Structural Insights into 5-HT1A/D4 Selectivity of WAY-100635 Analogues: Molecular Modeling, Synthesis, and in Vitro Binding

Article abstract of DOI:10.1021/acs.jcim.5b00753

The resurgence of interest in 5-HT_1A receptors as a therapeutic target requires the existence of highly selective 5-HT_1A ligands. To date, WAY-100635 has been the prototypical antagonist of these receptors. However, this compound als

Full text of DOI:10.1021/acs.jcim.5b00753

Article
pubs.acs.org/jcim
Structural Insights into 5‑HT1A/D4 Selectivity of WAY-100635
Analogues: Molecular Modeling, Synthesis, and in Vitro Binding
Sebastien Dilly*,†,‡,§ and Jean-Franco
†Laboratory of Medicinal Chemistry and C.I.R.M. and Laboratory of Pharmacology and GIGA-Neuroscience, University of Lieg
̧
is Liegeois†
́
́
‡
̀
e,
̀
avenue Hippocrate, 15 (B36), B-4000 Liege 1, Belgium
S
* Supporting Information
ABSTRACT: The resurgence of interest in 5-HT1A receptors as a therapeutic
target requires the existence of highly selective 5-HT1A ligands. To date, WAY-
100635 has been the prototypical antagonist of these receptors. However, this
compound also has significant affinity for and activity at D4 dopamine receptors.
In this context, this work was aimed at better understanding the 5-HT1A/D4
selectivity of WAY-100635 and analogues from a structural point of view. In
silico investigations revealed two key interactions for the 5-HT1A/D4 selectivity
of WAY-100635 and analogues. First, a hydrogen bond only found with the Ser
7.36 of D4 receptor appeared to be the key for a higher D4 affinity for newly
synthesized aza analogues. The role of Ser 7.36 was confirmed as the affinity of
aza analogues for the mutant D4 receptor S7.36A was reduced. Then, the
formation of another hydrogen bond with the conserved Ser 5.42 residue
appeared to be also critical for D4 binding.
by the in silico investigations as influencing the selectivity are
then evaluated by a mutagenesis approach and in vitro binding
assays. Furthermore, new compounds were also synthesized
and tested to validate our assumptions.
1. INTRODUCTION
As shown by some recent studies,1−4 5-HT1A receptors are a re-
emerging therapeutic target for various diseases as different as
central nervous system disorders and cancer. Therefore, their
pharmacological investigations require the existence of highly
selective ligands. Up to now, WAY-100635 (compound 1), N-
[2-[4-(2-methoxyphenyl)-1-piperazinyl]-ethyl]-N-(2-pyridin-
yl)-cyclohexane carboxamide (Table 1), has been the
prototypical antagonist of 5-HT1A receptors.5 Regarding
pharmacological studies and pharmacological tools, such a
reference molecule is expected to be highly selective for the
selected target. It is reported that this molecule shows affinity
for and activity at D4 dopamine receptors that could hamper its
use as a pharmacological tool, especially when both receptors
are close together.6 Moreover, in a behavioral model, WAY-
100635 leads to discriminative stimulus effects in rats mediated
by dopamine D4 receptor activation.7 In this context, diverse
chemical changes of the WAY-100635 structure were carried
out with the aim of improving its 5-HT1A versus D4
selectivity.8−13 In particular, the selectivity significantly
increased when the basic side chain of WAY-100635 was
replaced by a 4-phenylpiperazine (compound 3 in Table 1) or a
4-phenyl-1,2,3,6-tetrahydropyridine moiety (compound 5 in
Table 1).12 In the same study,12 interestingly, the presence of
nitrogen atoms in the acyl group both reduced the affinity for 5-
HT1A receptors and increased the affinity for D4 receptors, thus
reducing the 5-HT1A/D4 selectivity (compound 2, 4, and 6 in
Table 1). For the purpose of understanding these changes in 5-
HT1A/D4 selectivity, this present work explores the binding
modes of the compounds by docking analysis on homology
models of the two receptors. The structural features identified
2. MATERIALS AND METHODS
2.1. Molecular modeling. 2.1.1. Receptor Modeling. The
two receptor models were built by homology modeling by
means of the SYBYL 8.0 molecular modeling package14 and
according to several stages.
First, the human sequences of the 5-HT1A and D4 receptors
obtained from the Universal Protein Resource15 (code entries
P0890816 and P21917,17 respectively) were aligned with the
sequence of the turkey β1 adrenergic G-protein coupled
receptor (GPCR) (code entry P07700)18 by using the FUGUE
sequence alignment module.19 An adequate alignment of the
highly conserved residues of the GPCR superfamily according
to Baldwin et al.20 was carefully checked. For 5-HT1A, the
sequence alignment indicated homology rates of 83 and 73%,
respectively, with the full-length sequence and the trans-
membrane (TM) regions of the 5-HT1A sequence. For D4,
these homology rates were 80 and 66%, respectively.
