Identification of 2-aminobenzimidazole dimers as antibacterial agents

Article abstract of DOI:10.1016/S0960-894X(03)00245-2

The preparation and evaluation of 2-aminobenzimidazole dimers as antibacterial agents is described. Biological screening of the dimers indicated that compounds with multiple chloro substituents possessed optimal antibacterial activity.

Full text of DOI:10.1016/S0960-894X(03)00245-2

Bioorganic & Medicinal Chemistry Letters 13 (2003) 1669–1672
Identification of 2-Aminobenzimidazole Dimers
as Antibacterial Agents
Punit P. Seth,* Elizabeth A. Jefferson, Lisa M. Risen and Stephen A. Osgood
Ibis Therapeutics, A Division of Isis Pharmaceuticals, Inc., 2292 Faraday Avenue, Carlsbad, CA 92008, USA
Received 9January 2003; revised 10 March 2003; accepted 11 March 2003
Abstract—The preparation and evaluation of 2-aminobenzimidazole dimers as antibacterial agents is described. Biological screen-
ing of the dimers indicated that compounds with multiple chloro substituents possessed optimal antibacterial activity.
# 2003 Elsevier Science Ltd. All rights reserved.
The rapid emergence of multidrug resistant pathogenic
bacteria has become a serious health threat worldwide.1
It has been postulated that the development of resis-
tance to known antibiotics could be overcome by iden-
tifying new drug targets via genomics, improving
existing antibiotics and by identifying new antibacterial
agents with novel structures and modes of action.2ꢀ4 As
part of a program aimed at identifying new antibacterial
agents, 2-aminobenzimidazole dimer 5a was identified
as having weak bacterial growth inhibition activity (Fig.
1). A survey of the existing literature indicated that
there were no reports, which described the use of
2-aminobenzimidazole dimers as antibacterial agents.
Therefore, we initiated a study to further explore the
antibacterial activity of this class of compounds.
anilines 3a–i were reduced using Raney–Nickel to pro-
vide the corresponding phenylenediamines 4a–i, which
were cyclized using CNBr to provide the desired 2-ami-
nobenzimidazole dimers 5a–i in moderate yield (Scheme
1, 20–30% over three steps).5 All dimers were purified
by reverse phase preparative HPLC (>95% purity)
prior to biological testing.
Dimers 5a–i were evaluated for their ability to inhibit
bacterial growth against both Gram positive and Gram
negative organisms (Table 1).6ꢀ8 Initial screening
revealed that dimers 5b and 5c that possess a nonpolar
chloro and trifluoromethyl group, respectively, at the
5-position of the benzimidazole nucleus, showed a slight
improvement in antibacterial activity (entries 2–3) as
compared to the unsubstituted dimer 5a. However, the
antibacterial activity disappeared when the substituent
at C-5 was an electron withdrawing cyano (5e) or ester
(5f) group or a bromine (5d) atom (entries 4–6). Simi-
larly, dimers 5g and 5h that possess an electron donat-
ing substituent at C-6 were devoid of antibacterial
activity (entries 7–8). A dramatic improvement in anti-
bacterial activity was observed for dimer 5i, which has
chloro substituents at the 5- and 6-position of the ben-
zimidazole nucleus (entry 9).
Synthesis of 2-aminobenzimidazole dimers 5a–i was
accomplished by SNAr displacement of fluorine in 1a–i
with amine 2 to provide nitroanilines 3a–i. The nitro-
Due to the enhanced biological activity of dichloro
substituted dimer 5i, it was decided to maintain this
substitution pattern for further synthetic and biological
studies. To explore the importance of the tertiary nitro-
gen atom in the alkyl spacer, dimers 8 and 9 were pre-
pared as shown in Scheme 2. Screening of dimers 8 and
9 indicated that they did not inhibit bacterial growth at
100 mM. To ascertain the significance of the alkyl spacer
Figure 1. 2-Aminobenzimidazole dimer.
*Corresponding author. Tel.: +1-760-603-2587; fax: +1-760-603-
4653; e-mail: pseth@isisph.com
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00245-2

