Design, synthesis, and pharmacological characterization of N - And O-substituted 5,6,7,8-tetrahydro-4H-isoxazolo[4,5-d]azepin-3-ol analogues: Novel 5-HT2A/5-HT2C receptor agonists with pro-cognitive properties

Article abstract of DOI:10.1021/jm301656h

The isoxazol-3-one tautomer of the bicyclic isoxazole, 5,6,7,8-tetrahydro- 4H-isoxazolo[4,5-d]azepin-3-ol (THAZ), has previously been shown to be a weak GABA_A and glycine receptor antagonist. In the present study, the potential in this scaffold

Full text of DOI:10.1021/jm301656h

Article
pubs.acs.org/jmc
Design, Synthesis, and Pharmacological Characterization of N- and
O‑Substituted 5,6,7,8-Tetrahydro‑4H‑isoxazolo[4,5‑d]azepin-3-ol
Analogues: Novel 5‑HT2A/5-HT2C Receptor Agonists with Pro-
Cognitive Properties
Anders A. Jensen,*,† Niels Plath,‡ Martin H. F. Pedersen,§ Vignir Isberg,† Jacob Krall,†
Petrine Wellendorph,†,∥ Tine B. Stensbøl,‡ David E. Gloriam,† Povl Krogsgaard-Larsen,†
and Bente Frølund†
†Department of Drug Design and Pharmacology, Faculty of Health and Medical Sciences, University of Copenhagen,
Universitetsparken 2, Copenhagen, Denmark
‡H. Lundbeck A/S, Valby, Denmark
§Hevesy Laboratory, Radiation Research Division, Risø, Technical University of Denmark, Roskilde, Denmark
∥Neurobiology Research Unit and Center for Integrated Molecular Brain Imaging (Cimbi), Rigshospitalet and University of
Copenhagen, Copenhagen, Denmark
S
* Supporting Information
ABSTRACT: The isoxazol-3-one tautomer of the bicyclic isoxazole, 5,6,7,8-tetrahydro-4H-isoxazolo-
[4,5-d]azepin-3-ol (THAZ), has previously been shown to be a weak GABAA and glycine receptor
antagonist. In the present study, the potential in this scaffold has been explored through the synthesis
and pharmacological characterization of a series of N- and O-substituted THAZ analogues. The
analogues N-Bn-THAZ (3d) and O-Bn-THAZ (4d) were found to be potent agonists of the human 5-
HT2A and 5-HT2C receptors. Judging from an elaborate pharmacological profiling at numerous other
CNS targets, the 3d analogue appears to be selective for the two receptors. Administration of 3d
substantially improved the cognitive performance of mice in a place recognition Y-maze model, an effect
fully reversible by coadministration of the selective 5-HT2C antagonist SB242084. In conclusion, as novel
bioavailable cognitive enhancers that most likely mediate their effects through 5-HT2A and/or 5-HT2C
receptors, the isoxazoles 3d and 4d constitute interesting leads for further medicinal chemistry
development.
of the 5-HT2A, 5-HT2B, and 5-HT2C subtypes, which are Gq/11
-
INTRODUCTION
■
protein coupled receptors functionally linked to activation of
phospholipase C and the resulting formation of the second
messengers 1,4,5-trisphosphate and diacylglycerol and mobi-
lization of intracellular calcium, as well as to other intracellular
signaling cascades.14,15 The 5-HT2A subtype is heterogeneously
expressed in the CNS, and the receptor constitutes a major drug
target in psychiatric diseases and is believed to be the main
mediator of the psychotropic and psychotomimetic effects of
natural product compounds like lysergic acid diethylamide
(LSD) and psilocybin as well as several synthetic drugs.1,16,17 The
5-HT2B receptor is widely expressed in the periphery, where it for
example is involved in regulation of gastrointestinal motility and
bone metabolism, and the receptor has been linked with vascular
pathologies such as cardiac hypertrophy and pulmonary
hypertension.18,19 Finally, the 5-HT2C receptor is almost
exclusively found in the CNS where it is even more abundantly
expressed than 5-HT2A. 5-HT2C receptors are predominantly
found postsynaptically to serotonergic terminals, where they act
The neurotransmitter serotonin (5-hydroxytryptamine, 5-HT) is
widely distributed throughout the central nervous system (CNS)
and the peripheral nervous system.1 In the periphery, 5-HT
regulates a broad spectrum of functions in the cardiovascular,
gastrointestinal, endocrine, pulmonary, and genitourinary
systems.1−4 In the CNS, 5-HT plays key roles in processes
involved in mood, libido, aggression, anxiety, cognition, sleep,
appetite sensation, as well as pain perception and processing.
During the last five decades, serotonergic ligands have proven
effective drugs in the treatment of migraine, pain, obesity, and a
wide range of psychiatric and neurological disorders, and
intervention in central serotonergic neurotransmission continues
to be intensively pursued in the search for new therapeutics
against these indications.1,5−9
5-HT mediates its physiological effects through six classes of 7-
transmembrane (7-TM) receptors (5-HT1, 5-HT2, 5-HT4, 5-
HT5, 5-HT6, and 5-HT7) containing a total of 13 subtypes10−12
and through a class of cation-selective ligand-gated ion channels,
the 5-HT3 receptors.13 This heterogeneity is further increased by
the fact that several of the receptors are subjected to splice
variation or RNA editing.10−12 The 5-HT2 receptor class consists
Received: November 9, 2012
Published: January 9, 2013
© 2013 American Chemical Society
1211
dx.doi.org/10.1021/jm301656h | J. Med. Chem. 2013, 56, 1211−1227

Journal of Medicinal Chemistry
Article
as somatodendritic heteroreceptors on glutamatergic, dopami-
nergic, and GABAergic neurons and interneurons in several
regions.14,15,20−22 These modulatory role held by the receptor for
the activities in these three neurotransmitter systems have
prompted investigations of 5-HT2C receptors as putative targets
for the treatment of a wide range of neurological and psychiatric
disorders.22−26
tetrahydro-4H-isoxazolo[4,5-c]azepin-3-ol (THAO) analogues
(1b, 2b, 3a−i, 4a,b,d,i, 5a−c, 6a,c, 7c) is illustrated in Schemes 1
a
Scheme 1
In the present study, we report the discovery of a novel class of
subtype-selective and potent 5-HT receptor agonists based on
the 5,6,7,8-tetrahydro-4H-isoxazolo[4,5-d]azepin-3-ol (THAZ)
scaffold (Figure 1). THAZ is a seven-membered ring analogue of
a
Reagents and conditions: (a) K2CO3, R1X, acetone; (b) 33% HBr/
AcOH; (c) KOH/MeOH; (d) aq HCO2H, HCHO.
a
Scheme 2
Figure 1. Chemical structures of the THIP, THAZ, THPO, and THAO
analogues investigated in this study.
the γ-aminobutyric acid type A (GABAA) receptor agonist
4,5,6,7-tetrahydroisoxazolo[5,4-c]pyridin-3-ol (THIP)27 (Figure
1), and in previous studies THAZ has been shown to be a weak
antagonist of native glycine and GABAA receptors.28−30 Our
interest in the THAZ scaffold in the present study arose from its
structural similarity as the isoxazole-3-one tautomer to the 5-HT
receptor agonist N,N-diethyl-1-methylpiperidine-4-carboxa-
mide31 and the tetrahydropyridine carboxamide part of LSD
(Figure 1). We have synthesized a series of N- and O-substituted
THAZ analogues and identified a couple of relatively potent
agonists of the 5-HT2A and 5-HT2C receptors. One of these, N-
Bn-THAZ (3d), is shown to be a selective 5-HT2A/5-HT2C
agonist displaying negligible activity at other 5-HT receptors and
at a plethora of other CNS targets. In the search for other
putative CNS targets for 3d the binding distribution of an 125I-
labeled analogue of the compound in rat brain tissue has been
mapped, and the therapeutic potential in 3d has been
investigated in animal tests for cognitive function and
schizophrenia.
a
Reagents and conditions: (a) K2CO3, R1X, acetone; (b) 33% HBr/
AcOH or HCl/AcOEt; (c) KOH/MeOH.
and 2 and follow similar strategy. N-Methoxycarbonyl protected
THIP (8),32 THAZ (9),28 THPO (14),32 and THAO (15)28
were alkylated using the approriate alkyl or alkylaryl halide in the
presence of potassium carbonate. The alkylation step gave rise to
a mixture of the N-alkylated (10b, 12a−g, 16a−c) and O-
alkylated (11b, 13a−g, 17a−c) isomers, which were separated
using column chromatography. Deprotection was performed
using acidic or basic conditions. Reductive amination using
aqueous formic acid and formaldehyde led to the N-methylated
analogues 3h,i and 4h,i.
The synthesis of the stannyl precusor for the radioligand [125I]
N-Bn-THAZ) (21), was performed using the N-protected p-I-
Bn-THAZ analogue (12g), synthesized as described (Scheme 3).
Compound 12g was treated with hexabutyldistannane in the
RESULTS
■
Chemistry. The synthesis of the THAZ, THIP, 4,5,6,7-
tetrahydroisoxazolo[4,5-c]pyridin-3-ol (THPO), and 5,6,7,8-
1212
dx.doi.org/10.1021/jm301656h | J. Med. Chem. 2013, 56, 1211−1227

Journal of Medicinal Chemistry
Article
a
Scheme 3
a
Reagents and conditions: (a) Pd(PPh3)4, (SnBu3)2, toluene, reflux 22 h; (b) NaI, chloramine-T in a mixture of 94% EtOAc, 4% DMF, 1% AcOH,
1% H2O, room temp, 30 min, Na2S3O5; (c) 44.0 MBq Na125I, chloramine-T in a mixture of 94% EtOAc, 4% DMF, 1% AcOH, 1% H2O, room
temperature, 30 min; (d) 33% HBr/AcOH, room temperature 5 h.
presence of tetrakis(triphenylphosphine)Pd(0) in refluxing
toluene to furnish the stannylated derivative 19. To test the
efficiency of the iodination reaction, 19 was initially converted
into the cold ligand 12g by reaction with NaI under oxidative
conditions with chloramine-T in a mixture of ethyl acetate,
dimethylformamide, acetic acid, and water, followed by
deprotection with hydrobromide in acetic acid. As this strategy
proved functional, the radioligand was generated in a similar
fashion. Thus, radioiodination was carried out by treatment with
Na125I using the chloramine-T method followed by deprotection
with hydrobromide in acetic acid. The radioligand 21 was
purified by HPLC and analyzed to have a radiochemical purity of
100% and a total activity of 40.3 MBq (yield: 91% based on
Na125I).
Functional Characterization of THAZ, THIP, THPO, and
THAO Analogues at 5-HT2A and 5-HT2C in the Fluo-4/Ca2+
Assay. Preliminary testing suggested that some of the
synthesized THAZ analogues possessed agonist activity at 5-
HT2 receptors (data not shown). To characterize the functional
properties of the compounds in greater detail, two polyclonal
HEK293 cell lines stably expressing the human 5-HT2A and 5-
HT2C receptors were constructed. To validate the pharmaco-
logical properties exhibited by the receptors expressed in these
cells, 10 reference 5-HT2 receptor ligands were characterized
functionally at the cell lines in the fluorescence-based Fluo-4/
Ca2+ assay that measures the release of Ca2+ from intracellular
stores upon activation of the Gq/11-coupled receptors.14,15 The
rank orders of agonist potencies and to some extent also the
absolute EC50 values displayed by 5-HT, PNU-22394, Ro 60-
0175, CP 809101, MK-212, and WAY 161503 at the receptors in
this assay were in good agreement with findings from previous
studies (Table 1 and Figure 2A).33−36 Also, in concordance with
previous studies, the five selective 5-HT2 agonists were found to
be full agonists or high-efficacy partial agonists at both subtypes,
displaying maximal responses similar to or slightly lower than
those of 5-HT at the two receptors.33,34,36−38 The Ki values
obtained for four reference 5-HT2 receptor antagonists at the
receptors also correlated well with those obtained in previous
studies.33,38−42 Both mianserin and asenapine were found to be
equipotent antagonists at 5-HT2A and 5-HT2C, clozapine
displayed a preference (∼8-fold) for 5-HT2A over 5-HT2C, and
MDL 11,939 was a highly selective 5-HT2A antagonist exhibiting
a ∼4000-fold lower Ki at this receptor than at 5-HT2C (Table 1).
All agonists and antagonists displayed Hill slopes close to unity
(nH ∼ 1) at the two receptors (data not shown).
When characterized in the Fluo-4/Ca2+ assay the majority of
the 12 THAZ analogues (THAZ, 3a−d, 3h−i, 4a−b, 4d, 4h−i)
were found either to be weak agonists at the 5-HT2A and 5-HT2C
receptors or to be completely inactive as agonists and antagonists
at concentrations up to 1 mM (Table 1 and Figure 2A).
However, notable exceptions were the N- and O-benzylated
THAZ analogues 3d and 4d, which were potent agonists
displaying high nanomolar EC50 values at both receptors.
Whereas 3d was a high-efficacious partial agonist at both
subtypes, 4d was a pronounced partial agonist at 5-HT2A,
eliciting a maximal response of 59% of that of 5-HT at the
receptor (Table 1 and Figure 2A). The 5-HT2A and 5-HT2C
activity of the THAZ analogues prompted us to test series of N-
and O-substituted analogues of the structurally related scaffolds
THIP, THPO, and THAO at the receptors. However, none of
these compounds (1b, 2b, 5a−c, 6a, 6c, 7c) displayed substantial
pharmacological activity at the receptors (Table 1).
Functional Characterization of THAZ Analogues 3d and 4d
at 5-HT2A and 5-HT2C in the IP-One Assay. The agonist activity
displayed by 3d and 4d at the 5-HT2A and 5-HT2C receptors in
the Fluo-4/Ca2+ assay was verified in the IP-One assay, in which
the formation and accumulation of the second messenger
inositol monophosphate (IP1) upon activation of the receptors is
measured.43 The agonist potencies displayed by 3d and 4d in the
IP-One assay were slightly lower (2−5-fold) than those exhibited
by the agonists in the Fluo-4/Ca2+ assay (Tables 1 and 2).
However, the same was true for the reference agonists CP
809101 and MK-212 and even more pronounced differences in
agonist potencies between the two assays were observed for 5-
HT and PNU 22394 (Tables 1 and 2). The rank orders of
potencies displayed by the six agonists at 5-HT2A were 5-HT >
1213
dx.doi.org/10.1021/jm301656h | J. Med. Chem. 2013, 56, 1211−1227

