Design and Discovery of Functionally Selective Serotonin 2C (5-HT2C) Receptor Agonists

Article abstract of DOI:10.1021/acs.jmedchem.6b01194

On the basis of the structural similarity of our previous 5-HT_2C agonists with the melatonin receptor agonist tasimelteon and the putative biological cross-talk between serotonergic and melatonergic systems, a series of new (2,3-dihydro)benzofu

Full text of DOI:10.1021/acs.jmedchem.6b01194

Subscriber access provided by Northern Illinois University
Article
Design and Discovery of Functionally Selective
2C
Serotonin 2C (5-HT) Receptor Agonists
Jianjun Cheng, John D. McCorvy, Patrick M. Giguère, Hu
Zhu, Terry Kenakin, Bryan L Roth, and Alan P. Kozikowski
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b01194 • Publication Date (Web): 11 Oct 2016
Downloaded from http://pubs.acs.org on October 13, 2016
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 free 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 accessible to all
readers and 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.
Journal of Medicinal Chemistry 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 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Design and Discovery of Functionally
Selective Serotonin 2C (5ꢀHT ) Receptor
2
C
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Agonists
§
,†,˄
‡,†
‡,†,˅
‡,#
Jianjun Cheng,
John D. McCorvy, Patrick M. Giguere,
Hu Zhu, Terry
¶
,‡
,§
Kenakin, Bryan L. Roth,* and Alan P. Kozikowski*
§
Drug Discovery Program, Department of Medicinal Chemistry and Pharmacognosy,
College of Pharmacy, University of Illinois at Chicago, Chicago, IL 60612, United
States
‡
National Institute of Mental Health Psychoactive Drug Screening Program,
¶
Department of Pharmacology and Division of Chemical Biology and Medicinal
Chemistry, University of North Carolina Chapel Hill Medical School, Chapel Hill,
NC 27599, United States
†
These authors contributed equally.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 57
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
Based on the structural similarity of our previous 5ꢀHT2C agonists with the melatonin
receptor agonist tasimelteon, and the putative biological crossꢀtalk between
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
serotonergic
2,3ꢀdihydro)benzofuranꢀbased compounds were designed and synthesized. The
compounds were evaluated for their selectivity toward 5ꢀHT , 5ꢀHT , and 5ꢀHT
2C
and
melatonergic
systems,
a
series
of
new
(
2
A
2B
receptors in the calcium flux assay with the ultimate goal to generate selective 5ꢀHT2C
agonists. Selected compounds were studied for their functional selectivity by
comparing their transduction efficiency at the G protein signaling pathway versus
βꢀarrestin recruitment. The most functionally selective compound (+)–7e produced
weak βꢀarrestin recruitment and also demonstrated less receptor desensitization
compared to serotonin in both calcium flux and phosphoinositide (PI) hydrolysis
assays. We report for the first time that selective 5ꢀHT2C agonists possessing weak
βꢀarrestin recruitment can produce distinct receptor desensitization properties.
ACS Paragon Plus Environment

Page 3 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Introduction
The serotonin 2C (5ꢀHT2C) receptor has been found to be an invaluable drug target for
a variety of central nervous system (CNS) disorders, such as obesity, schizophrenia,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
ꢀ3
and drug addiction. The 5ꢀHT2C receptor is activated endogenously by serotonin
(
5ꢀHT, 1, Figure 1), which is a major neurotransmitter widely found in both the
4
periphery and the CNS. The 5ꢀHT receptor belongs to the family of serotonin
2C
receptors comprised of 14 subtypes (5ꢀHT1ꢀ7, some of these with further
5
subclassifications). All serotonin receptors belong to the G proteinꢀcoupled receptor
(
GPCR) superfamily except 5ꢀHT , which is a ligandꢀgated ion channel. The 5ꢀHT2C
3
receptor shares high sequence similarity with the other 5ꢀHT
2
ꢀfamily receptors:
5
ꢀHT2A and 5ꢀHT2B receptors (55% and 52% homology, respectively).
In the mammalian pineal gland, serotonin can be converted to melatonin (2), which is
another neurotransmitter that is involved in the control of circadian rhythms linked to
certain physiological functions including the timing of sleep, blood pressure
6
regulation, seasonal reproduction, immune function, etc. A close interrelationship
between the melatonergic and serotonergic systems has long been suspected, and
evidence for their crossꢀtalk has been recently reported. For example, melatonin
7
inhibits the ability of serotonin to phase shift the suprachiasmatic circadian clock. In
fact, agomelatine, which is on the market for major depressive disorder, exhibits
8
melatonin receptor agonist activity as well as 5ꢀHT2C antagonism.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 57
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 1. Bioꢀinspired design of 2,3ꢀdihydrobenzofuran (6) and benzofuran (7 and 8)
compounds as 5ꢀHT agonists.
2
C
In our previous work, we identified compounds 3 and 4 as highly selective 5ꢀHT2C
9, 10
agonists (Figure 1).
Both compounds display excellent selectivity against the
5
ꢀHT2A and 5ꢀHT2B receptors, which activation are associated with hallucinogenic
1
1ꢀ13
effects and cardiac valvulopathy, respectively.
Interestingly, although compounds
3
and 4 evolved from a hit compound which was identified through a high throughput
14
screening (HTS) campaign, they share the same 2ꢀarylcyclopropylmethylamine
backbone with tasimelteon (5), a melatonin receptor agonist that was approved
1
5
recently for the treatment of nonꢀ24ꢀhour sleepꢀwake disorder. The crossꢀtalk
between melatonergic and serotonergic signaling pathways, and the structural
similarity between compounds 3, 4, and tasimelteon, thus inspired us to design
compounds possessing the general structure 6 (Figure 1). As a matter of fact, both
benzofuran and 2,3ꢀdihydrobenzofuran have been reported as substructures of
ACS Paragon Plus Environment

