5-HT2C receptor selectivity and structure-activity relationship of N-methyl-N-(1-methylpiperidin-4-yl)benzenesulfonamide analogs

Article abstract of DOI:10.1016/j.bmcl.2011.11.001

Agonists of the 5-HT_2C receptor have attracted much attention as therapeutic agents for the treatment of obesity. Subtype selectivity against other 5-HT₂ receptors is one of the most important prerequisites for reducing side effects. We present the synthesis of N-methyl-N-(1- methylpiperidin-4-yl)benzenesulfonamide analogs and their structure-activity relationship studies on 5-HT_2A and 5-HT_2C receptors. Although the compounds showed nanomolar activity to the 5-HT_2C receptor, their selectivity against the 5-HT_2A receptor was modest to low. Molecular modeling studies using homology modeling and docking simulation revealed that selectivity originated from subtype specific residues. The observed binding modes and receptor-ligand interactions provided us a clue for optimizing the selectivity against the 5-HT_2A receptor.

Full text of DOI:10.1016/j.bmcl.2011.11.001

Bioorganic & Medicinal Chemistry Letters 22 (2012) 347–352
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
5-HT2C receptor selectivity and structure–activity relationship
of N-methyl-N-(1-methylpiperidin-4-yl)benzenesulfonamide analogs
Jae Wan Jang a,b, Je-sook Baek c, Gil Don Choi d, Woo-Kyu Park d, Yong Seo Cho a, Du-Jong Baek c,
,
⇑
Ae Nim Pae a,b,
⇑
a Center for Neuro-Medicine, Brain Science Institute, Korea Institute of Science and Technology, PO Box 131, Cheongryang, Seoul 130-650, Republic of Korea
b Department of Medicinal & Pharmaceutical Chemistry, School of Science, University of Science and Technology, 52 Eoeun dong, Yuseong-gu, Daejeon 305-333, Republic of Korea
c Department of Chemistry, College of Natural Sciences, Sangmyung University, Seoul 110-743, Republic of Korea
d Pharmaceutical Screening Research Team, Korea Research Institute of Chemical Technology, Daejeon 305-343, Republic of Korea
a r t i c l e i n f o
a b s t r a c t
Article history:
Agonists of the 5-HT2C receptor have attracted much attention as therapeutic agents for the treatment of
obesity. Subtype selectivity against other 5-HT2 receptors is one of the most important prerequisites for
reducing side effects. We present the synthesis of N-methyl-N-(1-methylpiperidin-4-yl)benzenesulfon-
amide analogs and their structure–activity relationship studies on 5-HT2A and 5-HT2C receptors. Although
the compounds showed nanomolar activity to the 5-HT2C receptor, their selectivity against the 5-HT2A
receptor was modest to low. Molecular modeling studies using homology modeling and docking simula-
tion revealed that selectivity originated from subtype specific residues. The observed binding modes and
receptor–ligand interactions provided us a clue for optimizing the selectivity against the 5-HT2A receptor.
Ó 2011 Elsevier Ltd. All rights reserved.
