Lyngby (Denmark)
E-mail: ten@kemi.dtu.dk
[b] T. H. Jakobsen, Prof. T. Tolker-Nielsen, Prof. M. Givskov
Costerton Biofilm Center
Department of International Health, Immunology and Microbiology
Faculty of Health and Medical Sciences, University of Copenhagen
Blegdamsvej 3B, 2200 Copenhagen (Denmark)
[c] Dr. S. Skovstrup, Prof. O. Taboureau
Center for Biological Sequence Analysis
Department of Systems Biology, Technical University of Denmark
Søltofts Plads 221, 2800 Kgs. Lyngby (Denmark)
[d] Prof. O. Taboureau
UMR-S973, Molecules Therapeutics in Silico
Department of Life Sciences, University of Paris Diderot
5 rue Thomas Mann, 75013 Paris (France)
[e] Prof. M. Givskov, Prof. T. E. Nielsen
Singapore Centre on Environmental Life Sciences Engineering
Nanyang Technological University
60 Nanyang Drive, SBS-01N-27 (Singapore)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cbic.201300533.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2014, 15, 460 – 465 460

CHEMBIOCHEM
FULL PAPERS
www.chembiochem.org
of these ligand/receptor pairs constitutes a potential
target for QS modulation.[16,17]
Successful searches for QS modulators have cen-
tered around AHL structural analogues; these fall
into one of two categories: either the lactone moiety
or the acyl tail is varied.[16–18] So far, the synthesis of
AHL analogues with non-native head groups has pro-
vided only a few active modulators of the LuxR-type
proteins in P. aeruginosa.[19–25] However, studies on
AHL analogues with structural variations in the acyl
chain have led to several very potent LasR antago-
nists and agonists.[16,19,26–34] By simply changing the
length or degree of saturation of the alkyl chain, ago-
nists with nanomolar activity have been discovered.
Both agonists and antagonists are valuable chemical
tools for the study of QS pathways.[17]
Recently, crystal structures of the LasR protein
have allowed computational and more rational
design of new QS modulators.[26,35,36] In silico screen-
ing to identify structures that display strong binding
Scheme 1. A) Generation of a virtual library of HSL-containing tripeptides with the 20
canonical amino acids and l-homoserine lactone. B) Structures of OdDHL and TP-1.
interactions with the ligand-binding domain of LasR would
allow the selective synthesis of compounds with a high proba-
bility of either activating or inhibiting LasR. Thus, we set out to
investigate the QS modulating activity of N-dipeptido-l-homo-
serine lactones as AHL analogues. In this structure, the acyl
chain of the natural ligand is replaced with a short peptide
chain (Scheme 1). We envisioned that peptides of this type
could be easily prepared by high-throughput solid-phase syn-
thesis strategies and display favorable cell-penetrating proper-
ties by keeping the peptide chain short. This is critically impor-
tant for novel QS modulators as the LasR protein so far has
resisted stable isolation for biochemical screening assays. We
report an in silico screening study of a virtual combinatorial
peptide AHL library against the ligand-binding domain of LasR,
and solid-phase synthesis and biological evaluation of com-
pounds that modulate QS in vivo.
mainly contained amino acids with small, nonaromatic side
chains at the AA1 position, whereas aromatic side chains domi-
nated at this position for 3IX4 structure. This can be explained
by displacement of Arg61 in the 3IX4 structure compared to
the 2UV0 structure.
Table 1. Docking hits for the two LasR crystal structures.
Compound
Structure
PDB ID
Docking score
OdDHL
1
–
2UV0
2UV0
2UV0
2UV0
2UV0
2UV0
2UV0
2UV0
2UV0
2UV0
2UV0
2UV0
2UV0
3IX4
3IX4
3IX4
3IX4
3IX4
3IX4
3IX4
3IX4
3IX4
3IX4
3IX4
3IX4
3IX4
3IX4
3IX4
3IX4
3IX4
3IX4
3IX4
À7.34
À11.19
À11.06
À11.04
À10.85
À10.84
À10.35
À9.93
Phe-Gly-HSL
N-Ac-Phe-Asp-HSL
N-Ac-His-Cys-HSL
Tyr-Ser-HSL
2
3
4
5
6
7
8
9
Tyr-Gly-HSL
Trp-Ala-HSL
N-Ac-Phe-Ser-HSL
Phe-Pro-HSL
Phe-Ala-HSL
N-Ac-Phe-Gly-HSL
N-Ac-Phe-Ala-HSL
Phe-Asp-HSL
–
Trp-Tyr-HSL
Trp-Phe-HSL
Lys-Trp-HSL
N-Ac-Trp-Trp-HSL
Trp-Lys-HSL
Tyr-Tyr-HSL
À9.92
À9.81
Results and Discussion
10
11
12
TP-1
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
À9.76
In silico screening
À9.65
À9.54
A virtual library based on l-homoserine lactone (HSL) and the
20 canonical amino acids (AA) as building blocks was generat-
ed (Scheme 1).