The next step was the copy of a set of constraints derived
from ligand-biased crystal structures of the turkey β1 adrenergic
GPCR to the corresponding residues of the sequences to be
modeled using the ORCHESTRAR protein structure modeling
module.21 Since the compounds are 5-HT1A antagonists,12 the
5-HT1A receptor model was built in its inactive form from a
Received: December 28, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.jcim.5b00753
J. Chem. Inf. Model. XXXX, XXX, XXX−XXX

Journal of Chemical Information and Modeling
Article
Table 1. In Vitro Binding Evaluation of Compounds 1−6 on Cloned 5-HT1A and D4 Receptors12
a
b
Ki in nanomolar; mean SEM, n ≥ 3. The selectivity ratio is calculated by dividing D4 affinity by 5-HT1A affinity.
crystal structure of the inactivated turkey β1 adrenergic
receptor (PDB code 2VT4).22 In the modeling protocol, the
cocrystallized antagonist cyanopindolol in the inactivated β1
crystal structure was replaced by WAY-100635. To this end, an
exhaustive analysis of conformational space (391 conforma-
tions) for WAY-100635 was carried out using the program
Random Search23 of SYBYL (including an energy cutoff of 10
kcal/mol). All the conformations were then aligned according
two common structural features for the ligand binding in
biogenic amine receptors by using a SPL (SYBYL programming
language) program developed by our team. The first key
binding feature is the protonated amine which is involved in a
hydrogen bond with Asp 3.32 (Ballesteros and Weinstein
numbering),24 an important residue for ligand binding among
all mammalian monoamine receptors.25 The second key
binding feature is the aromatic ring which interacts with
aromatic residues in helices V and VI.26−28 In the transfer
process, ORCHESTRAR treated WAY-100635 as a rigid group
by transferring it from the template into the model protein. In
the case of the D4 receptor, the compounds act as agonists.12
Thus, the D4 receptor model was generated in its active form
from a crystal structure of the activated turkey β1 adrenergic
receptor (PDB code 2Y00).29 The cocrystallized agonist
dobutamine in the activated β1 crystal structure was replaced
by WAY-100635 according the same protocol as previously
described in the 5HT1A model generation. In the generation of
both models, N- and C-terminal parts that are not aligned with
the template structures along with the very variable IL3 region
were removed and not regarded since they do not participate in
the ligand binding. Moreover, the conserved disulfide bond
between the cysteine Cys 3.25 of helix 3 and the residue
cysteine in the center of extracellular loop ECL2 was added and
was maintained as a constraint in the model refinement.
force field30 and the same protocol described in our previous
work.31
Lastly, the robustness of the complex models was assessed by
observing their Ramachandran plot using the RAMPAGE
program.32 High quality parameters with a very good
distribution of φ and ψ angles for both models were found.
Indeed, more than 100% and 98% of the residues were located
in the favored and allowed regions for 5-HT1A and D4 receptor
models, respectively. The residues composing the binding site
of the D4 receptor model do not fall into the 2% disallowed
regions.
2.1.2. Ligands Modeling. The protonated compounds were
modeled under the Sybyl 8.0 software14 by using a library of
standard fragments. Their geometry was then optimized using
the MMFF94 force field.30 A conformational analysis was
performed with the program Random Search23 of Sybyl to
identify the low-energy conformer for each compound.
2.1.3. Binding Modes of the Ligands. The binding mode of
the ligands for the two receptors was explored by flexible
docking simulations using the GOLD 5.1 program.33 A sphere
of 15 Å centered on the centroid of WAY-100635 was defined
as the receptor binding pockets. Flexibility of the ligands was
applied by applying torsion angle distributions extracted from
the Cambridge Structural Database (CSD) (http://www.ccdc.
cam.ac.uk/products/csd/). These distributions help GOLD to
find the correct solution by orienting the search toward ligand
torsion-angle values that are frequently found in crystal
structures. Thirty docking runs were performed. The best
docking models were selected on the basis of the high-scoring
conformation given by ChemPLP.34 The docking protocols
were validated by redocking WAY-100635 into the rigid 5-
HT1A and D4 models. The resulting positions of the redocked
WAY-100635 were found to be close to the initial positions
with a RMSD score of 0.75 Å for 5HT1A and 0.52 Å for D4. The
ligand−receptor complexes were further optimized by energy
The resulting WAY-100635−receptor complexes were then
optimized by energy minimization by means of the MMFF94
B
DOI: 10.1021/acs.jcim.5b00753
J. Chem. Inf. Model. XXXX, XXX, XXX−XXX

Journal of Chemical Information and Modeling
Article
minimization using the MMFF94 force field,30 and the same
2.3. Cell Culture, Cell, and Membrane Preparation. sf9
cells expressing wildtype human cloned D4.2 receptors were
employed as membrane preparations (purchased from
PerkinElmer 6110112). Mutant D4 receptor DNAs was
subcloned into the mammalian plasmid expression vector
pCDNA3.1/hygro (+) (Eurogentec, Seraing, Belgium). Mutant
D4 receptors were then transiently expressed in HEK293 cells.
Experimental procedures can be found in our previous
reports.12
protocol defined in our previous work.31
2.2. Chemistry. A general procedure for the preparation of
the target compounds (7−12) is summarized below and
illustrated in Scheme 1. The corresponding amide (∼1.5
a
Scheme 1. Synthesis of Compounds 7−12
2.4. In Vitro Binding. The procedure has been already
applied in the work of Graulich et al.35 and is described in the
Supporting Information section.
2.5. Data Analysis. Statistical analysis was carried out by
means of PRISM 5.02 (GraphPad Software, San Diego, CA,
USA). Data were analyzed with a Kruskal−Wallis test followed
by post hoc Mann−Whitney U test.
3. RESULTS AND DISCUSSION
3.1. Molecular Docking Studies. 3.1.1. 5-HT1A−Ligand
Binding Modes. As presented in Figure 1A, the binding site of
WAY-100635 (compound 1) is located in the top of receptor
helix bundle between TMD III, V, VI, VII and the extracellular
loop 2 and is found to be coherent with data from the literature.