Journal of Medicinal Chemistry
Article
Table 1. Functional Properties of Reference Ligands and THAZ, THIP, THPO, and THAO Analogues at Polyclonal h5-HT2A-
a
HEK293 and h5-HT2C-HEK293 Cell Lines in the Fluo-4/Ca2+ Assay
h5-HT2A
h5-HT2C
EC50 [pEC50 SEM]
Rmax SEM
EC50 [pEC50 SEM]
Rmax SEM
5-HT
0.0026 [8.59 0.06]
0.023 [7.63 0.12]
0.027 [7.57 0.10]
0.076 [7.12 0.17]
0.42 [6.37 0.04]
0.012 [7.93 0.05]
100
0.00087 [9.06 0.08]
0.0077 [8.11 0.13]
0.0047 [8.33 0.13]
0.0019 [8.72 0.18]
0.028 [7.55 0.11]
0.0073 [8.13 0.12]
100
PNU 22394
Ro 60−0175
CP 809101
MK-212
102
9
101
5
117 12
111 10
88
97
6
7
103
93
8
8
WAY 161503
110 12
95 10
Ki [pKi SEM]
Ki [pKi SEM]
mianserin
0.0081 [8.09 0.13]
0.0091 [8.04 0.09]
0.00091 [9.04 0.17]
0.0025 [8.61 0.09]
0.0080 [8.09 0.14]
0.073 [7.14 0.15]
0.0025 [8.61 0.09]
∼10 [∼5.0]
clozapine
asenapine
MDL 11,939
EC50 [pEC50 SEM]
Rmax SEM
EC50 [pEC50 SEM]
Rmax SEM
THAZ analogues
b
b
THAZ
>1000 [<3.0]
>1000 [<3.0]
c
c
3a
weak agonist
weak agonist
d
d
3b
agonist
agonist
d
d
3c
agonist
agonist
3d
0.55 [6.26 0.04]
80 6.5
0.42 [6.38 0.08]
90 5.6
b
b
3h
>1000 [<3.0]
>1000 [<3.0]
b
b
3i
>1000 [<3.0]
>1000 [<3.0]
d
4a
agonist
12 [4.92 0.08]
3.0 [5.53 0.12]
0.10 [6.99 0.13]
80 4.1
67 5.6
84 2.7
4b
7.9 [5.10 0.08]
0.24 [6.61 0.10]
42 3.1
59 5.8
4d
b
c
4h
>1000 [<3.0]
weak agonist
b
b
4i
>1000 [<3.0]
>1000 [<3.0]
THIP analogues
b
b
1b
>1000 [<3.0]
>1000 [<3.0]
b
b
2b
>1000 [<3.0]
>1000 [<3.0]
THPO analogues
b
b
5a
>1000 [<3.0]
>1000 [<3.0]
b
b
5b
>1000 [<3.0]
>1000 [<3.0]
b
b
5c
>1000 [<3.0]
>1000 [<3.0]
d
d
6a
agonist
agonist
c
c
6c
weak agonist
weak agonist
THAO analogue
7c
b
b
>1000 [<3.0]
>1000 [<3.0]
a
The EC50 values are given in μM with pEC50 SEM values in brackets, and Rmax SEM values are given in % of the Rmax obtained for 5-HT at the
same receptor at the same plate. In the antagonist experiments, EC75−EC85 concentrations of 5-HT were used. Data are the means of 4−10
b
c
individual experiments performed in duplicate. No significant agonist or antagonist activity at 1 mM. Weak agonist response at concentrations of
d
100−1000 μM. The concentration−response curve of the agonist was not complete at concentrations up to 1000 μM.
PNU 22394 > CP 809101 > 4d > MK-212 ∼ 3d in the Fluo-4/
Ca2+ assay and CP 809101−5-HT ∼ PNU 22394 > MK-212 ∼
4d > 3d in IP-One assay. At the 5-HT2C receptor, the potency
rank orders were 5-HT > CP 809101 > PNU 22394 > MK-212 >
4d > 3d in the Fluo-4/Ca2+ assay and CP 809101 > 5-HT > PNU
22394 ∼ MK-212 ∼ 4d > 3d in the IP-One assay. Thus, the rank
orders of agonist potencies at the two receptors were fairly
similar in the two assays, the major difference being the relatively
higher potencies of CP 809101 in the IP-One assay compared to
the Fluo-4/Ca2+ assay. The differences between the absolute
EC50 values measured for the agonists are thus likely to arise from
the inherent differences between the two assays, the one
measuring transient increases in intracellular Ca2+ concentrations
and the other accumulation of IP1 in the cells over a longer period
of time (see Experimental Section). In agreement with its profile
in the Fluo-4/Ca2+ assay, 4d displayed partial agonism at 5-HT2A
in the IP-One assay, whereas the maximal responses elicited by
4d at 5-HT2C and by 3d at both 5-HT2A and 5-HT2C were not
significantly different from those of 5-HT at the respective
receptors (Figure 2B and Table 2).
Binding Properties of the THAZ Analogues at the 5-HT2A
and 5-HT2C Receptors. The binding affinities of the THAZ
analogues at 5-HT2A and 5-HT2C were determined in [3H]-
ketanserin and [3H]mesulergine binding assays to membranes
from tsA201 cells transiently expressing the two receptors. The
KD values determined for [3H]ketanserin at 5-HT2A and for
[3H]mesulergine at 5-HT2C in saturation binding experiments
were in good agreement with previously reported values, as were
the Ki values determined for the antagonists mianserin and
clozapine (Table 3).44−46 The Ki values obtained for the
reference agonists 5-HT, CP 809101, and MK-212 at the two
receptors were considerably higher than their respective EC50
values in the Fluo-4/Ca2+ and IP-One assays (Tables 1−3). The
low binding affinities exhibited by the agonists are likely to arise
1214
dx.doi.org/10.1021/jm301656h | J. Med. Chem. 2013, 56, 1211−1227

Journal of Medicinal Chemistry
Article
Figure 2. Functional properties of THAZ analogues at human 5-HT2A and 5-HT2C receptors. (A) Concentration−response curves of selected reference
ligands and THAZ analogues at the polyclonal 5-HT2A- and 5-HT2C-HEK293 cell lines in the Fluo-4/Ca2+ assay. Error bars are omitted for reasons of
clarity (B) Concentration−response curves of 5-HT, PNU-22394, CP 809101, MK-212, and THAZ analogues 3d and 4d at the polyclonal 5-HT2A- and
5-HT2C-HEK293 cell lines in the IP-One assay. Data in B are given as mean SEM values. The figures in A and B depict representative experiments
performed in duplicate (Fluo-4/Ca2+) or triplicate (IP-One).
Table 2. Functional Properties of Four Reference Agonists and THAZ Analogues 3d and 4d at Polyclonal h5-HT2A-HEK293 and
a
h5-HT2C-HEK293 Cell Lines in the IP-One Assay
h5-HT2A
EC50 [pEC50 SEM]
h5-HT2C
EC50 [pEC50 SEM]
Rmax SEM
Rmax SEM
5-HT
0.15 [6.83 0.10]
0.21 [6.68 0.08]
0.14 [6.86 0.06]
1.2 [5.92 0.09]
2.4 [5.62 0.09]
1.3 [5.89 0.06]
100
0.044 [7.36 0.11]
0.12 [6.91 0.06]
0.0081 [8.09 0.11]
0.16 [6.79 0.10]
1.7 [5.76 0.08]
0.23 [6.64 0.05]
100
PNU 22394
CP 809101
MK-212
3d
98
90
5
2
97
5
109
93
7
101
95
4
5
7
7
2
92
4d
77
105
6
a
The EC50 values are given in μM with pEC50 SEM values in brackets, and Rmax SEM values are given in % of the Rmax obtained for 5-HT at the
same receptor at the same plate. Data are the means of 4−7 individual experiments performed in triplicate.
from the use of 5-HT2A and 5-HT2C antagonists as radioligands
in the assays, since previous studies of 5-HT2A47 and other 7TM
receptors48−51 have found binding affinities of agonists to be
highly dependent on the intrinsic activity of the radioligand used.
Analogously to the reference agonists, the Ki values
determined for the THAZ analogues at 5-HT2A and 5-HT2C in
the [3H]ketanserin and [3H]mesulergine binding assays were
markedly higher than their functional EC50 values. However, a
reasonable correlation existed between the binding affinities
determined for the analogues and the agonist potencies displayed
by them in the Fluo-4/Ca2+ assay (Tables 1 and 3). Whereas 3d
and 4d displayed low micromolar binding affinities to 5-HT2A
and 5-HT2C, the other analogues displayed substantially higher
Ki values at the receptors (Table 3).
Docking of Selected THAZ Analogues in a Homology
Model of the 5-HT2A Receptor. The THAZ analogues 3d, 4b, 4d,
and 4i were docked in our previously published homology model
of the 5-HT2A receptor (Figure 3).52 In the following, residues in
the 5-HT2A receptor will be presented with their respective
numbers in the amino acid sequence of 5-HT2A as well as with
their numbering according to the Ballesteros−Weinstein
nomenclature.53 In the case of both 3d and 4d, the charged
amino group of the THAZ analogue forms a salt bridge with the
D1553.32 residue in 5-HT2A, the exocyclic oxygen in 3d and 4d
serves as an acceptor in a hydrogen bond interaction with
S2395.43, and the O- or N-benzyl group on the isoxazole ring
occupies a hydrophobic region between the transmembrane
helices 5 and 6 formed by three aromatic residues (F2435.47
,
1215
dx.doi.org/10.1021/jm301656h | J. Med. Chem. 2013, 56, 1211−1227

Journal of Medicinal Chemistry
Article
opioid and neurokinin receptors. The compound was also found
to be inactive at the monoamine transporters, at miscellaneous
ligand-gated and voltage-gated ion channels, and at numerous
enzymes including phosphodiesterases and kinases (Supporting
Information Table 1).
Table 3. Binding Affinities of Reference Ligands and THAZ
Analogues in the [3H]Ketanserin and [3H]Mesulergine
Competition Binding Assays Using Membranes from tsA201
Cells Transiently Expressing Human 5-HT2A and 5-HT2C
a
Receptors, Respectively
Development of [125I]-N-(p-I-Bn)-THAZ and Preliminary
Characterization of Its Binding to Rat Brain Homogenate and
Slices. In an attempt to elucidate the distribution of N-Bn-THAZ
(3d) binding sites in rat CNS and thus enable the identification
of other putative targets for the compound, a [125I]-labeled
analogue of the compound was developed. The three
monoiodinated N-Bn-THAZ analogues 3e, 3f, and 3g were
synthesized, and their binding affinities to membranes from 5-
HT2C-expressing tsA201 cells were determined in the [3H]-
mesulergine binding assay. Whereas 3e and 3f exhibited similar
binding affinities to 5-HT2C as 3d, the N-(p-I-Bn)-THAZ
analogues (3g) displayed ∼7-fold higher binding affinity, and
therefore this analogue was chosen as the candidate for
radioligand development (Figure 4A).
h5-HT2A
h5-HT2C
radioligands
KD SEM
KD SEM
[3H]ketanserin
[3H]mesulergine
2.9 0.9 nM
nt
nt
1.4 0.8 nM
h5-HT2C
h5-HT2A
reference ligands
Ki [pKi SEM]
Ki [pKi SEM]
5-HT
0.32 [6.49 0.04]
0.081 [7.09 0.04]
21 [4.68 0.04]
0.027 [7.57 0.04]
0.036 [7.44 0.05]
h5-HT2A
0.079 [7.10 0.04]
0.0093 [8.03 0.03]
3.4 [5.47 0.08]
0.0017 [8.76 0.03]
0.0093 [8.03 0.04]
h5-HT2C
CP 809101
MK-212
mianserin
clozapine
THAZ analogues
Ki [pKi SEM]
Ki [pKi SEM]
Prior to experiments with rat brain tissue or slices, the binding
properties of [125I]-N-(p-I-Bn)-THAZ to membranes from
human 5-HT2C receptor-expressing tsA201 cells were inves-
tigated (Figure 4B). Because of the low molar concentration of
the [125I]-N-(p-I-Bn)-THAZ stock solution, saturation binding
experiments could not be performed. However, when using a
concentration of 6 nM [125I]-N-(p-I-Bn)-THAZ (a ∼50-fold
lower concentration than the Ki value of 3g in the [3H]-
mesulergine binding assay), the radioligand displayed specific
binding to the 5-HT2C-expressing tsA201 cell membranes when
100 μM N-Bn-THAZ (3d) or 10 μM mianserin were used as cold
ligands to determine nonspecific binding (Figure 4B).
THAZ
3a
>300 [<3.5]
>300 [<3.5]
∼300 [∼3.5]
∼100 [∼4.0]
8.8 [5.06 0.07]
nt
>300 [<3.5]
>300 [<3.5]
3b
3c
∼300 [∼3.5]
∼100 [∼4.0]
2.3 [5.64 0.06]
7.2 [5.14 0.04]
4.8 [5.32 0.05]
0.31 [6.51 0.04]
>300 [<3.5]
3d
3e
3f
nt
3g
3h
3i
nt
>300 [<3.5]
>300 [<3.5]
∼300 [∼3.5]
∼100 [∼4.0]
3.3 [5.48 0.04]
>300 [<3.5]
>300 [<3.5]
>300 [<3.5]
4a
∼300 [∼3.5]
∼100 [∼4.0]
1.2 [5.92 0.05]
>300 [<3.5]
4b
4d
4h
4i
The ability of [125I]-N-(p-I-Bn)-THAZ to bind to membranes
prepared from four different rat brain regions were investigated.
In homogenates of cerebellum, hippocampus, or midbrain, no
specific [125I]-N-(p-I-Bn)-THAZ binding could be determined,
neither when using 5-HT, mianserin, and N-Bn-THAZ (3d) nor
N-(p-I-Bn)-THAZ (3g) as cold ligand for the determination of
the nonspecific binding (data not shown). In cerebral cortex
membranes, total levels of [125I]-N-(p-I-Bn)-THAZ binding
were reduced significantly in a concentration-dependent manner
by 3d and 3g, displaying pKi SEM values of 6.14 0.06 (n = 3)
and 7.99 0.06 (n = 3), respectively (Figure 4C). In contrast,
neither 5-HT, mianserin, nor the selective 5-HT2C receptor
antagonist SB242084 were able to displace [125I]-N-(p-I-Bn)-
THAZ binding when tested at concentrations up to 50 μM
(Figure 4C and data not shown). Furthermore, total levels of
>300 [<3.5]
a
The Ki values are given in μM with pKi SEM values in brackets.
Data represents the means of 3−4 individual experiments performed
in duplicate. nt, not tested.
W3366.48, and F3406.52) and three aliphatic residues (V2355.39
,
L2365.40,and I3446.56) (Figure 3A). The insight into potential
reasons for the functional implications arising from methylation
of the amino group (Figure 3B) and from different substitutions
introduced on the isoxazole ring in the THAZ analogue (Figure
3C) gained from dockings of 4b, 4i, and 4d into the model will be
addressed in the Discussion section.
Selectivity Profile of N-Bn-THAZ at Numerous Putative CNS
Targets. The selectivity profile of N-Bn-THAZ (3d) was
determined in a broad screening of the ligand at a considerable
number of neurotransmitter receptors and transporters as well as
at other CNS targets. As can be seen from Supporting
Information Table 1, the compound appeared to be quite
selective for 5-HT2A and 5-HT2C. 3d was inactive at the third 5-
HT2 receptor, 5-HT2B, when tested as an agonist, an antagonist,
and as a modulator, and displayed no significant binding to and/
or functional activity at 5-HT1A, 5-HT1B, 5-HT1D, 5-HT3A, 5-
HT4, 5-HT6, or 5-HT7 receptors at concentrations up to 10 μM
(Supporting Information Table 1). Furthermore, when tested at
a concentrations of 10 μM in binding assays and/or functional
assays, 3d displayed no or negligible activity at a wide range of
other 7-TM receptors, including all human dopamine,
norepinephrine, histamine, muscarinic, acetylcholine, adenosine,
and melatonin receptors or various peptide receptors such as
[
125I]-N-(p-I-Bn)-THAZ binding to cerebral cortex membranes
were found not to be significantly reduced by 300 μM dopamine,
300 μM norepinephrine, 300 μM acetylcholine, 300 μM GABA,
300 μM (S)-glutamate, or 1 mM glycine (data not shown).
Autoradiography experiments were performed using 6 nM
[
125I]-N-(p-I-Bn)-THAZ binding to frontal levels of cortex and
hippocampus/midbrain. Because 5-HT, mianserin, and
SB242084 were unable to displace [125I]-N-(p-I-Bn)-THAZ
binding to cerebral cortex membranes to any significant degree
(Figure 4C), nonspecific binding was determined in the presence
of 100 μM N-Bn-THAZ. As shown in Figure 4D, total binding of
[
125I]-N-(p-I-Bn)-THAZ could be displaced by excess cold
ligand at cortical and midbrain levels indicative of specific
binding in these regions, whereas binding was not convincingly
reduced at hippocampal levels.
Effects of N-Bn-THAZ in Rodent Assays for Spatial Memory
and Psychosis. To examine the effect of N-Bn-THAZ (3d) on
1216
dx.doi.org/10.1021/jm301656h | J. Med. Chem. 2013, 56, 1211−1227

