Biological screening of a diverse set of AI-2 analogues in Vibrio harveyi suggests that receptors which are involved in synergistic agonism of AI-2 and analogues are promiscuous

Article abstract of DOI:10.1039/b909666c

C1-alkyl AI-2 analogues do not induce bioluminescence in V. harveyi on their own but enhance the bioluminescence induced by AI-2 in a synergistic fashion. A new facile synthesis of AI-2 facilitates the synthesis of a diverse set of AI-2 analogues and biol

Full text of DOI:10.1039/b909666c

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Biological screening of a diverse set of AI-2 analogues in Vibrio harveyi
suggests that receptors which are involved in synergistic agonism of
AI-2 and analogues are promiscuousw
Jacqueline A. I. Smith,a Jingxin Wang,a Sao-Mai Nguyen-Mau,b Vincent Leeb and
Herman O. Sintim*a
Received (in Cambridge, UK) 15th May 2009, Accepted 21st September 2009
First published as an Advance Article on the web 13th October 2009
DOI: 10.1039/b909666c
C1-alkyl AI-2 analogues do not induce bioluminescence in
V. harveyi on their own but enhance the bioluminescence induced
by AI-2 in a synergistic fashion. A new facile synthesis of AI-2
facilitates the synthesis of a diverse set of AI-2 analogues and
biological screening suggests that receptors that are involved in
the synergistic bioluminescence production in V. harveyi are
promiscuous.
concentrations, DPD dimerizes into an inactive triacetal
compound.14 The ‘‘instability’’ of DPD therefore places a
constraint on the synthetic strategies that can be used to access
this molecule.18,19 Crucially, the last step of the synthesis of
DPD has to be near quantitative and it is vital to employ
reagents that are easy to remove without the use of standard
chromatographic separation techniques. For the reported
synthesis of C1-linear alkyl chain analogues of AI-2, one of
the key steps involved a SN2 reaction with primary iodides.13
Difficulties in adapting reported AI-2 syntheses to make
C1-branched and cyclic alkyl chain AI-2 analogues are
envisioned (nucleophilic substitution reactions at 21 and 31 alkyl
halide centers are difficult to accomplish). Herein we report a new
and more facile synthesis of AI-2 that has allowed access to the
‘‘difficult-to-make’’ C1-tert-butyl, C1-isopropyl and cyclic alkyl
chain AI-2 analogues (see Scheme 1).
Many bacteria use chemical signals to regulate the expression
of important genes as a function of cell density. This
phenomenon, known as quorum sensing (QS), is a cell-to-cell
communication system that allows bacteria to assess their
local population density via the secretion and detection of
small, diffusible signal molecules called autoinducers and
regulate gene expression when a critical population density is
reached.1 QS regulates the expression of virulence factors,2
and other factors that are important for bacterial colonization
of higher organisms,3 susceptibility to antibiotics,4 the formation
of biofilms5,6 and the growth rates of some bacteria.7 Therefore
molecules that can inhibit proteins or other macromolecules
that are involved in the quorum-sensing process could in
principle be used to fight bacterial infection.8–11
For our strategy towards AI-2, the products obtained from
the condensation of diazocarbonyl 1 and aldehyde 2 in the
presence of DBU were not purified but taken to the next step
whereby the silyl group was deprotected with tetrabutyl-
ammonium fluoride (TBAF) in tetrahydrofuran (THF) to give
diol 3 in good yields after silica column chromatography.
Diazodiol 3 existed predominantly as an open chain and not as
a lactol (as would have been expected for a compound containing
The LuxS enzyme catalyzes the formation of (S)-4,5-
dihydroxy-2,3-pentanedione (DPD), the precursor of AI-2,
and it has been identified in over 55 bacterial strains.12 Because
AI-2 is produced and sensed by many bacterial strains, it has
been termed ‘‘the universal autoinducer’’.12 It therefore follows
that small molecules that antagonize the actions of AI-2 might
have broad spectrum antibiotic effects. Six C1-linear alkyl
analogues of AI-2 have been shown to be non-toxic to human
cells and they exhibit different biological activities in V. harveyi
and S. typhimurium; synergistic agonists for bioluminescence
in V. harveyi and antagonists for b-galactosidase activity in
S. typhimurium.13 Access to C1 branched and cyclic alkyl chain
AI-2 analogues will allow a more comprehensive study of how
both the size and shape of the C1 alkyl group of AI-2 analogues
affect their biological profile.