Indeed, in addition to the expected charge reinforced
hydrogen bond between the protonated nitrogen atom and
the residue Asp 116 (Asp 3.32 according to the Ballesteros and
Weinstein numbering24)25 and the edge-to-face CH−π
interaction between the phenyl ring and the residue Phe 362
(Phe 6.52),26 three hydrogen bonds (HB) are detected. First, a
hydrogen bond is found between the methoxy oxygen atom of
the ligand and the hydroxyl group of Ser 199 (Ser 5.42), thus
confirming the involvement of this residue in the binding
process.36 Then, a second hydrogen bond is detected between
the pyridine nitrogen atom and the hydroxyl group of Tyr 390
(Tyr 7.43).
a
Reagents and conditions: (a) Et3N, dioxane, 2−3 h, rt, (b) 1-
phenylpiperazine or 1-(2-methoxyphenyl)piperazine or 4-phenyl-
1,2,3,6-tetrahydropyridine, Et3N, dioxane, 1−2 h, reflux, (c) LiAlH4,
Et2O, 5 min at 0 °C and 1 h at reflux, (d) SOCl2, ACN, reflux, (e) R′-
COCl, Et3N, EtOAc, 2 h, rt.
An interaction with this residue was also found in the ligand
binding of other aryl piperazine derivatives.37 The last hydrogen
bond is detected between the carbonyl oxygen atom of the
ligand and the amide group of Asn 386 (Asn 7.39). This is
coherent with site directed mutagenesis data indicating that the
replacement of Asn 386 (Asn 7.39) by a valine leads to a
decrease in the 5-HT1A binding affinity of WAY100635
(compound 1).38 The geometrical parameters of the
interactions are indicated in Table S1.
The hypothetical binding site of WAY-100635 (compound
1) was then used for docking simulations of compounds 2−6.
The key interactions found for WAY-100635 (compound 1)
are maintained for compound 2 (Figure 1B and Table S1).
However, some interactions are less favorable, which could
explain the lower affinity found for the receptor (10-fold lower
than WAY-100635). Indeed, first, the charge reinforced
hydrogen bond formed with Asp 116 (Asp 3.32) appears
weaker with HB lengths and angles of 1.93 Å and 161° (versus
1.64 Å and 174° for WAY-100635). Second, the hydrogen
bonds formed with Asn 386 (Asn 7.39) and Tyr 390 (Tyr 7.43)
are also weaker due to poor HB angles (86° and 96°,
respectively). The presence of the bulky quinoxaline group
seems to prevent the molecule to adopt a conformation
favorable for optimal interactions with the binding site.
For compound 3 (Figure 1C), the removal of the methoxy
group obviously leads to an absence of interaction with Ser 199
(Ser 5.42) but has little impact on the other key interactions
with a good quality of their parameters (Table S1). It is
mmol) is dissolved in 25 mL anhydrous diethyl ether. Carefully,
an excess of LiAlH4 is added in one time under stirring with
cooling. The mixture is stirred for 5 min at 0 °C. After that, the
suspension is heated at reflux for 1 h. The end of the reaction is
checked by TLC. Then, the mixture is poured on ice with
stirring. The precipitate is removed by filtration and washed
with methylene chloride. The crude product is extracted with
methylene chloride. The organic layer is dried with anhydrous
MgSO4 and evaporated under reduced pressure. The residue is
used immediately in the next step without further purification.
In parallel, the acid chloride (≈1.5 mmol) is prepared by
treatment of the appropriate acid with an excess of SOCl2 in
freshly distilled acetonitrile at reflux. After removal of the
reagent, the crude acid chloride is used immediately.
The crude amine obtained in the previous step is dissolved in
ethyl acetate (20 mL) and Et3N (3 mL) is added. The
corresponding acid chloride is dissolved in a mixture of
methylene chloride and ethyl acetate and added dropwise. The
end of the reaction is checked by TLC. Then, the mixture is
evaporated under reduced pressure and the residue is mixed
with a 10% aqueous potassium carbonate solution. The
suspension is extracted with methylene chloride. The organic
layer is dried with anhydrous MgSO4 and evaporated under
reduced pressure. The product is purified on kieselgel with
ethyl acetate or acetone (or mixture of methanol in ethyl
acetate) as mobile phase and further recrystallized mostly in
diisopropylether.
C
DOI: 10.1021/acs.jcim.5b00753
J. Chem. Inf. Model. XXXX, XXX, XXX−XXX

Journal of Chemical Information and Modeling
Article
Figure 1. Binding modes of WAY-100635 (A) and compounds 2−6 (B−F) in the 5-HT1A receptor model. C atoms are colored gray for the receptor
and magenta for the ligands. O, N, and H are colored red, blue, and white, respectively. The hydrogen and ionic bonds are highlighted by orange
dashed lines.
probably due to the conserved CH-π interaction with Phe 362
(Phe 6.52) that stabilizes the binding.
the ligand and the hydroxy group of Thr 434 (Thr 7.39).
However, some interactions are less favorable than those for 5-
HT1A (Table S2), especially the hydrogen bond with Asp 115
(Asp 3.32) (L = 1.93 Å) and the aromatic stacking with Phe
410 (Phe 6.52) (A = 48°). This could explain the lower affinity
of WAY-100635 for this receptor.