Received 6 August 2011
Revised 31 October 2011
Accepted 1 November 2011
Available online 15 November 2011
Keywords:
5-HT2c ligands
Obesity
Selectivity
Homology modeling
5-Hydroxytryptamine (5-HT) is a neurotransmitter derived from
tryptophan that plays a variety of important roles in the human
body. Its receptor has various functions related to the central ner-
vous system, such as mood, appetite, and sleep.1,2 The 5-HT2C
receptor has drawn much attention from the pharmaceutical indus-
try because of its involvement in regulating body weight for both
humans and rats. 5-HT2C receptor knockout mice were severely ob-
ese and defective in food intake; appetite and body weight in hu-
mans were reduced by 5-HT2C agonists in clinical studies.3–5 This
receptor belongs to the 5-HT2 receptor family together with
5-HT2A and 5-HT2B, and is found in high concentrations exclusively
in the central nervous system (CNS). Various 5-HT2C agonists have
been developed and have shown weight loss in humans by inhibit-
ing food intake. For example, mCPP and PNU-22394, which are non-
selective 5-HT2C agonists, have caused considerable weight loss in
clinical trials.6 Selective 5-HT2C agonists, such as BVT933 and
APD-356, also have shown significant weight loss.7
is a need to develop 5-HT2C agonists with high selectivity against
5-HT2A. To this end, we carried out high-throughput screening
using a subset of Chemdiv GPCR focused library, and several
chemical scaffolds as a 5-HT2C agonist were uncovered. By modi-
fying one of these structures, we designed a novel N-methyl-N-(1-
methylpiperidin-4-yl)benzenesulfonamide as a 5-HT2C agonist. In
this study, we report the synthesis and biological evaluation of
analogs of this scaffold. Furthermore, to reveal the structural basis
of selectivity against 5-HT2A, binding modes to both receptors
were investigated through molecular modeling. First, due to the
unavailability of crystal structures for these receptors, compara-
tive homology modeling was used to construct the 3D coordi-
nates of the receptors. Then, the synthesized compounds were
docked into the active sites of the models. The study of recep-
tor–ligand interactions offered structural insight for selectivity,
allowing further optimization of the current scaffold.
The synthesis of the target compounds is shown in Scheme 1.
First, methyl-(1-methylpiperidin-4-yl)amine was synthesized from
1-methyl-4-piperidone by reductive amination with methylamine
using sodium cyanoborohydride in 45% yield.10 Then, this secondary
amine was reacted with R1-substituted benzenesulfonyl chlorides in
the presence of triethylamine to give sulfonamides in 72–96%
yield.11
One of the key requirements for developing more effective and
safe 5-HT2C agonists is selectivity against subtypes. The 5-HT2A
receptor is included in the same family as the 5-HT2C receptor
and is also widely expressed in the brain, especially in the cere-
bral cortex. It is implicated in hallucinogenic action induced by
drugs such as lysergic acid diethylamide (LSD).8,9 Therefore, there
The biological activity of the small focused library of benzene-
sulfonamide analogs was evaluated for their binding affinities to
the serotonin receptors 5-HT2A and 5-HT2C (Table 1). In this ser-
⇑
Corresponding authors. Tel.: +82 2 2297 5143 (D.-J.B.); tel.: +82 2 958 5185;
fax: +82 2 958 5189 (A.N.P.).
12
ies of compounds, studies were carried out to examine the effect of
E-mail addresses: djbaek@smu.ac.kr (D.-J. Baek), anpae@kist.re.kr (A.N. Pae).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.11.001

348
J. W. Jang et al. / Bioorg. Med. Chem. Lett. 22 (2012) 347–352
R1
Cl
S
CH3
NH
CH3
R1
O O
O
N
3
CH3NH2
S
O
O
N
N
N
NaBH3CN
Et3N
H3C
H3C
H3C
CH3OH
45%
THF
72~96%
4
2
1
Scheme 1. Synthesis of N-methyl-N-(1-methylpiperidin-4-yl)benzenesulfonamide analogs.
the phenyl ring substitution on the binding affinity for serotonin
receptors. This structural feature was identified on the basis of
the unsubstituted compound 4d, having IC50 values of 1020 and
4260 nM for 5-HT2A and 5-HT2C, respectively. First, the compounds
with hydrophobic alkyl or alkoxy substituents at the para-position
of the phenyl ring, such as compounds 4f–j, 4l, and 4x, were active
for both 5-HT2A and 5-HT2C with IC50 values within the range of
50–1000 nM. In particular, linear chains at the para-position, such
as the butoxy group in compound 4l, showed the best activity in
this series, with IC50 values of 77 and 55 nM for 5-HT2A and 5-
HT2C, respectively. Meanwhile, the bulkier substituents, such as
iso-propyl or tert-butyl in compounds 4f and 4g, exhibited reduced
affinities for the serotonin receptors (161–1060 nM). When the
para-substitution was changed to meta- or ortho-substitution, the
affinity values were generally reduced, as shown in compounds
4e (para-methyl, 2130 and 179 nM), 4s (meta-methyl, 3450 and
579 nM) and 4v (ortho-methyl, 1550 and 1820 nM). This trend
was also observed for F-substitution in compounds 4a, 4q, and
4t, as well as Cl-substitution in compounds 4b, 4r, and 4u. In this
case, when both para- and meta-positions of the phenyl ring were
substituted, as in compound 4w (3,4-dimethyl), the affinity was
further decreased to 5980 and 1940 nM.
