Characterization of new PPARγ agonists: Benzimidazole derivatives - Importance of positions 5 and 6, and computational studies on the binding mode

Article abstract of DOI:10.1016/j.bmc.2010.06.102

In this and previous studies we investigated the importance of partial structures of Telmisartan on PPARγ activation. The biphenyl-4-ylmethyl moiety at N1 and residues at C2 of the central benzimidazole were identified to be essential for receptor activation and potency of receptor binding. Now we focused our attention on positions 5 and 6 of the central benzimidazole and introduced bromine (3b-5/6, 3c), phenylcarbonyl (3d-5/6), hydroxy(phenyl)methyl (3g-5/6), hydroxymethyl (3h-5/6) and formyl (3i) groups. The selection of these moieties was inspired by the structure of Losartan and its metabolite EXP3179. In order to increase the hydrophobicity of the central part of the molecule, the benzimidazole was exchanged by a naphtho[2,3-d]imidazole (5). The compounds 3a-3i and 5 were tested in a differentiation assay using 3T3-L1 preadipocytes and a luciferase assay using COS-7 cells, transiently transfected with pGal4-hPPARγDEF, pGal5-TK-pGL3 and pRL-CMV, as established models for the assessment of cellular PPARγ activation. An enhanced effect on PPARγ activation could be observed if lipophilic moieties are introduced in these positions. 4′-[(2-Propyl-1H-naphtho[2,3-d]imidazol-1-yl)methyl]biphenyl-2- carboxylic acid (5) was identified as the most potent compound with an EC ₅₀ of 0.26 μM and the profile of a full agonist. Together with compounds of the former structure-activity relationship study (position 2-substituted benzimidazole derivatives 4a-4j), the binding mode of Telmisartan and its derivatives have been analyzed in 3D pharmacophore-driven docking experiments.

Full text of DOI:10.1016/j.bmc.2010.06.102

Bioorganic & Medicinal Chemistry 18 (2010) 5885–5895
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Characterization of new PPARc agonists: Benzimidazole
derivatives—importance of positions 5 and 6, and computational
studies on the binding mode
Matthias Goebel a, Gerhard Wolber a, Patrick Markt b, Bart Staels c, Thomas Unger d, Ulrich Kintscher d,
Ronald Gust a,e,
*
a Institute of Pharmacy, Freie Universität Berlin, Königin-Luise Strasse 2+4, 14195 Berlin, Germany
b Center for Molecular Biosciences and Institute of Pharmacy, Department of Pharmaceutical Chemistry, Computer-Aided Molecular Design Group, University of Innsbruck,
Innrain 52c, 6020 Innsbruck, Austria
c INSERM U545, Institut Pasteur de Lille, Faculté de Pharmacie et Faculté de Médecine, Université de Lille 2, 59000 Lille, France
d Center for Cardiovascular Research (CCR), Institute of Pharmacology, Charité-Universitätsmedizin Berlin, Hessische Strasse 3-4, 10115 Berlin, Germany
e Institute of Pharmacy, Department of Pharmaceutical Chemistry, University of Innsbruck, Innrain 52c, 6020 Innsbruck, Austria
a r t i c l e i n f o
a b s t r a c t
Article history:
In this and previous studies we investigated the importance of partial structures of Telmisartan on PPAR
c
Received 5 November 2009
Revised 28 June 2010
Accepted 29 June 2010
Available online 3 July 2010
activation. The biphenyl-4-ylmethyl moiety at N1 and residues at C2 of the central benzimidazole were
identified to be essential for receptor activation and potency of receptor binding. Now we focused our
attention on positions 5 and 6 of the central benzimidazole and introduced bromine (3b–5/6, 3c), phenyl-
carbonyl (3d–5/6), hydroxy(phenyl)methyl (3g–5/6), hydroxymethyl (3h–5/6) and formyl (3i) groups.
The selection of these moieties was inspired by the structure of Losartan and its metabolite EXP3179.
In order to increase the hydrophobicity of the central part of the molecule, the benzimidazole was
exchanged by a naphtho[2,3-d]imidazole (5). The compounds 3a–3i and 5 were tested in a differentiation
assay using 3T3-L1 preadipocytes and a luciferase assay using COS-7 cells, transiently transfected with
Keywords:
Benzimidazole derivatives
Partial agonism
Selective PPARc modulators
Structure–activity relationship
pGal4-hPPAR
PPAR activation. An enhanced effect on PPAR
introduced in these positions. 40-[(2-Propyl-1H-naphtho[2,3-d]imidazol-1-yl)methyl]biphenyl-2-carbox-
c
DEF, pGal5-TK-pGL3 and pRL-CMV, as established models for the assessment of cellular
c
c
activation could be observed if lipophilic moieties are
Molecular modeling
ylic acid (5) was identified as the most potent compound with an EC50 of 0.26 lM and the profile of a full
agonist.
Together with compounds of the former structure–activity relationship study (position 2-substituted
benzimidazole derivatives 4a–4j), the binding mode of Telmisartan and its derivatives have been ana-
lyzed in 3D pharmacophore-driven docking experiments.
Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
tion of new PPARc ligands that retain metabolic efficacy without
exerting adverse actions plays a central role in the development
of new therapeutic strategies for insulin resistance and type 2 dia-
betes mellitus.7 A promising new group of such ligands are selec-
The Peroxisome Proliferator-Activated Receptor
c (PPARc) is a
major regulator of adipogenesis and also involved in glucose
homoeostasis, cell differentiation and inflammation.1,2 The recep-
tor belongs to the nuclear receptor superfamily of ligand-activated
transcription factors.3–6 Glitazones or thiazolidinediones, such as
Pioglitazone or Rosiglitazone, are high affinity ligands and full ago-
nists for this receptor and are currently used in the treatment of
type 2 diabetes mellitus. Because of the side effects of glitazones
such as weight gain, edema and fluid retention, the characteriza-
tive PPARc modulators (SPPARcMs), compounds that activate
only a subset of the functions induced by cognate ligands or act
in a cell-type selective manner.8 It has been shown that a subgroup
of angiotensin II receptor blockers (ARBs) induced PPARc activity,
among which Telmisartan showed the highest activating proper-
ties.9,10 In addition, there is accumulating evidence for Telmisartan
and Irbesartan to act like SPPARc
Ms.11
We already demonstrated in a previous study that a ‘backbone’
part of Telmisartan containing the basic scaffold of the 1-(biphe-
nyl-4-ylmethyl)-1H-benzimidazole seems to be absolutely neces-
sary to maintain PPAR
alkyl chain in position 2 of the benzimidazole core, especially the
* Corresponding author. Address: Institut für Pharmazie, Abteilung für Phar-
mazeutische Chemie, Universität Innsbruck, Innrain 52c, 6020 Innsbruck, Austria.
Tel.: +43 512 507 5245; fax: +43 512 507 2940.
c
activation (Fig. 1).12 Furthermore, the
E-mail addresses: ronald.gust@uibk.ac.at, rgust@zedat.fu-berlin.de (R. Gust).
0968-0896/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2010.06.102

5886
M. Goebel et al. / Bioorg. Med. Chem. 18 (2010) 5885–5895
influence on
partial activity?
influence on
activation
and potency
Telmisartan. An additional N-methyl group at this substituent led
to PPAR
agonists even more potent than Telmisartan.12 Therefore,
c
we decided to evaluate the importance of these positions more de-
tailed using also partial structures of Losartan and its metabolite
EXP3179.14,15 As depicted in Figure 2 and Table 1 bromine, carbinol
and carbonyl derivatives were selected to study the hydrophobic
interactions as well as possible H-donor and H-acceptor binding
at the receptor. Bromine was favoured because of the large binding
pocket (see 3D modelling SAR studies) and the higher hydropho-
bicity compared to chlorine and fluorine.
influence on
activation
influence on
potency
N
N
N
N
CO2H
minimum
requirement
Telmisartan
2. Results
2.1. Chemistry
Compounds 1a–1f were commercially available. Cycloconden-
sation of the respective 1,2-benzenediamine (1a–1f) with tri-
methyl orthobutyrate by dropwise addition of concd HCl
(Scheme 1) yielded the 2-propyl substituted benzimidazoles 2a–
2f. The products 3a–3g were generated by N-alkylation of the
benzimidazole derivatives 2a–2d with 40-(bromomethyl)biphe-
nyl-2-carbonitrile in dry DMF and NaH, with a following nitrile
saponification with KOH in ethylene glycol at 185 °C. Under these
conditions a reduction of the carbonyl function to the hydroxyl
group took place resulting in compound 3g–5/6.
R1
R2
N
N
positions of
interest
CO2H
Figure 1. Central 1-(biphenyl-4-ylmethyl)-1H-benzimidazole structure of Telmi-
sartan as a basis for new PPAR
c
ligands with the focus of interest on position 5 and
6 of the benzimidazole.
