Discovery of β-Arrestin Biased Ligands of 5-HT7R

Article abstract of DOI:10.1021/acs.jmedchem.8b00642

Though many studies have been published about therapeutic potentials of selective 5-HT₇R ligands, there have been few biased ligands of 5-HT₇R. The development of potent and selective biased ligands of 5-HT₇R would be of great help in understanding the relationship between pharmacological effects and G protein/β-arrestin signaling pathways of 5-HT₇R. In order to identify 5-HT₇R ligands with biased agonism, we designed and synthesized a series of tetrahydroazepine derivatives 1 and 2 with arylpyrazolo moiety or arylisoxazolo moiety. Through several biological evaluations such as binding affinity, selectivity profile, and functions in G protein and β-arrestin signaling pathways, 3-(4-chlorophenyl)-1,4,5,6,7,8-hexahydropyrazolo[3,4-d]azepine 1g was discovered as the β-arrestin biased ligand of 5-HT₇R. In an electroencephalogram (EEG) test, 1g increased total non-rapid eye movement (NREM) sleep time and decreased total rapid eye movement (REM) sleep time.

Full text of DOI:10.1021/acs.jmedchem.8b00642

Subscriber access provided by University of Sussex Library
Article
Discovery of #-arrestin biased ligands of 5-HT7R
Youngjae Kim, Hyunguk Kim, Jieon Lee, Jae Kyun Lee, Sun-Joon Min, Jihye
Seong, Hyewhon Rhim, Jinsung Tae, Hyunjoo Jenny Lee, and Hyunah Choo
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b00642 • Publication Date (Web): 20 Jul 2018
Downloaded from http://pubs.acs.org on July 24, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery of βꢀarrestin biased ligands of 5ꢀHT R
7
†
,‡
§
†,
ꢀ
†
∇
Youngjae Kim, Hyunguk Kim, Jieon Lee, Jae Kyun Lee, SunꢁJoon Min,
†
,
ꢀ
,
#
ꢀ
,
‡
,§
,†,
ꢀ
¶
Jihye Seong,
Hyewhon Rhim, Jinsung Tae, Hyunjoo Jenny Lee* , Hyunah Choo*
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
†
Center for NeuroꢁMedicine, Brain Research Institute, Korea Institute of Science and
Technology, Seoul 02792, Republic of Korea
Department of Chemistry, Yonsei University, Seoul 03722, Republic of Korea
School of Electrical Engineering, Korea Advanced Institute of Science and Technology,
Daejeon 34141, Republic of Korea
Division of BioꢁMedical Science & Technology, KIST School, Korea University of Science
∇and Technology, Seoul 02792, Republic of Korea
Department of Chemical & Molecular Engineering/Applied Chemistry, Hanyang University,
Ansan, Gyeonggiꢁdo 15588, Republic of Korea
Convergence Research Center for Diagnosis Treatment Care of Dementia, Korea Institute of
Science and Technology, Seoul 02792, Republic of Korea
Center for Neuroscience, Brain Research Institute, Korea Institute of Science and
Technology, Seoul 02792, Republic of Korea
‡
§
ꢀ
#
¶
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Abstract. Though many studies have been published about therapeutic potentials of selective
5ꢁHT R ligands, there have been few biased ligands of 5ꢁHT R. The development of potent
7
7
and selective biased ligands of 5ꢁHT
7
R would be of great help in understanding the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
relationship between pharmacological effects and G protein/βꢁarrestin signaling pathways of
ꢁHT R. In order to identify 5ꢁHT R ligands with biased agonism, we designed and
5
7
7
synthesized a series of tetrahydroazepine derivatives 1 and 2 with arylpyrazolo moiety or
arylisoxazolo moiety. Through several biological evaluations such as binding affinity,
selectivity profile, and functions in G protein and βꢁarrestin signaling pathways, (3ꢁ(4ꢁ
chlorophenyl)ꢁ1,4,5,6,7,8ꢁhexahydropyrazolo[3,4ꢁd]azepine) 1g was discovered as the βꢁ
7
arrestin biased ligand of 5ꢁHT R. In electroencephalogram (EEG) test, 1g increased total nonꢁ
rapid eye movement (NREM) sleep time and decreased total rapid eye movement (REM)
sleep time.
2
ACS Paragon Plus Environment

Page 3 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
■ INTRODUCTION
G proteinꢁcoupled receptors (GPCRs) are seven transmembrane helical proteins and there are
1ꢁ4
over 800 members of GPCRs in the human genome with diverse structures and functions.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
These receptors are generally involved in regulation of various physiological signal
transductions through G proteins stimulated by diverse ligands including biogenic amines,
5
ꢁ7
amino acids, and peptides. Since GPCRs mediate many cellular signaling pathways, they
have become targets of numerous therapeutic drugs and still considered as one of the most
8
important classes of pharmacological targets. However, there have been difficulties in
elucidating in vivo pharmacological effects of GPCR ligands through an in vitro assay for G
9
ꢁ11
protein activation, which resulted in limitations in discovering GPCR drugs.
As GPCR
biology advances, these problems have been shown to be associated with the complexity of
the GPCRꢁmediated signaling. Particularly, not only G proteins but also nonꢁG protein
effectors such as βꢁarrestin are involved in GPCR signaling. Thus, to understand the
mechanism of GPCR drugs and to develop more efficacious drugs with fewer side effects,
ligands with functional selectivity or signaling bias has been actively pursued. The soꢁcalled
“
functionally selective” or “biased” ligands are believed to stimulate either G protein
signaling pathway or βꢁarrestin signaling pathway independently by binding and stabilizing
different conformations of GPCR resulting in ligandꢁspecific cellular responses.
5
ꢁHT receptor (5ꢁHT R) is the most recently identified serotonin receptor among serotonin
7 7
receptor subtypes, which is mainly detected in central nervous system (CNS) including
prefrontal cortex, spinal cord, thalamus, hippocampus, hypothalamus, and suprachiasmatic
12
nucleus. There are four isoforms of 5ꢁHT
7
R such as 5ꢁHT7aR, 5ꢁHT7bR, 5ꢁHT7cR, and 5ꢁ
HT7dR. All of 5ꢁHT R isoforms belong to GPCRs and are coupled to Gs protein, which
7
activate adenylyl cyclase (AC) with a resultant increase of the intracellular level of a
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
3ꢁ18
secondary messenger cyclic AMP (cAMP).
7
The pathophysiological function of 5ꢁHT R is
known to be associated with circadian rhythm, learning and memory as well as several
diseases such as migraine, mental fatigue, sleep disturbance, anxiety, depression, and
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
19ꢁ21
schizophrenia.
Until now, many potent 5ꢁHT
7
R ligands have been reported and classified according to
22,23
activities on G protein signaling pathway, the two main classes being agonists (ASꢁ19,
2
4
25
26
27
28,29
E55888, 8OHꢁDPAT, and LPꢁ211 ) and antagonists (SB269970, SB258719,
3
0
31
DR4485, and JNJ18038683 ) (Figure 1). It was reported that 5ꢁHT
7
R antagonists,
SB269970 and JNJ18038683, showed antidepressantꢁlike activities in forced swimming test
31
(
FST). In addition, administration of the 5ꢁHT
7
R antagonists caused modulation of sleep
patterns, especially decrease of rapid eye movement (REM) duration and increase of REM
32
latency. However, there have been controversial results about in vivo effects of 5ꢁHT R
7
agonists and antagonists. LPꢁ211, a 5ꢁHT
7
R agonist, induced REM sleep suppression like 5ꢁ
while both LPꢁ211 and SB269970 (an antagonist) exhibited anxiolyticꢁ
Although the role of 5ꢁHT R may depend on brain regions and neurochemical
environment, the controversial pharmacological effects of 5ꢁHT R ligands might be
explained by concurrent elucidation of their activities on βꢁarrestin signaling pathway: 5ꢁ
HT R ligands could be evaluated on signaling pathways of G protein and βꢁarrestin, and
categorized as G protein biased ligands, βꢁarrestin biased ligands, and balanced
agonists/antagonists on both pathways. It was reported that 5ꢁHT R indeed activates not only
33,34
7
HT R antagonists,
35,36
like effects.
7
7
7
7
37
G protein signaling pathway but also βꢁarrestin signaling pathway.
Recently, ergotamine and a compound with indoleꢁaripiprazole moiety were reported as βꢁ
38,39
arrestin biased ligand of 5ꢁHT
7
R,
which, however, were not selective 5ꢁHT
7
R ligands.
Biased ligands which selectively bind to 5ꢁHT
7
R have not been developed yet. Thus,
4
ACS Paragon Plus Environment