Page 5 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
6
5
ꢀHT2A/2C receptor ligands previously. We anticipated that compound 6 and its
analogs 7 and 8 would function as 5ꢀHT2C agonists.
Furthermore, functional selectivity, also known as ligand bias, is a phenomenon
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
whereby a ligand can possess multiple receptor signaling events (i.e G –linked
q
calcium flux, βꢀarrestin recruitment) and is thought to occur through a ligand’s ability
17
to stabilize certain receptor conformations. Biased ligands and functional selectivity
have emerged as improved therapeutics for several GPCRs including the dopamine D
2
18
19
20
receptor, the kappa opioid receptor, and the mu opioid receptor, and it is likely
that the interest in the discovery of novel GPCR ligands targeting specific
downstream pathways will continue to grow with time. Although functional
selectivity has been actively investigated for several GPCRs, especially for 5ꢀHT
21
receptors, there has been little characterization of biased agonism at the 5ꢀHT2C
receptor, especially examining βꢀarrestin recruitment. The recruitment of βꢀarrestin to
the GPCRs leads to G proteinꢀindependent signaling, desensitization and
sequestration of the receptor and eventually to GPCR internalization, and it is
considered to be an opposing and complementary signaling pathway to G protein
22
signaling. In this article, we report our recent work in synthesizing compounds
bearing the general structures 6–8, their pharmacological profiling at 5ꢀHT2C, 5ꢀHT2A
,
and 5ꢀHT2B receptors, and explore for the first time their G protein versus βꢀarrestin
signaling profiles.
Results and Discussion
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 57
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
Chemistry
We have previously developed a general synthetic approach to various
9, 23ꢀ25
2
ꢀphenylcyclopropylmethylamines,
which employs appropriate benzaldehydes
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
as the starting materials. Hence, the corresponding aldehydes 15a–c for compounds 6,
and 8 were prepared and the synthetic routes are depicted in Scheme 1. Methyl
2,5ꢀdihydroxybenzoate 9 was used as the starting material, the oxidation of which
with Ag O provided quinone 10. The benzofuran intermediate 11 was prepared via a
7
2
[3 + 2] cycloaddition of quinone 10 with butyl vinyl ether followed by an
26
aromatization reaction under acidic conditions. The phenol 11 was methylated to
afford compound 12, which was then reduced by hydrogenation to provide the
2
,3ꢀdihydrobenzofuran intermediate 13. This carboxylic ester was converted to
aldehyde 15a via a LiAlH reduction followed by a Swern oxidation. By omitting the
hydrogenation step, direct LiAlH reduction and Swern oxidation of 12 provided 15b
4
4
in good yields. To incorporate the ethyl substitution at position 2 of the benzofuran
scaffold, the benzyl alcohol 16 was protected as the TBS ether 17, which was treated
27
in turn with nꢀBuLi and EtI to provide intermediate 18. Desilylation and Swern
oxidation of 18 produced aldehyde 15c.
ACS Paragon Plus Environment

Page 7 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
Scheme 1. Synthesis of Aldehydes 15aꢀc.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Reagents and conditions: (a) Ag O, MgSO , Et O, rt, 3 h; (b) butyl vinyl ether,
2
4
2
toluene, 50 ºC, overnight; then TFA, 50 ºC, 3 h, 52% for 2 steps; (c) MeI, K
DMF, rt, 3 h, 96%; (d) H (50 psi), 10% Pd/C, MeOH, overnight, 79%; (e) LiAlH
THF, 0 ºC to rt, 1 h, 93–95%; (f) DMSO, (COCl) , DCM, –78 ºC, 45 min; then Et N,
5–89%; (g) DMF, TBSCl, imidazole, rt, 2 h; (h) nꢀBuLi, EtI, THF, –78 ºC to rt, 24 h,
4% for 2 steps; (i) TBAF, THF, rt, 1 h, 90%.
2 3
CO ,
2
4
,
2
3
7
9
With aldehydes 15a–c in hand, the syntheses of the desired target compounds 6a–8a
25
were accomplished following the same approach as we reported previously, as
depicted in Scheme 2. Wittig reaction of aldehydes 15a–c with the commercial
reagent
NꢀmethoxyꢀNꢀmethyl(triphenylphosphoranylidene)acetamide
provided
acrylamides 20a–c as their E isomers with complete selectivity. CoreyꢀChaykovski
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 57
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
cyclopropanation of 20a–c generated the cyclopropanes 21a–c in their trans
conformations, using the sulfur ylide generated from trimethylsulfoxonium iodide
upon treatment with sodium hydride. Next, sequential reduction with DIBALꢀH and
sodium borohydride provided alcohols 22a–c in good yields, which were then
converted by Mitsunobu reaction with phthalimide to afford the Gabriel imides 23a–c
in excellent yields. Deꢀprotection of the imides with hydrazine hydrate afforded the
primary amines, which were then protected as the Boc intermediates 24a–c.
Separation of 24a–c using chiral preparativeꢀHPLC followed by the removal of the
Boc group provided compounds (–)ꢀ6a, (+)ꢀ6a, (–)ꢀ7a, (+)ꢀ7a, (–)ꢀ8a, and (+)ꢀ8a as
the optically pure isomers.
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
Scheme 2. Synthesis of Compounds 6a–8a.
a
3 2 2
Reagents and conditions: (a) Ph P=CHC(O)N(OMe)Me, CH Cl , rt, overnight, 79–
96%; (b) trimethylsulfoxonium iodide, NaH, DMSO, rt, overnight, 83–91%; (c)
ACS Paragon Plus Environment