Polar substituents were detrimental to both activities. For
example, compounds 4n and 4o, with cyano and nitro groups
as polar substituents, showed very poor affinity (>10,000 nM),
as did other polar substituents including fluoro, trifluorometh-
oxy, and trifluoromethyl groups (compounds 4a, 4m, and 4p;
1680–5900 nM for 5-HT2A, and 2530–>10,000 nM for 5-HT2C).
With regard to the selectivity between 5-HT2A and 5-HT2C, com-
pound 4e (para-methyl) showed better affinity for 5-HT2C over
5-HT2A and had the highest selectivity value (15) in this series
(2130 vs 179 nM). Compounds 4s (meta-methyl, 3450 vs
579 nM) and 4x (1-naphthyl, 502 vs 99 nM) were also effica-
cious, but with lesser degrees of selectivity. However, opposite
results were observed in compounds 4f (para-isopropyl, 161 vs
1010 nM), 4m (para-trifluoromethoxy, 2450 vs >10000 nM), 4c
(para-iodo, 160 vs 766 nM), and 4d (unsubstituted, 1020 vs
4260 nM) with better affinity for 5-HT2A over 5-HT2C. Although
Table 1
Binding activity for N-methyl-N-(1-methylpiperidin-4-yl)benzenesulfonamides at 5-HT2A and 5-HT2C
3
O
S
R2
2
N
4
R1
O
N
N
O
S
5
6
N
O
4x,4y
4a~4w
Compd.
R1
R2
Experimental IC50 at 5-HT2A (nM)
Experimental IC50 at 5-HT2C (nM)
Subtype selectivitya
4a
4b
4c
4d
4e
4f
4g
4h
4i
4j
4k
4l
4m
4n
4o
4p
4q
4r
4-F
4-Cl
4-I
H
5900
537
160
1019
2132
161
519
452
329
281
2534
1208
766
4260
179
1012
1055
164
471
523
1863
55
>10000
>10000
>10000
5733
2422
603
2.3
0.44
0.21
0.24
11.9
0.16
0.49
2.8
0.70
0.54
0.45
1.4
<0.25
—
—
0.30
0.77
1.6
4-Methyl
4-Isopropyl
4-tert-Butyl
4-Ethyl
4-Propyl
4-Phenyl
4-Methoxy
4-Butoxy
4-Trifluoromethoxy
4-Cyano
4-Nitro
831
77
2448
>10000
>10000
1683
1859
986
4-Trifluoromethyl
3-F
3-Cl
4s
4t
3-Methyl
2-F
2-Cl
2-Methyl
3,4-Dimethyl
3451
>10000
3180
1552
5981
502
579
6.0
—
2.7
0.85
3.1
>10000
1195
1824
1944
99
4u
4v
4w
4x
4y
Naphthalene-1-yl
Benzo thiophen-7-yl
5.1
—
>10000
>10000
a
Subtype selectivity: IC50 at 5-HT2A/IC50 at 5-HT2C
.

J. W. Jang et al. / Bioorg. Med. Chem. Lett. 22 (2012) 347–352
349
compound 4l showed the best activity to the 5-HT2C receptor,
compounds 4e and 4x were more promising candidates for 5-
HT2C agonists when selectivity was considered. To find the struc-
tural basis for further optimization of selectivity, we set out to
use molecular modeling methods.
their residue type: position 4.56 (numbering scheme by Ballesteros
and Weinstein14), where isoleucine in 5-HT2A was substituted by
the smaller valine in 5-HT2C and position 5.46, where the serine
in 5-HT2A was replaced by non-polar alanine in 5-HT2C. Therefore,
these different residues might have different interactions with the
ligand. On the other hand, loop regions in protein structures mu-
tate in high probability throughout the same subtype family. Fur-
thermore, extracellular loop 2 (EL2) region in the GPCR family
has been considered important because they affect the protein–li-
gand interactions, thus influencing the docking results and virtual
screening performance.15 We thus carefully considered this region
for subtype-specific ligand binding, but both subtypes displayed
identical residues (leucine at Âl2.52 and aspartate at Âl2.54).