1670
P. P. Seth et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1669–1672
Scheme 1. Preparation of 2-aminobenzimidazole dimers.
length, dimer 11 was also prepared (Scheme 2).5
Biological screening indicated that dimer 11 also pos-
sessed moderate to good antibacterial activity against
both, Gram positive and Gram negative bacteria (entry
12). To examine the optimal position of the alkyl
spacer, N2 substituted dimers 13 and 14 were prepared
starting from 2,5,6-trichlorobenzimidazole9 (Scheme 2).
Biological screening indicated that the N2 substituted
dimers had good antibacterial activity, comparable to
the N1 substituted dimers 5i and 11 (entries 13 and 14).
Analysis of the biological data shows that the 2-amino
group (primary or secondary) and the dimeric structure
are crucial for antibacterial activity. Depending on the
organism, removal of the 2-amino group (17) results in
a 4- to 16-fold reduction of biological activity while
removal of the dimeric structure (15 and 16) abolishes
activity completely. Replacement of the tertiary amine
in the tether with an oxygen (9) or methylene (8) abol-
ishes biological activity. This result suggests that a
positively charged nitrogen (at physiological pH) in the
tether is important for antibacterial activity. Activity is
retained when the alkyl spacer chain is moved from the
N1 ring nitrogen to the 2-amino group on the benzimi-
dazole nucleus (13 and 14). The length of the spacer
between the aromatic rings also appears to have some
effect on the biological activity. In general, dimers with
a three carbon spacer (5i and 14) between the aromatic
ring and the basic amine possess better activity compared
to similar dimers (11 and 13) with a two carbon spacer.
Substitution at the 5-position of the benzimidazole
Lastly, importance of the 2-aminobenzimidazole dimer
pharmacophore was probed by evaluating 2-amino-
benzimidazoles 15 and 16 as well as benzimidazole
dimer 17 for antibacterial activity (Fig. 2). Screening
results indicated that 2-aminobenzimidazoles 15 and 16
were not biologically active. Benzimidazole dimer 17
missing the 2-amino group, showed only weak anti-
bacterial activity against both Escherichia coli and Sta-
phylococcus aureus (entry 15).
Table 1. In vitro antibacterial activity of 2-aminobenzimidazole dimers
Entry
Compda
MICb,c (mM) E. coli
MIC (mM) S. aureus
MIC (mM) S. pyogenes
MIC (mM) E. faecalis
1
2
3
4
5
6
7
8
5a (R1=H, R2=H)
5b (R1=H, R2=Cl)
5c (R1=H, R2=CF3)
5d (R1=H, R2=Br)
5e (R1=H, R2=CN)
5f (R1=H, R2=CO2Me)
5g (R1=Me, R2=H)
5h (R1=OMe, R2=H)
5i (R1=Cl, R2=Cl)
8 (R1=Cl, R2=Cl)
9 (R1=Cl, R2=Cl)
11
50–100
25–50
12–25
>100
>100
>100
>100
>100
3–6
>100
>100
12–25
6–12
50–100
25–50
25–50
>100
>100
>100
>100
>100
6–12
>100
>100
6–12
12–25
3–6
25–50
NTd
25–50
12–25
NT
NT
NT
NT
NT
3–6
NT
NT
12–25
25–50
NT
NT
NT
NT
NT
3–6
NT
9
10
11
12
13
14
15
NT
3–6
6–12
3–6
NT
1.5–3
1.5–3
3–6
13
14
17
6–12
50–100
NT
NT
aAll dimers were purified by preparative HPLC and tested as their trifluoroacetate or acetate salts.
bMinimum inhibitory concentration of compound that inhibited visible growth.
cAs a control, ciprofloxacin gave MIC values of 0.04–0.07 mM (E. coli), 0.3–0.6 mM (S. aureus), 1.2–2.5 mM (S. pyogenes) and 1.2–2.5 mM (E. fae-
calis) in our assay.
dNot tested.