Page 9 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
DIBALꢀH, THF, –78 ºC, 3 h; then MeOH, NaBH
4
, rt, 0.5 h, 76–90%; (d) phthalimide,
Ph
3
2 4 2
P, diethyl azodicarboxylate, rt, overnight, 92–97%; (e) N H ꢀH O, EtOH, reflux, 2
h; then DCM, Et
3 2
N, Boc O, rt, 30 min, 70–88%; (f) chiral preparative HPLC; then
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
M HCl in Et O, rt, 24–48h, 67–75% (compounds (–)ꢀ6a and (+)ꢀ6a were further
2
purified using preparative HPLC and obtained as TFA salts).
Preparation of the compounds bearing a 5ꢀalkoxy group other than methoxy was
accomplished by deꢀmethylation of intermediates 24a–c followed by alkylation with
the appropriate alkyl halides. The syntheses of these compounds are outlined in
Schemes 3–5 respectively. The demethylation and alkylation conditions are similar to
25
those we had previously reported for the 2ꢀphenylcyclopropylmethylamines.
Notably, chiral separation of the 2,3ꢀdihydrobenzofurans was accomplished using the
phenol intermediate 25, while for the benzofuran and 2ꢀethylbenzofuran compounds
the chiral separation was carried out with the ethers 28a–h and 30a–e. These
protocols were used due to the fact that chiral separation of intermediates 26a–26e
proved very difficult using the chiral columns we had available. Therefore the
resolution was performed one step earlier and the phenol intermediate 25 was
separated efficiently with a RegisPack column. Similarly, the final compounds were
obtained by removal of the Boc protecting group using HCl in diethyl ether.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 57
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 3. Synthesis of compounds 6b–6f.
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) CH
2
Cl
2
, BBr
3
, –78 ºC to rt, 3 h; then Et
3 2
N, Boc O, rt,
1
0
2
.5 h, 90%; (b) R X (X = Br or I), Cs
2
CO , DMF for 26a–26d, 59–94%;
3
ꢀfluoroethanol, Ph P, diethyl azodicarboxylate, THF for 26e, 19%; (c) 2M HCl in
3
2
Et O, rt, 24–48h, 62–84%.
ACS Paragon Plus Environment

Page 11 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
Scheme 4. Synthesis of compounds 7b–7i.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Reagents and conditions: (a) CH
2 2 3 3 2
Cl , BBr , –78 ºC to rt, 3 h; then Et N, Boc O, rt,
0
.5 h, 77%; (b) 2ꢀfluoroethanol, Ph P, diethyl azodicarboxylate, THF for 28d, 99%;
3
2
R X (X = Cl, Br, or I), Cs
2 3 2
CO , DMF for others, 74–95%; (c) 2M HCl in Et O, rt, 24–
4
8h, 57–70%.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 57
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 5. Synthesis of compounds 8b–8f.
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) CH
2
Cl
P, diethyl azodicarboxylate, THF for 30c, 91%;
R X (X = Br or I), Cs CO , DMF for others, 94–100%; (c) 2M HCl in Et O, rt, 24–
2 3 3 2
, BBr , –78 ºC to rt, 3 h; then Et N, Boc O, rt,
0
.5 h, 86%; (b) 2ꢀfluoroethanol, Ph
3
3
2
3
2
4
8h, 60–85%.
In our previous work, the absolute configurations of both enantiomers of
2ꢀphenylcyclopropylmethylamine were assigned by comparison of the optical
14
rotations of their synthetic intermediates to those of known compounds. Based on
that, we subsequently assigned the absolute configurations of other analogs (which
possess an extra 2ꢀalkoxy substitution on the benzene ring) based on the very good
correlation of compound potency with the direction of their optical rotations (namely,
9, 14, 25
the (+)ꢀenantiomer is always more potent than the (–)ꢀenantiomer).
Therefore,
we assume that an enantiomer with a negative optical rotation possesses the
1
R,2Rꢀconfiguration, while a compound showing a positive optical rotation has the
ACS Paragon Plus Environment

Page 13 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
S,2Sꢀconfiguration. An Xꢀray crystal structure would be an ideal way to test whether
this is correct. However, for the new compounds described in this work, we have
failed to obtain single crystals for Xꢀray diffraction. An alternative way to address this
problem is to use a wellꢀproven asymmetric synthesis method to prepare a
representative compound in the series, and then to compare its optical rotation with
that of the equivalent compound obtained from chiral HPLC separation. Since chiral
separation in the 2,3ꢀdihydrobenzofuran series was found to be relatively difficult, we
chose compound (+)ꢀ6d as a representative compound to test this approach.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
28ꢀ30
A camphorsultamꢀdirected asymmetric cyclopropanation reaction,
for which the
absolute configuration of the dominant product could be readily predicted, was used
as the key step for the asymmetric synthesis of compound (+)ꢀ6d. As shown in
Scheme 6 (see Supporting Information for experimental details), aldehyde 15a
obtained as described in Scheme 1 was converted to the cinnamic acid 31 through
Knoevenagel condensation with malonic acid, followed by the coupling with
(1R)ꢀ(+)ꢀ2,10ꢀcamphorsultam to produce cinnamamide 32. The cyclopropanation of
substrate 32 would favor a desired inward attack on the double bond. Twenty
equivalents of diazomethane was used in this reaction to achieve a greater than 95%
conversion of the cinnamamide, and the reaction proceeded in a 96.5/3.5
diastereoselectivity based on HPLC analysis of the crude product. Recrystallization of
the crude product from ethanol afforded compound 33 with greater than 99% de
purity and 58% yield. The higher diastereoselectivity of this reaction compared to
2
8, 29
other reported examples (which are mostly below 10 : 1),
is possibly due to the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 57
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
diꢀortho substitution pattern of substrate 32, which would force a higher πꢀface
selectivity. Subsequent hydrolysis of 33 with a solution of lithium hydroxide provided
acid 34 in quantitative yield, which was then converted to the corresponding amide 35
in excellent yield. Reduction of 35 was brought about using borane, which was
followed by Boc protection of the free amine to provide intermediate 36.
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
3
Deꢀmethylation of 36 using BBr afforded the phenol 37, which was subjected to the
same allylation and deꢀprotection procedures as described in Scheme 2 to provide
compound 39. The spectral data and optical rotation value of 39 were in agreement
with those of (+)ꢀ6d which was obtained from chiral HPLC separation. This result
would thus confirm that the absolute configuration of (+)ꢀenantiomers is (S, S) and
that of (–)ꢀenantiomers is (R, R). Based on this result, we thus assume that the
absolute configurations of compounds (–)ꢀ6a–6f, (–)ꢀ7a–7i, and (–)ꢀ8a–8f are (R,R)
while those of compounds (+)ꢀ6a–6f, (+)ꢀ7a–7i, and (+)ꢀ8a–8f are (S,S).
ACS Paragon Plus Environment

