Molecular determinants of ligand binding modes in the histamine H 4 receptor: Linking ligand-based three-dimensional quantitative structure-activity relationship (3D-QSAR) models to in silico guided receptor mutagenesis studies

Article abstract of DOI:10.1021/jm201042n

The histamine H₄ receptor (H₄R) is a G protein-coupled receptor (GPCR) that plays an important role in inflammation. Similar to the homologous histamine H₃ receptor (H₃R), two acidic residues in the H₄

Full text of DOI:10.1021/jm201042n

Article
pubs.acs.org/jmc
Molecular Determinants of Ligand Binding Modes in the Histamine
H4 Receptor: Linking Ligand-Based Three-Dimensional Quantitative
Structure−Activity Relationship (3D-QSAR) Models to in Silico
Guided Receptor Mutagenesis Studies
Enade P. Istyastono,†,§ Saskia Nijmeijer,†,§ Herman D. Lim,‡ Andrea van de Stolpe,† Luc Roumen,†
Albert J. Kooistra,† Henry F. Vischer,† Iwan J. P. de Esch,† Rob Leurs,† and Chris de Graaf*,†
†Department of Pharmacochemistry, Faculty of Exact Sciences, Leiden/Amsterdam Center for Drug Research (LACDR), Division of
Medicinal Chemistry, VU University Amsterdam, De Boelelaan 1083, 1081 HV Amsterdam, The Netherlands
‡Department of Medicinal Chemistry, Room P-246, Griffin Discoveries BV, De Boelelaan 1083, 1081 HV, Amsterdam, The
Netherlands
S
* Supporting Information
ABSTRACT: The histamine H4 receptor (H4R) is a G protein-
coupled receptor (GPCR) that plays an important role in
inflammation. Similar to the homologous histamine H3 receptor
(H3R), two acidic residues in the H4R binding pocket, D3.32 and
E5.46, act as essential hydrogen bond acceptors of positively
ionizable hydrogen bond donors in H4R ligands. Given the
symmetric distribution of these complementary pharmacophore
features in H4R and its ligands, different alternative ligand binding
mode hypotheses have been proposed. The current study focuses
on the elucidation of the molecular determinants of H4R−ligand binding modes by combining (3D) quantitative structure−activity
relationship (QSAR), protein homology modeling, molecular dynamics simulations, and site-directed mutagenesis studies. We have
designed and synthesized a series of clobenpropit (N-(4-chlorobenzyl)-S-[3-(4(5)-imidazolyl)propyl]isothiourea) derivatives to investigate
H4R−ligand interactions and ligand binding orientations. Interestingly, our studies indicate that clobenpropit (2) itself can bind to H4R in
two distinct binding modes, while the addition of a cyclohexyl group to the clobenpropit isothiourea moiety allows VUF5228 (5) to adopt
only one specific binding mode in the H4R binding pocket. Our ligand-steered, experimentally supported protein modeling method gives
new insights into ligand recognition by H4R and can be used as a general approach to elucidate the structure of protein−ligand complexes.
previously demonstrated for 1,12−14 2,15,16 4a (immepip),13
INTRODUCTION
■
4b (iodoproxyfan),16 and 5 (VUF5228)15 (Figure 1).
The discovery of the human histamine H4 receptor (H4R) in
In general, two classes of binding modes have been proposed: one
2000 has attracted attention from pharmaceutical industry and
where the imidazole moiety binds to E5.46 (mode I)12,13,15 and
academia because the receptor plays an important role in allergic
another where the imidazole moiety binds to D3.32 (mode
and inflammatory processes.1−5 Together with the histamine H1
II).13,14,16−18 Although several GPCR crystal structures have been
(H1R), H2 (H2R), and H3 (H3R), the H4R belongs to the
G protein-coupled histamine receptor family.2,6−9 The H4R shows
a sequence identity at the protein level of 23%, 22%, and 31%
solved in the past 3 years, including the related bioaminergic β1
adrenergic (ADRB1), β2 adrenergic (ADRB2), dopamine D3
(DRD3), and histamine H1 (H1R) receptors,19−22 experimental
structures of H3R or H4R are still lacking. Only by combination of
experimental information on both the ligand (e.g., structure−activity
relationships) and the protein target (e.g., site-directed mutagenesis
studies), reliable structural models of protein−ligand complexes can
be constructed.23 In the current study we present clobenpropit and
its analogues, dual active H3R/H4R ligands24 containing two basic
groups, as useful tools to identify and recognize the chemical
properties and features that determine the binding mode of ligands
in the H4R binding pocket. We have previously described
with H1R, H2R, and H3R, respectively.1,6 Moreover, the putative
histamine (1) binding pockets of H3R and H4R share a
sequence identity of 56% and sequence similarity of ∼80%,
explaining why the ligand pharmacophores for both receptors
are very similar.10 Some of the first H4R ligands were previously
known as selective H3R ligands, e.g., 2 (clobenpropit) and 3
(imetit) (Figure 1).1,2,6−9 Both H3R and H4R ligands typically
have two hydrogen bond donors,1,10 complementary with two
negatively residues in the putative binding pocket of H3R and
H4R, D3.32 and E5.46 (Ballesteros−Weinstein numbering).1,11,12
These symmetric distributions of complementary pharmaco-
phore features in the ligands and cavities of H3R and H4R make
the definition of a ligand binding mode challenging, as
Received: August 3, 2011
Published: October 17, 2011
© 2011 American Chemical Society
8136
dx.doi.org/10.1021/jm201042n|J. Med. Chem. 2011, 54, 8136−8147

Journal of Medicinal Chemistry
Article
Figure 1. Structures of H3R and H4R ligands 1−5. Biological activities are from refs 1 and 6 (1, 3, 4) and the current study (2, 5).
quantitative structure−activity relationship (QSAR) studies on
clobenpropit analogues that indicated that the stretch−bending
energy (which depends on the conformation of the compounds) is
an important property that explains the affinity of H4R ligands.24
This property also plays an important role in H4R/H3R selectivity. In
order to better understand the importance of this descriptor, a new
series of clobenpropit derivatives were synthesized and evaluated for
their H3R and H4R affinity. Addition of a cyclohexyl moiety to
compound 2 results in compound 5, which has higher affinity for
H3R and much lower affinity for H4R than 2 (Figure 1), resulting in
a 16-fold increase in H3R over H4R selectivity. This additional
lipophilic moiety is proposed to probe an additional lipophilic pocket
in both H3R15,25 and H4R.1 By combining classical QSAR,24
molecular interaction fields (MIF) based 3D-QSAR approaches,26
and in silico guided site-directed mutagenesis (SDM) studies,12,27,28
we were able to identify determinants of H3R/H4R selectivity and
determine ligand binding modes in the H4R binding pocket.
Figure 2. Flowchart of the approaches used in the current study. See
Methods for more details. (a) SK-N-MC cells stably expressing H3R or
H4R were used to determine ligand affinities for either the H3R or the
H4R. The data were subjected to the classical QSAR method using MOE
descriptors described previously24 and (b) GRID MIF-based 3D-
QSAR.26,53 (c) PLANTS docking against H3R and H4R.18,28 (d)
Molecular dynamics simulations with AMBER.38 (e) Site-directed
mutagenesis (SDM) studies were performed as described previously.12,28
RESULTS AND DISCUSSION
Addition of a Lipophilic Moiety Increases Selectivity
toward H3R. Compounds 2, 5, and 6a−l were investigated for
their H3R and H4R affinity (Table 1) and were subsequently
subjected to QSAR studies. Addition of a cyclohexyl moiety to
compound 2, resulting in compound 5, increases the H3R
affinity but decreases the H4R affinity. Further investigation
shows that smaller lipophilic substituents (compare 6a to 6b)
decrease the H3R affinity and slightly increase the H4R affinity.
Remarkably, a cyclopentyl substitution at R1 results in a
compound (6a) with similar affinity for both the H3R and the
H4R (ΔpKi = 0). Among the series, only compounds 6c and 6i
have a slightly favorable affinity toward the H4R (ΔpKi = −0.4).
These compounds are among the smallest compounds in the
series. On the other hand, compound 6d, one of the largest
compounds, shows the highest H3R selectivity over H4R (ΔpKi =
2.16). This is in line with earlier (Q)SAR, indicating that steric
properties play an important role in driving H3R over H4R
selectivity.13,24
■
The main aim of our study was to probe the important
molecular features and residues that determine the binding
orientation of ligands in the H4R pocket. For this purpose,
classical and three-dimensional (3D) quantitative structure−
activity relationship (QSAR), protein homology modeling,
molecular dynamics simulations, and site-directed mutagenesis
(SDM) studies were combined in an integrative workflow
(Figure 2) to investigate the binding modes of clobenpropit (2)
and (N-(4-chlorobenzyl)-S-[3-(4(5)-imidazolyl)propyl]-
isothiourea) clobenpropit analogues (5 and 6).
Design and Synthesis of Novel Clobenpropit (N-(4-
Chlorobenzyl)-S-[3-(4(5)-imidazolyl)propyl]isothiourea)
Derivatives. To probe the distinct hydrophobic pockets of
the receptors, a series of compounds (6a−l) was designed and
synthesized (Scheme 1). Intermediate 4-(3-bromo-propyl)-1H-
imidazole hydrobromide (7), which was synthesized according
to literature procedure,29,30 was treated with thiourea
compounds (8a−l). The corresponding thiourea compounds
were prepared from commercially available isothiocyanates and
the corresponding amines.
Steric QSAR Descriptors Determine the Activity of the
Ligands to the H3R and the H4R. QSAR models to describe
the binding of the ligands to both the H3R (eq 1) and the H4R
8137
dx.doi.org/10.1021/jm201042n|J. Med. Chem. 2011, 54, 8136−8147