P. P. Seth et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1669–1672
1671
lipophilic, dichloro substituted benzimidazole dimers
could be attributed to their ability to penetrate bacterial
cell membranes more efficiently and reach their cellular
targets.10 Interestingly, the dichlorosubstituted dimers 11
and 13 appeared to be especially effective against Enter-
ococcal bacterial strains with MIC values comparable to
those obtained with ciprofloxacin (Table 1).
In summary, the preparation and evaluation of two
types of 2-aminobenzimidazoles dimers (N1 and N2
substituted) as antibacterial agents is described. The
2-aminobenzimidazole dimers represent a novel scaffold
that possesses antibacterial activity and as such may aid
in the discovery of new antibacterial agents.
References and Notes
1. Bax, R.; Mullan, N.; Verhoef, J. Int. J. Antimicrob. Agents
2000, 16, 51.
2. Walsh, C. Nature 2000, 406, 775.
3. Ritter, T. K.; Wong, C.-H. Angew. Chem. Int. Ed. 2001, 40,
3508.
4. Wijkmans, J. C. H. M.; Beckett, R. P. DDT 2002, 7, 126.
5. Preparation for dimer 5i: 3,30-diamino-N-methyldipropyl-
amine 2 (2 mmol, 0.28 g) was added to a mixture of 1,2-
dichloro-3-fluoro-4-nitrobenzene 1i (0.26 mL, 1 mmol) and
CaCO3 (0.3 g, 4 mmol) in CH2Cl2 (2 mL). After stirring
for 12 h at rt, the reaction was filtered through Celite and
the filter bed was washed with additional CH2Cl2. The fil-
trate was concentrated under reduced pressure to provide 3i
as a yellow solid that was used without any further pur-
ification. Crude 3i (1 mmol) obtained above was hydro-
genated using Raney–Nickel (catalytic) and H2 gas
(balloon) in EtOH (10 mL) for 12 h at rt. The reaction was
filtered through Celite and the filter bed was washed with
additional EtOH. The filtrate was concentrated under reduced
pressure to provide phenylenediamine 4i as a dark oil that
was used without further purification. Crude phenylenedia-
mine 4i (1 mmol) was dissolved in EtOH (5 mL) and treated
with CNBr (3 mmol, 0.31 g). The reaction was stirred for 12
h at rt after which it was basified using 4M NaOH (pH >12).
The aqueous layer was extracted with EtOAc (2ꢁ) and the
combined organic layers were washed with brine, dried
(MgSO4) and concentrated under reduced pressure. Crude
dimer 5i was purified by reverse-phase preparative HPLC
using a Gilson 215 system with a Waters PrepPak (25ꢁ100
mm) C18 column, eluting with a linear gradient of 20–40%
mobile phase B for 30 min with a flow rate of 5 mL/min
(A=0.1% TFA in water, B=0.1% TFA in acetonitrile). 5i:
1H NMR (300 MHz, DMSO-d6) d 9.03 (s, 4H), 7.97 (s, 2H),
7.68 (s, 2H), 4.18 (t, 4H, J=6.9), 3.19 (m, 4H), 2.78 (s, 3H),
2.09(m, 4H). LC/MS: LC 2.44 min, m/z 515. 11: 1H NMR
(300 MHz, DMSO-d6) d 8.99 (s, 4H), 7.81 (s, 2H), 7.58 (s, 2H),
4.32 (m, 4H), 3.19(m, 4H), 2.78 (s, 3H). LC/MS: LC 2.75 min,
m/z 488.
Scheme 2. Preparation of dichlorosubstituted 2-aminobenzimidazole
dimers.
Figure 2. Other benzimidozole structures evaluated for antibacterial
activity.
nucleus with nonpolar groups such as chloro and tri-
fluoromethyl is tolerated and results in a modest
increase in MIC activity (5b–c). The same beneficial
effect was not seen when the substituent at C-5 was an
electron withdrawing cyano or ester group or a bromine
atom. Similarly, substitution at C-6 with a halogen such
as chlorine is beneficial (5i) while substitution with an
electron donating substituents such as methyl or meth-
oxy (5g–h) abolishes activity. Best results were seen
when both C-5 and C-6 were substituted with a chlorine
atom (5i). The improved biological activity of the more
6. Waterworth, P. M. In Laboratory Methods in Antimicrobial
Chemotherapy; Garrod, L., Ed.; Churchill Livingston: Edin-
burgh, 1978; pp 31–40.
7. Barry, A. L. The Antimicrobic Susceptibility Test: Principles
and Practice; Lea and Febiger: Philadelphia, 1976.
8. The MIC assays were carried out in a 150-mL volume in
duplicate in 96-well clear flat-bottom plates. The bacterial
suspension from an overnight culture growth in the appro-
priate medium was added to a solution of test compound in
0.5% DMSO in water. Final bacterial inoculum was approx-
imate 102–103 CFU/well. The percentage growth of the

1672
P. P. Seth et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1669–1672
bacteria in the test wells relative to that observed for a control
well containing no compound was determined by measuring
absorbance at 595 nm (A595) after 20–24 h at 37 ꢂC. The MIC
was determined as a range of concentrations where complete
inhibition of growth was observed at the higher concentration
and the bacterial cells were viable at the lower concentration.
The bacterial strains used for the assays include E. coli ATCC
25922, S. aureus ATCC 13709, E. faecalis ATCC 29212, S.
pyogenes ATCC 49399.
9. For a preparation of 2,5,6-trichlorobenzimidazole, see:
Hinkley, J. M.; Porcari, A. R.; Walker, J. A., II; Swayze, E. E.;
Townsend, L. B. Synth. Comm. 1998, 28, 1703. Experimental for
dimer 13: A mixture of 2,4,5-trichlorobenzimidazole 12 (0.15 g,
0.68 mmol), 2,20-diamino-N-methyl-diethylamine (0.029mL,
0.23 mmol), triethylamine (0.031 mL, 0.23 mmol) in EtOH (0.5
mL) was heated in a sealed tube at 130 ꢂC for 6 h. The reaction
was then concentrated and purified by reverse-phase preparative
1
HPLC to provide 13: H NMR (300 MHz, DMSO-d6) d 7.27
(s, 4H), 6.88 (m, 2H), 3.4 (m, 4H), 2.63 (t, 4H, J=6.3), 2.30 (s,
1
3H). LC/MS: LC 2.93 min, m/z 488. 14: H NMR (300MHz,
DMSO-d6) d 7.29(s, 4H), 7.15 (m, 2H), 3.34 (m, 4H), 2.77 (t,
4H, J=7.0), 2.46 (s, 3H), 1.85 (m, 4H). LC/MS: LC 214 min,
m/z 515.
10. All 2-aminobenzimidazole dimers prepared did not inhibit
bacterial transcription/translation (T/T) in a cell free assay at
100 mM. This result indicated that the dimers do not interfere
with bacterial T/T and therefore exert their antibacterial
activity by a different mode of action. For a description of the
cell free T/T assay, see: Lesley, S. A.; Brow, M. A. D.; Bur-
gess, R. R. J. Biol. Chem. 1991, 266, 2632.