Substituted 2-phenyl-benzimidazole derivatives: novel compounds that suppress key markers of allergy

Article abstract of DOI:10.1016/j.ejmech.2006.03.014

The pharmacotherapy of allergy and asthma has traditionally focused on the effecter molecules of the allergic cascade, while neglecting targets that play an early role in their development. Reasoning that IgE is central to the expansion of atopic diseases, we identified and extended a novel family of 2-(substituted phenyl)-benzimidazole inhibitors of IgE response. Pharmacological activity depends on an intact phenylbenzimidazole-bis-amide backbone, and is optimized by the presence of lipophilic terminal groups composed of either bis cycloalkyl or combinations of aliphatic and halogen-substituted aromatic groups. These compounds also inhibit IL-4 and IL-5 responses in T cells and CD23 expression on B cells, with potencies that parallel their inhibition of IgE. The broad profile of these compounds thus underscores their potential for treating the multifarious pathology of asthma.

Full text of DOI:10.1016/j.ejmech.2006.03.014

European Journal of Medicinal Chemistry 41 (2006) 950–969
http://france.elsevier.com/direct/ejmech
Original article
Substituted 2-phenyl-benzimidazole derivatives:
novel compounds that suppress key markers of allergy
Mark L. Richards a , Shirley Cruz Lio a , Anjana Sinha a, Homayon Banie a, Richard J. Thomas a,
Michael Major b, Mark Tanji a, Jagadish C. Sircar a
a Avanir Pharmaceuticals, 11388 Sorrento Valley Road, San Diego, CA 92121, USA
b Chembridge-Major Inc., W130 N10497 Washington Avenue, Germantown, WI 53022, USA
Received in revised form 9 March 2006; accepted 23 March 2006
Available online 02 May 2006
Abstract
The pharmacotherapy of allergy and asthma has traditionally focused on the effecter molecules of the allergic cascade, while neglecting
targets that play an early role in their development. Reasoning that IgE is central to the expansion of atopic diseases, we identified and extended
a novel family of 2-(substituted phenyl)-benzimidazole inhibitors of IgE response. Pharmacological activity depends on an intact phenylbenzi-
midazole-bis-amide backbone, and is optimized by the presence of lipophilic terminal groups composed of either bis cycloalkyl or combinations
of aliphatic and halogen-substituted aromatic groups. These compounds also inhibit IL-4 and IL-5 responses in T cells and CD23 expression on
B cells, with potencies that parallel their inhibition of IgE. The broad profile of these compounds thus underscores their potential for treating the
multifarious pathology of asthma.
© 2006 Elsevier SAS. All rights reserved.
Keywords: 2-(Substituted phenyl)-benzimidazoles; CD23; IL-4; IL-5; IgE
1. Introduction
The complexity of allergic disease, particularly asthma, pre-
sents a daunting problem for the development of effective med-
icines. Hypersensitivity of the immune system to a specific
antigen (allergen) plays a central role in the initiation of asthma
and allergic rhinitis. Key components of this process include
Th2 lymphocytes, which are a major cellular infiltrate in asth-
matic lung [7,8], and the antibody immunoglobulin E (IgE),
which is over produced in the majority of people who suffer
from allergies [9–14]. However, numerous other components
have also been implicated as indeed control of the IgE response
does not always translate to control of asthma [15–17]. For
example, the low affinity receptor for IgE (CD23) has been
reported to have direct effects on IgE regulation, antigen pre-
sentation, and airway hyper-responsiveness [18–26]. Interleu-
kin-4 (IL-4) and IL-13 also are required for IgE responses in
vitro and in vivo, and have other putative roles in the develop-
ment of allergy aside from their direct activation of IgE [27–
33]. There are also considerable data supporting a role for lung
and nasal eosinophils in the pathogenesis of allergic airway
diseases, as well as the cytokines and other chemo-attractants
that contribute to the elevated levels of eosinophils [34–36].
Allergic disorders such as rhinitis, sinusitis, dermatitis, asth-
ma, and food allergy are becoming one the most common and
costly of human diseases [1,2]. One popular theory for their
rapid expansion in industrialized countries implicates the di-
minished microbial exposure resulting in the loss of Th1 cell
polarization and, ultimately, Th2 cell expansion [3–6]. Despite
the near epidemic growth of asthma, however, there has been a
paucity of novel pharmacotherapeutic approaches for the treat-
ment of allergy and asthma introduced in the last 30 years. In-
deed, the favored strategy has been to develop improved ver-
sions of existing drugs such as anti-histamines.
Abbreviations: Ab, antibody; CD23, low affinity receptor for IgE; ConA,
concanavalin A; DNP, dinitrophenol; GAHE, goat anti-human IgE antibody;
GAME, goat anti-mouse IgE antibody; Ig, immunoglobulin; IL, interleukin;
i.p., intraperitoneal; KLH, Keyhole Limpet Hemocyanin; PBL, peripheral
blood lymphocytes.
* Corresponding author. Tel.: +1 858 622 5225; fax: +1 858 658 7447.
E-mail address: mrichards@avanir.com (M.L. Richards).
1
Present address: Neurocrine Biosciences, Inc., 12790 El Camino Real, San
Diego, CA 92130, USA.
0223-5234/$ - see front matter © 2006 Elsevier SAS. All rights reserved.
doi:10.1016/j.ejmech.2006.03.014

M.L. Richards et al. / European Journal of Medicinal Chemistry 41 (2006) 950–969
951
IL-5, for example, has been shown to play a prominent role in
the recruitment of eosinophils to the lung and the subsequent
pulmonary pathophysiology associated with allergen exposure
[37–39]. Other mediators too numerous to list have been re-
ported to contribute to the development of allergy and asthma
[13,14]. This complexity of allergic pathology impedes our ef-
forts to establish disease control.
While there are a number of pharmacological agents avail-
able for the treatment of asthma and allergic rhinitis, a major
shortcoming of many of these therapeutic alternatives is that
they impact the disease state by targeting a single mediator that
modifies a response at the target organ. By acting on effecter
molecules, these drugs provide some symptomatic relief but do
not modulate the course of the disease. Anti-histamines, for
example, continue to be the drugs of choice for allergic rhinitis
because they are somewhat effective and are linked to few side
effects. However, anti-histamines provide little benefit for most
cases of asthma, and require chronic dosing to achieve optimal
effectiveness in allergic rhinitis. Leukotriene receptor antago-
nists have more recently been developed for the treatment of
asthma but their focus on a single effecter molecule limits effi-
cacy to a minority of patients [40].
More recently an anti-IgE antibody, omalizumab (Xolair)
[15–17], has been approved by the FDA for the treatment of
asthma. The clinical activity of this anti-IgE antibody has pro-
vided a sound, clinically proven basis for IgE down-regulation
as a strategy for the treatment of allergy and asthma, and gilded
a path for the development of orally active small molecule IgE
inhibitors [41–45]. However, in spite of its central position in
the expansion of allergic response, IgE is not absolutely re-
quired for disease development in animal models of asthma
[46], and its suppression by the anti-IgE antibody has not been
shown to reverse asthma pathology [15]. Moreover for omali-
zumab to be effective, serum IgE levels must be depleted by
more than 95%, a level that has been unachievable by small
molecule IgE inhibitors in vivo [16,41,42,45]. Other antibo-
dies/proteins have been developed to modulate the regulation
of allergic responses, including anti-IL-5 antibody [47,48], the
Th1 cell response promoter IL-12 [49], soluble IL-4 receptor
[50,51], and anti-CD23 antibody [52]. However, as noted
above, the single protein directedness of these target-based ap-
proaches limits their effectiveness.
It is apparent that for a drug to be effective against allergic
asthma, an action on a target that influences multiple mediators
within the allergy cascade is ultimately required. As such, the
broadly acting corticosteroids remain the cornerstone of asthma
pharmacotherapy [53,54]. However, these agents are most ef-
fective when administered daily, a strategy that is associated
with severe side effects and thus is reserved for serious exacer-
bations of the disease. Although less toxic, inhaled corticoster-
oids cause untoward effects with chronic use, and their effec-
tiveness diminish with the increasing severity of asthma. Thus,
there is a clear need for improvements in the treatment of mod-
erate to severe asthma.
the IgE response would act on a target that is fundamentally
linked to allergy development. Implementation of this strategy
through screening a proprietary compound library for activity
in an ex vivo IgE assay resulted in the identification of a lead
structure, and this was subsequently expanded to a library of
derivatives sharing the ability to suppress IgE responses. Re-
cently, we reported preliminary data on IgE-suppressing activ-
ity of one series of 2-(substituted phenyl) benzimidazole deri-
vatives headlined by the lead compound AVP-13358, which is
currently in clinical trials [55]. Herein, we have greatly ex-
panded the structural characterization of the active 2-phenyl
benzimidazole derivatives. Analysis of oral bioavailability has
also been integrated in a detailed structure–activity relationship
(SAR) in order to identify a lead compound. Additionally, we
have broadened the pharmacological profile of these com-
pounds, including a quantification of their effects on other mar-
kers of allergic disease such as CD23 and Th2 cytokines, and
have shown that the inhibition of these markers of allergy are
suppressed in parallel with the IgE response. Finally, we show
that the 2-phenyl benzimidazole compounds suppress IgE in
cultures of human peripheral blood cells, and that they reduce
the IgE response in an in vivo mouse model. Their activity
against a diverse group of allergy mediators thus establishes
these compounds as powerful new tools for the treatment of
allergy-based diseases such as asthma and allergic rhinitis.
2. Chemistry
Screening an internal universal informer library of structu-
rally diverse compounds identified the 2-phenyl benzimidazole
1a (Table 1) as a down regulator of IgE ex vivo (IC50 = 8 nM).
The compound is the first in the series of 2-(substituted phe-
nyl)-benzimidazole derivatives and formed the structural basis
for the synthesis of a focused library of compounds to identify
the candidates with a more suitable drug-like profile. Analogs
of the benzimidazole 1a were synthesized in four different ser-
ies as shown in Schemes 1–4.
2.1. Scheme 1
The benzimidazole backbone 7 for series 1 was prepared by
heating 4-nitro-phenylenediamine (5) and 4-amino benzoic
acid (6) in PPA at 150 °C for 6 h. The amine 7 was then
coupled with an appropriate acyl and aryl acid chloride in pyr-
idine or THF/pyridine mixture to give 8. The nitro-amide 8
was reduced to the amine 9 by hydrogenation with Pd–C in
MeOH–THF and the resulting amine 9 was reacted with the
appropriate acyl or aryl chloride to yield compounds 1a to 1yy.
2.2. Scheme 2
Series 2 (double reverse amides of series 1) consists of com-
pounds containing a dicarbox-diamide linked benzimidazole 12
prepared by the coupling of 3,4-diaminobenzoic acid (10) and
4-carbomethoxy benzaldehyde (11) in hot nitrobenzene for
18 h. The acid 12 was converted to the amide 13 by CDI cou-
With the objective of filling this void in asthma pharma-
cotherapy we reasoned that a compound that interferes with

952
M.L. Richards et al. / European Journal of Medicinal Chemistry 41 (2006) 950–969
Table 1
Structure and biological activity of 2-phenyl benzimidazoles against IgE responses ex vivo and in vitro: series 1
IgE
IgE
Ex vivo
In vitro
IC50 (nM)
20
Compounds
1a
1b
1c
1d
1e
1f
1g
1h
1i
1j
1k
1l
1m
1n
1o
1p
1q
1r
1s
1t
1u
1v
1w
1x
1y
1z
1aa
1bb
1cc
1dd
1ee
1ff
1gg
1hh
1ii
R1
R2
X
Y
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
2-F
Formula
IC50 (nM)
Phenyl
Phenyl
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
C27H20N4O2
C27H18Br2N4O2
C27H18F2N4O2
C27H18Cl2N4O2
C27H18Cl2N4O2
C27H18Cl2N4O2
C27H16Cl4N4O2
C27H16Cl4N4O2
C27H16Cl4N4O2
C27H16Cl4N4O2
C27H16Cl4N4O2
C29H18F6N4O2
C27H10F10N4O2
C27H19ClN4O2
C27H18N6O6
C27H22N6O2
C29H18N6O2
C29H24N4O4
C31H28N4O6
C33H32N4O8
C33H32N4O8
C29H24N4O2S2
C29H24N4O2
C39H28N4O2
C35H24N4O2
C25H20N4O2S2
C27H28N4O2
C17H16N4O2
C27H26N4O2
C22H24N4O2
C27H24Cl2N4O2
C27H25ClN4O2
C27H24Cl2N4O2
C27H25ClN4O2
C31H29FN4O2
C31H29FN4O2
C30H29N5O2
C30H29N5O2
C27H32N4O2
C35H40N4O2
C29H36N4O2
C25H28N4O2
C23H24N4O2
C21H20N4O2
C29H36N4O2
C29H36N4O2
C23H26N6O2
C31H24N4O2
C31H28N4O2
C28H22N4O2
C30H36N4O4
C29H34N4O3
C27H31FN4O2
8
4-Bromophenyl
4-Fluorophenyl
4-Chlorophenyl
3-Chlorophenyl
2-Chlorophenyl
3,4-Dichlorophenyl
2,3-Dichlorophenyl
3,5-Dichlorophenyl
2,4-Dichlorophenyl
2,6-Dichlorophenyl
4-CF3-phenyl
Penta-fluoro-phenyl
Phenyl
4-Nitrophenyl
4-Aminophenyl
4-Cyanophenyl
4-Methoxyphenyl
3,4-Dimethoxyphenyl
4-Bromophenyl
4-Fluorophenyl
4-Chlorophenyl
3-Chlorophenyl
2-Chlorophenyl
3,4-Dichlorophenyl
2,3-Dichlorophenyl
3,5-Dichlorophenyl
2,4-Dichlorophenyl
2,6-Dichlorophenyl
4-CF3-phenyl
Penta-fluoro-phenyl
4-Chlorophenyl
4-Nitrophenyl
4-Aminophenyl
4-Cyanophenyl
4-Methoxyphenyl
3,4-Dimethoxyphenyl
20
9
200
12
5
5
2
0.7
6
3
50
2
25
45
40
10
70
30
400
0.5
15
6
4
90
150
17
13
4
100
30
150
> 2500
30
80
4
700
3,4,5-Trimethoxyphenyl 3,4,5-Trimethoxyphenyl
4-Methoxyphenyl
4-S-methyl-phenyl
4-Methylphenyl
4-Phenyl-phenyl
1-Naphthalene
CH2-2-thiophene
Cyclohex-3-ene
CH3
Phenyl-
CH3
3,4-Dichlorophenyl
4-Chlorophenyl
Cyclohexyl
Phenyl
4-S-methyl-phenyl
4-Methylphenyl
4-Phenyl-phenyl
1-Naphthalene
CH2-2-thiophene
Cyclohex-3-ene
CH3
Cyclohexyl
Cyclohexyl
Cyclohexyl
Cyclohexyl
3,4-Dichlorophenyl
4-Chlorophenyl
2-Fluorophenyl
4-Fluorophenyl
1-Adamantyl
1-Adamantyl
Cyclohexyl
1-Adamantyl
Cycloheptyl
Cyclopentyl
Cyclobutyl
Cyclopropyl
4-Methyl-cyclohexyl
2-Methyl-cyclohexyl
2-Pyrrolidine
Cinnamyl
150
20
> 2500
12
55
2
>2500
80
500
40
> 2500
7
>2500
10
50
0.8
1
100
0.8
6
0.8
4
0.4
8
Cyclohexyl
1-Adamantyl
1-Adamantyl
2-Pyridyl
3-Pyridyl
Cyclohexyl
1-Adamantyl
Cycloheptyl
Cyclopentyl
Cyclobutyl
Cyclopropyl
4-Methyl-cyclohexyl
2-Methyl-cyclohexyl
2-Pyrrolidine
Cinnamyl
10
10
6
20
1jj
1kk
1ll
1mm
1nn
1oo
1pp
1qq
1rr
1ss
1tt
1uu
1vv
1ww
1xx
1yy
1zz
1aaa
1
4
1
4
0.5
4
1.5
80
700
1
400
1000
4
0.5
> 2000
12
2000
500
3
70
Phenyl
Phenyl
Cyclohexyl
Cyclohexyl
Phenyl
Phenyl
Cyclohexyl
Cyclohexyl
n-Butyl
CH3
COOCH2CH3
COCH3
H
800
7
1
1.5
2
Cyclohexyl
Cyclohexyl