Page 5 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
7
developing potent and selective biased ligands against 5ꢁHT R would be of great help in
understanding the relationship between physiological effects and G protein/βꢁarrestin
signaling pathways of 5ꢁHT R. In this study, we designed and synthesized tetrahydroazepine
derivatives 1 and 2 to discover novel G protein or βꢁarrestin biased ligands, resulting in
discovery of the βꢁarrestin biased ligands of 5ꢁHT R which modulated sleep patterns like SBꢁ
69970.
7
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
2
7
Figure 1. Structures of representative 5ꢁHT R agonists and antagonists
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
■
RESULTS AND DISCUSSION
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Due to the lack of 5ꢁHT
7
R biased ligands, the approach to develop a novel biased ligand
R ligands. Two
entailed of identifying the common structural features of the known 5ꢁHT
7
agonists (ASꢁ19 and E55888) and one antagonist (JNJ18038683) (Figure 2) were chosen as
the representatives for 5ꢁHT R ligands, which showed two common features such as a biaryl
7
group (red color in Figure 2) and ethyleneamine moiety (blue in Figure 2). Based on these
common
features,
we
designed
tetrahydropyrazoloazepines
1
and
tetrahydroisoxazoloazepines 2 which have various aryl substituent (R) including hydrogen,
halogen and methyl group.
Synthesis. The tetrahydropyrazoloazepine derivatives 1 with R substituents were synthesized
40
in total 6 steps (R = H, 2ꢁF, 3ꢁF, 4ꢁF, 2ꢁCl, 3ꢁCl, 4ꢁCl, 2ꢁMe, 3ꢁMe, and 4ꢁMe) (Scheme 1).
1ꢁBocꢁ4ꢁpiperidone 3 underwent ring expansion by reaction with ethyldiazoacetate and
BF ·OEt at ꢁ20°C to give oxoazepane 4, which was reacted with hydrazine to afford
3
2
dihydropyrazolone 5 (97% yield in two steps). The dihydropyrazolone 5 underwent Boc
protection of the secondary amino group in dihydropyrazolone moiety in 63% yield, and the
2
resulting compound 6 was treated with Nꢁphenylꢁbis(trifluoromethanesulfonimide) (Tf NPh)
under basic conditions to give a triflate 7 in 93% yield. We confirmed the structure of the
compound 7 through Xꢁray crystallographic analysis (Supplementary Figure S1). Suzuki
coupling reactions between 7 and variously substituted phenylboronic acids were carried out
by using palladium catalyst, Pd(dppf)Cl , to afford Bocꢁprotected tetrahydropyrazoloazepines
2
8a~8j in 11ꢁ47% yields. Finally, 8a~8j were treated with 1N HCl to obtain the desired aryl
tetrahydropyrazoloazepines 1a~1j in 16ꢁ81% yields.
6
ACS Paragon Plus Environment

Page 7 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. Designed tetrahydropyrazoloazepines 1 and tetrahydroisoxazoloazepines 2
We also synthesized the aryl tetrahydroisoxazoloazepine derivatives 2 with various
substituents (R) in 7 steps starting from ethyl 4ꢁoxopiperidineꢁ1ꢁcarboxylate 9 (Scheme 2).
The 4ꢁoxopiperidine 9 underwent ring expansion followed by acetal protection to give 11 in
46% yield in two steps. The compound 11 was treated with hydroxylamine to afford 12 in 31%
yield, which underwent deprotection of acetal group followed by subsequent cyclization
40
under acidic conditions to give 3ꢁhydroxyisoxazole 13 in 94% yield.
The 3ꢁ
hydroxyisoxazole 13 was treated with TF NPh to give the corresponding triflate 14 in 86%
2
yield. Suzuki coupling of 14 with various phenylboronic acids afforded aryl
tetrahydroisoxazoloazepines 15a~15j in 19ꢁ33% yields. After deprotection of ethyl carbamate
group by treatment with HBr in acetic acid, the final compounds 2a~2j were obtained in 15ꢁ
70% yields.
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
a
Scheme 1. Synthesis of tetrahydropyrazoloazepine 1aꢀ1j
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Reagents and conditions: a) ethyldiazoacetate, BF
O, TEA, DCM, DMF, rt, 63%; d) Tf
, Pd(dppf)Cl , K PO , THF, reflux, 11ꢁ47%; f) i) 1 N HCl in
O, DCM, MeOH, ii) 10 N NaOH, rt, 16ꢁ81%.
3
OEt
2
, Et
2
O, ꢁ20 °C to rt; b) hydrazine,
EtOH, reflux, 97% in 2 steps; c) Boc
DCM, rt, 93%; e) RꢁPhB(OH)
Et
2
2
NPh, DIPEA,
2
2
2
4
2
8
ACS Paragon Plus Environment

