N-Hydroxybenzimidazole inhibitors of ExsA MAR transcription factor in Pseudomonas aeruginosa: In vitro anti-virulence activity and metabolic stability

Article abstract of DOI:10.1016/j.bmcl.2010.04.014

ExsA is a multiple adaptational response (MAR) transcription factor, regulating the expression of a virulence determinant, the type III secretion system (T3SS) in Pseudomonas aeruginosa. Non-cytotoxic, non-antibacterial N-hydroxybenzimidazoles were identified as effective inhibitors of ExsA-DNA binding, and their potential utility as anti-virulence agents for P. aeruginosa was demonstrated in a whole cell assay. Select N-hydroxybenzimidazole inhibitors were stable in an in vitro human liver microsomal assay.

Full text of DOI:10.1016/j.bmcl.2010.04.014

Bioorganic & Medicinal Chemistry Letters 20 (2010) 3380–3383
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
N-Hydroxybenzimidazole inhibitors of ExsA MAR transcription factor in
Pseudomonas aeruginosa: In vitro anti-virulence activity and metabolic stability
Mark C. Grier a, Lynne K. Garrity-Ryan a, Victoria J. Bartlett b, , Kevin A. Klausner a, Peter J. Donovan c,
,
Caroline Dudley a, Michael N. Alekshun d, , S. Ken Tanaka a, Michael P. Draper a, Stuart B. Levy a,
Oak K. Kim a,
*
a Paratek Pharmaceuticals, Inc. 75 Kneeland Street, Boston, MA 02111, United States
b Pfizer Research, 35 CambridgePark Drive, Cambridge, MA 02141, United States
c Akrivis Technologies, 1 Broadway, Cambridge, MA 02142, United States
d Merck, 2015 Galloping Hill Road, Kenilworth, NJ 07033, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 18 February 2010
Revised 4 April 2010
Accepted 7 April 2010
Available online 11 April 2010
ExsA is a multiple adaptational response (MAR) transcription factor, regulating the expression of a viru-
lence determinant, the type III secretion system (T3SS) in Pseudomonas aeruginosa. Non-cytotoxic, non-
antibacterial N-hydroxybenzimidazoles were identified as effective inhibitors of ExsA-DNA binding,
and their potential utility as anti-virulence agents for P. aeruginosa was demonstrated in a whole cell
assay. Select N-hydroxybenzimidazole inhibitors were stable in an in vitro human liver microsomal assay.
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
MAR
Antivirulence
ExsA
Pseudomonas aeruginosa
N-Hydroxybenzimidazoles
Life-threatening infections caused by multi-drug resistant
(MDR) pathogens continue to present serious public health issues
worldwide. About half of MDR bacteria are Gram-negative, includ-
ing Pseudomonas aeruginosa, Escherichia coli, and Acinetobacter spp.
P. aeruginosa, in particular, are associated with hospital-acquired
pneumonia (HAP) as well as the chronic lung infections in cystic
fibrosis patients.1,2 Ventilator-associated pneumonia (VAP) espe-
cially poses great challenges in antimicrobial chemotherapy, be-
cause of the high frequency of MDR strains including P. aeruginosa.
Currently marketed antibiotics are becoming increasingly inac-
tive against MDR P. aeruginosa mainly because of resistance.3,4 The
next generation of antibiotic derivatives (e.g., carbapenems and
cephalosporins), demonstrating good activities against clinical iso-
lates of resistant P. aeruginosa, are in the late stage of clinical devel-
opment.5 However, given the same mechanism of action and the
similar chemical structures, these newer antibiotics would also
be subject to already existing resistance mechanisms. This situa-
tion has fostered interest in pursuing novel approaches as alterna-
tives to traditional antimicrobial chemotherapy.
Previously we reported MAR (multiple adaptational response)
technology in which we developed small molecule inhibitors for
bacterial transcription factors SoxS, MarA and Rob in E. coli, and
LcrF in Yersinia pseudotuberculosis.6,7 MAR proteins, a subgroup of
AraC family proteins, regulate the virulence gene expression in
many clinically important bacterial species. This novel strategy of
targeting bacterial virulence, as compared to traditional antibiotics
targeting bacterial growth or survival, may offer an opportunity to
circumvent resistance development issues. Given that MAR pro-
teins are not required for bacterial survival outside of a host, inhib-
itors of MAR proteins are less likely to apply the selection pressure
for resistance development.
ExsA is a MAR transcription factor in P. aeruginosa which regu-
lates the expression of a virulence determinant, the type III secre-
tion system (T3SS).8 The exsA null mutants do not express the T3SS
and exhibit substantially attenuated virulence in whole cell assays
and animal models of P. aeruginosa infection.9,10 Furthermore,
many clinically important P. aeruginosa strains express the T3SS,
which is correlated with the increased severity of clinical pneumo-
nia cases such as HAP/VAP.11,12 Therefore, developing anti-viru-
lence agents by targeting ExsA would be a viable approach to
prevent infections caused by P. aeruginosa, especially among pa-
tients in a high-risk environment such as receiving mechanical
ventilation.
* Corresponding author. Tel.: +1 617 275 0040; fax: +1 617 275 0039.
E-mail address: okim@paratekpharm.com (O.K. Kim).
Present address.
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.04.014