Prior to the N-alkylation yielding the compounds 3h–5/6 and 3i,
the intermediate 2e was reduced with LiAlH4 in dry THF at ꢀ10 °C,
and the nitrile group of 2f was converted to a formyl group with
Ni–Al alloy (Raney Nickel) in formaldehyde at 95 °C.16
Finally, compound 5 was obtained in the same way starting
with the cyclocondensation of 2,3-diaminonaphthalene with tri-
methyl orthobutyrate as described above (Scheme 2).17
Cl
N
Cl
N
N
N
N
N
N
N
NH
NH
HO
O
N
N
CYP2C9
H
Separation of the regioisomers was carried out by a combina-
tion of silica gel column chromatography and fractional crystalliza-
tion. The separation failed in the case of compound 3i (ratio of the
5- to the 6-regioisomer: 6:4), also using preparative HPLC. Com-
pounds 3d–6 and 3g–6 could only be obtained as a 33% enriched
6-regioisomer. Compounds 3d–5 and 3g–5 were isomerically pure.
The assignment of the moieties to the position 5 or 6 (3b–5/6,
3d–5/6, 3g–5/6 and 3h–5/6) was performed by differential 1H
NMR Nuclear Overhauser Effect experiments (NOE-Diff), on the ba-
sis of a saturation transfer from the C-7 positioned proton (H-7) at
the central benzimidazole to the benzylic methylene group (see
Figs. S1–S16 at Supplementary data).
Losartan
EXP3179
N
N
N
N
HO
O
CO2H
CO2H
H
3h-5
3i
2.2. In vitro SAR studies
N
N
N
N
Compounds were evaluated for PPAR
c activation in vitro.
HO
O
PPAR is known as the ‘master regulator’ of adipocyte differentia-
c
CO2H
CO2H
tion. Its activity closely correlates with the degree of differentiation
analyzed by Oil Red O-staining. Therefore, 3T3-L1 pre-/adipocyte
differentiation was chosen as an established model for the assess-
ment of cellular PPAR
and 4B, concentrations of each compound: 1 and 10
(V): DMSO; positive controls: Pioglitazone (P) and Telmisartan
(T)). All test compounds were also investigated in a luciferase
transactivation assay, using COS-7 cells transiently transfected
c
activation (absorption bar chart in Figs. 3B
M; vehicle
3g-5
3d-5
l
Figure 2. One step of Losartan metabolism and derivation of compounds 3d–5, 3g–5,
3h–5 and 3i.
with pGal4-hPPAR
Pioglitazone, as a full PPAR
c
DEF and pGal5-Tk-pGL3 (Figs. 3A and 4A).
agonist, was used as a positive control
propyl group, was important for potency and activation.13 In con-
tinuation of our studies we investigated the significance of the
c
and its maximum activation was defined as 100%. The results of
both assays are summarized in Table 1 (EC50-values, percentage
of maximum luciferase activation and maximum adipocyte differ-
entiation). Additionally, a column with calculated log D-values
(ACD/Chemsketch 3.6) at physiological pH 7.4 was included, which
is helpful for interpretation of the receptor activity.
positions 5 and 6 of the benzimidazole for PPARc modulating
activity.
First hints on the meaning of these positions gave the results of
our first structure–activity relationship (SAR) study: compounds
with a benzimidazole at position 5 or 6 were less active than

M. Goebel et al. / Bioorg. Med. Chem. 18 (2010) 5885–5895
5887
Table 1
Overview of the in vitro data of the compounds 3a–3i and 5
Compounda
R1
R2
clog D (pH 7.4)b
EC50
(lM)
Amax (%)
Max. diff. at 10 l
Md (%)
c
c
Pioglitazone
Telmisartan
3a
3b–5
3b–6
—
—
H
Br
H
—
—
H
H
Br
Br
2.06
4.46
2.37
3.01
3.03
3.64
0.14
3.4
1.9
0.4
0.47
0.4
100
47
49
104
79
73
100
70
67
103
102
74
3c
Br
O
3d–5
3d–6
3g–5
3g–6
H
3.21
3.22
2.86
2.81
0.58
0.62
2.9
92
79
58
51
104
97
H
O
OH
H
100
H
3.5
88
OH
3i
5
Isomeric mixture of 5-/6-CH@O
1.68
3.61
10
37
45
0.26
109
105
a
b
c
Compounds 3d–6 and 3g–6 were tested as isomeric mixtures. Compounds 3h–5/6 were inactive
clog D-values were calculated with ACD/ChemSketch 3.6.
Data for EC50 and maximum activation (Amax) values were taken from the graphs of Figures 3A and 4A, calculated with GraphPad Prism 4.
d
Data for maximum differentiation values were taken from the bar chart of Figures 3B and 4B.
For the evaluation of the significance of the modifications com-
effect is nearly independent on the position (5 or 6) of the substi-
tuent: 3g–5 (EC50 = 2.9 M, max. activation: 58%) < 3d–5
(EC50 = 0.58 M, max. activation: 92%), 3g–6 (EC50 = 3.5 M, max.
activation: 51%) < 3d–6 (EC50 = 0.62 M, max. activation: 79%).
Replacement of the phenyl ring by H reduced the activity drasti-
cally: 3h–5/6 (no measurable activity, data not shown) < 3i
pound 3a was used for comparison. Its PPAR
c
activating property
l
was described recently.13
l
l
Figure 3 illustrates the results of compounds 3a, 3b–5, 3b–6, 3c
and 5 in the luciferase assay and the adipocyte differentiation as-
say. Compounds 3h–5 and 3h–6 showed no relevant effects (data
not shown). If a bromine substituent is attached to compound 3a
in position 6 (3b–6), the lipophilicity rises from a clog D value of
2.37 (3a) to 3.03 (3b–6), which also led to a higher potency
l
(EC50 = 10 lM, max. activation: 37%).
With substituents at position 5 it is possible to reach a higher
activation level than with those in position 6, keeping in mind that
3d–6, 3g–6 and 3i are isomeric mixtures. The concentration activ-
ity curves in the luciferase assay correlate well with the adipocyte
differentiation assay, indicating a higher potency with a lower dif-
ferentiation level for the 6 positioned compared to the 5 positioned
moieties (Fig. 4B, Table 1). Because the binding mode of Telmisar-
tan and related compounds is unknown we used these data and
those of previous papers for 3D modelling SAR studies.
(EC50 = 0.47 lM). However the profile of a partial agonist as dem-
onstrated for 3a (max. activation of 49%) is shifted more to that
of a full agonist (3b–6 max. activation of 79%). Adipocyte differen-
tiation reached 100%. The shift of the bromine substituent from po-
sition 6 to 5 (3b–5) increased the max. activation in the luciferase
assay to 104% but let the potency nearly unchanged
(EC50 = 0.4
positions was more lipophilic (clog D = 3.64) and caused results
(EC50 = 0.4 M; max. activation of 73%) like 3b–6 in the luciferase
lM). Compound 3c with bromine substituents in both
l
2.3. 3D modelling SAR studies
assay. Replacing the benzimidazole scaffold of 3a by an 1H-naph-
tho[2,3-d]imidazole yielded the most potent compound (5) in this
PPARc is a challenging target for molecular modeling due to
series (EC50 = 0.26
curve and the adipocyte differentiation behaviour were compara-
ble to Pioglitazone.
It is obvious from the results depicted in Figure 4 that the effi-
cacy is stronger in the case of benzophenone derivatives (3d–5 and
3d–6) compared to the carbinol congeners (3g–5 and 3g–6). The
l
M; max. activation of 109%). The activation
several reasons: (i) the active site is huge and consists of three
arms of which only two have to interact with the ligand to show
biological activity, (ii) the binding is considered flexible, that is,
especially side chains are known to move significantly upon ligand
binding, (iii) the binding site is highly lipophilic, which represents
a challenge for structure-based approaches.18–20 On the other

5888
M. Goebel et al. / Bioorg. Med. Chem. 18 (2010) 5885–5895
R1
R2
N
N
R1
R2
R1
R2
NH2
NH2
N
a
b, c
CO2H
N
H
1a-1f
2a-2f
3a-3i
R1
R2
R1
R2
R1
R2
H
1a
1b
H
H
H
2a
2b
H
H
H
3a
H
3b-5
3b-6
3c
Br
H
Br
Br
Br
Br
H
Br
Br
1c Br
2c Br
Br
Phenylcarbonyl
CO2H
Phenylcarbonyl
1d
1e
1f
H
2d
2e
2f
H
H
H
Phenylcarbonyl
3d-5
H
d, b, c
CO2H
CN
H
H
Phenylcarbonyl
3d-6
H
CN
3g-5 Hydroxy(phenyl)methyl
H
3g-6
Hydroxy(phenyl)methyl
H
Hydroxymethyl
3h-5
3h-6
3i
H
Hydroxymethyl
H
e, f, g
5-/6-Formyl (isomeric mixture)
Scheme 1. Reagents and conditions: (a) H3C(CH2)2C(OCH3)3, HCl, rt; (b) NaH, 40-(bromomethyl)-2-biphenylcarbonitrile, DMF, 0 °C–rt; (c) ethylene glycol, KOH, 185 °C;
(d) THF, LiAlH4, ꢀ10 °C to rt; (e) Ni–Al, formaldehyde, H2O, 95 °C; (f) NaH, methyl-40-(bromomethyl)-2-biphenylcarboxylate, DMF, 0 °C–rt; (g) MeOH, aq NaOH soln (10%), reflux.