M.L. Richards et al. / European Journal of Medicinal Chemistry 41 (2006) 950–969
Scheme 1. Synthesis of 2-substituted phenylbenzimidazole series 1.
Scheme 2. Synthesis of 2-substituted phenylbenzimidazole series 2.
953
pling with amines. The ester-amide 13 was hydrolyzed with aq.
LiOH in THF to afford the corresponding acid 14, which was
appended to the appropriate amine using standard CDI or
EDCI/HOBt/Et3N coupling to afford the amides 2a–2dd.
nitro-amide 21 was reduced to the amine 22, and finally
coupled with the appropriate acid chlorides to afford the
amides 4a–4nnn.
3. Results and discussion
2.3. Scheme 3
3.1. Ex vivo/in vitro IgE response
In series 3 (mono reverse amide of series 2) the nitrocarbo-
methoxy-benzimidazole 15 was prepared by heating 4-nitophe-
nylenediamine (5) and 4-carbomethoxy benzaldehyde (11) in
nitrobenzene at 155–160 °C for 18 h. The ester group of 15
was saponified with aq. LiOH in THF to the acid 16, which
was eventually converted to a nitro-amide 17 by CDI coupling
to an amine. The nitro-amide 17 was reduced to the amine 18
and coupled with the appropriate acid chlorides to afford the
amides 3a–3aa.
IgE is centrally positioned in the allergy cascade, and is
regulated by molecules that play a prominent role in allergy
such as IL-4, IL-13, and various co-stimulating molecules
within the Th2 response system. We reasoned that a compound
that inhibits IgE response would potentially interfere with a
molecule that is fundamental to the development or perpetua-
tion of allergic responses. Thus, a cell-based assay was used to
identify inhibitors of antigen-specific IgE response in vitro.
Although it contributes little understanding of the pharmacolo-
gical action, a cell-based approach is far more likely to identify
agents that act on unexplored response pathways than target-
based screening.
2.4. Scheme 4
In series 4 (mono reverse amide of series 1), the nitro-acid
benzimidazole core 20 was prepared by condensing 3,4-diami-
nobenzoic acid (10) with 4-nitrobenzaldehyde (19) in nitroben-
zene at 155–160 °C for 18 h. The nitro acid 20 was converted
to an amide 21, by CDI or EDCI coupling to an amine. The
The particular ex vivo IgE response assay was selected for
compound screening because of its capacity to reveal activity
against allergen responses in either B or T lymphocytes. The
assay utilizes in vivo priming with antigen followed by in vitro

954
M.L. Richards et al. / European Journal of Medicinal Chemistry 41 (2006) 950–969
Scheme 3. Synthesis of 2-substituted phenylbenzimidazole series 3.
Scheme 4. Synthesis of 2-substituted phenylbenzimidazole series 4.
antigen-specific IgE response in the presence of test com-
pounds [56]. Although IgE is the readout of the assay, its in-
hibition can result from an action on any of the processes
known to be important for the amplification of atopic diseases.
Thus, it can expose an agent with activity against any of a
number of processes in the secondary immune response includ-
ing antigen presentation, expression of co-stimulatory mole-
cules or cytokine receptors, or synthesis/release of IL-4, IL-
13, or IgE.
to specific antigen (DNP-KLH) [56]. Culturing B and T cells
from spleens of sensitized mice in the presence of DNP-KLH
during the in vitro portion of the protocol thus provides all the
components necessary for generating IgE. A similar response
can also be achieved by isolating B cells from mouse spleen
and culturing them with recombinant murine IL-4 and anti-
CD40 antibody (in vitro IgE response assay), thus negating
the need for T cell help in generating the IgE response. The
in vitro assay with isolated mouse splenic B cells thus directly
measures the B cell response to stimulus. Because IL-4 is sup-
plied exogenously, compounds such as suplatast tosilate that
inhibit IgE via inhibition of Th2 cytokines would not inhibit
IgE in the in vitro assay [41–45]. As shown in Tables 1–4,
all 2-phenyl benzimidazole compounds that are active in the
ex vivo IgE response assay also suppress IgE in the in vitro
assay, although two to fivefold higher concentrations are typi-
cally required in the latter. Thus, while the results in the IL-4/
anti-CD40 antibody (Ab) system indicate that the 2-(substi-
tuted phenyl)-benzimidazole derivatives act directly on B cells,
the lower potency of these compounds in the in vitro assay
relative to the ex vivo assay suggests that other components
(e.g. T cells) of the IgE response are affected. Although initial
development of the SAR was based on activity in both the ex
vivo and in vitro IgE assays, the latter was utilized in subse-
quent analyses because a) the results in the two assays follow
The ex vivo IgE response assay was used to screen an inter-
nal universal informer library that was generated from over
300,000 compounds. Of the compound hits generated from
the initial screen, a cluster of chemically similar compounds
containing the 2-(substituted phenyl)-benzimidazole core struc-
ture showed moderate potency for inhibiting IgE. The most
potent compound identified during this discovery phase is 1a,
which led to the synthesis of derivatives containing the core 2-
phenyl benzimidazole scaffold. This focused synthetic effort
resulted in the generation of compounds with IC50s in the ex
vivo IgE response assay that range from 400 pM to over 3 μM
(Tables 1–4).
IgE secretion by B cell-derived lymphocytes requires either
IL-4 or IL-13 and the engagement of CD40 on their cell sur-
face. In the ex vivo IgE response assay, these components are
provided by spleen T cells that were primed in vivo to respond

M.L. Richards et al. / European Journal of Medicinal Chemistry 41 (2006) 950–969
955
Table 2
Structure and biological activity of 2-phenyl benzimidazoles against IgE responses ex vivo and in vitro: series 2
IgE
IgE
Ex vivo
IC50 (nM)
55
200
40
In vitro
IC50 (nM)
Compounds
2a
2b
2c
2d
2e
2f
2g
2h
2i
2j
2k
2l
2m
2n
2o
2p
2q
2r
2s
2t
2u
2v
2w
2x
2y
2z
2aa
2bb
2cc
2dd
R1
R2
Y
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Formula
4-Methyl-phenyl
4-Fluorophenyl
4-Methoxyphenyl
Phenyl
Phenyl
Cyclohexyl
1-Adamantyl
Phenyl
1-Adamantyl
1-Adamantyl
2-Adamantyl
Cyclohexyl
4-Methyl-phenyl
4-Fluorophenyl
4-Methoxyphenyl
3,4-Dichlorophenyl
5-Methyl-2-pyridyl
Phenyl
C
C
C
29H24N4O2
27H18F2N4O2
29H24N4O4
40
150
100
100
300
9
25
8
9
1.5
16
5
3
7
40
40
10
10
20
40
40
40
20
80
16
35
8
C27H18Cl2N4O2
C27H21N5O2
C
C
27H26N4O2
31H30N4O2
Phenyl
1-Adamantyl
4-Chlorophenyl
3,4-Dichlorophenyl
3,4-Dichlorophenyl
4-Fluorophenyl
4-Chlorophenyl
4-Methoxyphenyl
1-Adamantyl
2-Adamantyl
2-Pyridyl
C31H30N4O2
C31H29ClN4O2
C
C
31H28Cl2N4O2
31H28Cl2N4O2
0.8
C27H25FN4O2
27H25ClN4O2
C32H32N4O3
32H32N4O3
C31H29FN4O2
30H29N5O2
Cyclohexyl
C
2-Adamantyl
4-Methoxyphenyl
4-Fluorophenyl
1-Adamantyl
2-Adamantyl
2-Adamantyl
2-Adamantyl
2-Pyridyl
2-Pyridyl
2-Adamantyl
Cyclohexyl
1-Adamantyl
4-Methyl-cyclohexyl
Cyclohexyl
1-Adamantyl
2-Adamantyl
2-Methyl-cyclohexyl
C
C
2-Pyridyl
3-Pyridyl
4-Pyridyl
C30H29N5O2
C30H29N5O2
C30H29N5O2
2-Adamantyl
1-Adamantyl
5-Methyl-2-pyridyl
Cyclohexyl
1-Adamantyl
4-Methyl-cyclohexyl
1-Adamantyl
Cyclohexyl
C30H29N5O2
C30H29N5O2
C31H31N5O2
C27H32N4O2
10
35
20
10
15
C
C
35H40N4O2
29H36N4O2
C31H36N4O2
31H36N4O2
C
35
4
4
2-Methyl-cyclohexyl
1-Adamantyl
C32H38N4O2
C32H38N4O2
in parallel, b) the ex vivo assay is complicated by the require-
ment of multiple cell types, and c) the in vitro system is more
robust, quicker, and less labor-intensive.
comparison with other series are shown in Table 5. Oral bioa-
vailability as a guide for choosing the lead compound is dis-
cussed in detail in Section 3.4.
3.3. Structure–activity relationships
3.2. Oral bioavailability
The SARs of the 2-(substituted phenyl)-benzimidazole com-
pounds were established through analyses of four series of
compounds that differ in the orientation of the bis-amide lin-
kages. The synthetic focus was also influenced by the relative
oral bioavailability, as discussed in next section.
The active 2-(substituted phenyl)-benzimidazole (2-PB)
moiety was initially identified based on the activity of 1a. This
compound contains the 2-PB backbone flanked by amide-
linked phenyl groups, and has moderate potency for suppres-
sing IgE response (IC50 = 20 nM in vitro; Table 1). The 2-
phenyl benzimidazole backbone showed little tolerance for
change as a substantial loss of activity resulted from modifica-
tions to the core structure; including replacement of the carbo-
nyl or the amine of the amide linkages with methyls, modifica-
tion of the benzimidazole or 2-phenyl groups, substitution of a
In addition to showing potent activity in an appropriate bio-
logical assay, an allergy drug must achieve adequate serum
levels following oral administration. This requirement was a
critical factor in evaluating candidate compounds for further
development. Oral bioavailability was assessed by administer-
ing 40 mg/kg of compound suspended at 10 mg/ml in 0.2 M
citric acid/PBS (pH 4) by oral gavage. Serum samples were
obtained at 1 and 4 h after dosing and quantified by compar-
ison of HPLC tracings from serum samples to a standard curve.
Compounds having active metabolites (determined by measur-
ing bioactivity of HPLC eluates) were evaluated on the basis of
total active compound in the serum. Overall, oral bioavailabil-
ity was low for all four benzimidazole series, although series 4
compounds achieved the highest serum concentrations of those
tested (Table 4). The data for selected compounds and their