Page 15 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
Scheme 6. CamphorsultamꢀDirected Asymmetric Synthesis of Compound (+)ꢀ6d.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Reagents and conditions: (a) malonic acid, pyrrolidine, pyridine, reflux, 2 h, 94%; (b)
, rt, 2 h; then (1R)ꢀ(+)ꢀ2,10ꢀcamphorsultam, NaH, THF, 0 °C to rt, overnight,
0%; (c) diazomethane, Pd(OAc) , CH Cl /Et O, 0 °C to rt, overnight, 58%; (d)
LiOH, H O/THF, 50 °C, 3 h, 100%; (e) SOCl , toluene, 80 °C, 2 h; then NH OH
28%), 1,4ꢀdioxane, 0 °C to rt, 0.5 h, 98%; (f) BH ꢀTHF (1.0 M in THF), THF, reflux,
SOCl
2
9
2
2
2
2
2
2
4
(
3
4
h; then Boc O, Et N, CH Cl , rt, 0.5 h, 74%; (g) BBr , CH Cl , –78 °C to rt, 3 h;
2
3
2
2
3
2
2
then Boc
2
O, Et
3
N, rt, 0.5 h, 95%; (h) allyl bromide, Cs
2 3
CO , DMF, microwave, 80
°
C, 0.5 h, 83%; (i) 2M HCl in Et O, rt, 48 h, 74%.
2
Pharmacology
5
2
ꢀHT Receptor Screening. All new compounds were screened employing
recombinant, stably expressed human 5ꢀHT2A, 5ꢀHT2B and 5ꢀHT2C receptors in the
HEKꢀ293 cell line, using a fluorescence imaging plate reader (FLIPR) assay as we
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 57
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
9
, 25
described previously.
Estimates of Emax, EC50, and log(Emax/EC50), which is a
relative activity calculation to account for partial agonist differences, are found in
Table 1. Serotonin was used as a positive control and its Emax values were normalized
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
to 100% for all receptors. The FDA approved 5ꢀHT2C agonist lorcaserin (EC = 2.64
50
nM, Emax = 100%, log(Emax/EC50) = 8.58) was used as a reference compound for
comparison purposes.
ACS Paragon Plus Environment

Page 17 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
Table 1. Pharmacological Profiling of Compounds (–)ꢀ/(+)ꢀ6a–6f, (–)ꢀ/(+)ꢀ7a–7i, and (–)ꢀ/(+)ꢀ8a–8f at 5ꢀHT
2
Receptors in Calcium Flux Assay.
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
a
h5ꢀHT2C
EC50
h5ꢀHT2B
EC50
(nM)
h5ꢀHT2A
STRUCTURE
ID
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
Log(Emax
Log(Emax
EC50
(nM)
Log(Emax
/ EC50)
pEC50
E
max
pEC50
E
max
pEC50
E
max
(nM)
/ EC50
)
/ EC50
)
ꢀ
ꢀ
serotonin
9.78 ± 0.02
0.17
2.64
19
100 ± 0.5
100 ± 0.7
102 ± 1.1
101 ± 0.8
105 ± 0.6
106 ± 0.7
94 ± 0.7
104 ± 0.6
90 ± 0.8
101 ± 0.7
65 ± 1.0
95 ± 0.7
79±3.7
9.77
8.84 ± 0.03 1.46 100 ± 1.1
8.84
8.63 ± 0.02 2.35 100 ± 0.7
8.63
lorcaserin 8.58 ± 0.01
–)ꢀ6a
(+)ꢀ6a
8.58
7.73
8.05
7.41
8.59
6.99
7.72
6.98
7.96
6.36
7.56
6.42
6.36 ± 0.03
6.56 ± 0.02
7.12 ± 0.01
6.11 ± 0.03
7.23 ± 0.02
NA
433
279
77
80 ± 1.6
45 ± 0.6
70 ± 0.5
30 ± 0.8
58 ± 0.6
NA
6.27
6.21
6.96
5.58
7.03
NA
6.61 ± 0.01
6.53 ± 0.01
6.93 ± 0.02
6.01 ± 0.02
6.95 ± 0.02
NA
248
295
117
976
113
NA
749
NA
618
NA
754
1960
68 ± 0.5
66 ± 0.7
90 ± 0.9
35 ± 0.6
81 ± 1.0
NA
6.44
6.35
6.89
5.55
6.86
NA
(
7.71 ± 0.02
8.05 ± 0.02
7.39 ± 0.01
8.57 ± 0.01
7.02 ± 0.01
7.69 ± 0.01
7.03 ± 0.01
7.96 ± 0.01
6.55 ± 0.02
7.58 ± 0.01
6.52±0.04
8.9
41
(
–)ꢀ6b
+)ꢀ6b
781
54
(
2.7
96
(
–)ꢀ6c
NA
258
NA
258
NA
269
NA
O
NH2
O
(+)ꢀ6c
(–)ꢀ6d
(+)ꢀ6d
20
6.59 ± 0.03
NA
31 ± 0.5
NA
6.08
NA
6.13 ± 0.01
NA
31 ± 0.3
NA
5.62
NA
94
11
6.59 ± 0.03
NA
35 ± 0.7
NA
6.13
NA
6.20 ± 0.06
NA
49 ± 0.7
NA
5.90
NA
(
–)ꢀ6e
+)ꢀ6e
–)ꢀ6f
283
26
(
6.57 ± 0.05
NA
33 ± 1.0
NA
6.09
NA
6.12 ± 0.04
5.71±0.08
17 ± 0.5
55±4.3
5.35
5.45
(
302
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 57
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
(
+)ꢀ6f
7.13±0.03
8.17 ± 0.02
8.79 ± 0.03
7.59 ± 0.02
8.59 ± 0.01
7.06 ± 0.02
7.99 ± 0.02
7.25 ± 0.01
8.27 ± 0.03
7.50 ± 0.02
9.22 ± 0.05
6.95 ± 0.01
7.91 ± 0.01
7.67 ± 0.02
8.21 ± 0.01
7.00 ± 0.01
73
6.6
1.6
26
76±3.4
96 ± 0.9
93 ± 0.9
107 ± 1.2
106 ± 0.7
88 ± 1.1
98 ± 1.1
101 ± 0.9
103 ± 1.6
98 ± 1.1
97 ± 1.5
87 ± 0.8
99 ± 0.8
97 ± 1.0
99 ± 0.6
93 ± 0.8
7.02
8.16
8.76
7.61
8.61
7.00
7.99
7.26
8.29
7.47
9.22
6.89
7.92
7.66
8.20
6.97
NA
NA
725
111
NA
84
NA
55 ± 8.8
64 ± 5.12
NA
NA
5.88
6.76
NA
5.63±0.09
6.96 ± 0.02
7.43 ± 0.02
6.23 ± 0.02
6.96 ± 0.02
NA
2320
109
37
63±4.5
66 ± 0.6
86 ± 1.0
15 ± 0.2
74 ± 1.0
NA
5.43
(–)ꢀ7a
6.14 ± 0.18
6.96 ± 0.12
NA
6.78
7.37
5.41
6.82
NA
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
(
(
+)ꢀ7a
–)ꢀ7b
585
111
NA
392
NA
213
393
22
O
NH2
O
(+)ꢀ7b
–)ꢀ7c
2.6
87
7.07 ± 0.03
NA
50 ± 0.8
NA
6.77
NA
(
NA
NA
NA
348
448
46
(+)ꢀ7c
(–)ꢀ7d
10
NA
NA
NA
6.41 ± 0.02
NA
21 ± 0.3
NA
5.73
NA
56
NA
NA
NA
(
+)ꢀ7d
–)ꢀ7e
+)ꢀ7e
5.3
33
6.46 ± 0.06
6.35 ± 0.04
7.34 ± 0.06
6.29 ± 0.06
6.47 ± 0.05
6.34 ± 0.24
6.76 ± 0.04
6.23 ± 0.05
31 ± 1.1
34 ± 0.9
103 ± 2.8
27 ± 1.2
27 ± 0.8
25 ± 4.2
39 ± 1.0
29 ± 1.2
5.95
5.88
7.35
5.71
6.18
5.74
6.35
5.70
6.67 ± 0.02
6.41 ± 0.02
7.65 ± 0.05
NA
46 ± 0.8
22 ± 0.4
98 ± 1.8
NA
6.33
5.75
7.65
NA
(
(
0.59
113
12
(–)ꢀ7f
(+)ꢀ7f
524
180
453
174
585
NA
456
938
298
NA
6.34 ± 0.03
6.03 ± 0.11
6.53 ± 0.03
NA
25 ± 0.6
38 ± 3.8
79 ± 1.4
NA
5.74
5.61
6.42
NA
(
–)ꢀ7g
21
(
(
+)ꢀ7g
–)ꢀ7h
6.2
99
ACS Paragon Plus Environment

