Ac2-DPD, the bis-(O)-acetylated derivative of 4,5-dihydroxy-2,3-pentanedione (DPD) is a convenient stable precursor of bacterial quorum sensing autoinducer AI-2

Article abstract of DOI:10.1016/j.bmcl.2006.11.076

Ac₂-DPD, the bis-(O)-acetylated derivative of 4,5-dihydroxy-2,3-pentanedione (DPD), was prepared both as a racemic mixture and in the optically active form found in naturally occurring DPD. It was shown to exhibit the same ability as DPD to induce bioluminescence in Vibrio Harveyi and β-galactosidase activity in Salmonella enterica Typhimurium, both Gram-negative bacteria. Likewise, it was also shown to inhibit biofilm formation in Gram-positive Bacillus cereus. The most likely hypothesis is that Ac₂-DPD activity is due to the release of DPD by in situ hydrolysis of the ester groups. Importantly, by contrast with DPD, Ac₂-DPD proved to be a stable compound which can be purified and stored.

Full text of DOI:10.1016/j.bmcl.2006.11.076

Bioorganic & Medicinal Chemistry Letters 17 (2007) 1428–1431
Ac2-DPD, the bis-(O)-acetylated derivative of 4,5-dihydroxy-2,3-
pentanedione (DPD) is a convenient stable precursor
of bacterial quorum sensing autoinducer AI-2
a
a
b
c
`
Marine Frezza, Laurent Soulere, Damien Balestrino, Michel Gohar,
Christian Deshayes,a Yves Queneau,a Christiane Forestierb and Alain Doutheaua,*
aLaboratoire de Chimie Organique, UMR 5181 CNRS, Universite´ Lyon 1, INSA, Institut National des Sciences Applique´es,
ˆ
Bat. Jules Verne, 20 avenue A. Einstein, 69621 Villeurbanne, France
bLaboratoire de Bacte´riologie Faculte´ de Pharmacie, Universite´ d’Auvergne Clermont-Ferrand 28, Place Henri Dunant,
BP 38. 63001 Clermont-Ferrand cedex 1, France
cMicrobiologie et Ge´ne´tique Mole´culaire, INRA-CNRS-INAPG, F-78850 Thiverval-Grignon, France
Received 17 July 2006; revised 28 November 2006; accepted 30 November 2006
Available online 2 December 2006
Abstract—Ac2-DPD, the bis-(O)-acetylated derivative of 4,5-dihydroxy-2,3-pentanedione (DPD), was prepared both as a racemic
mixture and in the optically active form found in naturally occurring DPD. It was shown to exhibit the same ability as DPD to
induce bioluminescence in Vibrio Harveyi and b-galactosidase activity in Salmonella enterica Typhimurium, both Gram-negative
bacteria. Likewise, it was also shown to inhibit biofilm formation in Gram-positive Bacillus cereus. The most likely hypothesis is
that Ac2-DPD activity is due to the release of DPD by in situ hydrolysis of the ester groups. Importantly, by contrast with
DPD, Ac2-DPD proved to be a stable compound which can be purified and stored.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Quorum sensing (QS) is a cell to cell communication sys-
tem used by bacteria to regulate gene expression in a
concerted way dependent upon the population density.1
This process is based on the synthesis of small, diffusible
signalling molecules called autoinducers (AI) which are
able to activate transcriptional regulators. In Gram-neg-
ative bacteria, AI are mainly acylated homoserine lac-
tone (AHLs), while linear and cyclic peptides are used
by Gram-positive bacteria. Besides these two families,
it has been recently shown that, either the hydrate 1 or
the furanosyl borate diester 2, mediates QS communica-
tion in both Gram-negative and Gram-positive bacteria
(Scheme 1).2 Compound 1 is issued from the hemiketalic
ring-closed form of the (4S)-4,5-dihydroxy-2,3-pentan-
edione (DPD) derived from (S)-adenosyl methionine.
Since DPD, 1 and 2 are in equilibrium, autoinducer of
type 2 (AI-2) has been suggested as a collective term
for DPD-derived compounds promoting bacterial
cross-communication.3
In the last few years, AI-2 has received a great deal of
interest not only from microbiologists but also from
chemists as an interesting synthetic target. Indeed, de-
spite its very simple structure, the synthesis of DPD is
not an easy task owing to its instability at high concen-
trations. As a consequence, in the reported synthesis of
DPD this compound was not purified and was obtained
as dilute aqueous solutions, either from acidic hydrolysis
of a orthoester4 or a ketal,5 or by reductive ozonolysis of
an a-methylene ketone.6,7
We envisaged to overcome this lack of practical avail-
ability of DPD by preparing a precursor with enhanced
stability that could be stored and used in biological as-
says, without prior chemical treatment. With this objec-
tive in mind, we decided to prepare and test the
corresponding mono and di-O-acetylated derivatives.
The acetyl was chosen as the OH masking group be-
cause of its moderate stability at physiological pH and
its known enzymo-lability.8
We first synthesised the 5-O-acetyl-4,5-dihydroxy-2,3-
pentanedione 3.7 This compound proved to be hardly
more stable than DPD and could not be concentrated
Keywords: Quorum sensing; Autoinducer; AI-2; DPD.
*
Corresponding author. Fax: +33 472 43 88 96; e-mail: alain.
doutheau@insa-lyon.fr
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.11.076