N
NH2
a, b, c
N
CO2H
NH2
4
5
Scheme 2. Reagents and conditions: (a) H3C(CH2)2C(OCH3)3, HCl, rt; (b) NaH,
40-(bromomethyl)-2-biphenylcarbonitrile, DMF, 0 °C–rt; (c) ethylene glycol, KOH,
185 °C.
hand, there is a large amount of published experimental data and
successful molecular modeling studies available. For example Chit-
tiboyina et al. and Mizuno et al. published a study involving molec-
ular docking to facilitate their research for Telmisartan
analogs.21,22 Their docking study was focused on a single ligand–
PPARc complex and relies on docking scores for compound rank-
ing, while here several structures and pharmacophoric information
are used to perform modeling studies.
55 PPARc structures with different co-crystallized ligands have
been deposited in the PDB at the time of submission of this paper.
Moreover, several successful modeling applications for PPAR have
been published that focus on known pharmacophoric patterns.23–
26
Based on this large amount of available experimental data, we
were able to apply a more sophisticated docking work-flow.
In order to address the difficulties and to profit from the large
amount of experimental data, 83 single chains were extracted from
55 partly polymeric PDB structures. For all of these chains, re-dock-
ing experiments were performed using CCDCs software GOLD.28 If
the ligand was geometrically placed within a root mean square
deviation (RMSD) of below 4 Å to the co-crystallized conformation
by the program, it was considered for further examination. From
this set those structures were selected, in which the re-docked li-
gand showed the same chemical interactions as the experimentally
Figure 3. (A) Compounds tested in luciferase activation assay using COS-7 cells
transiently transfected with pGal4-hPPAR DEF and pGal5-Tk-pGL3. Firefly lucifer-
ase activity was measured after 36 h and normalized with activity of cotransfected
renilla luciferase. Points expressed are the means SD of threefold determination in
c
a
single experiment. Activation (%) of the luciferase gene in COS-7 cells by
Pioglitazone (5 was defined as 100%), Telmisartan and the synthesized
compounds 3a–3c and 5. (B) Lipid accumulation in 3T3-L1 cells after nine days of
differentiation, the indicated compounds at 1 and 10 M, was assessed by the
lM
l
quantification of Oil Red O after staining in isopropanol (OD515). DMSO as vehicle
(V), Pioglitazone (P) and Telmisartan (T) as positive controls.

M. Goebel et al. / Bioorg. Med. Chem. 18 (2010) 5885–5895
5889
quite restrictive selection, the chosen structures still cover a broad
range of side chain movements and different ligand poses (Fig. 5).
All resulting poses that fulfill all important pharmacophoric inter-
actions were below 2 Å RMSD compared to the co-crystallized li-
gand (Table 2).
As a next step, the backbones of the isolated and selected chains
were three-dimensionally aligned using the tool MUSTANG29 in or-
der to obtain a common coordinate frame. All compounds were
docked into all of these aligned chains without further restrictions.
Built-in scoring algorithms for ranking ligand poses are unapt to
estimate binding energy, while it is common understanding that
the resulting pose ensemble covers ligand conformational space
in a reasonable way.30
Since Telmisartan was the lead structure for this study, we first
tried to find an adequate binding mode for this compound. We
compared all generated poses from the parallel docking experi-
ments and selected the most plausible by visual inspection consid-
ering known important interactions from previous studies.23
Telmisartan most likely forms hydrogen bonds to Ser342 at the
carboxylic acid moiety, and shows at the same time a relevant
charge transfer interaction with Arg288. These chemical features
anchor the compound in arm II allowing hydrophobic contacts
with Met348, Ile341, and Phe264. The central benzimidazole acts
as a linker stabilizing the lipophilic parts of the molecule that point
into arm I (Ile326, Leu469, Leu453) and arm III (Ala292, Leu333), so
that the lipophilic part of the molecule can fill the hydrophobic
portions of the pocket (Fig. 6). This predicted pose is in correspon-
dence with the assumptions of Benson et al.,10 although the charge
transfer with Arg288 has not been described in detail. However,
this interaction seems to be relevant, since also the endogenous
Figure 4. (A and B) Showing compounds 3d–5, 3d–6, 3g–5, 3g–6 and 3i. Methodical
details are described in the legend of Figure 3.
PPAR
location.31
To further analyze and understand the mechanism of receptor
c ligand alkylglycerophosphate was observed at this
activation by the synthesized compounds, a structure–activity
relationship with a 3D pharmacophore-driven molecular model-
ling study has been performed. All the in vitro tested compounds
from this (compounds 3a, 3b–5, 3b–6, 3c, 3d–5, 3d–6, 3g–5, 3g–
6, 3i, and 5) and a former study (2-substituted benzimidazoles
4a–4j) have been included13 (compound 3a is consistent with
4a). From the pose shown in Figure 6, a three-dimensional phar-
macophore model was derived using the software LigandScout.32,33
All molecules were docked into the 12 selected structures and sub-
sequently scored by a three-dimensional search using the Telmi-
sartan 3D pharmacophore model built in LigandScout. The
surmised binding poses are shown in Figure 7. Additionally, each
resulting compound conformation is depicted separately in the
Supplementary data.
Table 2
Selected PDB structures for docking experiments
PDB code
Chain identifier
RMSD (Å) of re-docked ligand
2fvj
A
A
B
A
B
A
C
A
B
B
A
A
0.83
1.58
0.44
0.95
1.10
1.44
1.93
0.46
1.17
1.58
0.87
1.42
2g0g
2g0h
2g0h
2om9
2om9
2om9
2q5s
2q61
2q6s
3b3k
4prg
Figure 5. PDB structures as described in Table 2 after the backbone alignment
procedure and removal of all ligands. Although the backbone can be overlaid
without major deviations in most parts, there is still considerable side chain
flexibility. Side-chains of arm I are colored in green, important arm II residues in
orange, and arm III residues in yellow. Visualization was performed using VMD.27
determined pose. This procedure resulted in a high quality set of
protein chains for subsequent docking experiments. Despite this
All re-docked structures reproduce the important pharmacophoric interactions that
can also be observed for the co-crystallized ligand.

5890
M. Goebel et al. / Bioorg. Med. Chem. 18 (2010) 5885–5895
Figure 7. Overlay of structures 4a–4j after parallel docking and 3D pharmacophore
scoring in the PPAR binding pocket (shown using PDB entry 2FVJ). All structures
c
show the same anchoring point of the carboxylic acid moiety, while the lipophilic
regions are mainly distributed between arms I and III.
pound 4i could be that the top scored pose is likely to trap a water
molecule between the side chains of His323, His449, and Tyr473,
which is entropically unfavorable. Although this hypothesis would
have to be proved in more detailed theoretical studies, the geom-
etry that was produced by rigid docking allows for plausible water
placement using a simple MMFF94s force field minimization in
LigandScout as shown in Figure 8.
2.3.2. SAR of positions 5 and 6 of the central benzimidazole
The increase in activity observed for the bromine substituted
compounds 3b–5, 3b–6, and 3c in comparison to the unsubstituted
compound 3a can be easily explained by the rising lipophilicity and
the larger area of hydrophobic contact possibilities. Compounds
3d–5 and 3d–6 both show a linear geometry that nicely binds into
both arm I and III. The carbonyl group of the phenone moiety of
3d–5 not only forms hydrogen bonds to Tyr327, but also to
Ser289 (Fig. 9), a PPAR residue which is known to be important
for full agonistic activity.18 The docking pose determined for
3d–6 is not able to form such hydrogen bonds. This could be an
Figure 6. The putative binding mode of Telmisartan was used as a starting point.
The LigandScout 3D pharmacophore consists of
5 lipophilic features (yellow
spheres), a hydrogen bond (red arrow) and a charge transfer interaction (3D
depiction on the top (A) and 2D illustration below (B) visualized with LigandScout).