For compounds 4 (Figure 1D), the absence of the methoxy
group coupled with the presence of the bulky quinoxaline
group leads to less favorable interactions, particularly for Asp
116 (Asp 3.32) (L = 1.85 Å) and Asn 386 (Asn 7.39; A = 63°).
This could explain the lower affinity for the receptor.
For compound 2, two binding modes are detected. In the
first one (Figure 2B1 and Table S2), the interactions found
with WAY-100635 (compound 1) are maintained. In addition,
a hydrogen bond is formed between the nitrogen N1 of the
quinoxaline group and the hydroxy group of Ser 431 (Ser 7.36).
In the second binding mode (Figure 2B2 and Table S2), the
same interactions are found but the residue Ser 431 (Ser 7.36)
interact with the ligand through a hydrogen bond formed with
the nitrogen N2 of the quinoxaline group. This interaction
could explain the higher affinity found for the receptor.
For compound 3, the absence of interaction with Ser 196
(Ser 5.42) due to the removal of the methoxy group (Figure 2C
and Table S2) coupled to a weak aromatic interaction with Phe
362 (Phe 6.52) (A = 47°) could explain the low D4 affinity
found for this molecule.
For compound 4, two binding modes are also detected. As
with compound 2, a hydrogen bond is found between the
hydroxy group of Ser 431 (Ser 7.36) and either the nitrogen N1
or the nitrogen N2 of the quinoxaline group (Figure 2D1 and
D2 and Table S2). This interaction seems compensate the
absence of interaction with Ser 196 (Ser 5.42) due to the
removal of the methoxy group and thus stabilize the ligand in
its binding site. This could explain the good D4 affinity found
for this ligand.
For compounds 5 (Figures 1E), the absence of the methoxy
group also reduces the affinity for the receptor, but in a lesser
degree than the other compounds. Indeed, as proposed in
previous work39 in the THP analogues, the phenyl ring presents
in preference an almost coplanar orientation relative to the
THP moiety, which is more favorable for an edge-to-face CH-π
interaction with Phe 362 (Phe 6.52).40 The dihedral angle
measured between the ring planes (92°) is very close to the
optimal value of 90°.41 On the contrary, for the piperazine
analogue 3, the more perpendicular orientation of the phenyl
ring is less favorable for the CH−π interaction with a dihedral
angle of 59°. The same observation can be made for compound
6 that has a better affinity than its piperazine analogue 4.
Indeed, a dihedral angle of 85° is found for compound 6
whereas compound 4 presents a value of 61°.
3.1.2. D4−Ligand Binding Modes. As shown in Figure 2A,
WAY-100635 (compound 1) is located in the top of receptor
helix bundle between TMD III, V, VI, VII, and the extracellular
loop 2 and is found to interact with the same conserved
residues between D4 and 5-HT1A receptor binding sites, i.e. Asp
115 (Asp 3.32), Ser 196 (Ser 5.42), Phe 410 (Phe 6.52), Tyr
438 (Tyr 7.43).
The replacement of Asn 434 (Asn 7.39) by a threonine
residue in D4 receptor does not seem unfavorable since a
hydrogen bond is formed between the carbonyl oxygen atom of
For compound 5, the removal of the methoxy group coupled
to the absence of interaction with the residue Ser 431 (Ser
D
DOI: 10.1021/acs.jcim.5b00753
J. Chem. Inf. Model. XXXX, XXX, XXX−XXX

Journal of Chemical Information and Modeling
Article
Figure 2. Binding modes of WAY-100635 (A) and compounds 2 (two binding modes B1 and B2), 3 (C), 4 (two binding modes D1 and D2), 5 (E),
and 6 (two binding modes F1 and F2) in the D4 receptor model. C atoms are colored gray for the receptor and magenta for the ligands. O, N, and H
are colored red, blue, and white, respectively. The hydrogen and ionic bonds are highlighted by orange dashed lines.
7.36) (Figure 2E) leads to a lower affinity for D4 receptor, but
in a lesser extent than its piperazine analogue (compound 3).
According to the geometrical parameters (Table S2), it is
probably due to the THP moiety that leads to a better CH-π
interaction with Phe 410 (Phe 6.52) (A = 82°) and a stronger
reinforced hydrogen bond formed with Asp 115 (Asp 3.32) (L
= 1.79 Å).
For compound 6, like for compounds 2 and 4, the residue
Ser 431 (Ser 7.36) can form a hydrogen bond either the
nitrogen N1 or the nitrogen N2 of the quinoxaline group
(Figure 2F1 and F2). As compound 5, the THP moiety also
leads to a stronger CH-π interaction with Phe 410 (Phe 6.52)
(A = 80°). These favorable interactions can explain the
relatively good affinity for the receptor despite the absence of
interaction with Ser 196 (Ser 5.42).
In summary, these docking studies first reveal that the
absence of interaction with Ser 5.42 due to the removal of the
methoxy group is much more unfavorable for the binding to D4
receptor than for the binding to 5-HT1A receptors, thus
explaining the higher selectivity for compounds 3 and 5 (782
and 755, respectively). Furthermore, the presence of nitrogen
in the acyl group leads to the formation of an additional
hydrogen bond with only the D4 receptors due to the presence
of the polar residue Ser 7.36 in the immediate vicinity of the
nitrogenous heterocycle (versus an alanine residue in 5HT1A
receptors). Thus, this discriminant hydrogen bond appears as
the key of the observed increase of D4 affinity and consequently
the reduced of 5-HT1A/D4 selectivity for compounds 2, 4, and
6.