Page 9 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
Scheme 2. Synthesis of tetrahydroisoxazoloazepines 2aꢁ2j
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
3 2 2
Reagents and conditions: a) ethyldiazoacetate, BF OEt , Et O, ꢁ20 °C to rt; b) ethylene
glycol, pꢁTsOH, benzene, reflux, 46% in 2 steps; c) NH OH/HCl, Na, MeOH, 40 °C to rt,
2
31%; d) HCl, MeOH, reflux, 94%; e) Tf
Pd(dppf)Cl , K PO , KBr, THF, reflux, 19ꢁ33%; g) i) HBr in AcOH, ii) 1 N NaOH, rt, 15ꢁ
0%.
2 2
NPh, DIPEA, DCM, rt, 86%; f) RꢁPhB(OH) ,
2
2
4
7
Binding affinities. Binding affinities (K ’s) of the synthesized tetrahydropyrazoloazepines 1
i
3
and tetrahydroisoxazoloazepines 2 against 5ꢁHT
7
R were determined by [ H]LSD radioligand
binding assay in transfected HEK293 cells (Table 1). SB269970 was used as a reference
compound. Compound 1a without any substituent in phenyl moiety showed moderate binding
i
affinity with a K value of 322 nM. Among 1bꢁ1j (R = F, Cl, Me), the orthoꢁsubstituted
compounds 1b, 1e, and 1h showed moderate binding affinities with K values of 714 nM, 386
i
nM, and 881 nM, respectively. Compounds 1c, 1f, and 1i with R’s at metaꢁposition were less
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
i
active than the corresponding orthoꢁsubstituted congeners, and they showed K values of
1,243 nM, 883 nM, and 1,436 nM, respectively. On the other hands, the paraꢁsubstituted
compounds 1d, 1g, and 1j showed significantly improved binding affinities with K values of
04 nM, 30 nM, and 47 nM, respectively. These results clearly show that the position of a
i
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
7
substituent in phenyl ring is very important in the binding affinity to 5ꢁHT R. Specifically,
introduction of a substituent at the paraꢁposition of the aromatic ring potentiated binding of
the resulting compounds (1d, 1g, and 1j) to 5ꢁHT R. Among the paraꢁsubstituted compounds,
7
1g (R = pꢁCl) and 1j (R = pꢁMe) showed better binding affinities than the paraꢁfluorinated
compound 1d. Considering the similar size of chlorine and methyl group, it could be
7
suggested that size of the paraꢁsubstituent matters for binding to 5ꢁHT R.
Next, we introduced an isoxazole group instead of the pyrazole group to investigate changes
in binding affinity to 5ꢁHT R. Most of the compounds 2 showed moderate to potent binding
7
7 i i
affinities against 5ꢁHT R with exception of 2h (K > 10,000 nM) and 2c (K = 1,407 nM). As
in the case of 1, paraꢁsubstituted compounds 2d, 2g, and 2j showed much better binding
affinities than orthoꢁsubstituted (2b, 2e, and 2h) and metaꢁsubstituted (2c, 2g, and 2j)
i
counterparts. The binding affinities (K ’s) of paraꢁsubstituted compounds 2d, 2g, and 2j were
53 nM, 70 nM, and 16 nM, respectively. Like 1a, the compound 2a with no substitution at the
phenyl group showed only moderate binding affinity against 5ꢁHT
nM.
7 i
R with a K value of 319
1
0
ACS Paragon Plus Environment

Page 11 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 1. Binding affinities of tetrahydropyrazoloazepines 1 and tetrahydroisoxazoloazepines
2
7
to 5ꢁHT R
5
7
ꢁHT R
Compound
X
R
a
b
pKi
K in nM
i
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1a
1b
H
2F
3F
6.49 ± 0.07
6.15 ± 0.05
5.91 ± 0.05
6.98 ± 0.06
6.41 ± 0.05
6.05 ± 0.05
7.50 ± 0.10
6.05 ± 0.05
5.84 ± 0.06
7.33 ± 0.06
6.50 ± 0.06
6.44 ± 0.08
5.85 ± 0.06
7.28 ± 0.06
6.37 ± 0.05
6.43 ± 0.05
322
714
1,243
104
386
883
30
881
1,436
47
319
366
1,407
53
425
369
1c
1d
4F
1e
2Cl
3Cl
4Cl
2Me
3Me
4Me
H
2F
3F
4F
2Cl
3Cl
4Cl
2Me
3Me
4Me
NH
1f
1g
1h
1i
1j
2a
2b
2c
2d
2e
O
2f
2g
7.15 ± 0.05
70
c
ꢁꢁ
c
ꢁꢁ
2h
2i
2j
6.30 ± 0.05
7.79 ± 0.05
9.47 ± 0.05
506
16
0.34
SB269970
a
Values are the mean ± SEM of at least three independent experiments performed in triplicate.
Values calculated from pKi.
b
c
7
Not determined due to <50% inhibition at 10 ꢂM of the compound against 5ꢁHT R.
7
Selectivity Profiles. The best 5ꢁHT R binders (1d, 1g, 1j, 2d, 2g, and 2j) were then evaluated
for their selectivity over other serotonin receptor subtypes such as 5ꢁHT1AR, 5ꢁHT1BR, 5ꢁ
HT R, 5ꢁHT R, 5ꢁHT R, 5ꢁHT R, 5ꢁHT R, 5ꢁHT R, 5ꢁHT R, 5ꢁHT R, and 5ꢁHT .
1
D
1E
2A
2B
2C
3
4
5A
6R
The selectivity for 5ꢁHT R over other serotonin receptor subtypes was expressed as
7
selectivity index (SI) which can be calculated by dividing K
subtypes by K for 5ꢁHT R (Table 2). Compound 1d showed good selectivity (SI > 5) over 5ꢁ
HT1AR, 5ꢁHT1BR, 5ꢁHT1DR, 5ꢁHT1ER, 5ꢁHT R, 5ꢁHT R, and 5ꢁHT5AR. However, 1d
exhibited no selectivity (SI < 2) or only moderate (2 < SI < 5) 5ꢁHT R selectivity over 5ꢁ
i
for other serotonin receptor
i
7
3
4
7
1
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
HT2AR, 5ꢁHT2BR, 5ꢁHT2CR, and 5ꢁHT
6
R. Compound 1g was very selective to almost 5ꢁHTR
subtypes except 5ꢁHT R, over which it showed only moderate selectivity for 5ꢁHT R.
2
B
7
Compound 1j also showed good selectivity over 5ꢁHTR subtypes, except 5ꢁHT2AR, 5ꢁHT2BR,
and 5ꢁHT R. In the case of the isoxazoles, compound 2d showed the best selectivity profiles
with all SI values over 5, while compounds 2g and 2j showed good selectivity to almost 5ꢁ
HTR subtypes, except 5ꢁHT2BR, 5ꢁHT2CR, and/or 5ꢁHT R. Compound 2g showed moderate
selectivity over 5ꢁHT R and 5ꢁHT R, and compound 2j showed very low selectivity over 5ꢁ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
6
3
2
B
3
HT2CR with an SI value of 1.3. Based on selectivity profiles, the compounds 1g and 2d were
considered as the most potent and selective ligands for 5ꢁHT R.
7
1
2
ACS Paragon Plus Environment