956
M.L. Richards et al. / European Journal of Medicinal Chemistry 41 (2006) 950–969
Table 3
Structure and biological activity of 2-phenyl benzimidazoles against IgE responses ex vivo and in vitro: series 3
IgE
IgE
Ex vivo
In vitro
Compounds
3a
3b
3c
3d
3e
3f
3g
3h
3i
3j
3k
3l
3m
3n
3o
3p
3q
3r
3s
3t
3u
3v
3w
3x
R1
Phenyl
3,4-Dichlorophenyl
Phenyl
Cyclohexyl
R2
Phenyl
Phenyl
Cyclohexyl
Phenyl
Y
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Formula
22H24N4O2
C27H18Cl2N4O2
27H26N4O2
C27H26N4O2
31H30N4O2
C31H30N4O2
C31H28Cl2N4O2
C28H24Cl2N4O2
IC50 (nM)
IC50 (nM)
400
50
C
200
30
35
70
30
1.5
1
C
70
130
150
4
0.7
40
1-Adamantyl
Phenyl
Phenyl
C
1-Adamantyl
1-Adamantyl
Bicycloheptyl
Cyclohexyl
CH3
3,4-Dichlorophenyl
3,4-Dichlorophenyl
3,4-Dichlorophenyl
Cyclohexyl-
CH3
20
7
C27H24Cl2N4O2
15
C22H24N4O2
C17H16N4O2
C22H24N4O2
1000
> 2000
300
15
3
> 2000
> 2000
> 2000
50
6
3
500
30
60
3
60
4
3
30
20
3
70
40
CH3
CH3
Cyclohexyl
Cyclohexyl
Cyclohexyl
1-Adamantyl
Cycloheptyl
Cyclopentyl
Bicycloheptyl
2-Methyl-cyclohexyl
4-Methyl-cyclohexyl
Cyclohexyl
1-Adamantyl
1-Adamantyl
Bicycloheptyl
Bicycloheptyl
Cycloheptyl
Cyclopentyl
1-Adamantyl
C27H32N4O2
1-Adamantyl
Cycloheptyl
Cyclopentyl
Bicycloheptyl
2-Methyl-cyclohexyl
4-Methyl-cyclohexyl
1-Adamantyl
Cyclohexyl
4-Methyl-cyclohexyl
Cyclohexyl
1-Adamantyl
1-Adamantyl
1-Adamantyl
2-Pyridyl
C26H28N4O2
C29H36N4O2
C25H28N4O2
2
220
70
9
6
8
1
1
25
60
7
C29H32N4O2
C29H36N4O2
C
C
29H36N4O2
31H36N4O2
C31H36N4O2
C32H38N4O2
C28H32N4O2
C23H24N4O2
3y
3z
3aa
C32H38N4O2
C30H34N4O2
C30H29N5O2
70
benzthiazole for the benzimidazole, replacement of the amide
linkages with sulfonamides, substitution on the imidazole ni-
trogen, and replacement of amide linkages with terminal halo-
gens, amines, nitros, or methyls. The following SAR analyses
thus focuses primarily on changes in both the amide orientation
and terminal group substitutions.
two or more methoxy groups resulted in much less active com-
pounds (1s, 1t). Compounds with terminal phenyl groups sub-
stituted with nitro (1o), S-methyl (1v), amino (1p), or cyano
(1q) also were less potent than the parent (1a). Methyl substi-
tution on the terminal aryl (1w) had little effect, but the addi-
tion of bulky aryls such as a p-phenyl (1x) resulted in a loss of
activity. The bis-naphthalene derivative (1y), however, was
nearly as potent as bis-phenyl compound (1a). Replacement
of aromatic groups with methyl-thiophenes (1z) or with
methyls at one or both R positions (1bb, 1dd) resulted in a
substantial loss of activity indicating the importance of bulky
hydrophobic groups for IgE suppression.
Replacement of terminal aryls with cycloalkyl groups gen-
erally improved the potency of series 1 benzimidazoles. Sub-
stitution of a cycloalkyl in place of one of the terminal aryls
resulted in approximately equivalent activity as the bis-aryl (1a
vs. 1cc). However, when halogen-substituted aryls are used to-
gether with terminal cycloalkyls at R1 or R2 (1ee, 1ff, 1gg),
potency was increased about 10-fold. Large bis-cycloalkyls
(1mm, 1nn, 1oo) or hetero-aryl in combination with a fused
cycloalkyl group (1kk) also provided higher potency than bis
aryls. Methyl substitutions on the cyclohexyl (1ss, 1tt) did not
substantially improve potency over the unsubstituted com-
Initial attempts to improve the potency of this series by 4-
halo substitutions on the phenyl rings at R1 and R2 generally
resulted in equivalent or less potent derivatives (1b, 1c, 1d).
Substitution at the 4-position prefers the order of F > Cl > Br.
Chloro-substitutions at the 2 or 3 position (1e, 1f) either re-
tained or improved the potency over the 4-position (1d). Bis-
halo substitutions generally provided higher potency over sin-
gle halogens, and chloro-substitution at the 2,3 positions (1h)
is preferred to the 3,4 (1g), the 3,5 (1i), or the 2,4 (1j) posi-
tions, all of which are preferred to the 2,6 position (1k). Addi-
tion of electro-negative groups, either penta-fluoro or p-CF3,
also improved the potency over the unsubstituted aryls (1l,
1m), which together with the 2,3-dichloro compound (1h) were
the most potent of the bis-aryls. Introduction of an electron
donating group such as methoxyl or methyl at the 4-position
of the aromatic at R1 and R2 produced compounds with no
improvement in potency (1r, 1w) whereas substitution with

M.L. Richards et al. / European Journal of Medicinal Chemistry 41 (2006) 950–969
957
Table 4
Structure and biological activity of 2-phenyl benzimidazoles against IgE responses ex vivo and in vitro: series 4
IgE
IgE
Ex vivo
IC50 (nM)
120
In vitro
Serum
BAa
Compounds
4a
4b
4c
4d
4e
4f
4g
4h
4i
4j
4k
4l
4m
4n
4o
4p
4q
4r
4s
4t
4u
4v
4w
4x
4y
4z
4aa
4bb
4cc
4dd
4ee
4ff
4gg
4hh
4ii
R1
Phenyl
2-Pyridyl
2-Pyridyl
2-Pyridyl
1-Adamantyl
Phenyl
R2
Phenyl
Phenyl
Y
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Formula
C27H20N4O2
IC50 (nM)
200
350
250
60
6
60
0.8
6
2
5
24
7
6
C
C
26H19N5O2
27H18F3N5O3
100
800
2-OCF3-phenyl
3,4-Dichlorophenyl
Phenyl
1-Adamantyl
4-Chlorophenyl
4-Nitrophenyl
3,4-Dichlorophenyl
Penta-fluoro-benzene
Phenyl
9240b
C26H17Cl2N5O2 90
C31H30N4O2
1.5
20
0.5
5
C
C
31H30N4O2
31H29ClN4O2
1-Adamantyl
1-Adamantyl
1-Adamantyl
1-Adamantyl
Cyclohexyl
Cyclohexyl
Cyclohexyl
Bicycloheptyl
Phenyl
4-Azido-phenyl
4-Amino-phenyl
1-Adamantyl
Cyclohexyl
Cyclohexyl
Cyclohexyl
1-Adamantyl
2-Fluorophenyl
4-Fluorophenyl
1-Adamantyl
2-Pyridyl
C31H29N5O4
C31H28Cl2N4O2 0.4
C
C
31H25F5N4O2
27H26N4O2
2
8
5
7
50
40
3,4-Dichlorophenyl
4-Chlorophenyl
Phenyl
C27H24Cl2N4O2
27H25ClN4O2
C28H26N4O2
27H26N4O2
C31H29N7O2
31H31N5O2
C
50
60
7
Cyclohexyl
C
1-Adamantyl
1-Adamantyl
2-Fluorophenyl
2-Fluorophenyl
2-OCF3-phenyl
4-OCF3-phenyl
2-OCF3-phenyl
Cyclohexyl
7
C
100
10
40
40
10
10
40
10
7
C31H29FN4O2
C27H25FN4O2
C28H25F3N4O3
250*
425
750
23
C28H25F3N4O3
C32H29F3N4O3
C27H25FN4O2
C27H25FN4O2
355*
450
100
20
Cyclohexyl
2-Pyridyl
C
C
30H29N5O2
30H29N5O2
3
5
20
35
25
1-Adamantyl
1-Adamantyl
1-Adamantyl
Cyclohexyl
3-Acetoxy-1-adamantyl
4-Keto-1-adamantyl
Bicycloheptyl
3-OH-1-adamantyl
2-OH-1-adamantyl
4-OH-1-adamantyl
4-O-methyloxime-1-
adamantyl
8
600b
150
100
3-Pyridyl
4-Pyridyl
2-Pyridyl
2-Pyridyl
2-Pyridyl
2-Pyridyl
2-Pyridyl
2-Pyridyl
C30H29N5O2
C30H29N5O2
C26H25N5O2
C32H31N5O4
C30H29N5O3
C27H25N5O2
35
35
50
3
10
60
90
80
5
155b
155b
10
80
C
30H29N5O3
25
C30H29N5O3
C30H29N5O3
2-Pyridyl
2-Pyridyl
40
1500
4jj
C
31H30N6O3
30
4kk
4ll
2-Pyridyl
2-Pyridyl
4,6-Dimethyl-2-pyridyl
2-Pyridyl
H
4-F-1-adamantyl
3-Br-1-adamantyl
1-Adamantyl
1-Methyl-adamantyl
1-Adamantyl
Ch3
Ch3
Cyclohexyl
Cyclopentyl
Cycloheptyl
1-Adamantyl
Bicycloheptane
Cyclohexyl
Cyclohexyl
T-butyl
1-Adamantyl
Bicycloheptyl
Cyclohexyl
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
C30H28FN5O2
30H28BrN5O2
C32H33N5O2
C31H31N5O2
C25H26N4O2
C26H28N4O2
2
2
3
230
15
C
4mm
4nn
4oo
4pp
4qq
4rr
40
220
90
>2000
9
80
3
5
7
60
20
350
10
120
1
2
4
0.5
0.4
55
3
1-Adamantyl
Cyclohexyl
Cyclohexyl
Cyclopentyl
Cycloheptyl
1-Adamantyl
1-Adamantyl
1-Adamantyl
2-Adamantyl
Cyclohexyl
Cyclohexyl
Cyclohexyl
Bicycloheptyl
10
C
22H24N4O2
C27H32N4O2
4ss
4tt
C
C
C
25H28N4O2
29H36N4O2
35H40N4O2
4uu
4vv
4ww
4xx
4yy
4zz
C32H36N4O2
C31H36N4O2
6
5
C
C
31H36N4O2
25H30N4O2
0.4
70
2
10
65
C31H36N4O2
C28H32N4O2
4aaa
4bbb
3
4
C
28H32N4O2
(continued)

958
M.L. Richards et al. / European Journal of Medicinal Chemistry 41 (2006) 950–969
Table 4 (continued)
IgE
IgE
Ex vivo
IC50 (nM)
20
6
In vitro
IC50 (nM)
80
8
Serum
BAa
Compounds
4ccc
4ddd
R1
2-Pyridyl
2-Pyridyl
R2
Y
Formula
C31H31N5O3
C41H43N5O4
1-Adamantyl
1-Adamantyl
2-OMe
2-(1-Adamantane
carboxylate)
2-OH
2-F
3-F
2-CF3
3-OH
3-OH
3-OH
2-F
205b
4eee
4fff
4ggg
4hhh
4iii
4jjj
4kkk
4lll
2-Pyridyl
2-Pyridyl
2-Pyridyl
2-Pyridyl
2-Pyridyl
2-Pyridyl
2-Pyridyl
2-Pyridyl
2-Pyridyl
2-Pyridyl
1-Adamantyl
1-Adamantyl
1-Adamantyl
1-Adamantyl
C
30H29N5O3
4
6
6
650
6
8
C30H28FN5O2
C30H28FN5O2
C31H28F3N5O2
C30H29N5O3
10
35
1000
12
1
1100b
900b
380b
<5
1-Adamantyl
3-OH-1-adamantyl
3-Acetoxy-1-adamantyl
4-Keto-1-adamantyl
4-OH-1-adamantyl
3-OH-1-adamantyl
C
30H29N5O4
C32H31N5O5
30H26FN5O3
<5
1
5
10
C
160b
4mmm
4nnn
a
2-F
2-F
C30H28FN5O3
C30H28FN5O3
3
60
Bioavailability: ng/ml of compound in serum 1 h after a 40 mg/kg oral dose.
Includes parent compound and active metabolite.
b
pound (1mm). The potency of compounds substituted with
smaller cycloalkyls diminished in proportion to their size; cy-
clopentyl (1pp) is much more potent than cyclobutyl (1qq) or
cylcopropyl (1rr). This further underscores the importance of
bulky hydrophobic groups for IgE suppression.
potent (2o, 2p). 2-Pyridines are more potent than 3- or 4-pyr-
idines (2r, 2s, 2t) or the 5-methyl substituted 2-pyridine (2w)
when paired with 2-adamantyl at R1. With a pyridine at R1 and
adamantyl at R2, however, potency decreases (2v, 2u). Substi-
tution of 4-methoxy phenyl at R1 and R2 (2c) or just R1 (2o)
gave less potent compounds than 2-adamantyl/4-methoxy-phe-
nyl at R1/R2 (2n). As noted above for series 1, halo-substituted
phenyls combined with large cycloalkyl groups at R1 (2i, 2j,
2k, 2l, 2m), and mixtures of large unsubstituted or methyl-sub-
stituted cycloalkyls (2aa, 2cc, 2dd) also resulted in potent
compounds. This series of benzimidazoles thus show a prefer-
ence for large cycloalkyls in R1 with halo-substituted aromatics
at R2.
As pointed out above, attempts to make substitutions on the
benzimidazole backbone generally resulted in loss of activity.
Modest changes such as the introduction of halogens at the 2-
position of phenyl ring Y retained compound activity (1aaa),
but substitutions on the imidazole hydrogen with aliphatic al-
kyl groups substantially reduced the activity (1ww, 1xx). Car-
bamate (1yy) and amide derivatives (1zz), on the other hand,
retained the potency most likely because they behave as pro-
drugs by hydrolyzing to the parent compound under the assay
conditions. Thus, series 1 compounds require an intact 2-phe-
nyl benzimidazole backbone and prefer large terminal hydro-
phobic substitutions, either cycloalkyl or aryl.
Series 3 compounds were synthesized and tested to assess
the functional role of mixing the amide linkage orientation. As
with series 2, however, these compounds were found to be less
potent than their congeners in series 1 (Table 3). This applies
to bis aryl, bis alkyl, and combinations of cycloalkyls with aro-
matics (3a, 3c, 3m, 3p). As the size of the aliphatic groups
increase, however, the discrepancy in potency between the ser-
ies diminishes (3o, 3n, 3s). Direct comparison of series 3 with
series 2 compounds gave variable results. Cylcoalkyls at R1
with phenyl at R2 favor series 2 compounds (2f vs. 3d), while
the reverse orientation favors series 3 (3f). Bis-cycloalkyls also
favor series 3 (3m, 3n) while the potency of mixed cycloalkyls
is similar in both series (3t, 3u). When bulky aliphatic groups
were inserted at R2 with aromatic groups at R1 (3f), potency of
series 3 compounds increased to a level that was comparable to
the most potent compounds of other series. A highly potent
compound also resulted when bis-halogenated aromatic groups
were used at R1 with a bulky aliphatic group at R2 (3g). Like-
wise, larger bis-cycloalkyl groups were preferred over smaller
ones (3m, 3n, 3o, 3p) although bicycloheptyls (3h, 3q, 3w, 3z)
were less potent than other bulky cycloalkyls. As with the
Series 2 compounds are the double reverse amides of series
1 2-phenyl benzimidazoles. Examination of the biological ac-
tivity of these compounds indicates that, with few exceptions,
series 2 compounds are significantly less potent than their con-
geners in series 1 (Table 2). This is particularly true when com-
paring methyl-, halo-, or methoxy-substituted aromatics (2a,
2b, 2c), or unsubstituted or methyl-substituted bis-cycloalkyls
(2v, 2x, 2y, 2z) in congeners of both series, wherein potency
strongly favors series 1 compounds. However, the cyclohexyl/
4-fluoro-phenyl combination at R1 and R2 (2l) show equivalent
potency in series 1 and 2. A similar compound 2j which com-
bines 1-adamantyl with 3,4 dichloro phenyl at the termini is as
potent as any compound in series 1. Moreover, insertion of
other aliphatics and either halogen- or methoxy-substituted aro-
matics at R1 and R2, respectively, also results in reasonably
potent compounds (2k, 2l, 2m). Compounds with similar term-
inal substitutions but in the reverse orientation are overall less