3.2. Experimental Validation of in Silico Results.
3.2.1. Mutation of Serine 7.36 of D4 Receptors Decreases
the Affinity of Aza Derivatives. In order to validate our
assumption about the importance of the residue Ser 7.36 in 5-
HT1A/D4 selectivity, a modified D4 receptor was prepared by
replacing this serine residue by an alanine. The affinity of WAY-
100635 and compounds 2−6 was determined in in vitro
binding assays. As presented in Table 2, WAY-100635 and its
related cyclohexyl analogues (3, 5) have a similar affinity for
both the native and the S7.36A mutant receptors (≈1−1.5-fold;
Table 2. In Vitro Binding Affinity of Compounds 1−6 for
Native and Mutant D4 Receptors
a
a
b
compounds
D4
D4 S7.36A
change ratio
1
2
3
4
5
6
52 9.4
15 1.5
946 144
56 2.9
385 28
80 12
65 11
1.5
4.3
1.1
8.8
1.1
6.5
1026 167
496 87
436 75
230 50
35
7
a
b
Ki in nanomolar, mean SEM, n ≥ 3. The change ratio is calculated
by dividing D4 S7.36A affinity by D4 affinity.
E
DOI: 10.1021/acs.jcim.5b00753
J. Chem. Inf. Model. XXXX, XXX, XXX−XXX

Journal of Chemical Information and Modeling
Article
Table 3. In Vitro Binding Affinity of WAY-100635 (1) and Compounds 7−12 for 5-HT1A Receptors, Native, and Mutant D4
Receptors
a
b
c
Ki in nanomolar; mean SEM, n ≥ 3. The 5-HT1A/D4 selectivity is obtained by dividing D4 affinity by 5-HT1A affinity. The change ratio is
calculated by dividing D4 S7.36A affinity by D4 affinity. NT: not tested due to a low D4 affinity.
Figure 3. Binding mode of compound 7 (A) and compound 10 (B) in D4 receptor. C atoms are colored gray for the receptor and magenta for the
ligands. O, N, and H are colored red, blue, and white, respectively. The hydrogen and ionic bonds are highlighted by orange dashed lines.
n ≥ 3, P > 0.05, Mann−Whitney test) while the affinity of
quinoxaline analogues (2, 4, 6) is significantly reduced for the
S7.36A mutant receptor (≈4−9; n ≥ 3, P < 0.05, Mann−
Whitney test). These changes may appear quite low but similar
impacts are reported in the literature.42 Indeed, although some
reported mutations had a strong effect on the interactions with
small molecules studied in this work such as dopamine or
norepinephrine, the impact on bigger molecules is lower and
interestingly in the same range than those observed in our
paper. This is probably due to the fact that bigger molecules
have more interactions, thus stabilizing their binding. In our
case, besides the hydrogen bond found with Ser 7.36, other
interactions are present (H bonds, salt bridge, CH-π
interaction).
Therefore, as previously proposed by in silico data, the polar
residue serine 7.36 plays a major role in the higher D4 affinity
found for aza analogues of WAY-100635.
3.2.2. Identification of the Nitrogen Atom Involved in the
Interaction with Ser 7.36. According to the docking studies of
aza analogues, the residue Ser 7.36 can form a hydrogen bond
either the nitrogen atom N1 or the nitrogen atom N2. In order
to identify the correct binding mode, i.e. to determine which
nitrogen atom is involved in the interaction, 2-quinoline, and 3-
quinoline analogues were synthesized (Scheme 1).
1
All synthesized compounds were subjected to H and 13C
NMR, IR, and elemental analysis before biological testing to
confirm their chemical structure and/or purity. Yields are not
optimized and are between 50 and 70%. Target compounds
(7−12) are evaluated in in vitro binding assays to get the
F
DOI: 10.1021/acs.jcim.5b00753
J. Chem. Inf. Model. XXXX, XXX, XXX−XXX

Journal of Chemical Information and Modeling
Article
Present Address
affinity for native and mutant D4 receptors and also for 5-HT1A
receptors to determine 5-HT1A/D4 selectivity (Table 3).
For 2-quinoline analogues (compounds 7−9 in Table 3), a
significant reduction of the D4 S7.36A affinity is observed (≈4−
6 fold; n ≥ 3, P < 0.05, Mann−Whitney test), thus supporting
the role of the nitrogen in the interaction with the residue Ser
431 (Ser 7.36). The binding mode of compound 7 is presented
in Figure 3A. However, the D4 affinity of these compounds is
lower than that of their quinoxaline analogues (Table 1), thus
suggesting that the presence of a nitrogen atom that position
(N1) is a necessary but not sufficient condition for an optimal
interaction with the D4 binding site.
§Laboratory of medicinal chemistry, Universite
́
Victor Segalen
Bordeaux 2, 146, rue Leo
́
Saignat, 33076 Bordeaux Cedex.
Funding
This study is supported in part by grants of the Fonds de la
Recherche Scientifique-FNRS (F.R.S.-FNRS) (1.5.295.08 and
1.5.208.10 F) and a 2012 grant of the Fonds Leo
the Faculty of Medicine of the University of Lieg
́
n Fred
́ ́
ericq of
e (Belgium).
̀
J.-F.L. is Research Director of the F.R.S.-FNRS.
Notes
The authors declare no competing financial interest.