Journal of Medicinal Chemistry
Article
a
Scheme 1
a
(a) Et2O, room temp, 1 h; (b) ethanol, microwave, 120 °C, 0.5 h.
Table 1. Affinity (pKi) of Clobenpropit Analogues at the Human H3R and the Human H4R
a
a
b
compd
R1
R2
pKi(H3R)
pKi(H4R)
ΔpKi
2
H
4-chlorobenzyl
4-chlorobenzyl
benzyl
8.55 ± 0.08
8.98 ± 0.09
7.37 ± 0.02
8.01 ± 0.05
7.33 ± 0.07
8.99 ± 0.11
8.31 ± 0.08
7.77 ± 0.04
7.49 ± 0.13
7.73 ± 0.10
7.17 ± 0.09
8.33 ± 0.10
8.79 ± 0.10
8.94 ± 0.10
8.00 ± 0.06
7.32 ± 0.14
7.37 ± 0.02
7.12 ± 0.03
7.79 ± 0.01
6.83 ± 0.10
7.05 ± 0.01
6.64 ± 0.09
6.42 ± 0.10
7.50 ± 0.04
7.53 ± 0.08
7.05 ± 0.09
6.90 ± 0.07
6.87 ± 0.05
0.55
1.66
0.00
0.89
−0.46
2.16
1.26
1.13
1.07
0.23
−0.36
1.28
1.89
1.88
5
cyclohexyl
6a
6b
6c
6d
6e
6f
6g
6h
6i
cyclopentyl
cyclohexyl
benzyl
cyclopentyl
n-butyl
3,4-dichlorobenzyl
4-chlorobenzyl
4-chlorobenzyl
3,4-dichlorobenzyl
cyclopentyl
cyclohexyl
n-butyl
benzyl
benzyl
cyclohexyl
tert-butyl
n-butyl
cyclopentyl
6j
3,4-dichlorobenzyl
4-cyanobenzyl
4-nitrobenzyl
6k
6l
cyclohexyl
cyclohexyl
a
SK-N-MC cells stably expressing H3R or H4R were used to determine ligand affinities for either the H3R or the H4R. Data shown are the mean ±
b
SEM of at least three independent experiments. ΔpKi = pKi(H3R) − pKi(H4R).
(eq 2) were generated, using a similar approach as reported
previously.24 Equation 1 shows that the ligand affinity to the
H3R is determined by a steric descriptor (Diameter) and a
combined steric−electronic descriptor (BCUT_SMR_0),31,32
while eq 2 shows that ligand affinity at H4R is determined by
solely steric descriptors (vdw_area and E_strain). Interestingly,
similar to the previous QSAR studies,24,32 an energy-related
descriptor (E_strain) emerges as the important descriptor
describing the ligand affinity to the H4R. Both descriptors
(E_stb in the previous QSAR studies and E_strain in this
study) are conformation-related descriptors, and E_stb was
proposed as the selectivity determining descriptor.24,32 This
indicates that clobenpropit analogues bind to the H4R binding
pocket in an energetically less favorable conformation than in
the H3R binding pocket.33
(2)
Linking Steric Descriptors to 3D-QSAR Hydrophobic
Probes. In order to correlate the information regarding
important global ligand features (step 1 in Figure 2) to ligand−
receptor binding modes in three-dimensional space, 3D-QSAR
models (step 2 in Figure 2) were constructed correlating molecular
interaction fingerprints (MIFs)26 with H3R (eq 3) and H4R (eq 4)
ligand binding affinities. The structure of 5 was generated using
CORINA, resulting in two 3D configurations (Figure 3).34 Ab
(1)
8138
dx.doi.org/10.1021/jm201042n|J. Med. Chem. 2011, 54, 8136−8147

Journal of Medicinal Chemistry
Article
in the H4R 3D-QSAR model (eq 4, Figure 4B), on the other hand,
matches the steric descriptors (vdw_area and E_strain) in the
classical H4R QSAR model (eq 2).
(3)
(4)
Figure 3. Depiction of two configurations of compound 5 generated
with CORINA.
initio (DFT/6-31G*) optimizations with Gaussian were per-
formed for both 3D configurations.35−37 The lower energy
configuration suggested by the ab initio calculations (Figure 3B)
was selected as the reference structure to align 2 and 6a−l with
ROCS to construct MIF-based 3D-QSAR models (Figure 4).38,39
Interestingly, a hydrophobic probe (DRY.1) and a polar probe
(O.1) are descriptors in the H3R 3D-QSAR model (eq 3, Figure
4A), which correspond to the steric (Diameter) and steric−
electronic (BCUT_SMR_0) descriptors in the classical H3R
QSAR model (eq 1). The single hydrophobic descriptor (DRY.2)
Since the (3D-)QSAR models (eqs 1−4) models indicated an
important role of steric/hydrophobic descriptors in determining
H3R and H4R affinity, a 3D-QSAR model predicting the
difference between H4R and H3R affinity (ΔpKi) was
constructed. The result (eq 5) reveals two hydrophobic probes
that determine H3R over H4R selectivity, one that is within 3.8 Å
from the hydrophobic centroid of the cyclohexyl moiety (DRY.4)
of compound 6 (Figure 4C). Another selectivity determining
probe is located within 3.9 Å from the hydrophobic centroid of
Figure 4. Structure of compound 5 in 3D with important probes explaining the ligand affinity to H3R (eq 3, A), important probes explaining the
ligand affinity to H4R (eq 4, B), and important probes explaining the H3R/H4R selectivity (eq 5, C). The distances between hydrophobic probes to
the nearest hydrophobic centroid of compound 6 (presented as black spheres) are presented in parts A−C. Graph between observed and calculated
ΔpKi is presented in part D.
8139
dx.doi.org/10.1021/jm201042n|J. Med. Chem. 2011, 54, 8136−8147

Journal of Medicinal Chemistry
Article
the aromatic moiety (DRY.3) of compound 6 (Figure 4C).
and E7.39 via its imidazole and isothiurea groups, respectively.43,44
The recent GPCR DOCK 2010 challenge demonstrated that this
symmetric distribution of complementary pharmacophore
features in ligand and protein made the prospective prediction
of CXCR4−ligand interactions highly challenging.45,46 In line
with the same philosophy we followed in our successful
CXCR4−ligand binding pose prediction in this community
wide GPCR modeling assessment,44 we investigated two putative
H4R binding mode hypotheses: one where the imidazole moiety
binds to E5.46 (mode I, i.e., models in Figure 5A and Figure 5C)12,13,15
(5)
The descriptors used in the (3D-)QSAR models (eqs 1−5)
are described in Table 2. All models have acceptable statistical
Table 2. Definition of the Descriptors in the QSAR Models
(Eqs 1−5)
a
descriptor
definition
Diameter
largest value in the distance matrix
BCUT_SMR_0
vdw_area
the BCUT descriptors using atomic contribution to
molar refractivity31,32,61
area of van der Waals surface calculated using a
connection table approximation
E_strain
DRY.1
local strain energy
interaction energy value of hydrophobic probe
DRY.8331 (4.0, −1.0, 2.0)
O.1
interaction energy value of sp2 carbonyl oxygen probe
O.6874 (7.0, −6.0, 1.0)
DRY.2
DRY.3
DRY.4
interaction energy value of hydrophobic probe
DRY.8195 (−2.0, −5.0, 2.0)
interaction energy value of hydrophobic probe
DRY.3376 (9.,; −6.0. −8.0)
interaction energy value of hydrophobic probe
DRY.9361 (5.0, −5.0, 4.0)
a
Descriptors diameter, BCUR_SMR_0, vdw_area, and E_strain gener-
ated with the QuaSAR descriptor module in MOE.24,32 Interaction energy
values of GRID26,53 interaction points DRY.8331 (DRY.1), O.6874 (O.1),
DRY.8195 (DRY.2), DRY.3376 (DRY.3), and DRY.9361 (DRY.4) at speci-
fic positions (Cartesian coordinates from the midpoint of compound 6).
Figure 5. Possible binding modes of compounds 2 (A, B) and 5
(C, D) in the H4R 3D homology models.1 Compound 2 is depicted as
ball-and-sticks with orange carbon atoms, while compound 5 is
depicted with magenta carbon atoms. The backbones of TM helices 4,
5, 6, and 7 are represented by yellow ribbons, and part of TM3 is
shown as ribbon (the top of the helix is not shown for clarity).
Important binding residues are depicted as ball-and-sticks with gray
carbon atoms. Residues studied by side-directed mutagenesis are
labeled (note that D3.32 has not been mutated in the current study but
has previously been shown to be an essential residue for H4R ligand
binding1,11,12 and is labeled in gray). If the Ki of the point mutant for
the specific ligand is decreased by 5-fold or more (compared to the Ki
of wild type H4R), the residue is labeled red.
significances (N = number of samples, F = test for quality of fit,
r = coefficient of correlation, R2 = coefficient of determination, S
= standard error of estimation, and q2 = leave-one-out cross-
validation).24,40 Moreover, a total of 1000 runs of Y-random-
ization was performed for each model to show the distribution of
the dependent variable values. The results show that the internal
predictivity (q2) of each model was better than any of the 1000
scrambled models. We are therefore confident that the number
of examples is sufficient and heterogeneous.24,41
Linking QSAR Studies to 3D Homology Models. In
parallel with the synthesis, pharmacological evaluation, and the
QSAR modeling of clobenpropit analogues as dual active H3R
and H4R ligands, ADRB2-based19 homology models of H4R
(step 3 in Figure 2) were generated to identify the selectivity
determining residues. These residues were subsequently
investigated by site-directed mutagenesis studies to elucidate
the binding modes of compounds 2 and 5 in the H4R binding
pocket. Compounds 2 and 5 have two moieties, imidazole and
isothiourea moieties, that can act as hydrogen bond donors to
interact with the receptor.1,15,24,25 The H4R binding pocket has
two residues, D3.32 and E5.46, that have been shown to be essential
for ligand binding1,12,28,42 and are proposed to act as
complementary negatively ionizable hydrogen bond acceptors.
A similar phenomenon is observed in the recently solved CXC
chemokine receptor 4 (CXCR4) crystal structure, where the two
small molecule antagonist 1T forms ionic interactions with D2.63
and another where the imidazole moiety binds to D3.32 (mode
II, i.e., in Figure 5B and Figure 5D).13,14,16,17
The hydrogen bond network with D3.32 (an essential H4R
hydrogen bond anchor as determined by previous SDM)1,28,42,47
was further investigated by molecular dynamics (MD)
simulations (step 4 in Figure 2). Analysis of the MD trajectories
(Supporting Information Table 5) showed that the H-bond
network between D3.32 and 2 remained stable in both binding
modes I (observed in 99% of the MD snapshots, starting from
the structure in Figure 5A) and II (observed in 97% of the
snapshots, starting from the structure in Figure 5B). Compound
5, however, only formed a stable H-bond with D3.32 in binding
mode II (observed in 99% of the MD snapshots, starting from
the structure in Figure 5D). This suggests that binding mode II is
the most plausible binding orientation for 5, while 2 is able to
adopt both binding modes I and II in the H4R binding pocket.
8140
dx.doi.org/10.1021/jm201042n|J. Med. Chem. 2011, 54, 8136−8147