Journal of Medicinal Chemistry
Article
Figure 3. Docking of selected THAZ analogues into a homology model of the 5-HT2A receptor. Receptor residues are given with their numbering
according to the Ballesteros−Weinstein nomenclature. (A) Proposed binding modes of 3d (left) and 4d (right) in the orthosteric site of 5-HT2A. (B) 4b
(left) and 4i (right) docked into the orthosteric site of 5-HT2A: Methylation of the amino group in the THAZ analogue is detrimental for its binding to 5-
HT2A. (C) 4b (left) and 4d (right) docked into the orthosteric site of 5-HT2A: Large hydrophobic substituents on the isoxazole ring of THAZ give rise to
higher 5-HT2A affinities by interacting with a hydrophobic region between TM5 and TM6.
learning and memory, we tested the compound in the mouse
two-trial place recognition Y-maze, a cognition assay that
addresses episodic-like spatial memory formation. In this assay,
animals can freely explore two spatially distinct arms of a
transparent Y-maze during the acquisition trial and are then
exposed, after a flexible inter trial interval (ITI), to the previously
explored (“familiar”) and one unexplored arm (“novel”) for the
retrieval trial. The degree to which animals explore the novel over
the familiar environment during retrieval is interpreted as an
index for the spatial memory formed for the familiar arms.
We first monitored the ability of normal mice to discriminate
between novelty and familiarity as a function of increased ITI (10
had occurred at this time point. Thus, we applied the 24 h ITI to
subsequently test the pro-cognitive potential of N-Bn-THAZ.
Animals treated with 0.43 mg/kg N-Bn-THAZ 30 min prior to
the acquisition trial showed significantly increased exploration of
the novel environment as demonstrated by increased distance
traveled (novel 40.4 2.7%, familiar 29.8 1.4%; One-Way
ANOVA on ranks: p < 0.012), increased number of arm entries
(novel 38.5 3%, familiar 30.8 1.5%; F1,16 = 7.513; p < 0.05;
NK) and increased time spent (novel 38.8 3.5%, familiar 30.6
1.7%; F1,18 = 5.302; p < 0.05; NK) in the novel arm (Figure
5B−D). To test whether this effect is mediated by the 5-HT2C
receptor component of N-Bn-THAZ, we treated another group
of animals with the selective 5-HT2C receptor antagonist
SB242084 (2.1 mg/kg) in addition to 0.43 mg/kg N-Bn-
THAZ (Figure 5B−D). Co-treatment with SB242084 fully
reversed the pro-cognitive effect of N-Bn-THAZ with regard to
distance traveled (novel 34.7 2.4%, familiar 32.6 1.2%) and
number of arm entries (novel 32.8 3.2%, familiar 33.6 1.6%)
and also affected the time spent in the novel versus familiar
environment (novel 36.6 3.8%, familiar 31.7 1.9%) (Figure
̈
min, 1 h, 24 h). As shown in Figure 5A, drug naive mice showed a
clear preference for the novel environment 10 min and 1 h after
the acquisition, as demonstrated by a significant increase in the
distance they traveled in the novel arm during retention (10 min:
novel 44.9 11.3%, familiar 25.7 6.5%; p < 0.05; t-test. 1 h:
novel 49.1 9.3%, familiar 26.1 5.1%; p < 0.05; t-test). This
preference was absent 24 h after acquisition (novel 37 5.8%,
familiar 31.5 2.9%), indicating that a natural forgetting process
1217
dx.doi.org/10.1021/jm301656h | J. Med. Chem. 2013, 56, 1211−1227

Journal of Medicinal Chemistry
Article
Figure 4. Characterization of the binding properties of [125I]N-(p-I-Bn)-THAZ. (A) Concentration−inhibition curves for N-Bn-THAZ (3d) and its
iodinated analogues N-(o-I-Bn)-THAZ (3e), N-(m-I-Bn)-THAZ (3f), and N-(p-I-Bn)-THAZ (3g) in the [3H]mesulergine binding assay using
membranes from 5-HT2C-expressing tsA201 cell membranes. The figure depicts a representative individual experiment (n = 3), and error bars are
omitted for clarity. (B) Binding of 6 nM [125I]N-(p-I-Bn)-THAZ to membranes from 5-HT2C-expressing tsA201 cell membranes in the absence (buffer)
or presence of 100 μM N-Bn-THAZ (3d) or 10 μM mianserin. (C) Concentration−inhibition curves for N-Bn-THAZ (3d) and N-(p-I-Bn)-THAZ
(3g) in the [125I]N-(p-I-Bn)-THAZ binding assay using rat cerebral cortex membranes. 5-HT and SB242084 did not displace [125I]N-(p-I-Bn)-THAZ
(6 nM) binding to the membranes at concentrations up to 50 μM. The figure depicts a representative individual experiment. (D) Autoradiograms
showing distribution of [125I]N-(p-I-Bn)-THAZ binding sites in rat brain. Coronal sections displaying cortical and midbrain regions (top) and cortical
and hippocampal regions (bottom) were incubated in the presence of 6 nM [125I]N-(p-I-Bn)-THAZ for 40 min at room temperature, either in the
absence (total binding) or presence of 100 μM N-Bn-THAZ (3d) (nonspecific binding). The autoradiograms shown are representative slides for two
independent experiments performed in triplicate. The sections are represented in a pseudocolor scale that is based on gray scale values. Purple represents
the highest levels of binding, while yellow represents the lowest levels. A gray scale image of D is given in Supporting Information Figure 1.
5B−D). These data indicate that N-Bn-THAZ bears a pro-
cognitive potential on spatial memory formation by strengthen-
ing encoding and/or retrieval of spatial information over time.
A few studies have demonstrated efficacy of 5-HT2C receptor
agonism in animal models predictive of antipsychotic-like
efficacy.36,54−57 Thus, we tested N-Bn-THAZ in the rat
conditioned avoidance response (CAR) task. However, we did
not observe any effect of the compound in this task at
concentrations up to 0.85 mg/kg (data not shown).
analogues is a key pharmacophore element for their 5-HT2A/5-
HT2C activity. Interestingly, direct comparison of the pharmaco-
logical properties of the N-Me, N-Et, and N-Bn analogues 3a, 3b,
and 3d and those of the corresponding O-substituted analogues
4a, 4b, and 4d reveals a shared SAR pattern: both N- and O-
methylation of THAZ induces weak agonistic activity (3a and
4a), which is enhanced further with the introduction of an ethyl
group (3b and 4b) and quite dramatically increased in the N-Bn
and O-Bn analogues (3d and 4d). We propose that the massive
increases in agonist potencies of 3d and 4d (≥30-fold compared
to the corresponding N-Et and O-Et THAZ analogues) are likely
to arise from specific interactions of the benzyl ring with the
receptor rather than simply introduction of bulk in these
positions. The O-Bn analogue is a slightly more potent agonist
than the N-Bn analogue, a trend also observed for the methyl (3a
and 4a) and ethyl (3b and 4b) analogues (Table 1). However,
the spatial orientations of the phenyl rings of 3d and 4d in the
orthosteric site could be very similar, with this ring system in the
two analogues forming many of the same interactions with the
receptor. Thus, the benzyl groups in 3d and 4d appears to be of
key importance for 5-HT2A/5-HT2C activity, and judging from
DISCUSSION
■
In the present study, we have applied the structure of the weak
glycine and GABAA receptor antagonist THAZ28−30 as a scaffold
in the development of a novel class of selective and fairly potent
5-HT2A/5-HT2C receptor agonists.
Structure−Activity Relationships of the THAZ Ana-
logues at 5-HT2A and 5-HT2C Receptors. The functional
properties exhibited by the 12 THAZ analogues at the 5-HT2A
and 5-HT2C receptors convey interesting SAR information about
this agonist scaffold (Table 1). The complete inactivity of THAZ
itself demonstrates that the O- or N-substituent in the active
1218
dx.doi.org/10.1021/jm301656h | J. Med. Chem. 2013, 56, 1211−1227

Journal of Medicinal Chemistry
Article
Figure 5. Pro-cognitive effect of N-Bn-THAZ (3d) on episodic-like spatial memory. (A) Time dependent decay of spatial memory in the 2-trial place
recognition Y-maze. Exploration of novel and (pooled) familiar arms measured as distance traveled in the respective areas and shown as mean
̈
percentage of total distance traveled. Drug naive, saline treated animals show significant preference for the novel environment 10 min and 1 h, but not 24
h after acquisition. * p < 0.05 versus familiar arm (t test (n = 10 per group). (B−D) Effect of N-Bn-THAZ (3d) on spatial discrimination in the 2-trial
place recognition Y-maze. Drugs were applied prior to acquisition and memory retention assessed 24 h later. Data are presented as mean percentage of
distance traveled (B), number of arms entries (C) and time spent (D) in the novel (white bars) versus familiar (black arms) environment. The effects of
N-Bn-THAZ (3d) for each respective readout were blocked by coadministration of the 5-HT2C selective antagonist SB242084. * p < 0.05 versus familiar
arms (one-way ANOVA followed by Student−Newman−Keuls (NK) post hoc test) (n = 10 per group).
the 7-fold increased 5-HT2C binding affinity of 3g compared to
3d in the [3H]mesulergine binding assay (Figure 4A), it may be
possible to improve the binding affinity and potency in future
THAZ analogues through the introduction of different
substituents in the benzyl ring or by substitution of the phenyl
ring with other aromatic groups.
The presence of a secondary amino group in the THAZ
analogues appears to be another key pharmacophore element for
their 5-HT2A/5-HT2C activity. Although tertiary amino group
analogues of the more potent THAZ analogues 3d and 4d were
not included in the series, the detrimental effects of the
introduction of a methyl group at the amino function on
receptor activity are clear when comparing the functional
properties of 3b and 3i and those of 4b and 4i (Table 1).
The bicyclic isoxazol-3-one tautomeric structure of the THAZ
analogue appears to be essential for its 5-HT2A/5-HT2C activity,
as none of the N- and O-substituted THIP, THPO, or THAO
analogues exhibit significant activities at the receptors (Table 1).
We propose that analogues with these bicyclic scaffolds are
unable to accommodate a low energy conformation in which the
5-HT2A/5-HT2C pharmacophore elements, i.e., the amino group,
the exocyclic oxygen, and the isoxazole ring substituent, are
positioned as favorable for receptor binding as in the
corresponding THAZ analogues. Interestingly, the N-substituted
THAZ analogue shares some structural similarity with a series of
heterocycle-fused azepines recently reported to be potent 5-
HT2C receptor agonists.58 However, in contrast to N-Bn-THAZ,
most of these analogues, including the 2-benzyl-6,7,8,9-
tetrahydro-2H-pyridazino[3,4-d]azepin-3(5H)-one analogue
closely related to 3d, do not exhibit significant selectivity for 5-
HT2C over 5-HT2B.58 This further underlines the subtle
structural determinants for these bicyclic compounds in terms
of activity and subtype-selectivity profiles at the 5-HT2 receptors.
Proposed Binding Modes of the THAZ Analogues to
the 5-HT2A Receptor. To elucidate the binding modes of the
THAZ analogues to the 5-HT2A receptor, 3d, 4b, 4d, and 4i were
1219
dx.doi.org/10.1021/jm301656h | J. Med. Chem. 2013, 56, 1211−1227

Journal of Medicinal Chemistry
Article
docked in our previously published homology model of the 5-
HT2A receptor.52 In view of the lack of selectivity of the THAZ
analogues between 5-HT2A and 5-HT2C and the highly conserved
orthosteric binding sites of these two receptors, the binding
conformations of these analogues in 5-HT2C are likely to be very
similar to those in 5-HT2A. Interestingly, the spatial orientations
of the two potent agonists from the series in the binding site were
found to be somewhat different. Whereas 3d and 4d appear to
interact with the same receptor residues in their respective
binding modes to 5-HT2A and although the orientation of the
benzyl groups in the two analogues in the binding site overlap,
the orientations of THAZ moieties in the two agonists differ
substantially (Figure 3A). The difference in binding mode arises
because both analogues span between transmembrane helices 3
and 5, but the alkylbenzyls are located at different positions of the
isoxazole rings. The observed detrimental effects of N-
methylation of the amino group in the THAZ analogue on 5-
HT2A/5-HT2C activity seems to be rooted in a disruption of the
salt bridge between this group and D1553.32 in the receptor
(Figure 3B). Thus, we speculate that the spatial orientation of the
amino group in the THAZ analogues in the binding site differs
from those of the tertiary amino groups in N,N-diethyl-1-
methylpiperidine-4-carboxamide, LSD, and several other 5-HT2
agonists and perhaps is more similar to the orientations of this
group in the binding conformations of other 5-HT2 agonists with
protonated secondary amino groups, including 5-HT.
The projection of the O- or N-substituents on the isoxazole
ring of THAZ into the hydrophobic region formed by residues in
transmembrane helices 5 and 6 in the 5-HT2A receptor (Figure
3C) is similar to what has been proposed to contribute to the
binding of several other 5-HT2 ligands. Interactions between
hydrophobic ligand substituents and this cavity in the receptor
have been associated with agonist activity at 5-HT2A,59 whereas
large aromatic ligand substituents projecting into region has been
proposed to lead to antagonism of the receptor.60 From the
proposed binding mode of the THAZ analogue to 5-HT2A, it
seems feasible that this cavity can accommodate the binding of
fairly large, hydrophobic substituents (Figure 3C). Furthermore,
introduction of a benzyl group in this region introduces more
hydrophobic receptor interactions with the receptor than small
aliphatic substituents, which again is in agreement with the
dramatically increased 5-HT2A/5-HT2C activity displayed by 3d
and 4d compared to the Me- and Et-substituted THAZ
analogues in the series (Table 1). The key role of the aromatic
substituent in 3d and 4d for their 5-HT2A/5-HT2C activity will be
explored in the design, synthesis, and characterization of future
series of THAZ analogues.
including the m1 subtype, which is closely linked with cognitive
functions (Supporting Information Table 1).62 However,
considering that 3d exhibits ∼20-fold higher EC50 values at 5-
HT2A and 5-HT2C than at m1 in comparable Ca2+ imaging assays
and the ability of SB242084 to fully reverse the effect of 3d in the
cognition model strongly suggest that this muscarinic
component is unlikely to contribute to the cognition enhancing
potential of the compound.
The binding distribution of [125I]-N-(p-I-Bn)-THAZ in rat
brain slices did not shed considerable more light on other
putative CNS targets targeted by N-Bn-THAZ. On one hand, the
preliminary nonquantitative experiments using autoradiography
could suggest presence of specific [125I]-N-(p-I-Bn)-THAZ
binding sites for in cortical layers and midbrain regions as the
binding here is partly displaceable by N-Bn-THAZ (Figure 4D).
Such a binding distribution pattern would be in concordance
with membrane homogenate binding studies, and the inability of
5-HT to displace [125I]-N-(p-I-Bn)-THAZ from its native
binding sites in homogenate could suggest that the observed
binding sites may be distinct from 5-HT2A and 5-HT2C receptors
(Figure 4C). On the other hand, the binding distribution could
also be ascribed to nonspecific binding of the radioligand, a
binding that the closely related N-Bn-THAZ is able to reduce,
whereas 5-HT and other 5-HT2 ligands are not. The apparent
low 5-HT2C binding affinity of N-(p-I-Bn)-THAZ (3g) (Ki ∼ 300
nM in the [3H]mesulergine binding assay) could constitute a
problem in terms of obtaining robust [125I]-N-(p-I-Bn)-THAZ
binding intensity as could a potential contamination of cold iodo
compound in the radioligand. All in all, this would make the
radioligand unsuited for quantitative experiments due to the long
exposure times needed (up to four days) and concomitant loss of
resolution. In conclusion, [125I]-N-(p-I-Bn)-THAZ appears not
to be a suitable tool radioligand for mapping native 5-HT2A/5-
HT2C binding sites. In view of these difficulties in interpreting the
data and because the pro-cognitive effects of N-Bn-THAZ appear
to be convincingly explained by an involvement of 5-HT2C
receptors (see below), we will refrain from speculations about
putative additional targets for the ligand.
N-Bn-THAZ Is an Efficient Bioavailable Cognitive
Enhancer. The impact of 5-HT2C receptor modulation on
cognition remains a matter of debate, with recent studies
providing rather conflicting evidence on the beneficial effect of
either agonism or antagonism.22 In the present study, we
investigated the effect of N-Bn-THAZ (3d) on spatial
recognition memory in mice in the two-trial place recognition
Y-maze (Figure 5). In contrast to other classical tests of spatial
learning and memory, such as the Morris water maze, this
learning task avoids the use of appetitive or aversive reinforcers
that may trigger nonspecific effects. Moreover, it does not require
extensive training procedures before addressing mnemonic
processes. Administering the compound to mice prior to the
encoding phase restored delay-induced spatial memory impair-
ment during the retrieval phase. This effect was fully reversed by
coadministration of a selective 5-HT2C antagonist, indicating that
N-Bn-THAZ mediates its effects via 5-HT2C receptor agonism
and suggesting a central role for 5-HT2C receptors in spatial
learning. As the compounds were dosed prior to the acquisition
trial, the pro-cognitive effect observed during retrieval is due to a
potentiation of either the encoding phase, the consolidation
phase, or both. Three different parameters were examined as
measures for the exploration of novelty versus familiarity, namely
“distance travelled”, “time spent”, and “number of arm entries”.
The factors time spent and distance traveled reflect “inspective”
N-Bn-THAZ Is a Selective 5-HT2A and 5-HT2C Receptor
Agonist. The elaborate in vitro profiling of N-Bn-THAZ (3d) at
numerous putative CNS targets was conducted in part to
delineate the selectivity profile of the ligand and in part to search
for potential alternative mechanisms underlying its pro-cognitive
properties. Interestingly, the compound appears to be quite
selective for 5-HT2A and 5-HT2C, exhibiting no significant
activity at the other 5-HT receptors tested, at a wide range of
other neurotransmitter receptors and transporters or at
numerous other CNS targets at a concentration of 10 μM
(Supporting Information Table 1). While the 5-HT2A/5-HT2C
activity of N-Bn-THAZ in itself could give rise to hallucinations
and adverse cardiovascular effects,4,61 the lack of activity of the
compound at the 5-HT2B subtype is advantageous from a
therapeutic perspective. Interestingly, N-Bn-THAZ exhibits
weak agonist activity at some muscarinic acetylcholine receptors,
1220
dx.doi.org/10.1021/jm301656h | J. Med. Chem. 2013, 56, 1211−1227