Despite the structural simplicity of AI-2, its chemical synthesis
is not trivial.14–17 DPD (the precursor of AI-2) is only stable at
dilute concentrations and it has been shown that at higher
a Department of Chemistry and Biochemistry, University of Maryland,
College Park, MD 20742, USA. E-mail: hsintim@umd.edu
b Department of Cell Biology and Molecular Genetics,
University of Maryland, College Park, MD 20742, USA
w Electronic supplementary information (ESI) available: Character-
ization of new compounds. See DOI: 10.1039/b909666c
Scheme 1 A two-pot synthesis of AI-2 and analogues; (a) cat. DBU,
MeCN; (b) TBAF (1–3 eq), THF; (c) dimethyl dioxirane; acetone.
DBU = 1,8-diazabicyclo[5.4.0]undec-7-ene.
ꢀc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 7033–7035 | 7033

View Article Online
hydroxyl and ketone functionalities in a 1,4-relationship). The
13C NMR showed a carbonyl peak at 192 ppm and there was
no peak between 100 and 120 ppm (which would have been
indicative of lactolization). Presumably, hyperconjugation
between the bonding electrons of the C–N bond and the p*
of the carbonyl bond makes the CQO bond less electrophilic
for lactolization to occur.
50 mM, only the ethyl analogue 4b and the cyclopropyl
analogue 4g induced significant bioluminescence, although
the bioluminescence intensities from both 4b and 4g were at
least an order of magnitude less than that from AI-2 after 8 h
(see Fig. 1b). The bioluminescence intensities from the rest of
the analogues (4c, 4d, 4e, 4f and 4h) were similar to when no
analogue was added (i.e. background).
With adequate quantities of diazodiol 3 in hand, after only a
one-pot operation, the stage was set to oxidize the diazo
functionality into a carbonyl moiety. At this stage, we
were cognizant of the fact that the oxidizing reagent or any
byproducts of the oxidizing reagent required for the end-game
of our synthesis had to be volatile because of the difficulty in
purifying DPD using column chromatography, vide supra.
Therefore, despite the availability of a wide range of oxidizing
reagents that have been shown to oxidize the diazo moiety into
a carbonyl,20 only a handful of reagents fitted this requirement.
We settled on dioxirane because it is volatile and importantly
its byproduct is acetone, which is also volatile.21 Pleasingly,
treatment of diazodiol 3 with an acetone solution of dioxirane
afforded DPD 4 and its isomers 5 and 6 in quantitative yield
after evaporation of the acetone solvent. The NMR of our
synthetic DPD and that of a quinoxaline derivative, which was
obtained by reacting the synthetic AI-2 with 1,2-phenyldiamine,
matched literature values.14–17
In the course of conducting our studies, another group also
reported that another diverse group of C1-alkyl analogues of
AI-2 enhanced AI-2-induced bioluminescence in V. harveyi
even though they did not induce bioluminescence on their own
(see ref. 13).23 Four of the compounds in this study (isopropyl
4e, tert-butyl 4f, cyclopropyl 4g and cyclohexyl 4h) are new
analogues and were not reported in the recent report.13 These
new analogues were designed to test how shape and size affect
AI-2 analogue induced bioluminescence or synergistically-
induced bioluminescence. We find that diverse shapes and
sizes of the C1-alkyl chain of AI-2 are all able to synergistically
induce bioluminescence in V. harveyi in the presence of AI-2;
receptors that bind to AI-2 analogues in order to promote
bioluminescence in V. harveyi (in synergism with AI-2) display
marked promiscuity of ligand binding.