Journal of Medicinal Chemistry
Article
By combining 3D-QSAR studies (Figure 4) and H4R receptor
modeling studies (Figure 5, Table 3), we proposed residues that
determinant of H4R−ligand binding based on binding mode I
(Figure 5A and Figure 5C), while residues V3.40, N4.57, T5.42, and
T6.55 are selected based on binding mode II (Figure 5D and
Figure 5D). We furthermore wanted to investigate the role of
D3.32 and E5.46 as putative ionic/H-bond interaction anchors of 2
and 5 (Figure 5). Unfortunately, a suitable D3.32 mutant to study
radioligand displacement is not available because mutation of
this residue is detrimental to binding of both histamine (3) and
JNJ-7777120 (5-chloro-2-[(4-methylpiperazin-1-yl)carbonyl]-
1H-indole, 4c).12 Our H4R models, however, suggest that Q7.42
forms an intramolecular H-bond with D3.32 (Figure 5) and might
indirectly modulate ligand interactions with this residue. The
previously reported H4R E5.46Q mutant has interesting ligand
dependent effects on binding affinity12,42 and was therefore
selected to study the role of the negatively ionizable E5.46 in
binding of 2 and 5.
In Silico Guided Site-Directed Mutagenesis Studies
Identify Molecular Determinants of H3R/H4R Selectivity
and Elucidate Ligand Binding Modes in H4R. Our
integrative ligand-based 3D-QSAR and protein-based H4R
modeling approach was used to design site-directed muta-
genesis studies to elucidate the binding mode of clobenpropit
and its analogues in H4R (step 5 in Figure 2). Table 3 shows
the binding affinities of 2 and 5 at the E5.46Q mutant and the
five H4R mutants mimicking the H3R binding pocket (V3.40A,
N4.57Y, T5.42A, T6.55 M, Q7.42L).
a
Table 3. The affinity (pKi) of 1 and 6 at the human H3R ,
b
c
the human H4R , and some H4R mutants
d
Mutant
pKi
2
5
H3R WT
H4R WT
9.2 ± 0.14
8.1 ± 0.08
8.0 ± 0.07
7.8 ± 0.07
8.6 ± 0.13
7.3 ± 0.10
8.8 ± 0.08
9.0 ± 0.17
9.3 ± 0.34
7.5 ± 0.11
7.2 ± 0.08
6.5 ± 0.05
7.9 ± 0.15
6.8 ± 0.07
7.9 ± 0.19
7.9 ± 0.14
H4R V3.40
H4R N4.57
A
Y
H4R T5.42
H4R E5.46
H4R T6.55
H4R Q7.42
A
Q
M
L
a
Measured by displacement of [3H]N-α-methylhistamine binding
using membranes of HEK293T cells transiently expressing the human
b
H3R. Measured by displacement of [3H]histamine binding using
membranes of HEK293T cells transiently expressing either the human
c
H4R or the H4R mutants. H4R-E5.46Q data were obtained by
d
displacement of [3H]-JNJ-7777120 (tritiated 4c). Data shown are
mean ± SEM of at least three independent experiments.
might play an important role in H4R-ligand binding. We
hypothesized that residues responsible for H3R/H4R selectivity
(Table 1) are H4R specific (i.e., are different residues in H3R)
and match the essential interaction probes derived from our 3D-
QSAR studies (i) within 4.5 Å from the hydrophobic centroid of
the cyclohexyl moiety (Figure 4A and Figure 4B), (ii) within
3.9 Å from the hydrophobic centroid of the aromatic moiety
(Figure 4C), or (iii) within 3.7 Å from the protonated nitrogen
(Figure 4A). In this way, residue Q7.42 is identified as a potential
The affinities of 2 and 5 at the E5.46Q mutant compared to
H4R WT are decreased by about 6-fold (Table 4, Figure 6A,B).
Interestingly, the same mutation results in an even larger
decrease in the affinity for small ligands containing two basic
groups (1 and 4d (VUF8430)) but does not affect the binding
affinity of the larger hydrophobic ligands with one basic moiety
(4c and 4f (clozapine)).12,42 Apparently, 2 and its analogue 5,
Figure 6. Radioligand displacement curves of compound 2 (A, C) and compound 5 (C, D) on a selected set of indicated H4R mutants E5.46Q (∗);
Q7.42L (○); N4.57Y (◇); T6.55 M (▽) and WT (●). E5.46Q mutant data were determined with [3H]JNJ-7777120 radioligand, whereas
[3H]histamine was used for H4R WT and all other H4R mutants. Data shown are representative specific binding curves of at least three experiments
performed in triplicate. Error bars indicate SEM values. To enable a better visualization of the pKi shifts, each curve is corrected for the radioligand
concentration and the Kd for WT or mutant H4Rs.
8141
dx.doi.org/10.1021/jm201042n|J. Med. Chem. 2011, 54, 8136−8147