2.3.1. SAR of position 2 of the central benzimidazole13
As mentioned above, alkyl chains such as propyl substituents at
position 2 of the benzimidazole moiety are of special interest for
potent PPAR agonists. Also other scaffolds for PPAR agonists con-
taining such hydrophobic substituents have been published.34,35
To rationalize the importance of propyl and other hydrophobic
moieties at this position, we predicted the binding modes for com-
pounds 4a–4j (see Supplementary data) comprising different
hydrophobic substituents at position 2. These binding poses ex-
plain the increase of activity from propyl to butyl (4a and 4b) sub-
stituents, since the hydrophobic contact area is increased. The
slight decrease in activity for compounds 4c and 4d could be re-
lated to hydrophobic contacts that can only be formed at the tran-
sition between arms I and III. Thus, the ligands cannot reach out to
Tyr327 that interacts with 4a and 4b. The aromatic substitutions in
4e and 4g also permit lipophilic contacts with Met348, which leads
to an increase in activity. The phenyl ring in compound 4f shows
an unfavorable conformation for the linear geometry of arm I
and forms no hydrophobic contacts, whereas the chlorine substitu-
tion in compound 4h can interact with Leu228 in arm III. The
methoxy group in compound 4i reaches out to arm III, potentially
even forming a weak hydrophobic interaction with Tyr327.
An explanation for the decreased activity of phenol substitution
in compound 4j compared to the methoxyphenyl moiety of com-
Figure 8. Compound 4j potentially traps a water molecule between His323, His449,
and Tyr327, which is entropically unfavourable. Red and green arrows indicate
hydrogen bonding.

M. Goebel et al. / Bioorg. Med. Chem. 18 (2010) 5885–5895
5891
Figure 9. Compound 3d–5 (left) interacts with Tyr327 and Ser289 (illustrated by red vectors) resulting in full agonistic activity, while 3d–6 (right) cannot form these
hydrogen bonds due to its different orientation.
explanation for the fact that in contrast to 3d–6, 3d–5 acts as a full
agonist.
tuent changes at the 5- and 6-positions play a major role in recep-
tor activation. On the one hand, the results demonstrate the
importance of lipophilicity and on the other hand the direction of
these positions, so that one could assume that the shift from posi-
tion 6 to 5 leads from a partial to a full agonist. Thus, this needs to
be elucidated in further studies. The partial to full agonism shift
could also be seen for the exchange of the hydroxyl by a carbonyl
function, that is, 3g–5?3d–5, which could be explained by hydro-
gen bonds of the carbonyl oxygen with Tyr327 and Ser289. Above
all, the rigidity and geometry of the structures have a positive
influence on potency and maximum receptor activation, which is
exemplified in the following row: 3g–6 < 3d–6 < 5, where the most
The lack of biological activity determined for 3i could not be ex-
plained by the docking studies in this work. One plausible explana-
tion could be that the formyl moiety of 3i acts as an ‘electrophilic
warhead’, which could form a covalent bond with a residue located
at the entrance of the PPAR ligand-binding pocket, such as known
for PPAR
pound 3i could be forced into a conformation that does not allow
proper ligand–protein interactions necessary for PPAR activation.
c
antagonists that bind covalently to Cys285.36 Thus, com-
c
Compound 5, finally, shows a perfect complementarity to the arm
I/III binding pocket with a high degree of rigidity, which reflects its
biological activity (Fig. 10). Each docking pose is provided as a sep-
arate image in the Supplementary data for further details.
rigid compound
5 emerged as the most efficacious one
(EC50 = 0.26 M, max. activation: 109%), which could also be dem-
l
onstrated in the modeling study.
The 3D pharmacophore-driven docking of both the 2-substi-
tuted and the 5/6-substituted benzimidazoles with the PPARc cor-
3. Discussion and conclusion
related nicely with the in vitro data. It has been shown that a linear
elongation with lipophilic moieties (4e and 4g) results in better
receptor activity. An interference of this linearity (4f and 4h) de-
creased the activity, while a hydrophilic group (i.e., hydroxyl) in
this position leads almost to inactivity (4j). Similar conclusions
can be drawn for the 5/6-positioned substituents. These com-
pounds only give a little insight to the possible opportunities of
modulating the receptors activation by certain moieties (fine-tun-
ing), which could be able to release a distinct conformational
change and therefore a specific gene expression profile.
In this SAR study some new benzimidazole derivatives could be
characterized as PPARc ligands and it has been shown that substi-
4. Experimental section
4.1. Chemistry
All reagents and solvents were purchased from Acros Organics,
Sigma–Aldrich, Alfa Aesar or Merck. All reactions were monitored
by TLC, performed on silica gel plates 60 F254 (Merck, Darmstadt/
Germany). Visualization on TLC was achieved by UV light. Column
chromatography was performed with Merck Silica Gel 60H, grain
size <0.063 mm, 230 mesh ASTM (Darmstadt/Germany). Melting
points: B 545 Büchi (Flawil/Schweiz) capillary melting point appa-
ratus. 1H NMR: Avance DPX-400 spectrometer (Bruker, Karlsruhe/
Germany) at 400 MHz (internal standard: TMS). Elemental analy-
ses: Microlaboratory of the Freie Universität Berlin with Elemen-
taranalysator Vario EL (Elementar, Hanau); the analytical results
are within 0.4% of theoretical values. EIMS spectra: CH-7A-Varian
MAT, 70 eV (Melbourne/Australia). Microplate Reader FLASHScan
S12 (Analytik Jena AG, Jena/Germany) Microlumat: Victor2 1420
Figure 10. Overlay of structures 3a, 3b–5, 3b–6, 3c, 3d–5, 3d–6, 3g–5, 3g–6, 3i and
5 in the PPAR binding pocket. All structures show the same anchoring point of the
carboxylic acid moiety, while the lipophilic regions are mainly distributed between
c
arms I and III.

5892
M. Goebel et al. / Bioorg. Med. Chem. 18 (2010) 5885–5895
Multilabel Counter (Wallac, Perkin–Elmer, Life sciences, Turku/
Finnland).
4.3. General procedure for N-alkylation with 40-
(bromomethyl)biphenyl-2-carbonitrile
Compounds 1a (1,2-benzenediamine), 1b (4-bromo-1,2-
benzenediamine), 1c (4,5-dibromo-1,2-benzenediamine), 1d
(3,4-diaminobenzophenone), 1e (3,4-diaminobenzoic acid), 1f
(3,4-diaminobenzonitrile) were commercially available.
To a stirred solution of the appropriate secondary amine
(1 mmol) in dry DMF (3 mL) NaH (2 mmol) was added at 0 °C
(ice cooling). After approx. 30 min (or after no more visible emer-
gence of hydrogen) 40-(bromomethyl)biphenyl-2-carbonitrile
(0.3 g, 1.1 mmol) was added slowly and first stirred for 1 h at
0 °C and then 2–5 h at room temperature. The reaction mixture
was poured in 6 N HCl (1 mL) with crushed ice (25 g) and extracted
with CHCl3 (3 ꢁ 15 mL). The organic layers were combined and
dried over Na2SO4. After filtration, the solvent was removed under
reduced pressure, and the resulting crude product was purified by
chromatography on silica gel with stepwise gradient elution
(CH2Cl2/methanol 99:1, 98:2, 95:5).
4.2. General procedure for the preparation of 2-substituted 1H-
benzimidazoles
To a stirred suspension the appropriate 1,2-benzenediamine
(10 mmol) in the trimethyl orthobutyrate (6.4 mL, 40 mmol) at
room temperature concentrated HCl was added dropwise till the
suspension was cleared up and started an exothermic reaction. A
5%-NaHCO3 solution (50 mL) was transferred to the reaction mix-
ture at pH 8 and extracted with CHCl3 (3 ꢁ 25 mL). The organic lay-
ers were combined and dried over NaSO4. After filtration, the
solvent was removed under reduced pressure, and the resulting
crude product was purified by chromatography on silica gel with
CH2Cl2/methanol (95:5).
4.4. General procedure for saponification of carbonitriles
A stirred solution of the respective carbonitrile (1 mmol), KOH
(0.28 g, 5 mmol) in H2O (0.055 mL, 1 mmol) and ethylene glycol
(4 mL) were heated at 185 °C. Every hour of the reaction time per-
iod some H2O (1 mmol) has to be added cautiously. After 5–6 h the
reaction mixture was cooled to 100 °C, and H2O (8 mL) was added
subsequently. The solution was acidified with 6 N HCl (pH 5–6)
and stirred for 15 min to complete the precipitation. The obtained
solid was purified by chromatography on silica gel with stepwise
gradient elution (CH2Cl2/methanol 95:5, 9:1, 8:2) and recrystalliza-
tion from methanol.