M.L. Richards et al. / European Journal of Medicinal Chemistry 41 (2006) 950–969
959
Table 5
Selected oral bioavailability comparisons of 2-phenyl benzimidazoles
Core
Amide
[Serum]
(ng/ml)*
250
155
153
Bio-
activity#
25
16
15
Compounds Structure
X
o-F
o-F
p-F
linkage
Series 4
Series 1
Series 1
Series 4
IC50
10
10
10
10
I
4r
1ii
1jj
4v
o-OCF3
355
36
II
4s
4t
4u
o-F
o-OCF3
p-OCF3
Series 4
Series 4
Series 4
40
40
10
425
750
23
11
19
2
III
4w
4x
o-F
p-F
Series 4
Series 4
40
10
450
100
11
10
IV
4z
4aa
4bb
2-pyridine
3-pyridine
4-pyridine
Series 4
Series 4
Series 4
8
35
35
600
150
100
75
4
3
V
1kk
2v
3aa
4z
Series 1
Series 2
Series 3
Series 4
6
100
410
450
600
16
41
11
75
10
40
8
VI
4z
4fff
4ggg
H
2-F
3-F
Series 4
Series 4
Series 4
8
10
35
600
1100
900
75
110
26
* Serum obtained at 1 h after a 40 mg/kg dose to five female BALB/c mice (per drug). The bleeds were pooled prior to processing and quantifying by HPLC.
#
Bioactivity = peak serum concentration (nM) divided by the in vitro IC50.
other series, replacement of aryls or cyclo-alkyls with small
aliphatic groups at the termini of series 3 compounds (3j, 3k,
3l), drastically reduced their potency (IC50 > 200–2000 nM).
Thus the mixed amide linkage orientation of series 3 com-
pounds optimally favors aryls or large cyloalkyls at R1 with
bulky cyloalkyls at R2 (3s, 3u, 3v, 3y, 3n, 3o, 3f, 3g, 3i).
Series 4 compounds, while not the most potent series, pro-
vided the required elements of a drug in that they achieve high-
er bioavailability than congeners in other series while retaining
reasonable pharmacological potency. This series of compounds
developed a similar SAR as observed with the other series (Ta-
ble 4) but the low potency of bis-aryl compounds in this series
(e.g. 4a) shifted the emphasis to compounds with alkyl substi-
tutions. Acyl and aryl combinations at R1 and R2 provided the
most active compounds in series 4, with better potency than
their congeners in series 2 and equal potency to those in series
1. The few compounds available for direct comparisons be-
tween series 4 and series 1 indicate that the potency of com-
pounds with aromatics at R1 and cycloalkyls at R2 in general
favored series 1. Large bis-cycloalkyl-substituted compounds
are of similar potency in congeners of both series (4rr, 4tt,
4uu), while the smaller cyclopentyl (4ss) is substantially less
potent in series 4. The available comparisons with series 2
compounds reveal that the potency of compounds with cy-
cloalkyls at R1 and aromatics at R2 (4e, 4g, 4i, 4k, 4m, 4y)
were equivalent in the two series, while the reverse combina-
tions or bis- or mixed cycloalkyl substitutions favored series 4
(4z, 4rr, 4uu, 4ww, 4zz). Series 3 compounds were less potent

960
M.L. Richards et al. / European Journal of Medicinal Chemistry 41 (2006) 950–969
than congeners in series 4 when comparing cycloalkyls/aro-
matics at R1/R2 (4e, 4k), whereas similar activity resulted when
comparing bis-phenyl or phenyl/cycloalkyl congeners (4a, 4o).
Substitution with smaller cycloalkyls favored series 4 com-
pounds (4rr, 4ss) but larger alkyls were comparable in series
3 and 4 (4tt, 4uu). Combinations of cycloalkyls again favored
series 4 over series 3 when smaller groups or bicycloheptane
were inserted in R2 (4ww, 4vv), but again the distinction is lost
when more bulky groups are inserted in R2 (4zz).
R2 (4iii), 3-OH-adamantyl (4jjj) and 3-acetoxy adamantyl
(4kkk) were more potent. When 2-F was substituted in the Y
position little change in potency was observed for either the 4-
keto- or 4-hydroxyl-adamantyl compounds (4iii, 4mmm).
However, a greater than 10-fold loss of potency resulted from
substituting a hydroxyl in the 3-position of the adamantyl
(4nnn).
In summary, although there are series-specific variations,
compounds of all series generally share the same SARs. Of
the four series of compounds, series 1 are the most potent fol-
lowed by series 4, series 2, and then series 3. This is revealed
by direct comparisons of the IgE IC50s of congeners that are
present in each pair of series: i.e. series 1 versus series 2, 12 vs.
58 nM (number of available direct comparisons [n] = 8); series
1 versus series 3, 18 vs. 388 nM (n = 8); series 1 versus series
4, 8 vs. 40 nM (n = 10); series 2 versus series 3, 30 vs. 58 nM
(n = 7); series 2 versus series 4, 22 vs. 7 nM (n = 11); and
series 3 versus series 4, 92 vs. 32 nM (n = 10). Cross-series
comparisons also reveal that compounds with small aliphatic
Because of their high potency when attached to the amide
linkages that characterize series 4, cycloalkyl combinations
were studied in detail. As noted with other series, a comparison
of the activity of mixed aromatics and cycloalkyls favored lar-
ger alkyl groups such as adamantyl over cyclohexyl or bicyclo-
heptyl (4e, 4k, 4n). Phenyls substituted with electron with-
drawing nitro moieties (4h) were of equivalent potency to the
unsubstituted aryls (4e, 4k), which were of lower potency that
halogen-substituted phenyls (4g, 4l, 4m). A mixture of R1 aryl
and R2 alkyl (4o) is less active than the reverse (4k). Azide
substitution on the 4-position of the phenyl in R1 (4p) is con-
siderably more potent than amino substitutions (4q). When 2-
amino pyridines were used in combination with bulky aliphatic
group in either orientation, the resulting compounds suppressed
IgE with moderate-high potency (4z, 4y). However, as noted in
series 2, potency was reduced when 3- or 4-aminopyridines
were used at R1 or R2 (4aa, 4bb) in place of 2-pyridine (4z).
Dimethyl substitution on the 2-pyridine improved the potency
over unsubstituted 2-pyridine (4mm). When paired with 2-pyr-
idine at R1, adamantyl at R2 is more potent than smaller alkyls
(4cc, 4ff). Thus, substitutions were made on the adamantyl at
R2 in an effort to improve solubility without sacrificing po-
tency. Compounds with acetylated or 2-OH, or 3-OH adaman-
tane (4gg, 4hh, 4nn) were of lower potency, while 4-OH- or
halo-substituted adamantyls (4ii, 4nn, 4kk, 4ll) were more po-
tent than unsubstituted adamantane (4z). Combination of 2-
pyridine at R1 with the 4-O-methyloxime of adamantane at
R2 yielded a compound (4jj) with moderate potency, but toxi-
city was a concern. Insertion of hydrogen or methyl in R1 or R2
resulted in a substantial loss of activity (4oo, 4pp, 4qq). As
with other series, bis-cycloalkyls of series 4 display a prefer-
ence for bulkier terminal structures (4rr, 4ss, 4tt, 4uu). Mix-
tures of cycloalkyls (adamantyl, cyclohexyl, or bicycloheptyl)
appeared to demonstrate little preference for R1 versus R2 in
series 4 (4vv, 4ww, 4xx, 4yy, 4zz, 4aaa, 4bbb).
or aromatic groups at R1 or R2 exhibit low potency (IC50
>
90 nM) (2d, 3j, 3k, 3l, 4pp, 4qq), but this significantly im-
proves when aliphatic cyclic-alkyl groups or di-halogenated
aromatic rings are incorporated at R1 and R2 in bis (e.g. 1h)
or in combinations (1gg, 2j, 3g, 4g). Symmetrical and non-
symmetrical bis-acyl or aryl amides often resulted in potent
compounds (1h, 1mm, 1nn, 2cc, 3f, 4dd), but their oral bioa-
vailability (as described below) was low. Electron donating
aromatic groups or hetero-aryl groups at R1 and R2 resulted
in low activity (1z, 2c, 2o, 4c). When the imidazole hydrogen
was blocked with small aliphatic chains or even CH3, the po-
tency decreased significantly (IC50 = 2000 nM) (1ww, 1xx) in-
dicating that imidazole free hydrogen is crucial for activity.
The impact of the systemic bioavailability after oral adminis-
tration of these compounds is integrated into the discussion
below.
3.4. Selecting the lead compound
Small molecule drugs for the treatment of chronic diseases
optimally should be active following oral administration, and
thus systemic bioavailability after oral administration must be
factored into the SAR analysis. Moreover, although in vitro
systems exist to estimate this character, the intact animal al-
lows for a more accurate assessment of oral bioavailability
since serum levels of drug also reflect the impact of the “first
pass” effect as the absorbed compound passes through the liver
into the bloodstream. Although series 1 compounds achieved
the highest potency as a group, these highly potent halogen-
substituted aryls (1ee, 1gg) and large bis cyclo-alkyls (1mm,
1nn, 1oo) were poorly bioavailable following oral administra-
tion. On the other hand, series 4 emerged as the group with the
highest serum bioactivity (determined by dividing the peak ser-
um concentration by the IC50) after oral administration. Thus
bioavailability data are shown for series 4 (Table 4), while the
analyses of series 1–3 are limited to selected comparisons in
Table 5 and a discussion in the text.
A further effort to enhance potency and/or hydrophilicity
focused on substitutions in the phenyl ring Y. Several com-
pounds having the combination of 2-pyridine at R1 and ada-
mantyl at R2, and with substitutions on the Y phenyl allow a
direct comparison of the impact of the latter substitutions. 2-
Fluoro (4fff) or 3-fluoro (4ggg) substitution on the aromatic
ring (Y) retained or reduced potency, respectively, compared
with its unsubstituted parent (4z). Potency was substantially
reduced by substituting 2-CF3 (4hhh). Hydroxyl substitution
either at the 2-Y (4eee) or 3-Y (4iii) position did not affect
the relative potency of their counterparts that lack the Y-sub-
stitution (4z). In comparison to the unsubstituted adamantyl at