Page 19 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
+)ꢀ7h
8.03 ± 0.01
6.58 ± 0.03
7.60 ± 0.02
6.99 ± 0.01
8.57 ± 0.02
6.94 ± 0.01
7.84 ± 0.02
9.2
261
25
103 ± 0.8
70 ± 0.9
99 ± 0.9
88 ± 0.7
101 ± 1.0
76 ± 0.8
100 ± 0.8
8.05
6.43
7.60
6.94
8.57
6.82
7.82
6.89 ± 0.03
6.14 ± 0.07
6.82 ± 0.03
NA
128
733
150
NA
60
38 ± 0.6
33 ± 1.7
40 ± 0.6
NA
6.47
5.65
6.43
NA
6.45 ± 0.02
NA
358
NA
260
NA
939
NA
49 ± 0.7
NA
6.14
NA
(–)ꢀ7i
(+)ꢀ7i
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
6.59 ± 0.02
NA
19 ± 0.3
NA
5.86
NA
(
–)ꢀ8a
101
2.7
115
15
(
(
+)ꢀ8a
–)ꢀ8b
7.22 ± 0.03
NA
55 ± 0.7
NA
6.96
NA
6.03 ± 0.02
NA
72 ± 1.6
NA
5.88
NA
NA
309
(+)ꢀ8b
–)ꢀ8c
+)ꢀ8c
6.51 ± 0.04
36 ± 0.9
6.07
5.58 ± 0.01 2598
35 ± 0.5
5.13
(
NA
NA
81
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
(
7.09 ± 0.01
96 ± 0.6
7.07
(–)ꢀ8d
7.09 ± 0.01
81
84 ± 1.7
7.02
NA
NA
NA
NA
NA
NA
NA
NA
(
+)ꢀ8d
–)ꢀ8e
(+)ꢀ8e
–)ꢀ8f
+)ꢀ8f
7.96 ± 0.01
NA
11
NA
23
106 ± 0.9
NA
7.98
NA
6.26 ± 0.19
NA
550
NA
433
NA
NA
27 ± 3.7
NA
5.69
NA
5.64
NA
NA
5.66 ± 0.04 2210
44 ± 2.0
NA
5.30
NA
NA
NA
NA
(
NA
NA
NA
NA
NA
NA
NA
NA
7.65 ± 0.02
6.35 ± 0.02
6.86 ± 0.01
102 ± 0.9
69 ± 1.0
101 ± 0.6
7.65
6.19
6.86
6.36 ± 0.15
NA
19 ± 2.1
NA
NA
(
446
138
NA
(
NA
NA
NA
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 57
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
a
All new compounds were tested as HCl salts except compounds (
–)ꢀ6a and (+)ꢀ6a which are TFA salts; pharmacological data were acquired
with recombinant, stably expressed human 5ꢀHT2C, 5ꢀHT2A, and 5ꢀHT2B receptors in the HEKꢀ293 cell line, using a fluorescence imaging plate
reader (FLIPR) assay; pEC50 and Emax values are shown as mean ± SEM (n ≧ 2); EC50 values were calculated from averaged pEC50s; “NA”, no
activity up to 10 ꢁM; “ꢀ”, structures of serotonin and lorcaserin are not shown.
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
ACS Paragon Plus Environment