Journal of Medicinal Chemistry
Article
which also contain two basic groups but are relatively larger and
more hydrophobic than 3 and 4d, depend only partly on ionic
interactions with E5.46. The H4R Q7.42L mutant has a
significantly increased affinity for 2 (9-fold compared to H4R
WT), resulting in pKi comparable to that for H3R, but has a
much smaller (2-fold) effect on the affinity of 5 (Table 3,
Figure 6A,B). This residue, which was identified by probe O.1
in 3D-QSAR studies (Figure 4A; step 2 in Figure 2) and
homology modeling (Figure 5A and Figure 5C), acts as a clear
selectivity switch for 2.
CONCLUSIONS
■
By combining ligand-based QSAR models, protein-based H4R
modeling studies, and in silico guided site-directed mutagenesis
experiments, we have systematically identified the molecular
determinants that drive H3R/H4R selectivity and used this
information to elucidate the binding modes of clobenpropit (2)
and its analogues (5, 6) in the H4R binding pocket. QSAR
studies on a novel series of clobenpropit derivatives containing
two lipophilic moieties indicated that descriptors related to
ligand size and conformation energy were correlated with H4R
binding affinity and identified molecular interaction probes that
drive H3R over H4R selectivity. Linking these essential 3D-
QSAR probes to specific residues in H4R receptor models
enabled the design of site-directed mutagenesis studies to
elucidate H4R−ligand binding modes. Our studies indicate that
clobenpropit (2) can adopt two distinct binding modes in H4R,
while the addition of a cyclohexyl group to the clobenpropit
isothiourea moiety allows compound 5 to adopt only one
specific binding orientation in the H4R binding cavity. Our
experimentally supported ligand-steered protein modeling
method gives new insights into ligand recognition by H4R
and can be used as a general approach to elucidate the structure
of protein−ligand complexes.
Similar to the H4R Q7.42L mutation, the T6.55 M mutation has
a larger effect on the affinity for 2 than for 5 (Table 3, Figure
6C,D). Residues T5.42 and T6.55 were identified by probe DRY.3
in 3D-QSAR studies (Figure 3C) and 3D homology models
(Figure 5B and Figure 5D). Since the H4R T6.55 M mutant
shows more pronounced effect in determining H3R selectivity
toward H4R compared to H4R T5.42A mutant, we propose that
DRY.3 matches residue T6.55. Interestingly, the N4.57Y mutation
significantly reduces the affinity of 5 by 10-fold but only slightly
affects the affinity of 2 (Table 3, Figure 6C,D). The residue at
position 4.57 has been shown to be an important molecular
determinant of H4R species differences in ligand binding.28 In
pig and dog H4R the H4.57 residue is proposed to influence the
orientation of E5.46 by forming direct H-bond interactions.28
We postulate the same mechanism for the N4.57Y mutant and
hypothesize that the larger Y4.57 residue fixes the orientation of
the E5.46 side chain in a way that is not compatible with the
binding mode of 5 in the H4R pocket. While clobenpropit can
adopt the position of its imidazole (mode I) or isothiurea
(mode II) group to this alternative orientation of E5.46, the
binding orientation of clobenpropit analogue 5 is relatively rigid
because of its bulky cyclohexyl group which binds deep in the
hydrophobic pocket between TM3, TM4, TM5, and TM6
(Figure 5). The N4.57 residue was identified by probes DRY.2
and DRY.4 in the 3D-QSAR studies (Figure 4B and Figure 4C)
based on binding mode II (Figure 5B and Figure 5D). The H4R
V3.40A mutant was studied, since it was indicated by similar
probes identifying residue N4.57. However, despite the fact that
the residue in position 3.40 is a determinant in H3R species
METHODS
■
Synthetic Methods. Chemicals were obtained from commercial
suppliers and were used without further purification. 1H NMR and 13
C
NMR spectra were recorded on a Bruker AC-200 (200 MHz)
spectrometer or a Bruker Avance 400 (400 MHz) spectrometer.
Microwave assisted chemistry was performed with a Biotage Initiator
typically using 2 or 5 mL vials obtained from Biotage (Sweden). Purity
of the synthesized compounds 2, 5, and 6a−l was verified to be at least
95% by liquid chromatography−mass spectrometry (LCMS, see
Supporting Information Table 1). LCMS analyses were performed
with a Shimadzu LCMS-2010 EV mass spectrometer, a Shimadzu LC-
20 AB pump system, a SPD-M20 A diode array detector, and a CTO-
20 AC column oven using an XBrigde column (C18, 5 μm, 4.6 mm ×
50 mm). An aqueous buffer (pH 8) of 0.04% NH4HCO3 (solvent A)
and a mixture of 90% MeCN and 10% of a 0.4% NH4HCO3 buffer
(pH 8, solvent B) were used. The runs started with 5% B with a linear
gradient to 90% B in 4.5 min, then continuing for 1.5 min with 90% B
and finally a linear gradient to 5% B in 0.5 min. Total run time was 8
min. J. T. Baker silica gel was used for flash chromatography. All
melting points are uncorrected and were measured on an Optimelt
automated melting point system from Stanford Research Systems.
General Method A. The following procedure was used to obtain
the intermediate thiourea compounds 8a−l. A solution of 30 mmol of
the amine in 20 mL of diethyl ether was added dropwise to a solution
of 30 mmol of the isothiocyanate in 25 mL of diethyl ether. The
reaction mixture was stirred for 1 h. The precipitated product was
filtered off and washed 3 times with diethyl ether. Isolated yields are
between 85% and 95%.
differences and influences H1R activation,48−50 the V3.40
A
mutant has almost no effect on the affinity of 2 and 5 (Table 3).
Therefore we propose that probes DRY.2 and DRY.4 in the 3D-
QSAR studies (Figure 4B and Figure 4C) pinpoint residue N4.57
as a determinant of H3R/H4R selectivity for this compound.
In conclusion, while binding of compound 2 to H4R is
affected by mutation of the Q7.42 residue (corresponding
binding mode I in Figure 5A) and by mutation of the T6.55
residue identified (corresponding with binding mode II in
Figure 5B), binding affinity of 5 for H4R is only significantly
affected by mutated residue N4.57 (corresponding with binding
mode II in Figure 5D). Therefore, we propose that compound
2 can adopt two different binding modes (I and II) in the H4R
binding pocket, while 5 only adopts binding mode II. These
data are furthermore in line with our molecular dynamics
studies (Supporting Information Table 5), indicating that 5 can
only form a stable H-bond network with D3.32 in this binding
orientation. An alternative explanation for the role of Q7.42 in
clobenpropit binding is that it modulates the clobenpropit
binding orientation by forming an intramolecular H-bond with
the essential D3.32 residue. The binding mode of 5 is more rigid
because its bulky cyclohexyl group binds in the hydrophobic
pocket between TM3, 5, and 6 (Figure 5D) and is therefore
less affected by the Q7.42L mutation.
1-Benzyl-3-cyclopentylthiourea (8a). 1H NMR(CDCl3): δ 1.34−
1.75 (m, 6H); 1.82−2.08 (m, 2H); 4.07 (bs, 1H); 4.69 (d, J = 4.97 Hz,
2H); 5.91 (bs, 2H).
1-Benzyl-3-cyclohexylthiourea (8b). 1H NMR (CDCl3): δ 1.01−
1.41 (m, 5H); 1.48−1.75 (m, 4H); 1.95 (d, J = 11.83 Hz, 2H); 3.80
(bs, 1H); 4.62 (d, J = 4.94 Hz, 2H); 5.68 (bs, 1H); 6.00 (bs, 1H);
7.28−7.35 (m, 5H).
1-Butyl-3-cyclopentylthiourea (8c). 1H NMR (CDCl3): δ 0.92
(t, J = 7.2 Hz, 3H); 1.20−1.80 (m, 10H); 1.95−2.15 (m, 2H); 3.43
(m, 2H); 4.10 (bs, 1H); 5.62 (bs, 1H); 5.80 (bs, 1H).
1-Cyclohexyl-3-(3,4-dichlorobenzyl)thiourea (8d). 1H NMR
(CDCl3): δ 1.11−1.40 (m, 5H); 1.50−1.76 (m, 3H), 1.90−1.99 (m,
2H); 3.73 (bs, 1H); 4.66 (d, J = 5.57 Hz, 2H); 5.89 (bs, 1H); 6.12 (bs,
1H); 7.12 (dd, J1 = 6.25 Hz, J2 = 1.95 Hz, 1H); 7.31−7.42 (m, 2H).
8142
dx.doi.org/10.1021/jm201042n|J. Med. Chem. 2011, 54, 8136−8147