Page 13 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 2. Selectivities of compounds 1d, 1g, 1j, 2d, 2g, and 2j over other serotonin receptor subtypes
1
d
1g
1j
2d
2g
2j
5
ꢁHTR
subtypes
ꢁHT
b
SI
b
SI
b
SI
b
SI
b
SI
b
SI
a
a
a
a
a
a
K in nM
i
K in nM
i
K in nM
i
K in nM
i
K in nM
i
K in nM
i
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
1
.0
c
16
1.0
c
5
7
104
1.0
30
1.0
c
47
1.0
c
53
1.0
c
70
>
>
10,000
10,000
> 10,000
> 10,000
> 10,000
> 10,000
5ꢁHT1A
1,866
> 10,000
636
18.0
ꢁꢁ
ꢁꢁ
ꢁꢁ
c
ꢁꢁ
ꢁꢁ
c
c
c
c
c
c
c
> 10,000
> 10,000
> 10,000
> 10,000
> 10,000
> 10,000
> 10,000
> 10,000
147
5
ꢁHT1B
ꢁHT1D
ꢁꢁ
ꢁꢁ
ꢁꢁ
c
ꢁꢁ
c
ꢁꢁ
ꢁꢁ
ꢁꢁ
70.5
9.2
111.7
8.6
5
6.1
7.6
1.7
0.9
2,112
1,366
187
165
890
159
106
3.5
18.9
3.4
ꢁꢁ
ꢁꢁ
4
5.5
1,787
138
5ꢁHT1E
793
ꢁꢁ
6.2
3.3
5.7
3.5
5ꢁHT2A
5ꢁHT2B
5ꢁHT2C
175
471
505
785
703
8.9
9.5
399
247
1
00
150
97
2.3
9.4
1.3
799
26.6
c
ꢁꢁ
c
15.7
20
200
1.9
c
296
6.3
c
14.8
1,099
>
10,000
> 10,000
> 10,000
> 10,000
248
> 10,000
420
26.3
5
ꢁHT3
ꢁHT
ꢁHT5A
ꢁHT
ꢁꢁ
ꢁꢁ
13.3
303
4.3
c
c
ꢁꢁ
c
c
ꢁꢁ
c
c
c
> 10,000
> 10,000
> 10,000
109
> 10,000
> 10,000
5
4
ꢁꢁ
> 10,000
1,231
ꢁꢁ
ꢁꢁ
> 10,000
1,135
ꢁꢁ
c
c
>
10,000
416
70.9
5
ꢁꢁ
ꢁꢁ
ꢁꢁ
23.2
ꢁꢁ
c
8.3
2.3
5
6
4.0
> 10,000
360
5.1
492
30.8
ꢁꢁ
a
b
c
Values are the mean of at least three independent experiments performed in triplicate.
Selectivity indexes = K for other serotonin receptor subtype/Ki for 5ꢁHT R.
value of the compound against each 5ꢁHTR.
i
7
Not determined due to > 10 ꢂM K
i
1
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Activities on G Protein/βꢀArrestin Signaling Pathway. We tested functional activities of G
protein pathway and βꢁarrestin pathway, respectively, of the selected compounds 1d, 1g, 1j,
2d, 2g, and 2j. Bioluminescenceꢁbased assays of all selected ligands have been carried out by
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
1,42
measuring Gsꢁmediated cAMP production (cAMP assay)
and βꢁarrestin recruitment
R antagonist
In cAMP assay for measuring G protein activation, all
(
Tango assay). An endogenous agonist 5ꢁHT (serotonin) and a selective 5ꢁHT
7
3
7,43
SB269970 were used as controls.
compounds tested (1d, 1g, 1j, 2d, 2g, and 2j) acted as inverse agonists with the same result of
SB269970, while the endogenous agonist 5ꢁHT showed agonistic effect in G protein pathway
(Figure 3a). In 5ꢁHT (100 nM)ꢁstimulated antagonist assay, the test compounds showed
moderate antagonistic potencies with IC50 values of from 844 nM to 7,788 nM (Figure 3b,
Table 3). The IC50 value means the concentration of a compound as antagonist required to
produce 50% reduction of maximal response of a control agonist (in this receptor, 5ꢁHT).
Maximum %ꢁinhibition (Imax) was calculated based on maximum inhibition by SB269970 as
100%. Because all the Imax values were over 94.76%, the compounds 1d, 1g, 1j, 2d, 2g, and
2j could be defined as full antagonists in G protein signaling pathway.
Tango assay is an experimental tool to monitor protein interactions between GPCR and βꢁ
44
arrestin in a living cell, and this assay requires interaction of the two proteins to stimulate a
cellular response. For Tango assay, HEK293ꢁderived cell line containing stable interactions of
a tTAꢁdependent luciferase reporter and a βꢁarrestin2ꢁTEV fusion gene (HTLA cells) was
used. SB269980 showed a slight increase in the cellular response in Tango assay with
negligible (less than 20%) activation, which cannot be considered as an agonist in βꢁarrestin
signaling pathway (Figure 3c). However, all the test compounds (1d, 1g, 1j, 2d, 2g, and 2j)
showed partial agonistic effect in Tango assays with EC50 values of 188 nM, 162 nM, 165 nM,
135 nM, 113 nM, 40.8 nM, and 48 nM, respectively, and with Emax values of between 38.23%
1
4
ACS Paragon Plus Environment

Page 15 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
and 62.75% (Figure 3c, Table 3). They showed no antagonistic activities in 5ꢁHT (1 ꢂM)ꢁ
stimulated Tango assays, but SB269970 acted as an antagonist in Tango assay.
Taken together, Compounds 1d, 1g, 1j, 2d, 2g, and 2j with tetrahydropyrazoloazepine or
tetrahydroisoxazoloazepine scaffold act as potent partial agonists in βꢁarrestin signaling
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
pathway of 5ꢁHT R, and as full antagonists in G protein signaling pathway. Among them,
compound 1g, 3ꢁ(4ꢁchlorophenyl)ꢁ1,4,5,6,7,8ꢁhexahydropyrazolo[3,4ꢁd]azepine, is the most
effective βꢁarrestin partial agonist (Emax 62.75%, EC50 162 nM), which is the βꢁarrestin biased
7
ligand of 5ꢁHT R.
Figure 3. Doseꢁresponse curves of a) cAMP agonist assay, b) cAMP antagonist assay, c)
Tango agonist assay, and d) Tango antagonist assay of compounds 1d, 1g, 1j, 2d, 2g, and 2j.
Activation of Gs protein was determined by bioluminescenceꢁbased cAMP assay using Dꢁ
luciferin. Tango assay was selected to determine potency and efficacy for βꢁarrestin
recruitment. Data were normalized relative to the response of 5ꢁHT (100%) or SB269970
(100% at antagonist assays) and vehicle (0%) and represent mean ± SEM from three
independent experiments, each performed at least in triplicate.
1
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 3. IC50 and Imax values of compounds 1d, 1g, 1j, 2d, 2g, and 2j in cAMP assay and
their EC50 and Emax values in Tango assay
cAMP
βꢁarrestin
Compound
a
b
a
b
pIC50
IC (nM)
50
Imax (%)
pEC50
EC (nM)
50
Emax (%)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1d
5.16 ± 0.10
5.11 ± 0.08
5.59 ± 0.05
5.61 ± 0.06
5.40 ± 0.07
6.07 ± 0.07
6,978
7,788
2,556
2,463
3,973
844
98.41
94.76
97.41
97.96
94.92
95.79
6.73 ± 0.15
6.79 ± 0.16
6.78 ± 0.11
6.87 ± 0.13
6.95 ± 0.21
7.39 ± 0.12
7.32 ± 0.06
188
162
165
135
113
40.8
48
53.86
62.75
50.94
38.23
47.11
40.33
100
1g
1j
2d
2g
2j
c
c
c
5ꢁHT
ꢁꢁ
ꢁꢁ
0.9
ꢁꢁ
100.5
c
c
SB269970 9.03 ± 0.04
Values are the mean ± SEM of three independent experiments performed at least in triplicate.
Values calculated from pIC or pEC
ꢁꢁ
ꢁꢁ
< 20%
a
b
5
0
50
.
c
Not determined.
Molecular Docking Study. We analyzed docking poses of compound 1g into the binding site
of 5ꢁHT R to identify residues implicated in βꢁarrestin signaling pathway in relation to G
7
protein signaling pathway. We constructed a comparative model structure of 5ꢁHT R by the
7
45ꢁ47
SwissModel server using the 5ꢁHT1B receptor structure (PDB: 4IAR) as a template.
Recently, it has been uncovered that signal biases of several kinds of aminergic GPCRs are
induced by competitive interactions between a ligand and Ile residue in extracellular loop 2
38,48
(EL2) or Ser residue in transmembrane helix 5 (TM5).
That is, the closer the distance
between a ligand and EL2 is, the more βꢁarrestin biased activity is enhanced. Also, a
hydrogen bond of a ligand with Ser residue in TM5 stabilizes the conformation, resulting in
moving intracellular loop 2 (IL2) and TM6 regions that are involved in G protein activation.
Until now, there has been no biased ligand of 5ꢁHT
between a biased ligand and the binding site of 5ꢁHT
tried to explain the βꢁarrestin bias of 1g by specifying residues in the binding site of 5ꢁHT R.
7
R and thus, no explanation of interactions
7
R. In this molecular docking study, we
7
49ꢁ53
According to the literature,
7
the binding site of 5ꢁHT R consisted of two hydrophobic
binding pockets, HPP1 (Val163/Thr214/Phe344) and HPP2 (Leu232/Ile233/Arg367), and an
1
6
ACS Paragon Plus Environment