Therefore, the EL2 region does not play any role in ligand
selectivity.
The final alignment and templates were submitted as input in
MODELLER 9v416 implemented in Discovery Studio 2.5
(http://accelrys.com). We chose the best final homology models
for each subtype of receptor by the following method. In total,
100 candidate structures for 5-HT2A and for 5-HT2C were initially
generated. Among them, 10 models having the lowest total prob-
ability density function (PDF) value of MODELLER were selected
based on the assumption that the lowest energy conformation
is near the native structure.17 The side chains of these models
were refined using local rotamer libraries.18 Then, minimization
of these structures was performed with the CHARMm forcefield.19
During the minimization, the protein backbones were held fixed
to prevent collapse of the small GPCR receptor binding site. The
optimized structures were evaluated by their ability to accommo-
date the known ligand structures from docking simulation. The
interactions between the ligands and receptors were considered
Structure-based molecular modeling methods were adopted for
the purpose of investigating receptor–ligand interactions. In the
absence of crystal structures for our target proteins, homology
modeling is a useful way of building the three-dimensional protein
structures. The active form of human b2-adrenergic receptor was
used as a template crystal structure (PDB code: 3P0G). The se-
quences of the human 5-HT2A (P28223) and 5HT2C (P28335) recep-
tors were retrieved from the Swiss-Prot database as a target.
Automated sequence alignment of the template with the targets
was performed using the ClustalW method. The obtained results
were manually adjusted to properly align the highly conserved res-
idues13 of the second extracellular loop (EL) and the fifth TM re-
gion. Full sequence alignments of the targets and template
generated by the methods above are shown in Figure 1. The se-
quence identity and similarity between the targets and template
(5-HT2A: 21.2%/36.7%, 5-HT2C: 24.4%/41.5%) appeared to be too
close to the border line (20% identity) to rely on the homology
model. However, the statistics in the TM region were almost
two-fold higher than those of the proteins as a whole, indicating
good reliability for application to homology modeling (5-HT2A
:
39.1%/62.8%, 5-HT2C: 41.4%/64.7%).
The sequence alignment also revealed the possible origin of
selectivity—regions in which residues were different among the
subtypes. Most of the residues in the proximity of the putative li-
gand binding pocket were well conserved across the subtypes
and even with the template. However, two positions differed in
Figure 1. Sequence alignment of human 5-HT2A and 5-HT2C relative to b2 adrenergic receptor. Identical residues are dark blue, and closely related residues are light blue. The
most conserved residues among the GPCR super-family, cysteine residues of the disulfide bridge between the second extracellular loop region and the third transmembrane
region, are shown in bold. Residues involved in ligand binding are pink and are numbered according to the Ballesteros–Weinstein scheme.

350
J. W. Jang et al. / Bioorg. Med. Chem. Lett. 22 (2012) 347–352
reliable if they were consistent with site-directed mutagenesis
data20–22 (Supplementary data). After discarding the models that
missed interactions, the models were submitted to PROCHECK23
and QPACK24 analysis to validate the geometry of the residues
and assess the quality of the final structures. The Ramachandran
plot of PROCHECK showed that 91.7% and 92.8% of residues were
in the most favored regions for the 5-H2A and 5-HT2C homology
models, respectively. This result indicated the reasonable accu-
racy of the obtained structures (Supplementary data).
Molecular docking was carried out using the final homology
models to study the receptor–ligand interaction. GOLD5.025 molec-
ular docking software was used to produce a protein-ligand com-
plex. The active site was defined using all the residues within a
radius of 10 Å around the highly conserved Asp3.32, which has
an important role in charge-assisted hydrogen bond interaction.
