C4-alkoxy-HPD: A potent class of synthetic modulators surpassing nature in AI-2 quorum sensing

Article abstract of DOI:10.1021/ja305532y

Bacteria have developed cell-to-cell communication mechanisms, termed quorum sensing (QS), that regulate bacterial gene expression in a cell population-dependent manner. Autoinducer-2 (AI-2), a class of QS signaling molecules derived from (4S)-4,5-dihydroxy-2,3-pentanedione (DPD), has been identified in both Gram-negative and Gram-positive bacteria. Despite considerable interest in the AI-2 QS system, the biomolecular communication used by distinct bacterial species still remains shrouded. Herein, we report the synthesis and evaluation of a new class of DPD analogues, C4-alkoxy-5-hydroxy-2, 3-pentanediones, termed C4-alkoxy-HPDs. Remarkably, two of the analogues were more potent QS agonists than the natural ligand, DPD, in Vibrio harveyi. The findings presented extend insights into ligand-receptor recognition/signaling in the AI-2 mediated QS system.

Full text of DOI:10.1021/ja305532y

Communication
pubs.acs.org/JACS
C4-Alkoxy-HPD: A Potent Class of Synthetic Modulators Surpassing
Nature in AI‑2 Quorum Sensing
Kyoji Tsuchikama, Jie Zhu, Colin A. Lowery, Gunnar F. Kaufmann, and Kim D. Janda*
The Skaggs Institute for Chemical Biology and Departments of Chemistry, Immunology & Microbial Science, and Worm Institute for
Research and Medicine (WIRM), The Scripps Research Institute, 10550 North Torrey Pines Road, La Jolla, California 92037, United
States
S
* Supporting Information
ABSTRACT: Bacteria have developed cell-to-cell com-
munication mechanisms, termed quorum sensing (QS),
that regulate bacterial gene expression in a cell population-
dependent manner. Autoinducer-2 (AI-2), a class of QS
signaling molecules derived from (4S)-4,5-dihydroxy-2,3-
pentanedione (DPD), has been identified in both Gram-
negative and Gram-positive bacteria. Despite considerable
interest in the AI-2 QS system, the biomolecular
communication used by distinct bacterial species still
remains shrouded. Herein, we report the synthesis and
evaluation of a new class of DPD analogues, C4-alkoxy-5-
hydroxy-2,3-pentanediones, termed C4-alkoxy-HPDs. Re-
markably, two of the analogues were more potent QS
agonists than the natural ligand, DPD, in Vibrio harveyi.
The findings presented extend insights into ligand−
receptor recognition/signaling in the AI-2 mediated QS
system.
Figure 1. A plausible equilibrium of DPD in aqueous solution.
Figure 2. Structures of DPD analogues for the modulation of AI-2
based QS.
acteria have developed unique systems to coordinate their
Bbehavior in a cell density-dependent manner, a process
that has been termed quorum sensing (QS).1,2 QS is mediated
by a series of signal production/secretion/response events
utilizing diffusible signaling molecules called autoinducers.
When the bacterial population, and thus concentration of
autoinducer, reaches a threshold level, gene expression is
synchronized in a concerted manner. In many species, this
process often results in pathogenic events such as biofilm
formation and virulence factor production. Therefore, the
modulation of QS has emerged as a potential therapeutic
approach for combating microbial infection.3−6
Autoinducer-2 (AI-2), one of the classes of autoinducers, is
produced by both Gram-negative and Gram-positive bacterial
species.7−10 AI-2 signaling molecules share a single common
precursor, (4S)-4,5-dihydroxy-2,3-pentanedione (DPD), that
exists in a multiplexed equilibrium between its linear and cyclic
forms (Figure 1).11 Collectively, these findings have led to the
hypothesis that bacteria use AI-2 as a “universal language” for
interspecies monitoring as well as intraspecies communica-
tion.10
minor structural alterations of DPD can have a significant
impact on the function of AI-2 in QS mediated processes.
However, even with these studies, a clear gap still exists
between our understanding of DPD’s complex chemical
equilibrium and leveraging structural information for agonist/
antagonist design between bacterial species.11
To help better define the chemical determinants critical
within the sphere of the AI-2 QS system, we were drawn to the
QS activity seen with the C1-alkyl derivatives, particularly their
oscillatory agonist/antagonist activity based upon the addition
of single methylene units. Thus, we designed and synthesized a
new group of synthetic DPD analogues, C4-alkoxy-5-hydroxy-
2,3-pentanedione (C4-alkoxy-HPD, Figure 2). In this commu-
nication, we present data demonstrating that these C4-alkoxy-
HPDs are capable of more potent QS agonism than the natural
signal used in Vibrio harveyi.