Page 17 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ionic interaction residue (Asp162) (Figure 4). An agonist generally docked into only one
hydrophobic pocket (HPP1), while an antagonist occupied both the hydrophobic pockets
HPP1 and HPP2. In docking study of compound 1g, it occupied only one hydrophobic pocket
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(HPP1) of the two hydrophobic pockets by forming ionic interaction with Asp162 and πꢁalkyl
interaction with Ile233 in EL2 (Figure 4, 5a). Interestingly, it did not occupy the other
hydrophobic pocket (HPP2) even though the compound 1g was identified as a full antagonist
in cAMP assays. This might be explained by the lack of interaction of compound 1g with
Ser243 in TM5 and other residues in TM6, which are known to be critical in activation of G
protein signaling pathway. Collectively, these results show that, if a ligand is not close
7
enough to TM5 or TM6, it can block G protein signaling of 5ꢁHT R even in the absence of
occupation of the HPP2 region.
To verify an effect of the key residue Ile233 in EL2 on βꢁarrestin signaling pathway, we also
studied docking pose of ASꢁ19 to compare with that of 1g (Figure 5b). ASꢁ19 contains
7
pyrazole group and tetralin moiety, which acts as a partial agonist of 5ꢁHT R to enhance G
protein as well as βꢁarrestin signaling (Supplementary Figure S2, Table S1). In the docking
mode of ASꢁ19, there is no interaction with Ser243 in TM5, but Tꢁshape πꢁπ interaction with
Arg350 direct affects TM6 which is involved in activation of G protein signaling pathway.
The hydrophobic interaction with Ile233 in EL2 is also found in the docking mode, which
makes ASꢁ19 a partial βꢁarrestin agonist. Overall, compound 1g might be able to stimulate
only βꢁarrestin signaling pathway because it makes the hydrophobic interaction with Ile233
in EL2 and no interaction with Ser243 in TM5 and other residues in TM6. Taken together, it
could be proposed that engagement of 1g with Ile233 in 5ꢁHT R lead to activation of βꢁ
7
arrestin signaling pathway.
1
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. Docking pose of 1g (pink) and hydrophobic pockets (HPP1 and HPP2) at ligand
binding site of 5ꢁHT R constructed by homology modeling using the 5ꢁHT R structure
7
1B
(PDB: 4IAR). The two hydrophobic pockets HPP1 and HPP2 are shown in molecular surface
representation, respectively highlighted with red dotted halfꢁcircles. HPP1 consists of
Val163/Thr214/Phe344 and HPP2 includes Leu232/Ile233/Arg367 residues, which are
expressed by blue color sticks. Compound 1g formed ionic interaction with Asp162 that is
shown by the green color stick.
1
8
ACS Paragon Plus Environment

Page 19 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. a) Binding interaction between 1g and several amino acids in binding site of 5ꢁ
HT R and b) superimposed structures of 1g and ASꢁ19 shown in red and yellow, respectively.
7
The important amino acid residues such as Ser243 and Thr244 in TM5, Ile233 in EL2 and
Asp162 in TM3 were expressed by orange, blue, and green color, respectively.
1
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Cytochrome P450 (CYP) Isozyme Activity and Microsomal Stability. To avoid drugꢁdrug
interactions, inhibitory activities of 1g against cytochrome P450 (CYP) isozymes such as
CYP1A2, CYP2C9, CYP2C19, CYP2D6, and CYP3A4 were examined (Table 4). Compound
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1g has little effects on five CYP isozymes (%ꢁremaining activities > 75.7%) implying that 1g
has no high risk of drugꢁdrug interaction. To test metabolic stability of 1g, human liver
microsomal stability assay was carried out. The microsomal stability was shown in %ꢁ
remaining concentration after incubation with human liver microsomes for 30 min. The
stability is a factor that affects pharmacokinetics (PK) parameters such as clearance, a halfꢁ
life and oral bioavailability of the drug, which can be easily tested in in vitro assays before in
vivo animal study. Compound 1g showed 82.4% recovery after incubation with human liver
microsomes. Overall, 1g displayed excellent in vitro pharmacological properties regarding
selectivity over other serotonin receptor subtypes, CYP isozyme activity, and microsomal
stability.
PK parameters for 1g were then evaluated after intravenous and oral administration in male
ICR mice (Table 5). No mortality occurred when 1g was administered orally and
intravenously. After intravenous injection, 1g exhibited drugꢁlike PK parameters. It was
present in the blood at the highest concentration at 0.11 h and the maximum plasma
concentration (Cmax) was 2198.00 ng/ml. In case of clearance, the mean clearance rate (CL)
was 0.96 l/hr/kg and the terminal halfꢁlife was 1.78 h. In comparison, after oral
administration, the oral bioavailability of 1g was excellent (F = 90.79%). Thus, the PK
properties of 1g strongly support its favorable druggable properties.
2
0
ACS Paragon Plus Environment

Page 21 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 4. Results of CYP450 and microsomal stability tests
a
Human
Microsomal
CYP450 screening
Compound
stability
CYP1A2
86.0
CYP2C9
> 100
CYP2C19 CYP2D6 CYP3A4
> 100 75.7 95.4
b
(%)
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1g
82.4
a
b
%ꢁremaining activity.
Human liver microsomal stability (% remaining duing 30 min).
a
Table 5. Mean (± SD ) pharmacokinetic parameters after intravenous (IV, n = 5) and oral (PO,
n = 4) administration (5 mg/kg for IV, 50 mg/kg for PO) of 1g to ICR mice
Plasma
Intravenous
Oral
±
±
±
±
±
±
±
±
±
AUC0ꢁ∞ (hr ng/ml)
AUClast (hr ng/ml)
Terminal halfꢁlife (hr)
5210.48
4897.50
1.78
297.09
391.73
0.35
64319.56
53688.13
2.77
3601.43
4433.79
0.55
±
±
±
C
max (ng/ml)
max (h)
2198.00
0.11
211.12
0.08
11782.50
1.37
1773.48
0.75
±
T
ꢁꢁ
CL (l/hr/kg)
MRT (hr)
0.96
0.06
ꢁꢁ
2.50
0.19
ꢁꢁ
V
dss (l/kg)
2.47
0.52
ꢁꢁ
±
F (%)
90.79
12.69
AUC0ꢁ∞, total area under the plasma concentrationetime curve from time zero to time infinity;
AUC , total area under the plasma concentrationetime curve from time zero to last
last
measured time; Cmax, peak plasma concentration; Tmax, time to reach Cmax; CL, timeꢁ
averaged total body clearance; MRT, mean residence time; V , apparent volume of
dss
distribution at steady state; F, oral bioavailability
a
SD: Standard deviations.
2
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Effect on NREM/REM Sleep. REM sleep pattern could be changed in several psychiatric
5
4,55
disorders such as depression and autism.
amount of REM sleep was reduced in the 5ꢁHT
selective 5ꢁHT R antagonists and selective serotonin reuptake inhibitors (SSRIs) could
increase the latency of REM sleep and decrease the total time of REM sleep with little change
Recently, preclinical data indicated that the total
56,57
7
R knockꢁout models.
In addition,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
31,56,58
in nonꢁrapid eye movement (NREM) sleep pattern.
Thus, we evaluated the effect of 1g
on electroencephalogram (EEG) sleep architecture to investigate a correlation between βꢁ
arrestin biased activity of 1g and NREM/REM sleep pattern in male mice with SB269970 as
a control (Figure 6, Table 6). All of the tested drugs were intraperitoneally (ip) administered
at the beginning of the light phase. After administration of SB269970 at a dose of 30 mg/kg
to EEG mouse models, the duration of NREM sleep was not altered compared with NREM
sleep in the absence of a drug. Compound 1g, however, elevated NREM sleep duration
during all of the time intervals and total NREM time about 21.4%.
Next, the total REM sleep after injection of SB269970 was decreased to 14.4 min, which was
consistent with previous pharmacological data (Figure 6b, Table 6). In total, about 37.1% of
total REM sleep was reduced for tested 8 h compared to 22.9 min of total REM sleep in the
absence of a drug. In sleep EEG experiments of 1g, a more notable decrease in REM sleep
duration was measured during the first 4 h after administration of 1g (30 mg/kg). Only 3.3
min were spent for REM sleep in these time intervals, which means that about 69.4 % of
REM sleep time was decreased compared to the vehicleꢁinjected mice in the same time. This
suppression of REM sleep was prolonged to third time interval (5~6 h) and not observed at
fourth time interval (7~8 h). Total REM sleep time (14.5 min) was also decreased by about
36.7%. Taken together, the βꢁarrestin partial agonist 1g increased NREM sleep duration and
decreased REM sleep duration, while the antagonist SB269970 in both pathways decreased
2
2
ACS Paragon Plus Environment