Journal of Medicinal Chemistry
Article
1-Butyl-3-(4-chlorobenzyl)thiourea (8e). 1H NMR (CDCl3): δ
0.88 (t, J = 7.18 Hz, 3H); 1.30 (s, 2H), 1.50 (dd, J1 = 7.6 Hz, J2 = 6.87
Hz, 2H); 3.30 (bs, 2H); 4.64 (d, J = 5.29 Hz, 2H); 5.84 (bs, 1H); 6.00
(bs, 1H); 7.21−7.32 (m, 4H).
26.12; 29.66; 30.69; 32.85; 61.87; 116.56; 128.41; 130.71; 132.15;
133.10; 133.80 (3C); 136.09; 167.83.
3-(1H-Imidazol-4-yl)propyl N-(4-Chlorobenzyl)-N′-butylcarbami-
midothioate (6e). 1H NMR (MeOD, 400 MHz): δ 0.97 (bs, 3H);
1.25−1.55 (m, 2H); 1.58−1.79 (m, 2H); 2.06 (bs, 2H); 2.83 (bs, 2H);
3.28−3.37 (m, 2H); 3.45−3.52 (m, 2H); 4.71 (bs, 2H), 7.24 (s, 1H);
7.32−7.36 (m, 2H); 7.38−7.43 (m, 2H); 8.51 (s, 1H). 13C NMR
(MeOD, 200 MHz): δ 14.03; 20.6; 24.53 (2C); 28.84 (2C); 32.37;
45.62; 117.23; 130.072 (2C); 130.29 (2C); 134.63; 135.21; 168.54.
3-(1H-Imidazol-4-yl)propyl N-(4-Chlorobenzyl)-N′-benzylcarba-
mimidothioate (6f). 1H NMR (MeOD, 200 MHz): δ 2.02 (q, J1 =
7.26 Hz, J2 = 7.62 Hz, 2H); 2.81 (t, J = 7.56 Hz, 2H); 3.35 (t, J = 7.43
Hz, 2.0H); 4.72 (s, 2H); 4.74 (s, 2H); 7.06−7.55 (m, 10H); 8.64 (s,
1H). 13C NMR (MeOD, 500 MHz): δ 24.13; 28.72; 32.49; 61.44
(2C); 117.20; 129.69 (9C); 134.01 (3C); 134.91; 169.02.
3-(1H-Imidazol-4-yl)propyl N-(3,4-Dichlorobenzyl)-N′-benzylcar-
bamimidothioate (6g). 1H NMR (MeOD, 400 MHz)): δ 2.04 (q,
J1 = 6.44 Hz, J2 = 8044 Hz, 2H); 2.84 (t, J = 7.8 Hz, 2H); 3.40 (t, J =
7.37 Hz, 2H); 4.76 (bs, 4H); 7.11−7.68 (m, 9 h); 8.81 (s, 1H). 13C
NMR (MeOD, 400 MHz): δ 24.05; 28.68; 32.51; 43.58; 44.73;
117.24; 127.97; 128.08; 128.41; 128.78; 129.15; 129.46; 130.06;
130.40; 130.84; 131.45; 132.04; 133.70; 133.81; 134.70; 134.88;
169.29.
3-(1H-Imidazol-4-yl)propyl N-Cyclohexyl-N′-cyclopentylcarbami-
midothioate (6h). 1H NMR (400 MHz, MeOD): δ 1.18−1.59 (m,
5H); 1.69 (bs, 5H); 1.82 (bs, 4H); 1.95 (bs, 2H); 2.05−2.15 (m, 4H);
2.93 (t, J = 3.8 Hz, 2H); 3.33−3.39 (m, 2H); 3.84 (bs, 1H); 4.15−4.45
(bs, 1H); 7.45 (s, 1H), 8.88 (s, 1H). 13C NMR (MeOD, 400 MHz): δ
24.22; 24.83; 25.98 (3C); 28.74; 32.98 (5C); 117.36; 133.95; 135.04;
166.29.
3-(1H-Imidazol-4-yl)propyl N-Cyclopentyl-N′-(tert-butyl)-
carbamimidothioate (6i). 1H NMR (MeOD, 400 MHz, T = 330
K): δ 1.52 (s, 9H); 1.62−1.76 (m, 4H); 1.76−1.91 (m, 2H); 2.93 (t,
J = 8.0 Hz, 2H); 3.35 (t, J = 8.0 Hz, 2H); 4.30−4.40 (m, 1H), 7.39 (s,
1H); 8.73 (s, 1H). 13C NMR (MeOD, 200 MHz): δ 24.49; 24.88
(2C); 28.85; 29.00; 29.63 (3C); 33.28 (2C); 34.18; 58.17; 117.37;
134.20; 135.04; 167.93.
3-(1H-Imidazol-4-yl)propyl N-(3,4-Dichlorobenzyl)-N′-butylcar-
bamimidothioate (6j). 1H NMR (MeOD, 400 MHz): δ 0.87−1.04
(m, 3H); 1.20−1.51 (m, 2H); 1.52−1.79 (m, 2H); 1.99−2.18 (m,
2H); 2.78−2.98 (m, 2H); 3.30−3.46 (m, 2H); 3.53 (t, J = 7.5 Hz,
2H); 4.75 (s, 2H); 7.27−7.43 (m, 2H); 7.55 (d, J = 7.80 Hz, 2H); 8.80
(s, 1H). 13C NMR (MeOD, 400 MHz): δ 14.40; 21.41; 24.62; 29.12;
31.42; 32.94; 46.10; 49.00; 117.72; 128.97 (2C); 132.00; 132.53;
134.16; 134.40; 169.06.
3-(1H-Imidazol-4-yl)propyl N-(4-Chlorobenzyl)-N′-cyclohexylcar-
bamimidothioate (5). 1H NMR (MeOD, 200 MHz): δ 1.22 (bs,
1H); 1.31−1.51 (m, 4H); 1.68 (d, J = 12.0 Hz, 1H); 1.82 (d, J = 12.2
Hz, 2H); 1.89−2.12 (m, 4H); 2.80 (t, J = 7.2 Hz, 2H); 3.72−3.79 (m,
1H); 4.72 (s, 2H); 7.25 (s, 1H); 7.34 (d, J = 8.5 Hz, 2H); 7.39 (d, J =
8.5 Hz, 2H); 8.56 (s, 1H). 13C NMR (MeOD, 200 MHz): δ 24.55;
25.87 (2C); 25.93; 28.82; 48.87; 55.20; 117.09; 130.06 (6C); 134.73;
167.59.
3-(1H-Imidazol-4-yl)propyl N-(4-Cyanobenzyl)-N′-cyclohexylcar-
bamimidothioate (6k). 1H NMR (MeOD, 500 MHz): δ 1.14−1.60
(m, 5H); 1.68−2.20 (m, 7H); 2.74−3.00 (m, 2H); 3.30−3.45 (m,
2H); 3.82 (s, 1H); 4.86 (s, 2H); 7.38 (s, 1H); 7.57 (d, J = 8.25 Hz,
2H)); 7.77 (d, J = 8.03 Hz, 2H); 8.80 (d, J = 1.23 Hz, 1H). 13C NMR
(MeOD, 500 MHz): δ 21.87; 24.10; 25.89; 28.56 (2C); 32.11; 55.74;
61.37; 117.23; 119.27; 129.30; 133.76 (4C); 133.92 (2C); 134.93;
167.72.
3-(1H-Imidazol-4-yl)propyl N-(4-Nitrobenzyl)-N′-cyclohexylcar-
bamimidothioate (6l). 1H NMR (MeOD, 500 MHz): δ 1.06−1.62
(m, 5H); 1.7 (bs, 1H); 1.76−2.33 (m, 6H)2.74−3.00 (m, 2H); 3.31−
3.56 (m, 2H); 3.85 (s, 1H)7.37 (s, 1H); 7.63 (d, J = 8.78 Hz, 2H);
8.25 (d, J = 8.52 Hz, 2H); 8.79 (s, 1H). 13C NMR (MeOD, 500
MHz): δ 24.19; 25.91; 25.97; 28.66 (2C); 32.73; 55.86; 61.45; 117.30;
124.98 (4C); 129.57; 133.99; 134.16; 134.98; 167.83.
1-Benzyl-3-(4-chlorobenzyl)thiourea (8f). 1H NMR (CDCl3): δ
4.57 (d, J = 6.67 Hz, 2 h); 4.60 (d, J = 5.94 Hz, 2H); 6.00 (bs, 1H);
6.15 (bs, 1H), 7.08−7.40 (m, 9H).
1-Benzyl-3-(3,4-dichlorobenzyl)thiourea (8g). 1H NMR (CDCl3):
δ 4.56 (d, J = 5.25 Hz, 2H); 4.65 (d, J = 5.65 Hz, 2H); 5.97 (bs, 1H);
6.25 (bs, 1H); 7.02 (dd, J1 = 8.12 Hz, J2 = 0 Hz, 1 h); 7.12−7.48
(m, 7H).
1-Cyclohexyl-3-cyclopentylthiourea (8h). 1H NMR (CDCl3): δ
1.04−1.78 (m,16H); 1.82−2.12 (m, 2H); 4.03 (bs, 2H); 5.53 (bs,
1H); 5.83 (bs, 1H).
1-(tert-butyl)-3-cyclopentylthiourea (8i). 1H NMR (CDCl3): δ
1.39 (s, 10H); 1.40−1.73 (m, 7H); 1.94−2.16 (m, 2H); 4.35 (bs, 1H);
5.64 (bs, 1H); 5.84 (bs, 1H).
1-Butyl-3-(3,4-dichlorobenzyl)thiourea (8j). 1H NMR (CDCl3): δ
0.87 (t, J = 7.1 Hz, 3H); 1.30 (m, J1 = 8.09 Hz, J2 = 7.02 Hz, J3 = 7.15
Hz, 2H); 1.50 (q, J1 = 3.75 Hz, J2 = 7.43 Hz, 2H); 3.30 (bs, 2H); 4.65
(d, J = 5.57 Hz, 2H); 6.10 (bs, 1H); 6.28 (bs, 1H); 7.11 (dd, J1 = 6.26
Hz, J2 = 2.01 Hz, 1H); 7.34 (d, J = 2.98 Hz, 1H); 7.36 (d, J = 3.22 Hz,
1H).
1-(4-Cyanobenzyl)-3-cyclohexylthiourea (8k). 1H NMR (CDCl3,
200 MHz): δ 0.99−1.43 (m, 5H); 1.50−1.82 (m, 3H); 1.87−2.08
(m, 2H); 3.76 (bs, 1H), 4.78 (d, J = 5.68 Hz, 2H); 6.00 (bs, 1H); 6.33
(bs, 1H); 7.36 (d, J = 8.20 Hz, 2H); 7.51 (d, J = 8.30 Hz, 2H).
1-Cyclohexyl-3-(4-nitrobenzyl)thiourea (8l). 1H NMR (CDCl3): δ
1.04−1.47 (m, 5H); 1.52−1.82 (m, 3H); 1.83−2.08 (m, 2H); 3.71 (bs,
1H); 4.89 (d, J = 5.80 Hz, 2H); 5.89 (bs, 1H), 6.03 (bs, 1H); 7.45 (d,
J = 8.59 Hz, 2H); 8.14 (d, J = 8.68 Hz, 2H).
General Method B. The following procedure was used to obtain
the clobenpropit derivatives 6a−l.
4-(3-Bromopropyl)-1H-imidazole hydrobromide (400 mg, 1.48
mmol) and an excess of the corresponding thiourea (500 mg) were
dissolved in 3 mL of ethanol in a microwave tube. The reaction
mixture was heated in the microwave to 120 °C during 30 min. The
ethanol was evaporated, and the residue was purified with flash column
chromatography. After the solvents were evaporated, the residue was
crystallized from ethanol. The 1H and 13C NMR data are given below,
whereas isolated yields, purity by LCMS, and HRMS data are
presented in the Supporting Information.
3-(1H-Imidazol-4-yl)propyl N-Benzyl-N′-cyclopentylcarbamimi-
dothioate (6a). 1H NMR (MeOD, 400 MHz): δ 1.54−1.82 (m,
6H); 1.95−2.21 (m, 4H); 2.78−2.98 (m, 2H); 3.33−3.41 (m, 2H);
4.24 (bs, 1H); 7.33 (s, 1H); 7.39 (bs, 5H); 8.84 (s, 1H). 13C NMR
(MeOD, 200 MHz): δ 24.08; 24.52; 24.90; 28.67; 32.96; 34.10; 44.85;
53.35; 57.64; 117.31; 128.19; 128.41; 129.14; 129.53; 130.06 (2C);
133.86; 134.98.
3-(1H-Imidazol-4-yl)propyl N-Benzyl-N′-cyclohexylcarbamimido-
thioate (6b). 1H NMR (MeOD, 400 MHz): δ 1.20 (bs, 1H); 1.32−
1.51 (m, 4H); 1.63−1.72 (m, 1H); 1.76−1.85 (m, 2H); 1.86−2.09 (m,
4H); 2.65−2.79 (m, 2H); 3.24 (t, J = 7.10 Hz, 2H); 3.73 (bs, 1H);
4.72 (s, 2H); 6.95 (s, 1H); 7.28−7.43 (m, 5H); 7.85 (bs, 1H). 13C
NMR (MeOD, 400 MHz): δ 26.18; 26.30; 26.40 (2C); 29.88; 49.40;
55.80; 117.10; 128.92 (3C); 129.69; 130.46 (2C); 136.41; 168.16.
3-(1H-Imidazol-4-yl)propyl N-Butyl-N′-cyclopentylcarbamimido-
thioate (6c). 1H NMR (MeOD, 200 MHz): δ 0.97 (t, J = 7.20 Hz,
3H); 1.39 (sx, J1 = 8.20 Hz, J2 = 7.10 Hz, J3 = 7.20 Hz, 2H); 1.57−1.94
(m, 8H); 1.95−2.03 (m, 4H); 2.77 (t, J = 7.10 Hz, 2H); 3.26 (t, J =
7.50 Hz, 2H); 4.08 (bs, 1H); 6.94 (s, 1H); 7.75 (s, 1H). 13C NMR
(MeOD, 200 MHz): δ 14.82; 14.89; 21.74; 24.97; 25.30; 25.72; 29.50;
33.62; 33.90; 34.38; 34.72; 42.00; 118.15; 134.69; 135.79.
3-(1H-Imidazol-4-yl)propyl N-(3,4-Dichlorobenzyl)-N′-cyclohexyl-
carbamimidothioate (6d). 1H NMR (MeOD, 200 MHz): δ 1.11−
1.58 (m, 6H); 1.63−2.12 (m, 6H); 2.71 (t, J = 7.23 Hz, 2H); 3.25 (t,
J = 7.06 Hz,2H); 3.76 (bs, 1H); 4.71 (s, 2H); 6.90 (s, 1H); 7.30 (dd,
J1 = 6.15, J2 = 2.10, 1H); 7.53 (s, 1H); 7.55 (d, J = 8.25 Hz, 1H); 7.73
(s, 1H). 13C NMR (MeOD, 500 MHz): δ 25.94 (2C); 26.03 (2C);
Pharmacology. DNA Constructs and Site-Directed Muta-
genesis. The wild-type human H3R and H4R cDNA (pcDNA3.1)
were purchased from the Missouri S & T cDNA Resource Center
8143
dx.doi.org/10.1021/jm201042n|J. Med. Chem. 2011, 54, 8136−8147