The synthesis of the C4-alkoxy-HPDs is shown in Scheme 1,
which began from readily available racemic 1-((tert-
butyldimethylsilyl)oxy)pent-3-yn-2-ol (1).19 Alkyl groups
(methyl, ethyl, propyl, hexyl, and benzyl) were installed on
the C4-hydroxy group under standard conditions of ether
synthesis; however, incomplete consumption of the starting
material and significant degradation of substrate and products
Our group and others have reported upon the synthesis and
potency of a series of DPD analogues including C1-alkyl-
DPDs,12−16 2,5-dihydroxy-2-methylcyclopentanones
(DHMPs),17 and C5-methyl-DPDs18 as QS modulators
(Figure 2). In sum, these reports have provided evidence that
Received: June 7, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja305532y | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
dependent on the length of the C4-hydroxy alkylation. Thus,
while C4-MeO-HPD 4a presented modest QS activity,
remarkably, C4-EtO-HPD 4b and C4-PrO-HPD 4c appeared
to be better ligands than DPD. Notably, analogue 4c, a
racemate, was almost 10-fold more potent than the natural
ligand DPD with a submicromolar EC50 value of 0.15 0.03
μM, whereas the EC50 value of DPD was 1.07 0.06 μM under
our assay conditions (Table 1).24 It is reasonable to speculate
that a single enantiomer of analogue 4c may provide even
greater enhanced activity.
Scheme 1. Synthesis of C4-Alkoxy-HPDs 4a−e
a
Table 1. Summary of QS Modulation by C4-Alkoxy-HPDs
compound
(C4-RO-
HPD)
EC50 in V. harveyi assay
β-galactosidase activity in S.
Typhimurium (mean SD, %)
b
(mean SD, μM)
(4S)-DPD
1.07 0.06
7.60 0.45
0.79 0.05
0.15 0.03
100.0 2.9
R = Me (4a)
R = Et (4b)
R = Pr (4c)
R = Hex (4d)
R = Bn (4e)
<5
<5
<5
<5
were observed and propargyl ethers 2a−e were obtained in low
to moderate yields. Subsequently, ruthenium-catalyzed oxida-
tion of ethers 2a−e was accomplished to give diketones 3a−e.
Finally, precursors 3a−e were solubilized in dilute sulfuric acid
to afford C4-alkoxy-HPDs 4a−e in quantitative yield. NMR
analysis of compounds 4a−e showed that they exist as a
mixture of linear/cyclic forms in a ratio of approximately 10:90.
Given the reported linear/cyclic ratio of DPD (approximately
20:80),11,20,21 these results illustrate that installation of
“capping” substituents on the C4-hydroxy group within DPD
does not significantly impact cyclization.
The obtained C4-alkoxy-HPDs 4a−e were evaluated for the
modulation of QS signaling using two established reporter
assays: bioluminescence emission in V. harveyi22 and induction
of β-galactosidase expression in Salmonella enterica serovar
Typhimurium.23 First, we examined the effects of the synthetic
analogues on bioluminescence of the V. harveyi strain MM32
(ATCC BAA-1117, ΔluxS, ΔluxN), a strain unable to produce
its own DPD due to the lack of the LuxS synthase and also
lacking the acylhomoserine lactone receptor LuxN. The use of
this strain allows for the measurement of QS responses only via
the AI-2 pathway in the presence of exogenous autoinducers.
Cultures were incubated in the presence of DPD or each
synthetic analogue at varying concentrations (20−0.02 μM).
Analogues 4a−c induced bioluminescence emission, whereas
C4-HexO-HPD 4d and C4-BnO-HPD 4e exhibited weak or no
induction at 20 μM (Figure 3). The induction was strongly
c
N.D.
c
N.D.
<5
a
b
All assays were performed in triplicate. All EC50 values were
c
calculated from the sigmoidal curves obtained in Figure 3. Assays
were performed in the presence of 50 μM compound. SD, standard
deviation; N.D., not determined.
To our knowledge, these compounds are the first examples
of synthetic DPD analogues that are better at initiating an AI-2
mediated process than the natural ligand.25 In the AI-2 based
QS of V. harveyi, it is a borate diester derived from the cyclic
isomer (2S,4S)-THMF that the periplasmic AI-2 receptor
protein LuxP recognizes (Figure 1), which triggers a protein
phosphorylation cascade that ultimately results in altered gene
expression.26,27 With this mechanism in mind, the slightly
higher ratio of cyclic forms in the equilibrium of C4-alkoxy-
HPDs (Scheme 1) may help drive their enhanced activity;
however, this is unlikely to be the sole contributing factor given
the varying agonist activity of the panel of C4-alkoxy-HPDs
examined. We surmise that LuxP has an additional structural
hydrophobic modulatory site, which analogue 4c has
advantageously accessed to produce superior QS activity.