Page 23 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
REM sleep duration without change of NREM sleep. Though there are more experimental
data needed on sleep patterns of 5ꢁHT
7
R biased ligands, it could be suggested that βꢁarrestin
activity of 5ꢁHT R might modulate NREM sleep patterns.
7
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6. Effect of 1g on a) NREM sleep and b) REM sleep duration. Data bars represent
means ± SEM. REM sleep duration was compiled for the 8 h following the intraperitoneal
**
***
administration of vehicle, SB269970 and 1g: *P < 0.05, P < 0.01, and P < 0.001 versus
vehicle based on twoꢁway ANOVA followed by Bonferroni post hoc test.
Table 6. Duration of REM sleep and total time spent in REM sleep after treatment with
vehicle, SB269970 and 1g (30 mg/kg) for 8 hours measured in healthy subjects (n = 8, 6, and
9, respectively). All data are expressed as mean ± SEM (minutes).
NREM duration
Total NREM
1ꢁ2 h
3ꢁ4 h
5ꢁ6 h
7ꢁ8 h
Vehicle
SB269970
62.7 ± 5.1
65.5 ± 7.0
75.3 ± 6.2
65.0 ± 5.7
69.4 ± 6.4
82.7 ± 4.1
69.5 ± 2.5
61.1 ± 6.6
80.9 ± 5.5
57.0 ± 3.5
55.9 ± 7.4
69.5 ± 4.6
254.2 ± 12.7
252.0 ± 22.7
*
*
*
1g
308.5 ± 14.5
Total REM
22.9 ± 3.9
REM duration
3ꢁ4 h
1ꢁ2 h
5ꢁ6 h
7ꢁ8 h
Vehicle
SB269970
3.9 ± 0.7
6.9 ± 1.3
5.0 ± 1.2
7.5 ± 1.5
3.8 ± 0.9
4.6 ± 1.3
*
*
*
1.4 ± 0.5
0.6 ± 0.3
4.2 ± 0.5
14.4 ± 2.3
14.5 ± 2.7
***
**
*
**
1g
2.7 ± 0.9
4.9 ± 0.8
6.3 ± 1.0
*
**
***
P < 0.05, P < 0.01, and P < 0.001 versus vehicle based on twoꢁway ANOVA followed by
Bonferroni post hoc test.
2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
■ CONCLUSIONS
We designed and synthesized a series of tetrahydroazepines 1 and 2 with arylpyrazolo moiety
or arylisoxazolo moiety. With total 20 synthesized compounds (1 and 2), biological
evaluations such as binding affinity, selectivity profile, and functions in G protein/βꢁarrestin
signaling pathway were carried out. SAR studies of 1 and 2 revealed that compounds 1d, 1g,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
1j, 2d, 2g, and 2j with paraꢁsubstitutent in phenyl group more strongly bind to 5ꢁHT R than
the corresponding ortho or metaꢁsubstituted congeners. In cAMP and Tango assays, we have
discovered that compounds 1d, 1g, 1j, 2d, 2g, and 2j acted as βꢁarrestin biased ligands of 5ꢁ
HT
7
R (EC50 < 188 nM, Emax > 38.23%). Among those 6 compounds, the compound 1g was
ꢁarrestin partial agonist with 5ꢁHT R selectivity over other serotonin
the most effective
β
7
receptor subtypes. Molecular docking studies suggested that 1g could activate only βꢁarrestin
recruitment, where 1g has πꢁalkyl interaction with Ile233, a key residue for βꢁarrestin
activation, but no interaction with residues for G protein activation such as Ser243 in TM5
and other residues in TM6. According to assessment on EEG sleep architecture, the βꢁarrestin
biased ligand 1g increased NREM sleep duration during all of the time intervals after
intraperitoneal administration unlike a balanced antagonist SB269970, while both 1g and
SB269970 decreased REM sleep time. The novel βꢁarrestin biased ligand 1g with good
druggable properties would be a lead compound to develop drugs targeting 5ꢁHT
treatment of sleep disorders.
7
R for
2
4
ACS Paragon Plus Environment