The origin of the concentration-dependent synergistic
enhancement (by 4b, 4c, 4d, 4e, 4f and 4g) of AI-2-induced
bioluminescence remains unknown. The analogues might act
on a different target protein and the action of the second target
sensitizes the LuxP/LuxQ system that is involved in V. harveyi
bioluminescence or (b) the analogues may activate a protein
which is further downstream of the LuxP/LuxQ AI-2 signaling
pathway. It is also plausible that the enhancement of
AI-2-induced bioluminescence by C1-alkyl analogues of AI-2
is due to one AI-2 molecule binding to one active site in the
LuxP/LuxQ dimer whereas the analogue binds to the other
active site; resulting in a desymmetrized complex that is more
active at eliciting bioluminescence than the complex that
contains two identical AI-2 or AI-2 analogue molecules;
Hughson et al. have previously shown that two AI-2 molecules
bind to the LuxP/LuxQ dimer, in an asymmetric manner, to
elicit bioluminescence in V. harveyi.24
We selected bioluminescence production in V. harveyi for
biological evaluation of our AI-2 analogues because of the
rapid readout and reproducibility of this particular assay.
Using the V. harveyi LuxNÀ and LuxSÀ strain, MM32, we
monitored the bioluminescence of each analogue for up to
8 hours. Our synthetic AI-2 4a induced bioluminescence in
V. harveyi (MM32) at various concentrations (from nanomolar
to micromolar).22
Our group is interested in deciphering the role of conformation
of C1-alkyl analogues of AI-2 on enhancing or inhibiting
AI-2-mediated quorum sensing processes. Towards this end,
we synthesized a panel of AI-2 analogues with different ring
and linear sizes and shapes. Because AI-2 and analogues exist
as an equilibrium mixture of compounds, we confirmed the
identity of the analogues by converting them into quinoxaline
derivatives and fully characterizing these derivatives (see the
ESIw). None of the C1-alkyl analogues of AI-2 induced
bioluminescence at 2 mM. At AI-2 analogue concentration of
In conclusion, we have provided
a straight-forward
synthesis of the universal autoinducer, AI-2. This development
has facilitated the synthesis of a panel of seven C1-alkyl
analogues of AI-2, four of which are novel. The analogues
Fig. 1 (a) Bioluminescence induction in V. harveyi MM32 (at 8 h) by the addition of a DPD mixture containing analogue (2 mM), AI-2 (12 nM)
and boric acid (100 mM); synergistic agonism. Fold-activations are as follows: ethyl (4.3), propyl (2.6), butyl (2.7), isopropyl (2.4), tert-butyl (2.9),
cyclopropyl (3.1), cyclohexyl (9.1). The analogues (2 mM) did not induce bioluminescence on their own. (b) Bioluminescence induction in V. harveyi
MM32 by the addition of 50 mM of each analogue and boric acid (100 mM).
ꢀc
This journal is The Royal Society of Chemistry 2009
7034 | Chem. Commun., 2009, 7033–7035

View Article Online
were screened for bioluminescence induction in V. harveyi and
they enhanced AI-2-mediated bioluminescence, although they
did not induce bioluminescence in V. harveyi on their own.
Interestingly, a clear trend does not emerge as regards to the
size or shape of an AI-2 analogue and its fold enhancement of
AI-2-mediated bioluminescence in V. harveyi. This suggests
that the receptors that mediate the AI-2 and analogue syner-
gistic agonism may be promiscuous. Our new and expeditious
synthesis of AI-2 analogues could be used to prepare an
expanded set of AI-2 analogues for further biological testing,
such as investigating promiscuity or lack-thereof in proteins
that signal AI-2 binding into other quorum sensing processes.
We thank the University of Maryland, National Science
Foundation and the GAANN fellowship (to JS) for funding
our projects.
11 P. B. Kapadnis, E. Hall, M. Ramstedt, W. R. J. D. Galloway,
M. Welch and D. R. Spring, Chem. Commun., 2009, 538.
12 C. M. Waters and B. L. Bassler, Annu. Rev. Cell Dev. Biol., 2005,
21, 319.