Journal of Medicinal Chemistry
Article
(Rollo, MO)/Guthrie cDNA Resource Center (Sayre, PA). The
human H4R cDNA was subcloned into mammalian expression vector
pcDEF3 (a gift from Dr. J. Langer) using BamHI and XbaI restriction
sites. Site-directed mutagenesis was performed with the fusion PCR
method using oligonucleotide primers containing the indicated
mutations. The mutations were verified by sequence analysis at
ServiceXS (Leiden, The Netherlands).
Cell Culture, Transfection, and Membrane Preparation. HEK293T
cells were cultured in Dulbecco’s modified Eagle medium (DMEM)
supplemented with 10% fetal bovine serum (FBS), 50 IU/mL
penicillin, and 50 μg/mL streptomycin at 37°C and 5% CO2. Cells
were seeded 2 × 106 per 10 cm dish 1 day prior to transfection.
Approximately 4 × 106 cells were transfected with 5 μg of cDNA using
the PEI method. Briefly, 5 μg of cDNA was mixed with 20 μg of 25
kDa linear polyethyleneimine in 500 μL of 150 mM NaCl. This
transfection mix was incubated at room temperature for 10−30 min
and subsequently added dropwise to a 10 cm dish with 6 mL of fresh
culture medium. Two days after transfection, transfected cells were
washed once with phosphate buffered saline (PBS) and subsequently
scraped from their culture dish in 1 mL of PBS. Crude membrane
extracts were collected by centrifugation at ∼2000g for 10 min at 4°C
and stored at −20 °C until further use.
Radioligand Binding Assays. Binding assays were performed using
crude membrane extracts from transfected cells in 50 mM Tris-HCl
binding buffer (pH 7.4 at room temperature). For heterologous
displacement studies, crude membrane extracts were co-incubated
with increasing (1 pM to 0.1 mM) concentration of compounds and
∼10 nM [3H]histamine, ∼10 nM [3H]JNJ-7777120 (tritiated 4c), or
∼1 nM [3H]N-α-methylhistamine (NAMH) in a total volume of 100
μL/well. The reaction suspensions were incubated for 1−1.5 h at room
temperature on a shaking table (750 rpm). Bound radioligand was
separated from free radioligand via rapid filtration over a 0.5% PEI-
presoaked glass fiber C plate (GF/C, Perkin-Elmer). GF/C plates
were subsequently washed three times with ice-cold 50 mM Tris-HCl
wash buffer (pH 7.4 at 4 °C). The retained radioactivity on the GF/C
plates was counted by liquid scintillation counting in a Wallac
Microbeta (Perkin-Elmer).
method from the Weka 3.6.2 data-mining software package followed
by the stepwise method from the SPSS 14.0 for Windows was
subsequently used to select the important probes and generate QSAR
models, with the dependent variables being pKi(H3R) and
pKi(H4R).24,32,51
Homology Modeling Procedure. The H4R model was constructed
as previously described.28 Very recently, during the preparation of the
current paper, the doxepin-bound crystal structure of the histamine H1
receptor was published. A preliminary H4R homology model made
based on this H1R crystal structure is very similar to the H4R model
used in the current study, which is derived from the ADRB2 crystal
structure41 (data not shown). This is in line with the fact that the H1R
and ADRB2 crystal structures have a similar structure (Supporting
Information Figure 2).53 Compound 6 was docked into the H4R
receptor model without constraint using PLANTS.55 The first pose
showing adjacency between the imidazole moiety and E5.46 was
subjected to energy minimization in MOE, version 2009.10,32 using
restrained H-bonds between the protonated imidazole nitrogen atom
of the ligand and one of the carboxylate oxygen atoms (OE2) of E5.46
and between the protonated isothiourea moiety of the ligand and one
of the carboxylate oxygen atoms (OD2) of D3.32 (model 6A, Figure
3C). On the other hand, the first pose showing adjacency between
the imidazole moiety and D3.32 was subjected to energy minimization
in MOE, version 2009.10,32 using restrained H-bonds between
the protonated imidazole nitrogen atom of the ligand and one of
the carboxylate oxygen atoms (OD2) of D3.32 and between the
protonated isothiourea moiety of the ligand and one of the carboxylate
oxygen atoms (OE2) of E5.46 (model 6B, Figure 3D). Compound 1
was docked without constraint into model 6A.55 The first ranked pose
was subjected to energy minimization in MOE, version 2009.10,32
using restrained H-bonds between the protonated imidazole nitrogen
atom of the ligand and one of the carboxylate oxygen atoms (OE2) of
E5.46 and between the protonated isothiourea moiety of the ligand and
one of the carboxylate oxygen atoms (OD2) of D3.32 (model 1A,
Figure 3A). Compound 1 was also docked without constraint into the
model 6B.55 The first ranked pose was subjected to energy
minimization in MOE, version 2009.10,32 using restrained H-bonds
between the protonated imidazole nitrogen atom of the ligand and one
of the carboxylate oxygen atoms (OD2) of D3.32 and between the
protonated isothiourea moiety of the ligand and one of the carboxylate
oxygen atoms (OE2) of E5.46 (model 1B, Figure 3B).
Molecular Dynamics Procedure. The models 1A, 1B, 6A, and 6B
were then minimized using AMBER 10 to relax the structure.56 Force-
field parameters for the ligands were derived using the Antechamber
program, and partial charges for the ligands were computed using the
AM1-BCC procedure in Antechamber.56 Upper-bound distance
restraint of 3.5 Å to maintain the interaction of the ligand to D3.32
3.5 Å was applied. The minimized model was subsequently embedded
in a pre-equilibrated lipid bilayer consisting of molecules of 1-
palmitoyl-2-oleoylphosphatidylcholine (POPC) and solvated with
TIP3P water molecules (box dimensions of 79.2 Å × 72.0 Å × 76.9 Å)
as described by Urizar et al.57 The complexes embedded in the
hydrated lipid bilayer were minimized shortly using AMBER 10.56 The
hydrogen bond to D3.32 constraint and a positional harmonic
constraint of 50 kcal mol−1 Å−1 on Cα carbon atoms were applied.
The entire system was then subjected to a 1.1 ns constant pressure
molecular dynamics (MD) simulation. All bonds involving hydrogen
atoms were frozen with the SHAKE algorithm. During the first 100 ps,
the Cα carbon atoms were constrained and the hydrogen bond of the
ligand to D3.32 was restrained as previously described and the
temperature was linearly increased from 0 to 300 K. During the last
1000 ps, the temperature was kept constant at 300 K and the pressure
at 1 bar, using a coupling constant of 0.2 ps and the Berendsen
approach. Interactions were calculated according to the AMBER 03
force field, using particle-mesh-Ewald (PME) summation to include
the long-range electrostratic forces. van der Waals interactions were
calculated using a cutoff of 8.0 Å. MD snapshots were extracted at
every 1 ps.58,59 This yielded 1000 snapshot for each model. The
binding pocket regions of MD snapshots were then fitted to the
corresponding binding pocket regions of the initial 3D model. The
Data Analysis. pKi values were calculated using nonlinear
regressions for a single binding site model using GraphPad Prism
4.0. Kd values were previously determined.12,28
Materials. [Ring, methylenes-3H(N)]histamine (10−40 Ci/mmol)
and N-α-[methyl-3H]methylhistamine (45−90 Ci/mmol) were
purchased from Perkin-Elmer. [3H]JNJ-7777120 (56.1 Ci/mmol)
was a kind gift from R. Thurmond (J&J, La Jolla, CA).
Computational Methods. Classical QSAR Procedure. The in
silico building, conformational search, and molecular descriptor
calculations of 2, 5, and 6a−l were performed using Molecular
Operating Environment (MOE) software, version 2006.08, developed
by Chemical Computing Group, Inc. (Montreal, Canada) as described
previously. The stepwise method from the SPSS 14.0 for Windows was
subsequently used to select the important descriptors and generate
QSAR models, with the dependent variables being pKi(H3R) and
pKi(H4R).24,32,51
3D-QSAR Procedure. The structures and the major protonation
states of compounds 2, 5, and 6a−l at pH 7.4 were built and
determined using MarvinSketch, version 5.2.5.1, from ChemAxon.52
The 3D structure of compound 6 was subsequently generated using
CORINA 3.46, resulting in two 3D configurations.34 Ab initio
optimizations using DFT/6-31G* by employing the Gaussian 03
software package were performed to both 3D configurations.35−37 The
lower energy configuration suggested by the ab initio calculation was
selected as the reference 3D structure to align protonated 1 and 7a−l
by using OMEGA, version 2.3.2, and ROCS, version 2.3.1.38,39 The
alignment was refined subsequently using MOE, version 2009.10.32
The MIF probes (DRY, N1, O, and O::) were then calculated using
the GRID package, version 22.0.3, from Molecular Discovery.26,53 The
probes in a radius of 5 Å around aligned compounds were calculated
using a grid resolution of 1 Å. The probes values were normalized, and
probes with standard deviation of less than 1.0 were filtered out by
employing R statistical package, version 2.7.1.54 The GreedyStepwise
8144
dx.doi.org/10.1021/jm201042n|J. Med. Chem. 2011, 54, 8136−8147