M. Frezza et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1428–1431
1429
OH
O
HO
HO
O
OH
OH
CH3
B
OH
O
HO
HO
H2O
O
OH
CH3
CH3
HO
HO
HO
O
O
O
O
(S)-DPD
1
2
AI-2
Scheme 1. Structures of bacterial QS autoinducers of type-2 (AI-2).
OH
O
OR
O
OR
OAc O
c
5 or 7
OAc O
OAc
OR
O
OAc O
( ) 8
S) -8
3
4 R = H
6 R = H
a
b
5 R = Ac
7 R = Ac
(
Scheme 2. Synthesis of 8. Reagents and conditions: (a) acetic anhydride, 4-dimethylaminopyridine (DMAP), CH2Cl2, 70%; (b) acetic anhydride,
pyridine, Et3N (0.1 equiv), CH2Cl2, 0 ꢁC, 90%; (c) O3, MeOH, ꢀ78 ꢁC then dimethyl sulfide (DMS), ꢀ78 ꢁC–rt.
without decomposition. We then prepared the bis-O-ac-
etyl-4,5-dihydroxy-2,3-pentanedione 8 (Scheme 2). The
synthesis of racemic ( ) 8 was achieved, using our meth-
odology,7 by acetylation of the b-hydroxy enone 4 to
give 5 and subsequent ozonolysis. To prepare (S)-8, we
used the methodology reported by Vanderleyden et al.
for the synthesis of (S)-DPD from the dihydroxy enone
(S)-6.6 The latter compound was first acetylated to give
7 which was then submitted to reductive ozonolysis.
Both crude ( ) 8 and (S)-8 were purified using column
chromatography on silica gel without notable degrada-
tion and were obtained as yellow oils (Scheme 2).
about the same activity as (S)-DPD, with IC50 values
being of 2.1 and 2.6 lM, respectively. Similar results
were obtained for ( ) 8 compared with ( ) DPD with
IC50 values of 2.4 and 3.3 lM, respectively (data not
shown).10
We anticipated that Ac2-DPD would have the same
activity as DPD by releasing the latter compound after
in situ hydrolysis of the ester groups. To support this
hypothesis, we studied the chemical behaviour of (S)-8
in an aqueous medium at room temperature and at the
pH of biological tests (100 mM phosphate buffer, pH
7.4) by recording 1H NMR spectra every 10 min
(Fig. 2).11 The monoacetylated compound 3 resulting
from hydrolysis of the secondary 4-OAc group of (S)-
8 is rapidly observed in the medium (Fig. 2). The con-
centration of 3 increases to reach a maximum after
about 1 h and then slowly decreases due to the further
hydrolysis of the primary acetate function giving rise
to DPD. After about 280 min of reaction side products
arising from DPD degradation are also detected pre-
venting an accurate analysis of the mixture.
The diacetate (8) was first evaluated for its ability to in-
duce bioluminescence in Vibrio harveyi bacteria known
to use borate 2 as autoinducer.9 As shown by results
depicted in Figure 1, (S)-8 strongly induces light with
1,00E+07
5,00E+06
0,00E+00
Comparison of the rate of this in vitro formation of
DPD from (S)-8 is not fully consistent with the
100
50
0
0,001
0,01
0,1
1
10
100
0
100
200
300
Concentration (µM)
time (min)
Figure 1. Bioluminescence induction in V. harveyi after 4 h of
incubation at 30 ꢁC with increasing concentration of (S)-8 (d) and
(S)-DPD (j).
Figure 2. Hydrolysis of diacetate (S)-8 (j) into monoacetate 3 (d) and
DPD (·) in buffered (pH 7.4) aqueous medium.