À12.63
À12.43
À12.16
À12.12
À12.03
À12.01
À11.78
À11.44
À11.42
À11.42
À11.40
À11.33
À11.09
À11.08
À10.76
À10.75
À10.71
À10.66
À10.64
By varying the two amino acids, 400 different structures
were obtained. These were investigated both as the free
amines and as the N-acetylated analogues. The docking was
performed to the ligand-binding domain of LasR obtained
from two crystal structures (PDB IDs: 2UV0 and 3IX4). As refer-
ence, docking of the cocrystallized ligand OdDHL was included
in the screen against 2UV0, and the triphenyl mimic TP-1 was
used in the screen against 3IX4.
Phe-Trp-HSL
Met-Trp-HSL
Phe-Tyr-HSL
Trp-Ile-HSL
Trp-His-HSL
Trp-Leu-HSL
N-Ac-Phe-Met-HSL
Gln-Trp-HSL
Trp-Val-HSL
Met-Tyr-HSL
N-Ac-Phe-Tyr-HSL
N-Ac-His-Tyr-HSL
Generally, the docking scores observed for the virtual library
members were high compared to OdDHL for the 2UV0 struc-
ture, and similar to the score of TP-1 for the 3IX4 structure
(Table 1). Structures with a free N terminus tended to give the
highest docking scores. Hits for the 2UV0 crystal structure
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2014, 15, 460 – 465 461

CHEMBIOCHEM
FULL PAPERS
www.chembiochem.org
The docked compounds generally displayed two different
binding modes (Figure 1): in one the HSL head group was
placed similarly to the natural ligand and interacted with
Tyr93, Asp73, and Trp60 (Figure 1C and D); in the other the
HSL group was shifted to a close-by subpocket, thereby facili-
tating interactions between the terminal amine and Asp73
(Figure 1A and B). In both cases, an aromatic group was locat-
ed in a pocket formed by generally hydrophobic residues (e.g.,
Gly38, Leu40, Tyr47, Val76, Cys79, Leu125, and Gly126) thereby
forming p-stacking interactions with Tyr47 (for a detailed over-
view of the LasR ligand binding site please see Skovstrup
et al.[37]). Interestingly, there was no overlap of the hits be-
tween the different crystal structures. Of the 40 best hits, 30
were selected for synthesis and biological evaluation.
Chemistry
With the aim of synthesizing a peptide library, it was decided
to develop a solid-phase synthesis strategy based on amino-
functionalized PEG-based resins (PEGA800 and ChemMatrix). The
resin was prepared for synthesis by attachment of the HMBA
linker by using an N,N,N’,N’-tetramethyl-O-(benzotriazol-1-yl)-
uronium tetrafluoroborate (TBTU)-mediated amide coupling
procedure (Scheme 2).[37] The first step in the synthesis was
Scheme 2. Synthesis of solid-supported tripeptides. Reagents and condi-
tions: a) N-Fmoc-O-Trt-l-Hse-OH, MSNT, MeIm, CH2Cl2, 1 h (ꢁ2); b), d), f) 20%
piperidine (DMF), 20 min; c), e) Fmoc-AA-OH, TBTU, NEM, DMF, 2 h; g) 20%
Ac2O (DMF), 30 min.
coupling of N-Fmoc-O-Trt-HSe-OH by using an optimized
MSNT protocol for ester formation.[38] Fmoc deprotection with
20% piperidine in DMF was followed by two rounds of TBTU-
mediated couplings with selected AAs to give 33. For the N-
acetylated analogues the resin was treated with 20% Ac2O in
DMF to give 34.