M.L. Richards et al. / European Journal of Medicinal Chemistry 41 (2006) 950–969
961
Pharmacological activity of the 2-phenyl benzimidazole
compounds is dependent on the lipophilicity of the terminal
substitutions, a character that negatively affects aqueous solu-
bility and oral absorption. This is evident in the bioavailability
data described above and provided in Table 4. The most potent
compounds such as those containing bis-cycloalkyls or bis-ar-
yls achieved very low serum levels, while compounds substi-
tuted with hetero-aryls (4c) or containing polar groups such as
O-methyloxime (4jj) achieved high serum concentrations but
exhibited low potency. To improve serum bioactivity following
oral administration a strategy was needed that would incorpo-
rate features of the molecule that translate to potency on one
hand while improving solubility on the other. Because large
cycloalkyl groups appear to contribute the most to potency,
they were paired in various combinations with bulky but more
hydrophilic groups.
fluorine on the Y phenyl (comparison VI) improves oral bioa-
vailability but caused little change in potency (4fff, 4ggg).
However, the fluorinated compounds showed a tendency for
increased toxicity in vivo and thus were not considered for
further evaluation.
On the strength of its bioactivity following oral dosing and
lower toxicity, 4z was selected as the lead compound. The oral
bioavailability of 4z was further augmented by making its bis-
methansulfonic salt, resulting in a threefold increase of the
peak serum concentration after oral dosing. The 4-hydroxy-
lated adamantane compound (4ii) was found to be a prominent
active metabolite of this compound.
3.5. IgE response in human peripheral blood lymphocytes
To examine the breadth of their pharmacological profile,
compounds from each structural series were selected for addi-
tional testing and comparison with mouse IgE response. To
demonstrate that the effect on IgE translates across species bar-
riers, lymphocytes were isolated from fresh human peripheral
blood (PBL) and cultured with human IL-4 and anti-human
CD40 antibody for 12 days. Thus the assay was carried out
similar to the in vitro IgE assay in mouse cells. Suppression
of IgE production was observed in parallel cultures that con-
tained 2-phenyl benzimidazole derivatives (Table 6) and with
similar potencies as observed in the in vitro mouse model (r2
linear goodness-of-fit = >0.99). These results suggest that the
target of these compounds performs a similar role in both
mouse and human lymphocytes.
Acyl and aryl combinations at R1 and R2 provided the best
series 4 compounds, with better potency than their congeners
in series 2 and with equivalent potency and better oral bioavail-
ability than those in series 1. To improve dissolution and oral
bioavailability, several combinations of adamantyl groups and
hetero-aryls were tested. When 2-amino pyridines were used at
R1 in combination with a bulky aliphatic group at R2 (4z), a
compound with moderate potency and acceptable oral bioavail-
ability resulted (Table 4). Reversing the groups retained the
potency but drastically reduced oral bioavailability (4y). At-
tempts to improve hydrophilicity by substituting the terminal
adamantyl with hydroxyl or keto groups (4gg, 4ii, 4iii, 4jjj,
4kkk, 4lll) resulted in a loss of oral bioavailability likely be-
cause they were readily conjugated. The fluoro substituted
compounds (4fff, 4ggg) provided similar or reduced in vitro
potency, and provided a modestly improved serum concentra-
tion profile over their unsubstituted parent (4z). The best oral
bioavailability was obtained by coupling 2-OCF3-substituted
phenyl at R2 with the 2-pyridine at R1 (4c). Thus, although
incorporation of hetero-aryl groups (S or N) at R1 and R2 re-
sulted in less active compounds, they improved the dissolution
rate and solubility, thereby increasing oral bioavailability.
Moreover, they are amenable to preparing salts (4z).
3.6. In vivo antibody response
To assess whether the in vitro results translate to the intact
animal, BALB/c mice were subjected to an in vivo antigen
challenge model while receiving selected compounds from
the 2-phenyl benzimidazole series. As a secondary screening
tool, it was necessary that this procedure be short, not labor-
intensive, and require a minimum quantity of compound. Thus,
a protocol was chosen that employed T cell priming in irra-
diated mice followed by challenge with specific antigen during
a 3-day course of drug [57,58]. Three weeks following antigen
challenge, mice were bled and serum antibody concentrations
determined. Irradiation is known to preferentially enhance the
IgE response while the combination of T cell priming and anti-
gen challenge provides a robust IgE stimulus, an important is-
sue when generating an IgE response in an animal that is natu-
rally resistant to developing allergies. Moreover, because the
mice were not exposed to antigen prior to drug treatment this
minimizes the background problem that would occur if mice
were primed with specific antigen instead of the carrier protein
alone. This further expedites compound testing during the dis-
covery phase of development.
When factoring the in vitro potency of these compounds, 4z
and 4fff achieve the highest serum bioactivity after oral admin-
istration. This is shown in Table 5, which compares the influ-
ence of halogen substitution, the substitution site of the pyri-
dine, and the orientation of the amide linkages. Comparisons I–
IV show that o-halo-substituted phenyls promote improved
bioavailability over p-substitution in series 4 but not in series
1 compounds (1ii, 1jj, 4r, 4v, 4s, 4t, 4u, 4w, 4x), although the
loss of bioavailability by the p-substitution is at least partially
compensated by an increase of potency (4u, 4x). Comparison
II shows that o-OCF3 substitutions provide better oral bioavail-
ability than o-F in series 4 compounds (4s, 4t). Comparison IV
shows the benefit of the 2-pyridines over the 3- and 4-pyridines
(4z, 4aa, 4bb), while comparison V demonstrates the superior-
ity of series 4 over the other series when comparing serum
bioactivity of compounds containing 2-pyridine at R1 and 1-
adamantyl at R2 (1kk, 2v, 3aa, 4z). Finally, substitution of
Intraperitoneal (i.p.) administration of 1mm suppressed the
IgE response to specific antigen (DNP) while not affecting
DNP-specific IgG1 and IgG2a levels in the serum of primed
BALB/c mice 20 days following immunization with DNP6-

962
M.L. Richards et al. / European Journal of Medicinal Chemistry 41 (2006) 950–969
Table 6
Comparison of potencies versus mouse IgE, human IgE, cytokines, and CD23
Mouse IgEa
IC50 (nM)
5
4
0.8
80
16
40
Human IgEb
IC50 (nM)
2
3
0.8
50
25
40
IL-4c
IC50 (nM)
IL-5c
IC50 (nM)
6
12
4
60
45
40
CD23d
IC50 (nM)
25
30
15
525
250
400
Compounds
R1
R2
1mm
1nn
1gg
2x
2y
2v
Cyclohexyl
1-Adamantyl
Cyclohexyl
Cyclohexyl
1-Adamantyl
2-Pyridine
Cyclohexyl
1-Adamantyl
3,4-Dichlorophenyl
Cyclohexyl
1-Adamantyl
1-Adamantyl
2.5
2
0.4
30
10
15
3n
3m
3a
4rr
4z
1-Adamantyl
Cyclohexyl
Phenyl
Cyclohexyl
2-Pyridine
1-Adamantyl
1-Adamantyl
1-Adamantyl
Cyclohexyl
Phenyl
6
8
3
12
60
225
12
20
15
5
50
50
400
9
8
5
35
250
8
10
4
12
180
5
4
2
400
> 2000
80
80
35
Cyclohexyl
1-Adamantyl
1-Adamantyl
3,4-Dichlorophenyl
4uu
4i
2
1.5
0.4
7
a
In vitro IgE from Tables 1 to 4.
Human IgE response measured in PBL.
ConA-stimulated murine IL-4 and IL-5.
Murine spleen cell CD23.
b
c
d
KLH/alum (Fig. 1A). Dexamethasone also depressed the IgE
response to specific antigen similar to previous reports but ex-
hibits no selectivity over its effect on DNP-specific IgG1 or
IgG2a. Although not shown, similar results were achieved with
4z as observed for 1mm, and follow-up experiments have
shown the ED50 for 1mm and 4z to be 1 and 2 mg/kg per
day i.p., respectively. The potency of 53 phenyl benzimidazole
derivatives tested in both the ex vivo (IC50) and in vivo
(ED50) IgE response assays were compared in Fig. 1B.
Although many of the in vivo ED50 values are estimates based
on partial responses, the result demonstrates a positive relation-
ship between IgE responses in vitro and in vivo. This validates
the ex vivo screening assay as a means for identifying com-
pounds with the capacity to suppress antigen-specific IgE re-
sponse in the intact animal.
on an earlier stage such as antigen presentation, or a T cell
function such as cytokine production.
Finally, it should be pointed out that 2-phenyl benzimida-
zole compounds are capable of suppressing the level of IgE in
this in vivo assay to only 20% of that achieved in vehicle-trea-
ted mice. While this might be expected to provide some benefit
for controlling allergic manifestations in the clinical environ-
ment, it has nonetheless been demonstrated that greater than
95% reduction of circulating IgE is required for significant
therapeutic benefit from the anti-IgE antibody omalizumab
[15,16]. Moreover, we have further shown that 1mm is effec-
tive in suppressing asthma in an animal model in doses that do
not quantitatively affect antigen-specific IgE [55]. This issue
was directly addressed by an additional experiment that tested
the response to exogenous administration of IgE–antigen com-
plexes (anti-DNP IgE – DNP-OVA) to OVA-sensitized mice.
Sensitized mice experienced strong pulmonary eosinophil infil-
tration in response to intranasal instillation of IgE–antigen
complexes, and this was inhibited by administration of 1mm
(data not shown). These results suggest that 2-substituted ben-
zimidazole compounds suppress airway responses to allergen
challenge beyond their effect on IgE production/release. Thus
while these compounds clearly suppress IgE responses in spe-
cific protocols, the modest reduction of IgE in mice treated
with 2-phenyl benzimidazole derivatives suggests that the
ED50s generated by this approach, however, are somewhat
artificial in that our protocol provides drug over a 3–4-day per-
iod that is completed 2–3 weeks prior to measuring serum IgE.
This delay between dosing and IgE measurement provides an
intriguing insight into its possible mechanism. Because the
compounds are eliminated within 24 h of dosing and minimal
IgE is found in the serum of control animals during the first
week following antigen challenge (not shown), this suggests
that the benzimidazole compounds probably do not act on
IgE synthesis/release in this model. This implicates drug action

M.L. Richards et al. / European Journal of Medicinal Chemistry 41 (2006) 950–969
963
Fig. 1. Antigen-specific antibody responses in vivo. A. Female BALB/c mice
(5–10 mice per group) were irradiated (250R), primed with KLH/alum, and
7 days later challenged with DNP6-KLH/alum. Four doses of either 1mm
(20 mg/kg), dexamethasone (100 μg per injection), or vehicle (0.25%
methylcellulose) were administered i.p. over a 3-day period starting 1 day
prior to antigen challenge. Three weeks following challenge with DNP6-KLH,
the mice were bled and DNP-specific IgE, IgG1, and IgG2a were quantified by
ELISA. Control mice produced serum levels of approximately 15, 1500, and
3 μg/ml of DNP-specific IgE, IgG1, and IgG2a, respectively. Subsequent dose–
response studies have yielded an ED50 for 1mm of 1 mg/kg per day for
suppression of IgE; * P < 0.01 by Bonferroni’s multiple comparison test. B.
Fig. 2. IL-4 and IL-5 responses in spleen cells from antigen-sensitized mice.
Spleens were removed from female BALB/c mice that had been sensitized by i.
p. injection of DNP6-KLH. Cell cultures were established in the presence of
DNP6-KLH and drug, as noted for the ex vivo IgE response assay. After 4 days
of culture, ConA (5 μg/ml) was added and the cells cultured for an additional
4 h. IL-4 and IL-5 levels in the culture supernatants were quantified by ELISA
(Panel A and B, respectively). Control responses were determined from cultures
that received stimulus but contained no drug.
The potencies of 53 2-phenyl benzimidazole compounds in the ex vivo (IC50
nM) and in vivo (ED50, mg/kg per day i.p.) IgE response assays are compared.
,
compounds act directly on B cells to block IgE production/re-
lease. However, the higher potency displayed by most of these
compounds in the ex vivo relative to the in vitro assay suggests
that in addition to the B cell, they are acting on accessory cells
as well. Since T cells comprise about half the population of
lymphocytes in BALB/c mouse spleen, these cells are likely
the source of both the cytokines and CD40 ligand (CD154).
A similar response is observed when substituting bacterial li-
popolysaccharide (LPS), a co-stimulatory molecule commonly
used for initiating polyclonal IgE responses in vitro [59,60], for
the CD40 engagement in the in vitro IgE assay (data not
shown). Because the inhibitory potencies of 2-phenyl benzimi-
dazoles are similar regardless of whether anti-CD40 antibody
anti-atopic activity of these agents are not dependent on IgE
suppression.
3.7. Cytokine response
In the in vitro IgE response assay, the two stimuli minimally
required by B cells to produce IgE are provided exogenously,
including cytokine (either IL-4 or IL-13) and CD40 engage-
ment (anti-CD40 antibody). Thus the 2-phenyl benzimidazole

964
M.L. Richards et al. / European Journal of Medicinal Chemistry 41 (2006) 950–969
or LPS are used, this suggests that these compounds may pre-
vent cytokine release by the T cell.
To directly examine the activity of 2-phenyl benzimidazole
compounds against cytokine response in T cells, two assays
were developed with BALB/c mouse spleen cells. Fig. 2 shows
the inhibitory activity of 1mm and 4z against the responses of
IL-4 and IL-5 to the combination of specific antigen (DNP6-
KLH) and ConA. This assay closely follows the ex vivo IgE
response assay except that ConA is added during the final 48 h
to boost the cytokine response. Cultures containing antigen
alone or ConA alone (during the last 48 h) yielded low and
variable cytokine levels, as measurable by ELISA.
A second assay was established with isolated T cells from
the spleens of naive mice using ConA alone as stimulus. The
culture period was shortened to 40 h to minimize secondary
cytokine responses to other mediators that might have been
released into the culture. To enhance the sensitivity of the as-
say, IL-4 and IL-5 were detected using fluorescent beads and
quantified by a Luminex 100 flow analyzer. Examination of
cytokine responses in isolated T cells thus provides a less am-
biguous means for evaluating direct effects on T cell. As per-
formed above for human IgE responses, selected compounds
from each of the four structural series were tested for activity
against IL-4 and IL-5 responses (Table 6). All compounds sup-
pressed both cytokines with potencies that paralleled the results
with the ex vivo cytokine assay and the in vitro IgE responses
in mouse cells. Plotting the IC50s for inhibition of mouse IgE
and IL-4 (or IgE and IL-5) for the compounds in Table 6
yielded a r2 linear goodness-of-fit value of > 0.99. Inhibition
of cytokine responses by T cells isolated from mouse spleen
indicates that in addition to a B cell-specific action, 2-phenyl
benzimidazole compounds also act directly on the T cell.
Moreover, the similar potency that the 2-(substituted phenyl)-
benzimidazole derivatives exhibit against B cell IgE and T cell
cytokine responses suggest that all compounds are acting on
the same target for both responses and in both cell types. Inter-
estingly, as observed for the IgE response assays, the potency
against IL-4 and IL-5 in the in vitro assay with naive cells is
about threefold lower than for the ex vivo assay with sensitized
cells. Thus it is possible that in an assay system that requires
more than a single cell type, the effects of 2-phenyl benzimi-
dazole compounds on each cell type involved in the measured
response are additive.
Fig. 3. CD23 expression on mouse spleen cells. BALB/c mouse spleen cells
were depleted of red blood cells and cultured in the presence of IL-4 (10 ng/ml)
and anti-CD40 Ab (100 ng/ml), with or without 4z (100 or 1000 ng/ml). After
24 h, cells were harvested, washed, stained with anti-CD23 antibody, and
analyzed by flow cytometry. The histograms of the control and drug-treated
samples are overlaid: control (blue shaded), 4z 100 ng/ml (green shaded), 4z
1000 ng/ml (red line). The inset shows the original histograms.
in parallel with IgE, requiring CD40 engagement and either IL-
4 or IL-13 for maximal expression [62,63]. To test for effects
on CD23 expression, spleen cells from BALB/c mice were de-
pleted of red blood cells and cultured with IL-4 and anti-CD40
antibody for 24 h. Cells cultured in the presence of 4z showed
a concentration-dependent suppression of CD23 on their cell
surface (Fig. 3). Repeating this assay in the presence of a full
concentration range of all compounds (Table 6) resulted in
CD23 suppression with IC50s that paralleled those required
for inhibition of mouse IgE (r2 linear goodness-of-fit, > 0.98).
The similar potency of all compounds against multiple re-
sponses in different cells (Table 6) suggests that a highly con-
served target is integral to the response in all assay systems.
Moreover, the target for these compounds does not reside in
a compartment dedicated to protein secretion but perhaps in
an upstream compartment such as the Golgi or endoplasmic
reticulum. Indeed, separate work has identified the Golgi as
the likely site of action for the 2-(substituted phenyl)-benzimi-
dazole derivatives (data not shown). These compounds, more-
over, act via a mechanism that is clearly distinct from other
agents (such as brefeldin A and monensin) known to disrupt
this organelle. This provides a conceptual basis in which to
frame the diverse responses to these compounds.
3.8. CD23 expression
4. Conclusions
IgE and cytokines are secreted proteins, and their inhibition
by the 2-(substituted phenyl)-benzimidazole derivatives irre-
spective of the stimuli suggests that the compounds are acting
downstream of the inductive pathway. To assess whether these
compounds perturb the expression of molecules at the plasma
membrane, their effect on the expression of CD23 was tested.
CD23 is the low affinity receptor for IgE expressed on dendri-
tic cells, macrophages, B cells, and eosinophils [61], and in-
volved in the regulation of IgE, presentation of antigen, and
airway responsiveness [18–26]. Moreover, CD23 is regulated
A comparison of four novel series of 2-(substituted phenyl)-
benzimidazole derivatives reveal a distinct SAR for the inhibi-
tion of IgE response in vitro. In addition to the core 2-(substi-
tuted phenyl)-benzimidazole backbone, the active structure has
a strong preference for terminal large lipophilic groups such as
aliphatic cycloalkyls or aromatics flanking the amide linkages.
Compounds from series 1 were favored for potency but series 4
compounds showed better serum bioactivity after oral dosing,
and a compound from the latter group (4z, AVP-13358) was