For 3-quinoline analogue 10 (Table 3), a significant
reduction of the D4 S7.36A affinity is observed but in a lesser
extent than that for their 2-quinoline analogues 7 (n ≥ 3, P <
0.05, Mann−Whitney test). For 3-quinoline analogue 12
(Table 3), a reduction is also found but it is not significant
(n ≥ 3, P = 0.100, Mann−Whitney test). Compound 11 (Table
3) was not tested due to a low D4 affinity. These data show that
an interaction is possible between the nitrogen N2 and Ser 431
(Ser 7.36) (e.g., compounds 7 and 10 in Figure 3B). However,
the geometrical parameters of the hydrogen bond formed with
Ser 431 (Ser 7.36) appears less favorable than those of analogue
7 (L = 3.70 Å and A = 152° versus L = 2.60; A = 170°,
respectively). This could explain why the affinity of compound
10 is less affected by the mutation than that of its 2-quinoline
analogue.
ACKNOWLEDGMENTS
■
The technical assistance of S. Counerotte and C. Gillissen for
elemental analyses and cell culture, respectively, is gratefully
acknowledged. Recordings of NMR spectra were kindly
performed by E. Goffin.
ABBREVIATIONS
■
TMD, transmembrane domain; HB, hydrogen bond; THP,
tetrahydropyridine; GPCR, G protein coupled receptor
REFERENCES
■
(1) Brindisi, M.; Butini, S.; Franceschini, S.; Brogi, S.; Trotta, F.; Ros,
S.; Cagnotto, A.; Salmona, M.; Casagni, M.; Andreassi, M.; Saponara,
S.; Gorelli, B.; Weikop, P.; Mikkelsen, J. D.; Scheel-Kruger, J.;
Sandager-Nielsen, K.; Novellino, E.; Campiani, G.; Gemma, S.
Targeting Dopamine D3 and Serotonin 5-HT1A and 5-HT2A
Receptors for Developing Effective Antipsychotics: Synthesis, Bio-
logical Characterization, and Behavioral Studies. J. Med. Chem. 2014,
57, 9578−9597.
In summary, these results indicate that two binding modes
are possible, but the one involving the nitrogen N1 appears
more favorable.
4. CONCLUSION
(2) Fiorino, F.; Severino, B.; Magli, E.; Ciano, A.; Caliendo, G.;
Santagada, V.; Frecentese, F.; Perissutti, E. 5-HT1A Receptor: An Old
Target as a New Attractive Tool in Drug Discovery from Central
Nervous System to Cancer. J. Med. Chem. 2014, 57, 4407−4426.
(3) Lian, P.; Li, L.; Geng, C.; Zhen, X.; Fu, W. Higher-Affinity
Agonists of 5-HT1AR Discovered through Tuning the Binding-Site
Flexibility. J. Chem. Inf. Model. 2015, 55, 1616−1627.
(4) Luo, M.; Wang, X. S.; Roth, B. L.; Golbraikh, A.; Tropsha, A.
Application of Quantitative Structure−Activity Relationship Models of
5-HT1A Receptor Binding to Virtual Screening Identifies Novel and
Potent 5-HT1A Ligands. J. Chem. Inf. Model. 2014, 54, 634−647.
(5) Forster, E. A.; Cliffe, I. A.; Bill, D. J.; Dover, G. M.; Jones, D.;
Reilly, Y.; Fletcher, A. A Pharmacological Profile of the Selective Silent
5-HT1A Receptor Antagonist, WAY-100635. Eur. J. Pharmacol. 1995,
281, 81−88.
This work aimed at explaining the impact of some chemical
modifications of WAY-100635 on 5-HT1A/D4 selectivity. Our
investigations identified two interactions with a significant role
in that selectivity. The absence of hydrogen bond with Ser 5.42
combined to the presence of an acyl group capable of forming a
hydrogen bond with Ser 7.36 appeared as the key for the
observed increase of selectivity of WAY-100635 analogues.
These results are a step forward for understanding the 5-HT1A/
D4 selectivity. In addition, through this work, we generated and
validated robust models of 5-HT1A and D4 receptors that will be
useful for a future virtual screening to identify new ligands with
a high 5-HT1A/D4 selectivity.
(6) Chemel, B. R.; Roth, B. L.; Armbruster, B.; Watts, V. J.; Nichols,
D. E. WAY-100635 is a Potent Dopamine D4 Receptor Agonist.
Psychopharmacology (Berl) 2006, 188, 244−251.
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jcim.5b00753.
(7) Marona-Lewicka, D.; Nichols, D. E. WAY 100635 Produces
Discriminative Stimulus Effects in Rats Mediated by Dopamine D(4)
Receptor Activation. Behav. Pharmacol. 2009, 20, 114−118.
(8) Al Hussainy, R.; Verbeek, J.; van der Born, D.; Booij, J.;
Herscheid, J. K. Design, Synthesis and In Vitro Evaluation of
Bridgehead Fluoromethyl Analogs of N-{2-[4-(2-Methoxyphenyl)-
Piperazin-1-yl]Ethyl}-N-(Pyridin-2-yl)Cyclohexanecarboxamide
(WAY-100635) for the 5-HT(1A) Receptor. Eur. J. Med. Chem. 2011,
46, 5728−5735.
(9) Al Hussainy, R.; Verbeek, J.; van der Born, D.; Braker, A. H.;
Leysen, J. E.; Knol, R. J.; Booij, J.; Herscheid, J. K. Design, Synthesis,
Radiolabeling, and In Vitro and In Vivo Evaluation of Bridgehead
Iodinated Analogues of N-{2-[4-(2-Methoxyphenyl)Piperazin-1-yl]-
Ethyl}-N-(Pyridin-2-yl)Cyclohexanecarboxamide (WAY-100635) as
Potential SPECT Ligands for the 5-HT1A Receptor. J. Med. Chem.