Page 21 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
As can be seen from Table 1, most of the new compounds show potent agonism of the
5ꢀHT2C receptor and display selectivity against the 5ꢀHT2A and 5ꢀHT2B receptors. Overall the
(+)ꢀisomers are more potent than their (–)ꢀisomers within each pair of enantiomers, and
increasing the size of the alkoxyl group at position 5 of the 2,3ꢀdihydrobenzofuran or
benzofuran scaffold decreases potency in most cases, while enhancing compound selectivity
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
9, 25
for the 5ꢀHT2C receptor. These observations are consistent with our previous findings.
For
compounds 6a–f which bear the 2,3ꢀdihydrobenzofuran scaffold 6, ethoxy substitution
provides the best potency (compound (+)ꢀ6b: EC50 = 2.7 nM, Emax = 106%, log(Emax/EC50) =
8.59), whereas the elongation of carbon chain leads to a significant loss of potency as
exemplified by compound (+)ꢀ6c (EC = 20 nM). Maintaining the fourꢀatom length of the
5
0
alkoxy chain, allyloxy substitution in compound (+)ꢀ6d (EC50 = 11 nM) increases compound
potency while cyclopropylmethyoxy slightly decreases the activity (compound (+)ꢀ6e, EC50
6 nM). The presence of a 2ꢀfluoroethoxy substituent as in compound (+)ꢀ6f results in a
=
2
reduction in potency at the 5ꢀHT2C receptors (EC = 73 nM), but a gain of selectivity against
50
the 5ꢀHT2B receptors was achieved as (+)ꢀ6f shows no 5ꢀHT2B activation at concentrations up
to 10 ꢁM.
Changing from the 2,3ꢀdihydrobenzofuran backbone to benzofuran leads to an overall
enhancement of 5ꢀHT2C potency, as observed for compound 7a–7i. Compared to compound
(
+)ꢀ6a, (+)ꢀ7a shows improved potency at 5ꢀHT2C (EC = 1.6 nM, E = 93%) as well as
50 max
better selectivity against 5ꢀHT2B (111/1.6 = 69ꢀfold) and 5ꢀHT2A (37/1.6 = 23ꢀfold). For the
propyl compound 7c, the (+)ꢀenantiomer shows 10 nM potency at 5ꢀHT2C receptors, no
activity at 5ꢀHT2B, and good selectivity against 5ꢀHT2A (EC50 = 392 nM). Compared to
lorcaserin, (+)ꢀ7c represents one of the best ligands in this series of compounds in terms of
both potency and selectivity. Our efforts in introducing other alkoxy substitutions as shown
for compounds 7d–7i did not lead to any significant improvements in selectivity as compared
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 57
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
to compound (+)ꢀ7c, but the majority of these compounds showed high potency and moderate
selectivity. Compound (+)ꢀ7e which bears the 2ꢀfluoroethoxy substitution, displays the best
potency among all compounds (EC50 = 0.59 nM).
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
The medium sized ethyl group was introduced to position 2 of the benzofuran backbone to
further explore the ligand chemical space at this position. As can be seen from compounds
8a–8f, the additional ethyl group slightly decreases the 5ꢀHT2C potency of these compounds
compared to compounds 7a–7i, but provides much better selectivity against the 5ꢀHT2B and
5ꢀHT2A receptors. Among them, compound (+)ꢀ8d showed an EC50 value of 11 nM at 5ꢀHT2C
receptors, 50ꢀfold selectivity against 5ꢀHT2B (EC = 550 nM and over 200ꢀfold selectivity
50
against 5ꢀHT2A (EC50 = 2210 nM). Given that log(Emax/EC ) values of (+)ꢀ8d at both the
50
5
ꢀHT2B and 5ꢀHT2A receptors are much lower (5.69 and 5.30, respectively), we would
consider this ligand to be more selective than lorcaserin (5ꢀHT2B log(Emax/EC50) = 6.27;
ꢀHT2A log(Emax/EC50) = 6.44). Compound (+)ꢀ8e (EC50 = 23 nM) is 2ꢀfold weaker than
+)ꢀ8d at 5ꢀHT , but shows similar weak 5ꢀHT activity (EC = 433 nM, E = 19%) and
5
(
2C
2B
50
max
no 5ꢀHT2A activation. Compounds (+)ꢀ8c and (+)ꢀ8f are weaker in terms of their activation of
ꢀHT2C receptors (EC50 = 81 and 138 nM respectively), but both of them show excellent
selectivity against the other two receptors as no agonist activity was observed for either
5
5ꢀHT2B or 5ꢀHT2A.
Taken together, the majority of the compounds bearing the new scaffolds 6–8 are selective
ꢀHT2C agonists showing moderate to high selectivity against both the 5ꢀHT2A and 5ꢀHT2B
receptors. Given the fact that the Emax ranges of most of these compounds at 5ꢀHT2A and
ꢀHT2B receptors are much lower than those at 5ꢀHT2C receptors, a number of compounds
5
5
such as (+)ꢀ6d, (+)ꢀ7c, (+)ꢀ8d, and (+)ꢀ8e represent good candidates for further studies. It is
important to note that some of these compounds bind to the 5ꢀHT2A and 5ꢀHT2B receptors at
moderate to weak affinities (see Supporting Information, Table S1), but their low intrinsic
ACS Paragon Plus Environment