M.L. Richards et al. / European Journal of Medicinal Chemistry 41 (2006) 950–969
965
chosen as a development candidate for evaluation in clinical
trials. The molecular target of these compounds has not been
characterized, although they act directly on B cells to block
secretion of IgE and expression of CD23, and directly on T
cells to inhibit the release of IL-4 and IL-5. Moreover, these
diverse responses were suppressed with similar potency sug-
gesting that a common target mediates all pharmacological ac-
tivities. Their broad-based activity against several allergy mar-
kers and their novel mechanism of action thus suggests that
these compounds may offer important new tools for the treat-
ment of allergic disorders.
8.10 (d, J = 8.0 Hz, 1H), 7.82 (d, J = 8.0 Hz, 2H), 7.73 (br s,
1H), 2.45 (m, 1H), 1.87–1.65 (m, 5H), 1.48–1.20 (m, 5H);
ESI-MS (m/z): 365 (M + H).
5.1.1.3. Cyclohexanecarboxylic acid [4-(5-amino-1H-benzoi-
midazol-2-yl)-phenyl]-amide (9, R2 = cyclohexyl, Y = H). A so-
lution of 8 (1.25 g, 3.43 mmol) in MeOH/THF (8:2; 40 ml)
was added with 5% Pd/C and the reaction mixture was stirred
under H2 at 80 °C for 16 h. The amine 9 was freed from cat-
alyst and dried under vacuum (1.09 g; 95%); 1H NMR
(DMSO-d6, 400 MHz) δ 12.11 (br s, 1H), 10.07 (s, 1H), 8.06
(d, J = 8.80 Hz, 2H), 7.72 (d, J = 8.80 Hz, 2H), 7.29 (br s, 1H),
4.95 (br s, 2H), 2.39–2.31 (m, 1H), 1.88–1.60 (m, 5H), 1.50–
2.15 (m, 5H); ESI-MS (m/z): 335.6 (M + H).
5. Experimental protocols
The analogs were made according to following process for
each series. General preparation of lead compounds from each
series- 1–4 are discussed as follows.
5.1.1.4. Cyclohexanecarboxylic acid [4-(5-cyclohexylamino-
1H-benzoimidazol-2-yl)-phenyl]-amide (1mm, Y = H). Cyclo-
hexane carbonyl chloride (0.650 g, 4.5 mmol) was added to a
stirring solution of amine 9 (1.00 g, 3 mmol) in anhydrous
THF (15 ml) and pyridine (5 ml), and the reaction mixture
stirred at room temperature for 16 h. The reaction mixture
was poured into water and the solid (1mm) was filtered and
re-crystallized from EtOAc/hexane, m.p. 336–337 °C; 1H
NMR (DMSO-d6, 300 MHz) δ 12.6 (br s, 1H), 10.1 (s, 1H),
9.8 (s, 1H), 8.15 (br s, 3H), 7.80 (br s, 2H), 7.45 (br s, 1H),
7.25 (br s, 1H), 2.34 (m, 2H), 1.90–1.10 (m, 22H). Analysis:
C27H32N4O2·0.60 H2O (C = 71.13; H = 7.19 N = 12.14).
5.1. Syntheses
5.1.1. Method A (series 1): cyclohexanecarboxylic acid [4-(5-
cyclohexylamino-1H-benzoimidazol-2-yl)-phenyl]-amide
(1mm, Y = H)
5.1.1.1. 4-(5-Nitro-1H-benzoimidazol-2-yl)-phenylamine
(7,
Y = H). To a suspension of polyphosphoric acid (115 g) in xy-
lene (100 ml) at 60 °C, 4-nitro-1,2 phenylenediamine (5)
(11.2 g, 73.2 mmol) and 10.2 g (74.3 mmol) 4-aminobenzoic
acid (6) were added. The temperature was raised to 130 °C for
1 h before it was raised to 150 °C and stirred for 6 h. The
reaction mixture was cooled and diluted with hot water with
stirring. The hot reaction mixture was filtered through a Buch-
ner funnel and solid was isolated. Mother liquor also afforded
additional solid. Both solids were taken in water (500 ml) and
neutralized with solid NaHCO3. The yellowish green solid was
filtered and washed with hot water (5 × 500 ml). The solid was
crystallized from hot THF (600 ml). This process yielded a
5.1.2. Method B (series 2): 2-(4-cyclohexylcarbamoyl-phenyl)-
1H-benzoimidazole-5-carboxylic acid cyclohexylamide (2f, Y
= H)
5.1.2.1. 2-(4-Methoxycarbonyl)-phenyl)-1H-benzimidazole-5-
carboxylic acid (12, Y = H). A mixture of 3,4-diaminobenzoic
acid (10) (3.0 g, (19.7 mmol) and methyl-4-formyl benzoate
(11) (3.24 g, 19.7 mmol) in nitrobenzene (80 ml) was heated at
150–155 °C for 18 h. The reaction mixture was cooled to room
temperature and diluted with Et2O (500 ml). The solids were
filtered, washed with Et2O, and vacuum dried to get 3.2 g solid
12 (3.2 g, 52% yield); 1H NMR (DMSO-d6, 400 MHz) δ 13.40
(br s, 1H), 8.33 (br d, J = 7.60 Hz, 2H), 8.21 (br s, 1H), 8.13
(br d, J = 7.60 Hz, 2H), 7.94 (br s, 1H), 7.87–7.80 (dd,
J = 20.0 and 8.0 Hz, 1H), 7.69 (br d, J = 20.0 Hz, 1H), 3.91
(s, 3H); ESI-MS (m/z): 297.4 (M + H).
1
yellow solid (7) (8.7 g, 47%), m.p. 283–285 °C; H NMR
(DMSO-d6, 400 MHz) δ 13.14 (br s, 1H), 8.38 (br s, 1H),
8.16 (d, J = 3.2 Hz, 1H), 7.98 (d, J = 2.4 Hz, 2H) 7.68 (br s,
1H), 6.88 (d, J = 5.6 Hz, 2H), 5.87 (br s, 2H); ESI-MS (m/z):
255.6 (M + H).
5.1.1.2. Cyclohexanecarboxylic acid [4-(5-nitro-1H-benzoimi-
dazol-2-yl)-phenyl]-amide (8, R2 = cyclohexyl, Y = H). To a
solution of 7 (1.26 g, 4.97 mmol) in anhydrous pyridine
(10 ml) at room temperature, cylohexane carbonyl chloride
(1.0 ml, 5.14 mmol) was added and the reaction mixture was
stirred at room temperature for 3 h, quenched with water, and
solid was filtered. The gummy solid was taken in EtOH
(0.020 ml) and stirred with 10% NaOH (1 ml) at 90 °C for
30 min. The EtOH was evaporated and water was added. The
solid 8 was filtered, washed with water, and dried (1.36 g,
5.1.2.2. 4-(5-Cyclohexylcarbamoyl-1H-benzoimidazol-2-yl)-
benzoic acid methyl ester (13, R1 = cyclohexyl, Y = H). A so-
lution of acid 12 (5.0 g, 16.9 mmol) in dry DMF (250 ml) and
CDI (7.1 g, 44 mmol, 2.5 eq.), was stirred at room temperature
for 3 h. Cyclohexyl amine (2.0 g, 23 mmol, 1.35 eq.) was
added, refluxed for 96 h, poured into water (2.0 l), and stirred
at room temperature for 16 h. The resulting solid was filtered
and washed with water. The solid 13 was re-crystallized from
THF–MeOH–Et2O (0.4 g, 6% yield) and used for the next step.
1
73%); H NMR (DMSO-d6, 400 MHz) δ 13.60 (br s, 1H),
10.18 (br s, 1H), 8.48 (br s, 1H), 8.14 (br d, J = 12 Hz, 2H),

