Amidino benzimidazole inhibitors of bacterial two-component systems

Article abstract of DOI:10.1016/S0960-894X(01)00024-5

Amidino benzimidazoles have been identified as inhibitors of the bacterial KinA/Spo0F two-component system (TCS). Many of these inhibitors exhibit good in vitro antibacterial activity against a variety of susceptible and resistant Gram-positive organisms.

Full text of DOI:10.1016/S0960-894X(01)00024-5

Bioorganic & Medicinal Chemistry Letters 11 (2001) 1545±1548
Amidino Benzimidazole Inhibitors of Bacterial
Two-Component Systems
Michele A. Weidner-Wells, Kwasi A. Ohemeng,y Van N. Nguyen,y
Stephanie Fraga-Spano, Mark J. Macielag, Harvey M. Werblood,
Barbara D. Foleno, Glenda C. Webb, John F. Barretty and Dennis J. Hlasta*
Drug Discovery, The R. W. Johnson Pharmaceutical Research Institute, 1000 Route 202, Raritan, NJ 08869, USA
Received 7 August 2000; accepted 28 September 2000
AbstractÐAmidino benzimidazoles have been identi®ed as inhibitors of the bacterial KinA/Spo0F two-component system (TCS).
Many of these inhibitors exhibit good in vitro antibacterial activity against a variety of susceptible and resistant Gram-positive
organisms. The moiety at the 2-position of the benzimidazole was extensively modi®ed. In addition, the regioisomeric benzoxazoles,
heterocyclic replacements for the benzimidazole, have been synthesized and their activity against the TCS evaluated. # 2001
Elsevier Science Ltd. All rights reserved.
The number of cases of multidrug resistant bacterial
infections is increasing at an alarming rate. Clinicians
have become reliant on vancomycin as the antibiotic for
serious infections resistant to traditional agents. Due to
this, the incidence of vancomycin-resistant Enterococcus
faecium (VRE) infections in intensive-care units of US
hospitals has increased from 0.5% in 1989 to 22% in
1997.1 New classes of antibacterial agents with novel
mechanisms of action are urgently needed.
Screening of our chemical library in the model KinA/
Spo0F TCS4,5 from Bacillus subtilis resulted in the
identi®cation of the bisamidine indole 1 as a lead struc-
ture (Table 1, IC50=31 mM). Indole 1 is a known anti-
bacterial/antifungal agent, structurally related to the
bisbenzamidine class of DNA minor-groove binders.6,7
The monoamidine analogue 2 was synthesized to reduce
the potential for DNA binding. This compound showed
a 4-fold increase in potency (IC50=7.7 mM), though the
antibacterial activity was diminished (Table 6).
Bacterial two-component systems (TCSs), which include
a histidine protein kinase (HPK) and a response reg-
ulator (RR), are signal transduction pathways that
sense and respond to environmental changes thus
allowing the bacteria to adapt and survive within the
host.2 Inhibitors of TCS should block these signaling
pathways and may lead to agents which would either
slow bacterial cell growth or result in bacterial cell
death.
Since the synthesis of this class of indoles is tedious and
low yielding, the indole nucleus was replaced with a
benzimidazole. This modi®cation has the advantage
that a large variety of compounds with diverse sub-
stitution patterns can be easily synthesized and, there-
fore, the SAR could be rapidly delineated.
Table 1. KinA inhibitory activity of initial leads
HPKs respond to environmental signals by undergoing
autophosphorylation of a conserved histidine residue.
The histidine N-phosphate is then transferred to an
aspartic acid residue in the active site of the RR that
regulates gene transcription.3
Compd
X
Y
IC50 (mM)
*Corresponding author at present address. RWJPRI, Welsh
&
McKean Roads, Spring House, PA 19477, USA. Fax: +1-215-628-
3297; e-mail: dhlasta@prius.jnj.com
yCurrent address: Bristol-Myers Squibb Institute, Wallingford, CT,
USA.
1
2
3
CH
CH
N
C(ˆNH)NH2
31
7.7
42
H
H
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00024-5