13 C. A. Lowery, J. Park, G. F. Kaufmann and K. D. Janda, J. Am.
Chem. Soc., 2008, 130, 9200.
14 M. M. Meijler, G. F. Kaufmann, L. G. Hom, K. McKenzie,
C. Sun, J. A. Moss, M. Matsushita and K. D. Janda, Angew.
Chem., Int. Ed., 2004, 43, 2106.
15 M. F. Semmelhack, S. R. Campagna, M. J. Federle and
B. L. Bassler, Org. Lett., 2005, 7, 569.
16 M. Frezza, L. Soulere, Y. Queneau and A. Doutheau, Tetrahedron
Lett., 2005, 46, 6495.
17 S. C. J. De Keersmaecker, C. Varszegi, N. van Boxel, L. W. Habel,
K. Metzger, R. Daniels, K. Marchal, D. De Vos and
J. Vanderleyden, J. Biol. Chem., 2005, 280, 19563.
18 S. Schauder, K. Shokat, M. G. Surette and B. L. Bassler, Mol.
Microbiol., 2001, 41, 463.
19 K. Winzer, K. R. Hardie, N. Burgess, N. Doherty, D. Kirke, M. T.
G. Holden, R. Linforth, K. A. Cornell, A. J. Taylor, P. J. Hill and
P. Williams, Microbiol., 2002, 148, 909.
20 T. Ye and A. McKervey, Chem. Rev., 1994, 94, 1091.
21 R. W. Murray and R. J. Jeyaraman, J. Org. Chem., 1985, 50, 2847.
22 We find that AI-2 prepared using our method is stable and
biologically active after several months of storage, even at
millimolar solutions in DMSO/water.
23 Although synergistic agonism assays of AI-2 analogues have been
previously done without added boric acid (see ref. 13), we added
boric acid (100 mM) to our assay. We thank an anonymous
reviewer who pointed out that when boric acid is not added to
the culture media for the bioluminescence assay, the analogues
might scavenge for adventitious borate; thereby affecting the
results of different assays that contained different amounts of
adventitious borate. We nonetheless confirm Janda’s report that
AI-2 analogues cause fold enhancement of AI-2 induced biolumi-
nescence in V. harveyi, although that study did not add boric acid
to the media for the bioluminescence assay.
Notes and references
1 W. C. Fuqua, S. C. Winans and E. P. Greenberg, J. Bacteriol.,
1994, 176, 269.
2 J. M. Zhu, M. B. Miller, R. E. Vance, M. Dziejman, M. B. Bassler
and J. J. Mekalanos, Proc. Natl. Acad. Sci. U. S. A., 2002, 99, 3129.
3 M. D. Koutsoudis, D. Tsaltas, T. D. Minogue and S. B. Bodman,
Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 5983.
4 N. A. A. Ahmed, F. C. Petersen and A. A. Scheie, J. Antimicrob.
Chemother., 2007, 60, 49.
5 D. G. P. Davies, M. R. Parsek, J. P. Pearson, B. H. Iglewski,
J. W. Costerton and E. P. Greenberg, Science, 1998, 280, 295.
6 P. K. Singh, A. L. Schaefer, M. R. Parsek, T. O. Moninger,
M. J. Welsh and P. E. Greenberg, Nature, 2000, 407, 762.
7 K. M. Sandoz, S. M. Mitzimberg and M. Schuster, Proc. Natl.
Acad. Sci. U. S. A., 2007, 104, 15876.
8 A. E. P. Clatworthy and D. T. Hung, Nat. Chem. Biol., 2007, 3, 541.
9 K. M. Smith, Y. Bu and H. Suga, Chem. Biol., 2003, 10, 81.
10 G. D. Geske, R. J. Wezeman, A. P. Siegel and H. E. Blackwell,
J. Am. Chem. Soc., 2005, 127, 12762.
24 M. B. Neiditch, M. J. Federle, A. J. Pompeani, R. C. Kelly,
D. L. Swern, P. D. Jeffrey, B. L. Bassler and F. M. Hughson,
Cell, 2006, 126, 1095.
ꢀc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 7033–7035 | 7035