Page 25 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
■ EXPERIMENTAL SECTION
Chemistry. All reactions were carried out under dry nitrogen unless otherwise indicated.
Commercially available reagents were used without further purification. Solvents and gases
were dried according to standard procedures. Organic solvents were evaporated with reduced
pressure using a rotary evaporator. Analytical thin layer chromatography (TLC) was
performed using glass plates preꢁcoated with silica gel (0.25 mm). TLC plates were
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
visualized by exposure to UV light and then were visualized with a KMnO , ninhydrin, and
4
pꢁanisaldehyde staining solution followed by brief heating on a hot plate. Flash column
chromatography was performed using silica gel 60 (230ꢁ400 mesh, Merck) with the indicated
1
13
solvents. H and C NMR spectra were recorded on Bruker 300, or 400 NMR spectrometers.
1
H NMR spectra are represented as follows: chemical shift, multiplicity (s = singlet, brs =
broad singlet, d = doublet, t = triplet, q = quartet, m = multiplet), integration, and coupling
constant (J) in Hertz (Hz). LC/MS and HRMS analyses were performed on Agilent 6410
+
Triple Quad system and Bruker Compact ESI positive mode, respectively. The purities of all
final compounds were ≥95% and checked on a Waters HPLC e2695 instrument equipped
with a UV/Vis 2489 detector and a Capcell Pak® 3 µm C18 MGꢁII (4.6 x 75 mm) column.
Standard conditions were as follows: solvent A (CH CN), solvent B (0.1% AcOH in H O), a
3
2
flow rate of 1 mL/min, and a gradient of solvent A (10–100%) in solvent B over 20 min. The
detector wavelength was set at 254 nm or 280 nm.
3ꢀPhenylꢀ1,4,5,6,7,8ꢀhexahydropyrazolo[3,4ꢀd]azepine (1a). To a solution of compound 8a
(0.27 mmol) in DCM/MeOH (9:1, 5 ml), the excess amount of 1 N HCl in Et O (3 ml) was
2
added. The reaction mixture was stirred at room temperature for 12 h. After reaction
termination, the reaction solution extracted with a minimum volume of water, and basified by
2
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
10 N NaOH. The aqueous layer was extracted with DCM again, and dried over MgSO ,
filtered, and concentrated to obtain desired final compounds 1a (0.042 mmol, 16% yield).
1
HPLC: purity 99 %, t
R
=6.1 min; HNMR (300 MHz, CDCl
3
) δ 7.47ꢁ7.35 (m, 5H), 3.00ꢁ2.93
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
13
(
m, 4H), 2.86ꢁ2.82 (m, 2H), 2.78ꢁ2.75 (m, 2H); CNMR (75 MHz, CDCl
3
) δ 149.75, 144.84,
+
1
31.85, 128.64, 128.12, 127.90, 116.79, 50.51, 48.69, 32.13, 28.12; HRMS (ESI ) calcd for
+
+
C
13
16 3
H N [M+H] 214.1344, found 214.1349.
3ꢀ(2ꢀFluorophenyl)ꢀ1,4,5,6,7,8ꢀhexahydropyrazolo[3,4ꢀd]azepine (1b). Compound 1b was
synthesized according to the procedure of 1a using a solution of 8b (0.093 mmol) and 1N
HCl in Et
2
O (2 ml) in DCM/MeOH (9:1, 2 ml). Yield: 28% (0.026 mmol); HPLC: purity
1
9
2
8 %, t
R
=6.1 min; HNMR (300 MHz, CDCl
3
) δ 7.43ꢁ7.31 (m, 2H), 7.21ꢁ7.11 (m, 2H), 3.02ꢁ
13
3
.93 (m, 4H), 2.87ꢁ2.84 (m, 2H), 2.67ꢁ2.63 (m, 2H); CNMR (75 MHz, CDCl ) δ 159.77 (d,
J = 246 Hz), 149.09, 139.44, 131.02 (d, J = 3.7 Hz), 129.92 (d, J = 8.2 Hz), 124.29 (d, J = 3.0
Hz), 119.65 (d, J = 15 Hz), 118.50, 116.05 (d, J = 22.5 Hz), 50.32, 48.67, 32.12, 28.37;
+
+
+
HRMS (ESI ) calcd for C13
H
15FN
3
[M+H] 232.1250, found 232.1252.
3ꢀ(3ꢀFluorophenyl)ꢀ1,4,5,6,7,8ꢀhexahydropyrazolo[3,4ꢀd]azepine (1c). Compound 1c was
synthesized according to the procedure of 1a using a solution of 8c (0.43 mmol) and 1N HCl
in Et O (10 ml) in DCM/MeOH (9:1, 10 ml). Yield: 72% (0.31 mmol); HPLC: purity 99 %,
2
1
t
R
=6.3 min; HNMR (300 MHz, CDCl
3
) δ 7.42ꢁ7.15 (m, 3H), 7.05ꢁ6.99 (m, 1H), 2.98ꢁ2.92
) δ 162.75 (d, J = 224.5 Hz), 147.36,
13
(
m, 4H), 2.79ꢁ2.70 (m, 4H); CNMR (75 MHz, CDCl
3
145.70, 134.83 (d, J = 7.5 Hz), 130.02 (d, J = 8.25 Hz), 124.00 (d, J = 2.25 Hz), 116.77,
+
1
15.14 (d, J = 22.5 Hz), 114.49 (d, J = 21 Hz), 50.33, 48.41, 31.19, 28.02; HRMS (ESI )
+
+
calcd for C H FN [M+H] 232.1250, found 232.1247.
13
15
3
3ꢀ(4ꢀFluorophenyl)ꢀ1,4,5,6,7,8ꢀhexahydropyrazolo[3,4ꢀd]azepine (1d). Compound 1d was
synthesized according to the procedure of 1a using a solution of 8d (0.23 mmol) and 1N HCl
2
6
ACS Paragon Plus Environment