966
M.L. Richards et al. / European Journal of Medicinal Chemistry 41 (2006) 950–969
5.1.2.3. 4-(5-Cyclohexylcarbamoy)-1H-benz imidazol-2yl)ben-
zoic acid (14, R1 = cyclohexyl, Y = H). A mixture of methyl
4-(5-(cyclohexylcarbonyl)-1H-benzoimidazole-2-yl)benzoate
(13) (2.5 g, 6.6 mmol) in THF (30 ml), water (25 ml), and
LiOH (1.06 g) was stirred at room temperature for 2 h, mixed
with charcoal, stirred with slight heating, and then filtered. The
filtrate was chilled in ice and acidified with concentrated HCl
to pH 1.0. The acid 14 was filtered, washed with water and re-
crystallized from THF/MeOH/Et2O to get a white solid (1.9 g,
80%); m.p. 275–280 °C and used without further purification
for the next step.
The reaction mixture was poured into ice water (100 ml), and
stirred at room temperature for 4–5 h. The nitro-amide 17 was
filtered, washed with water and ether and dried in vacuum
(yield 1.9 g, 71%) and used without further purification for
the next step.
5.1.3.3. 4-(5-Amino-1H-benzoimidazol-2-yl)-N-cyclohexyl-ben-
zamide (18, R2 = cyclohexyl, Y = H). The nitrocyclohexyla-
mide 17 (1.9 g, 5.0 mmol) was hydrogenated in presence of
5% Pd–C (0.475 g) in MeOH (40 ml) at room temperature
for 18 h. The amine 18 was filtered, and solvent was evapo-
rated. The residue was treated with ether (20 ml), and solid was
filtered, re-crystallized from THF/MeOH/hexane and dried in
vacuum. Yield: 1.05 g, 60%.
5.1.2.4. 2-(4-Cyclohexylcarbamoyl-phenyl)-1H-benzoimida-
zole-5-carboxylic acid cyclohexylamide (2f, Y = H). A reaction
mixture of acid 14 (0.2 g, 0.56 mmol) in anhydrous DMF
(2.0 ml), EDCI (0.107 g, 1 mmol), HOBt (0.076 g, 1 mmol),
and Et3N (0.3 ml) was stirred at room temperature for 5 min. A
solution of aniline (0.052 g, 1 mmol) in DMF (0.015 ml) was
added and stirred for 21 h. The reaction was quenched with ice
water (0.010 ml) and the solids were filtered and washed with
ice water. The solids were re-crystallized from EtOH/H2O as
5.1.3.4. 4-[5-(Cyclohexanecarbonyl-amino)-1H-benzoimida-
zol-2-yl]-N-cyclohexyl-benzamide (3m, Y = H). A mixture of
amine 18 (0.5 g, 1.50 mmol) in dry THF and pyridine
(0.54 g) was heated to a clear solution. Then a solution of cy-
clohexyl carbonyl chloride (0.26 g, 1.77 mmol) in THF (10 ml)
was added and refluxed for 18 h. The reaction was poured on
water and stirred for 1 h. The solid 3m was filtered, washed
with water and ether, re-crystallized from THF:MeOH
(3 ml:0.5 ml) and diluted with Et2O. Yield 0.423 g, 63%, m.p.
1
being solid of 2f (0.168 g, 34%); m.p. 324–325 °C; H NMR
(DMSO-d6, 300 MHz) δ 13.27 (br s, 1H), 10.37(s, 1H), 8.46–
8.13 (br m, 6H), 7.90–7.70 (br m, 4H), 7.41–7.34 (t,
J = 9.0 Hz, 2H), 7.16–7.09 (m, 1H), 3.80 (m, 1H), 1.84–1.78
(br m, 5H), 1.34–1.05 (br m, 5H). Analysis: C27H26N4O2·1.68
H2O (C = 69.17; H = 6.14; N = 12.01).
1
325–328 °C; H NMR (DMSO-d6, 300 MHz) δ 13.0 (br s,
1H), 9.80 (br s, 1H), 8.31 (d, J = 9.0 Hz, 1H), 8.20 (br d,
J = 6.0 Hz, 2H), 8.12 (br s, 1H), 7.98 (br d, J = 6.0 Hz, 2H),
7.51 (br s, 1H), 7.27 (br s, 1H), 2.73 (m, 2H). Analysis:
C27H32N4O2·1.60 H2O (C = 68.12; H = 7.25; N = 11.00).
5.1.3. Method C (series 3): 4-[5-(cyclohexanecarbonyl-
amino)-1H-benzoimidazol-2-yl]-N-cyclohexyl-benzamide (3m,
Y = H)
5.1.4. Method D (series 4): 2-(4-adamantanecarboxamido)
phenyl)-N-(pyridin-2-yl)-1H-benzoimidazole-5-carboxamide
(4z, Y = H)
5.1.3.1. 4-(5-Nitro-1H-benzoimidazol-2-yl)-benzoic
acid
methyl ester (15, Y = CH3). A mixture of 4 nitrophenylene-
1,2-diamine (5) (6.34 g, 41.4 mmol) and methyl 4-formyl-
benzoate (11) (6.80 g, 41.4 mmol) was heated in nitrobenzene
(100 ml) at 155–160 °C for 18 h. The reaction mixture was
cooled to room temperature and diluted with Et2O (500 ml),
and solid 15 was filtered and washed with ether (8.0 g, 56%).
5.1.4.1. 2-(4-Nitrophenyl)-1H-benzoimidazole-5-carboxylic
acid (20, Y = H). A solution of 12.5 g (82.15 mmol) 3,4-dia-
minobenzoic acid (10) and 4-nitrobenzaldehyde (19) (12.5 g
82.7 mmol) in nitrobenzene (250 ml) was heated at
155–160 °C for 18 h. The reaction mixture was cooled at room
temperature and diluted with diethyl ether (1 l). The yellow
solids were collected by filtration, dissolved in hot DMF, trea-
ted with charcoal, and filtered. The filtrate was diluted with
5.1.3.2. N-Cyclohexyl-4-(5-nitro-1H-benzoimidazol-2-yl)-ben-
zamide (17, R2 = cyclohexyl, Y = H). To a solution of nitro es-
ter 15 (8.0 g, 26.9 mmol) in THF (27 ml) and water (26 ml),
LiOH (3.39 g) was added and stirred at room temperature until
the hydrolysis was complete. The solvent was evaporated and
ice water was added. The reaction mixture was acidified with
3 N HCl and solid was filtered, washed with water, dissolved
in hot DMF and chilled to 4 °C. The yield of the acid 16 was
1
Et2O to precipitate out the nitro acid 20 (11.45 g, 55%); H
NMR (DMSO-d6, 400 MHz) δ 8.42 (m, 3H), 8.25 (br s, 1H),
7.94 (s, 1H), 7.88 (1H, d, J = 6.8 Hz, 1H), 7.71 (br s, 1H); ESI-
MS (m/z): 284.2 (M + H) and 282.4 (M – H).
1
5.37 g (70%); H NMR (DMSO-d6, 400 MHz) δ 13.80 (br s,
5.1.4.2. 2-(4-Nitrophenyl)-1H-benzoimidazole-5-carboxylic
acid pyridin-2-ylamide (21, R1 = 2-pyridine, Y = H). A solu-
tion of nitro acid 20 (3.4 g, 12.0 mmol) and CDI (2.90 g,
18.0 mmol) in dry DMF (30 ml) was heated at 80 °C for 3 h.
One equivalent of adamantaneamine hydrochloride (2.25 g,
12 mmol) was added twice at an interval of 2 h at room tem-
perature, and then heated at 100 °C for 18 h. The reaction
mixture was cooled to room temperature and diluted with
1H), 13.20 (br s, 1H), 8.55 (br s, 1H), 8.34 (m, 2H), 8.15 (m,
2H), 7.95 (br d, J = 16.0 Hz, 1H), 7.81 (br s, 1H); ESI-MS (m/
z): 284.3 (M + H).
To a solution of acid 16 (2.0 g, 7.0 mmol) in DMF (8.0 ml),
CDI (1.70 g, 10.5 mmol) was added and stirred at room tem-
perature for 2 h. Cyclohexyl amine (1.60 ml, 14 mmol) was
added to the reaction mixture and heated at 100 °C for 18 h.

M.L. Richards et al. / European Journal of Medicinal Chemistry 41 (2006) 950–969
967
water. The yellow nitro-amide 21 was filtered and sequentially
washed with water and MeOH. The yield was 2.80 g (65%); 1H
NMR (DMSO-d6, 400 MHz) δ 13.79 (br s, 1H), 10.56 (br s,
1H), 8.46 (m, 2H), 8.29 (m, 2H), 8.13 (br s, 1H), 7.85 (br d,
1H), 7.78–7.71 (br m, 3H), 7.18 (br s, 1H), 7.05 (dt, J = 5.60
and 4.40 Hz, 1H). ESI-MS (m/z): 360.9 (M + H).
ml) in DMEM/10% fetal bovine serum (FBS) at 37 °C, 10%
CO2.
5.2.2. Ex vivo IgE assay
The ex vivo Ig response assay involves in vivo antigen
priming and measures secondary antibody responses in vitro.
The basic protocol has been thoroughly documented and opti-
mized for a range of parameters including: antigen dose for
priming and time span following priming; number of cells cul-
tured in vitro; antigen concentrations for eliciting secondary
IgE (and other Igs) response in vitro; FBS batch that will per-
mit optimal IgE response in vitro; the importance of primed
CD4-positive T cells and hapten-specific B cells; and specifi-
city of the ELISA assay for IgE [22]. Female BALB/cByj mice
were immunized intraperitoneally (i.p.) with 10 μg dinitrophe-
nol-conjugated keyhole limpet hemocyanin (DNP6-KLH) ad-
sorbed onto 4 mg alum and sacrificed after 14–20 days.
Spleens were removed and homogenized in a tissue grinder,
washed twice, and spleen cell cultures (3 × 106 lymphocytes
per ml) established in 96-well round-bottom plates in the pre-
sence or absence of DNP6-KLH (10 ng/ml). Test compounds
were added to spleen cell cultures immediately prior to addi-
tion of antigen. Cultures were prepared in quadruplicate and
incubated for 7–12 days.
5.1.4.3. 2-(4-Amino-phenyl)-1H-benzoimidazole-5-carboxylic
acid pyridin-2-ylamide (22, R1 = 2-pyridine, Y = H). To a so-
lution of 2.5 g (7.26 mmol) nitro pyridine-benzimidazole-car-
boxamide (21) in EtOH/DMF (9:1, 50 ml), 5% Pd/C (0.625 g)
catalyst was added and the reaction mixture was heated under
H2 atmosphere at 90 °C for 3 h and at 70 °C for additional 3 h.
The amine was freed from catalyst and EtOH was evaporated.
The residue was diluted with water (100 ml) and product was
filtered and air-dried. The amine 22 was re-crystallized from
1
MeOH (yield: 0.80 g, 64%); H NMR (DMSO-d6, 400 MHz)
δ 12.71 (br s, 1H), 10.64 (br d, 1H), 8.38 (d, J = 4.0 Hz, 1H),
8.27 (br s, 1H), 8.21 (d, J = 6.40 Hz, 1H) 8.10 (br s, 1H), 7.89–
7.81 (m, 3H), 7.60–7.48 (dd, J = 6.80 Hz, 1H), 7.15 (dd,
J = 5.60 and 4.0 Hz, 1H), 6.68 (d, J = 6.80 Hz, 2H), 5.69–
5.66 (br d, J = 12.0 Hz 2H); ESI-MS (m/z): 330 (M + H).
5.1.4.4. 2-{4-[(Adamantane-1-carbonyl)-amino]-phenyl}-1H-
benzoimidazole-5-carboxylic acid pyridin-2-ylamide (4z,
Y = H). To a solution of amine 22 (1.2 g, 3.65 mmol) in
THF (15 ml) and pyridine (0.5 ml), adamantane-1-carbonyl
chloride (1.09 g, 5.19 mmol) was added and stirred at 80 °C
for 4 h. The resulting suspension was cooled and diluted with
water. The amide 4z was filtered and stirred with saturated
NaHCO3 solution for 2 h. The solid was filtered and washed
with water and re-crystallized from MeOH (yield: 0.80 g,
5.2.3. In vitro IgE assay
Spleen cells were obtained from naïve female BALB/c
mice, and T cells were depleted prior to culture by incubating
spleen cells first with a cocktail of anti-Thy1 ascites (10%),
anti-CD4 Ab (0.5 μg/ml) and anti-CD8 Ab (0.5 μg/ml), fol-
lowed by guinea pig complement (agarose adsorbed). Cells
were cultured at 3 × 106 per ml in quadruplicate. Culture con-
ditions were followed as described in the ex vivo protocol ex-
cept that instead of antigen, Ig production was stimulated by
IL-4 (10 ng/ml) and anti-CD40 antibody (100 ng/ml). After 5–
7 days of culture, 2 μl of supernatant was removed for quanti-
fying by ELISA. The ELISA protocol essentially follows as
described in the ex vivo protocol except that since the IgE is
not produced in response to specific antigen, the polyclonal
IgE is retained on the plates by coating them with GAME
(1 μg/ml).
1
45%); m.p. 328–330 °C; H NMR (DMSO-d6, 300 MHz) δ
13.10 (br s, 1H), 10.73 (s, 1H), 9.38 (s, 1H), 8.40 (br d,
J = 3.0 Hz), 8.29 (br s, 1H), 8.21 (d, J = 6.0 Hz, 1H), 8.14
(d, J = 9.0 Hz), 7.93–7.82 (m, 4 H), 7.64 (br d, J = 9.0 Hz,
1H), 7.17 (dt, J = 6.0 and 3.0 Hz, 1H), 2.04 (br s, 3H), 1.95
(s, 6H), 1.73 (s, 6H). ESI-MS (m/z): 492.6 (M + H). Analysis:
C30H29N5O2·0.60 H2O (C = 71.43; H = 5.90; N = 14.13).
5.2. Biological assays
5.2.1. Materials
Goat anti-mouse IgE (GAME) and biotinylated GAME (b-
GAME) were either prepared in-house or purchased from
Bethyl Labs and The Binding Site, respectively. Murine and
human recombinant IL-4 was obtained from Peprotech and
Boehringer-Manheim, respectively. Anti-mouse and anti-hu-
man CD40 antibodies were obtained from Pharmingen. Female
8–12-week-old BALB/cByj mice were purchased from Jackson
Laboratories and used in all mouse cell-based experiments.
Spleen cells were obtained by sacrificing the mice by cervical
dislocation, removal of the spleens, dispersing cells in a tissue
grinder, filtering through nytex, washing twice with phosphate-
buffered saline (PBS), and culturing (5 × 106 lymphocytes per
5.2.4. Oral bioavailability
Female BALB/c mice (five mice per compound) were ad-
ministered a single dose of 40 mg/kg suspended at 10 mg/ml in
0.2 M citric acid (pH 4). After 1 h mice were bled and the
samples pooled. Serum was obtained by centrifuging twice to
eliminate cells, adding an equal volume of acetonitrile, vortex-
ing, centrifuging, and collecting the eluate. The solution phase
was dried in a vacufuge (Eppendorf), sonicated in mobile
phase (0.1% TFA in water/acetonitrile, 30:70), centrifuged,
and analyzed by reverse-phase HPLC. Sample compounds
were quantified by comparison to a standard curve.

