Synthesis of carbon-14 labelled (5Z)-4-bromo-5-(bromomethylene)-2(5H)- furanone: A potent quorum sensing inhibitor

Article abstract of DOI:10.1002/jlcr.846

The potent quorum sensing inhibitor (5Z)-4-bromo-5-(bromomethylene)-2(5H)- [2-¹⁴C]furanone has been prepared in five steps in 7.7% overall yield starting from bromo[1-¹⁴C]acetic acid. Condensation of ethyl bromo[1-¹⁴C]acetate with ethyl acetoacetate followed by decarboxylation was accelerated by microwave heating to afford [1- ¹⁴C]levulinic acid. Subsequently, bromination and oxidation gave the targeted furan-2-one with a radiochemical purity of > 97% and a specific activity of 57 mCi/mmol. Copyright

Full text of DOI:10.1002/jlcr.846

JOURNAL OF LABELLED COMPOUNDS AND RADIOPHARMACEUTICALS
J Label Compd Radiopharm 2004; 47: 627–634.
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jlcr.846
Research Article
Synthesis of carbon-14 labelled (5Z)-4-bromo-5-
(bromomethylene)-2(5H)-furanone: a potent quorum
sensing inhibitor
Tobias Persson1,2, Steen K. Johansen3,*, Lars Martiny3, Michael Givskov4
and John Nielsen1
1 Department of Chemistry, Royal Veterinary and Agricultural University, DK-1871
Frederiksberg C, Denmark
2 Department of Chemistry, Technical University of Denmark, DK-2800 Kgs. Lyngby,
Denmark
3 Department of Isotope Chemistry, Novo Nordisk A/S, Novo Nordisk Park, DK-2760
(
Maløv, Denmark
4 Department of Microbiology, Technical University of Denmark, DK-2800 Kgs.
Lyngby, Denmark
Summary
The potent quorum sensing inhibitor (5Z)-4-bromo-5-(bromomethylene)-2(5H)-
[2-14C]furanone has been prepared in five steps in 7.7% overall yield starting from
bromo[1-14C]acetic acid. Condensation of ethyl bromo[1-14C]acetate with ethyl
acetoacetate followed by decarboxylation was accelerated by microwave heating to
afford [1-14C]levulinic acid. Subsequently, bromination and oxidation gave the
targeted furan-2-one with a radiochemical purity of >97% and a specific activity of
57 mCi/mmol. Copyright # 2004 John Wiley & Sons, Ltd.
Key Words: carbon-14; quorum sensing inhibitor; levulinic acid; microwave;
decarboxylation; condensation
Introduction
Several gram-negative bacteria control expression of virulence factors by a
phenomenon termed quorum sensing.1 The bacteria produces extracellular
signal molecules, which accumulate during growth, and at a certain threshold
level, interaction with the receptor (LuxR homologue) leads to a burst in
(
*Correspondence to: S.K. Johansen, Novo Nordisk A/S, Novo Nordisk Park, DK-2760 Malꢀv, Denmark.
E-mail: skjo@novonordisk.com
Contract/grant sponsor: The Danish Technical Research Council; contract grant number: 56-00-0134
Copyright # 2004 John Wiley & Sons, Ltd.
Received 26 March 2004
Revised 21 April 2004
Accepted 26 April 2004