Page 27 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
in Et
2
O (10 ml) in DCM/MeOH (9:1, 5 ml). Yield: 75% (0.17 mmol); HPLC: purity 97 %,
1
t =6.3 min; HNMR (300 MHz, CDCl ) δ 7.41ꢁ7.36 (m, 2H), 7.07ꢁ7.01 (m, 2H), 2.95ꢁ2.90
R
3
13
(m, 4H), 2.72ꢁ2.69 (m, 4H); CNMR (75 MHz, CDCl
3
) δ 162.45 (d, J = 245.25 Hz), 147.91,
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
45.50, 129.99 (d, J = 8.25 Hz), 128.55, 116.60, 115.43 (d, J = 21 Hz), 50.46, 48.58, 31.54,
+
+
+
8.14; HRMS (ESI ) calcd for C13
3
H15FN [M+H] 232.1250, found 232.1247.
ꢀ(2ꢀChlorophenyl)ꢀ1,4,5,6,7,8ꢀhexahydropyrazolo[3,4ꢀd]azepine (1e). Compound 1e was
synthesized according to the procedure of 1a using a solution of 8e (0.23 mmol) and 1N HCl
in Et
2
O (10 ml) in DCM/MeOH (9:1, 5 ml). Yield: 81% (0.19 mmol); HPLC: purity 97 %,
1
t
R
=6.4 min; HNMR (300 MHz, CDCl
3
) δ 7.45ꢁ7.42 (m, 1H), 7.36ꢁ7.24 (m, 3H), 2.94ꢁ2.90
13
(m, 4H), 2.69 (t, J = 4.9 Hz, 2H), 2.51 (t, J = 4.9 Hz, 2H); CNMR (75 MHz, CDCl
3
) δ
1
2
47.62, 143.42, 133.92, 132.12, 131.45, 129.76, 129.57, 126.61, 118.38, 50.13, 48.64, 31.57,
+
+
+
8.42; HRMS (ESI ) calcd for C H ClN [M+H] 248.0955, found 248.0958.
13
15
3
3ꢀ(3ꢀChlorophenyl)ꢀ1,4,5,6,7,8ꢀhexahydropyrazolo[3,4ꢀd]azepine (1f). Compound 1f was
synthesized according to the procedure of 1a using a solution of 8f (0.56 mmol) and 1N HCl
in Et
2
O (15 ml) in DCM/MeOH (9:1, 5 ml). Yield: 49% (0.27 mmol); HPLC: purity 99 %,
1
t
R
=6.8 min; HNMR (300 MHz, CDCl
3
) δ 7.43 (s, 1H), 7.32ꢁ7.27 (m, 3H), 2.95ꢁ2.90 (m, 4H),
13
2.74ꢁ2.68 (m, 4H); CNMR (75 MHz, CDCl ) δ 147.35, 145.62, 134.53, 134.37, 129.72,
3
+
128.26, 127.68, 126.45, 116.87, 50.39, 48.50, 31.32, 28.14; HRMS (ESI ) calcd for
+
+
C
13 3
H15ClN [M+H] 248.0955, found 248.0959.
3ꢀ(4ꢀChlorophenyl)ꢀ1,4,5,6,7,8ꢀhexahydropyrazolo[3,4ꢀd]azepine (1g). Compound 1g was
synthesized according to the procedure of 1a using a solution of 8g (0.78 mmol) and 1N HCl
in Et O (15 ml) in DCM/MeOH (9:1, 5 ml). Yield: 73% (0.57 mmol); HPLC: purity 99 %,
2
1
t
R
= 6.5 min; HNMR (300 MHz, CDCl
3
) δ 7.41ꢁ7.30 (m, 4H), 2.99ꢁ2.94 (m, 4H), 2.80ꢁ2.71
13
(m, 4H); CNMR (75 MHz, CDCl
3
) δ 147.88, 145.42, 133.77, 130.88, 129.47, 128.72,
2
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 58
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
+
+
+
116.81, 50.45, 48.56, 31.55, 28.14; HRMS (ESI ) calcd for C13
H
15ClN
3
[M+H] 248.0955,
found 248.0962.
3
ꢀ(oꢀTolyl)ꢀ1,4,5,6,7,8ꢀhexahydropyrazolo[3,4ꢀd]azepine (1h). Compound 1h was
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
synthesized according to the procedure of 1a using a solution of 8h (0.50 mmol) and 1N HCl
in Et
2
O (15 ml) in DCM/MeOH (9:1, 5 ml). Yield: 55% (0.27 mmol); HPLC: purity 96 %,
1
t
R
= 6.4 min; HNMR (300 MHz, CDCl
3
) δ 7.26ꢁ7.17 (m, 4H), 2.83 (q, J = 4.9 Hz, 4H), 2.53
1
3
(t, J = 4.7 Hz, 2H), 2.41 (t, J = 4.6 Hz, 2H), 2.17 (s, 3H); CNMR (75 MHz, CDCl ) δ
3
149.32, 143.75, 137.51, 131.35, 130.58, 130.12, 128.41, 125.50, 117.53, 50.40, 48.86, 31.83,
+
+
+
2
8.10, 19.87; HRMS (ESI ) calcd for C14
18 3
H N [M+H] 228.1501, found 228.1504.
3
ꢀ(mꢀTolyl)ꢀ1,4,5,6,7,8ꢀhexahydropyrazolo[3,4ꢀd]azepine (1i). Compound 1i was
synthesized according to the procedure of 1a using a solution of 8i (0.28 mmol) and 1N HCl
in Et O (10 ml) in DCM/MeOH (9:1, 5 ml). Yield: 33% (0.092 mmol); HPLC: purity 98 %,
2
1
t
R
= 6.7 min; HNMR (300 MHz, CDCl
3
) δ 7.35ꢁ7.30 (m, 1H), 7.26ꢁ7.17 (m, 3H), 3.05ꢁ3.02
1
3
(m, 2H), 2.98ꢁ2.90 (m, 4H), 2.79ꢁ2.76 (m, 2H), 2.40 (s, 3H); CNMR (75 MHz, CDCl
3
) δ
1
2
3
49.30, 144.43, 138.46, 131.29, 128.87, 128.69, 128.63, 125.18, 116.09, 49.81, 48.00, 30.61,
+
+
+
6.60, 21.45; HRMS (ESI ) calcd for C14
18 3
H N [M+H] 228.1501, found 228.1501.
ꢀ(pꢀTolyl)ꢀ1,4,5,6,7,8ꢀhexahydropyrazolo[3,4ꢀd]azepine (1j). Compound 1j was
synthesized according to the procedure of 1a using a solution of 8j (0.95 mmol) and 1N HCl
in Et
2
O (15 ml) in DCM/MeOH (9:1, 5 ml). Yield: 58% (0.55 mmol); HPLC: purity 100 %,
1
t
R
= 6.6 min; HNMR (300 MHz, CDCl
3
) δ 7.51 (brs, 1H), 7.29 (d, J = 7.8 Hz, 2H), 7.12 (d, J
1
3
=
7.5 Hz, 2H), 2.87 (brs, 4H), 2.71ꢁ2.69 (m, 4H), 2.32 (s, 3H); CNMR (75 MHz, CDCl
3
) δ
1
48.81, 145.19, 137.36, 129.37, 129.16, 128.22, 116.30, 50.44, 48.58, 31.77, 28.16, 21.26;
+
+
+
HRMS (ESI ) calcd for C14
18 3
H N [M+H] 228.1501, found 228.1505.
3
ꢀPhenylꢀ5,6,7,8ꢀtetrahydroꢀ4Hꢀisoxazolo[4,5ꢀd]azepine (2a). To a solution of HBr in
2
8
ACS Paragon Plus Environment

Page 29 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
AcOH (33%, 0.15 ml), compound 15a (0.15 mmol) was added. The reaction mixture was
stirred at room temperature for 2 h. During the reaction time, HBr in AcOH (33%, 0.15 x 2
ml) was additionally added twice. After reaction termination, the reaction solution extracted
with a minimum volume of water, and basified by 10 N NaOH. The aqueous layer was
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
extracted with DCM again, and dried over MgSO
4
, filtered, and concentrated to obtain
1
desired final compounds 2a (0.042 mmol, 28%). HPLC: purity 96 %, t
R
= 6.5 min; HNMR
(300 MHz, CDCl ) δ 7.62ꢁ7.35 (m, 5H), 3.20ꢁ2.95 (m, 4H), 2.94ꢁ2.80 (m, 2H), 2.79ꢁ2.71 (m,
3
13
3
2H); CNMR (75 MHz, CDCl ) δ 161.89, 141.20, 128.95, 128.55, 127.12, 120.46, 112.11,
+
+
+
4
2
3
7.05, 46.58, 40.50, 36.0; HRMS (ESI ) calcd for C13
15.1181.
ꢀ(2ꢀFluorophenyl)ꢀ5,6,7,8ꢀtetrahydroꢀ4Hꢀisoxazolo[4,5ꢀd]azepine (2b). Compound 2b
15 2
H N O [M+H] 215.1184, found
was synthesized according to the procedure of 2a using a solution of 15b (0.31 mmol) and
R
HBr in AcOH (33%, 0.15 x 3 ml). Yield: 79% (0.10 mmol); HPLC: purity 98 %, t = 6.4 min;
1
HNMR (300 MHz, CDCl
3
) δ 7.40ꢁ7.26 (m, 2H), 7.25ꢁ7.07 (m. 2H), 3.15ꢁ2.92 (m, 4H), 2.90ꢁ
.62 (m 4H); ; CNMR (75 MHz, CDCl ) δ 159.00 (d, J = 236.25 Hz), 130.72 (d, J = 8.25
Hz), 129.57, 129.53, 129.26 (d, J = 14.25 Hz), 124.29 (d, J = 3 Hz), 119.61, 116.05 (d, J = 21
13
2
3
+
+
+
Hz), 115.25, 47.04, 46.38, 40.40, 36.18; HRMS (ESI ) calcd for C H FN O [M+H]
13
14
2
233.1090, found 233.1084.
3ꢀ(3ꢀFluorophenyl)ꢀ5,6,7,8ꢀtetrahydroꢀ4Hꢀisoxazolo[4,5ꢀd]azepine (2c). Compound 2c
was synthesized according to the procedure of 2a using a solution of 15c (0.20 mmol) and
R
HBr in AcOH (33%, 0.15 x 3 ml). Yield: 95% (0.19 mmol); HPLC: purity 98 %, t = 6.6 min;
1
HNMR (400 MHz, CDCl ) δ 7.38ꢁ7.06 (m, 4H), 2.99 (brs, 4H), 2.83 (brs, 2H), 2.72 (brs,
3
13
2
H); CNMR (100 MHz, CDCl
3
) δ 162.51 (d, J = 245 Hz), 160.48, 143.30 (d, J = 7 Hz),
130.22 (d, J = 8 Hz), 122.97 (d, J = 3 Hz), 120.01, 115.78 (d, J = 21 Hz), 114.19 (d, J = 22
2
9
ACS Paragon Plus Environment