1546
M. A. Weidner-Wells et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1545±1548
Benzimidazole 3 was about 5 times less potent than the
analogous indole 2 in the biochemical assay, and exhib-
ited moderate potency against Gram-positive bacteria,
including methicillin-resistant Staphylococcus aureus
(MRSA) and VRE. Encouraged by this data, the
synthesis of benzimidazole analogues was undertaken.
The HPK inhibitory potency and antibacterial activity
of these compounds are reported in this paper.
By continuing to employ diverse aldehydes in the benz-
imidazole forming reaction, it was discovered that the 2-
hydroxy-3,5-di-t-butyl analogue 8 was 4-fold more
potent than phenoxyphenyl analogue 3, but, more
importantly, it exhibited superior antibacterial activity
(Tables 3 and 6).
Compounds 9±14 were synthesized to probe the SAR
for this subseries of amidino benzimidazoles. As is evi-
dent from the data in Table 3, removal of the hydroxyl
group resulted in a decrease in activity (8 vs 10 and 11
vs 14). Methylation of the hydroxyl was also detri-
mental to activity (8 vs 9). However, the hydroxyl group
itself was not sucient for HPK inhibitory activity.
Compound 13, which lacks both t-butyl groups, was
essentially inactive. To explore the signi®cance of the
individual t-butyl groups, the regioisomeric mono-t-
butyl compounds 11 and 12 were synthesized. The
results suggested that KinA contains a hydrophobic
pocket near the 3-position of the phenyl ring, since the
3-t-butyl moiety is necessary for good potency (8 vs 11
and 11 vs 12).
The target amidines I were synthesized in one step by
condensing the bisul®te adduct of the appropriately
substituted benzaldehyde with 4-amidinophenylenedi-
amine in warm ethanol.8 The compounds were either
screened as the free base or the hydrochloride salts
(formed by recrystallization from 1 N HCl). In this
manner, over 200 compounds were prepared in a parallel
synthetic e€ort (Scheme 1).9
Table 3. KinA inhibitory activity of 2-hydroxyphenyl series
Compd
R1
R2
R3
IC50 (mM)
8
9
OH
OMe
H
OH
OH
OH
H
t-Butyl
t-Butyl
t-Butyl
t-Butyl
H
t-Butyl
t-Butyl
t-Butyl
H
t-Butyl
H
12
76
25
6
64
470
163
Scheme 1.
10
11
12
13
14
The 4-phenoxyphenyl analogue 3 was slightly more
potent than the 3-phenoxyphenyl analogue with IC50's
of 42 and 53 mM, respectively. In addition, substitutions
on the phenoxyphenyl portion did not signi®cantly
a€ect HPK inhibitory potency (data not shown).
Replacement of the phenoxy moiety was brie¯y
explored (Table 2). Biphenyl analogue 4 was less potent
than phenoxy derivative 3, as was the isopropyl analo-
gue 5. Removal of the phenoxy moiety completely
abolished all HPK inhibitory activity (6 vs 3). Intro-
duction of a diphenylamino group a€orded compound
7 which was about 2.5 times more potent than parent
benzimidazole 3.
H
t-Butyl
H
Modi®cation of the amidine moiety was also examined
(Table 4). Cyclic amidine 15 was equipotent to com-
pound 8 in the biochemical assay. Replacement of the
basic amidine with a carboxylic acid resulted in
approximately a 7-fold decrease in potency (8 vs 16).
The analogous methyl ester 17 was essentially inactive.
However, the aminoamide 18 retained good HPK
Table 4. Modi®cation of the charged substituent
Table 2. Modi®cation of phenyl substituents
Compd
X
IC50 (mM)
8
15
C(ˆNH)NH2
12
10
Compd
R
IC50 (mM)
3
4
5
6
7
OPh
Ph
CH(CH3)2
H
NPh2
42
88
220
>500
17
16
17
18
CO2H
CO2CH3
C(O)NHCH2C(CH3)2NH2
86
380
10