Solid-supported peptides with an ester-linked O-Trt-l-homo-
serine residue in the first position would, after Trt deprotec-
tion, allow acid-mediated cyclative release of the desired prod-
uct from the solid phase. A similar strategy by Blackwell and
Figure 1. A) and C) 3D binding conformations of 31 and 24 in the LasR bind-
ing pocket from the 3IX4 crystal structure. B) and D) 2D ligand–protein inter-
actions. Pink arrows represent hydrogen bonds to a protein side chain;
green lines represents p–p interactions.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2014, 15, 460 – 465 462

CHEMBIOCHEM
FULL PAPERS
www.chembiochem.org
co-workers used TFA vapor to release HSLs from an 3,5-di-
chloro-4-hydroxybenzoic acid linker attached to a planar solid
support.[31] We needed a more robust linker construct to meet
this particularly challenging part of the synthetic process to
obtain preparative amounts sufficient for HPLC purification
and scale-up. As discussed below, the HMBA linker proved very
well suited for this synthetic approach, thus allowing multistep
peptide coupling and deprotection reactions.
Table 2. Optimization study for selective Trt deprotection (selected re-
sults).
For standard synthesis, beads, various solvent systems, TFA
concentrations, and reaction times were investigated in a
model system (Scheme 3, 35) in order to find conditions meet-
Aqueous TFA [%]
treact [min]
Ratio 37:38[a,b]
Ratio 39:40[a]
1
2
3
4
5
6
7
8
5
5
10
30
10
30
10
30
10
30
>15:1
>15:1
>15:1
10:1
10:1
7:1
5:1
16:1
1:1
10
10
20
20
50
50
1:3
<1:15
<1:15
<1:15
<1:15
5:1
3:1
[a] Determined by RP-HPLC. [b] Corrected for UV absorption at 215 nm.
product. The developed protocol was applied to the library of
solid-supported tripeptides (Scheme 4). After the deprotection
step, the resin was washed once with acetonitrile, followed by
treatment with 95% aqueous TFA for 2 h to provide the
desired product by cyclative release. The selective Trt depro-
tection proved to be somewhat substrate dependent, albeit
bringing about cyclative release in synthetically useful yields
(26–95%; see the Supporting Information for details). The syn-
thesized compounds were easily purified by preparative RP-
HPLC prior to biological screening.
Scheme 3. Model system for the identification of optimal conditions for se-
lective Trt-deprotection.
ing these requirements. We expected that acidic treatment
with TFA in low concentrations would be sufficient to remove
the Trt protecting group from the homoserine (Scheme 3,
lower pathway). The released Trt alcohol 37 could then be re-
moved by filtration; subsequent treatment of the resin with
a high concentration of TFA would then provide 38. For this
strategy to be successful, it was important than the Trt remov-
al could be performed at TFA concentrations that did not
affect cyclative release of the product as this would result in
loss of product during filtration (Scheme 3, upper pathway).
Following TFA treatment, the filtrate was analyzed by HPLC
to determine the ratio between the Trt alcohol 37 and the
released HSL 38. To investigate whether Trt deprotection was
complete, the resin was subsequently subjected to aqueous
sodium hydroxide to release the substrate from the resin and
to determine the ratio between the carboxylic acids
39 and 40 that result from cleavage of 35 and 36,
respectively.
Biological screening
To determine their QS modulating activity, the synthesized
compounds were screened by means of cell-based LasR inhibi-
tion and activation assays. Six compounds were identified as
LasR activators with EC50 values in the low micromolar range
(Table 3); none of the compounds displayed LasR-inhibiting ac-
tivity. The most active compounds were 19 and 20 (EC50 4 and
6 mm, respectively). A clear trend was observed in the struc-
tures of the active compounds: an amino acid with an aromat-
ic side chain in the first position and a free N terminus were
important for the activation of LasR.
Selected results from the screening of reaction
conditions are reported in Table 2. The best results
were observed when using water as solvent, and the
optimal conditions were concluded to be treatment
with 20% aqueous TFA for 10 min (entry 5). Lower
TFA concentrations resulted in incomplete removal of
the Trt group, whereas higher concentrations result-
Scheme 4. Trt deprotection and cyclative release of library compounds. Reagents and
ed in both Trt deprotection and cyclative release of conditions: a) 20% aqueous TFA, 10 min; b) 95% aqueous TFA, 2 h.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2014, 15, 460 – 465 463

CHEMBIOCHEM
FULL PAPERS
www.chembiochem.org
Table 3. EC50 values of the most active N-dipeptido l-homoserine lac-
tones showing in vitro QS activation (LasR agonism) in an E. coli reporter
strain.