2011, 54, 3480−3491.
Ramachandran plots and 3D coordinates of the 5-HT1A
and D4 receptor models; additional tables containing the
geometrical parameters of the interactions found
between compounds 1−6 and the receptor models;
details about spectroscopic data, elemental analysis
results, and melting points for compounds 7−12; details
about in vitro binding procedure (PDF)
AUTHOR INFORMATION
■
Corresponding Author
*Tel.: +33 5 57 57 47 01. E-mail: sebastien.dilly@gmail.com.
G
DOI: 10.1021/acs.jcim.5b00753
J. Chem. Inf. Model. XXXX, XXX, XXX−XXX

Journal of Chemical Information and Modeling
Article
(10) Childers, W. E., Jr.; Havran, L. M.; Asselin, M.; Bicksler, J. J.;
Chong, D. C.; Grosu, G. T.; Shen, Z.; Abou-Gharbia, M. A.; Bach, A.
C., 3rd; Harrison, B. L.; Kagan, N.; Kleintop, T.; Magolda, R.;
Marathias, V.; Robichaud, A. J.; Sabb, A. L.; Zhang, M. Y.; Andree, T.
H.; Aschmies, S. H.; Beyer, C.; Comery, T. A.; Day, M.; Grauer, S. M.;
Hughes, Z. A.; Rosenzweig-Lipson, S.; Platt, B.; Pulicicchio, C.; Smith,
D. E.; Sukoff-Rizzo, S. J.; Sullivan, K. M.; Adedoyin, A.; Huselton, C.;
Hirst, W. D. The Synthesis and Biological Evaluation of Quinolyl-
Piperazinyl Piperidines as Potent Serotonin 5-HT1A Antagonists. J.
Med. Chem. 2010, 53, 4066−4084.
Serotonin Receptors Revealed by Mutation and Molecular Modeling
of Conserved Residues in Transmembrane Region 5. Mol. Pharmacol.
2000, 58, 877−886.
(29) Warne, T.; Moukhametzianov, R.; Baker, J. G.; Nehme, R.;
Edwards, P. C.; Leslie, A. G.; Schertler, G. F.; Tate, C. G. The
Structural Basis for Agonist and Partial Agonist Action on a Beta(1)-
Adrenergic Receptor. Nature 2011, 469, 241−244.
(30) Halgren, T. A. Merck Molecular Force Field. I. Basis, Form,
Scope, Parameterization, and Performance of MMFF94. J. Comput.
Chem. 1996, 17, 490−519.
́ ́
(31) Dilly, S.; Scuvee-Moreau, J.; Wouters, J.; Liegeois, J.-F. The 5-
(11) Khatri, M.; Rai, S. K.; Alam, S.; Vij, A.; Tiwari, M. Synthesis and
Pharmacological Evaluation of New Arylpiperazines N-[4-[4-(Aryl)
Piperazine-1-yl]-Phenyl]-Amine Derivatives: Putative Role of 5-HT1A
Receptors. Bioorg. Med. Chem. 2009, 17, 1890−1897.
HT1A agonism potential of substituted-piperazine-ethyl-amide de-
rivatives is conserved in the hexyl homologues: molecular modeling
and pharmacological evaluation. J. Chem. Inf. Model. 2011, 51, 2961−
2966.
(12) Mangin, F.; Dilly, S.; Joly, B.; Scuvee
́
-Moreau, J.; Evans, J.;
(32) Lovell, S. C.; Davis, I. W.; Arendall, W. B.; de Bakker, P. I. W.;
Word, J. M.; Prisant, M. G.; Richardson, J. S.; Richardson, D. C.
Structure Validation by Calpha Geometry: Phi, Psi and Cbeta
Deviation. Proteins: Struct., Funct., Genet. 2003, 50, 437−450.
(33) Jones, G.; Willett, P.; Glen, R. C.; Leach, A. R.; Taylor, R.
Development and Validation of a Genetic Algorithm for Flexible
Docking. J. Mol. Biol. 1997, 267, 727−748.
(34) 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.
Setola, V.; Roth, B. L.; Liegeois, J.-F. Moderate Chemical
́
Modifications of WAY-100635 Improve the Selectivity for 5-HT1A
versus D4 Receptors. Bioorg. Med. Chem. Lett. 2012, 22, 4550−4554.
(13) Marchais-Oberwinkler, S.; Nowicki, B.; Pike, V. W.; Halldin, C.;
Sandell, J.; Chou, Y. H.; Gulyas, B.; Brennum, L. T.; Farde, L.;
Wikstrom, H. V. N-oxide Analogs of WAY-100635: New High Affinity
5-HT(1A) Receptor Antagonists. Bioorg. Med. Chem. 2005, 13, 883−
893.
(14) SYBYL, version 8.0; Tripos Inc.: St. Louis, MO, 2008.
(15) Consortium, T. U. O. a. f. d. a. t. U. P. R. N. A. R., 39, D214-
D219. Ongoing and future developments at the Universal Protein
Resource. Nucleic Acids Res. 2011, 39, D214−D219.
(16) The Universal Protein Resource. http://www.uniprot.org/
uniprot/P08908 (accessed July 1, 2016).
(17) The Universal Protein Resource. http://www.uniprot.org/
uniprot/P21917 (accessed July 1, 2016).
(18) The Universal Protein Resource. http://www.uniprot.org/
uniprot/P07700 (accessed July 1, 2016).