Indeed, a docking model of LuxP and analogue 4c appears to
support our speculation (Figure S5).28 On the other hand,
congeners with large hydrophobic groups such as the hexyl
(4d) and benzyl (4e) derivatives may be too bulky for a
productive interaction.
To continue our studies, analogues 4a−e were screened for
QS modulation in S. Typhimurium strain Met844 (ΔluxS, lsr-
lacZ fusion), a strain incapable of producing its own DPD.29
The lsr-lacZ fusion encodes β-galactosidase under the AI-2-
regulated lsr promoter, and enables the monitoring of AI-2-
dependent lsr expression based on the residual β-galactosidase
activity. Assays were performed using the analogues at 50 μM
in the absence of DPD (agonist assay) or in the presence of 50
μM DPD (antagonist assay). These experiments revealed no
significant agonist or antagonist activity (Table 1 and Figure
S3). This is in contrast to previous reports on a panel of C1-
alkyl-DPD analogues, in which several compounds exhibited
antagonist effects in AI-2 based QS of S. Typhimurium.12,15,16
In the AI-2 QS circuit of S. Typhimurium, DPD is
transported into the cytoplasm and the C5-hydroxy group
within the linear form is phosphorylated by the kinase LsrK to
Figure 3. Bioluminescence of V. harveyi strain MM32 (ΔluxN, ΔluxS)
in the presence of the analogues. Bioluminescence was normalized to
cell density. All assays were performed in triplicate, and error bars
represent standard deviation. All curves were fit to variable-slope
sigmoidal curves using GraphPad Prism software. BnO-HPD 4e
showed no agonistic effect at 20 μM.
B
dx.doi.org/10.1021/ja305532y | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
afford the chemical species thought to be responsible for
regulating gene expression.30,31 A recent report highlights how
LsrK also phosphorylates a series of unnatural C1-alkyl-DPDs
to provide active modulators of QS acting on the lsr operon.15
For the purpose of further dissecting the lack of analogue
activity, LsrK-mediated phosphorylation assays of C4-alkoxy-
HPDs were performed (Figures 4 and S4). As anticipated based
REFERENCES
■
(1) Engebrecht, J.; Nealson, K.; Silverman, M. Cell 1983, 32, 773−81.
(2) Fuqua, W. C.; Winans, S. C.; Greenberg, E. P. J. Bacteriol. 1994,
176, 269−75.
(3) Suga, H.; Smith, K. M. Curr. Opin. Chem. Biol. 2003, 7, 586−91.
(4) Bjarnsholt, T.; Jensen, P. O.; Rasmussen, T. B.; Christophersen,
L.; Calum, H.; Hentzer, M.; Hougen, H. P.; Rygaard, J.; Moser, C.;
Eberl, L.; Hoiby, N.; Givskov, M. Microbiology 2005, 151, 3873−80.
(5) Lowery, C. A.; Salzameda, N. T.; Sawada, D.; Kaufmann, G. F.;
Janda, K. D. J. Med. Chem. 2010, 53, 7467−89.
(6) Zhu, P.; Li, M. Curr. Med. Chem. 2012, 19, 174−86.
(7) Bassler, B. L.; Wright, M.; Silverman, M. R. Mol. Microbiol. 1994,
13, 273−86.
(8) Bassler, B. L.; Greenberg, E. P.; Stevens, A. M. J. Bacteriol. 1997,
179, 4043−5.
(9) Vendeville, A.; Winzer, K.; Heurlier, K.; Tang, C. M.; Hardie, K.
R. Nat. Rev. Microbiol. 2005, 3, 383−96.
(10) Waters, C. M.; Bassler, B. L. Annu. Rev. Cell. Dev. Biol. 2005, 21,
319−46.
Figure 4. LsrK-mediated phosphorylation of DPD and C4-alkoxy-
HPDs.
(11) Globisch, D.; Lowery, C. A.; McCague, K. C.; Janda, K. D.
Angew. Chem., Int. Ed. 2012, 51, 4204−8.
(12) Lowery, C. A.; Park, J.; Kaufmann, G. F.; Janda, K. D. J. Am.
Chem. Soc. 2008, 130, 9200−1.
(13) Lowery, C. A.; Abe, T.; Park, J.; Eubanks, L. M.; Sawada, D.;
Kaufmann, G. F.; Janda, K. D. J. Am. Chem. Soc. 2009, 131, 15584−5.
(14) Ganin, H.; Tang, X.; Meijler, M. M. Bioorg. Med. Chem. Lett.
2009, 19, 3941−4.
on their QS inactivity, none of the analogues were
phosphorylated by LsrK, suggesting that the C4-hydroxy
group in DPD is a critical factor in the mechanism of LsrK-
mediated phosphorylation.