In addition, soft potential was applied to the residues in the active
site to make an induced fit effect during docking. As shown in Fig-
ure 2, all of the subtype-specific residues are exposed to the ligand
binding pocket, implying their importance for interacting with li-
gands. Furthermore, the derivatizing groups were located near
these residues, improving the chances for optimizing selectivity.
It was noted that the selectivity determinant regions were mostly
hydrophobic with residues such as Gly5.42, Thr3.37, and Val3.33,
suggesting the necessity of hydrophobic substituents for activity.
This observation was consistent with our observations from SAR
studies that all of the compounds with polar substituents in the
R1 phenyl ring had no activity to either receptor. All other residues
had similar or identical conformations.
It was suggested by the docking results of compound 4e and 4f
that the bulkier substituent would be harmful to 5-HT2C selectivity
over 5-HT2A. Figure 2 shows that compounds 4e and 4f, which dis-
played the highest selectivity for each receptor, have different
binding conformations and interactions with each active site.
While highly conserved ionic interactions between the protonated
nitrogen of piperidine and residue Asp3.32 occurred in both cases,
the sulfone groups of the ligands interacted with different resi-
dues: Asn6.55 in 5-HT2A and Ser3.36 in 5-HT2C. This result was
probably due to different residues located below the R1 phenyl
ring. In the case of 5-HT2C, Ala5.46 is positioned to be directly fac-
ing the R1 phenyl ring. Therefore, the ligand adopted a binding
mode where the R1 phenyl ring was located deeper in the hydro-
phobic cleft (surrounded by Val3.33, Thr3.37, Val5.45, and
Ala5.46). As a result, the methyl substituent of compound 4e
was well positioned without any steric hindrance, whereas the
bulkier isopropyl group of compound 4f had a poorer fit in the
hydrophobic cleft due to repulsion of surrounding residues. As
for 5-HT2A, the mutation from Ala5.46 to the polar Ser5.46 caused
the hydrophobic cleft to become less hydrophobic, thereby forcing
the R1 phenyl away from position 5.46 due to unfavorable polar–
non-polar interactions. As a result, the sulfone group formed
hydrogen bonds with Asn6.55 instead of Ser3.36. The resulting
binding orientation caused the para-substituent to come into clo-
ser contact with Ile4.56 located in front of the substituent. There-
fore, the bulkier isopropyl substituent of compound 4f would be
more stabilized than the smaller methyl group of compound 4e,
probably because of well-established van der Waals interactions.
The binding mode suggested above also makes it possible to ex-
plain activities and selectivities of other compounds between two
subtypes. The ethyl substituent of compound 4h, which is not
bulky, but linear, showed high activity to 5-HT2C and lower activity
to 5-HT2A, although the selectivity was not significant. This result
is in line with the explanation that the ligands exhibit different
binding conformation to each receptor, and 5-HT2C hydrophobic
pocket can accommodate smaller and more linear component of
para-substituent of phenyl ring than that of 5-HT2A. Compound
4d, which has no substituent at the para-position, showed the
reversal selectivity compared to 4e, but both activities were de-
creased simultaneously, which supports the existence of hydro-
phobic cavity near the para-position of R1 phenyl ring where
filling the substituents with appropriate size and volume in the
space is needed for a stabilization of binding energy. On the other
hand, the activities of compounds with halogen substituents such
as 4a, 4b, and 4c can be accounted for by the binding mode above.
The smallest electron size, F-substituent, has better affinity to 5-
HT2C, whereas the largest electron size, I-substituent, was more
stabilized in the active site of 5-HT2A. This is consistent with the
result that the smaller para-methyl substituent of compound 4e
has higher affinity to 5-HT2C than that of 4f, whereas bulkier iso-
propyl group of 4f binds the pocket of 5-HT2A more favorably than
that of 5-HT2C. It was noticed that the activities of compound 4c to
Figure 2. Results from docking of compounds 4e and 4f in each binding site of the homology models. (A) Homology model of 5-HT2A and docking result to its active site. (B)
Homology model of 5-HT2C and docking result to its active site. The ligand structures are displayed in the stick form and colored orange and green for 4e and 4f, respectively.