4.2.1. 2-Propyl-1H-benzo[d]imidazole (2a)
From 1,2-benzenediamine (1 g, 9.24 mmol) and trimethyl ort-
hobutyrate (5.9 mL, 37 mmol). Colorless solid, 89%. 1H NMR
(DMSO-d6): d = 12.05 (s, 1H), 7.48–7.35 (m, 2H), 7.13–6.93 (m,
2H), 2.75 (t, J = 7.5 Hz, 2H), 1.77 (tq, J = 7.5 Hz, J = 7.4 Hz, 2H),
0.91 (t, J = 7.4 Hz, 3H).
4.4.1. 40-[(2-Propyl-1H-benzo[d]imidazol-1-yl)methyl]biphenyl-
2-carboxylic acid (3a)
4.2.2. 6-Bromo-2-propyl-1H-benzo[d]imidazole (2b)
From 4-bromo-1,2-benzenediamine (2 g, 10.7 mmol) and tri-
methyl orthobutyrate (6.9 mL, 43 mmol). Brown solid, 93%. 1H
NMR (DMSO-d6): d = 12.36 (s, 1H), 7.65 (s, 1H), 7.42 (d, J = 8.4 Hz,
1H), 7.24 (dd, J = 8.5 Hz, 1.9 Hz, 1H), 2.77 (t, J = 7.5 Hz, 2H), 1.77
(tq, J = 7.5 Hz, J = 7.4 Hz, 2H), 0.93 (t, J = 7.4 Hz, 3H).
N-Alkylation of 2a (0.5 g, 3.12 mmol) with 40-(bromo-
methyl)biphenyl-2-carbonitrile (0.933 g, 3.43 mmol). The resulting
40-[(2-propyl-1H-benzo[d]imidazol-1-yl)methyl]biphenyl-2-car-
bonitrile (91%, 1 g, 2.85 mmol) was treated with KOH (0.785 g,
14 mmol) as described for general procedure for saponification to
give a colorless solid, 43% (39% total); mp: 255 °C; 1H NMR
(DMSO-d6): d = 12.74 (s, 1H), 7.69 (d, J = 7.7 Hz, 1H), 7.60–7.58
(m, 1H), 7.54–7.49 (m, 2H), 7.43 (t, J = 7.5 Hz, 1H), 7.33 (d,
J = 7.7 Hz, 1H), 7.29 (d, J = 8.2 Hz, 2H), 7.18–7.16 (m, 2H), 7.12 (d,
J = 8.1 Hz, 2H), 5.54 (s, 2H), 2.84 (t, J = 7.5 Hz, 2H), 1.79 (tq,
J = 7.5 Hz, J = 7.4 Hz, 2H), 0.96 (t, J = 7.4 Hz, 3H); MS (EI, 225 °C):
370 [M+Å]. Anal. Calcd for C24H22N2O2: C, 77.81; H, 5.99; N, 7.56.
Found: C, 77.86; H, 6.17; N, 7.59.
4.2.3. 5,6-Dibromo-2-propyl-1H-benzo[d]imidazole (2c)
From 4,5-dibromo-1,2-benzenediamine (1 g, 3.76 mmol) and
trimethyl orthobutyrate (3 mL, 18.8 mmol). Brown solid, 90%. 1H
NMR (DMSO-d6): d = 12.49 (s, 1H), 7.87 (s, 2H), 2.78 (t, J = 7.5 Hz,
2H), 1.77 (tq, J = 7.5 Hz, J = 7.4 Hz, 2H), 0.93 (t, J = 7.4 Hz, 3H).
4.2.4. Phenyl(2-propyl-1H-benzo[d]imidazol-6-yl)methanone
(2d)
From 3,4-diaminobenzophenone (1 g, 4.7 mmol) and trimethyl
orthobutyrate (3 mL, 18.8 mmol). Colorless solid, 85%. 1H NMR of
the tautomeric mixture (DMSO-d6): d = 12.61–12.54 (m, 1H),
7.89–7.80 (m, 1H), 7.75–7.72 (m, 2H), 7.69–7.63 (m, 2H), 7.60–
7.55 (m, 3H), 2.87–2.80 (m, 2H), 1.87–1.76 (m, 2H), 0.99–0.93
(m, 3H).
4.4.2. 40-[(5-Bromo-2-propyl-1H-benzo[d]imidazol-1-yl)methyl]-
biphenyl-2-carboxylic acid (3b–5)
N-Alkylation of 2b (1 g, 2.85 mmol) with 40-(bromo-
methyl)biphenyl-2-carbonitrile (0.844 g, 3.1 mmol). After isomer
separation on silica gel chromatography the resulting 40-[(5-bro-
mo-2-propyl-1H-benzo[d]imidazol-1-yl)methyl]biphenyl-2-carbo-
nitrile (0.5 g, 1.16 mmol) was treated with KOH (0.325 g, 5.8 mmol)
as described for general procedure for saponification to give a color-
less solid, 73%; mp: 248 °C; 1H NMR (DMSO-d6): d = 12.70 (s, 1H),
7.78 (d, J = 1.8 Hz, 1H), 7.69 (dd, J = 7.6, J = 1.2 Hz, 1H), 7.53 (td,
J = 7.6, J = 1.4 Hz, 1H), 7.50–7.46 (m, 1H), 7.42 (td, J = 7.5, J = 1.2 Hz,
1H), 7.32–7.25 (m, 4H), 7.12–7.07 (m, 2H), 2.83 (t, J = 7.5 Hz, 2H),
1.75 (tq, J = 7.5 Hz, J = 7.4 Hz, 2H), 0.93 (t, J = 7.4 Hz, 3H); MS (EI,
180 °C): 448 [M+Å]. Anal. Calcd for C24H21BrN2O2: C, 64.15; H, 4.71;
N, 6.23. Found: C, 64.22; H, 4.97; N, 6.23.
4.2.5. 2-Propyl-1H-benzo[d]imidazole-6-carboxylic acid (2e)
From 3,4-diaminobenzoic acid (1 g, 6.57 mmol) and trimethyl
orthobutyrate (4.2 mL, 26.3 mmol). Colorless solid, 94%. 1H NMR
of the tautomeric mixture (DMSO-d6): d = 12.68–12.54 (m, 1H),
12.55–12.47 (m, 1H), 8.10–8.00 (m, 1H), 7.78–7.73 (m, 1H), 7.58–
7.46 (m, 1H), 2.85–2.77 (m, 2H), 1.84–1.75 (m, 2H), 0.98–0.92
(m, 3H).
4.2.6. 2-Propyl-1H-benzo[d]imidazole-6-carbonitrile (2f)
From 3,4-diaminobenzonitrile (1 g, 7.5 mmol) and trimethyl
orthobutyrate (4.8 mL, 30 mmol). Colorless solid, 96%. 1H NMR of
the tautomeric mixture (DMSO-d6): d = 12.80–12.71 (m, 1H),
8.07–7.93 (m, 1H), 7.69–7.58 (m, 1H), 7.55–7.48 (m, 1H), 2.86–
2.79 (m, 2H), 1.84–1.75 (m, 2H), 0.92–0.89 (m, 3H).
4.4.3. 40-[(6-Bromo-2-propyl-1H-benzo[d]imidazol-1-yl)methyl]-
biphenyl-2-carboxylic acid (3b–6)
N-Alkylation of 2b (1 g, 2.85 mmol) with 40-(bromo-
methyl)biphenyl-2-carbonitrile (0.844 g, 3.1 mmol). After isomer
separation on silica gel chromatography the resulting 40-[(6-bro-

M. Goebel et al. / Bioorg. Med. Chem. 18 (2010) 5885–5895
5893
mo-2-propyl-1H-benzo[d]imidazol-1-yl)methyl]biphenyl-2-carbo-
nitrile (0.5 g, 1.16 mmol) was treated with KOH (0.325 g,
5.8 mmol) as described for general procedure for saponification
to give a colorless solid, 81%; mp: 258 °C; 1H NMR (DMSO-d6):
d = 12.71 (s, 1H), 7.81 (d, J = 1.8 Hz, 1H), 7.69 (dd, J = 7.8,
J = 1.2 Hz, 1H), 7.55–7.51 (m, 2H), 7.42 (td, J = 7.5, J = 1.3 Hz, 1H),
7.33–7.27 (m, 4H), 7.12–7.02 (m, 2H), 5.54 (s, 2H), 2.79 (t,
J = 7.5 Hz, 2H), 1.75 (tq, J = 7.5 Hz, J = 7.4 Hz, 2H), 0.93 (t,
J = 7.4 Hz, 3H); MS (EI, 200 °C): 448 [M+Å]. Anal. Calcd for
J = 1.3 Hz, 1H), 7.41–7.38 (m, 3H), 7.32 (dd, J = 7.7 Hz, J = 0.9 Hz,
1H), 7.30–7.25 (m, 4H), 7.20–7.16 (m, 2H), 7.11–7.07 (m, 2H),
5.82 (d, J = 4.0 Hz, 1H), 5.78 (d, J = 4.0 Hz, 1H), 5.49 (s, 2H), 2.81
(t, J = 7.5 Hz, 2H), 1.75 (tq, J = 7.5 Hz, J = 7.4 Hz, 2H), 0.94 (t,
J = 7.4 Hz, 3H). MS (EI, 300 °C): 476 [M+Å]. Anal. Calcd for
C31H28N2O3: C, 78.13; H, 5.92; N, 5.88. Found: C, 78.27; H, 6.14;
N, 6.06.