M. C. Grier et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3380–3383
3381
OH
N
OH
N
O2N
O2N
H
N
H
N
F
O
N
N
O
O
1
2
OH
N
O2N
H
N
F
N
O
3
Figure 1. The initial N-hydroxybenzimidazole inhibitors for ExsA-DNA binding.
We have identified, and developed small molecule inhibitors of
LcrF, a MAR protein of Yersinia spp., namely N-hydroxybenzimidaz-
oles as exemplified in Figure 1.7 Given the high degree of homology
between LcrF and ExsA in their DNA binding domain (85% identity
and 92% similarity), the inhibitors of LcrF (1, 2 and 3) were
considered potential inhibitors of ExsA. In fact, ExsA IC50 values of
compounds 1–3 were very similar to those of LcrF.7 However, sur-
prisingly, compounds 1, 2 and 3 were inactive in the T3SS dependent
P. aeruginosa macrophage whole cell cytotoxicity assay, whereas
they were highly active in the corresponding Y. pseudotuberculosis
based whole cell assay.7 For this reason, we further expanded our
investigation in the acrylic amide series of N-hydroxybenzimidaz-
oles (Table 1) in order to identify effective inhibitors of ExsA in the
whole cell assay.
Compounds listed in Table 1 were synthesized by following pre-
viously described synthetic routes with modifications (Schemes 1
and 2). For compounds 12–22, acid chlorides 5 were prepared by
the reaction of oxalyl chloride with the corresponding acid deriva-
tives that were obtained from a palladium(II)-catalyzed Heck-type
olefination of aryl halides 4 with an acrylic acid (Scheme 1). In
Scheme 1, the substitutions at the amide nitrogen were introduced
to intermediate 7 which underwent subsequent reactions to pro-
vide compounds 23–25. For R4 substituted derivatives, 2-chlor-
ophenylacetic acid 9 was nitrated to provide intermediate 10
which was then subjected to the established reaction conditions
to produce compounds 26 and 27 (Scheme 2).
The inhibitory activity of N-hydroxybenzimidazole derivatives
in Table 1 was measured using an in vitro cell-free ExsA-DNA bind-
ing assay, and is reported as IC50 values. Compounds 12–19, pos-
sessing Y = COCH3, SCH3, and CF3 (X = C or N; R6 = NO2 or CN),
exhibited reasonably good inhibitory activities in the range of
3.0–13.0 lM. Compounds possessing basic amino groups, 20 and
21, showed weak to no-measurable inhibitory activity. An attempt
to introduce a COCF3 group at Y resulted in compound 22, possess-
ing a C(OH)2CF3 group under the reaction conditions in Scheme 1,
which proved to be devoid of any measurable activity. For the sub-
stitution effect of W at the amide nitrogen, a carbamoyl methyl
group in 23 and 24, or an acetic acid group in 25 was unfavorable
for activity. On the other hand, substitution of an acetic acid group
at R4 resulted in two active compounds 26 and 27 (IC50 = 3.7 and
4.6
l
M, respectively).
Compounds exhibiting IC50 values of less than 20
lM (12–19,
26 and 27) in an ExsA-DNA binding assay were further tested in
a whole cell assay (Table 1). As control experiments, the inhibitors
were tested for intrinsic cell cytotoxicity toward mammalian cells
as well as antibacterial activity, and they were non-cytotoxic and
Table 1
In vitro activities of N-hydroxybenzimidazole inhibitors: cell-free ExsA-DNA binding assay and whole cell assay
OH
R6
W
N
N
Y
N
X
O
R4
Compd
R6
R4
W
X
Y
ExsAc IC50
(lM)
% WT cytotoxicityd (at 100
lM)
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
NO2
CN
NO2
CN
NO2
NO2
CN
NO2
NO2
NO2
NO2
NO2
CN
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
C
C
C
C
N
N
N
N
N
N
C
C
C
C
C
N
COCH3
COCH3
SCH3
SCH3
COCH3
SCH3
3.0a
22.4 3.6
19.2 3.1
>100
13.0a
10.4b
7.3b
>100
7.5a
12.1 9.5
52.7 6.3
46.6 7.3
78.5 5.0
—
—
—
—
—
6.9b
SCH3
CF3
3.5b
6.7b
H
H
H
1H-Imidazole
N(CH3)2
C(OH)2CF3
COCH3
COCH3
COCH3
COCH3
COCH3
31.2b
>100b
>100b
>47.5b
>49.0b
>51.0b
3.7b
CH2CONH2
CH2CONH2
CH2COOH
H
CN
NO2
NO2
—
CH2COOH
CH2COOH
65.4 7.5
33.1 6.7
H
4.6b
a
b
c
An ExsA protein of 298 amino acids from P. aeruginosa strain 388.
An ExsA protein of 278 amino acids from P. aeruginosa strain PAO1.
Most IC50 data are mean or average values of at least two independent experiments (see supplementary data for individual values).
Data are mean values standard deviation from at least four replicates.
d