1430
M. Frezza et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1428–1431
Figure 3. (a) Hydrolysis of the secondary acetyl group of (S)-8, in the absence (j) or in the presence (d) of PLE. (b) (S)-DPD release from (S)-8 in
the absence (j) or in the presence (d) of PLE.
biological activity of the latter compound on V. harveyi
bacterium. Indeed, after 4 h only about 30% of (S)-8 has
been transformed into DPD (Fig. 2) while, after a same
4 h period of incubation, the activity of 8 is identical to
that of pure DPD (Fig. 1). A combination of different
factors can explain this discrepancy such as the presence
of mineral components in the biological medium
increasing the rate of hydrolysis, or the acceleration of
DPD formation at the temperature of incubation. Also,
in the case of biological assays using DPD, concomitant
degradation all over the incubation period might lower
its actual concentration. Finally, bacterial hydrolytic en-
zymes could also participate in the acceleration of the
hydrolysis of acetate groups during the biological assay.
In order to detect the eventual contribution of an enzy-
matic catalysis, we studied the hydrolysis of 8 in the
presence of pig liver esterase (PLE), a common used
hydrolytic enzyme.12 As shown by results depicted in
Figure 3, this enzyme is able to increase significantly
the formation of (S)-DPD though resulting from moder-
ate rate enhancements of both the first and second
hydrolysis of acetyl groups of (S)-8.13
compound proved to be a potent inducer of b-galactosi-
dase production.16 The higher activity of (S)-8 com-
pared to DPD observed in this case is likely due to
partial degradation of this latter during the biological
assays.
Finally, the effect of (S)-8 on biofilm formation was as-
sayed in the strain 407 of the Bacillus cereus which is a
Gram-positive bacterium.17 At 8 lM and after 24 h of
incubation, both (S)-8 and (S)-DPD inhibited strongly
and significantly (p < 0.01, t-test) biofilm formation
(Fig. 5).18
In conclusion, we have observed that Ac2-DPD induces
the same biological effects as DPD on two Gram-nega-
tive and one Gram-positive bacteria most probably
through in situ release of DPD. Indeed, these bacteria
using two different derivatives of DPD as natural quo-
rum sensing autoinducers (either a hydrate or a borate),
it would be very unlikely that Ac2-DPD be active by it-
self or after hydrolysis of only one of its ester groups.
The fact that the same behaviour of Ac2-DPD is ob-
served in three different microorganisms comforts us in
the idea that it could be a precursor of DPD of general
use for biochemical, microbiological or biotechnological
Compound (S)-8 was then tested for its ability to induce
b-galactosidase production in Salmonella enterica
Typhimurium bacteria which use hydrate 1 as autoin-
ducer.2b,15 As shown by results depicted in Figure 4, this
2,0E+05
1,0E+05
0,0E+00
0,001
0,01
0,1
1
10
100
1000
Concentration (µM)
Figure 4. b-Galactosidase production in Salmonella enterica Typhimu-
rium with increasing concentration of (S)-8 (d) and (S)-DPD (j).
Figure 5. Biofilm inhibition in B. cereus after 24 h of incubation with
(S)-8 or (S)-DPD at 8lM.