Page 23 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
efficacy exclude them from mediating side effects through activation of those receptors.
Although these compounds possess close structural similarity with the melatonin agonist
tasimelteon, no activity at the melatonin receptors was anticipated, as the melatonin receptor
31
agonists bear an amide functional group rather than a primary amine. We did in fact test
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
some of these compounds at melatonin MT and MT receptors examining G function, and
1
2
i/o
observed either no activity or very weak agonist activity (EC50 > 1 µM) at both receptors (see
Supporting Information, Figures S1 and S2).
5ꢀHT2C Functional Selectivity
Agomelatine has been investigated for its antidepressant potential and shown to be a neutral
8
antagonist blocking both G
q
activity and βꢀarrestin recruitment at 5ꢀHT2C receptors. Despite
this study, few 5ꢀHT2C ligands have been thoroughly and systematically investigated for
βꢀarrestin recruitment and ligand bias. Therefore, we sought to investigate βꢀarrestin
recruitment efficacy with the above 5ꢀHT2Cꢀselective benzofuran ligands and to compare
their functional selectivity profile to lorcaserin and 5ꢀHT. We measured βꢀarrestin
32
recruitment activity using a modified reporter assay, and compared activities to
q
G ꢀmediated calcium flux for our newly synthesized 5ꢀHT2Cꢀselective benzofuran ligands,
which was performed in parallel using the same drug dilutions and conditions. Our results
indicate that lorcaserin recruited βꢀarrestin with almost full efficacy compared to 5ꢀHT (Table
2, Figure 2). By contrast, all tested benzofuran compounds exhibited diminished βꢀarrestin
recruitment efficacy (range 22–49%) compared to lorcaserin (89%) and 5ꢀHT (100%),
indicating that the 5ꢀHT2Cꢀselective benzofuran ligands exhibit functional selectivity with
preference for Gqꢀmediated calcium flux (Figure 2). To quantify the extent of preference for
17
Gq signaling, we calculated bias factors using relative activities (log Emax/EC50) for each
respective pathway. The relative activity (log Emax/EC ) of compounds can be used as a
50
A
surrogate for the Black and Leff operational model estimation of log (τ/K ) provided that the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 57
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
3
3
hill slope does not substantially deviate from 1, which in the case of these compounds it did
not. Using lorcaserin as the reference ligand to distinguish our newly synthesized
5
ꢀHT2Cꢀselective benzofuran ligands from it, we found that compound (+)ꢀ7e showed the best
preference for Gq signaling (bias factor = 3.8, Table 2). This Gq preference of (+)ꢀ7e is
mainly driven by its extremely weak arrestin recruitment efficacy (Emax = 28%), thus
demonstrating functionally selective effects compared to lorcaserin and 5ꢀHT.
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
ACS Paragon Plus Environment

Page 25 of 57
Journal of Medicinal Chemistry
5
ꢀHT
Lorcaserin
Ro 60ꢀ0175
1
25
00
125
100
75
50
25
0
125
100
75
50
25
0
Gq-Calcium Flux
β-Arrestin2
1
1
2
3
4
5
6
7
8
9
7
5
2
5
0
5
0
ꢀ
13 ꢀ12 ꢀ11 ꢀ10
ꢀ9
ꢀ8
ꢀ7
ꢀ6
ꢀ5
ꢀ4
ꢀ13 ꢀ12 ꢀ11 ꢀ10
ꢀ9
ꢀ8
ꢀ7
ꢀ6
ꢀ5
ꢀ4
ꢀ13 ꢀ12 ꢀ11 ꢀ10
ꢀ9
ꢀ8
ꢀ7
ꢀ6
ꢀ5
ꢀ4
log [drug] M
log [drug] M
log [drug] M
(
+)ꢀ6d
(+)ꢀ6f
(+)ꢀ7d
1
25
125
100
75
50
25
0
125
100
75
50
25
0
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
100
7
5
2
5
0
5
0
ꢀ13 ꢀ12 ꢀ11 ꢀ10
ꢀ9
ꢀ8
ꢀ7
ꢀ6
ꢀ5
ꢀ4
ꢀ13 ꢀ12 ꢀ11 ꢀ10
ꢀ9
ꢀ8
ꢀ7
ꢀ6
ꢀ5
ꢀ4
ꢀ13 ꢀ12 ꢀ11 ꢀ10
ꢀ9
ꢀ8
ꢀ7
ꢀ6
ꢀ5
ꢀ4
log [drug] M
log [drug] M
log [drug] M
(+)ꢀ7e
(+)ꢀ8d
(+)ꢀ8e
1
25
125
100
75
50
25
0
125
100
75
50
25
0
100
75
50
25
0
ꢀ13 ꢀ12 ꢀ11 ꢀ10
ꢀ9
ꢀ8
ꢀ7
ꢀ6
ꢀ5
ꢀ4
ꢀ13 ꢀ12 ꢀ11 ꢀ10
ꢀ9
ꢀ8
ꢀ7
ꢀ6
ꢀ5
ꢀ4
ꢀ13 ꢀ12 ꢀ11 ꢀ10
ꢀ9
ꢀ8
ꢀ7
ꢀ6
ꢀ5
ꢀ4
log [drug] M
log [drug] M
log [drug] M
Figure 2. Parallel 5ꢀHT screening for functional selectivity of 5ꢀHT ꢀselective compounds examining Gqꢀmediated calcium flux
2C
2C
(
FLIPR, red) and βꢀarrestinꢀ2 recruitment (Tango, blue). Data were acquired with the human 5ꢀHT INI receptor isoform measuring
2C
Gqꢀcalcium flux (FLIPR) or βꢀarrestinꢀ2 recruitment (Tango) performed in triplicate. Assays were conducted in parallel with the same
drug dilutions
2
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 57
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
a
Table 2. Assessment of functional selectivity for 5ꢀHT2Cꢀselective compounds.
GqꢀCalcium Flux
βꢀarrestinꢀ2
BIAS
Factor Factor
Gq) (βꢀArr)
BIAS
Compound
EC50
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
pEC50
E
max
Log(Emax / EC50
)
pEC50
EC50 (nM)
E
max
Log(Emax / EC50)
(
(nM)
lorcaserin
8.46 ± 0.03
3.4
96 ± 1.1
100 ± 0.8
99 ± 1.2
93 ± 1.2
93 ± 1.3
92 ± 1.3
94 ± 0.8
93 ± 1.3
94 ± 0.8
8.45
7.23 ± 0.03
59
21
91 ± 1.2
7.19
1.0
8.2
0.3
1.5
0.9
0.5
3.8
0.4
0.2
1.0
0.12
3.3
0.7
1.1
2.0
0.3
2.5
5.0
(reference)
5
ꢀHT
9.87 ± 0.03
0.14
2.2
26
9.85
8.65
7.55
7.36
8.02
8.89
7.63
7.58
7.66 ± 0.02
7.97 ± 0.04
6.57 ± 0.06
6.49 ± 0.04
7.71 ± 0.01
7.60 ± 0.06
7.02 ± 0.04
7.52 ± 0.08
100 ± 1.0
87 ± 1.2
35 ± 1.0
42 ± 0.8
25 ± 0.9
28 ± 0.7
50 ± 1.0
35 ± 1.1
7.68
7.94
6.12
6.12
7.10
7.05
6.72
7.07
Ro 60ꢀ0175 8.65 ± 0.04
10
(
+)ꢀ6d
+)ꢀ6f
(+)ꢀ7d
+)ꢀ7e
+)ꢀ8d
+)ꢀ8e
7.58 ± 0.03
7.39 ± 0.03
8.06 ± 0.04
8.92 ± 0.03
7.65 ± 0.04
7.60 ± 0.04
267
318
20
(
41
8.7
1.2
22
(
25
(
96
(
25
30
a
Data were acquired with the human 5ꢀHT2C INI receptor isoform measuring Gqꢀcalcium flux (FLIPR) and βꢀarrestinꢀ2 recruitment (Tango);
pEC and E values are shown as mean ± SEM (n = 3) performed in triplicate and assays were conducted in parallel with the same drug
50
max
2
6
ACS Paragon Plus Environment