628
T. PERSSON ET AL.
O
O
O
O
O
N
N
H
H
O
O
1
2
Br
Br
Br
Br
O
O
O
O
OH
3
4
Figure 1. Signal molecules N-butanoyl-l-homoserine lactone (1) and N-(3-
oxododecanoyl)-l-homoserine lactone (2) and quorum sensing inhibitors (10R,
5Z)-4-bromo-3-(10-hydroxybutyl)-5-(bromomethylene)-2(5H)-furanone (3) and
(5Z)-4-bromo-5-(bromomethylene)-2(5H)-furanone (4)
expression of the target genes. In Pseudomonas aeruginosa, the major signal
molecules are N-butanoyl-l-homoserine lactone (1) and N-(3-oxododecanoyl)-
l-homoserine lactone (2) (Figure 1) with the components of the dual quorum
sensing apparatus encoded by rhlI/rhlR and lasI/lasR, respectively.1 The
lactone moiety of the signal molecule (e.g. homoserine lactone in 1 and 2) is a
generic feature of ligands bound to the transcriptional regulator, and
interestingly, several other five-membered lactones have been shown to
antagonise quorum sensing.2,3 One example is fimbrolide 3, discovered as a
secondary metabolite of red seaweed Delisea pulchra, expressing antimicrobial
and antifungal activity.4–6 More recent findings demonstrate that fimbrolides
are potent inhibitors of quorum sensing in P. aeruginosa.7
In order to investigate the biological activity by means of reactivity and
mode of action, a suitable labelled inhibitor was needed. One of the most
convenient tools for labelling of biological compounds is the use of radioactive
isotopes such as tritium or carbon-14. Compared to other labelling techniques,
this methodology has the advantage of not altering the biological activity, as
the labelled compound is identical to the native compound except that one or
more atoms has been replaced by their radioactive counterparts. Furthermore,
radioactivity can be measured and quantified in trace amounts by simple
methods. While labelling with tritium offers the advantage of high specific
activity and short synthetic routes in general, it also offers considerably less
radiochemical and metabolic stability compared to carbon-14. For this reason,
a carbon-14 labelled tracer is normally used in more complex studies, even
though the synthetic route often proves longer and more complex. Attempts to
use tritium-labelled fimbrolide 3 have been reported as unsuccessful,8 but since
then, another more efficient inhibitor has been discovered. The five-membered
.
9
cyclic product obtained by Wolff and Rudel in 1896 by bromination and
oxidation of levulinic acid was reassigned as (5Z)-4-bromo-5-(bromomethy-
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 627–634

SYNTHESIS OF CARBON-14
629
lene)-2(5H)furanone (4) in 1963.10 Substance 4, a non-natural representative
of the fimbrolides, proved to have superior in vitro biological activity
compared to the natural fimbrolides,8 and recently, this was extended to an
in vivo mouse model showing attenuation of quorum sensing.7 Here, we
describe the synthesis of carbon-14 labelled 4 to be used in future binding-
assays and animal tests.
Results and Discussion
A number of different strategies for a small-scale radiochemical synthesis of
(5Z)-4-bromo-5-(bromomethylene)-2(5H)-furanone (4) were considered. The
.
9
last two steps of the procedure described by Wolff and Rudel (bromination
followed by simultaneous dehydration and oxidation of levulinic acid) were, as
it turned out, easily transferred to a radiolabelling strategy. With this protocol
in hand, a retrosynthetic C–C disconnection of levulinic acid was needed in
order to bring the strategy to a level at which carbon-14 labelled starting
materials were available. Carbon-14 labelled ethyl bromoacetate could be
envisioned as the synthon together with an enol provided either as the
corresponding stannane or masked in ethylacetoacetate. The latter route has
been described by Joubert et al. for the synthesis of carbon-13 labelled
levulinic acid.11 Carbon-14 labelled ethyl bromoacetate, in turn, is easily
prepared from commercially available carbon-14 labelled bromoacetic acid by
esterification in acidic ethanol. The synthetic route employing the stannane
methodology was one step shorter and offered a faster and more selective
bromination procedure.12 However, it turned out that even though the
coupling between ethyl bromoacetate and acetonyltributylstannane worked
well in preliminary unlabelled experiments run in THF, the stannane was
unstable to the esterification mixture. As ethyl bromoacetate is highly volatile,
removal of the solvent after the esterification proved very difficult, and
therefore, this approach was abandoned. Instead, we turned to the ethyl
acetoacetate approach (Scheme 1), by which the condensation reaction could
be made in the same solvent as the esterification. Even though this was a
longer route, it was in part negated by careful optimization of the reaction
conditions for all three steps leading to levulinic acid. As part of the
optimization procedure we included microwave-assisted chemistry, which has
gained significant attention lately with numerous examples from a wide range
of application areas.13 Starting from bromo[1-14C]acetic acid (5) (74%
purityy), less than an hour of microwave irradiation in all was needed for
the three steps leading to [1-14C]levulinic acid (7), which was obtained in 73%
y The acid/ester mixture of bromo[1-14C]acetic acid used as starting-material was 74% pure according to
HPLC-analysis. Although the extra impurities obviously could interfere in the synthesis, no simple
procedure for purifying bromoacetic acid or the corresponding ethyl ester exists. Instead, we were confident
that the impurities could be separated at the final purification stage.
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 627–634

630
T. PERSSON ET AL.
O
O
O
O
O
14C
OEt
O
14C
5
O
14C
HCl
NaOEt
EtOH
EtO
Br
Br
OEt
OH
OEt
EtOH
6
HCl (aq)
Br
Br
Br
O
14C
O
14C
Br
H2SO4
Br2, HBr
CH2Cl2
O
OH
OH
14C
9
8
7
O
O
O
Scheme 1. Synthesis of (5Z)-4-bromo-5-(bromomethylene)-2(5H)-[2-14C]fura-
none (9)
yield and 53% radiochemical purity. The purity at this stage, although
seemingly on the low side, was quite acceptable considering the purity of the
starting material. The [1-14C]levulinic acid obtained was used without further
purification. However, this direct approach made the subsequent bromination
difficult to control due to the impurities present. The reaction had to be
carefully monitored by HPLC analysis to ensure an optimum distribution of
3,5-dibromo[1-14C]levulinic acid (8) in preference to mono- and tri-substituted
analogues as well as the 5,5-disubstituted analogue. To facilitate the final
reaction step, some of the polar impurities were separated at this stage by
solid-phase extraction of the reaction mixture. This provided 8 (64%) with a
radiochemical purity of 37%, which is satisfying considering the distribution
of formed products in the reaction.
The lactonization and oxidation step was performed in concentrated sulfuric
acid. We found that most of the impurities did not propagate through this
step, the reason most likely being that the brominated analogues are not
lactonized or oxidized in a similar manner. In preliminary experiments, we
observed that a bromide shift was seen under certain reaction conditions, so
despite the lack of 5,5-dibromolevulinic acid in the mixture, the final product
rearranged to the undesired 5-(dibromomethylene)-2(5H)-[2-14C]furanone.
This has been described in the literature, when 20% oleum was used as
oxidizing agent.10 On the other hand, we observed this bromide shift during
early attempts to promote the efficiency in the reaction by microwave heating
at higher temperatures. The time of heating together with the temperature was
found to be critical to the regioselectivity, and therefore, termination of the
reaction was made by simply pouring the reaction mixture onto ice to impede
any bromide shifts. Using this approach, a clean reaction was observed
making it straightforward to purify the product by HPLC. It was very
surprising though to discover that the product disappeared when evaporating
to dryness in vacuo. Instead, 9 was concentrated by solid-phase extraction and
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 627–634

SYNTHESIS OF CARBON-14
631
obtained in 13% yield with a radiochemical purity of >97%, the only
impurity seemingly being the 5-(dibromomethylene)-2(5H)-[2-14C]furanone.
The overall yield of 9 starting from 5 was 5.9% (7.7% corrected for the purity
of the starting material) and the specific activity was 57 mCi/mmol as
determined by mass spectroscopy, which also confirmed the identity of the
compound. All other identifications were made by HPLC based on
characterisation performed during unlabelled optimization experiments using
NMR and MS. No assignment of stereochemistry concerning the exocyclic
bond in 9 was made, but characterisation on similar compounds and 4 implies
it is the Z-isomer.14
Conclusion
Quorum sensing inhibitor (5Z)-4-bromo-5-(bromomethylene)-2(5H)-furanone
(4), labelled with carbon-14 in the carboxylic position, has been synthesised
from bromo[1-14C]acetic acid (5) in 7.7% corrected overall yield with a
radiochemical purity of >97% and a specific activity of 57 mCi/mmol. The
initial three steps of the synthesis were accelerated by microwave heating
offering a fast route to carbon-14 labelled levulinic acid from simple building
blocks. Levulinic acid, a well-used synthetic intermediate, was converted in
two steps into (5Z)-4-bromo-5-(bromomethylene)-2(5H)-[2-14C]furanone (9).
The successful implementation of microwave heating in this and other projects
underlines the value of this technology in the radiochemistry laboratory.
Currently, compound 9 is being applied in bioassays to elucidate the biological
activity and mechanism-of-action by this potent quorum sensing inhibitor.
Experimental
General
Bromo[1-14C]acetic acid (specific activity: 58 mCi/mmol) was supplied by
Amersham Biosciences, UK and had a 74% radiochemical purity (as a free
acid/ethyl ester mixture) at the time of use. All other reagents and solvents
were of analytical grade and used without further purification. HPLC was
performed using a Merck Hitachi Intelligent Pump L6200A equipped with a
Supertherm Column Thermostat (Mikrolab Aarhus) (set at 408C), a Merck
LC Organizer with a Rheodyne injector, and a Merck Hitachi UV Detector
L4000 (detection at 220 nm). Detection of carbon-14 was performed on a
Canbarra Packard flow detector 500TR.
Analytical HPLC was performed on a RP C18 column (4.6 ꢀ 250 mm, 5 mm,
(
OdDMeSi 120 A, Novo Nordisk) with a flow of 1.0 ml/min using one of the
two following systems. System 1: 100% ! 40% A 0-30 min and system 2: 80%
A 0-30 min followed by 0% A for 10 min (A: 10% acetonitrile in 0.1% aq.
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 627–634

632
T. PERSSON ET AL.
TFA; B: 90% acetonitrile in 0.1% aq. TFA). Purifications were performed
using system 2 employing a 10 ꢀ 250 mm column (5.0 ml/min).
Radioactivity measurements were performed on a Packard Tri-Carb 1000
liquid scintillation analyzer using Ultima FloTM M (PerkinElmer) as liquid
scintillation cocktail. Specific activity was determined on a ThermoFinnigan
LCQ mass spectrometer operating in the APCI mode. Microwave heating was
performed on an EmrysTM Creator from Personal Chemistry, Sweden.
Concentrations were made on a rotary evaporator at a temperature below
408C.
1,4-Diethyl-acetyl[1-14C]succinate (6)
A solution of bromo[1-14C]acetic acid (5) (9.88 mCi, 74% radiochemical
purity, specific activity: 58 mCi/mmol) in abs. EtOH (2 ml) was added to a
microwave vessel containing abs. EtOH (0.5 ml) acidified with AcCl (2 ml) and
the mixture was heated by microwaves at 1108C for 0.5 h. Then, ethylacetoa-
cetate (0.18 mmol, 23 ml) dissolved in an EtOH/NaOEt-solution (0.20 mmol,
0.5 ml, 2.4 M) was added in one portion, and the mixture was heated by
microwaves at 60–658C for a total of 16 min (4+4+8 min). The solvent was
removed using a gentle stream of nitrogen giving 6 as an oily residue with a
radiochemical purity of 39% (System 1).
[1-14C]Levulinic acid (7)
1,4-Diethyl-acetyl[1-14C]succinate (6) (9.20 mCi, 62 mmol) was emulsified with
2 M HCl (3 ml) carefully providing for the oil being covered by the solvent.
The mixture was heated by microwaves at 1308C for 10 min. Evaporation to
dryness afforded 7 (7.17 mCi, 73% from 5) as a semi-solid with a
radiochemical purity of 53% (System 1).
3,5-Dibromo[1-14C]levulinic acid (8)
A solution of [1-14C]levulinic acid (7) (7.17 mCi, 66 mmol) in dichloromethane
(1 ml) was cooled to 08C. Then, HBr (10 ml, 33% in AcOH) was added
followed by dropwise addition of Br2 (10 ml) dissolved in dichloromethane
(0.5 ml). The reaction mixture was heated to room temperature and then to
reflux while following the progress of the reaction by HPLC analysis (System
1). Additional equivalents of HBr and Br2 were added to the reaction mixture,
whenever the reaction seemed to slow down. After 10 h of reflux, the loss of
product seemed to be increasing faster than the formation and the reaction
was stopped. In total, 72 ml Br2 (1.4 mmol) and 130 ml HBr (33% in AcOH)
were added. The reaction mixture was eluted through a Sep-Pak1 cartridge
(Waters, SiO2, 2 g) and concentrated to give 8 (4.59 mCi, 64%) as a colorless
oil with a radiochemical purity of 37% (System 1).
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 627–634

SYNTHESIS OF CARBON-14
633
(5Z)-4-Bromo-5-(bromomethylene)-2(5H)-[2-14C]furanone (9)
3,5-Dibromo [1-14C]levulinic acid (8) (4.59 mCi, 29 mmol) was dissolved in
sulfuric acid (95–97%, 1 ml) and the reaction mixture heated at 1008C for
precisely 10 min in an oil bath. Then, the mixture was poured onto crushed ice
followed by extraction with dichloromethane (4 ꢀ 5 ml). The combined
organic phases were eluted through a Sep-Pak1 cartridge (Waters, SiO2,
0.5 g) and concentrated giving crude 9 (1.77 mCi) with a radiochemical purity
of 60% (System 1). The crude product was purified by HPLC (System 2). As
the product was found to disappear when concentrated in vacuo, the combined
fractions containing 9 were diluted with water to an acetonitrile content of 5%
and then appliedz to a pre-activated Sep-Pak1 cartridge (Waters, RP-C18, 2 g)
followed by washing with water (20 ml). Eluting with acetonitrile (20 ml)
afforded 9 (0.58 mCi, 13%) in 5.9% overall yield (7.7% corrected for the
purity of the starting material and the final product). The radiochemical purity
was greater than 97% as determined by HPLC (System 2) and the specific
activity 57 mCi/mmol as determined by mass spectrometry.
Acknowledgements
We are grateful to Dr Henrik Olsen for performing the mass spectroscopy
analysis.
References
1. Whitehead NA, Barnard AML, Slater H, Simpson NJL, Salmond GPC. FEMS
Microbiol Rev 2001; 25: 365–404.
2. Hjelmgaard T, Persson T, Rasmussen TB, Givskov M, Nielsen J. Bioorg Med
Chem 2003; 11: 3261–3271.
3. Olsen JA, Severinsen R, Rasmussen TB, Hentzer M, Givskov M, Nielsen J.
Bioorg Med Chem Lett 2002; 12: 325–328.
4. Kazlauskas R, Murphy PT, Quinn RJ, Wells RJ. Tetrahedron Lett 1977; 37–40.
5. Ohta K. Agric Biol Chem 1977; 41: 2105–2106.
6. Givskov M, Nys Rd, Manefield M, Gram L, Maximilien R, Eberl L, Molin S,
Steinberg P, Kjelleberg S. J Bacteriol 1996; 178: 6618–6622.
7. Hentzer M, Wu H, Andersen JB, Riedel K, Rasmussen TB, Bagge N, Kumar N,
Schembri MA, Song Z, Kristoffersen P, Manefield M, Costerton JW, Molin S,
Eberl L, Steinberg P, Kjelleberg S, Hꢀiby N, Givskov M. EMBO J 2003; 22:
3803–3815.
8. Manefield M, Rasmussen TB, Henzter M, Andersen JB, Steinberg P, Kjelleberg S,
Givskov M. Microbiol 2002; 148: 1119–1127.
.
9. Wolff L, Rudel F. Ann Chemie 1896; 293–294: 183–205.
10. Wells PR. Aust J Chem 1963; 16: 165–169.
z Since the product slowly elutes from the Sep-Pak cartridge when it is applied in 5% aq. acetonitrile, the
cartridge was changed to a new one, whenever the product could be detected in the effluent.
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 627–634

634
T. PERSSON ET AL.
11. Joubert C, Beney C, Marsura A, Luu-Duc C. J Label Compd Radiopharm 1995;
36: 745–754.
12. Zimmer R, Collas M, Roth M, Reibig H-U. Liebigs Ann Chem 1992; 92: 709–714.
13. Lidstrom P, Tierney J, Wathey B, Westman J. Tetrahedron 2001; 57: 9225–9283.
.
14. Nys Rd, Wright AD, Konig GM, Sticher O. Tetrahedron 1993; 49: 11213–11220.
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 627–634