M. Frezza et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1428–1431
1431
scans) for 10 h at room temperature. The proportion of
each compound, (S)-8, 3, and DPD, was determined on
the basis of integration of the signals (in D2O) corre-
sponding to the 2 diastereotopic H of the methylene group
(C-5) at 4.24 and 4.44 ppm for (S)-8 and at 4.14 ppm for3
or to the hemiketalic methyl groups (2 diasteroisomers) at
1.39 and 1.42 ppm for DPD. The precision for integration
measurements is about 5%. At low concentrations of DPD
(ca. 10%) no accurate values can be given by this method
because the intensity of DPD signals is weak due to its
equilibrium between three species.
applications. Also, this work opens new perspectives for
the development of prodrug type derivatives as quorum
sensing modulators.
Acknowledgments
This research was supported by the MENESR and
CNRS. M.F. thanks the MENESR for a scholarship.
Authors thank Dr B. Bassler for very kindly providing
V. harveyi BB170 and MM30 as well as Salmonella
typhimurium MET 844 strains.
12. Tamm, C. Pure Appl. Chem. 1992, 64, 1187.
13. Kinetic data showed that the hydrolysis of the 4-OAc
group of 8 to give3 is first-order with respect to 8
(k1 = 0.029 minꢀ1). The hydrolysis of the 5-OAc group of
3 giving rise to DPD appeared also to be first-order with
respect to 3. From these series first-order reaction,14 the
corresponding rate constant k2 was found to be about
0.0021 minꢀ1. In the presence of PLE (EC 3.1.1.1, 19 U/
mg, 10 mg), a similar treatment of kinetic data of Fig. 3a
and b allowed us to estimate the corresponding first-order
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.bmcl.
2006.11.076.
constants: k01 ꢁ 0:045 minꢀ1; k02 ꢁ 0:0047 minꢀ1
.
14. Moore, J. W.; Pearson, R. G. In Kinetics and Mechanism;
John Wiley & Sons: New York, 1981; p 290.
References and notes
15. Taga, M. E.; Miller, S. T.; Bassler, B. L. Mol. Microbiol.
2003, 50, 1411.
1. For recent reviews (a) Miller, M. B.; Bassler, B. L. Annu.
Rev. Microbiol. 2001, 55, 165; (b) Whitehead, N. A.;
Barnard, A. M.; Slater, H.; Simpson, N. J.; Salmond, G.
P. FEMS Microbiol. Rev. 2001, 25, 365; (c) Lyon, G. J.;
Muir, T. W. Chem. Biol. 2003, 10, 1007; (d) Waters, C. M.;
Bassler, B. L. Annu. Rev. Cell Dev. Biol. 2005, 21, 319; (e)
Reading, N. C.; Sperandio, V. FEMS Microbiol. Lett.
2006, 254, 1.
2. (a) Chen, X.; Schauder, S.; Potier, N.; Van Dorsselaer, A.;
Pelczer, I.; Bassler, B. L.; Hughson, F. M. Nature 2002,
415, 545; (b) Miller, S. T.; Xavier, K. B.; Campagna, S. R.;
Taga, M.; Semmelhack, M. F.; Bassler, B. L.; Hughson, F.
M. Mol. Cell. 2004, 15, 677.
16. The bioluminescence- and b-galactosidase-inducing activ-
ities of ( )-DPD or (S)-8 were measured using a V. harveyi 9
and a Salmonella enterica Typhimurium15 reporter strains,
respectively. The luminescent reporter strain V. harveyi
BB170 was grown for 18 h at 30 ꢁC, with aeration, in
AB medium and was diluted 1:5000 into fresh AB medium.
The S. typhimurium MET844 strain was grown for 18 h
at 37 ꢁC in Luria Bertani (LB) medium and diluted 1:100
in LB. Next, DPD or 8 was added to each diluted bacterial
cell suspension at a 10% (v/v) final concentration, and
then shaken at 30 ꢁC (V. harveyi) or 37 ꢁC (S. typhimurium)
for 4 h. The backgrounds were determined by the
addition of sterile medium to diluted bacterial cell suspen-
sions. After the incubation period, the resulting light or
b-galactosidase production was measured with a lumino-
meter, directly or using the beta-Gal reporter gene assay
(chemiluminescent) (Roche, Mannheim, Germany), and
activity was expressed as relative units of luminescence
(RLU) per second.
3. De Keersmaecker, S. C.; Sonck, K.; Vanderleyden, J.
Trends Microbiol. 2006, 14, 114.
4. Meijler, M. M.; Hom, L. G.; Kaufmann, G. F.; McKen-
zie, K. M.; Sun, C.; Moss, J. A.; Matsushita, M.; Janda,
K. D. Angew. Chem., Int. Ed. 2004, 43, 2106.
5. Semmelhack, M. F.; Campagna, S. R.; Federle, M. J.;
Bassler, B. L. Org. Lett. 2005, 7, 569.
6. De Keersmaecker, S. C.; Varszegi, C.; van Boxel, N.;
Habel, L. W.; Metzger, K.; Daniels, R.; Marchal, K.; De
Vos, D.; Vanderleyden, J. J. Biol. Chem. 2005, 280, 19563.
7. Frezza, M.; Soulere, L.; Queneau, Y.; Doutheau, A.
Tetrahedron Lett. 2005, 46, 6495.
8. (a) Potter, P. M.; Wadkins, R. M. Curr. Med. Chem. 2006,
13, 1045; (b) Kadereit, D.; Waldmann, H. Chem. Rev.
2001, 101, 3367.
9. Bassler, B. L.; Greenberg, E. P.; Stevens, A. M. J.
Bacteriol. 1997, 179, 4043.
10. Racemic and (S)-DPD were prepared according to Refs.
7,6, respectively.
17. Auger, S.; Krin, E.; Aymerich, S.; Gohar, M. Appl.
Environ. Microbiol. 2006, 72, 937.
18. Biofilm inhibition was tested in a 96-well polyvinylchlo-
ride microtiter plate assay.17 Briefly, microtiter plates
were inoculated with a diluted overnight preculture of
the strain 407 of B. cereus. Synthetic (S)-DPD or (S)-8
were subsequently added to the microtiter wells and
biofilm density was measured after 24 h of incubation as
follows: the microtiter plate wells were washed with
phosphate-buffered saline, and bound cells were stained
with a 1% (wt/vol) crystal violet solution at room
temperature for 20 min. The wells were then washed and
the dye was solubilized with a 20:80 acetone/ethanol
mixture. The absorbance at 600 nm of the solubilized
dye was then determined. Assays were performed
triplicate.
11. NMR samples were prepared as follow: (S)-8 (1.8 mg,
8.33 lmol) was dissolved in a 100 mM Na2 HPO4 buffer in
D2O (0.6 mL, pH 7.4). 1H NMR spectra were recorded on
a Bruker spectrometer DRX 500 MHz each 10 min (64