The receptor residues are represented in line form and information is listed for residues involved in ligand interactions.

J. W. Jang et al. / Bioorg. Med. Chem. Lett. 22 (2012) 347–352
351
Figure 3. Results from docking compound 4x in each binding pocket of the homology model. (A) Homology model of 5-HT2A and docking result to its active site. (B)
Homology model of 5-HT2C and docking result to its active site. The ligand structures are displayed in stick form and the receptor residues are in line form.
Figure 4. Results from the docking of compound 4l into each binding pocket of the homology models. (A) Homology model of 5-HT2A and docking result to its active site. (B)
Homology model of 5-HT2C and docking result to its active site. The ligand structure is displayed in stick form and the receptor residues are in line form.
both subtypes are similar to that of 4f, although properties of iodo
and isopropyl substituents are quite different each other. It seems
that similar activities are attributable to a similar steric effect to
the hydrophobic pocket of both moieties, which contributes more
dominantly than other electronic effects that are different between
these two substituents.
Compound 4x also seemed to be a promising candidate for fur-
ther optimization. Although the compound displayed high activity
to the 5-HT2C receptor (99 nM for 5-HT2C), the selectivity was
determined to be moderate (IC50 at 5-HT2A/IC50 at 5-HT2C = 5.1).
Contrary to the compounds with a R1 phenyl ring, this compound
showed similar binding modes for both active sites (Figure 3).
The naphthyl substituent was leaning against TM5 and faced the
subtype specific residues (Ser for 5-HT2A and Ala for 5-HT2C). It
was therefore clear that the activity difference of the compound
was attributable to the changed hydrophobic interaction that orig-
inated from the mutation of Ala to Ser at position 5.46. The binding
mode analysis of compounds 4e, 4f, and 4x prompted us to modify
the naphthyl substituent such that it had an additional methyl or
ethyl group to better fit the hydrophobic active site.
Compound 4l showed the highest activity for the 5-HT2C recep-
tor, but with no selectivity against the 5-HT2A receptor. Our homol-
ogy model and docking experiment also provided structural insight
for the underlying mechanism of such biological activity. As shown
in Figure 4, while hydrogen bond interactions with Asp3.32 and
Ser3.35 were similar to previous compounds, there were additional
hydrogen bond interactions between the butoxy oxygen and
Ser5.43 in both cases. These extra hydrogen bond interactions
would explain why the compound is equipotent on 5-HT2A and
5-HT2C. Therefore, the strategy of attaching an alkoxy substituent

352
J. W. Jang et al. / Bioorg. Med. Chem. Lett. 22 (2012) 347–352
(1-methylpiperidin-4-yl)amine (0.200 g, 1.56 mmol) in THF (2 mL),
triethylamine (0.118 g, 1.17 mmol) was added, and the mixture was cooled
to 0 °C. p-Butoxybenzenesulfonyl chloride (0.200 g, 0.780 mmol) was added
and the reaction mixture was stirred for 3 h at room temperature. Brine
solution (1 mL) and water (1 mL) were added to the mixture, and the mixture
was extracted three times with methylene chloride (6 mL each). The organic
layer was dried with anhydrous magnesium sulfate, filtered, and concentrated
in vacuo. The resulting oil was purified by column chromatography
(methanol:ethyl acetate = 3:1) to give 0.212 g (80%) of the title compound:
1H NMR (250 MHz, CDCl3) d 7.70 (2H, dt, J = 8.9, 2.9 Hz), 6.93 (2H, dt, J = 8.9,
2.8 Hz), 4.00 (2H, t, J = 6.5 Hz), 3.77 (1H, m), 2.85 (2H, t, J = 11.7 Hz), 2.72 (3H,
s), 2.25 (3H, s), 2.00 (2H, t, J = 10.8 Hz), 1.73 (4H, m), 1.46 (4H, m), 0.97 (3H, t,
J = 7.3 Hz). HRMS (ESI, M+H); 340.1821 calculated for C17H28N2O3S, found
341.1906.
or other substituents that can act as hydrogen bond acceptors is
unsuitable for ligand selectivity, which is an essential requirement
for the development of a 5-HT2C agonist.