Journal of Medicinal Chemistry
Article
interaction fingerprint (IFP) for each snapshot was calculated
subsequently.60 Snapshots with hydrogen bond to D3.32 were counted.
(6) Lim, H. D.; van Rijn, R. M.; Ling, P.; Bakker, R. A.; Thurmond,
R. L.; Leurs, R. Evaluation of histamine H1-, H2-, and H3-receptor
ligands at the human histamine H4 receptor: identification of 4-
methylhistamine as the first potent and selective H4 receptor agonist.
J. Pharmacol. Exp. Ther. 2005, 314, 1310−1321.
(7) Liu, C.; Wilson, S. J.; Kuei, C.; Lovenberg, T. W. Comparison of
human, mouse, rat, and guinea pig histamine H4 receptors reveals
substantial pharmacological species variation. J. Pharmacol. Exp. Ther.
2001, 299, 121−130.
(8) Nguyen, T.; Shapiro, D. A.; George, S. R.; Setola, V.; Lee, D. K.;
Cheng, R.; Rauser, L.; Lee, S. P.; Lynch, K. R.; Roth, B. L.; O’Dowd, B.
F. Discovery of a novel member of the histamine receptor family. Mol.
Pharmacol. 2001, 59, 427−433.
(9) Oda, T.; Morikawa, N.; Saito, Y.; Masuho, Y.; Matsumoto, S.
Molecular cloning and characterization of a novel type of histamine
receptor preferentially expressed in leukocytes. J. Biol. Chem. 2000,
275, 36781−367816.
(10) Gemkow, M. J.; Davenport, A. J.; Harich, S.; Ellenbroek, B. A.;
Cesura, A.; Hallett, D. The histamine H3 receptor as a therapeutic drug
target for CNS disorders. Drug Discovery Today 2009, 14, 509−515.
(11) Uveges, A. J.; Kowal, D.; Zhang, Y.; Spangler, T. B.; Dunlop, J.;
Semus, S.; Jones, P. G. The role of transmembrane helix 5 in agonist
binding to the human H3 receptor. J. Pharmacol. Exp. Ther. 2002, 301,
451−459.
(12) Jongejan, A.; Lim, H. D.; Smits, R. A.; de Esch, I. J.; Haaksma,
E.; Leurs, R. Delineation of agonist binding to the human histamine
H4 receptor using mutational analysis, homology modeling, and ab
initio calculations. J. Chem. Inf. Model. 2008, 48, 1455−1463.
(13) Ishikawa, M.; Watanabe, T.; Kudo, T.; Yokoyama, F.; Yamauchi,
M.; Kato, K.; Kakui, N.; Sato, Y. Investigation of the histamine H3
receptor binding site. Design and synthesis of hybrid agonists with a
lipophilic side chain. J. Med. Chem. 2010, 53, 6445−6456.
(14) Kiss, R.; Noszal, B.; Racz, A.; Falus, A.; Eros, D.; Keseru, G. M.
Binding mode analysis and enrichment studies on homology models of
the human histamine H4 receptor. Eur. J. Med. Chem. 2008, 43, 1059−
1070.
(15) Lorenzi, S.; Mor, M.; Bordi, F.; Rivara, S.; Rivara, M.; Morini,
G.; Bertoni, S.; Ballabeni, V.; Barocelli, E.; Plazzi, P. V. Validation of a
histamine H3 receptor model through structure−activity relationships
for classical H3 antagonists. Bioorg. Med. Chem. 2005, 13, 5647−5657.
(16) Schlegel, B.; Laggner, C.; Meier, R.; Langer, T.; Schnell, D.;
Seifert, R.; Stark, H.; Holtje, H. D.; Sippl, W. Generation of a
homology model of the human histamine H3 receptor for ligand
docking and pharmacophore-based screening. J. Comput.-Aided Mol.
Des. 2007, 21, 437−453.
(17) Rai, B. K.; Tawa, G. J.; Katz, A. H.; Humblet, C. Modeling G
protein-coupled receptors for structure-based drug discovery using
low-frequency normal modes for refinement of homology models:
application to H3 antagonists. Proteins 2009, 78, 457−473.
(18) Wijtmans, M.; de Graaf, C.; de Kloe, G.; Istyastono, E. P.; Smit,
J.; Lim, H.; Boonnak, R.; Nijmeijer, S.; Smits, R. A.; Jongejan, A.;
Zuiderveld, O.; de Esch, I. J.; Leurs, R. Triazole ligands reveal distinct
molecular features that induce histamine H4 receptor affinity and
subtly govern H4/H3 subtype selectivity. J. Med. Chem. 2011, 54,
1693−1703.
ASSOCIATED CONTENT
* Supporting Information
■
S
Purity and HRMS data for compounds 2, 5, and 6a−l as
determined by LCMS; alignment of compounds 2, 5, and 6a−l
used in 3D-QSAR and the values of the most influential
descriptors in eqs 1-5; analysis of H4R-ligand H-bond
interactions MD simulations; comparison of ligand binding
modes in ADRB2 and H1R crystal structures and H4R models.
This material is available free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*Phone: +31(0)205987553. Fax: +31(0)205987610. E-mail:
C.de.Graaf@vu.nl.
■
Author Contributions
§These authors contributed equally.
ACKNOWLEDGMENTS
■
This work was supported by the Top Institute Pharma [Project
Number D1.105: The GPCR Forum], The Netherlands
Organization for Scientific Research [NWO VENI Grant
700.59.408], and COST Action BM0806. Sven Uleman is
acknowledged for technical assistance.
ABBREVIATIONS USED
■
3D-QSAR, three-dimensional quantitative structure−activity
relationship; ADRB1, β1-adrenergic; ADRB2, β2-adrenergic;
DRD3, dopamine D3; GPCR, G protein-coupled receptor;
H1R, histamine H1 receptor; H3R, histamine H3 receptor; H4R,
histamine H4 receptor; SDM, site-directed mutagenesis
ADDITIONAL NOTE
■
During the preparation of this manuscript, a crystal structure of
the human histamine H1 receptor (H1R) was solved (PDB code
3RZE, released June 15, 2011) by Iwata et al. (Nature 2011,
475, 65−70 ). The overallstructure of H1R is shown to be
similar to the β2-adrenergic (ADRB2) crystal structure. As a
result, H4R models based on the new H1R crystal structure are
very similar to the ADRB2-based H4R models described in the
current study (C. de Graaf, unpublished results).
REFERENCES
■
(1) Istyastono, E. P.; de Graaf, C.; de Esch, I. J. P.; Leurs, R.
Molecular determinants of selective agonist and antagonist binding to
the histamine H4 receptor. Curr. Top. Med. Chem. 2011, 11, 661−679.
(2) Liu, C.; Ma, X.; Jiang, X.; Wilson, S. J.; Hofstra, C. L.; Blevitt, J.;
Pyati, J.; Li, X.; Chai, W.; Carruthers, N.; Lovenberg, T. W. Cloning
and pharmacological characterization of a fourth histamine receptor
(H(4)) expressed in bone marrow. Mol. Pharmacol. 2001, 59, 420−
426.
(19) Cherezov, V.; Rosenbaum, D. M.; Hanson, M. A.; Rasmussen,
S. G.; Thian, F. S.; Kobilka, T. S.; Choi, H. J.; Kuhn, P.; Weis, W. I.;
Kobilka, B. K.; Stevens, R. C. High-resolution crystal structure of an
engineered human β2-adrenergic G protein-coupled receptor. Science
2007, 318, 1258−1265.
(20) Chien, E. Y.; Liu, W.; Zhao, Q.; Katritch, V.; Han, G. W.;
Hanson, M. A.; Shi, L.; Newman, A. H.; Javitch, J. A.; Cherezov, V.;
Stevens, R. C. Structure of the human dopamine D3 receptor in
complex with a D2/D3 selective antagonist. Science 2010, 330, 1091−
1095.
(3) Smits, R. A.; Leurs, R.; de Esch, I. J. Major advances in the
development of histamine H4 receptor ligands. Drug Discovery Today
2009, 14, 745−753.
(4) Thurmond, R. L.; Gelfand, E. W.; Dunford, P. J. The role of
histamine H1 and H4 receptors in allergic inflammation: the search for
new antihistamines. Nat. Rev. Drug Discovery 2008, 7, 41−53.
(5) Zampeli, E.; Tiligada, E. The role of histamine H4 receptor in
immune and inflammatory disorders. Br. J. Pharmacol. 2009, 157,
24−33.
(21) Shimamura, T.; Shiroishi, M.; Weyand, S.; Tsujimoto, H.;
Winter, G.; Katritch, V.; Abagyan, R.; Cherezov, V.; Liu, W.; Han, G.
W.; Kobayashi, T.; Stevens, R. C.; Iwata, S. Structure of the human
8145
dx.doi.org/10.1021/jm201042n|J. Med. Chem. 2011, 54, 8136−8147