Journal of Medicinal Chemistry
Article
performed on Merck silica gel (0.040−0.063 mm), and for dry column
vacuum chromatography (DCVC)32 Merck silica gel (0.015−0.040
mm) was applied. Unless otherwise stated, DCVC was performed using
the solvents EtOAC and heptane with concentrations of EtOAc varying
from 0−100% with 5% increments. Melting points were recorded on a
exploration, whereas the number of arm entries (encounter with
novelty) corresponds more to “inquisitive” exploration.63 While
it has been suggested that drugs can affect these parameters in
opposite ways,63 N-Bn-THAZ exhibited beneficial effects on
both parameters. One disadvantage of measuring the time spent
in a given arm of the Y-maze is that it does not distinguish
between active exploration and nonrelated behavior (e.g.,
grooming, freezing). Thus, analyzing distance traveled is a
more accurate readout of inspective behavior.
A cognition enhancing potential of 5-HT2C agonists has been
addressed recently utilizing different compounds in rodents with
CP 809101 improving object recognition memory36 and Ro 60-
0175 enhancing attention.64 The present study is thus in line with
the recent literature but is, to our knowledge, the first report of a
beneficial effect of a 5-HT2C agonist on spatial learning. In
general, increased levels of acetylcholine and dopamine elicited
by 5-HT2C agonism have been ascribed to such cognition
promoting efficacy,65 but it should be noted that 5-HT2C
antagonists have been reported to bear pro-cognitive potential
as well.66 Thus, the efficacy and underlying mechanisms might
depend on the brain regions involved in the task and the chosen
animal model of cognitive impairment. In any case, the effect
described here underlines the mnemonic potential of N-Bn-
THAZ on hippocampus-dependent spatial memory, and it will
be of interest to address other cognitive domains as well as the
potential to N-Bn-THAZ and other THAZ analogues to reverse
cognitive deficits in animal disease models of cognitive
impairment and dementia in future studies.
Buchner instrument or on a SRS optimelt apparatus and were
̈
1
uncorrected. H and 13C NMR was recorded on a 300 MHz Varian
Mercury 300BB equipped with a 5 mm 1H{BB} probe, a 300 Varian
Gemini 2000BB spectrometer equipped with a 5 mm 31P, 13C{1H, 19F}
probe or a Bruker Avance 400 MHz spectrometer equipped with a 5 mm
PABBO BB{1H, 19F} Z-GRD probe, at 300 K. Data are tabulated in the
following order: chemical shift (δ) [multiplicity (b, broad; s, singlet; d,
doublet; t, triplet; q, quartet; sept, septet; m, multiplet), coupling
constant(s) J (Hz), number of protons]. The solvent residual peak was
used as internal reference.69 Elemental analyses were performed at
Analytical Research Department, H. Lundbeck A/S Denmark, or by J.
Theiner, Department of Physical Chemistry, University of Vienna,
Austria. Preparative and analytical reverse phase HPLC of the
radioligand were performed on a Waters Breeze HPLC system equipped
with a binary pump system, dual wavelength detector, and a Carroll
Ramsey S105 in-line radiodetector based on a silicon diode. Separation
was achieved using a Phenomenex Luna (2) C-18 (250 mm × 4.6 mm)
column and a standard binary buffer system with a flow of 1 mL/min:
buffer A (4.9% acetonitrile, 95% water, 0.1% trifluoroacetic acid), buffer
B (99.9% acetonitrile, 0.1% trifuoroacetic acid). Preparative separation
was achieved using the following gradient: 0−3 min (0% B); 3−35 min
(gradient to 60% B); 35−37 min (gradient to 100% B); 37−40 min
(100% B); 40−41 min (gradient to 0% B); 41−45 min (0% B).
Analytical separation and analysis was performed using the gradient: 0−
15 min (gradient from 0% to 100% B); 15−20 min (100% B); 20−21
min (gradient to 0% B); 21−25 min (0% B). The purity of all tested
compounds was established using combustion analysis or analytical
HPLC. The elemental analysis calculated values are within 0.4% of the
found values, and the HPLC purity is ≥95% unless otherwise stated.
General Procedure for O- and N-Alkylation. To a solution of 832
(for 10b, 11b), 928 (for 12a−g, 13a−g), 14,28 (for 16a−c, 17a−c), or
1532 (for 18c) (1 equiv) in acetone (14 mL/g) was added potassium
carbonate (2.5 equiv), and the mixture was refluxed for 30 min. To this
mixture was added the appropriate alkyl or phenylalkyl halide (1.1
equiv). The reaction mixture was refluxed for 24 h, filtered, and
evaporated. The resulting residue was subjected to FC or DCVC.
2-Ethyl-3-oxo-2,3,4,5,6,7-hexahydroisoxazolo[5,4-c]pyridin-6-
ium Fumarate (1b). Using the general procedure for alkylation of 8 and
ethyl bromide as the alkyl halide, the two isomeric products: Methyl 2-
ethyl-3-oxo-2,3,4,5-tetrahydroisoxazolo[5,4-c]pyridine-6(7H)-carboxy-
late (10b) (0.45 g, 21%) and methyl 3-ethoxy-4,5-dihydroisoxazolo[5,4-
c]pyridine-6(7H)-carboxylate (11b) (0.29 g, 14%) were obtained. A
solution of 10b (0.84 g, 3.1 mmol) in HCl in AcOEt (2.8N, 4 mL) was
left at room temperature until a clear solution appeared and then
evaporated. The residue was added aqueous KOH (2N, 3 mL) and water
(10 mL). The mixture was extracted with CH2Cl2 (3 × 15 mL). The
combined organic phases were dried (MgSO4), filtered, and evaporated.
This residue was dissolved in Et2O (10 mL), and a solution of fumaric
acid (0.22 g, 1.9 mmol) in 2-propanol (4 mL) was added to give the
product as colorless crystals (0.41 g, 55%): mp 136−137 °C. 1H NMR
(400 MHz, MeOD-d4): δ 6.69 (s, 1H), 3.98−3.92 (m, 4H), 3.16 (t, J =
5.8 Hz, 2H), 2.45 (tt, J = 5.8, 1.7 Hz, 2H), 1.29 (t, J = 7.1 Hz, 3H). 13C
NMR (101 MHz, MeOD-d4): δ 170.7, 167.7, 166.0, 136.0, 107.2, 42.77,
42.64, 41.6, 19.1, 12.9. Anal. (C8H12N2O2·0.5C4H4O4·0.33H2O) C, H,
N.
CONCLUSION
■
In summary, the N- and O-benzyl-substituted THAZ analogues
constitute a novel class of bioavailable and CNS accessible 5-
HT2A/5-HT2C receptor agonists characterized by pronounced
selectivity for these receptors over a plethora of other targets.
This adds yet another interesting chapter to the story of the
medicinal chemistry versatility of the bicyclic 3-isoxazole, which
also constitutes the scaffold of several interesting pharmaco-
logical tool compounds targeting ionotropic glutamate receptors,
GABAA receptors, and GABA transporters.27,67,68 While more
potent and equally selective agonists for 5-HT2A and 5-HT2C
receptors certainly exist, it may very well be possible to increase
the agonist potencies of N-Bn-THAZ (3d) and O-Bn-THAZ
(4d) by future development, whereas it remains to be seen if the
5-HT2A and 5-HT2C activity in the THAZ analogue can be
effectively separated. The cognition enhancement displayed by
3d in the spatial memory assay is highly interesting, and the
apparent key role of 5-HT2C for this pro-cognitive potential in
the compound supports the notion of this receptor being
involved in processes underlying learning and memory proposed
in previous studies. Besides the obvious therapeutic potential in
this cognitive enhancement, it could be interesting to apply 3d,
4d or analogues of these as pharmacological tools in future
studies aimed at unraveling the physiological roles and
therapeutic prospects in 5-HT2A and 5-HT2C receptors.
3-Ethoxy-4,5,6,7-tetrahydroisoxazolo[5,4-c]pyridin-6-ium Fuma-
rate (2b). A solution of 11b (0.6 g, 2.2 mmol) in HCl in EtOH (4N, 11
mL) was left at room temperature until a clear solution appeared and
then evaporated. To the residue was added aqueous sodium carbonate
(2N, 15 mL), and the mixture was extracted with CH2Cl2 (3 × 25 mL).
The combined organic phases were dried (MgSO4), filtered, and
evaporated. This residue was dissolved in Et2O (10 mL), and a solution
of fumaric acid (0.23 g, 2.0 mmol) in 2-propanol (5 mL) was added. The
precipitated crude of the product was recrystallized (MeCN−EtOH) to
give the product (0.37 g, 72%): mp 137−141 °C. 1H NMR (400 MHz,
EXPERIMENTAL SECTION
■
Chemistry. General Procedures. All reagents and solvents were
purchased from Aldrich and used without further purification. All air-
and moisture-sensitive reactions were performed under a nitrogen
atmosphere using syringe-septum cap techniques and flame-dried
glassware. Thin layer chromatography was carried out on Merck silica
gel 60 F254 plates, and detection took place using UV (254 and 366 nm)
or KMnO4 spray reagent. Flash column chromatography (FC) was
1221
dx.doi.org/10.1021/jm301656h | J. Med. Chem. 2013, 56, 1211−1227