Compound
EC50 [mm]
Compound
EC50 [mm]
1
2
3
4
OdDHL
14
18
0.009
20
16
5
6
7
20
21
28
6
12
25
19
4
Based on the initial screening results, follow-up compounds
were designed and synthesized. Analogues of the QS agonists
(Table 3) were explored by replacing tryptophan with unnatu-
ral aromatic and heteroaromatic amino acids, for example 3-(3-
thienyl)alanine, 3-(3-benzothienyl)alanine and 3-(1-naphthyl)al-
anine. N-acetylated analogues of 19 and 20 and analogues in
which the amino acid in the AA2 position was varied with both
canonical and unnatural amino acids were also investigated. Fi-
nally, the library also included diastereoisomeric variants of 19
and 20, synthesized from the corresponding d-amino acids
(synthesis protocols in Schemes 2 and 4). By subjecting the
ChemMatrix resin to extensive washings with water, DMF, and
dichloromethane after the Trt deprotection step, products (in-
cluding 41–46) were obtained with high purity (>95%) after
the cyclative release, thereby illustrating the usefulness of the
overall synthesis route, and the advantage of the cyclization–
cleavage strategy. The compounds were evaluated in a LasR
activation assay (Table 4), and the screening identified another
Scheme 5. Structures of active compounds from follow up studies.
and thereby identified several compounds with potential for
modulation of QS in P. aeruginosa. To assist these efforts, we
developed an efficient protocol for solid-phase synthesis of N-
peptido-homoserine lactones on ChemMatrix resin. All com-
pounds were screened for their effects on activation and inhib-
ition of LasR in cell-based assays. The structures of hits identi-
fied in the LasR activation assay were used to design further
sets of analogues, some of which were highly potent (low mi-
cromolar EC50 for QS activation). Intense research efforts in aca-
demic laboratories have recently been directed towards QS in
Gram-negative bacteria, and several non-native autoinducers
capable of antagonizing or agonizing QS in these organisms
have been reported. The dipeptido homoserine lactones in this
study are among the most potent agonists reported to date.
Table 4. EC50 values of structurally related N-dipeptido l-homoserine lac-
tones showing in vitro QS activation (LasR agonism) in an E. coli reporter
strain.
Compound
EC50 [mm]
Compound
EC50 [mm]
1
2
3
41
42
43
30
6
5
4
6
44
45
46
20
1.5
1.3
six LasR activators (including 45 and 46) with improved EC50
values compared to the initial library (Scheme 5). Neither N-
acetylated analogues nor compounds containing d-amino acid
residues showed any activity.
Based on the screening results, it is concluded that a large
aromatic amino acid in the AA1 position is favored for LasR ac-
tivation. Such amino acids include tryptophan and the unnatu-
ral benzothienyl- and naphthyl-substituted alanines. When the
AA1 position was occupied by one of these amino acids, com-
pounds with various functional groups in the side chain of AA2
displayed activating properties; the most potent were those
with a small heterocyclic motif at this position.
Experimental Section
In silico screening: The LasR crystal structures (PDB IDs: 2UV0 and
3IX4) were prepared for molecular docking by the Protein Prepara-
tion Wizard of Maestro software.[39] (This involves the addition of
hydrogen atoms and optimization of the hydrogen bonding net-
works throughout the protein structure.) Only one protein mono-
mer was retained (chain E; the four monomers are, however, very
similar around the ligand), and all water molecules were removed.
The peptide library was manually created and docked into the two
LasR ligand binding sites by Glide (Maestro) by using SP precision
and default settings, but with rotatable hydroxyl groups.[40]
Conclusions
We screened in silico a virtual library of N-dipeptido l-homo-
serine lactones against two different crystal structures of LasR,
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2014, 15, 460 – 465 464

CHEMBIOCHEM
FULL PAPERS
www.chembiochem.org
[15] D. S. Wade, M. W. Calfee, E. R. Rocha, E. A. Ling, E. Engstrom, J. P. Cole-
man, E. C. Pesci, J. Bacteriol. 2005, 187, 4372–4380.