3382
M. C. Grier et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3380–3383
O
a. acrylic acid, (n-Bu)3N,
Pd(OAc)2, DMF, 100oC
(Br)I
Cl
b. (COCl)2, THF
W
Y
W
Y
4
5
W = C, N
Y = COCH3, SCH3, C(OH)2CF3, N(CH3)2, 1H-imidazole, CF3
OH
N
R6
a. BrCH2COR
DMF, NaHCO3
R6
O
NO2
R6
4-aminobenzylamine,
NaHCO3, DMF, RT
NO2
H
N
R
N
b. Base
EtOH/DMF
60oC
N
H
F(Cl)
8 (R = OH, NH2)
NH2
7
6
a. 5, pyridine
R6 = NO2 or CN
b. 3M NaOH
a. Base, EtOH/DMF
b. 5, pyridine
c. 3M NaOH
R
O
OH
R6
OH
N
N
R6
Y
N
O
H
N
Y
W
N
N
W
O
compounds = 23-25
compounds = 12-22
(W = C, Y = COCH3)
Scheme 1. Preparation of substituted N-hydroxybenzimidazoles.
non-antibacterial, respectively (data not shown). In the whole cell
assay, P. aeruginosa strains cause lysis of infected macrophages
through the ExsA-regulated delivery of type 3 secreted toxins
(e.g., ExoU, ExoT, and ExoY).8,9 Therefore, effective inhibitors of
ExsA are expected to protect macrophage cells from the cytotoxic-
ity induced by wild-type (WT) P. aeruginosa. In Table 1, the in vitro
whole cell activities of inhibitors are expressed as % cytotoxicity
relative to that of WT P. aeruginosa (100% cytotoxicity).13 While
Before evaluating in vivo efficacy of lead compounds, we were
interested in investigating potential metabolic liabilities of phar-
macophores in N-hydroxybenzimidazole inhibitors. For prelimin-
ary in vitro phase I metabolism, select compounds 1, 12, and 13
were incubated with pooled human liver microsomes. The micro-
somal stabilities of test compounds were determined quantita-
tively by monitoring the amount of parent compounds remaining
in the microsomal supernatant at the end of each timed interval
using LC/MS/MS analytical methods. Testosterone, a known sub-
strate for in vitro metabolism, was tested as a positive control com-
pound. In a separate control experiment, the test compounds were
found to be chemically stable in the absence of microsomes (data
not shown). In the in vitro microsomal assay, while testosterone
was rapidly degraded as expected, compound 1 remained un-
changed throughout the incubation period. Compounds 12 and
13 appeared to undergo enzymatic degradation at the beginning
of incubation; however, they did not show significant changes after
the first interval (0–10 min) (Fig. 2). This result implies that the –
NO2 group at R6 as well as the acrylic amide linker may not be sub-
strates for phase I metabolism in vitro.
the
DexsA null mutant typically induced almost no cytotoxicity
to macrophage cells, compounds 12, 13, and 16 effectively attenu-
ated the cytotoxicity of WT P. aeruginosa. Compound 16, in partic-
ular, inhibited cytotoxicity (12.1% of WT cytotoxicity) nearly
reaching the level of
DexsA null mutant (<10% of WT cytotoxicity).
The following structure–activity relationship (SAR) emerged
from the whole cell assay: (i) Compounds with a nitrogen atom
at X exhibited better whole cell activity than the corresponding
carbon analogs (e.g., 16, 17, 18, and 27 vs 12, 14, 15, and 26). (ii)
Both –NO2 group and –CN group at R6 demonstrated similar whole
cell activity (e.g., 12 vs 13; 14 vs 15; and 17 vs 18). (iii) Acetic acid
substitution of compounds at R4 decreased the whole cell activity
of their corresponding hydrogen analogs (e.g., 26 vs 12; and 27
vs 16).
In summary, we identified new N-hydroxybenzimidazole deriv-
atives as inhibitors of the MAR transcription factor ExsA in P. aeru-
O2N
NO2
Cl
a. 4-aminobenzylamine
DMF, NaHCO3
HNO3
H2SO4
Cl
b. Base
EtOH/DMF
O
O
OH
OH
9
10
OH
OH
O2N
O
N
O2N
O
H
N
a. 5, pyridine
b. 3M NaOH
N
Y
NH2
W
N
N
O
OH
11
compounds = 26 and 27
OH
(Y = COCH3)
Scheme 2. Preparation of R4 substituted N-hydroxybenzimidazoles.