Journal of Medicinal Chemistry
Article
1
MeOD-d4): δ 4.36 (t, J = 1.6 Hz, 2H), 4.32 (q, J = 7.1 Hz, 2H), 3.49 (t, J
= 6.1 Hz, 2H), 2.72 (tt, J = 6.1, 1.6 Hz, 2H), 1.42 (t, J = 7.1 Hz, 3H). 13C
NMR (101 MHz, MeOD-d4): δ 170.5, 161.6, 102.9, 67.4, 42.7, 41.6,
16.9, 14.8. Anal. (C8H12N2O2·0.5C4H4O4·0.5H2O) C, H, N.
Et2O) to give the product (0.87 g, 70%): mp 189 °C (decomp). H
NMR (D2O): δ 6.61 (s, 2H), 3.87 (t, J = 6 Hz, 2H), 3.3−3.5 (m, 4H),
3.07 (t, J = 5 Hz, 2H), 2.65 (t, J = 5 Hz, 2H), 1.5−1.8 (m, 2H), 0.80 (t, J =
5 Hz, 2H). Anal. (C10H16N2O2·C4H4O4) C, H, N.
2-Methyl-3-oxo-3,4,5,6,7,8-hexahydro-2H-isoxazolo[4,5-d]-
azepin-6-ium Bromide (3a). Using the general procedure for alkylation
of 9 and methyl iodide as the alkyl halide, the two isomeric products:,
methyl 2-methyl-3-oxo-4,5,7,8-tetrahydro-2H-isoxazolo[4,5-d]azepine-
6(3H)-carboxylate (12a) (5.30 g, 41%) and methyl 3-methoxy-7,8-
dihydro-4H-isoxazolo[4,5-d]azepine-6(5H)-carboxylate (13a) (5.35 g,
42%), were obtained. A solution of 12a (3.50 g, 1.6 mmol) in a solution
of HBr in AcOH (33%, 35 mL) was left at room temperature for 48 h
and then evaporated. The residue was recrystallized (MeOH−Et2O) to
give 3.19 g (83%) of the desired product: mp 171.0−173.0 °C. 1H NMR
(300 MHz, CDCl3/MeOD-d4): δ 3.85 (b s, 2H), 3.43 (t, J = 6.6 Hz, 2H),
3.17 (t, J = 5.7 Hz, 2H), 2.81 (t, J = 5.7 Hz, 2H). Anal.
(C8H12N2O2·HBr) C, H, N.
2-Benzyl-3-oxo-3,4,5,6,7,8-hexahydro-2H-isoxazolo[4,5-d]-
azepin-6-ium Fumarate (3d). Using the general procedure for
alkylation of 9 and benzyl iodide as the alkyl halide, the two isomeric
products, methyl 2-benzyl-3-oxo-4,5,7,8-tetrahydro-2H-isoxazolo[4,5-
d]azepine-6(3H)-carboxylate (12d) (0.94 g, 49%) and methyl 3-
benzyloxy-7,8-dihydro-4H-isoxazolo[4,5-d]azepine-6(5H)-carboxylate
(13d) (0.58 g, 30%) were obtained. A solution of 12d (3.75g, 12 mmol)
in HBr in AcOH (33%, 8 mL) was left at room temperature for 72 h,
evaporated, and worked up as described for 3c. To a solution of crude
amine (2.39 g, 10 mmol) in 2-propanol (5 mL) was added a solution of
fumaric acid (1.16 g, 10 mmol) in 2-propanol (10 mL). The precipitate
was recrystallized (MeOH−Et2O) to give the title compound as crystals
(3.14 g, 70%): mp 152 °C (decomp). 1H NMR (300 MHz, MeOD-d4):
δ 7.29−7.33 (m, 5H), 6.66 (s, 2H), 5.06, (s, 2H), 3.31−3.39 (m, 4H),
3.00 (t, J = 5.7 Hz, 2H), 2.69 (t, J = 5.7 Hz, 2H). Anal.
(C14H16N2O2·C4H4O4) C, H, N.
3-Benzyloxy-7,8-dihydro-4H-isoxazolo[4,5-d]azepine-6-ium
Chloride (4d). Prepared as described for 4a from 13d. The product was
isolated and recrystallized (EtOH−AcOEt) (0.6 g, 53%): mp 139−140
°C. 1H NMR (400 MHz, MeOD-d4): δ 7.47−7.42 (m, 2H), 7.41−7.34
(m, 3H), 5.26−5.24 (s, 2H), 3.42−3.34 (m, 4H), 3.16−3.12 (m, 2H),
2.73−2.69 (m, 2H). Anal. (C14H16N2O2·HCl·0.5H2O) C, H, N.
2-(2-Iodobenzyl)-3-oxo-3,4,5,6,7,8-hexahydro-2H-isoxazolo[4,5-
d]azepin-6-ium Bromide (3e). Using the general procedure for
alkylation of 9 and 2-iodobenzyl bromide as the alkyl halide, the two
isomeric products, methyl 2-(2-iodobenzyl)-3-oxo-4,5,7,8-tetrahydro-
2H-isoxazolo[4,5-d]azepine-6(3H)-carboxylate (12e) (0.26 g, 42%)
and methyl 3-(2-iodobenzyloxy)-7,8-dihydro-4H-isoxazolo[4,5-d]-
azepine-6(5H)-carboxylate (13e) (0.33 g, 54%), were obtained. A
mixture of 12e (0.33 g, 0.78 mmol) in HBr in AcOH (33%, 5 mL) was
refluxed for 20 h at 110 °C. The reaction mixture was evaporated and
recrystallized (MeOH−ether) to give (0.33, 79%). 1H NMR (300 MHz,
CDCl3): δ 7.89 (d, J = 7.8 Hz, 1H), 7.37 (t, J = 7.5 Hz, 1H), 7.22 (dd, J =
1.5, 7.8 Hz, 1H), 7.06 (dt, J = 1.5, 7.5 Hz, 1H), 5.08 (s, 2H), 3.38−3.44
(m, 4H), 3.01 (t, J = 5.7 Hz, 2H), 2.71 (t, J = 6.0 Hz, 2H). 13C NMR (75
MHz, CDCl3): δ 167.4, 166.5, 140.4, 136.2, 130.9, 130.3, 129.4, 109.3,
97.9, 67.1, 54.9, 47.3, 44.3, 25.0, 18.4 ppm. Anal.
(C14H15N2O2I·HBr·0.5H2O) C, H, N.
3-Methoxy-7,8-dihydro-4H-isoxazolo[4,5-d]azepine-6-ium Chlor-
ide (4a). A solution of 13a (0.68 g, 3.0 mmol) in a solution of KOH in
MeOH (4N, 6 mL) was left at room temperature for 8 h and evaporated.
To the residue was added water (10 mL) and extracted with CHCl3 (5 ×
30 mL). The combined organic phases were dried (MgSO4), filtered,
and evaporated. The residue was dissolved in a solution of HCl in AcOEt
(2.8N, 15 mL) and stirred at room temperature. The product was
1
isolated as crystals (297 mg, 44%): mp 201−202 °C. H NMR (400
MHz, MeOD-d4): δ 3.96 (s, 3H), 3.47−3.44 (m, 2H), 3.43−3.40 (m,
2H), 3.21−3.17 (m, 2H), 2.77−2.73 (m, 2H). 13C NMR (101 MHz,
MeOD-d4): δ 171.7, 169.2, 105.6, 57.7, 48.6, 45.8, 25.8, 18.8. Anal.
(C8H12N2 O2·HCl) C, H, N.
2-Ethyl-3-oxo-3,4,5,6,7,8-hexahydro-2H-isoxazolo[4,5-d]azepin-
6-ium Fumarate (3b). Using the general procedure for alkylation of 9
and ethyl iodide as the alkyl halide, the two isomeric products, methyl 2-
ethyl-3-oxo-4,5,7,8-tetrahydro-2H-isoxazolo[4,5-d]azepine-6(3H)-car-
boxylate (12b) (0.94 g, 49%) and methyl 3-ethoxy-7,8-dihydro-4H-
isoxazolo[4,5-d]azepine-6(5H)-carboxylate (13b) (0.58 g, 30%), were
obtained. A solution of 12b (5.15 g, 24 mmol) in a solution of HBr in
AcOH (33%, 30 mL) was left at room temperature for 72 h and then
evaporated. The residue was dissolved in water (30 mL), and upon
addition of sodium carbonate (6 g) the mixture was extracted with
CHCl3 (3 × 50 mL). The combined organic phases were dried
(MgSO4), filtered, and evaporated. This residue was dissolved in a
mixture of EtOH (2 mL) and Et2O (20 mL), and a solution of fumaric
acid (1.86 g, 16.0 mmol) in 2-propanol (25 mL) was added. The
precipitated crude of the product was recrystallized (MeOH−Et2O) to
give the product (4.12 g, 65%): mp 174−176 °C. 1H NMR (300 MHz,
CDCl3): δ 6.73 (s, 2H), 3.92 (q, J = 7.1 Hz, 2H), 3.24 (t, J = 5.1 Hz, 2H),
3.18 (t, J = 5.7 Hz, 2H), 2.96 (t, J = 6.3 Hz, 2H), 2.64 (t, J = 5.7 Hz, 2H).
Anal. (C9H14N2O2·C4H4O4) C, H, N.
3-Ethoxy-7,8-dihydro-4H-isoxazolo[4,5-d]azepine-6-ium Chloride
(4b). Prepared as described for 4a from 13b. The product was isolated as
crystals (0.60 g, 66%): mp 202−203 °C. 1H NMR (400 MHz, MeOD-
d4): δ 4.27 (q, J = 7.1 Hz, 2H), 3.47−3.41 (m, 4H), 3.21−3.17 (m, 2H),
2.77−2.73 (m, 2H), 1.40 (t, J = 7.1 Hz, 3H). 13C NMR (101 MHz,
MeOD-d4): δ 171.0, 168.9, 105.7, 67.1, 48.6, 45.8, 25.8, 18.8, 14.8. Anal.
(C9H14N2O2·HCl) C, H, N.
2-Propyl-3-oxo-3,4,5,6,7,8-hexahydro-2H-isoxazolo[4,5-d]-
azepin-6-ium Fumarate (3c). Using the general procedure for
alkylation of 9 and propyl bromide as the alkyl halide, the two isomeric
products, methyl 2-propyl-3-oxo-4,5,7,8-tetrahydro-2H-isoxazolo[4,5-
d]azepine-6(3H)-carboxylate (12c) (0.49 g, 64%) and methyl 3-
propoxy-7,8-dihydro-4H-isoxazolo[4,5-d]azepine-6(5H)-carboxylate
(13c) (0.21 g, 28%), were obtained. A solution of 12c (1.01 g, 4.0
mmol) in HBr in AcOH (33%, 8 mL) was left at room temperature for
72 h and then evaporated. The residue was suspended in aqueous
NaOH (1N, 20 mL), and the mixture was extracted with CH2Cl2 (3 × 25
mL). The combined organic phases were washed with a saturated
aqueous solution of NaHCO3 (25 mL), dried (MgSO4), filtered, and
evaporated. This residue was dissolved in 2-propanol (4 mL), and a
solution of fumaric acid (0.46 g, 3.9 mmol) in 2-propanol (4 mL) was
added. The precipitate of the crude product was recrystallized (MeOH−
2-(3-Iodobenzyl)-3-oxo-3,4,5,6,7,8-hexahydro-2H-isoxazolo[4,5-
d]azepin-6-ium Bromide (3f). Using the general procedure for
alkylation of 9 and 3-iodobenzyl bromide as the alkyl halide, the two
isomeric products, methyl 2-(3-iodobenzyl)-3-oxo-4,5,7,8-tetrahydro-
2H-isoxazolo[4,5-d]azepine-6(3H)-carboxylate (12f) (0.20 g, 33%) and
methyl 3-(3-iodobenzyloxy)-7,8-dihydro-4H-isoxazolo[4,5-d]azepine-
6(5H)-carboxylate (13f) (0.40 g, 66%), were obtained. A mixture of
12f (0.40 g, 0.93 mmol) in HBr in AcOH (33%, 5 mL) was refluxed for
20 h at 110 °C. Recrystallization (MeOH−Et2O) gave 3f as colorless
1
crystals (0.43 g, 74%). H NMR (300 MHz, D2O): δ 6.56 (b s, 1H,),
6.50 (b d, J = 7.8 Hz, 1H), 6.21 (b d, J = 8.1 Hz, 1H), 6.02 (t, J = 7.8 Hz,
1H), 3.81 (s, 2H), 2.30−2.42 (m, 4H), 1.95 (b t, J = 5.4 Hz, 2H), 1.68 (b
t, J = 5.7 Hz, 2H). 13C NMR (75 MHz, CDCl3): δ 167.6, 166.9, 137.9,
137.1, 137.0, 131.5, 127.9, 109.3, 95.5, 67.2, 49.5, 47.3, 44.3, 25.2, 18.5
ppm. Anal. (C14H15N2O2I·HBr·0.5H2O) C, H, N.
2-(4-Iodobenzyl)-3-oxo-3,4,5,6,7,8-hexahydro-2H-isoxazolo[4,5-
d]azepin-6-ium Bromide (3g). Using the general procedure for
alkylation of 9 and 4-iodobenzyl bromide as the alkyl halide, the two
isomeric products, methyl 2-(4-iodobenzyl)-3-oxo-4,5,7,8-tetrahydro-
2H-isoxazolo[4,5-d]azepine-6(3H)-carboxylate (12g) (0.32 g, 53%)
and methyl 3-(4-iodobenzyloxy)-7,8-dihydro-4H-isoxazolo[4,5-d]-
azepine-6(5H)-carboxylate (13g) (0.40 g, 33%), were obtained. A
mixture of 12g (0.40 g, 0.93 mmol) in HBr in AcOH (33%, 5 mL) was
refluxed for 20 h at 110 °C. Recrystallization (MeOH−Et2O) gave 3g as
colorless crystals (0.26 g, 59%). 1H NMR (300 MHz, D2O): δ 7.72 (b d,
J = 8.1 Hz, 2H), 7.06 (b d, J = 8.1 Hz, 2H), 5.03 (s, 2H), 3.38−3.50 (m,
4H), 3.06 (b t, J = 6.0 Hz, 2H), 2.71 (b t, J = 5.7 Hz, 2H). 13C NMR (75
1222
dx.doi.org/10.1021/jm301656h | J. Med. Chem. 2013, 56, 1211−1227