[16] M. E. Mattmann, H. E. Blackwell, J. Org. Chem. 2010, 75, 6737–6746.
[17] W. R. J. D. Galloway, J. T. Hodgkinson, S. D. Bowden, M. Welch, D. R.
Spring, Chem. Rev. 2011, 111, 28–67.
[18] G. D. Geske, J. C. O’Neill, H. E. Blackwell, Chem. Soc. Rev. 2008, 37, 1432–
1447.
[19] L. Passador, K. D. Tucker, K. R. Guertin, M. P. Journet, A. S. Kende, B. H.
Iglewski, J. Bacteriol. 1996, 178, 5995–6000.
Chemistry: Experimental details for the synthesis of N-dipeptido l-
homoserine lactones are provided in the Supporting Information.
Reporter strains: Two bacterial reporter strains were used for
measuring modulation of QS. For measuring antagonist activity we
used a P. aeruginosa PAO1 strain harboring a plasmid pMHLAS,[41]
which contain PlasB-gfp(ASV) and Plac-lasR fusions divergently and
separately transcribed. For measuring agonist activity an E. coli
MT102 strain harboring the same plasmid was used.[41] The strains
were grown in ABT minimal medium (B medium with thiamine
(2.5 mgLÀ1))[42] supplemented with A10 (10%), glucose (0.5%, w/v),
and Casamino acids (0.5%, w/v) for 16 h at 308C with shaking
(180 rpm).
[20] T. Ishida, T. Ikeda, N. Takiguchi, A. Kuroda, H. Ohtake, J. Kato, Appl. Envi-
ron. Microbiol. 2007, 73, 3183–3188.
[21] K. M. Smith, Y. Bu, H. Suga, Chem. Biol. 2003, 10, 81–89.
[22] K. M. Smith, Y. Bu, H. Suga, Chem. Biol. 2003, 10, 563–571.
[23] G. J. Jog, J. Igarashi, H. Suga, Chem. Biol. 2006, 13, 123–128.
[24] C. Kim, J. Kim, H.-Y. Park, J.-H. Lee, H.-J. Park, C. K. Kim, J. Yoon, Appl. Mi-
crobiol. Biotechnol. 2009, 83, 1095–1103.
[25] L. Y. W. Lee, T. Hupfield, R. L. Nicholson, J. T. Hodgkinson, X. Su, G. L.
Thomas, G. P. C. Salmond, M. Welch, D. R. Spring, Mol. BioSyst. 2008, 4,
505–507.
[26] M. J. Bottomley, E. Muraglia, R. Bazzo, A. Carfi, J. Biol. Chem. 2007, 282,
13592–13600.
[27] C. M. Oliver, A. L. Schaefer, E. P. Greenberg, J. R. Sufrin, J. Med. Chem.
2009, 52, 1569–1575.
Biological assays: The biological assays were conducted in 96-
wells microtiter dish (Black Isoplate, PerkinElmer) as previously de-
scribed.[43] Growth and green fluorescent protein (GFP) expression
were monitored by using a Victor X4 multilabel plate reader (Perki-
nElmer; 348C, measurement every 15 min over 14 h). Growth was
measured as absorbance at 450 nm. GFP expression was measured
as fluorescence at lex =485 nm and lem =535 nm. The EC50 values
were calculated from curves showing the GFP expression.
[28] N. Amara, R. Mashiach, D. Amar, P. Krief, S. A. Spieser, M. J. Bottomley, A.
Aharoni, M. M. Meijler, J. Am. Chem. Soc. 2009, 131, 10610–10619.
[29] G. D. Geske, J. C. O’Neill, D. M. Miller, M. E. Mattmann, H. E. Blackwell, J.
Am. Chem. Soc. 2007, 129, 13613–13625.
Acknowledgements
[30] G. D. Geske, M. E. Mattmann, H. E. Blackwell, Bioorg. Med. Chem. Lett.
2008, 18, 5978–5981.
[31] T. Praneenararat, G. D. Geske, H. E. Blackwell, Org. Lett. 2009, 11, 4600–
4603.
Financial support was provided by the DSF Center for Antimicro-
bial Research. M.G. is grateful to Novo Seeds for a Pre-seed grant.