968
M.L. Richards et al. / European Journal of Medicinal Chemistry 41 (2006) 950–969
5.2.5. IgE response in human cells
were typically 20, 2000, and 10 μg/ml, respectively, on day
21 after antigen challenge.
Freshly drawn human peripheral blood was diluted 1:1 with
sterile PBS and layered over 10 ml of Histopaque. The tubes
were centrifuged at 1800 rpm (400 × g) for 30 min at room
temperature. The plasma layer (yellow) was aspirated and the
white cells collected by removing the interphase. This layer
was transferred to a new 50 ml conical tube containing 20 ml
of sterile PBS. White blood cells were counted, centrifuged at
1000 rpm (250 × g) for 10 min, and re-suspended in 15 ml of
sterile PBS. Human PBL cultures (5 × 105 cells per ml) were
established in quadruplicate in the presence of human IL-4
(Boehringer-Manheim) and anti-human CD40 Ab (Pharmin-
gen), and incubated at 37 °C; 10% CO2 for 10–14 days before
harvesting supernatants for the ELISA.
Human IgE was quantified from PBL culture supernatants
on round-bottom plates that were coated overnight with 1 μg/
ml GAHE (Sigma) in PBS. The plates were washed and
blocked with 200 μl per well of 1% BSA in PBS/azide. After
washing, a solution of BSA/PBS/Tween/azide was added to all
wells, and 15 μl of the PBL culture supernatant was added to
the corresponding plates. DE-52-purified IgE from a plasmacy-
toma (PS) was used as standard. Samples and standards were
allowed to bind overnight at 4 °C. The plates were washed and
incubated with 50 μl of 1 μg/ml biotin anti-human IgE (the
binding site) for 1.5 h at 37 °C. After washing the plates,
50 μl of the 1:1000 dilution of AP-streptavidin was added
and incubated an additional 1.5 h. After washing, 100 μl of a
40 mg/ml working solution of 26 × PPMP was added and the
absorbance read at 540λ.
5.2.6. In vivo IgE response
Female BALB/c mice (5–10 per group) were administered
250 rads (except control group) followed by immunizing with
2 μg KLH in alum (4 mg) i.p. 2 h later [59,60]. 1mm was
injected i.p. on 3 consecutive days starting 6 days later. On
day 2 of the drug injections, 2 μg of DNP6-KLH with alum
(4 mg) was injected i.p. Periorbital bleeds were obtained
21 days following DNP6-KLH challenge and the serum was
processed for quantification of antigen-specific antibody by
ELISA.
5.2.8. Cytokine assays
Drug activity was tested in two distinct assay systems utiliz-
ing spleen cells from female BALB/c mice. For the in vitro
assay, T cells were isolated using Pan-T (CD3ε) magnetic bead
kit (Miltenyi Biotec) and cultured for 48 h in the presence of
ConA (5 μg/ml) in the presence and absence of drug. Culture
supernatants were tested for IL-4 and IL-5 content using fluor-
escent beads from a cytokine detection kit (Upstate) and ana-
lyzed by a Luminex 100 flow analyzer. IL-4 and IL-5 were
quantified by comparison to standard curves. The results ob-
tained with this method were found to agree well with ELISA
assays using reagents and protocols obtained from BD-Phar-
mingen (not shown).
For the ex vivo cytokine response assay, spleens were re-
moved from mice that had been sensitized to DNP6-KLH, as
described for the ex vivo IgE response assay. Two weeks fol-
lowing sensitization, mice were sacrificed and spleen cell cul-
tures established in the presence of DNP6-KLH. After 4 days,
ConA (5 mg/ml) was added to cultures and the cells incubated
for an additional 4 h. Cytokine levels in the culture superna-
tants were quantified by ELISA using reagents and protocols
supplied by Pharmingen. Control responses were determined
from cultures that contained no drug.
5.2.7. Ig ELISA
To detect DNP-specific antibodies in cultures of mouse
spleen cells, ELISA plates are prepared by coating with speci-
fic antigen (DNP-OVA) overnight. After washing and blocking
the plates with 200 μl bovine serum albumin (BSA) in PBS, an
aliquot (1:4 dilution in PBS with 1% BSA 0.1% azide 0.5%
Tween 20) of each culture supernatant was transferred directly
to the ELISA plates and incubated overnight in a humidified
box at 4 °C. IgE was quantified following successive 90 min
incubations with biotinylated GAME (prepared in-house), alka-
line phosphatase-streptavidin (Zymed), and 100 μl of phe-
nolphthalein monophosphate (PPMP, DCHA salt, 40 mg/ml).
Absorption was measured at 540λ. The level of detection for
IgE is about 200–400 pg/ml and there is less than 0.001%
cross-reactivity with any other Ig isotype in the ELISA for
IgE [56].
5.2.9. Murine CD23 expression
To measure DNP-specific antibodies from serum, ELISA
plates were prepared as with the ex vivo assay by coating with
50 μl DNP15-conjugated ovalbumin (DNP15-OVA, 5 μg/ml).
Serum samples were diluted in saline and tested in triplicate
as follows: for antigen-specific IgE and IgG2a, dilute 1:200,
1:400, 1:800, and 1:1600; for IgG1, dilute 1:16,000,
1:32,000, 1:64,000, and 1:128,000. Plates were stored over-
night at 4 °C in a humidified chamber, washed, and treated
as described for the ex vivo ELISA. Antigen-specific IgG1
and IgG2a were measured similarly except that biotinylated
(b)-GAMG1 and b-GAMG2a (Southern Biotechnology As-
sociates) were substituted for b-GAME. Quantification of each
isotype was determined by comparison to a standard curve. In
control animals, serum DNP-specific IgE, IgG1, and IgG2a
Spleen cells (106 per ml) were dispersed from the spleens of
non-sensitized BALB/c mice and cultured for 24 or 48 h. Ex-
pression of CD23 was stimulated with IL-4 (10 ng/ml) and
anti-CD40 antibody (100 ng/ml), while parallel cultures con-
tained the 2-phenyl benzimidazole compound. Flow cytometry
was carried out on a FACScan (Becton-Dickinson) using
FITC-conjugated monoclonal anti-CD23 antibody (B3B4,
Pharmingen).
Acknowledgements
We thank Dr. Sunil Kumar KC for critical review of the
manuscript.

M.L. Richards et al. / European Journal of Medicinal Chemistry 41 (2006) 950–969
969
[31] G. Brusselle, J. Kips, G. Joos, H. Bluethmann, R. Pauwels, Am. J. Re-
spir. Cell Mol. Biol. 12 (1995) 254–259.
References
[32] D.B. Corry, H.G. Folkesson, M.L. Warnock, D.J. Erle, M.A. Matthay,
J.P. Wiener-Kronish, et al., J. Exp. Med. 183 (1996) 109–117.
[33] D.C. Webb, A.N. McKenzie, A.M. Koskinen, M. Yang, J. Mattes, P.S.
Foster, J. Immunol. 165 (2000) 108–113.
[34] G.J. Gleich, J. Allergy Clin. Immunol. 85 (1990) 422–436.
[35] M. Wills-Karp, C.L. Karp, Science 305 (2004) 1726–1729.
[36] D.Y. Wang, P. Clement, Am. J. Rhinol. 14 (2000) 325–333.
[37] J.M. Wang, A. Rambaldi, A. Biondi, Z.G. Chen, C.J. Sanderson, A.
Mantovani, Eur. J. Immunol. 19 (1989) 701–705.
[1] G.J. Rodrigo, C. Rodrigo, J.B. Hall, Chest 125 (2004) 1081–1102.
[2] H.W. Kelly, Pharmacotherapy 17 (1997) 13S–21S.
[3] F.D. Martinez, Respir. Res. 2 (2001) 129–132.
[4] S. Romagnani, J. Allergy Clin. Immunol. 113 (2004) 395–400.
[5] J.J. McIntire, D.T. Umetsu, R.H. DeKruyff, Springer Semin. Immuno-
pathol. 25 (2004) 335–348.
[6] D.T. Umetsu, Nat. Med. 10 (2004) 232–234.
[7] C. Walker, E. Bode, L. Boer, T.T. Hansel, K. Blaser, J.C. Virchow Jr.,
Am. Rev. Respir. Dis. 146 (1992) 109–115.
[38] H.D. Campbell, W.Q. Tucker, Y. Hort, M.E. Martinson, G. Mayo, E.J.
Clutterbuck, et al., Proc. Natl. Acad. Sci. USA 84 (1987) 6629–6633.
[39] B. Gregory, A. Kirchem, S. Phipps, P. Gevaert, C. Pridgeon, S.M. Ran-
kin, et al., J. Immunol. 170 (2003) 5359–5366.
[40] J. Drazen, Am. J. Respir. Crit. Care Med. 157 (1998) S233–S237 (dis-
cussion S247–S248).
[41] M. Hasegawa, K. Takenouchi, K. Takahashi, T. Takeuchi, K. Komoriya,
Y. Uejima, et al., J. Med. Chem. 40 (1997) 395–407.
[42] Y. Tada, I. Yamawaki, S. Ueda, H. Matsumoto, N. Matsuura, M. Yasu-
moto, et al., J. Med. Chem. 41 (1998) 3330–3336.
[8] D.S. Robinson, Q. Hamid, S. Ying, A. Tsicopoulos, J. Barkans, A.M.
Bentley, et al., N. Engl, J. Med. 326 (1992) 298–304.
[9] S. Matsushita, M.L. Richards, D.H. Katz, in: E.B. Weiss, M. Stein
(Eds.), Bronchial Asthma, Mechanisms and Therapeutics, third ed, Little,
Brown and Company, Boston, 1993, pp. 57–67.
[10] H.J. Gould, B.J. Sutton, A.J. Beavil, R.L. Beavil, N. McCloskey, H.A.
Coker, et al., Annu. Rev. Immunol. 21 (2003) 579–628.
[11] J.G. Scadding, K.M. Moser, in: E.B. Weiss, M. Stein (Eds.), Bronchial
Asthma, Mechanisms and Therapeutics, third ed, Little, Brown and Com-
pany, Boston, 1993, pp. 3–14.
[12] B.A. Berman, J. Allergy Clin. Immunol. 81 (1988) 980–984.
[13] P.J. Barnes, K.F. Chung, C.P. Page, Pharmacol. Rev. 50 (1998) 515–
596.
[14] C. Taube, A. Dakhama, E.W. Gelfand, Int. Arch. Allergy Immunol. 135
(2004) 173–186.
[15] H. Milgrom, R.B. Fick Jr., J.Q. Su, J.D. Reimann, R.K. Bush, M.L. Wa-
trous, et al., N. Engl, J. Med. 341 (1999) 1966–1973.
[16] M. Haak-Frendscho, K. Robbins, R. Lyon, R. Shields, J. Hooley, M.
Schoenhoff, et al., Immunology 82 (1994) 306–313.
[17] H. Lin, K.M. Boesel, D.T. Griffith, C. Prussin, B. Foster, F.A. Romero,
et al., J. Allergy Clin. Immunol. 113 (2004) 297–302.
[18] P. Yu, M. Kosco-Vilbois, M.L. Richards, G. Kohler, M.C. Lamers, Nat-
ure 369 (1994) 753–756.
[19] L. Flores-Romo, J. Shields, Y. Humbert, P. Graber, J.P. Aubry, J.F. Gau-
chat, et al., Science 261 (1993) 1038–1041.
[20] A. Saxon, M. Kurbe-Leamer, K. Behle, E.E. Max, K. Zhang, J. Immu-
nol. 147 (1991) 4000–4006.
[21] A. Stief, G. Texido, G. Sansig, H. Eibel, G. Le Gros, H. van der Putten,
J. Immunol. 152 (1994) 3378–3390.
[22] G. Texido, H. Eibel, G. Le Gros, H. van der Putten, J. Immunol. 153
(1994) 3028–3042.
[23] A. Haczku, K. Takeda, E. Hamelmann, A. Oshiba, J. Loader, A. Joe-
tham, et al., Am. J. Respir. Crit. Care Med. 156 (1997) 1945–1955.
[24] R.M. Ten, M.J. McKinstry, G.D. Bren, C.V. Paya, J. Allergy Clin. Im-
munol. 104 (1999) 376–387.
[25] M.R. Kehry, L.C. Yamashita, Proc. Natl. Acad. Sci. USA 86 (1989)
7556–7560.
[26] U. Pirron, T. Schlunck, J.C. Prinz, E.P. Rieber, Eur. J. Immunol. 20
(1990) 1547–1551.
[27] E.A. Clark, J.A. Ledbetter, Nature 367 (1994) 425–428.
[28] L.B. Bacharier, R.S. Geha, J. Allergy Clin. Immunol. 105 (2000) S547–
S558.
[43] G. Hochhaus, L. Brookman, H. Fox, C. Johnson, J. Matthews, S. Ren, Y.
Deniz, Curr. Med. Res. Opin. 19 (2003) 491–498.
[44] Y. Yanagihara, M. Kiniwa, K. Ikizawa, T. Shida, N. Matsuura, A. Koda,
Jpn. J. Pharmacol. 61 (1993) 31–39.
[45] Y. Yanagihara, M. Kiniwa, K. Ikizawa, H. Yamaya, T. Shida, N. Mat-
suura, et al., Jpn. J. Pharmacol. 61 (1993) 23–30.
[46] P.D. Mehlhop, M. van de Rijn, A.B. Goldberg, J.P. Brewer, V.P. Kurup,
T.R. Martin, et al., Proc. Natl. Acad. Sci. USA 94 (1997) 1344–1349.
[47] M.J. Leckie, A. ten Brinke, J. Khan, Z. Diamant, B.J. O’Connor, C.M.
Walls, et al., Lancet 356 (2000) 2144–2148.
[48] C. Buttner, A. Lun, T. Splettstoesser, G. Kunkel, H. Renz, Eur. Respir. J.
21 (2003) 799–803.
[49] S.A. Bryan, B.J. O’Connor, S. Matti, M.J. Leckie, V. Kanabar, J. Khan,
et al., Lancet 356 (2000) 2149–2153.
[50] H. Renz, K. Enssle, L. Lauffer, R. Kurrle, E.W. Gelfand, Int. Arch. Al-
lergy Immunol. 106 (1995) 46–54.
[51] T.A. Sato, M.B. Widmer, F.D. Finkelman, H. Madani, C.A. Jacobs, K.H.
Grabstein, et al., J. Immunol. 150 (1993) 2123–2717.
[52] L.J. Rosenwasser, W.W. Busse, R.G. Lizambri, T.A. Olejnik, M.C. To-
toritis, J. Allergy Clin. Immunol. 112 (2003) 563–570.
[53] F.E. Hargreave, J. Dolovich, M.T. Newhouse, J. Allergy Clin. Immunol.
85 (1990) 1098–1111.
[54] E.A. Mitchell, J.M. Bland, J.M. Thompson, Thorax 49 (1994) 33–36.
[55] M.L. Richards, S.C. Lio, A. Sinha, K.K. Tieu, J.C. Sircar, J. Med. Chem.
47 (2004) 6451–6454.
[56] J.F. Marcelletti, D.H. Katz, Cell. Immunol. 135 (1991) 471–489.
[57] N. Chiorazzi, D.A. Fox, D.H. Katz, J. Immunol. 117 (1976) 1629–1637.
[58] D.A. Fox, N. Chiorazzi, D.H. Katz, J. Immunol. 117 (1976) 1622–1628.
[59] K. Zhang, E.A. Clark, A. Saxon, J. Immunol. 146 (1991) 1836–1842.
[60] R.L. Coffman, J. Ohara, M.W. Bond, J. Carty, A. Zlotnik, W.E. Paul, J.
Immunol. 136 (1986) 4341–4538.
[61] M.L. Richards, D.H. Katz, Crit. Rev. Immunol. 11 (1991) 65–86.
[62] J. Punnonen, G. Aversa, B.G. Cocks, J.E. de Vries, Allergy 49 (1994)
576–586.
[29] G. Grunig, M. Warnock, A.E. Wakil, R. Venkayya, F. Brombacher,
D.M. Rennick, et al., Science 282 (1998) 2261–2263.
[30] N. Miyahara, K. Takeda, T. Kodama, A. Joetham, C. Taube, J.W. Park,
et al., J. Immunol. 172 (2004) 2549–2558.
[63] M.L. Richards, D.H. Katz, J. Immunol. 158 (1997) 263–272.