In conclusion, we have detailed the synthesis of a panel of
DPD derivatives, C4-alkoxy-HPDs. These DPD analogues were
evaluated for the modulation of bacterial QS in established
assays, and unanticipated, yet remarkable, effects on AI-2 QS
circuits were observed in a species-dependent manner. Indeed,
the present work describes the discovery of the most potent
agonist known for V. harveyi QS signaling. While potent
antagonistic activity was not uncovered with this series of
structures, the nuance between agonist and antagonist is a fine
line and thus our findings should enable future chemical
structure-based approaches for accessing such alternative
activity. In a broader sense, the present work, coupled with
previous findings with C1-alkyl-DPD analogues, highlights how
manipulation of the DPD scaffold can provide valuable tools for
in-depth studies of the ligand−receptor interactions making up
AI-2-mediated QS.
(15) Roy, V.; Smith, J. A.; Wang, J.; Stewart, J. E.; Bentley, W. E.;
Sintim, H. O. J. Am. Chem. Soc. 2010, 132, 11141−50.
(16) Gamby, S.; Roy, V.; Guo, M.; Smith, J. A.; Wang, J.; Stewart, J.
E.; Wang, X.; Bentley, W. E.; Sintim, H. O. ACS Chem. Biol. 2012, 7,
1023−30.
(17) Tsuchikama, K.; Lowery, C. A.; Janda, K. D. J. Org. Chem. 2011,
76, 6981−9.
(18) Rui, F.; Marques, J. C.; Miller, S. T.; Maycock, C. D.; Xavier, K.
B.; Ventura, M. R. Bioorg. Med. Chem. 2012, 20, 249−56.
(19) Mohamed, Y. A. M.; Inagaki, F.; Takahashi, R.; Mukai, C.
Tetrahedron 2011, 67, 5133−41.
(20) Meijler, M. M.; Hom, L. G.; Kaufmann, G. F.; McKenzie, K. M.;
Sun, C.; Moss, J. A.; Matsushita, M.; Janda, K. D. Angew. Chem., Int.
Ed. 2004, 43, 2106−8.
(21) Semmelhack, M. F.; Campagna, S. R.; Federle, M. J.; Bassler, B.
L. Org. Lett. 2005, 7, 569−72.
(22) Taga, M. E.; Xavier, K. B. Curr. Protoc. Microbiol. 2011, Chapter
1, No. Unit1C.1.
(23) Griffith, K. L.; Wolf, R. E., Jr. Biochem. Biophys. Res 2002, 290,
397−402.
(24) A similar EC50 value of DPD was reported: Winzer, K.; Hardie,
K. R.; Burgess, N.; Doherty, N.; Kirke, D.; Holden, M. T.; Linforth, R.;
Cornell, K. A.; Taylor, A. J.; Hill, P. J.; Williams, P. Microbiology 2002,
148, 909−22.
(25) (4S,5R)-4,5-Dihydroxy-2,3-hexanedione (ref 18) was the most
potent agonist of the synthetic analogues reported to date in the AI-2
based QS of V. harveyi, although it showed approximately 10-fold
higher EC50 value than that of DPD.
(26) Neiditch, M. B.; Federle, M. J.; Pompeani, A. J.; Kelly, R. C.;
Swem, D. L.; Jeffrey, P. D.; Bassler, B. L.; Hughson, F. M. Cell 2006,
126, 1095−108.
ASSOCIATED CONTENT
* Supporting Information
Synthetic protocols, assay methods, additional biological assays,
and characterization of new compounds. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
kdjanda@scripps.edu
Notes
■
The authors declare no competing financial interest.
(27) Chen, X.; Schauder, S.; Potier, N.; Van Dorsselaer, A.; Pelczer,
I.; Bassler, B. L.; Hughson, F. M. Nature 2002, 415, 545−9.
(28) The existence of an additional cavity proximal to the DPD
binding site within LuxP has also been discussed in ref 18.
(29) Taga, M. E.; Miller, S. T.; Bassler, B. L. Mol. Microbiol. 2003, 50,
1411−27.
(30) Xavier, K. B.; Miller, S. T.; Lu, W.; Kim, J. H.; Rabinowitz, J.;
Pelczer, I.; Semmelhack, M. F.; Bassler, B. L. ACS Chem. Biol. 2007, 2,
128−36.
(31) Xue, T.; Zhao, L.; Sun, H.; Zhou, X.; Sun, B. Cell Res. 2009, 19,
1258−68.
ACKNOWLEDGMENTS
■
We gratefully acknowledge Prof. Bonnie Bassler (Princeton
University) for providing us with S. Typhimurium strain
Met844, Dr. Daniel Globisch for the LuxP-analogue docking
model, and Dr. Karen C. McCague for insightful discussions.
We also acknowledge financial support from the NIH (Grant
AI077644 to K.D.J.), the Skaggs Institute for Chemical Biology,
and the Japan Society for the Promotion of Science (fellowship
to K.T.).
C
dx.doi.org/10.1021/ja305532y | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX