Design and synthesis of fluorescent pilicides and curlicides: Bioactive tools to study bacterial virulence mechanisms

Article abstract of DOI:10.1002/chem.201103936

Pilicides and curlicides are compounds that block the formation of the virulence factors pili and curli, respectively. To facilitate studies of the interaction between these compounds and the pili and curli assembly systems, fluorescent pilicides and curl

Full text of DOI:10.1002/chem.201103936

DOI: 10.1002/chem.201103936
Design and Synthesis of Fluorescent Pilicides and Curlicides: Bioactive Tools
to Study Bacterial Virulence Mechanisms
Erik Chorell–,[a, c] Jerome S. Pinkner,[b, c] Christoffer Bengtsson,[a] Sofie Edvinsson,[a]
Corinne K. Cusumano,[b, c] Erik Rosenbaum,[a] Lennart B. ꢀ. Johansson,[a]
Scott J. Hultgren,*[b, c] and Fredrik Almqvist*[a, d]
4522
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 4522 – 4532

FULL PAPER
Abstract: Pilicides and curlicides are
compounds that block the formation of
the virulence factors pili and curli, re-
spectively. To facilitate studies of the
interaction between these compounds
and the pili and curli assembly systems,
fluorescent pilicides and curlicides
have been synthesized. This was ach-
ieved by using a strategy based on
structure–activity knowledge, in which
key pilicide and curlicide substituents
on the ring-fused dihydrothiazolo 2-
pyridone central fragment were re-
placed by fluorophores. Several of the
resulting fluorescent compounds had
improved activities as measured in pili-
and curli-dependent biofilm assays. We
created fluorescent pilicides and curli-
cides by introducing coumarin and 4,4-
difluoro-4-bora-3a,4a-diaza-s-indacene
(BODIPY) fluorophores at two posi-
tions on the peptidomimetic pilicide
and curlicide central fragment. Fluores-
cence images of the uropathogenic Es-
cherichia coli (UPEC) strain UTI89
grown in the presence of these com-
pounds shows that the compounds are
strongly associated with the bacteria
with a heterogeneous distribution.
Keywords: antivirulence · biologi-
cal activity · coumarin · fluores-
cence · structure–activity relation-
ships
Introduction
been dissected.[6–8] Our understanding of the structural basis
of CUP pili biogenesis has led to the design of pilicides that
bind to the chaperone and block critical functions thus pre-
venting pilus assembly.[9–11] Numerous (>100) CUP systems
are now known to be encoded in gram-negative genomes[12]
but little is known about their function. A typical E. coli
genome encodes approximately 10 such systems.[13] Further
detailed studies of the function and regulation of CUP pili
and other extracellular fibers are an important route to un-
derstanding the bacterial adaptation and survival strategies
that may be particularly relevant to human infections and
providing targets for the development of new therapeutics.
Here we describe the development of new compounds that
will serve as a strong foundation to support investigation of
novel anti-fiber therapeutics targeting critical assembly and
adhesion functions of fibers required for the determination
of tropism and the organization of bacterial communities
during infection.
Multidrug resistant (MDR) bacterial strains present a grow-
ing global health problem. As a consequence, the search for
new antibacterial agents and new methods to deal with bac-
terial resistance is urgent.[1] Toward this end, understanding
the details of the uropathogenic E. coli (UPEC) pathogenic
cascade is revealing ways to target critical pathways to de-
velop anti-virulence therapeutics. We have discovered that
type-1 pili, an adhesive pili assembled by the chaperone/
usher pathway (CUP), play an essential role in invasion of
bladder cells and in the formation of biofilm-like intracellu-
lar bacterial communities (IBCs) that protect bacteria from
host defenses and antibiotics.[2–5] Further, CUP pili play
a critical role in biofilm formation, mediating not only inter-
actions with host tissue, but also colonization of catheters
and other surfaces in nosocomial settings. Also, an amyloid
fiber called curli is critical in UPEC biofilm formation and
the molecular machine that mediates curli assembly has
Ring-fused 2-pyridones are peptidomimetics that can
target protein–protein interactions in macromolecular as-
sembly. We have previously shown that ring-fused dihydro-
thiazolo 2-pyridones (1) provide an excellent central frag-
ment for design and synthesis of compounds that block the
formation of pili and curli.[10,11,14] Pilicides (2a,b) are com-
pounds that block pilus biogenesis (exemplified in UPEC),
whereas curlicides (such as 3) prevent curli fiber biogenesis.
Development of traceable pilicides and curlicides could po-
tentially be obtained using a biomolecular labeling strategy,
for example, by radiolabeling or by the introduction of a flu-
orescent label. In the latter case, a fluorescent probe is usu-
ally attached by using a linker to the biomolecule to avoid
any interference with the biomolecular interactions. The low
molecular weight of ligands such as the pilicides 2a,b and
curlicide 3, implies that this technique could change the
overall molecular composition to a great extent and thereby
potentially reduce the bioactivity of these compounds. An
alternative approach would be to replace key substituents
for bioactivity by a fluorophore. To increase the likelihood
of succeeding by using this approach, and thus both retain
the biological effect and gain fluorescent properties, the
structure–activity knowledge on the central fragment could
[a] Dr. E. Chorell–, C. Bengtsson, S. Edvinsson, Dr. E. Rosenbaum,
Prof. Dr. L. B. ꢁ. Johansson, Prof. Dr. F. Almqvist
Department of Chemistry
Umeꢂ University, 90187 Umeꢂ (Sweden)
Fax : (+46)907-867-655
E-mail: fredrik.almqvist@chem.umu.se
[b] J. S. Pinkner, Dr. C. K. Cusumano, Prof. Dr. S. J. Hultgren
Department of Molecular Microbiology
Washington University School of Medicine
St.Louis, MO 63110 (USA)
E-mail: hultgren@borcim.wustl.edu
[c] Dr. E. Chorell–, J. S. Pinkner, Dr. C. K. Cusumano,
Prof. Dr. S. J. Hultgren
Center for Womenꢃs Infectious Disease Research
Washington University School of Medicine
St.Louis, MO 63110 (USA)
[d] Prof. Dr. F. Almqvist
Umeꢂ Centre for Microbial Research
Umeꢂ University, 90187 Umeꢂ (Sweden)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201103936.
Re-use of this article is permitted in accordance with the Terms and
Conditions set out at http://onlinelibrary.wiley.com/journal/10.1002/AHCTUNG-
TRENNUNG(ISSN)1521-3765/homepage/2111_onlineopen.html.
Chem. Eur. J. 2012, 18, 4522 – 4532
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4523

F. Almqvist, S. J. Hultgren et al.
be used for both the choice of fluorophore and its position-
ing on the central fragment. One potential problem with
this method is a higher probability for fluorescent quenching
of the fluorophore due to its close proximity to the bioactive
central fragment. Even so, exchange of one of the substitu-
ents on the peptidomimetic central fragment 1 to a fluoro-
phore could render attractive compounds for uptake/distri-
bution studies, development of competition-based assays,
Fçrster resonance energy transfer (FRET) studies on bind-
ing interactions, and to specifically image conserved pili and
curli assembly machineries in bacterial populations. Initial
studies of the structure–activity relationships on the ring-
fused dihydrothiazolo 2-pyridone central fragment have
shown that the C7 and C8 positions are highly important for
bioactivity and should preferably carry larger lipophilic sub-
stituents.[10,11,15–17] Consequently, we have in the present
study exchanged the substituents in the C7 and C8 positions
on the central fragment 1 with coumarin (4) or BODIPY
(5) fluorophores (Figure 1). The use of these particular fluo-
this approach, coumarins could also be introduced directly
on a bromomethyl-substituted central fragment using depro-
tonated 4-methyl coumarins. In total, 14 new fluorophore-
substituted derivatives of the central fragment 1 have been
synthesized and the photophysical measurements of these
compounds revealed compounds with high quantum yields.
In addition, biological evaluation of these compounds as pi-
licides and curlicides showed a great biological effect of sev-
eral compounds, with some being both potent inhibitors of
pili- and curli-dependent biofilm formation and having fluo-
rescent properties. Finally, treatment of the UPEC strain
UTI89 with the compounds under pili producing conditions
shows that the compounds are associated to the bacteria
and seem to discriminate between different bacteria in
a population.
Results and Discussion
Coumarins substituted with electron-donating groups in po-
sition 7 such as 7-methoxy and 7-diethylamino coumarins
are frequently used fluorophores (Figure 1). The linker to
the pilicide/curlicide central fragment was preferred through
the 4 position on the coumarins to resemble the geometry of
the C7 naphthyl substituent in 2 and 3. Consequently, the 7-
methoxy coumarin-4-yl acetic acid was first coupled to its
corresponding acyl Meldrumꢃs acid derivative 8 using stan-
dard conditions.[10] On the basis of previous structure–activi-
ty relationships of the pilicide/curlicide central fragment,
a phenyl,[9,15] a 3-trifluoromethylphenyl,[10,22] a 2-thiophen-
yl,[17] and a cyclopropyl[9,15] were selected as substituents on
the D2-thiazolines (9a–d). Reacting 8 with the 9a–d in the
acyl ketene imine cyclocondensation gave coumarin-substi-
tuted thiazolo ring-fused 2-pyridones 10a–d in 68–86%
yield (Scheme 1). Compounds 10a–d were next subjected to
hydrolysis to give the corresponding carboxylic acids 11a–d
in 53–66% yield.
Figure 1. The dihydrothiazolo ring-fused 2-pyridone central fragment
1 (R=H or Li for biological activity), pilicide (2a,b), curlicide (3), cou-
marin fluorophore 4, BODIPY fluorophore 5. Synthesis of the thiazolo
ring-fused 2-pyridone central fragment (1a) is performed by using 2-thia-
zolines (6) and Meldrumꢃs acid derivatives (7).
rophores could be justified by their good (compound 4) to
high (compound 5) quantum yields, absorption/emission
wavelengths, lipophilicity, lack of net ionic charge, photosta-
bility, different emission colors and relatively small size.[18–20]
The synthesis of the ring-fused dihydrothiazolo 2-pyridone
central fragment is based upon an acyl-ketene imine cyclo-
condensation between D2-thiazolines 6 and acyl Meldrumꢃs
acid derivatives 7 (Figure 1).[21] Carboxylic acids can be used
as starting materials in the synthesis of both 6 and 7 and car-
boxylic acid functionalized fluorophores have therefore
been used as building blocks for further incorporation in the
C7 and C8 positions on the central fragment. In addition to
Scheme 1. a) TFA, dichloroethane, microwave irradiation (MWI): 1208C,
140 s (10a–d: 77%, 86%, 77 and 68%, respectively); b) i) 0.1m aq LiOH,
THF/MeOH (4:1); ii) AcOH (11a–d: 66, 58, 53 and 66%, respectively).
The photophysical evaluation of 11a–d showed relatively
low quantum yields of fluorescence (FF, Table 1). The C8
phenyl-substituted 11a displayed highest quantum yield
(FF =5%), followed by the C8 3-trifluomethylphenyl-substi-
tuted 11b (FF =1%), the C8 2-thiophenyl-substituted 11c
4524
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 4522 – 4532

Design and Synthesis of Fluorescent Pilicides and Curlicides
FULL PAPER
Table 1. The fluorophore-substituted compounds photophysical properties and abilities to inhibit pili- and curli-dependent biofilm formation.
[a]
[a]
[b,c]
[b,d]
ID
R1 (C8)
R2 (C7)
X
EC50
EC50
labs
lfl
Quantum Yield[b]
Pili [mm]
Curli [mm]
[nm]
[nm]
FF [%] (lex nm)
11a
-S-
>200
NA[e]
328
394
5 (330)[f]
1 (330)[f]
11b
11c
11d
16a
16b
16c
16d
-S-
-S-
-S-
-S-
-S-
-S-
-S-
>200
>200
>200
65
NA[e]
NA[e]
NA[e]
175
330
328
328
329
327
393
392
420
396
413
430
393
474
484
0.7 (330)[f]
0.4 (330)[f]
0.6 (355)[f]
0.5 (330)[f]
15 (390)[g]
6 (330)[f]
156
NA[e]
18
25
12
NA[e]
23
28
-S-
-S-
5
4
17
14
396
506
478
524
11 (346)[f]
10 (470)[h]
Chem. Eur. J. 2012, 18, 4522 – 4532
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4525

F. Almqvist, S. J. Hultgren et al.
Table 1. (Continued)
[a]
[a]
[b,c]
[b,d]
ID
R1 (C8)
R2 (C7)
X
EC50
EC50
labs
lfl
Quantum Yield[b]
Pili [mm]
Curli [mm]
[nm]
[nm]
FF [%] (lex nm)
31
-S-
14
12
502
514
67 (470)[h]
34
37
-S-
10
13
24
14
498
497
509
531
11 (480)[h]
-O-
27 (480)[h]
38
3
-S(O)-
-S-
29
17
40
38
498
–
516
–
71 (480)[h]
–
[a] Estimated from 16–32 data points on every concentration. [b] All substances were dissolved in DMSO and subsequently diluted in phosphate buffer
at pH 7.0. The samples DMSO concentrations never exceed 5 wt%. The sample concentrations in the DMSO stock solutions are adjusted so that the
final samples never have a peak absorbance higher than 0.1. [c] Wavelengths of the peak absorption. [d] The peak fluorescence. [e] Not active. [f] Refer-
ence: POPOP in MeOH. [g] Reference: Perylene in cyclohexane. [h] Reference: Rhodamine 6G in water.
(FF =0.7%), and finally the C8 cyclopropyl-substituted 11d
(FF =0.4%). These results may be due to a quenching effect
of the fluorophore by the neighboring dihydrothiazolo ring-
fused 2-pyridone pilicide/curlicide central fragment. In an
attempt to circumvent this, the linker between the fluoro-
phore and the 2-pyridone central fragment was increased by
a one carbon extension. This was realized by using depro-
tonated 4-methyl coumarins as nucleophiles on bromometh-
yl-substituted 2-pyridone central fragments. By using this
strategy the introduction of both a 7-methoxy-substituted
coumarin and a 7-diethylamine-substituted coumarin could
be accomplished (Scheme 2). The intermediate bromometh-
yl-substituted 2-pyridone central fragments (13a and 13b)
were synthesized following a previously published proce-
dure.[23] Addition of lithiated coumarins (14a and 14b) to
13a and 13b generated the one carbon extended coumarin-
substituted ring-fused 2-pyridones 15a–d in 74–83% yield.
Subsequent hydrolysis rendered 16a–d in high yields (81–
91%; Scheme 2).
The one carbon extended linker did not influence the
Scheme 2. a) TFA, dichloroethane, MWI: 1208C, 140 s (13a,b: 92 and
quantum yield (16a, FF =0.6% and 16b FF =0.5%;
87%, respectively); b) 14a or 14b, LiHMDS, THF, ꢀ358C, 1 h (15a–d:
74, 82, 83 and 75%, respectively); c) For 16a,c: 0.1m aq LiOH, THF/
MeOH (4:1) 1.5 h; ii) AcOH (16a,c: 86 and 82%, respectively), For
16b,d: i) LiBr, TEA, MeCN (+2 v/v% H2O), RT, 3 h; ii) AcOH (16b,d:
91 and 81%, respectively).
Table 1). However, replacing the 7-(methoxy)coumarin by
a 7-(diethylamino)coumarin as in 16c and 16d increased the
quantum yields (FF =15 and 6%, respectively, Table 1) and
gave C7 coumarin-substituted compounds with useful fluo-
4526
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 4522 – 4532

Design and Synthesis of Fluorescent Pilicides and Curlicides
FULL PAPER
rescent properties. As a consequence, the 7-(diethylamino)-
coumarin was used in the introduction of a coumarin in the
C8 position of the pilicide/curlicide central fragment. The
coumarin-substituted D2-thiazoline 20 was synthesized in
three steps starting from 14b and ethyl 2-bromoacetate
(Scheme 3).[24] Subsequent reaction with acyl Meldrumꢃs
been synthesized in 21% yield.[25] From 24, the introduction
of a BODIPY substituent in the C8 position of the pilicide/
curicide central fragment could be pursued. Coupling 24
using a standard coupling procedure with methylester-pro-
tected cysteine gave the intermediate 25 in 64% yield
(Scheme 4). Ring closure of 25 to give 26 followed by acyl-
Scheme 4. a) i) (COCl)2, CH2Cl2, RT, 17 h; ii) TEA, cysteine methyl ester,
CH2Cl2, 08C to RT, 4.5 h, 64%; b) i) TiCl4, CH2Cl2, 08C to RT, 4.5 h;
ii) BF3·OEt2, CH2Cl2, RT, 40 min, 75%; c) 21, TFA, dichloroethane,
MWI: 1208C, 140 s, 64%; d) i) LiI, pyridine, MWI: 1408C, 15 min;
ii) TEA, BF3·OEt2, dichloroethane, 808C, 15 min, 29%.
ketene imine cyclocondensation with Meldrumꢃs acid deriva-
tive 21 gave the BODIPY-substituted dihydrothiazolo ring-
fused 2-pyridone 27 in 64% yield. The following hydrolysis
proved to be problematic and the harsh conditions needed
for this transformation ultimately gave 28 in 29% yield as
a racemate (Scheme 4). However, on the basis of previous
reports showing that both enantiomers of the pilicide central
fragment are biologically active, it was still entirely possible
that racemic 28 could display interesting bioactivity.[9,26]
The photophysical evaluation of 28 gave a quantum yield
of 10%, which is surprisingly low for a BODIPY-substituted
compound (Table 1). We hypothesized that photo-quenching
was a possible reason for this and thus the use of an aryl
linker between the fluorophore and the central fragment
could circumvent the problem. This would not only increase
the distance between the fluorophore and the pilicide/curli-
cide central fragment but it would also restrict the rotation
of the 1,3,5,7-tetramethyl-substituted BODIPY, which is
known to generate higher quantum yields.[20] Exchange of
the methylene linker for a phenyl results in the need for a re-
vised synthetic approach. A recent publication shows that it
is possible to introduce a benzoic acid in the C8 position of
the pilicide/curlicide central fragment by Suzuki–Miyaura
couplings.[27] From this benzoic acid derivative, the transfor-
mation into the desired C8 BODIPY-substituted central
fragment seemed feasible. Consequently, 2-pyridone 29 was
treated with oxalyl chloride followed by reaction with 2,4-di-
methylpyrrole in the presence of BF3·OEt2 and triethyl-
Scheme 3. a) LiHMDS, DMPU, ethyl 2-bromoacetate, THF, ꢀ208C,
35 min, 77%; b) i) 1m (aq) LiOH, THF, RT, 16 h, ii) HBTU, 18, DMF,
RT, 17 h, 60%; c) TiCl4, CH2Cl2, 08C to RT, 16 h, 40%; d) 21, TFA, di-
chloroethane, MWI: 1208C, 3 min, 73%; e) 0.1m aqueous LiOH, THF,
RT, 1 h; ii) AcOH, 91%.
acid derivative 21 rendered the C8 coumarin-substituted
central fragment 22 in 73% yield (Scheme 3). After hydroly-
sis of 22 into the target compound 23 the photophysical
properties were evaluated. From this a quantum yield com-
parable with the other 7-(diethylamino)coumarin-substituted
compounds was observed for 23 (FF =11%, Table 1). Thus,
compounds with a coumarin fluorophore in both the C7 and
C8 positions of the pilicide/curlicide central fragment have
been synthesized.
To increase the probability of identifying a bioactive com-
pound with useful fluorescence properties, the possibility of
introducing a different fluorophore on the central fragment
was investigated. 4,4-difluoro-4-bora-3a,4a-diaza-s-indacene
(BODIPY) is a known fluorophore that normally gives high
quantum yields, carries no net charges, is relatively insensi-
tive to the choice of solvent, and should give a different
emission color as compared to the coumarins.[20] The synthe-
sis of the BODIPY core structure is often accompanied by
low yields. The desired 8-propanoic acid-functionalized
1,3,5,7-tetramethyl-substituted BODIPY (24) has previously
Chem. Eur. J. 2012, 18, 4522 – 4532
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4527

F. Almqvist, S. J. Hultgren et al.
amine to give the desired C8 BODIPY-substituted central
fragment 30 in 15% yield. Subsequent hydrolysis was
straightforward for this BODIPY derivative, giving the cor-
responding carboxylic acid 31 in 84% yield (Scheme 5). As
hypothesized, introduction of the phenyl spacer in the C8
position increased the quantum yield of 31 to a satisfactory
67% (Table 1).
Scheme 5. a) i) (COCl)2, DMF, CH2Cl2, RT, 1 h; ii) TEA, BF3·OEt2, 2,4-
dimethylpyrrole, dichloroethane, MWI: 1408C, 50 min, 15%; b) 0.1m
aqueous LiOH, THF, RT, 1 h; ii) H+, 84%.
Scheme 6. a) DCC, Meldrumꢃs acid, DMAP, CH2Cl2, 08C RT, 8 h, 82%;
b) 9a, TFA, dichloroethane, MWI: 1208C, 140 s, 85%; c) i) 35,
(NH4)2MoO4, toluene, reflux soxhlet (3MS), 6 h; ii) 32, TFA, toluene,
reflux soxhlet (3MS), 2 h, 73%; d) starting from 34, mCPBA, CH2Cl2,
RT, 15 min, 80%; e) i) 0.1m aq LiOH, THF/MeOH (4:1); ii) AcOH,
88%; f) i) LiBr, TEA, MeCN (+2 v/v% H2O), RT, 3 h; ii) AcOH, 88%.
With these two C8 BODIPY-substituted compounds (28
and 31) in hand, introduction of BODIPY in position C7 on
the central fragment was next investigated. Coupling the
BODIPY carboxylic acid 24 with Meldrumꢃs acid using stan-
dard coupling conditions and a simple MeOH trituration as
purification gave the BODIPY-substituted acyl Meldrumꢃs
acid derivative 32 in 82% yield (Scheme 6). From this versa-
tile intermediate the following acylketene imine cyclocon-
densation to the C7 BODIPY substituted pilicide/curlicide
central fragment 33 was performed in 88% yield. Sequential
hydrolysis gave the corresponding carboxylic acid 34
(Scheme 6). The photophysical properties of 34 suggest, in
analogy with the coumarins, that the BODIPY fluorophore
also seems to be partially quenched in this position (FF =
11%, Table 1).
Even though the quantum yields for many of the synthe-
sized compounds are acceptable and, in the case of 31, high,
the underlying reason for the observed photo-quenching
was still unknown. Thus, we further investigated the role of
the sulfur in position 1 of the pilicide/curlicide central frag-
ment on the photophysical properties. The sulfur in the cen-
tral fragment was first replaced by an oxygen by reacting 32
and serine derivative 35 in a previously reported two-step
one-pot procedure to give 36 in 73% yield (Scheme 6).[28,29]
Hydrolysis of 36 was straightforward and gave 37 in 88%
yield. The quantum yield of 37 (FF =29%, Table 1) was
almost three times greater than that of its sulfur-containing
counterpart 34 (FF =11%, Table 1). This suggests that the
sulfur at position 1 of the pilicide/curlicide central fragment
is at least partially responsible for the photo-quenching. To
further verify this, the electronic properties of the sulfur
were altered through oxidation. Treating 34 with mCPBA
gave the sulfoxide 38 in 80% yield as a 4:1 mixture of dia-
stereomers (Scheme 6). Photophysical evaluation of sulfox-
ide 38 showed a great increase of quantum yield (FF =71%,
Table 1) compared to the parent sulfide 34 (FF =11%,
Table 1) thus supporting the theory that the sulfide in posi-
tion 1 of the central fragment is involved in the observed
partial photo-quenching.
The 14 fluorophore-substituted compounds were biologi-
cally evaluated for their ability to block pili and curli forma-
tion as measured by pili- or curli-dependent biofilm assays
on polyvinylchloride surfaces.[10,11,30] The amount of biofilm
that is formed in the E. coli clinical isolate UTI89 grown in
the presence of pilicide or curlicide, is related to the potency
of the compound in blocking piliation or curliation. The re-
sults are summarized in Table 1.
The strategy of generating fluorescent pilicides by replac-
ing key pilicide substituents with fluorophores based on
structure–activity knowledge was successful and many of the
tested compounds were both highly potent and fluorescent.
Several of the compounds were able to inhibit formation of
both pili- and curli-dependent biofilm, which is highly valua-
ble from a therapeutic perspective. From the pili-dependent
biofilm evaluation, C8 substitution with a methylene spacer
using both coumarin (23) and BODIPY (28) fluorophores
gave the most potent pilicides with estimated half-maximal
effective concentrations (EC50) of 5 and 4 mm, respectively.
Interestingly, compound 28 is also among the most potent
inhibitors of curli-dependent biofilm formation with an esti-
4528
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 4522 – 4532

Design and Synthesis of Fluorescent Pilicides and Curlicides
FULL PAPER
mated EC50 of 14 mm and is thus a potent dual pilicide–curli-
cide. This should be compared with the best reported di-sub-
stituted compound 3 (known as FN075), which exhibited
EC50 values of 17 mm in pili-dependent biofilm inhibition
and 38 mm in curli-dependent biofilm inhibition (Table 1).
The phenyl spacer in 31 resulted in a slight decreased pili-
cide activity compared to the methylene linker in 28, but is
still in the same range as 3. The C7 BODIPY-substituted de-
rivatives all displayed good inhibitory properties of pili- and
curli-dependent biofilm formation. The sulfide (34) and
oxygen analogue (37) are more potent than the sulfoxide
(39) both as pilicides and curlicides (Table 1). However, the
fact that the sulfoxide retains much of the biofilm-inhibition
activity might be of importance from a therapeutic and drug
metabolism perspective. For the coumarin-substituted com-
pounds with ethylene linker (16a–d), the 7-(diethylamino)-
coumarins 16c,d are more active pilicides than the 7-(me-
thoxy)coumarins 16a,b (Table 1). Furthermore, the com-
pounds having an ethylene linker are more potent pilicides
than the compounds with a methylene linker (16a and 16b
compared with 11b and 11d). None of the four 7-(methoxy)-
coumarins with a methylene linker 11a–d showed any activi-
ty (Table 1).
Finally, to study the compounds distribution in a bacterial
population, UPEC strain UTI89 was grown under pili-pro-
ducing conditions (24 h at 378C static) with compounds 28
and 34 and examined by fluorescence microscopy (Figure 2).
Compounds 28 and 34 appeared to bind bacteria differently,
with 34 appearing more punctate and 28 appearing more
diffuse. Further experiments are working to examine the
reason for these observed differences. We know that even
under pilus-inducing conditions, not all bacteria are express-
ing pili. Thus it is likely that the pilicide is only binding bac-
teria expressing pili. Experiments are underway to confirm
this hypothesis.
Figure 2. The UPEC strain UTI89 grown for 24 h with and without
100 mm of dual pilicide-curlicide 34 and 28. Blue is DAPI-nuclear stain,
green is the compound. A) Bacteria grown in the presence of 100 mm of
compound 34. B) Bacteria grown in the presence of 100 mm of compound
28. C) Bacteria grown without any added compound (scale bar is 2 mm).
fluorescence properties. Biological evaluation using whole
bacterial pili- and curli-biofilm assays revealed several com-
pounds that are both fluorescent and highly active. Finally,
treatment of UPEC strain UTI89 with the compounds
under pili-producing conditions shows that the compounds
are associated to the bacteria with a heterogeneous distribu-
tion over a population. The use of these compounds to
study the biological systems they interact with, for example,
the relation between the observed staining pattern of the
compounds and the heterogeneous distribution of pili pro-
duction in bacterial populations, is a matter of future studies
within our laboratories.
Conclusion
We have herein developed compounds that are both fluores-
cent and highly active inhibitors of pili- and curli-dependent
biofilm formation. The synthesis of these compounds was in-
itiated to facilitate profound studies of the pilicides and cur-
licides and to gain information about the complex systems
involved in the formation of pili and curli. To create these
fluorescent and bioactive compounds we implemented
a strategy based on structure–activity information, in which
important substituents for bioactivity on the pilicide/curli-
cide central fragment were replaced by fluorophores. Syn-
thetic methods were developed to enable a total of 14 com-
pounds with either a coumarin or a BODIPY motif in the
C7 or the C8 position. Consequently, new interesting inter-
mediates and new reactions to introduce BODIPY fluoro-
phores (e.g., via Meldrumꢃs acid derivatives and subsequent
acyl ketene cyclocondensations) have been developed.
Photo-quenching was frequently observed but its origin was
elucidated and circumvented to give compounds with useful
Experimental Section
General synthesis: All reactions were carried out under an inert atmos-
phere with dry solvents under anhydrous conditions, unless otherwise
stated. CH2Cl2 and 1,2-dichloroethane (DCE) and was distilled from cal-
cium hydride and THF was distilled from potassium. DMF was distilled
and dried over 3 ꢁ molecular sieves. All microwave reactions were car-
ried out in a monomode reactor (Smith Synthesizer, Biotage AB) using
Smith process vials sealed with a Teflon septa and an aluminum crimp
top. Reaction times refer to the irradiation time at the target temperature
not the total irradiation time. The temperature was measured with an IR
sensor. Flash column chromatography (eluent given in brackets) em-
ployed normal phase silica gel (Matrex, 60 ꢁ, 35–70 mm, Grace Amicon).
The 1H and 13C NMR spectra were recorded at 298 K with a Bruker
DRX-400 spectrometer in CDCl3 (residual CHCl3 (dH =7.26 ppm) or
CDCl3 (dC =77.0 ppm) as internal standard), [D6]DMSO (residual
[D5]DMSO (dH =2.50 ppm) or [D6]DMSO (dC =40.0 ppm) as internal
Chem. Eur. J. 2012, 18, 4522 – 4532
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4529

F. Almqvist, S. J. Hultgren et al.
standard), [D6]Acetone (residual [D5]Acetone (dH =2.05 ppm) or
[D6]Acetone (dC =29.9 ppm) as internal standard), and [D4]MeOH (re-
sidual [D3]MeOH (dH =3.30 ppm) or [D4]MeOH (dC =49.0 ppm) as inter-
nal standard). IR spectra were recorded on an ATI Mattson Genesis
Series FTIR spectrometer. Optical rotations were measured with a Perki-
nElmer 343 polarimeter at 208C. MS data were recorded using electron
spray (ES+) ionization on a Waters Micromass ZQ 2000 spectrometer.
HRMS was obtained Micromass Q-Tof Ultima mass spectrometer and
[M+H]+ molecular ions were generated by electrospray ionization.
1141, 1041, 1022, 987, 840, 705 cmꢀ1
; HRMS (EI): m/z: calcd for
C22H20NO6S: 426.1011 [M+H]; found: 426.1016.
(3R)-7-(2-(7-Methoxy-2-oxo-2H-chromen-4-yl)ethyl)-5-oxo-8-(3-(trifluor-
omethyl)phenyl)-3,5-dihydro-2H-thiazoloACTHUNTGRNEUNG[3,2-a]pyridine-3-carboxylic acid
(16a): LiOH (0.1m, 0.36 mL, 1 equiv) was added drop wise to a stirred
solution of 15a (20 mg, 36 mmol, 1 equiv) dissolved in THF/MeOH
(1 mL, 4:1). The reaction mixture was left for approximately one and
a half hours at room temperature. The solution was then concentrated,
and co-concentrated from MeOH three times. Purification by silica gel
chromatography (CH2Cl2/MeOH, 92:8, 90:8 + 2% AcOH), concentrated
and co-concentrated from CH2Cl2 and chloroform and then lyophilized
from MeCN/H2O (1:3) giving 16a as a white powder (17 mg, 86%).
[a]D =2 (c=0.51 in CHCl3); 1H NMR (400 MHz, [D4]MeOH/CDCl3
11:9): d=2.62–2.73 (m, 2H), 2.75–2.84 (m, 2H), 3.55–3.62 (m, 1H), 3.75–
3.84 (m, 1H), 3.84 (s, 3H), 5.64–5.70 (m, 1H), 5.89 (s, 1H), 6.34 (s, 1H),
6.9 (dd, J1 =2.19 Hz, J2 =8.64 Hz, 1H), 6.78 (d, J=2.42 Hz, 1H), 6.91 (t,
J=8.54 Hz, 1H), 7.44–7.51 (m, 1H), 7.52–7.64 (m, 2H), 7.67–7.73 ppm
(m, 1H); 13C NMR (100 MHz, [D4]MeOH/CDCl3 11:9): d=32.46, 32.49,
32.9 (split), 56.1, 64.7, 101.7, 111.4, 112.7, 113.3, 114.6, 115.7, 124.4 (q, J=
272.22 Hz), 125.6 (split), 126.0 (split), 127.5 (split), 130.7 (d, J=5.08 Hz),
132.2 (q, J=32.56 Hz), 134.5 (d, J=30.19 Hz), 137.5, 150.0, 154.8, 156.05,
156.07, 162.8, 163.7, 169.9, 170.0 ppm; IR: n˜ =1712, 1612, 1484, 1438,
1388, 1284, 1207, 1145, 1045, 1022, 987, 840, 705 cmꢀ1; HRMS (EI): m/z:
calcd for C27H21F3NO6S: 544.1042 [M+H]; found: 544.1049.
(3R)-7-((7-Methoxy-2-oxo-2H-chromen-4-yl)methyl)-5-oxo-8-phenyl-3,5-
dihydro-2H-thiazolo
ACHTUNGTRENNUNG
1.5 mL, equiv) was added dropwise to
1
a
(71.3 mg, 0.15 mmol, 1 equiv) dissolved in THF/MeOH (2 mL, 4:1). The
reaction mixture was stirred for two hours at room temperature before
being concentrated. Purification by silica gel chromatography (CH2Cl2/
MeOH/AcOH, 90:8:2) and then lyophilized from MeCN/H2O (1:3) to
give 11a (47 mg, 66%). [a]D =ꢀ17 (c=0.25 in CHCl3); 1H NMR
(400 MHz, [D6]DMSO): d=3.49 (dd, J1 =1.79 Hz, J2 =11.82 Hz, 1H),
3.72–3.88 (m, 6H), 5.48 (dd, J1 =1.63 Hz, J2 =9.01 Hz, 1H), 5.93 (s, 1H),
6.01 (s, 1H), 6.88 (dd, J1 =2.50 Hz, J2 =8.83 Hz, 1H), 6.96 (d, J=2.49,
1H), 7.20–7.28 (m, 1H), 7.29–7.36 (m, 2H), 7.36–7.44 ppm (m, 3H);
13C NMR (100 MHz, [D6]DMSO): d=32.4, 35.7, 56.8, 64.6, 101.9, 112.8,
113.1, 115.0, 115.3 (2C), 127.0, 129.2, 129.8 (2C), 130.8 (broad, 2C), 136.8,
149.6, 150.8, 154.0, 155.8, 160.79, 160.83, 163.3, 170.4 ppm; IR: n˜ =1712,
1608, 1554, 1484, 1423, 1388, 1257, 1207, 1025, 836, 802, 701 cmꢀ1; HRMS
(EI): m/z: calcd for C25H20NO6S: 462.1011 [M+H]; found: 462.1018.
(3R)-8-Cyclopropyl-7-(2-(7-methoxy-2-oxo-2H-chromen-4-yl)ethyl)-5-
oxo-3,5-dihydro-2H-thiazoloACHTNUTRGNEUNG[3,2-a]pyridine-3-carboxylic acid (16b): By
following a previously published procedure,[31] 15b (25 mg, 0.056 mmol)
was hydrolyzed to its corresponding carboxylic acid 16b (91% yield).
(3R)-7-((7-Methoxy-2-oxo-2H-chromen-4-yl)methyl)-5-oxo-8-(3-(trifluor-
omethyl)phenyl)-3,5-dihydro-2H-thiazoloACTHUNTGRNEUNG[3,2-a]pyridine-3-carboxylic acid
1
[a]D =ꢀ4 (c=0.3 in CHCl3/MeOH 2:1); H NMR (400 MHz, CDCl3): d=
(11b): Prepared according to the procedure described for compound 11a
starting from 10b (49 mg, 0.09 mmol, 1 equiv), giving 11b (10 mg, 58%).
[a]D =ꢀ4 (c=0.51 in CHCl3); 1H NMR (400 MHz, [D4]MeOH/CDCl3
1:1): d=3.60–3.68 (m, 1H), 3.68–3.78 (m, 3H), 3.83 (s, 3H), 5.57 (d, J=
8.11 Hz, 1H), 5.97 (s, 1H), 6.22 (s, 1H), 6.73–6.81 (m, 2H), 7.21 (d, J=
8.73, 1H), 7.39–7.61 ppm (m, 5H); 13C NMR (100 MHz, [D4]MeOH/
CDCl3 1:1): d=33.1, 35.8, 56.2, 66.4, 101.5, 112.6, 112.8, 113.3, 115.6,
115.9, 124.3 (q, J=272.34 Hz), 125.9, 126.0, 127.4 (d, J=29.1 Hz), 130.4,
131.9 (q, J=32.42 Hz), 134.2 (d, J=34.73 Hz), 137.1, 150.9, 151.0, 154.1,
155.8, 162.3, 162.7, 163.8, 172.5 ppm; IR: n˜ =1716, 1609, 1484, 1434, 1388,
0.56–0.66 (m, 2H), 0.85–1.02 (m, 2H), 1.50–1.60 (m, 1H), 2.96–3.16 (m,
4H), 3.64–3.74 (m, 1H), 3.75–3.83 (m, 1H), 3.87 (s, 3H), 5.65–5.73 (m,
1H), 6.14 (s, 1H), 6.41 (s, 1H), 6.81–6.91 (m, 2H), 7.51–7.59 (m, 1H),
8.9–9.4 ppm (bs, 1H); 13C NMR (100 MHz, CDCl3): d=7.5, 7.9, 11.0,
30.7, 30.9, 31.0, 55.8, 64.2, 101.2, 111.0, 112.4, 112.6, 113.0, 116.2, 125.1,
150.0, 154.7, 155.5, 157.9, 161.3, 162.4, 162.8, 168.4 ppm; IR: n˜ =1714,
1609, 1486, 1207, 1145, 1024, 835 cmꢀ1; HRMS (EI): m/z: calcd for
C23H22NO6S: 440.1168 [M+H]; found: 440.1169.
(3R)-7-(2-(7-(Diethylamino)-2-oxo-2H-chromen-4-yl)ethyl)-5-oxo-8-(3-
(trifluoromethyl)phenyl)-3,5-dihydro-2H-thiazoloACHTNUTRGNE[NUG 3,2-a]pyridine-3-car-
1330, 1288, 1211, 1126, 848, 705 cmꢀ1
; HRMS (EI): m/z: calcd for
boxylic acid (16c): Prepared according to the procedure described for
compound 16a, starting from 15c (20 mg, 0.036 mmol, 1 equiv) gave 16c
as a pale yellow powder (41 mg, 82%). [a]D =3 (c=0.60 in CHCl3);
1H NMR (400 MHz, [D4]MeOH): d=1.18 (t, J=6.99 Hz, 6H), 2.59–2.78
(m, 4H), 3.38 (q, J=7.03 Hz, 4H), 3.52–3.80 (m, 2H), 5.48–5.74 (m, 2H),
6.27–6.50 (m, 3H), 6.65–6.81 (m, 1H), 7.42–7.75 ppm (m, 4H); 13C NMR
(100 MHz, [D6]DMSO): d=12.2 (2C), 31.1, 32.0 (broad, split), 32.7
(broad, split), 43.9, 65.3 (broad), 79.2, 96.9, 106.7, 106.9, 108.3, 112.2,
113.7, 124.0 (q, J=272.59 Hz), 124.8, 125.1, 126.8 (d, J=9.48 Hz), 129.7
(q, J=31.61 Hz), 130.2, 134.7, 137.7, 148.7, 150.1, 152.0, 155.4, 155.8,
160.3, 160.7, 167.7 ppm; IR: n˜ =1708, 1596, 1527, 1484, 1415, 1330, 1272,
C26H19F3NO6S: 530.0885 [M+H]; found: 530.0880.
(3R)-7-((7-Methoxy-2-oxo-2H-chromen-4-yl)methyl)-5-oxo-8-(thiophen-
2-yl)-3,5-dihydro-2H-thiazoloACHTNUTRGNEUNG[3,2-a]pyridine-3-carboxylic acid (11c): Pre-
pared according to the procedure described for compound 11a starting
from 10c (72.2 mg, 0.15 mmol, 1 equiv) to give 11c (38 mg, 53%). [a]D =
ꢀ17 (c=0.69 in CHCl3); 1H NMR (400 MHz, [D6]DMSO): d=3.51 (dd,
J1 =1.77 Hz, J2 =11.61 Hz, 1H), 3.78–3.83 (m, 1H), 3.85 (s, 2H), 5.43 (dd,
J1 =1.52 Hz, J2 =9.16 Hz, 1H), 5.91 (s, 1H), 6.04 (s, 1H), 6.91 (dd, J1 =
2.56 Hz, J2 =8.88 Hz, 1H), 6.98 (d, J=2.54 Hz, 1H), 7.03–7.07 (m, 2H),
7.48 (d, J=8.88 Hz, 1H), 7.56–7.60 ppm (m, 1H). 13C NMR (100 MHz,
[D6]DMSO): d=32.6, 35.8, 56.9, 65.1, 101.9, 107.1, 112.78, 112.81, 113.1,
115.5, 126.9, 128.4, 128.9, 130.4, 136.8, 151.5, 152.4, 154.2, 155.8, 160.7,
160.9, 163.3, 170.4 ppm; IR: n˜ =1712, 1646, 1608, 1481, 1438, 1388, 1280,
1160, 1122, 1072, 802, 705 cmꢀ1
; HRMS (EI): m/z: calcd for
C30H28F3N2O5S: 585.1671 [M+H]; found: 585.1667.
(3R)-8-Cyclopropyl-7-(2-(7-(diethylamino)-2-oxo-2H-chromen-4-yl)eth-
1207, 1145, 1018, 987, 840, 705 cmꢀ1
; HRMS (EI): m/z: calcd for
C23H18NO6S2: 468.0576 [M+H]; found: 468.0587.
yl)-5-oxo-3,5-dihydro-2H-thiazolo
ACHTUNGTRENUN[NG 3,2-a]pyridine-3-carboxylic acid (16d):
By following
a
previously published procedure,[31] 15d (27.5 mg,
(3R)-8-Cyclopropyl-7-((7-methoxy-2-oxo-2H-chromen-4-yl)methyl)-5-
oxo-3,5-dihydro-2H-thiazoloACTHNUTRGNE[NUG 3,2-a]pyridine-3-carboxylic acid (11d): Pre-
0.056 mmol) was hydrolyzed to its corresponding carboxylic acid 16d
(81% yield). [a]D =5 (c=0.13 in MeOH); 1H NMR (400 MHz, CDCl3):
d=0.55–0.67 (m, 2H), 0.86–1.03 (m, 2H), 1.21 (t, J=7.10 Hz, 6H), 1.51–
1.61 (m, 1H), 2.90–3.13 (m, 4H), 3.41 (q, J=7.07 Hz, 4H), 3.61–3.70 (m,
1H), 3.82–3.89 (m, 1H), 5.64–5.71 (m, 1H), 5.94 (s, 1H), 6.33 (s, 1H),
6.49–6.53 (s, 1H), 6.57–6.62 (m, 1H), 7.38–7.43 ppm (m, 1H); 13C NMR
(100 MHz, CDCl3): d=7.4, 7.9, 11.0, 12.4 (2C), 30.4, 30.6, 31.3, 44.7 (2C),
64.2, 97.9, 107.6, 107.8, 108.7, 113.3, 115.9, 124.9, 149.5, 150.6, 154.9,
156.3, 158.0, 162.2, 162.7, 168.4 ppm; IR: n˜ =1710, 1613, 1596, 1487, 1415,
pared according to the procedure described for compound 11a starting
from 10d (49.0 mg, 0.11 mmol, 1 equiv) to give 11d (31.2 mg, 66%).
[a]D =ꢀ58 (c=0.20 in CHCl3); 1H NMR (400 MHz, [D6]DMSO): d=
0.51–0.71 (m, 2H) 0.76–0.88 (m, 2H), 1.56–1.66 (m, 1H), 3.52 (dd, J1 =
1.76 Hz, J2 =11.86 Hz, 1H), 3.82 (dd, J1 =9.08 Hz, J2 =11.87 Hz, 1H),
3.86 (s, 3H), 4.10–4.27 (m, 2H), 5.40 (dd, J1 =1.75 Hz, J2 =9.08 Hz, 1H),
5.79 (s, 1H), 6.03 (s, 1H), 6.97 (dd, J1 =2.55 Hz, J2 =8.87 Hz, 1H), 7.03
(d, J=2.51 Hz, 1H), 7.66 ppm (d, J=8.88 Hz, 1H); 13C NMR (100 MHz,
[D6]DMSO): d=7.6, 7.8, 11.2, 32.2, 34.9, 56.5, 63.8, 101.6, 111.8, 112.4,
112.7, 112.9, 114.5, 127.0, 149.6, 152.9, 154.3, 155.6, 160.4, 160.6, 163.0,
170.1 ppm; IR: n<ꢄ= >1712, 1643, 1608, 1481, 1438, 1388, 1280, 1207,
1139, 825 cmꢀ1
; HRMS (EI): m/z: calcd for C26H29N2O5S: 481.1797
[M+H]; found: 481.1794.
(3R)-8-((7-(Diethylamino)-2-oxo-2H-chromen-4-yl)methyl)-7-((naphtha-
len-1-yloxy)methyl)-5-oxo-3,5-dihydro-2H-thiazoloACHTNUTRGNE[NUG 3,2-a]pyridine-3-car-
4530
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 4522 – 4532

Design and Synthesis of Fluorescent Pilicides and Curlicides
FULL PAPER
boxylic acid (23): Compound 22 (0.34 mmol, 0.2 g) was dissolved in THF
(20 mL) and LiOH (0.44 mmol, 0.1m, 4.4 mL) was added, the reaction
was stirred at RT for 1 h. The solvent was evaporated and the crude
product was purified with column chromatography on silica gel (first
pure EtOAc then CH2Cl2/MeOH/AcOH 90:5:5) to give 23 as a yellow
solid (180 mg, 91% yield). [a]D =ꢀ10 (c=0.5 in CHCl3); 1H NMR
(400 MHz, [D6]DMSO): d=7.99 (d, J=8.5 Hz, 1H), 7.82 (d, J=8.1 Hz,
1H), 7.60 (d, J=9.1 Hz, 1H), 7.52–7.30 (m, 4H), 6.96 (d, J=7.7 Hz, 1H),
6.62 (dd, J=9.2, 2.4 Hz, 1H), 6.48 (s, 1H), 6.44 (d, J=2.4 Hz, 1H), 5.55–
5.49 (m, 1H), 5.47 (s, 1H), 5.17–5.06 (m, 2H), 4.05–3.87 (m, 3H), 3.65–
3.58 (m, 1H), 3.45–3.34 (m, 4H), 1.10 ppm (t, 6H); 13C NMR (100 MHz,
[D6]DMSO): d=169.4, 160.8, 160.1, 155.6, 153.1, 153.0, 150.3, 150.1,
149.5, 133.9, 127.4, 126.4, 125.8, 125.5, 125.3, 124.7, 121.1, 120.5, 113.2,
108.5, 107.3, 105.6, 105.4, 105.2, 96.8, 66.5, 63.3, 44.0 (2C), 31.6, 30.6,
12.3 ppm (2C); HRMS (EI): m/z: calcd for C33H30N2O6S: 583.1903-
cation by column chromatography (CH2Cl2/MeOH 97:3 to CH2Cl2/
MeOH/AcOH 96:3:1) gave 34 as a red non-crystalline solid (25.5 mg,
88%). [a]D =34 (c=0.05 in MeOH); 1H NMR (400 MHz, [D6]DMSO):
d=2.25 (s, 6H), 2.36 (s, 6H), 2.39–2.49 (m, 2H), 3.14–3.25 (m, 2H), 3.43–
3.50 (m, 1H), 3.73–3.82 (m, 1H), 5.40–5.47 (m, 1H), 6.19 (s, 2H), 6.30 (s,
1H), 7.18–7.24 (m, 1H), 7.25–7.36 (m, 2H), 7.36–7.43 ppm (m, 2H);
13C NMR (100 MHz, [D6]DMSO): d=14.1 (m, 2C), 15.7 (2C), 25.5, 31.7,
33.4, 64.0, 111.5, 113.8, 121.8 (broad), 128.2, 129.0 (2C), 129.9 130.1,
130.5, 136.2, 140.9 (broad), 144.9, 147.9, 152.0, 153.6 (broad), 160.2,
169.5 ppm; IR: n˜ =1636, 1548, 1509, 1489, 1407, 1198, 981, 704 cmꢀ1
;
HRMS (EI): m/z: calcd for C29H29BF2N3O3S: 548.1991 [M+H]; found:
548.1972.
(3R)-5-Oxo-8-phenyl-7-(2-(1,3,5,7-tetramethyl-4,4-difluoro-4-bora-3a,4a-
diaza-s-indacene-8-yl)ethyl)-3,5-dihydro-2H-oxazoloACTHNUTRGNE[NUG 3,2-a]pyridine-3-car-
boxylic acid (37): Compound 37 was synthesized by following a previously
published procedure,[29] starting from 36 (51 mg, 0.094 mmol). Purifica-
tion by column chromatography (CH2Cl2/MeOH 97:3 to CH2Cl2/MeOH/
AcOH 96:3:1) gave 37 as a red non-crystalline solid (27 mg, 54%).
[a]D =ꢀ26 (c 0.1, CHCl3); 1H NMR (400 MHz, [D6]DMSO): d=2.23 (bs,
6H), 2.35 (s, 6H), 2.44–2.68 (m, 2H), 3.05–3.29 (m, 2H), 4.58 (dd, J1 =
3.78 Hz, J2 =8.92 Hz, 1H), 4.78 (t, J=9.02 Hz, 1H), 4.96 (dd, J1 =
3.72 Hz, J2 =9.21 Hz, 1H), 6.08 (s, 1H), 6.16 (s, 2H), 7.19–7.28 (m, 3H),
7.29–7.37 ppm (m, 2H); 13C NMR (100 MHz, [D6]DMSO): d=14.1 (m,
2C), 15.6 (2C), 26.0, 33.3, 58.3, 72.8, 97.8, 107.1, 121.8 (broad), 127.3,
128.4 (2C), 130.5 (2C), 130.8 (broad), 132.8, 140.8 (broad), 145.0, 153.6
(broad), 153.9, 154.4, 158.3, 169.6 ppm; IR: n˜ =1659, 1549, 1510, 1197,
981, 703 cmꢀ1; HRMS: m/z: calcd for C29H29BF2N3O4: 532.2219 [M+H];
found: 532.2213.
ACHTUNGTRENNUNG
[M+H+]; found: 583.1900.
7-((Naphthalen-1-yloxy)methyl)-5-oxo-8-((1,3,5,7-tetramethyl-4,4-di-
fluoro-4-bora-3a,4a-diaza-s-indacene-8-yl)methyl)-3,5-dihydro-2H-
thiazoloACHTUNGTRENNUNG[3,2-a]pyridine-3-carboxylic acid (28): LiI (111 mg, 0.83 mmol)
was added to a solution 27 (52 mg, 0.08 mmol) in dry pyridine (2 mL)
while stirring. The reaction mixture was heated to 1408C for 15 min by
microwave irradiation. The reaction mixture was allowed to attain RT, di-
luted with CH2Cl2 and washed with 2% KHSO4. The aqueous layer was
extracted three times with CH2Cl2 and the combined organic layers were
dried over Na2SO4, filtered and concentrated in vacuo. The resulting
dark oil was diluted in 1,2-dichloroethane (2 mL) and TEA (58 mL,
0.42 mmol) was added dropwise while stirring. After 5 min BF3·OEt2 was
added and the reaction mixture was heated to 808C for 15 min. After
being allowed to attain RT the reaction mixture was diluted with CH2Cl2
and washed with water. The aqueous layer was extracted five times with
CH2Cl2 and the combined organic layers were concentrated in vacuo. Pu-
rification by column chromatography (CH2Cl2, 2% MeOH!CH2Cl2, 2%
MeOH, 1% AcOH) gave 28 as an orange solid (15 mg, 29%). [a]D =0;
1H NMR (400 MHz, [D6]DMSO): d=2.21 (s, 3H), 2.29 (s, 3H), 2.43 (s,
6H), 3.39 (dd, J1 =1.26 Hz, J2 =11.88 Hz, 2H), 3.56 (dd, J1 =8.76 Hz, J2 =
11.88 Hz, 2H), 4.27 (d, J=16.73 Hz, 1H), 4.35 (d, J=16.73 Hz, 1H),
5.36–5.40 (m, 3H), 6.18–6.24 (m, 2H), 6.58 (s, 1H), 7.17–7.23 (m, 1H),
7.43–7.49 (m, 1H), 7.53–7.61 (m, 3H), 7.89–7.95 (m, 1H), 8.24–8.29 ppm
(m, 1H); 13C NMR (100 MHz, [D6]DMSO): d=14.7, 14.72, 15.3, 15.8,
28.53, 32.1, 62.0, 66.8, 105.1, 106.7, 113.4, 121.2, 121.7, 122.1, 122.1, 125.2,
126.2, 127.1, 128.1, 134.1 (broad), 134.6, 138.9, 142.4 (broad), 144.8, 150.6,
153.4, 154.5, 154.9 (broad), 160.5, 169.8 ppm; IR: n˜ =1647, 1553, 1506,
1,5-Dioxo-8-phenyl-7-(2-(1,3,5,7-tetramethyl-4,4-difluoro-4-bora-3a,4a-
diaza-s-indacene-8-yl)ethyl)-3,5-dihydro-2H-thiazoloACHTNUTRGNEUNG[3,2-a]pyridine-3-car-
boxylic acid (38): mCPBA (70%, 9 mg, 0.37 mmol) was added to 34
(18.6 mg, 0.034 mmol) in CH2Cl2 (1.5 mL) at RT. The solution was stirred
for 15 min before quenching with Na2S2O5 (aq. saturated). The resulting
solution was extracted three times with CH2Cl2, dried with Na2SO4, fil-
tered, and concentrated. Purification by column chromatography
(CH2Cl2/acetone 9:1 to CH2Cl2/acetone/AcOH 88:10:2) gave 38 as a red
foam (15.4 mg, 80% yield). [a]D =ꢀ10 (c=0.1 in DMSO); 1H NMR
(400 MHz, [D6]DMSO): d=2.26 (s, 6H), 2.39 (s, 6H), 2.54–2.64 (m, 2H),
3.14–3.32 (m, 2H), 3.53–3.62 (m, 2H), 5.40–5.49 (m, 1H), 6.02 (s, 2H),
6.79 (s, 1H), 7.30–7.49 ppm (m, 5H); 13C NMR (100 MHz, [D6]DMSO):
d=14.1 (2C), 15.6 (2C), 25.4, 33.2, 51.0, 65.2, 119.9, 120.5, 121.9 (broad),
128.47, 128.54, 128.6 130.4, 130.5 (broad), 131.1, 133.4, 140.9 (broad),
144.6, 149.0, 151.8, 153.7 (broad), 159.0, 170.1 ppm; IR: n˜ =1650, 1549,
1509, 1407, 1197, 981, 704 cmꢀ1; HRMS: m/z: calcd for C29H29BF2N3O4S:
564.1940 [M+H]; found: 564.1928.
1198, 1159, 982 cmꢀ1
; HRMS (EI): m/z: calcd for C33H31BF2N3O4S:
614.2096 [M+H]; found: 614.2097.
(3R)-7-(Naphthalen-1-ylmethyl)-5-oxo-8-(4-(1,3,5,7-tetramethyl-4,4-di-
fluoro-4-bora-3a,4a-diaza-s-indacene-8-yl)phenyl)-3,5-dihydro-2H-
thiazolo
[3,2-a]pyridine-3-carboxylic
acid
(31):
Compound
30
(0.045 mmol, 30 mg) was dissolved in THF (5 mL) and LiOH (0.09 mmol,
0.1m, 0.9 mL) was added, the reaction was stirred at RT for 1 h. The re-
action mixture was diluted with water and pH was set to 1 with 1m HCl
(aq). The waterphase was extracted with EtOAc (2ꢅ50 mL), the com-
bined organic phases was dried (Na2SO4), filtered and concentrated. The
crude product was purified with HPLC, (C18, 250ꢅ21.2 mm, 5 mm, 0–
100% MeCN over 1 h) and lyophilized to give 31 as a red solid (25 mg,
84% yield). [a]D =20 (c 0.05, CHCl3); 1H NMR (400 MHz, [D6]DMSO):
d=7.96–7.88 (m, 1H), 7.85–7.78 (m, 1H), 7.78–7.71 (m, 1H), 7.64–7.34
(m, 7H), 7.31–7.25 (m, 1H), 6.18 (s, 1H), 6.03 (s, 1H), 5.70 (s, 1H), 5.55–
5.46 (m, 1H), 4.09 (s, 2H), 3.92–3.83 (m, 1H), 3.57–3.49 (m, 1H), 2.44 (s,
3H), 2.41 (s, 3H), 1.42 (s, 3H), 0.93 ppm (s, 3H); 13C NMR (100 MHz,
[D6]DMSO): d=169.5, 160.0, 154.9 (2C), 152.8, 148.2, 142.6 (2C), 141.4,
137.2, 134.0, 133.8, 133.3, 131.2 (4C), 130.6, 128.6, 128.2 (2C), 127.3,
126.9, 126.3, 125.8, 125.5, 123.4, 121.5 (2C), 114.2, 113.9, 63.4, 35.5, 31.4,
14.2 (2C), 13.8, 13.3 ppm; HRMS (EI): m/z: calcd for C38H32BF2N3O3S:
660.2304 [M+H+]; found: 660.2303.
Acknowledgements
This work was financially supported by the Swedish Research Council
(621–2010–4730), the Knut and Alice Wallenberg Foundation, JC Kempe
Foundation (SJCKMS), S.H. was supported by AI048689, AI049950,
AI029549, and DK086378 grants.
[1] H. W. Boucher, G. H. Talbot, J. S. Bradley, J. E. Edwards, D. Gilbert,
L. B. Rice, M. Scheld, B. Spellberg, J. Bartlett, Clin. Infect. Dis.
2009, 48, 1.
[2] G. G. Anderson, J. J. Palermo, J. D. Schilling, R. Roth, J. Heuser,
S. J. Hultgren, Science 2003, 301, 105.
[3] M. A. Mulvey, Y. S. Lopez-Boado, C. L. Wilson, R. Roth, W. C.
Parks, J. Heuser, S. J. Hultgren, Science 1998, 282, 1494.
[4] M. A. Mulvey, J. D. Schilling, S. J. Hultgren, Infect. Immun. 2001, 69,
4572.
(3R)-5-Oxo-8-phenyl-7-(2-(1,3,5,7-tetramethyl-4,4-difluoro-4-bora-3a,4a-
diaza-s-indacene-8-yl)ethyl)-3,5-dihydro-2H-thiazoloACHTNUTRGNEUNG[3,2-a]pyridine-3-car-
boxylic acid (34): Compound 34 was synthesized by following the same
procedure described for 11a starting from 33 (30 mg, 0.053 mmol). Purifi-
[5] J. D. Schilling, R. G. Lorenz, S. J. Hultgren, Infect. Immun. 2002, 70,
7042.
Chem. Eur. J. 2012, 18, 4522 – 4532
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4531

F. Almqvist, S. J. Hultgren et al.
[6] M. R. Chapman, L. S. Robinson, J. S. Pinkner, R. Roth, J. Heuser,
M. Hammar, S. Normark, S. J. Hultgren, Science 2002, 295, 851.
[7] A. A. Nenninger, L. S. Robinson, S. J. Hultgren, Proc. Natl. Acad.
Sci. USA 2009, 106, 900.
[8] L. S. Robinson, E. M. Ashman, S. J. Hultgren, M. R. Chapman, Mol.
Microbiol. 2006, 59, 870.
[17] E. Chorell, J. S. Pinkner, C. Bengtsson, Sainte-Luce Banchelin, T.; S.
Edvinsson, A. Linusson, S. J. Hultgren, F. Almqvist, Bioorg. Med.
Chem. 2012, DOI: 10.1016/j.bmc.2012.01.048.
[18] M. S. T. GonÅalves, Chem. Rev. 2009, 109, 190.
[19] L. D. Lavis, R. T. Raines, ACS Chem. Biol. 2008, 3, 142.
[20] A. Loudet, K. Burgess, Chem. Rev. 2007, 107, 4891.
[21] H. Emtenꢆs, L. Alderin, F. Almqvist, J. Org. Chem. 2001, 66, 6756.
[22] V. ꢁberg, F. Norman, E. Chorell, A. Westermark, A. Olofsson,
A. E. Sauer-Eriksson, F. Almqvist, Org. Biomol. Chem. 2005, 3,
2817.
[23] E. Chorell, C. Bengtsson, T. S. Banchelin, P. Das, H. Uvell, A. K.
Sinha, J. S. Pinkner, S. J. Hultgren, F. Almqvist, Eur. J. Med. Chem.
2011, 46, 1103.
[24] P. Raman, H. Razavi, J. W. Kelly, Org. Lett. 2000, 2, 3289.
[25] D. C. Wang, J. L. Fan, X. Q. Gao, B. S. Wang, S. G. Sun, X. J. Peng,
J. Org. Chem. 2009, 74, 7675.
[26] V. Aberg, M. Hedenstrom, J. S. Pinkner, S. J. Hultgren, F. Almqvist,
Org. Biomol. Chem. 2005, 3, 3886.
[27] C. Bengtsson, F. Almqvist, J. Org. Chem. 2010, 75, 972.
[28] E. Chorell, S. Edvinsson, F. Almqvist, Tetrahedron Lett. 2010, 51,
2461.
[29] N. Pemberton, J. S. Pinkner, S. Edvinsson, S. J. Hultgren, F. Almqv-
ist, Tetrahedron 2008, 64, 9368.
[30] G. A. OꢃToole, R. Kolter, Mol. Microbiol. 1998, 28, 449.
[31] S. Mattsson, M. Dahlstrom, S. Karlsson, Tetrahedron Lett. 2007, 48,
2497.
[9] V. Aberg, F. Almqvist, Org. Biomol. Chem. 2007, 5, 1827.
[10] L. Cegelski, J. S. Pinkner, N. D. Hammer, C. K. Cusumano, C. S.
Hung, E. Chorell, V. Aberg, J. N. Walker, P. C. Seed, F. Almqvist,
M. R. Chapman, S. J. Hultgren, Nat. Chem. Biol. 2009, 5, 913.
[11] J. S. Pinkner, H. Remaut, F. Buelens, E. Miller, V. berg, N. Pember-
ton, M. Hedenstrçm, A. Larsson, P. Seed, G. Waksman, S. J. Hultg-
ren, F. Almqvist, Proc. Natl. Acad. Sci. USA 2006, 103, 17897.
[12] S. P. Nuccio, A. J. Baumler, Microbiol. Mol. Biol. Rev. 2007, 71, 551.
[13] S. L. Chen, C. S. Hung, J. A. Xu, C. S. Reigstad, V. Magrini, A. Sabo,
D. Blasiar, T. Bieri, R. R. Meyer, P. Ozersky, J. R. Armstrong, R. S.
Fulton, J. P. Latreille, J. Spieth, T. M. Hooton, E. R. Mardis, S. J.
Hultgren, J. I. Gordon, Proc. Natl. Acad. Sci. USA 2006, 103, 5977.
[14] I. Horvath, C. F. Weise, E. K. Andersson, E. Chorell, M. Sellstedt,
C. Bengtsson, A. Olofsson, S. J. Hultgren, M. R. Chapman, M. Wolf-
Watz, F. Almqvist, P. Wittung-Stafshede, J. Am. Chem. Soc. 2012,
134, 3439 DOI: 10.1021/ja209829m.
[15] H. Emtenꢆs, K. Hlin, J. S. Pinkner, S. J. Hultgren, F. Almqvist, J.
Comb. Chem. 2002, 4, 630.
[16] A. Svensson, A. Larsson, H. Emtenas, M. Hedenstrom, T. Fex, S. J.
Hultgren, J. S. Pinkner, F. Almqvist, J. Kihlberg, ChemBioChem
2001, 2, 915.
Received: December 15, 2011
Published online: March 16, 2012
4532
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 4522 – 4532