M. C. Grier et al. / Bioorg. Med. Chem. Lett. 20 (2010) 3380–3383
3383
120
100
80
60
40
20
0
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2010.04.014.
testosterone
compd 1
References and notes
compd 12
compd 13
1. Talbot, G. H.; Bradley, J.; Edwards, J. E., Jr.; Gilbert, D.; Scheld, M.; Bartlett, J. G.
Clin. Infect. Dis. 2006, 42, 657.
2. Arias, C. A.; Murray, B. E. N. Eng. J. Med. 2009, 360, 439.
3. Livermore, D. M. Clin. Infect. Dis. 2002, 34, 634.
4. Meyer, A. L. Curr. Opin. Pharmacol. 2005, 5, 490.
5. Talbot, G. H. Expert Rev. Anti Infect. Ther. 2008, 6, 39.
6. Bowser, T. E.; Bartlett, V. J.; Grier, M. C.; Verma, A. K.; Warchol, T.; Levy, S. B.;
Alekshun, M. N. Bioorg. Med. Chem. Lett. 2007, 17, 5652.
7. Kim, O. K.; Garrity-Ryan, L. K.; Bartlett, V. J.; Grier, M. C.; Verma, A. K.; Medjanis,
G.; Donatelli, J. E.; Macone, A. B.; Tanaka, S. K.; Levy, S. B.; Alekshun, M. N. J.
Med. Chem. 2009, 52, 5626.
8. Yahr, T. L.; Wolfgang, M. C. Mol. Microbiol. 2006, 62, 631.
9. Finck-Barbancon, V.; Goranson, J.; Zhu, L.; Sawa, T.; Wiener-Kronish, J. P.;
Fleiszig, S. M. J.; Wu, C.; Mende-Mueller, L.; Frank, D. W. Mol. Microbiol. 1997,
25, 547.
0
10
20
30
40
50
60
Time (min)
Figure 2. In vitro microsomal assay of select N-hydroxybenzimidazole inhibitors.
All data points are average values of two independent experiments with less than
10% deviations.
10. Laskowski, M. A.; Osborn, E.; Kazmierczak, B. I. Mol. Microbiol. 2004, 54, 1090.
11. Roy-Burman, A.; Savel, R.; Racine, S.; Swanson, B.; Revadigar, N.; Fujimoto, J.;
Sawa, T.; Frank, D.; Wiener-Kronish, J. J. Infect. Dis. 2001, 183, 1767.
12. Schulert, G. S.; Feltman, H.; Rabin, S. D. P.; Martin, C. G.; Battle, S. E.; Rello, J.;
Hauser, A. R. J. Infect. Dis. 2003, 188, 1695.
ginosa from a cell-free ExsA-DNA binding assay. Among those,
compounds 12, 13, and 16 effectively attenuated the cytotoxicity
of P. aeruginosa in the T3SS dependent whole cell assay. Select
compounds demonstrated good metabolic stability from in vitro
human liver microsomal assay. This work suggests that these
non-cytotoxic and non-antibacterial N-hydroxybenzimidazoles
may be further developed as anti-virulence agents to prevent
infections caused by Pseudomonas spp.
13. In
shown) inactive in this whole cell assay were tested against P. aeruginosa
strain K1119 (an efflux pump mutant
mexAB-oprM), or K767 strain (with
a control experiment, select N-hydroxybenzimidazoles (structures not
D
efflux pump MexAB-OprM). Test compounds were also inactive against both
the K1119 and K767 strains, indicating that lack of activity observed in the
whole cell cytotoxicity assay was not likely to be caused by the constitutive
efflux pump MexAB-OprM in P. aeruginosa (data not shown).
Acknowledgment
The authors thank Janice Donatelli and Ann Macone for testing
in vitro antimicrobial susceptibility of compounds.