[32] D. M. Stacy, S. T. Le Quement, C. L. Hansen, J. W. Clausen, T. Tolker-Niel-
sen, J. W. Brummond, M. Givskov, T. E. Nielsen, H. E. Blackwell, Org.
Biomol. Chem. 2013, 11, 938–954.
[33] M. E. Mattmann, P. M. Shipway, N. J. Heth, H. E. Blackwell, ChemBioChem
2011, 12, 942–949.
Keywords: cyclative release · homoserine lactones · peptides ·
Pseudomonas aeruginosa
· quorum sensing · solid-phase
synthesis
[34] C. Kim, J. Kim, H.-Y. Park, H.-J. Park, C. K. Kim, J. Yoon, J.-H. Lee, Mol.
Cells 2009, 28, 447–453.
[35] Y. Zou, S. K. Nair, Chem. Biol. 2009, 16, 961–970.
[36] S. Skovstrup, S. T. Le Quement, T. Hansen, T. H. Jakobsen, M. Harmsen, T.
Tolker-Nielsen, T. E. Nielsen, M. Givskov, O. Taboureau, ChemMedChem
2013, 8, 157–163.
[37] R. Knorr, A. Trzeciak, W. Bannwarth, D. Gillessen, Tetrahedron Lett. 1989,
30, 1927–1930.
[38] B. Blankemeyer-Menge, M. Nimtz, R. Frank, Tetrahedron Lett. 1990, 31,
1701–1704.
[39] Maestro, Version 9.0, Schçdinger LCC, New York, NY, 2009.
[40] Glide, Version 5.5, Schrçdinger LLC, New York, NY, 2009.
[41] M. Hentzer, K. Riedel, T. B. Rasmussen, A. Heydorn, J. B. Andersen, M. R.
Parsek, S. A. Rice, L. Eberl, S. Molin, N. Høiby, S. Kjelleberg, M. Givskov,
Microbiology 2002, 148, 87–102.
[42] D. J. Clark, O. Maaloe, J. Mol. Biol. 1967, 23, 99–112.
[43] T. Bjarnsholt, M. van Gennip, T. H. Jakobsen, L. D. Christensen, P. Ø.
Jensen, M. Givskov, Nat. Protoc. 2010, 5, 282–293.
[1] T. R. de Kievit, B. H. Iglewski, Infect. Immun. 2000, 68, 4839–4849.
[2] M. E. Hibbing, C. Fuqua, M. R. Parsek, S. B. Peterson, Nat. Rev. Microbiol.
2010, 8, 15–25.
[3] O. Ciofu, L. F. Mandsberg, H. Wang, N. Høiby, FEMS Immunol. Med. Mi-
crobiol. 2012, 65, 215–225.
[4] T. Bjarnsholt, M. Givskov, Anal. Bioanal. Chem. 2007, 387, 409–414.
[5] M. Cooley, S. R. Chhabra, P. Williams, Chem. Biol. 2008, 15, 1141–1147.
[6] T. R. de Kievit, B. H. Iglewski, Infect. Immun. 2000, 68, 4839–4849.
[7] C. N. Wilder, G. Allada, M. Schuster, Infect. Immun. 2009, 77, 5631–5639.
[8] L. Passador, J. M. Cook, M. J. Gambello, L. Rust, B. H. Iglewski, Science
1993, 260, 1127–1130.
[9] D. G. Davies, M. R. Parsek, J. P. Pearson, B. H. Iglewski, J. W. Costerton,
E. P. Greenberg, Science 1998, 280, 295–298.
[10] B. Spellberg, R. Guidos, D. Gilbert, J. Bradley, H. W. Boucher, W. M.
Scheld, J. G. Bartlett, J. Edwards Jr., Clin. Infect. Dis. 2008, 46, 155–164.
[11] W. R. J. D. Galloway, J. T. Hodgkinson, S. Bovvden, M. Welch, D. R. Spring,
Trends Microbiol. 2012, 20, 449–458.
[12] B. L. Bassler, R. Losick, Cell 2006, 125, 237–246.
[13] W.-L. Ng, B. L. Bassler, Annu. Rev. Genet. 2009, 43, 197–222.
[14] C. Fuqua, E. P. Greenberg, Nat. Rev. Mol. Cell Biol. 2002, 3, 685–695.
Received: August 17, 2013
Published online on January 16, 2014
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2014, 15, 460 – 465 465