(35) Graulich, A.; Leo
́
nard, M.; Res
́
imont, M.; Huang, X. P.; Roth, B.
L.; Lieg
́
eois, J.-F. Chemical Modifications on 4-Arylpiperazine-Ethyl
Carboxamide Derivatives Differentially Modulate Affinity for 5-HT1A,
D4.2, and α2A Receptors: Synthesis and In Vitro Radioligand Binding
Studies. Aust. J. Chem. 2010, 63, 56−67.
(36) Ho, B. Y.; Karschin, A.; Branchek, T.; Davidson, N.; Lester, H.
A. The Role of Conserved Aspartate and Serine Residues in Ligand
Binding and in Function of the 5-HT1A Receptor: a Site-Directed
Mutation Study. FEBS Lett. 1992, 312, 259−262.
(37) Nowak, M.; Kolaczkowski, M.; Pawlowski, M.; Bojarski, A. J.
Homology Modeling of the Serotonin 5-HT1A Receptor using
Automated Docking of Bioactive Compounds with Defined Geometry.
J. Med. Chem. 2006, 49, 205−214.
(19) Shi, J.; Blundell, T. L.; Mizuguchi, K. FUGUE: Sequence-
Structure Homology Recognition Using Environment-Specific Sub-
stitution Tables and Structure-Dependent Gap Penalties. J. Mol. Biol.
2001, 310, 243−257.
(20) Baldwin, J. M.; Schertler, G. F.; Unger, V. M. An Alpha-Carbon
Template for the Transmembrane Helices in the Rhodopsin Family of
G-Protein-Coupled Receptors. J. Mol. Biol. 1997, 272, 144−164.
(38) Kuipers, W.; Link, R.; Standaar, P. J.; Stoit, A. R.; Van
Wijngaarden, I.; Leurs, R.; Ijzerman, A. P. Study of the Interaction
between Aryloxypropanolamines and Asn386 in Helix VII of the
Human 5-Hydroxytryptamine1A Receptor. Mol. Pharmacol. 1997, 51,
889−896.
́
(21) Dilly, S.; Liegeois, J.-F. Interaction of Clozapine and its
Nitrenium Ion with Rat D2 Dopamine Receptors: In Vitro Binding
and Computational Study. J. Comput.-Aided Mol. Des. 2011, 25, 163−
169.
́
(39) Dilly, S.; Graulich, A.; Liegeois, J.-F. Molecular Modeling Study
of 4-Phenylpiperazine and 4-Phenyl-1,2,3,6-Tetrahydropyridine De-
rivatives: a New Step Towards the Design of High-Affinity 5-HT1A
Ligands. Bioorg. Med. Chem. Lett. 2010, 20, 1118−1123.
(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 Beta1-Adrenergic G-Protein-Coupled Receptor.
Nature 2008, 454, 486−491.
(23) Saunders, M. Stochastic Exploration of Molecular Mechanics
Energy Surfaces. Hunting for the Global Minimum. J. Am. Chem. Soc.
1987, 109, 3150−3152.
(24) Ballesteros, J. A.; Weinstein, H. [19] Integrated Methods for the
Construction of Three-Dimensional Models and Computational
Probing of Structure-Function Relations in G Protein-Coupled
Receptors. Methods Neurosci. 1995, 25, 366−428.
(25) Bywater, R. P. Location and Nature of the Residues Important
for Ligand Recognition in G-Protein Coupled Receptors. J. Mol.
Recognit. 2005, 18, 60−72.
(40) Lieg
́
eois, J.-F.; Lespagnard, M.; Meneses Salas, E.; Mangin, F.;
Scuvee-Moreau, J.; Dilly, S. Enhancing a CH-pi Interaction to Increase
́
the Affinity for 5-HT1A Receptors. ACS Med. Chem. Lett. 2014, 5,
358−362.
(41) Burley, S. K.; Petsko, G. A. Aromatic-Aromatic Interaction: a
Mechanism of Protein Structure Stabilization. Science 1985, 229, 23−
28.
(42) Cummings, D. F.; Ericksen, S. S.; Goetz, A.; Schetz, J. A.
Transmembrane Segment Five Serines of the D4 Dopamine Receptor
Uniquely Influence the Interactions of Dopamine, Norepinephrine,
and Ro10−4548. J. Pharmacol. Exp. Ther. 2010, 333, 682−695.
(26) Choudhary, M. S.; Craigo, S.; Roth, B. L. A Single Point
Mutation (Phe340 → Leu340) of a Conserved Phenylalanine
Abolishes 4-[125I]iodo-(2,5-dimethoxy)phenylisopropylamine and
[3H]mesulergine but not [3H]ketanserin Binding to 5-Hydroxytrypt-
amine2 Receptors. Mol. Pharmacol. 1993, 43, 755−761.
(27) Salom, D.; Lodowski, D. T.; Stenkamp, R. E.; Le Trong, I.;
Golczak, M.; Jastrzebska, B.; Harris, T.; Ballesteros, J. A.; Palczewski,
K. Crystal Structure of a Photoactivated Deprotonated Intermediate of
Rhodopsin. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 16123−16128.
(28) Shapiro, D. A.; Kristiansen, K.; Kroeze, W. K.; Roth, B. L.
Differential Modes of Agonist Binding to 5-Hydroxytryptamine(2A)
H
DOI: 10.1021/acs.jcim.5b00753
J. Chem. Inf. Model. XXXX, XXX, XXX−XXX