4.4.8. 40-[(5-/6-(Hydroxy(phenyl)methyl)-2-propyl-1H-benzo[d]
imidazol-1-yl)methyl]biphenyl-2-carboxylic acid (3g–6),
isomeric ratio is 4:6
C24H21BrN2O2: C, 64.15; H, 4.71; N, 6.23. Found: C, 64.10; H,
5.03; N, 6.43.
Colorless solid, 5% (4% total); 1H NMR of the isomeric mixture
(DMSO-d6): d = 12.76 (s, 1H), 7.71 (td, J = 7.7 Hz, J = 1.2 Hz, 1H),
7.62–7.23 (m, 11H), 7.21–7.08 (m, 4H), 5.91–5.77 (m, 2H), 5.54–
5.47 (m, 2H), 2.84–2.77 (m, 2H), 1.82–1.71 (m, 2H), 0.98–0.89
(m, 3H). MS (EI, 250 °C): 476 [M+Å]. Anal. Calcd for
C31H28N2O3 ꢁ 1.5H2O: C, 73.94; H, 6.20; N, 5.56. Found: C, 73.61;
H, 6.15; N, 5.68.
4.4.4. 40-[(5,6-Dibromo-2-propyl-1H-benzo[d]imidazol-1-yl)-
methyl]-biphenyl-2-carboxylic acid (3c)
N-Alkylation of 2c (1 g, 3.15 mmol) with 40-(bromo-
methyl)biphenyl-2-carbonitrile (0.944 g, 3.47 mmol). The resulting
40-[(5,6-dibromo-2-propyl-1H-benzo[d]imidazol-1-yl)methyl]-
biphenyl-2-carbonitrile (87%, 1.4 g, 2.75 mmol) was treated with
KOH (0.772 g, 13.75 mmol) as described for general procedure for
saponification to give a colorless solid, 81% (71% total); mp:
318 °C; 1H NMR (DMSO-d6): d = 12.73 (s, 1H), 8.09 (s, 1H), 8.02
(s, 1H), 7.71 (dd, J = 7.7 Hz, 1.3 Hz, 1H), 7.55 (td, J = 7.6 Hz,
J = 1.4 Hz, 1H), 7.44 (td, J=7.6 Hz, J = 1.3 Hz, 1H), 7.34 (dd,
J = 7.7 Hz, J = 1.0 Hz, 1H), 7.31–7.28 (m, 2H), 7.11–7.07 (m, 2H),
5.57 (s, 2H), 2.81 (t, J = 7.5 Hz, 2H), 1.75 (tq, J = 7.5 Hz, J = 7.4 Hz,
2H), 0.93 (t, J = 7.4 Hz, 3H). MS (EI, 250 °C): 526 [M+Å]. Anal. Calcd
for C24H20Br2N2O2 ꢁ 0.25H2O: C, 54.11; H, 3.88; N, 5.26. Found:
C, 54.34; H, 4.01; N, 5.16.
4.4.9. 40-[(5-(Hydroxymethyl)-2-propyl-1H-benzo[d]imidazol-1-
yl)-methyl]biphenyl-2-carboxylic acid (3h–5) and 40-[(6-(hy-
droxymethyl)-2-propyl-1H-benzo[d]imidazol-1-yl)methyl]biphe-
nyl-2-carboxylic acid (3h–6)
To a stirred solution of 2e (1 g, 4.9 mmol) in dry THF (30 mL)
LiAlH4 (0.56 g, 14.7 mmol) was added over 10 min under ice cool-
ing. After stirring for 30 min on ice the reaction mixture was stirred
at room temperature for 2 h and then heated under reflux for an-
other 1 h. H2O was added slowly till no more visible emergence
of hydrogen was observed. After filtration, the solvent was re-
moved under reduced pressure, and the resulting crude product
was purified by chromatography on silica gel with stepwise gradi-
ent elution (CH2Cl2/methanol 98:2, 95:5). The resulting (2-propyl-
1H-benzo[d]imidazol-6-yl)methanol (54%, 0.5 g, 2.63 mmol) was
N-alkylated by general procedure with 40-(bromomethyl)biphe-
nyl-2-carbonitrile (0.79 g, 2.9 mmol). The resulting isomeric mix-
ture of 40-[(5-/6-(hydroxymethyl)-2-propyl-1H-benzo[d]imidazol-
1-yl)methyl]biphenyl-2-carbonitrile (33%, 0.33 g, 0.87 mmol) was
treated with KOH (0.244 g, 4.35 mmol) as described for general
procedure for saponification to give a colorless solid. The 5-/6-
regioisomers were separated by silica gel column chromatography.
Compounds 3h–5: colorless solid, 31% (6% total); mp: 175 °C;
1H NMR (DMSO-d6): d = 12.76 (s, 1H), 7.69 (d, J = 7.2 Hz, 1H),
7.54–7.50 (m, 2H), 7.44–7.40 (m, 2H), 7.34–7.27 (m, 3H), 7.14 (d,
J = 8.3 Hz, 1H), 7.11–7.09 (m, 2H), 5.52 (s, 2H), 5.12–5.09 (m, 1H),
4.57 (d, J = 3.9 Hz, 2H), 2.83 (t, J = 7.4 Hz, 2H), 1.78 (tq,
J = J = 7.5 Hz, J = 7.4 Hz, 2H), 0.96 (t, J = 7.3 Hz, 3H). MS (EI,
200 °C): 400 [M+Å]. Anal. Calcd for C25H28N2O5 ꢁ 2H2O: C, 68.79;
H, 6.47; N, 6.42. Found: C, 68.99; H, 6.36; N, 6.28.
The compounds 3d–5, 3d–6, 3g–5 and 3g–6 were prepared by
general procedure for N-alkylation of 2d (1.05 g, 4 mmol) with
40-(bromomethyl)biphenyl-2-carbonitrile (0.944 g, 3.47 mmol).
The resulting isomeric mixture (90%, 1.6 g, 3.6 mmol) of 40-[(5-
(phenylcarbonyl)-2-propyl-1H-benzo[d]imidazol-1-yl)methyl]-
biphenyl-2-carbonitrile and 40-[(6-(phenylcarbonyl)-2-propyl-1H-
benzo[d]imidazol-1-yl)methyl]biphenyl-2-carbonitrile was treated
with KOH (0.772 g, 13.75 mmol) as described for general proce-
dure for saponification to give all four compounds as a colorless
solid after silica gel column chromatography and recrystallization.
For compounds 3d–6 and 3g–6 the isomeric ratio of the 5- to the
6-regioisomer is 4:6.
4.4.5. 40-[(5-(Phenylcarbonyl)-2-propyl-1H-benzo[d]imidazol-1-
yl)methyl]biphenyl-2-carboxylic acid (3d–5)
Colorless solid, 8% (7% total); mp: 216 °C; 1H NMR (DMSO-d6):
d = 12.74 (s, 1H), 7.95 (s, 1H), 7.77–7.65 (m, 6H), 7.59–7.52 (m,
3H), 7.44 (td, J = 7.6 Hz, 1.2 Hz, 1H), 7.36–7.30 (m, 3H), 7.17–7.12
(m, 2H), 5.62 (s, 2H), 2.88 (t, J = 7.5 Hz, 2H), 1.79 (tq, J = 7.5 Hz,
J = 7.4 Hz, 2H), 0.97 (t, J = 7.3 Hz, 3H). MS (EI, 200 °C): 474 [M+Å].