M. A. Weidner-Wells et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1545±1548
1547
inhibitory activity. These data indicate that a basic
functional group is the preferred functionality on the
benzimidazole portion of the substrate.
corresponding nitro compound 28, to a€ord amide 30.
This amide was then cyclized to benzoxazole 31 with
PPTS in re¯uxing xylenes. Elaboration of nitrile 31 into
targeted amidines 20 and 21 was readily accomplished
in two steps via the imidate. The desired amidines were
converted to the hydrochloride salt by recrystallization
from hydrochloric acid (1 N). The 6-amidinobenzox-
azoles 22 and 23 were synthesized by an analogous
route utilizing 4-nitro-3-hydroxybenzonitrile.
Variation of the heterocyclic nucleus was examined in
the 4-phenoxyphenyl and 2-hydroxy-3,5-di-t-butylphen-
yl subseries in order to determine the importance of the
benzimidazole NH on TCS and antibacterial activity.
To this end, the methylated benzimidazole 19 and ben-
zoxazole derivatives 20±23 were prepared and screened
for biological activity.
Table 5. Modi®cation of the heterocycle
The N-methylbenzimidazole 19 was synthesized accord-
ing to Scheme 2. Nitro derivative 25, prepared by reac-
tion of aqueous methylamine with 4-chloro-3-
nitrobenzonitrile (24), was reduced to aniline 26. This
was reacted with the bisul®te adduct of 4-phenoxy-
benzaldehyde to a€ord the benzimidazole nucleus 27.
Pinner reaction of nitrile 27 a€orded amidine 19.
Compd
R1
R2
R3
X
KinA/Spo0F
(IC50, mM)
3
Ama
H
H
Am
Am
H
H
Am
H
4-OPh
4-OPh
4-OPh
NH
NMe
O
O
NH
O
42
44
138
49
9.8
14
15.4
19
20
22
8
21
23
H
Am
Am
H
4-OPh
2-OH; 3,5-t-Bu
2-OH; 3,5-t-Bu
2-OH; 3,5-t-Bu
Am
O
aAm is an amidine moiety.
N-Methyl derivative 19 was equipotent to the unalkyl-
ated benzimidazole 3 (Table 5) thus indicating that the
benzimidazole hydrogen was not absolutely necessary
for activity in the biochemical assay. Within the phen-
oxyphenyl series, the 6-amidino benzoxazole 22 was
approximately 3-fold more potent than the 5-amidino
isomer 20. However, in the di-t-butyl series, both iso-
mers 21 and 23 are equipotent to the analogous benz-
imidazole 8. These data con®rm that neither the
hydrogen bond donating ability nor the basicity of the
benzimidazole is essential for TCS inhibitory activity. In
addition, since oxygen can only accept a hydrogen
bond, it is possible that a hydrogen bond acceptor at
that position of the heterocycle is sucient for activity.
Scheme 2.
The synthesis of the regioisomeric benzoxazole analo-
gues for each subseries was undertaken. The 5-amidi-
nobenzoxazoles 20 and 21 were synthesized as depicted
in Scheme 3. The appropriately substituted aryl acid
chloride was coupled5 with 3-amino-4-hydroxy-
benzonitrile (29), prepared by hydrogenation of the
Table 6. MIC (mg/mL) for select compoundsa
Compd
S. aureus
ATCC 29213 OC 2089 OC 3041 (VRE) OC 3312
MRSA E. faecalis
E. faecium
1
2
1
4
0.5
4
1
4
0.5
4
3
7
8
4
16
16
16
4
16
4
8
1
2
2
1
9
10
11
12
15
19
21
23
4
2
1
8
4
4
1
8
1
8
4
2
16
1
>64
1
2
16
4
4
4
2
16
1
32
1
16
1
2
0.25
1
32
1
2
>64
2
1
1
Oxacillin
Vancomycin
>64
>128
aThe variance in the determination of MIC values is 2-fold such that
an MIC di€erence of more than two dilutions is signi®cant.10
Scheme 3.