Page 27 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
dilutions. Bias factors toward Gq or βꢀarrestinꢀ2 were calculated using change in respective relative activity, log (Emax/EC50), as described
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
17
previously.
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
2
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 57
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
5
ꢀHT2C Desensitization. Considering that arrestin recruitment is reported to cause
22
desensitization at numerous GPCRs and that compound (+)ꢀ7e exhibited weak arrestin
recruitment efficacy, we sought to examine the 5ꢀHT desensitization properties comparing full
2
C
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
arrestin agonist 5ꢀHT to weak partial arrestin agonist (+)ꢀ7e. To measure Gqꢀmediated 5ꢀHT2C
desensitization, we first preꢀtreated 5ꢀHT ꢀexpressing HEK cells with increasing concentrations
2
C
of either 5ꢀHT or (+)ꢀ7e for 30 min and then challenged cells after washout to release calcium
upon further drug stimulation. Preꢀtreatment with 5ꢀHT concentrations as low as 32 nM
depressed the absolute Emax without affecting 5ꢀHT’s potency, and a 10 µM 5ꢀHT preꢀtreatment
concentration completely abolished any additional 5ꢀHTꢀstimulated calcium flux (Figure 3a). By
contrast, (+)ꢀ7e did not show Emax depression at preꢀtreatment concentrations below 1 µM
(Figure 3b), suggesting that (+)ꢀ7e was not as efficient to desensitize 5ꢀHT2C Gqꢀmediated
calcium flux compared to similar 5ꢀHT concentrations.
Although calcium flux is commonly investigated as a functional readout of G ꢀcoupled receptor
q
desensitization, calcium flux does not allow full receptor occupancy given its short time
duration. In addition, calcium flux is downstream from PLCꢀmediated phosphoinositide (PI)
hydrolysis, which has been investigated as functional readout of 5ꢀHT2C desensitization by
3
34
measuring Hꢀinositol phosphates (IP) accumulation over time. Therefore, we compared full
arrestin agonist 5ꢀHT to partial arrestin agonist (+)ꢀ7e expecting to observe a loss of IP
accumulation over time indicative of 5ꢀHT2C desensitization. First, 5ꢀHT ꢀexpressing cells
2C
3
produced expected elevated basal levels of Hꢀinositol phosphates (Figure 3e, f), consistent with
35, 36
a high level of constitutive activity at this receptor isoform.
Despite the elevated level of
3
Hꢀphosphoinositides, both 5ꢀHT and (+)ꢀ7e at maximum receptor occupancy (1 µM, Figure S3
3
in Supporting Information) produced a rapid and robust increase in Hꢀphosphoinositides after
28
ACS Paragon Plus Environment

Page 29 of 57
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
only 5 min drug incubation peaking around 60 min (Figure 3c). In fact, IP accumulation for both
5
ꢀHT and (+)ꢀ7e was sustained up to 5 hours (300 min) with no depression in absolute Emax
(Figure 3e, f). 5ꢀHT incubation at longer time points, however, showed decreasing IP
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
accumulation at 600 min (10 h) with a gradual reduction back to basal levels at 40 hours (Figure
3d). Although the 5ꢀHT Emax was completely abolished by at least 30 hours of incubation (Figure
3e), (+)ꢀ7e produced a sustained IP accumulation and only began to show a significant decrease
in Emax after 20 hours of drug incubation (Figure 3d). Comparison of the desensitization rates of
5ꢀHT and (+)ꢀ7e revealed that (+)ꢀ7e at 1 µM produced slower reduction in IP accumulation over
time (Figure S4). In fact, (+)ꢀ7e still retained a doseꢀresponse at 30 h (Figure 3f), whereas 5ꢀHT
did not. Given that (+)ꢀ7e was previously found to produce weaker arrestin recruitment
compared to 5ꢀHT, these results support the finding that weaker desensitization is a unique
contributor to (+)ꢀ7e’s functional selectivity profile.
29
ACS Paragon Plus Environment