Anal. Calcd for C31H26N2O3 ꢁ 0.125H2O: C, 78.09; H, 5.55; N,
5.88. Found: C, 78.07; H, 5.78; N, 5.85.
Compounds 3h–6: colorless solid, 22% (4% total); mp: 200 °C;
1H NMR (DMSO-d6): d = 12.73 (s, 1H), 7.69 (d, J = 7.2 Hz, 1H),
7.56–7.51 (m, 2H), 7.46–7.40 (m, 2H), 7.33 (d, J = 8.3 Hz, 1H),
7.31–7.26 (m, 2H), 7.13 (dd, J = 8.3, J = 1.5 Hz, 1H), 7.11–7.07 (m,
2H), 5.52 (s, 2H), 5.16 (t, J = 5.1 Hz, 1H), 4.57 (d, J = 4.8 Hz, 2H),
2.81 (t, J = 7.5 Hz, 2H), 1.77 (tq, J = 7.5 Hz, J = 7.4 Hz, 2H), 0.95 (t,
J = 7.3 Hz, 3H). MS (EI, 225 °C): 400 [M+Å]. Anal. Calcd for
4.4.6. 40-[(5-/6-(Phenylcarbonyl)-2-propyl-1H-benzo[d]imidazol-
1-yl)methyl]biphenyl-2-carboxylic acid (3d–6), isomeric ratio is
4:6
Colorless solid, 5% (4% total); 1H NMR of the isomeric mixture
(DMSO-d6): d = 12.49 (s, 1H), 7.97–7.93 (m, 1H), 7.77–7.49 (m, 9H),
7.46–7.41 (m, 1H), 7.35–7.31 (m, 3H), 7.18–7.11 (m, 2H), 5.62 (s,
2H), 2.96–2.87 (m, 2H), 1.86–1.77 (m, 2H), 1.01–0.95 (m, 3H). MS
(EI, 200 °C): 474 [M+Å]. Anal. Calcd for C31H26N2O3 ꢁ 0.25H2O: C,
77.72; H, 5.58; N, 5.85. Found: C, 77.91; H, 5.69; N, 5.73.
C
25H28N2O5 ꢁ 0.65H2O: C, 72.85; H, 6.19; N, 6.80. Found: C,
72.86; H, 6.41; N, 7.13.
4.4.10. 40-[(5-/6-Formyl-2-propyl-1H-benzo[d]imidazol-1-
yl)methyl]biphenyl-2-carboxylic acid (3i), isomeric ratio 6:4
To a solution of 2f (1 g, 5.4 mmol) in 60 mL formic acid and
20 mL of H2O, Ni–Al alloy (4 g) was added in several portions.
The reaction mixture was heated at 95 °C for 4 h. The hot mixture
was filtered with the aid of Celite and washed thrice with water.
The solution was concentrated and the pH was adjusted to 9 by
the dropwise addition of 2 N NaOH. The product was obtained by
4.4.7. 40-[(5-(Hydroxy(phenyl)methyl)-2-propyl-1H-benzo[d]im-
idazol-1-yl)methyl]biphenyl-2-carboxylic acid (3g–5)
Colorless solid, 9% (7% total); mp: 208 °C; 1H NMR (DMSO-d6):
d = 12.72 (s, 1H), 7.69 (dd, J = 7.7 Hz, J = 1.2 Hz, 1H), 7.57 (d,
J = 1.4 Hz, 1H), 7.53 (td, J = 7.6 Hz, 1.4 Hz, 1H), 7.43 (dd, J = 7.6 Hz,

5894
M. Goebel et al. / Bioorg. Med. Chem. 18 (2010) 5885–5895
extraction with CH2Cl2 (3 ꢁ 50 mL). The organic layers were com-
bined, dried over Na2SO4, concentrated in vacuo and purified by
silica gel column chromatography to give a colorless solid. The
resulting 2-propyl-1H-benzo[d]imidazole-6-carbaldehyde (35%,
0.36 g, 1.89 mmol) was N-alkylated by general procedure with 40-
(bromomethyl)biphenyl-2-carbonitrile (0.57 g, 2.08 mmol). The
resulting isomeric mixture of 40-[(5-/6-formyl-2-propyl-1H-
benzo[d]imidazol-1-yl)methyl]biphenyl-2-carbonitrile (84%, 0.6 g,
1.58 mmol) was treated with KOH (0.44 g, 7.9 mmol) as described
for general procedure for saponification to give a colorless solid,
71% (21% total); 1H NMR of the isomeric mixture (DMSO-d6):
d = 12.73 (s, 1H), 10.05–10.01 (m, 1H), 8.20–8.15 (m, 1H), 7.80–
7.69 (m, 3H), 7.55 (td, J = 7.6 Hz, J = 1.3 Hz, 1H), 7.44 (td,
J = 7.6 Hz, J = 1.1 Hz, 1H), 7.35–7.27 (m, 3H), 7.21–7.09 (m, 2H),
5.68–5.61 (m, 2H), 2.92–2.81 (m, 2H), 1.85–1.75 (m, 2H), 0.99–
0.94 (m, 3H). MS (EI, 200 °C): 398 [M+Å]. Anal. Calcd for
C25H28N2O5 ꢁ 0.75H2O: C, 72.89; H, 5.75; N, 6.80. Found: C,
72.94; H, 5.50; N, 6.71.
COS-7 cells (8 ꢁ 105/well in a 96-well plate) were seeded the day
before and transfected for each well with the use of 0.25
l
L Lipo-
DEF, 45 ng
L Opti-MEM (Gibco).37
fectamine 2000 (Invitrogen) with 4.5 ng pGal4-hPPAR
c
pGal5-TK-pGL3 and 3 ng pRL-CMV in 25
l
After 4 h, a sixth part of the transfection medium volume of DMEM
(+10% FCS) plus the compounds or vehicle (DMSO) was added;
luciferase activity was measured after 36 h.
Acknowledgements
This work was supported by Grants Gu 285/7–1/2 and Ki 712/
3–1/2 from the Deutsche Forschungsgemeinschaft (DFG).
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2010.06.102.
References and notes
4.4.11. 40-[(2-Propyl-1H-naphth[2,3-d]imidazol-1-yl)methyl]-
biphenyl-2-carboxylic acid (5)
1. Michalik, L.; Auwerx, J.; Berger, J. P.; Chatterjee, V. K.; Glass, C. K.; Gonzalez, F.
J.; Grimaldi, P. A.; Kadowaki, T.; Lazar, M. A.; O0Rahilly, S.; Palmer, C. N.;
Plutzky, J.; Reddy, J. K.; Spiegelman, B. M.; Staels, B.; Wahli, W. Pharmacol. Rev.
2006, 58, 726.
2. Spiegelman, B. M. Diabetes 1998, 47, 507.
3. Willson, T. M.; Brown, P. J.; Sternbach, D. D.; Henke, B. R. J. Med. Chem. 2000, 43,
527.
4. Berger, J.; Moller, D. E. Annu. Rev. Med. 2002, 53, 409.
5. Mangelsdorf, D. J.; Thummel, C.; Beato, M.; Herrlich, P.; Schütz, G.; Umesono,
K.; Blumberg, B.; Kastner, P.; Mark, M.; Chambon, P.; Evans, R. M. Cell 1995, 83,
835.
From 2,3-diaminonaphthalene (0.5 g, 3.2 mmol) and trimethyl
orthobutyrate (2 mL, 12.8 mmol) as described for general procedure
for the preparation of 2-substituted 1H-benzimidazoles. The result-
ing 2-propyl-1H-naphth[2,3-d]imidazol (94%, 0.63 g, 3 mmol) was
N-alkylated by general procedure with 40-(bromomethyl)biphenyl-
2-carbonitrile (0.898 g, 3.3 mmol). The resulting 40-[(2-propyl-1H-
naphth[2,3-d]imidazol-1-yl)methyl]biphenyl-2-carbonitrile (88%,
1 g, 2.6 mmol) was treated with KOH (0.73 g, 13 mmol) as described
for general procedure for saponification to give a colorless solid, 91%
(74% total); mp: 271 °C; 1H NMR (DMSO-d6): d = 12.71 (s, 1H), 8.15
(s, 1H), 8.04–7.97 (m, 2H), 7.96–7.92 (m, 1H), 7.70 (dd, J = 7.6 Hz,
J = 1.1 Hz, 1H), 7.53 (td, J = 7.5 Hz, J = 1.2 Hz, 1H), 7.43 (td,
J = 7.6 Hz, J = 1.1 Hz, 1H), 7.40–7.27 (m, 5H), 7.17 (d, J = 8.2 Hz, 2H),
5.64 (s, 2H), 2.92 (t, J = 7.5 Hz, 2H), 1.85 (tq, J = 7.5 Hz, J = 7.4 Hz,
2H), 1.00 (t, J = 7.4 Hz, 3H). MS (EI, 230 °C): 420 [M+Å]. Anal. Calcd
for C28H24N2O2: C, 79.98; H, 5.75; N, 6.66. Found: C, 80.14; H, 5.99;
N, 6.62.