Journal of Medicinal Chemistry
Article
MHz, CDCl3): δ 167.5, 166.7, 138.4, 134.3, 130.4, 109.3, 94.3, 44.4,
25.1, 18.5 ppm. Anal. (C14H15N2O2I·HBr·1.5H2O) C, H, N.
[4,5-c]pyridine-5(4H)-carboxylate (17b) (3.37 g, 50%), were obtained.
5b was prepared from 16b as described for 3d. The product was isolated
and recrystallized (MeOH−EtO2) (1.9 g, 48%): 161.4−163.1. 1H NMR
(400 MHz, MeOD-d4): δ 6.70 (s, 3H), 4.01−3.96 (m, 4H), 3.56 (t, J =
6.2 Hz, 2H), 2.97 (tt, J = 6.2, 1.6 Hz, 2H), 1.31 (t, J = 7.1 Hz, 3H). 13C
NMR (101 MHz, MeOD-d4): δ 170.4, 166.2, 165.5, 135.8, 103.5, 42.7,
41.7, 38.9, 21.4, 13.0. Anal. (C8H12N2O2·1.5C4H4O4) C, H, N.
2,6-Dimethyl-3-oxo-3,4,5,6,7,8-hexahydro-2H-isoxazolo[4,5-d]-
azepin-6-ium Chloride (3h). A solution of 3a (2.50 g, 10 mmol) in
water (30 mL), adjusted to pH ca. 12 by addition of K2CO3, was
extracted with CH2Cl2 (3 × 50 mL). The combined organic phases were
evaporated and the residue dissolved in a mixture of aqueous solutions
of formic acid (98%, 25 mL) and formaldehyde (40%, 25 mL). This
solution was refluxed for 4 h and then evaporated. A mixture of the
residue and water (8 mL) adjusted to pH 12 by addition of K2CO3, was
extracted with CHCl3 (3 × 100 mL). The combined organic phases were
dried (MgSO4), filtered, and evaporated. This residue was dissolved in
EtOAc (5 mL), and upon addition of a solution of HCl in EtOAc (1N,
10 mL), crude product precipitated. Recrystallization (EtOH−Et2O)
afforded 1.45 g (66%) of the title compound: mp 215 °C (decomp). 1H
NMR (300 MHz, MeOD-d4): δ 6.71 (s, 2H), 3.61 (b t, J = 1.5 Hz, 2H),
3.50 (s, 3H), 2.88 (t, J = 5.7 Hz, 2H), 2.59 (s, 3H), 2.45 (dt, J = 1.5, 5.7
Hz, 2H). Anal. (C9H14N2O2·HCl) C, H, N.
6-Methyl-3-methoxy-7,8-dihydro-4H-isoxazolo[4,5-d]azepine-6-
ium Chloride (4h). Prepared as described for 3h. The product was
isolated and recrystallized (MeOH−AcOEt) (0.24 g, 62%): mp 179 °C
(decomp). 1H NMR (400 MHz, MeOD-d4): δ 3.96 (s, 3H), 3.54 (b s,
4H), 3.22 (t, J = 5.8 Hz, 2H), 3.03 (s, 3H), 2.78 (t, J = 5.7 Hz, 2H). 13C
NMR (101 MHz, MeOD-d4): δ 171.5, 168.7, 105.3, 57.83, 57.73, 54.9,
44.1, 24.4, 17.6.Anal. (C9H14N2O2·HCl) C, H, N.
2-Ethyl-6-methyl-3-oxo-3,4,5,6,7,8-hexahydro-2H-isoxazolo[4,5-
d]azepin-6-ium Fumarate (3i). The N-methylation of 3b (0.89 g, 3.0
mmol) was accomplished following a procedure analogous with that
described for 3h. To a solution of the crude free amine of 3i (0.54 g, 2.8
mmol) in 2-propanol (2 mL) was added a solution of fumaric acid (0.32
g, 2.8 mmol) in 2-propanol (5 mL). Upon addition of Et2O (5 mL),
crude fumaric salt slowly precipitated. Recrystallization (MeOH−Et2O)
afforded 3i as crystals (0.99 g, 66%): mp 126.0−128.0 °C. 1H NMR (300
MHz, MeOD-d4): δ 6.68 (s, 2H), 3.92, (q, J = 7.2 Hz, 2H), 3.28 (b t, J =
6.3 Hz, 2H), 3.21 (b t, J = 5.7 Hz, 2H), 3.01 (t, J = 5.7 Hz, 2H), 2.79 (s,
3H), 2.61 (t, J = 5.7 Hz, 2H). Anal. (C10H16N2O2·C4H4O4) C, H, N.
6-Methyl-3-ethoxy-7,8-dihydro-4H-isoxazolo[4,5-d]azepine-6-
ium Fumarate (4i). Prepared as described for 3i. The product was
isolated and recrystallized (MeOH−Et2O) (0.52 g, 61%): mp 129−130
°C. 1H NMR (400 MHz, MeOD-d4): δ 6.70 (s, 2H), 4.26 (q, J = 7.1 Hz,
2H), 3.40−3.32 (m, J = 5.0 Hz, 4H), 3.17−3.12 (m, 2H), 2.88 (s, 3H),
2.72−2.68 (m, 2H), 1.40 (t, J = 7.1 Hz, 3H). 13C NMR (101 MHz,
MeOD-d4): δ 170.9, 170.6, 169.3, 135.9, 105.8, 67.0, 57.7, 54.9, 44.1,
25.1, 18.2, 14.9. Anal. (C10H16N2O2·C4H4O4) C, H, N.
2-Methyl-3-oxo-2,3,4,5,6,7-hexahydroisoxazolo[4,5-c]pyridin-5-
ium Chloride (5a). To a solution of 14 (0.96 g, 4 mmol) in Et2O (20
mL) was added CH2N2 (1.65 mmol). The reaction mixture was stirred
overnight at room temperature added AcOH (2 mL) and evaporated.
FC gave the two isomers: methyl 2-methyl-3-oxo-2,3,6,7-
tetrahydroisoxazolo[4,5-c]pyridine-5(4H)-carboxylate (16a) (0.33 g,
32%) and methyl 3-methoxy-6,7-dihydroisoxazolo[4,5-c]pyridine-
5(4H)-carboxylate (17a) (0.39 g, 38%). 5a was prepared from 16a
using the method described for 3h. The product was isolated and
recrystallized (MeOH−AcOEt) (0.14 g, 86%): 152.1−153.2. 1H NMR
(400 MHz, MeOD-d4): δ 4.01 (t, J = 1.7 Hz, 2H), 3.60 (t, J = 6.2 Hz,
2H), 3.55 (s, 3H), 2.98 (tt, J = 6.2, 1.6 Hz, 2H). 13C NMR (101 MHz,
MeOD-d4): δ 165.84, 165.80, 103.0, 41.8, 39.0, 33.2, 21.2. Anal.
(C7H12N2O2·HCl) C, H, N.
2-Benzyl-3-oxo-2,3,4,5,6,7-hexahydroisoxazolo[4,5-c]pyridin-5-
ium Fumarate (5c). Using the general procedure for alkylation of 14
and benzyl bromide as the alkyl halide, the two isomeric products,
methyl 2-benzyl-3-oxo-2,3,6,7-tetrahydroisoxazolo[4,5-c]pyridine-
5(4H)-carboxylate (16c) (10.0 g, 69%) and methyl 3-benzyloxy-6,7-
dihydroisoxazolo[4,5-c]pyridine-5(4H)-carboxylate (17c) (2.3 g, 16%),
were obtained. 5c was prepared from 16c as describe for 1b. The
product was isolated and recrystallized (MeOH−EtO2) (0.82 g, 45%):
1
mp 166−168 °C. H NMR (400 MHz, MeOD-d4): δ 7.36−7.31 (m,
5H), 6.69 (s, 2H), 5.09 (s, 2H), 3.91 (t, J = 1.7 Hz, 2H), 3.44 (t, J = 6.1
Hz, 2H), 2.83 (tt, J = 6.1, 1.7 Hz, 2H). 13C NMR (101 MHz, MeOD-d4):
δ 170.7, 167.2, 166.0, 136.00, 135.94, 129.8, 129.40, 129.23, 104.4, 50.9,
41.8, 39.0, 21.9.Anal. (C13H14N2O2·C4H4O4) C, H, N.
3-Benzyloxyoxy-4,5,6,7-tetrahydroisoxazolo[4,5-c]pyridin-5-ium
Chloride (6c). Prepared from 17c as describe for 4a. The product was
isolated and recrystallized (MeCN−EtO2) (1.4 g, 34%): mp 196−197
°C. 1H NMR (400 MHz, MeOD-d4): δ 7.47−7.44 (m, 2H), 7.41−7.35
(m, 3H), 5.29 (s, 2H), 4.10 (t, J = 1.5 Hz, 2H), 3.58 (t, J = 6.2 Hz, 2H),
3.05 (tt, J = 6.2, 1.5 Hz, 2H). 13C NMR (101 MHz, MeOD-d4): δ 169.5,
166.7, 136.9, 129.73, 129.60, 129.58, 99.3, 73.1, 42.2, 39.3, 21.7.Anal.
(C13H14N2O2·HCl) C, H, N.
2-Benzyl-3-oxo-3,4,5,6,7,8-hexahydro-2H-isoxazolo[4,5-c]azepin-
5-ium Fumarate (7c). Using the general procedure for alkylation of 15
and benzyl bromide as the alkyl halide, the two isomeric products,
methyl 2-benzyl-3-oxo-4,6,7,8-tetrahydro-2H-isoxazolo[4,5-c]azepine-
5(3H)-carboxylate (18c) (0.4 g, 47%) and the O-benzyl isomer methyl
3-(benzyloxy)-7,8-dihydro-4H-isoxazolo[4,5-c]azepine-5(6H)-carboxy-
late (0.21 g, 28%), were obtained. 7c was prepared from 18c using the
method describe for 1b. The product was isolated and recrystallized
(MeOH−Et2O) (0.28 g, 51%): mp 139−142 °C. 1H NMR (400 MHz,
MeOD-d4): δ 7.38−7.30 (m, 5H), 6.69 (s, 2H), 5.08 (s, 2H), 4.00 (s,
2H), 3.50−3.47 (m, 2H), 2.90−2.86 (m, 2H), 2.10−2.04 (m, 2H). 13C
NMR (101 MHz, MeOD-d4): δ 173.1, 170.9, 166.9, 136.04, 135.88,
129.9, 129.4, 129.2, 105.4, 50.9, 50.2, 40.2, 27.4, 23.4. Anal.
(C14H16N2O2·C4H4O4) C, H, N.
Methyl 2-(Tributylstannylbenzyl)-3-oxo-4,5,7,8-tetrahydro-2H-
isoxazolo[4,5-d]azepine-6(3H)-carboxylate (19). 12g (0.28 g, 0.7
mmol) and tetrakis(triphenylphosphine)palladium(0) (0.084 mg, 0.06
mmol) was dissolved in dry toluene (40 mL) and bis(tributyltin) (0.66
mL, 1.3 mmol) was added under a N2 atmosphere. The reaction mixture
was refluxed for 22 h followed by evaporation and DCVC, which gave
1
the title compound as a colorless oil (157 mg, 41%). H NMR (300
MHz, CDCl3): δ 7.42 (b. d, J = 7.8 Hz, 2H), 7.24 (b d, J = 8.1 Hz, 2H),
4.95 (s, 2H), 3.72 (s, 3H), 3.52−3.66 (m, 4H), 2.68−2.80 (m, 2H),
2.51−2.62 (m, 2H). 13C NMR (75 MHz, CDCl3): δ 167.4, 167.2, 155.9,
142.0, 136.8, 134.3, 127.5, 109.8, 109.4, 52.9, 49.8, 48.4, 45.6, 29.2, 28.7,
27.5, 22.4, 13.9, 10.2. Anal. HPLC purity ≥99%.
Methyl 2-(4-Iodobenzyl)-3-oxo-4,5,7,8-tetrahydro-2H-isoxazolo-
[4,5-d]azepine-6(3H)-carboxylate (12g from 19). Chloramine-T
trihydrate (0.36 g, 1.3 mmol) was suspended in a solution of EtOAc−
DMF−AcOH (50 mL, 94:5:1) and water (0.4 mL). The solution was
added to 19 (27 mg, 0.045 mmol), followed by a solution of NaI (7 mg,
0.047 mmol) in phosphate buffered saline (2 mL). The reaction mixture
was stirred at room temperature for 25 min and quenched by adding
Na2S2O3 (0.25 g, 1.3 mmol). The reaction mixture was washed with
water (6 × 10 mL), dried (MgSO4), filtered, and evaporated. DCVC
gave 12g (13.5 mg, 71%). 1H NMR as described for 12g from 9.
3-Methoxy-4,5,6,7-tetrahydroisoxazolo[4,5-c]pyridin-5-ium Fu-
marate (6a). Prepared from 17a using the method described for 1b.
The product was isolated and recrystallized (MeOH−EtO2) (0.127 g,
62%): 192−195 (decomp). 1H NMR (400 MHz, MeOD-d4): δ 4.09 (t, J
= 1.5 Hz, 2H), 3.99 (s, 3H), 3.58 (t, J = 6.2 Hz, 2H), 3.06 (tt, J = 6.2, 1.5
Hz, 2H). 13C NMR (101 MHz, MeOD-d4): δ 170.2, 166.7, 99.1, 57.9,
42.1, 39.2, 21.6. Anal. (C7H12N2O2·C4H4O4) C, H, N.
[
125I]2-(4-Iodobenzyl)-3-oxo-3,4,5,6,7,8-hexahydro-2H-isoxazolo-
[4,5-d]azepin-6-ium Bromide (21). Chloramine-T trihydrate (41 mg)
was suspended in a mixture of EtOAc (4.75 mL), DMF (0.25 mL), acetic
acid (50 μL), and water (50 μL). Of this solution, 1 mL was added to the
precursor 19 (2.1 mg) in a 2 mL HPLC vial, and shortly after carrier-free
2-Ethyl-3-oxo-2,3,4,5,6,7-hexahydroisoxazolo[4,5-c]pyridin-5-
ium Fumarate (5b). Using the general procedure for alkylation of 14
and ethyl bromide as the alkyl halide, the two isomeric products, methyl
2-ethyl-3-oxo-2,3,6,7-tetrahydroisoxazolo[4,5-c]pyridine-5(4H)-car-
boxylate (16b) (2.8 g, 41%) and methyl 3-ethoxy-6,7-dihydroisoxazolo-
[
125I] NaI (20 μL, 44.0 MBq) was added. When the reaction was
1223
dx.doi.org/10.1021/jm301656h | J. Med. Chem. 2013, 56, 1211−1227

Journal of Medicinal Chemistry
Article
finished, NaS2O3 (9.4 mg in 300 μL water) was added to quench excess
of chloramine-T and the reaction mixture was washed with water (3 ×
300 μL). The organic layer was evaporated to dryness under a gentle
helium flow to give 20. Subsequently, a solution of HBr in acetic acid
(33%, 500 μL) was added to the vial, and the reaction was left for 5 h.
The reaction mixture was cooled on ice and slowly neutralized with a
solution of aqueous NaOH (4N, 500 μL), and the pH was adjusted to
approximately 7. This mixture was subjected to reverse phase HPLC for
purification. The desired product was collected in the interval 22.0−22.3
min. The product, 21, was measured to an activity of 40.3 MBq (RCY:
91%) and analyzed to have a radiochemical purity of 100% and a specific
activity of 94 Ci/mmol. The moderate specific activity was due to traces
of cold iodide compound used in the preparation of the stannyl
precursor. The HPLC fraction was diluted with 5× ultrapurified water
then applied onto a preactivated C-18 Sep-PAK (160 mg). It was then
washed with ultra purified water (5 mL), eluted with 1% AcOH in EtOH
(5 mL), and subsequently diluted to a concentration of approximately 5
MBq/ml with water.
Pharmacology. Materials. Culture media, serum, antibiotics and
buffers for cell culture were obtained from Invitrogen (Paisley, UK). The
Fluo−4/AM dye was obtained from Molecular Probes (Eugene, OR),
and the IP-One assay kit was purchased from Cisbio (Bagnols, France).
[3H]Mesulergine ([N6-methyl-3H]mesulergine) and [3H]ketanserine
were purchased from GE Healthcare (Buckinghamshire, UK). 5-HT and
probenecid were purchased from Sigma-Aldrich (St. Louis, MO),
asenapine was purchased from Ascent Scientific (Bristol, UK), and
PNU-22394, Ro 60-0175, CP 809101, MK-212, WAY 161503,
clozapine, mianserin, and MDL 11,939 were purchased from Tocris
Cookson (Bristol, UK). SB242084 were synthesized in-house at H.
Lundbeck A/S. The cDNAs encoding for the human 5-HT2A (variant 1)
and human 5-HT2C (isoform INI) were generous gifts from Dr. Hans
20 mM HEPES, 1 mM CaCl2, and 1 mM MgCl2, pH 7.4) at a
concentration of 1 × 107 cells/mL. Ligand solutions were prepared in
HBSS supplemented with 1 mM CaCl2, 1 mM MgCl2, and 40 mM LiCl.
Then 5 μL of ligand solution was mixed with 5 μL of cell suspension in a
white 384-well OptiPlate (PerkinElmer, Waltham, MA, USA). The plate
was sealed and incubated at 37 °C for 1 h, followed by 15 min incubation
at room temperature. Next, 10 μL of detection reagents (lysis buffer
containing 2.5% Eu3+-anti-IP1 antibody and 2.5% IP1-d2) was added and
the plate was incubated for 1 h at room temperature. The plate was read
on an Envision (PerkinElmer, Waltham, MA) exciting the cells with light
at 340 nm and measuring emitted light at 615 and 665 nm. In this assay,
the time-resolved-fluorescence resonance energy transfer 665 nm/615
nm ratio is inversely proportional to the IP1 accumulation in the cells
upon 5-HT2A or 5-HT2C receptor activation. The compounds were
characterized in triplicate at least four times at each cell line.
[3H]Ketanserin, [3H]Mesulergine, and [125I]-N-(p-I-Bn)-THAZ
Binding Assays. Competition binding to membranes from tsA201
cells transiently expressing human 5-HT2A and 5-HT2C receptors using
[3H]ketanserin and [3H]mesulergine as radioligands, respectively, was
performed essentially as previously described.44,72 The cells were
transfected with pCDNA3.1-h5-HT2A or pCDNA3.1-h5-HT2C using
PolyFect as a DNA carrier 36−48 h prior to the experiments. On the day
of the assay, the cells were harvested and scraped into assay buffer [50
mM Tris-HCl supplemented with 5 mM CaCl2 (pH 7.4) for 5-HT2A and
50 mM Tris-HCl (pH 7.4) for 5-HT2C], homogenized using an Ultra-
Turrax for 10 s, and centrifuged for 20 min at 50000g. The resulting
pellets were resuspended in fresh assay buffer, homogenized, and
centrifuged at 50000g for another 20 min, after which the pellet was
resuspended in assay buffer and used for the binding experiments.
In the saturation binding experiments, the cell membranes were
incubated with various [3H]ketanserin or [3H]mesulergine concen-
trations in the absence (total binding) or in the presence of 10 μM
mianserin (nonspecific binding). In the competition binding experi-
ments, cell membranes were incubated with a fixed radioligand
concentration (2 nM [3H]ketanserin or 0.5 nM [3H]mesulergine) and
various concentrations of the test compounds. A total binding assay
volume of 1000 μL was used, and nonspecific binding was determined in
reactions with 10 μM mianserin. The binding reactions were incubated
for 1 h at 37 °C. Whatman GF/C filters were presoaked for 1 h in a 0.2%
polyethyleneimine solution, and binding was terminated by filtration
through these filters using a 48-well cell harvester and washing with 3 × 4
mL ice-cold isotonic NaCl solution. Following this, the filters were dried,
3 mL of Opti-Fluor (Packard) was added, and the amount of bound
radioactivity was determined in a scintillation counter. The fraction of
specifically bound radioligand was always <5% of the total amount of
radioligand. The binding experiments were performed in duplicate at
least three times for each compound at each receptor.
Brauner-Osborne.
̈
Construction of Stable Cell Lines and General Cell Culture.
Polyclonal h5-HT2A- and h5-HT2C-HEK293 cell lines were generated
essentially as previously described.70 Cells were transfected with
pCDNA3.1-h5-HT2A or pCDNA3.1-h5-HT2C plasmids using PolyFect
(Qiagen, West Sussex, UK) as a DNA carrier and cultured in culture
medium [Dulbecco’s Modified Eagle Medium supplemented with
penicillin (100 U/mL), streptomycin (100 μg/mL), and 5% dialyzed
fetal bovine serum] supplemented with 2 mg/mL G418 for four weeks
before use. The tsA201 cells and the stable h5-HT2A- and h5-HT2C-
HEK293 cell lines were maintained at 37 °C in a humidified 5% CO2
incubator in culture medium (for the stable cell lines the medium was
supplemented with 1 mg/mL G-418).
Fluo-4/Ca2+ Assay. The functional characterization of 10 reference
ligands and the THAZ analogues at the h5-HT2A- and h5-HT2C-
HEK293 cell lines in the Fluo-4/Ca2+ assay was performed essentially as
previously described.71 The cells were split into poly-D-lysine-coated
black 96-well plates with clear bottom (6 × 104 cells/well). The
following day, the culture medium was aspirated and the cells were
incubated in 50 μL of assay buffer [Hank’s Buffered Saline Solution
(HBSS) containing 20 mM HEPES, 1 mM CaCl2, 1 mM MgCl2, and 2.5
mM probenecid, pH 7.4] supplemented with 6 mM Fluo−4/AM at 37
°C for 1 h. Then the buffer was aspirated, the cells were washed once
with 100 μL of assay buffer, and then 100 μL of assay buffer was added to
the cells (in the antagonist experiments the antagonist was added at this
point). The 96-well plate was assayed in a FLEXStation3 (Molecular
Devices, Crawley, UK) measuring emission [in fluorescence units (FU)]
at 525 nm caused by excitation at 485 nm before and up to 90 s after
addition of 33.3 μL of agonist solution in assay buffer. The compounds
were characterized in duplicate at least three times at each cell line. The
antagonist experiments were performed using an EC75−EC85
concentrations of 5-HT as agonist.
Competition binding studies with [125I]-N-(p-I-Bn)-THAZ (specific
activity 94 Ci/mmol) was performed to membranes from tsA201 cells
transiently expressing human 5-HT2C or to rat brain membranes
prepared from cortex, cerebellum, hippocampus, and midbrain
dissections and prepared using standard procedures.73 The binding
assay was performed exactly as the [3H]ketanserin and [3H]mesulergine
binding assays, except for the use of a different assay buffer [50 mM Tris-
HCl, 10 mM MgSO4, 0.5 mM EDTA, 0.1% ascorbic acid, 0.1% BSA (pH
7.4)] and an assay concentration of 6 nM [125I]-N-(p-I-Bn)-THAZ as
radioligand. Nonspecific binding was determined with 10 μM mianserin
in the experiments with the tsA201 cell membranes, whereas 10 μM
mianserin or 1 mM N-Bn-THAZ were used to determine nonspecific
binding in the experiments with the rat brain membranes.
Cryostat Sectioning and [125I]-N-(p-I-Bn)-THAZ Autoradiog-
raphy. Brains from male Sprague−Dawley rats (250−300 g; Charles
River, Germany) were sectioned on a cryostat in 20 μm coronal sections
at −25 °C (corresponding to bregma 4.5−4.8 mm),74 thaw-mounted
onto gelatinized glass slides, allowed to dry, and stored at −80 °C until
analyzed. Autoradiography was carried out using a protocol highly
similar to previously described using the same buffer as described for
homogenate binding studies.75 Frozen tissue sections were warmed to
room temperature, then preincubated for 30 min at room temperature in
buffer under constant gentle shaking. For determination of total binding,
IP-One Assay. The functional charaterization of ligands at the h5-
HT2A- and h5-HT2C-HEK293 cell lines in the IP-One assay was
performed essentially as described previously.43 On the day of the
experiment, subconfluent h5-HT2A- and h5-HT2C-HEK293 cells were
washed one time with phosphate-buffered saline and detached from the
cell culture plate using dissociation buffer (Sigma-Aldrich). Cells were
centrifuged and resuspended in assay buffer (HBSS supplemented with
1224
dx.doi.org/10.1021/jm301656h | J. Med. Chem. 2013, 56, 1211−1227