In summary, thorough investigation of R1 and R2 substituents of
N-methyl-N-(1-methylpiperidin-4-yl)benzenesulfonamide analogs
revealed a set of compounds with nanomolar activity to the 5-HT2C
receptor. In an effort to improve the selectivity against the 5-HT2A
receptor, three-dimensional structures of both subtypes were con-
structed using homology modeling. Molecular docking of synthe-
sized compounds into receptor models and the resulting binding
modes guided us in understanding the structural determinants
for selectivity. These models also provided us a clue for modifying
the structures to improve selectivity. Further modification and
optimization of the current scaffold as 5-HT2C agonists will be pub-
lished in the near future.
12. Biological assays: Radioligands [3H]ketanserin and [3H]imipramine were
purchased from PerkinElmer (PerkinElmer Life and Analytical Sciences,
Boston, USA), and [3H] mesulergine was obtained from Amersham
Biosciences (Buckinghamshire, UK). Cloned human recombinant serotonin 5-
HT2A and 5-HT2C receptors were obtained from Euroscreen (Brussels, Belgium).
Competition binding assays at the serotonin 5-HT2A receptor were performed
using 1 nM [3H]ketanserin by the protocol provided by the supplier of CHO-K1
membranes (Euroscreen, Brussels, Belgium) with minor modifications. Briefly,
Supplementary data
receptor membranes (15 lg/well) were incubated at 25 °C for 60 min in a final
volume of 0.25 mL reaction mixture containing [3H]ketanserin and various
concentrations of the drug in 50 mM Tris–HCl (pH 7.4) buffer containing 5 mM
CaCl2, 0.1% ascorbic acid and 10 lg/mL saponin. Then, the incubations were
terminated by rapid filtration using an Innotech cell harvester (Innotech
Biosystems, Switzerland) through Whatman GF/C glass fiber filter presoaked in
0.05% Brij. The filter was covered with MeltiLex, sealed in a sample bag
followed by drying in a microwave oven, and counted using MicroBeta Plus
(Wallac, Finland). Non-specific binding was determined in the presence of
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2011.11.001.
References and notes
1. Hoyer, D.; Martin, G. Neuropharmacology 1997, 36, 419.
2. Roth, B. L. Ann. Clin. Psychiatry 1994, 1994, 12.
mianserin (0.5 lM). Competition binding studies were carried out with 7–8
varied concentrations of the test compounds run in duplicate tubes, and
isotherms from three assays were calculated by computerized non-linear
regression analysis (GraphPad Prism Program, San Diego, USA) to yield IC50
values. Frozen membranes from stable CHO-K1 cell line expressing the human
recombinant 5-HT2C receptor were used. For the binding assay,
3. Tecott, L. H.; Sun, L. M.; Akana, S. F.; Strack, A. M.; Lowenstein, D. H.; Dallman,
M. F.; Julius, D. Nature 1995, 374, 542.
4. Nonogaki, K.; Strack, A. M.; Dallman, M. F.; Tecott, L. H. Nat. Med. 1998, 4, 1152.
5. Tecott, L. H.; Abdallah, L. CNS Spectr. 2003, 8, 5.
6. Sargent, P. A.; Sharpley, A. L.; Williams, C.; Goodall, E. M.; Cowen, P. J.
Psychopharmacology 1997, 133, 309.
7. Halford, J.; Harrold, J.; Lawton, C.; Blundell, J. Curr. Drug Targets 2005, 6, 13.
8. Appel, N. M.; Mitchell, W. M.; Garlick, R. K.; Glennon, R. A.; Teitler, M.; De
Souza, E. B. J. Pharmacol. Exp. Ther. 1990, 255, 843.