6. Committee, N. R. N. Cell 1999, 97, 161.
7. Willson, T. M.; Wahli, W. Curr. Opin. Chem. Biol. 1997, 1, 235.
8. Berger, J. P.; Petro, A. E.; Macnaul, K. L.; Kelly, L. J.; Zhang, B. B.; Richards, K.;
Elbrecht, A.; Johnson, B. A.; Zhou, G.; Doebber, T. W.; Biswas, C.; Parikh, M.;
Sharma, N.; Tanen, M. R.; Thompson, G. M.; Ventre, J.; Adams, A. D.; Mosley, R.;
Surwit, R. S.; Moller, D. E. Mol. Endocrinol. 2003, 17, 662.
9. Schupp, M.; Janke, J.; Clasen, R.; Unger, T.; Kintscher, U. Circulation 2004, 109,
2054.
10. Benson, S. C.; Pershadsingh, H. A.; Ho, C. I.; Chittiboyina, A.; Desai, P.; Pravenec,
M.; Qi, N.; Wang, J.; Avery, M. A.; Kurtz, T. W. Hypertension 2004, 43, 993.
11. Schupp, M.; Clemenz, M.; Gineste, R.; Witt, H.; Janke, J.; Helleboid, S.;
Hennuyer, N.; Ruiz, P.; Unger, T.; Staels, B.; Kintscher, U. Diabetes 2005, 54,
3442.
12. Goebel, M.; Clemenz, M.; Staels, B.; Unger, T.; Kintscher, U.; Gust, R.
ChemMedChem 2009, 4, 445.
5. Biology
13. Goebel, M.; Staels, B.; Unger, T.; Kintscher, U.; Gust, R. ChemMedChem 2009, 4,
1136.
14. Schupp, M.; Lee, L. D.; Frost, N.; Umbreen, S.; Schmidt, B.; Unger, T.; Kintscher,
U. Hypertension 2006, 47, 586.
Telmisartan and Pioglitazone were obtained from the pharmacy
extracted with CHCl3 and purified by chromatography on silica gel
and recrystallized from methanol.
15. Schmidt, B.; Schieffer, B. J. Med. Chem. 2003, 46, 2261.
16. Tawar, U.; Jain, A. K.; Dwarakanath, B. S.; Chandra, R.; Singh, Y.; Chaudhury, N.
K.; Khaitan, D.; Tandon, V. J. Med. Chem. 2003, 46, 3785.
17. Narr, B.; Bomhard, A. D.; Hauel, N. D.; Van, D.; Jacques, M. EP Patent 0,392,317,
1990.
5.1. Differentiation assay
18. Markt, P.; Schuster, D.; Kirchmair, J.; Laggner, C.; Langer, T. J. Comput. Aided Mol.
Des. 2007, 21, 575.
Murine 3T3-L1 preadipocytes were cultured in DMEM (+10%
FCS) and differentiated by a modified previously described proto-
col:9 after two days postconfluent preadipocytes were treated for
19. Cronet, P.; Petersen, J. F. W.; Folmer, R.; Blomberg, N.; Sjoblom, K.; Karlsson, U.;
Lindstedt, E. L.; Bamberg, K. Structure 2001, 9, 699.
20. Xu, H. E.; Lambert, M. H.; Montana, V. G.; Parks, D. J.; Blanchard, S. G.; Brown, P.
J.; Sternbach, D. D.; Lehmann, J. M.; Wisely, G. B.; Willson, T. M.; Kliewer, S. A.;
Milburn, M. V. Mol. Cell. 1999, 3, 397.
21. Chittiboyina, A. G.; Mizuno, C. S.; Desai, P. V.; Patny, A.; Kurtz, T. W.;
Pershadsingh, H. A.; Speth, R. C.; Karamyan, V.; Avery, M. A. Med. Chem. Res.
2009, 18, 589.
three days with complete medium containing 1
sone and 0.17 M insulin. After medium replacement they were
incubated for further three days with insulin (0.17 M), and for
lM Dexametha-
l
l
the last three days with only complete medium. Cells were treated
the whole time period with either vehicle (DMSO) as negative con-
trol, Pioglitazone and Telmisartan as positive controls or the syn-
thesized compounds until day 9 of differentiation, after which
the cells were washed with phosphate-buffered saline and stained
with Oil Red O.9 For quantification, the dye was extracted with iso-
propanol (80% V/V) and the absorption was measured at 515 nm.
22. Mizuno, C. S.; Chittiboyina, A. G.; Patny, A.; Kurtz, T. W.; Pershadsingh, H. A.;
Speth, R. C.; Karamyan, V. T.; Avery, M. A. Med. Chem. Res. 2009, 18, 611.
23. Markt, P.; Petersen, R. K.; Flindt, E. N.; Krjstiansen, K.; Kirchmair, J.; Spitzer, G.;
Distinto, S.; Schuster, D.; Wolber, G.; Laggner, C.; Langer, T. J. Med. Chem. 2008,
51, 6303.
24. Xu, X. Y.; Cheng, F.; Shen, J. H.; Luo, X. M.; Chen, L. L.; Yue, L. D.; Du, Y.; Ye, F.;
Jiang, S. H.; Zhu, D. Y.; Jiang, H. L.; Chen, K. X. Int. J. Quantum Chem. 2003, 93,
405.
25. Bharatam, P. V.; Patel, D. S.; Adane, L.; Mittal, A.; Sundriyal, S. Curr. Pharm. Des.
2007, 13, 3518.
26. Fracchiolla, G.; Laghezza, A.; Piemontese, L.; Carbonara, G.; Lavecchia, A.;
Tortorella, P.; Crestani, M.; Novellino, E.; Loiodice, F. ChemMedChem 2007, 2,
641.
27. Humphrey, W.; Dalke, A.; Schulten, K. J. Mol. Graphics 1996, 14. 333-8, 27-8.
28. CCDC Gold Suite, Cambridge, UK, 2008.
5.2. PPARc transactivation assay
Transient transfection (Invitrogen) and luciferase assays (Pro-
mega) were performed as described in the manufacture protocol.

M. Goebel et al. / Bioorg. Med. Chem. 18 (2010) 5885–5895
5895
29. Konagurthu, A. S.; Whisstock, J. C.; Stuckey, P. J.; Lesk, A. M. Proteins 2006, 64,
559.
35. Mahindroo, N.; Huang, C. F.; Peng, Y. H.; Wang, C. C.; Liao, C. C.; Lien, T. W.;
Chittimalla, S. K.; Huang, W. J.; Chai, C. H.; Prakash, E.; Chen, C. P.; Hsu, T. A.;
Peng, C. H.; Lu, I. L.; Lee, L. H.; Chang, Y. W.; Chen, W. C.; Chou, Y. C.; Chen, C. T.;
Goparaju, C. M.; Chen, Y. S.; Lan, S. J.; Yu, M. C.; Chen, X.; Chao, Y. S.; Wu, S. Y.;
Hsieh, H. P. J. Med. Chem. 2005, 48, 8194.
36. Leesnitzer, L. M.; Parks, D. J.; Bledsoe, R. K.; Cobb, J. E.; Collins, J. L.; Consler, T.
G.; Davis, R. G.; Hull-Ryde, E. A.; Lenhard, J. M.; Patel, L.; Plunket, K. D.; Shenk, J.
L.; Stimmel, J. B.; Therapontos, C.; Willson, T. M.; Blanchard, S. G. Biochemistry
2002, 41, 6640.
30. Warren, G. L.; Andrews, C. W.; Capelli, A.-M.; Clarke, B.; LaLonde, J.; Lambert,
M. H.; Lindvall, M.; Nevins, N.; Semus, S. F.; Senger, S.; Tedesco, G.; Wall, I. D.;
Woolven, J. M.; Peishoff, C. E.; Head, M. S. J. Med. Chem. 2005, 49, 5912.
31. Li, Y.; Zhang, J.; Schopfer, F. J.; Martynowski, D.; Garcia-Barrio, M. T.; Kovach,
A.; Suino-Powell, K.; Baker, P. R. S.; Freeman, B. A.; Chen, Y. E.; Xu, H. E. Nat.
Struct. Mol. Biol. 2008, 15, 865.
32. Wolber, G.; Dornhofer, A. A.; Langer, T. J. Comput. Aided Mol. Des. 2006, 20, 773.
33. Wolber, G.; Langer, T. J. Chem. Inf. Model. 2005, 45, 160.
34. Santini, C.; Berger, G. D.; Han, W.; Mosley, R.; MacNaul, K.; Berger, J.; Doebber,
T.; Wu, M.; Moller, D. E.; Tolman, R. L.; Sahoo, S. P. Bioorg. Med. Chem. Lett. 2003,
13, 1277.
37. Raspé, E.; Madsen, L.; Lefebvre, A. M.; Leitersdorf, I.; Gelman, L.; Peinado-
Onsurbe, J.; Dallongeville, J.; Fruchart, J. C.; Berge, R.; Staels, B. J. Lipid Res. 1999,
40, 2099.