Journal of Medicinal Chemistry
Article
sections were incubated in buffer containing 6 nM [125I]-N-(p-I-Bn)-
THAZ for 40 min at room temperature under constant gentle shaking.
Sections used for nonspecific binding included 100 μM 2-Bn-THAZ.
Sections were washed for 2 × 20 s in ice-cold buffer followed by 20 s in
ice-cold double-distilled H2O. Slides were dried at room temperature in
a gentle air stream for 1 h. They were then fixed in paraformaldehyde
vapor overnight at 4 °C. The sections were dried for 3 h in a desiccator at
room temperature, followed by exposure to a 125I-sensitive BAS-2040
phosphor imaging plate (Science Imaging Scandinavia AB, Nacka,
Sweden) for 4 days at 4 °C. The imaging plate was scanned on a BAS-
AUTHOR INFORMATION
Corresponding Author
*Phone: +45 39179650. Fax: +45 35336041. E-mail: aaj@sund.
ku.dk.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This study was supported by the Aase and Ejner Danielsen
Foundation, the Direktør Ib Henriksen Foundation, the
Lundbeck Foundation, the Carlsberg Foundation, the Danish
Medical Research Council, and the Novo Nordisk Foundation.
2500 bioimaging analyzer (Fujifilm Europe GmbH, Dusseldorf,
̈
Germany). The autoradiograms were analyzed with ImageJ V.1.38j
(http://rsbweb.nih.gov/ij/).
Two-Trial Place Recognition Y-Maze. Eight weeks old, male
C57Bl/6J mice (Charles River, Germany) were used in the place
recognition Y-maze task. Animals were housed in pairs in a temperature
and humidity controlled environment on a 12 h (6:00−18:00 h) light
cycle with free access to water and food. The task has been adapted from
earlier reports with slight adaptations described in the following.76,77
The maze consisted of transparent plastic shaped to a Y with each arm
arranged in a 120° angle to each other and measuring 32 cm × 14 cm × 8
cm (L × H × W). The maze was placed on an airflow table. Guillotine
doors were mounted at the center of the maze to allow blockade of each
arm independently. Moreover, fixed and nontransparent walls were
mounted on the outside of the maze in between each two arms to block
free view from one arm into the other. Different visual cues were
mounted outside of the arms, thus out of reach for the animals, making
each arm’s visual environment distinct from the others. During the
acquisition trial, subjects were placed in the middle of the maze with all
three doors closed and subsequently allowed to freely explore two arms
(“familiar”) for 3 min. Animals were then removed manually and
returned to their home cage for the duration of the intertrial interval
(ITI). Following the ITI, subjects were placed back into the center of the
maze (with all three doors initially closed) for the retention trial and
then allowed to freely explore all three arms for 3 min. The arm
previously closed during acquisition was designated as “novel” during
retention. All movement was videotracked by a ceiling-mounted CCD
camera and data transferred to a tracking software EthoVision3.0
(Noldus, Netherlands) for trace analysis. N-Bn-THAZ (dissolved in
0.9% NaCl) was applied orally (po) and SB242084 (dissolved in 10%
HP-betaCD) subcutaneously (sc), both 30 min prior to the acquisition
trial. The vehicle group was dosed with 10% HP-betaCD, po, 30 min
before acquisition. All doses stated in the results part refer to free base.
Data are expressed as the mean value SEM, and a one-way ANOVA
was applied to test for statistical differences between exploration of
familiar and novel arms (SigmaStat 3.0, Aspire Software, USA). Data
from the two familiar arms were not significantly different from each
other and pooled. All in vivo experiments were performed in accordance
with the Danish legislation, and animals were treated with adherence to
the guidelines for the care of experimental animals.
Dr. Hans Brauner-Osborne is thanked for providing us with the
̈
pCDNA3.1-h5-HT2A and pCDNA3.1-h5-HT2C plasmids, and
Dr. Gitte Moos Knudsen is thanked for providing access to
autoradiography equipment. Ulla Geneser is thanked for
technical assistance with the synthetic work, and M.Sc. Lenea
Nørskov-Lauritsen and M.Sc. Stine Engesgaard Jacobsen are
thanked for their kind guidance in connection with the
performance of the IP-One assay.
ABBREVIATIONS USED
■
5-HT, 5-hydroxytryptamine (serotonin); 7-TM, 7-transmem-
brane; CAR, conditioned avoidance response; GABAA, γ-
aminobutyric acid type A; CNS, central nervous system; IP1,
inositol monophosphate; ITI, inter trial interval; LSD, lysergic
acid diethylamide; NK, Newman−Keuls; THAO, 5,6,7,8-
tetrahydro-4H-isoxazolo[4,5-c]azepin-3-ol; THAZ, 5,6,7,8-tetra-
hydro-4H-isoxazolo[4,5-d]azepin-3-ol; THIP, 4,5,6,7-
tetrahydroisoxazolo[5,4-c]pyridin-3-ol; THPO, 4,5,6,7-
tetrahydroisoxazolo[4,5-c]pyridin-3-ol
REFERENCES
■
(1) Berger, M.; Gray, J. A.; Roth, B. L. The expanded biology of
serotonin. Annu. Rev. Med. 2009, 60, 355−366.
(2) Monti, J. M.; Jantos, H. The roles of dopamine and serotonin, and
of their receptors, in regulating sleep and waking. Prog. Brain Res. 2008,
172, 625−646.
(3) Gershon, M. D.; Tack, J. The serotonin signaling system: from
basic understanding to drug development for functional GI disorders.
Gastroenterology 2007, 132, 397−414.
(4) Ramage, A. G.; Villalon, C. M. 5-Hydroxytryptamine and
cardiovascular regulation. Trends Pharmacol. Sci. 2008, 29, 472−481.
(5) Moltzen, E. K.; Bang-Andersen, B. Serotonin reuptake inhibitors:
the corner stone in treatment of depression for half a centurya
medicinal chemistry survey. Curr. Top. Med. Chem. 2006, 6, 1801−1823.
(6) Gray, J. A.; Roth, B. L. The pipeline and future of drug development
in schizophrenia. Mol. Psychiatry 2007, 12, 904−922.
Docking of Ligands into a 5-HT2A Receptor Homology Model.
The THAZ analogues 3d, 4b, 4d, and 4i were docked in our 5-HT2A
receptor homology model.52 A docking grid was generated using Glide
5.8.78 An enclosed box of default size was defined around the Asp155
(D1553.32) residue in the receptor. Docking was carried out using Glide
5.8,78 with XP precision. Ten poses per ligand were kept. For each
ligand, the highest ranked pose where interactions with transmembrane
helices 3 and 5 were observed was selected for each compound.
(7) Katzman, M. A. Current considerations in the treatment of
generalized anxiety disorder. CNS Drugs 2009, 23, 103−120.
(8) Goadsby, P. J. Serotonin receptor ligands: treatments of acute
migraine and cluster headache. Handb. Exp. Pharmacol. 2007, 129−143.
(9) Halford, J. C.; Harrold, J. A.; Boyland, E. J.; Lawton, C. L.; Blundell,
J. E. Serotonergic drugs: effects on appetite expression and use for the
treatment of obesity. Drugs 2007, 67, 27−55.
(10) Kroeze, W. K.; Kristiansen, K.; Roth, B. L. Molecular biology of
serotonin receptors structure and function at the molecular level. Curr.
Top. Med. Chem. 2002, 2, 507−528.
ASSOCIATED CONTENT
■
(11) Hannon, J.; Hoyer, D. Molecular biology of 5-HT receptors.
Behav. Brain Res. 2008, 195−213, 198.
S
* Supporting Information
Supporting Information Table 1: Data from the screening of N-
Bn-THAZ (3d) at numerous putative targets. Supporting
Information Figure 1: A grey scale version of Figure 4D. This
material is available free of charge via the Internet at http://pubs.
acs.org.
(12) Millan, M. J.; Marin, P.; Bockaert, J.; la Cour, C. M. Signaling at G-
protein-coupled serotonin receptors: recent advances and future
research directions. Trends Pharmacol. Sci. 2008, 29, 454−464.
(13) Jensen, A. A.; Davies, P. A.; Brauner-Osborne, H.; Krzywkowski,
̈
K. 3B but which 3B? And that’s just one of the questions: the
1225
dx.doi.org/10.1021/jm301656h | J. Med. Chem. 2013, 56, 1211−1227