Journal of Medicinal Chemistry
Article
histamine H1 receptor complex with doxepin. Nature 2011, 475,
65−70.
(22) Warne, T.; Serrano-Vega, M. J.; Baker, J. G.; Moukhametzianov,
R.; Edwards, P. C.; Henderson, R.; Leslie, A. G.; Tate, C. G.; Schertler,
G. F. Structure of a β1-adrenergic G-protein-coupled receptor. Nature
2008, 454, 486−491.
(23) de Graaf, C.; Rognan, D. Customizing G protein-coupled
receptor models for structure-based virtual screening. Curr. Pharm.
Des. 2009, 15, 4026−4048.
Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J.
J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.;
Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman,
J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.;
Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen,
W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03; Gaussian
Inc.: Pittsburgh, PA, 2003.
(24) Lim, H. D.; Istyastono, E. P.; van de Stolpe, A.; Romeo, G.;
Gobbi, S.; Schepers, M.; Lahaye, R.; Menge, W. M.; Zuiderveld, O. P.;
Jongejan, A.; Smits, R. A.; Bakker, R. A.; Haaksma, E. E.; Leurs, R.; de
Esch, I. J. Clobenpropit analogs as dual activity ligands for the
histamine H3 and H4 receptors: synthesis, pharmacological evaluation,
and cross-target QSAR studies. Bioorg. Med. Chem. 2009, 17, 3987−
3994.
(25) De Esch, I. J.; Mills, J. E.; Perkins, T. D.; Romeo, G.; Hoffmann,
M.; Wieland, K.; Leurs, R.; Menge, W. M.; Nederkoorn, P. H.; Dean,
P. M.; Timmerman, H. Development of a pharmacophore model for
histamine H3 receptor antagonists, using the newly developed
molecular modeling program SLATE. J. Med. Chem. 2001, 44,
1666−1674.
(26) Dezi, C.; Carotti, A.; Magnani, M.; Baroni, M.; Padova, A.;
Cruciani, G.; Macchiarulo, A.; Pellicciari, R. Molecular interaction
fields and 3D-QSAR studies of p53-MDM2 inhibitors suggest
additional features of ligand−target interaction. J. Chem. Inf. Model.
2010, 50, 1451−1465.
(27) Lim, H. D.; Jongejan, A.; Bakker, R. A.; Haaksma, E.; de Esch, I.
J.; Leurs, R. Phenylalanine 169 in the second extracellular loop of the
human histamine H4 receptor is responsible for the difference in
agonist binding between human and mouse H4 receptors.
J. Pharmacol. Exp. Ther. 2008, 327, 88−96.
(28) Lim, H. D.; de Graaf, C.; Jiang, W.; Sadek, P.; McGovern, P. M.;
Istyastono, E. P.; Bakker, R. A.; de Esch, I. J.; Thurmond, R. L.; Leurs,
R. Molecular determinants of ligand binding to H4R species variants.
Mol. Pharmacol. 2010, 77, 734−743.
(29) Bloemhoff, W.; Kerling, K. Recl. Trav. Chim. Pays-Bas 1970, 89,
1181.
(30) Kivits, G. A. A.; Hora, J. A convenient preparation of 3-(1H-
imidazol-4-yl)propanol. J. Heterocycl. Chem. 1975, 12, 577.
(31) Wildman, S. A.; Crippen, G. M. Prediction of physiochemical
parameters by atomic contributions. J. Chem. Inf. Comput. Sci. 1999,
39, 868−873.
(32) MOE, version 2009.10; Chemical Computing Group (CCG):
Montreal, Canada, 2009; http://www.chemcomp.com/software.htm:.
(33) Muzammil, S.; Armstrong, A. A.; Kang, L. W.; Jakalian, A.;
Bonneau, P. R.; Schmelmer, V.; Amzel, L. M.; Freire, E. Unique
thermodynamic response of tipranavir to human immunodeficiency
virus type 1 protease drug resistance mutations. J. Virol. 2007, 81,
5144−5154.
(34) Sadowski, J.; Gasteiger, J.; Klebe, G. Comparison of automatic
three-dimensional model builders using 639 X-ray structures. J. Chem.
Inf. Comput. Sci. 1994, 34, 1000−1008.
(35) Iqbal, P.; Patel, D. S.; Bharatam, P. V. Ab initio study on
N,N2,N3-triaminoguanidine. J. Phys. Org. Chem. 2007, 20, 1072−1080.
(36) Netzeva, T. I.; Aptula, A. O.; Benfenati, E.; Cronin, M. T.; Gini,
G.; Lessigiarska, I.; Maran, U.; Vracko, M.; Schuurmann, G.
Description of the electronic structure of organic chemicals using
semiempirical and ab initio methods for development of toxicological
QSARs. J. Chem. Inf. Model. 2005, 45, 106−114.
(37) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.;
Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.;
Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.;
Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.;
Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao,
O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J.
B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R.
E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.;
(38) ROCS, version 2.3.1; OpenEye Scientific Software Inc.: Santa
Fe, NM , 2009; http://www.eyesopen.com.
(39) OMEGA, version 2.3.2; OpenEye Scientific Software Inc.: Santa
Fe, NM, 2008; http://www.eyesopen.com.
(40) Eriksson, L.; Jaworska, J.; Worth, A. P.; Cronin, M. T.;
McDowell, R. M.; Gramatica, P. Methods for reliability and
uncertainty assessment and for applicability evaluations of classi-
fication- and regression-based QSARs. Environ. Health Perspect. 2003,
111, 1361−1375.
(41) Rucker, C.; Rucker, G.; Meringer, M. y-Randomization and its
variants in QSPR/QSAR. J. Chem. Inf. Model. 2007, 47, 2345−2357.
(42) Shin, N.; Coates, E.; Murgolo, N. J.; Morse, K. L.; Bayne, M.;
Strader, C. D.; Monsma, F. J. Jr. Molecular modeling and site-specific
mutagenesis of the histamine-binding site of the histamine H4
receptor. Mol. Pharmacol. 2002, 62, 38−47.
(43) Wu, B.; Chien, E. Y.; Mol, C. D.; Fenalti, G.; Liu, W.; Katritch,
V.; Abagyan, R.; Brooun, A.; Wells, P.; Bi, F. C.; Hamel, D. J.; Kuhn,
P.; Handel, T. M.; Cherezov, V.; Stevens, R. C. Structures of the
CXCR4 chemokine GPCR with small-molecule and cyclic peptide
antagonists. Science 2010, 330, 1066−1071.
(44) Scholten, D. J.; Canals, M.; Maussang, D.; Roumen, L.; Smit, M.
J.; Wijtmans, M.; de Graaf, C.; Vischer, H. F.; Leurs, R.
Pharmacological modulation of chemokine receptor function. Br. J.
Pharmacol. [Online early access]. DOI: 10.1111/j.1476-
5381.2011.01551.x. Published Online: Jun 23, 2011.
(45) Kufareva, I.; Rueda, M.; Katritch, V.; Stevens, R. C.; Abagyan, R.
Status of GPCR modeling and docking as reflected by community-
wide GPCR Dock 2010 assessment. Structure 2011, 19, 1108−1126.
(46) Roumen, L.; Sanders, M. P. A.; Vroling, B.; de Esch, I. J. P.; De
Vlieg, J.; Leurs, R.; Klomp, J.; Nabuurs, S. B.; de Graaf, C. In silico
veritas: the pitfalls and challenges of predicting GPCR−ligand
interactions. Pharmaceuticals 2011, 4, 1196−1215.
(47) Leurs, R.; Chazot, P. L.; Shenton, F. C.; Lim, H. D.; de Esch, I.
J. Molecular and biochemical pharmacology of the histamine H4
receptor. Br. J. Pharmacol. 2009, 157, 14−23.
(48) Sansuk, K.; Deupi, X.; Torrecillas, I. R.; Jongejan, A.; Nijmeijer,
S.; Bakker, R. A.; Pardo, L.; Leurs, R. A structural insight into the
reorientation of transmembrane domains 3 and 5 during family A G
protein-coupled receptor activation. Mol. Pharmacol. 2011, 79, 262−9.
(49) Schnell, D.; Strasser, A.; Seifert, R. Comparison of the
pharmacological properties of human and rat histamine H3-receptors.
Biochem. Pharmacol. 2010, 80, 1437−1449.
(50) Yao, B. B.; Hutchins, C. W.; Carr, T. L.; Cassar, S.; Masters, J.
N.; Bennani, Y. L.; Esbenshade, T. A.; Hancock, A. A. Molecular
modeling and pharmacological analysis of species-related histamine
H(3) receptor heterogeneity. Neuropharmacology 2003, 44, 773−786.
(51) SPSS 14.0 for Windows; SPSS Inc.: Chicago, IL, 2005; http://
spss.com/.
(52) MarvinSketch, version 5.2.5.1; ChemAxon: Budapest, Hungary,
2009.
(53) Goodford, P. J. A computational procedure for determining
energetically favorable binding sites on biologically important
macromolecules. J. Med. Chem. 1985, 28, 849−857.
(54) R: A Language and Environment for Statistical Computing; R
Development Core Team: Vienna, Austria, 2008; http://www.R-
project.org.
(55) Korb, O.; Stutzle, T.; Exner, T. E. Empirical scoring functions
̈
for advanced protein−ligand docking with PLANTS. J. Chem. Inf.
Model. 2009, 49, 84−96.
8146
dx.doi.org/10.1021/jm201042n|J. Med. Chem. 2011, 54, 8136−8147

Journal of Medicinal Chemistry
Article
(56) Case, D. A.; Darden, T. A.; Cheatham, T. E., III; Simmerling, C.
L.; Wang, J.; Duke, R. E.; Luo, R.; Crowley, M.; Walker, R. C.; Zhang,
W.; Merz, K. M.; Wang, B.; Hayik, S.; Roitberg, A.; Seabra, G.;
́
Kolossvary, I.; Wong, K. F.; Paesani, F.; Vanicek, J.; Wu, X.; Brozell, S.
R.; Steinbrecher, T.; Gohlke, H.; Yang, L.; Tan, C.; Mongan, J.;
Hornak, V.; Cui, G.; Mathews, D. H.; Seetin, M. G.; Sagui, C.; Babin,
V.; Kollman, P. A. AMBER 10, University of California: San Francisco,
2008.
(57) Urizar, E.; Claeysen, S.; Deupi, X.; Govaerts, C.; Costagliola, S.;
Vassart, G.; Pardo, L. An activation switch in the rhodopsin family of
G protein-coupled receptors: the thyrotropin receptor. J. Biol. Chem.
2005, 280, 17135−17141.
(58) de Graaf, C.; Foata, N.; Engkvist, O.; Rognan, D. Molecular
modeling of the second extracellular loop of G-protein coupled
receptors and its implication on structure-based virtual screening.
Proteins 2008, 71, 599−620.
(59) Van Der Spoel, D.; Lindahl, E.; Hess, B.; Groenhof, G.; Mark,
A. E.; Berendsen, H. J. GROMACS: fast, flexible, and free. J. Comput.
Chem. 2005, 26, 1701−1718.
(60) Marcou, G.; Rognan, D. Optimizing fragment and scaffold
docking by use of molecular interaction fingerprints. J. Chem. Inf.
Model. 2007, 47, 195−207.
(61) Stanton, D. T. Evaluation and use of BCUT Descriptors in
QSAR and QSPR studies. J. Chem. Inf. Comput. Sci. 1999, 39, 11−20.
8147
dx.doi.org/10.1021/jm201042n|J. Med. Chem. 2011, 54, 8136−8147