1548
M. A. Weidner-Wells et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1545±1548
Several of the more potent analogues were tested for in
vitro antibacterial activity against important Gram-
positive pathogens (Table 6). The TCS inhibitory activ-
ity of the benzimidazoles generally correlated with
antibacterial activity within each subseries. In parti-
cular, there was a uniform 8-fold di€erence in the MIC
values for the organisms studied between the regioiso-
meric mono-t-butyl compounds 11 and 12. This is con-
sistent with the IC50 values observed for these two
compounds. Benzimidazole 8 and benzoxazoles 21 and
23 displayed comparable antibacterial activity thus
indicating that the NH is not essential for activity.
However, for N-methyl analogue 19, a free NH may
contribute to antibacterial activity, since there is a 2- to
>4-fold decrease in the MIC values.
Acknowledgements
The authors wish to thank Dr. Jamese Hilliard for techni-
cal contributions and Drs. Karen Bush and Raul Gold-
schmidt for their critical assessment of the manuscript.
References and Notes
1. Fridkin, S. K.; Gaynes, R. P. Clinics in Chest Medicine
1999, 20, 303.
2. Stock, J. B.; Ninfa, A. J.; Stock, A. M. Microbiol. Rev.
1989, 53, 450.
3. Two-Component Signal Transduction; Hoch, J. A., Silhavy,
T. J., Eds.; ASM: Washington, DC, 1996 and references cited
within.
4. This assay measures the ability of the test compounds to
inhibit phosphorylation of KinA. IC50's were determined at
least twice and reproducibility was within 50%.
5. Macielag, M.; Demers, J.; Hlasta, D.; Johnson, S.; Kanojia,
R.; Fraga-Spano, S.; Sui, Z.; Weidner-Wells, M.; Werblood,
H.; Foleno, B.; Goldschmidt, R.; Loelo€, M.; Webb, G.; Bar-
rett, J. F. J. Med. Chem. 1998, 41, 2939.
6. Anne, J.; DeClercq, E.; Eyssen, H.; Dann, O. Antimicrob.
Agents Chemother. 1980, 18, 231.
7. Cory, M.; Tidwell, R. R.; Fairley, T. A. J. Med. Chem.
1992, 35, 431.
Generally, the compounds exhibited consistent anti-
bacterial activity against both susceptible and resistant
Gram-positive organisms compared to vancomycin and
oxacillin. In spite of the apparent correlation between
HPK inhibitory potency and antibacterial activity,
recent studies have shown that bacterial killing may be
due to multiple mechanisms of action for compound
8.11
In conclusion, a series of amidino benzimidazoles has
been identi®ed as potent inhibitors of the HPK KinA.
The optimum group at the 2-position of the benzimida-
zole is 2-hydroxy-3-t-butylphenyl. The benzimidazole
NH is not essential for activity in the biochemical assay
nor is it essential for antibacterial activity. Many of
these compounds exhibited good in vitro antibacterial
activity against susceptible as well as resistant strains of
several Gram-positive organisms. They exhibited good
activity against VRE, a signi®cant nosocomial pathogen
especially among immunocompromised patients. The
SAR results from this study will be applied to the design
of future compounds.
8. Ohemeng, K. A.; Nguyen, V. N. US Patent 5,942,532,
1999.
9. Ohemeng, K. A.; Nguyen, V. N.; Foleno, B. D.; Grant, E.
B.; Hlasta, D. J.; Webb, G. C.; Weidner-Wells, M. A.; Wer-
blood, H. M.; Barrett, J. F. 26th National Medicinal Chem-
istry Symposium, Richmond, VA; American Chemical
Society: Washington, DC, 1998.
10. National Committee for Clinical Laboratory Standards,
1997. Methods for dilution antimicrobial susceptibility tests
for bacteria that grow aerobically, 4th ed. Approved standard.
NCCLS Document M7-A4. National Committee for Clinical
Laboratory Standards, Villanova, PA.
11. Hilliard, J. J.; Goldschmidt, R. M.; Licata, L.; Baum,
E. Z.; Bush, K. Antimicrob. Agents Chemother. 1999, 43, 1693.