[3H]mesulergine (1 nM), receptor membrane (4
were added into 50 mM Tris–HCl (pH 7.7) buffer containing 0.1% ascorbic acid
and 10 pargyline. Non-specific binding was determined using 0.5 lM
lg/well) and test compounds
lM
mianserin. The incubations were performed for 30 min at 37 °C, then
terminated by rapid filtration through Whatman GF/C glass fiber filters
presoaked in 1% BSA.
9. Nichols, D. E. Pharmacol. Ther. 2004, 101, 131.
10. Borch, R. F. Org. Synth. 1972, 52.
11. General procedure for preparation of 4a–y, N-methyl-N-(1-methylpiperidin-4-
13. Baldwin, J. M.; Schertler, G. F. X.; Unger, V. M. J. Mol. Biol. 1997, 272, 144.
14. Ballesteros, J. A.; Weinstein, H. In Methods in Neurosciences; Stuart, C. S., Ed.;
Academic Press, 1995; Vol. 25, p 366.
yl)-p-butoxybenzenesulfonamide (4d): To
a
solution of methylamine
hydrochloride (17.7 g, 262 mmol) in methanol (150 mL), potassium
hydroxide (4.00 g, 71.3 mmol) was added and the mixture was stirred until
the potassium hydroxide was completely dissolved. 1-Methyl-4-piperidone
(22.6 g, 200 mmol) was then added and the reaction mixture was stirred for
15 min. Next, a solution of sodium cyanoborohydride (4.74 g, 75.4 mmol) in
methanol (30 mL) was added dropwise over 30 min and the reaction mixture
was stirred for an additional 30 min. Potassium hydroxide (15.0 g, 267 mmol)
was added, the mixture was stirred for 30 min, and the resulting precipitate
was filtered off. The filtrate was concentrated to 80 mL in vacuo at 30 °C and
water (10 mL) and brine solution (25 mL) were added. The separated organic
layer was saved and the aqueous layer was extracted twice with diethyl ether
(50 mL each). The combined organic layer was extracted three times with cold
6 M hydrochloric acid (20 mL each), and then potassium hydroxide was slowly
added to the acidic aqueous layer at 0 °C until the solution reached pH 12. This
solution was extracted three times with diethyl ether (40 mL each). The
organic layer was dried with anhydrous magnesium sulfate, filtered, and
concentrated in vacuo at 30 °C. The resulting oil was purified by Kugelrohr
distillation at 53 °C at 10 mmHg to give 11.5 g (45%) of methyl-(1-
methylpiperidin-4-yl)amine as a clear oil. To a solution of obtained methyl-
15. Reynolds, K.; Katritch, V.; Abagyan, R. J. Comput. Aided Mol. Des. 2009, 23, 273.
16. Sali, A.; Blundell, T. L. J. Mol. Biol. 1993, 234, 779.
17. Shortle, D.; Simons, K. T.; Baker, D. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 11158.
18. Ponder, J. W.; Richards, F. M. J. Mol. Biol. 1987, 193, 775.
19. Brooks, B. R.; Bruccoleri, R. E.; Olafson, B. D.; States, D. J.; Swaminathan, S.;
Karplus, M. J. Comput. Chem. 1983, 4, 187.
20. Almaula, N.; Ebersole, B. J.; Ballesteros, J. A.; Weinstein, H.; Sealfon, S. C. Mol.
Pharmacol. 1996, 50, 34.
21. Shapiro, D. A.; Kristiansen, K.; Kroeze, W. K.; Roth, B. L. Mol. Pharmacol. 2000,
58, 877.
22. Manivet, P.; Schneider, B. t.; Smith, J. C.; Choi, D.-S.; Maroteaux, L.; Kellermann,
O.; Launay, J.-M. J. Biol. Chem. 2002, 277, 17170.
23. Laskowski, R. A.; MacArthur, M. W.; Moss, D. S.; Thornton, J. M. J. Appl.
Crystallogr. 1993, 26, 283.
24. Gregoret, L. M.; Cohen, F. E. J. Mol. Biol. 1990, 211, 959.
25. Jones, G.; Willett, P.; Glen, R. C.; Leach, A. R.; Taylor, R. J. Mol. Biol. 1997, 267,
727.