Journal of Medicinal Chemistry
Article
heterogeneity of human 5-HT3 receptors. Trends Pharmacol. Sci. 2008,
29, 437−444.
(35) Smith, B. M.; Smith, J. M.; Tsai, J. H.; Schultz, J. A.; Gilson, C. A.;
Estrada, S. A.; Chen, R. R.; Park, D. M.; Prieto, E. B.; Gallardo, C. S.;
Sengupta, D.; Dosa, P. I.; Covel, J. A.; Ren, A.; Webb, R. R.; Beeley, N.
R.; Martin, M.; Morgan, M.; Espitia, S.; Saldana, H. R.; Bjenning, C.;
Whelan, K. T.; Grottick, A. J.; Menzaghi, F.; Thomsen, W. J. Discovery
and structure−activity relationship of (1R)-8-chloro-2,3,4,5-tetrahydro-
1-methyl-1H-3-benzazepine (Lorcaserin), a selective serotonin 5-HT2C
receptor agonist for the treatment of obesity. J. Med. Chem. 2008, 51,
305−313.
(36) Siuciak, J. A.; Chapin, D. S.; McCarthy, S. A.; Guanowsky, V.;
Brown, J.; Chiang, P.; Marala, R.; Patterson, T.; Seymour, P. A.; Swick,
A.; Iredale, P. A. CP-809,101, a selective 5-HT2C agonist, shows activity
in animal models of antipsychotic activity. Neuropharmacology 2007, 52,
279−290.
(37) Porter, R. H.; Benwell, K. R.; Lamb, H.; Malcolm, C. S.; Allen, N.
H.; Revell, D. F.; Adams, D. R.; Sheardown, M. J. Functional
characterization of agonists at recombinant human 5-HT2A, 5-HT2B,
and 5-HT2C receptors in CHO-K1 cells. Br. J. Pharmacol. 1999, 128, 13−
20.
(38) Jerman, J. C.; Brough, S. J.; Gager, T.; Wood, M.; Coldwell, M. C.;
Smart, D.; Middlemiss, D. N. Pharmacological characterisation of
human 5-HT2 receptor subtypes. Eur. J. Pharmacol. 2001, 414, 23−30.
(39) Weiner, D. M.; Burstein, E. S.; Nash, N.; Croston, G. E.; Currier,
E. A.; Vanover, K. E.; Harvey, S. C.; Donohue, E.; Hansen, H. C.;
Andersson, C. M.; Spalding, T. A.; Gibson, D. F.; Krebs-Thomson, K.;
Powell, S. B.; Geyer, M. A.; Hacksell, U.; Brann, M. R. 5-
Hydroxytryptamine2A receptor inverse agonists as antipsychotics. J.
Pharmacol. Exp. Ther. 2001, 299, 268−276.
(14) Roth, B. L.; Willins, D. L.; Kristiansen, K.; Kroeze, W. K. 5-
Hydroxytryptamine2-family receptors (5-hydroxytryptamine2A, 5-hy-
droxytryptamine2B, 5-hydroxytryptamine2C): where structure meets
function. Pharmacol. Ther. 1998, 79, 231−257.
(15) Leysen, J. E. 5-HT2 receptors. CNS Neurol. Disord.: Drug Targets
2004, 3, 11−26.
(16) Gonzalez-Maeso, J.; Sealfon, S. C. Psychedelics and schizo-
phrenia. Trends Neurosci. 2009, 32, 225−232.
(17) Roth, B. L. Irving Page Lecture: 5-HT2A serotonin receptor
biology: interacting proteins, kinases and paradoxical regulation.
Neuropharmacology 2011, 61, 348−354.
(18) Brea, J.; Castro-Palomino, J.; Yeste, S.; Cubero, E.; Parraga, A.;
Dominguez, E.; Loza, M. I. Emerging opportunities and concerns for
drug discovery at serotonin 5-HT2B receptors. Curr. Top. Med. Chem.
2010, 10, 493−503.
(19) Hutcheson, J. D.; Setola, V.; Roth, B. L.; Merryman, W. D.
Serotonin receptors and heart valve diseaseit was meant 2B.
Pharmacol. Ther. 2011, 132, 146−157.
(20) Clemett, D. A.; Punhani, T.; Duxon, M. S.; Blackburn, T. P.; Fone,
K. C. Immunohistochemical localisation of the 5-HT2C receptor protein
in the rat CNS. Neuropharmacology 2000, 39, 123−132.
(21) Serrats, J.; Mengod, G.; Cortes, R. Expression of serotonin 5-
HT2C receptors in GABAergic cells of the anterior raphe nuclei. J. Chem.
Neuroanat. 2005, 29, 83−91.
(22) Jensen, N. H.; Cremers, T. I.; Sotty, F. Therapeutic potential of 5-
HT2C receptor ligands. Sci.World J. 2010, 10, 1870−1885.
(23) Wacker, D. A.; Miller, K. J. Agonists of the serotonin 5-HT2C
receptor: preclinical and clinical progression in multiple diseases. Curr.
Opin. Drug Discovery Dev. 2008, 11, 438−445.
(24) Millan, M. J. Serotonin 5-HT2C receptors as a target for the
treatment of depressive and anxious states: focus on novel therapeutic
strategies. Therapie 2005, 60, 441−460.
(25) Isaac, M. Serotonergic 5-HT2C receptors as a potential therapeutic
target for the design antiepileptic drugs. Curr. Top. Med. Chem. 2005, 5,
59−67.
(26) Halford, J. C.; Harrold, J. A. 5-HT2C receptor agonists and the
control of appetite. Handb. Exp. Pharmacol. 2012, 349−356.
(27) Krogsgaard-Larsen, P.; Frølund, B.; Liljefors, T.; Ebert, B. GABAA
agonists and partial agonists: THIP (Gaboxadol) as a non-opioid
analgesic and a novel type of hypnotic. Biochem. Pharmacol. 2004, 68,
1573−1580.
(28) Krogsgaard-Larsen, P.; Hjeds, H. Organic hydroxylamine
derivatives. X. Structural analogues of γ-aminobutyric acid (GABA) of
the Isoxazole enol-betaine type. Synthesis of 5,6,7,8-tetrahydro-4H-
isoxazolo[4,5-d]azepin-3-ol zwitterion and 4,5,6,7-tetrahydrisoxazolo-
[4,5-c]pyridin-3-ol zwitterion. Acta Chem. Scand. 1974, B28, 533−538.
(29) Krogsgaard-Larsen, P.; Hjeds, H.; Curtis, D. R.; Leah, J. D.; Peet,
M. J. Glycine antagonists structurally related to muscimol, THIP, or
isoguvacine. J. Neurochem. 1982, 39, 1319−1324.
(30) Krogsgaard-Larsen, P.; Mikkelsen, H.; Jacobsen, P.; Falch, E.;
Curtis, D. R.; Peet, M. J.; Leah, J. D. 4,5,6,7-Tetrahydroisothiazolo[5,4-
c]pyridin-3-ol and related analogues of THIP. Synthesis and biological
activity. J. Med. Chem. 1983, 26, 895−900.
(31) Carniaux, J.-F.; Cummins, J. Compositions and methods of
synthesis of pyridinoylpiperidine 5-HT1F agonists. WO/2011/123654
A1, 2011.
(32) Krogsgaard-Larsen, P. Muscimol Analogues. II Synthesis of Some
Bicyclic 3-Isoxazolol Zwitterions. Acta Chem. Scand. 1977, B31, 584−
588.
(33) Schlag, B. D.; Lou, Z.; Fennell, M.; Dunlop, J. Ligand dependency
of 5-hydroxytryptamine 2C receptor internalization. J. Pharmacol. Exp.
Ther. 2004, 310, 865−870.
(34) Acuna-Castillo, C.; Villalobos, C.; Moya, P. R.; Saez, P.; Cassels, B.
K.; Huidobro-Toro, J. P. Differences in potency and efficacy of a series of
phenylisopropylamine/phenylethylamine pairs at 5-HT2A and 5-HT2C
receptors. Br. J. Pharmacol. 2002, 136, 510−519.
(40) Shahid, M.; Walker, G. B.; Zorn, S. H.; Wong, E. H. Asenapine: a
novel psychopharmacologic agent with a unique human receptor
signature. J. Psychopharmacol. 2009, 23, 65−73.
(41) Aloyo, V. J.; Harvey, J. A. Antagonist binding at 5-HT2A and 5-
HT2C receptors in the rabbit: high correlation with the profile for the
human receptors. Eur. J. Pharmacol. 2000, 406, 163−169.
(42) Lameh, J.; McFarland, K.; Ohlsson, J.; Ek, F.; Piu, F.; Burstein, E.
S.; Tabatabaei, A.; Olsson, R.; Bradley, S. R.; Bonhaus, D. W. Discovery
of potential antipsychotic agents possessing pro-cognitive properties.
Naunyn Schmiedeberg’s Arch. Pharmacol. 2012, 385, 313−323.
(43) Thomsen, A. R. B.; Hvidtfeldt, M.; Brauner-Osborne, H. Biased
̈
agonism of the calcium-sensing receptor. Cell Calcium 2012, 51, 107−
116.
(44) Knight, A. R.; Misra, A.; Quirk, K.; Benwell, K.; Revell, D.;
Kennett, G.; Bickerdike, M. Pharmacological characterisation of the
agonist radioligand binding site of 5-HT2A, 5-HT2B and 5-HT2C
receptors. Naunyn Schmiedeberg’s Arch. Pharmacol. 2004, 370, 114−123.
(45) Muntasir, H. A.; Takahashi, J.; Rashid, M.; Ahmed, M.;
Komiyama, T.; Hossain, M.; Kawakami, J.; Nashimoto, M.;
Nagatomo, T. Site-directed mutagenesis of the serotonin 5-hydroxy-
tryptamine2c receptor: identification of amino acids responsible for
sarpogrelate binding. Biol. Pharm. Bull. 2006, 29, 1645−1650.
(46) Xie, J.; Dernovici, S.; Ribeiro, P. Mutagenesis analysis of the
serotonin 5-HT2C receptor and a Caenorhabditis elegans 5-HT2
homologue: conserved residues of helix 4 and helix 7 contribute to
agonist-dependent activation of 5-HT2 receptors. J. Neurochem. 2005,
92, 375−387.
(47) Sleight, A. J.; Stam, N. J.; Mutel, V.; Vanderheyden, P. M.
Radiolabelling of the human 5-HT2A receptor with an agonist, a partial
agonist and an antagonist: effects on apparent agonist affinities. Biochem.
Pharmacol. 1996, 51, 71−76.
(48) Rosenkilde, M. M.; Cahir, M.; Gether, U.; Hjorth, S. A.; Schwartz,
T. W. Mutations along transmembrane segment II of the NK-1 receptor
affect substance P competition with non-peptide antagonists but not
substance P binding. J. Biol. Chem. 1994, 269, 28160−28164.
(49) Hjorth, S. A.; Thirstrup, K.; Schwartz, T. W. Radioligand-
dependent discrepancy in agonist affinities enhanced by mutations in
the kappa-opioid receptor. Mol. Pharmacol. 1996, 50, 977−984.
(50) Sagan, S.; Beaujouan, J. C.; Torrens, Y.; Saffroy, M.; Chassaing,
G.; Glowinski, J.; Lavielle, S. High affinity binding of [3H]propionyl-
[Met(O2)11]substance P(7−11), a tritiated septide-like peptide, in
1226
dx.doi.org/10.1021/jm301656h | J. Med. Chem. 2013, 56, 1211−1227

Journal of Medicinal Chemistry
Article
Chinese hamster ovary cells expressing human neurokinin-1 receptors
and in rat submandibular glands. Mol. Pharmacol. 1997, 52, 120−127.
(51) Rosenkilde, M. M.; Schwartz, T. W. Potency of ligands correlates
with affinity measured against agonist and inverse agonists but not
against neutral ligand in constitutively active chemokine receptor. Mol.
Pharmacol. 2000, 57, 602−609.
(52) Isberg, V.; Balle, T.; Sander, T.; Jørgensen, F. S.; Gloriam, D. E. G
protein- and agonist-bound serotonin 5-HT2A receptor model activated
by steered molecular dynamics simulations. J. Chem. Inf. Model. 2011,
51, 315−325.
(53) Ballosteros, J. A.; Weinstein, H. Integrated methods for the
construction of three dimensional models and computational probing of
structure function relations in G protein-coupled receptors. In Methods
in Neurosciences; Sealfon, S. C., Conn, P. M., Eds.; Academic Press: San
Diego, CA, 1995; Vol. 25, pp 366−428.
(54) Meltzer, H. Y.; Nash, J. F. Effects of antipsychotic drugs on
serotonin receptors. Pharmacol. Rev. 1991, 43, 587−604.
(55) Martin, G. E.; Elgin, R. J., Jr.; Mathiasen, J. R.; Davis, C. B.;
Kesslick, J. M.; Baldy, W. J.; Shank, R. P.; DiStefano, D. L.; Fedde, C. L.;
Scott, M. K. Activity of aromatic substituted phenylpiperazines lacking
affinity for dopamine binding sites in a preclinical test of antipsychotic
efficacy. J. Med. Chem. 1989, 32, 1052−1056.
(56) Martin, G. E.; Elgin, R. J., Jr.; Kesslick, J. M.; Baldy, W. J.;
Mathiasen, J. R.; Shank, R. P.; Scott, M. K. Block of conditioned
avoidance responding in the rat by substituted phenylpiperazines. Eur. J.
Pharmacol. 1988, 156, 223−229.
(57) Meltzer, H. Y.; Massey, B. W.; Horiguchi, M. Serotonin receptors
as targets for drugs useful to treat psychosis and cognitive impairment in
schizophrenia. Curr. Pharm. Biotechnol. 2012, 13, 1572−1586.
(58) Brennan, P. E.; Whitlock, G. A.; Ho, D. K.; Conlon, K.;
McMurray, G. Discovery of a novel azepine series of potent and selective
5-HT2C agonists as potential treatments for urinary incontinence. Bioorg.
Med. Chem. Lett. 2009, 19, 4999−5003.
(59) Nichols, D. E. Structure−activity relationships of serotonin 5-
HT2A agonists. WIREs: Membr. Transp. Signaling 2012, DOI: 10.1002/
wmts.42.
Structure−activity relationships of selective GABA uptake inhibitors.
Curr. Top. Med. Chem. 2006, 6, 1861−1882.
(69) Gottlieb, H. E.; Kotlyar, V.; Nudelman, A. NMR Chemical Shifts
of Common Laboratory Solvents as Trace Impurities. J. Org. Chem.
1997, 62, 7512−7515.
(70) Jensen, A. A.; Christesen, T.; Bølcho, U.; Greenwood, J. R.;
Postorino, G.; Vogensen, S. B.; Johansen, T. N.; Egebjerg, J.; Brauner-
̈
Osborne, H.; Clausen, R. P. Functional characterization of Tet-AMPA
[tetrazolyl-2-amino-3-(3-hydroxy-5-methyl- 4-isoxazolyl)propionic
acid] analogues at ionotropic glutamate receptors GluR1-GluR4. The
molecular basis for the functional selectivity profile of 2-Bn-Tet-AMPA.
J. Med. Chem. 2007, 50, 4177−4185.
(71) Trattnig, S. M.; Harpsøe, K.; Thygesen, S. B.; Rahr, L. M.; Ahring,
P. K.; Balle, T.; Jensen, A. A. Discovery of a novel allosteric modulator of
5-HT3 receptors: inhibition and potentiation of Cys-loop receptor
signaling through a conserved transmembrane intersubunit site. J. Biol.
Chem. 2012, 287, 25241−25254.
(72) Krogsgaard-Larsen, N.; Jensen, A. A.; Kehler, J. Novel 7-
phenylsulfanyl-1,2,3,4,10,10a-hexahydro-pyrazino[1,2-a]indoles as dual
serotonin 5-HT2C and 5-HT6 receptor ligands. Bioorg. Med. Chem. Lett.
2010, 20, 5431−5433.
(73) Ransom, R. W.; Stec, N. L. Cooperative modulation of [3H]MK-
801 binding to the N-methyl-D-aspartate receptor-ion channel complex
by L-glutamate, glycine, and polyamines. J. Neurochem. 1988, 51, 830−
836.
(74) Paxinos, G.; Watson, C. The Rat Brain in Stereotaxic Coordinates.
4th ed.; Academic Press: San Diego, 1998.
(75) Wellendorph, P.; Høg, S.; Sabbatini, P.; Pedersen, M. H.; Martiny,
L.; Knudsen, G. M.; Frølund, B.; Clausen, R. P.; Brauner-Osborne, H.
̈
Novel radioiodinated γ-hydroxybutyric acid analogues for radiolabeling
and photolinking of high-affinity γ-hydroxybutyric acid binding sites. J.
Pharmacol. Exp. Ther. 2010, 335, 458−464.
(76) Dellu, F.; Fauchey, V.; Le Moal, M.; Simon, H. Extension of a new
two-trial memory task in the rat: influence of environmental context on
recognition processes. Neurobiol. Learn. Mem. 1997, 67, 112−120.
(77) Dellu, F.; Mayo, W.; Vallee, M.; Le Moal, M.; Simon, H.
Facilitation of cognitive performance in aged rats by past experience
depends on the type of information processing involved: a combined
cross-sectional and longitudinal study. Neurobiol. Learn. Mem. 1997, 67,
121−128.
(60) Trachsel, D.; Nichols, D. E.; Kidd, S.; Hadorn, M.; Baumberger, F.
4-Aryl-substituted 2,5-dimethoxyphenethylamines: synthesis and sero-
tonin 5-HT2A receptor affinities. Chem Biodiversity 2009, 6, 692−704.
(61) Nichols, D. E. Hallucinogens. Pharmacol. Ther. 2004, 101, 131−
181.
(62) Sellin, A. K.; Shad, M.; Tamminga, C. Muscarinic agonists for the
treatment of cognition in schizophrenia. CNS Spectrosc. 2008, 13, 985−
996.
(78) Friesner, R. A.; Banks, J. L.; Murphy, R. B.; Halgren, T. A.; Klicic, J.
J.; Mainz, D. T.; Repasky, M. P.; Knoll, E. H.; Shelley, M.; Perry, J. K.;
Shaw, D. E.; Francis, P.; Shenkin, P. S. Glide: a new approach for rapid,
accurate docking and scoring. 1. Method and assessment of docking
accuracy. J. Med. Chem. 2004, 47, 1739−1749.
(63) Dellu, F.; Mayo, W.; Cherkaoui, J.; Le Moal, M.; Simon, H. A two-
trial memory task with automated recording: study in young and aged
rats. Brain Res. 1992, 588, 132−139.
(64) Calcagno, E.; Carli, M.; Baviera, M.; Invernizzi, R. W. Endogenous
serotonin and serotonin2C receptors are involved in the ability of
M100907 to suppress cortical glutamate release induced by NMDA
receptor blockade. J. Neurochem. 2009, 108, 521−532.
(65) Marquis, K. L.; Sabb, A. L.; Logue, S. F.; Brennan, J. A.; Piesla, M.
J.; Comery, T. A.; Grauer, S. M.; Ashby, C. R., Jr.; Nguyen, H. Q.;
Dawson, L. A.; Barrett, J. E.; Stack, G.; Meltzer, H. Y.; Harrison, B. L.;
Rosenzweig-Lipson, S. WAY-163909 [(7bR,10aR)-1,2,3,4,8,9,10,10a-
Octahydro-7bH-cyclopenta-[b][1,4]diazepino[6,7,1hi]indole]: a novel
5-hydroxytryptamine 2C receptor-selective agonist with preclinical
antipsychotic-like activity. J. Pharmacol. Exp. Ther. 2007, 320, 486−496.
(66) Landolt, H. P.; Wehrle, R. Antagonism of serotonergic 5-HT2A/
2C receptors: mutual improvement of sleep, cognition and mood? Eur. J.
Neurosci. 2009, 29, 1795−1809.
(67) Frydenvang, K.; Pickering, D. S.; Greenwood, J. R.; Krogsgaard-
Larsen, N.; Brehm, L.; Nielsen, B.; Vogensen, S. B.; Hald, H.; Kastrup, J.
S.; Krogsgaard-Larsen, P.; Clausen, R. P. Biostructural and Pharmaco-
logical Studies of Bicyclic Analogues of the 3-Isoxazolol Glutamate
Receptor Agonist Ibotenic Acid. J. Med. Chem. 2010, 53, 8354−8361.
(68) Høg, S.; Greenwood, J. R.; Madsen, K. B.; Larsson, O. M.;
Frølund, B.; Schousboe, A.; Krogsgaard-Larsen, P.; Clausen, R. P.
1227
dx.doi.org/10.1021/jm301656h | J. Med. Chem. 2013, 56, 1211−1227