Discovery of 4-Phenylpiperidine-2-Carboxamide Analogues as Serotonin 5-HT2CReceptor-Positive Allosteric Modulators with Enhanced Drug-like Properties

Article abstract of DOI:10.1021/acs.jmedchem.9b01953

Targeting the serotonin (5-HT) 5-HT2C receptor (5-HT2CR) allosteric site to potentiate endogenous 5-HT tone may provide novel therapeutics to alleviate the impact of costly, chronic diseases such as obesity and substance use disorders. Expanding upon our recently described 5-HT2CR-positive allosteric modulators (PAMs) based on the 4-alkylpiperidine-2-carboxamide scaffold, we optimized the undecyl moiety at the 4-position with variations of cyclohexyl- or phenyl-containing fragments to reduce rotatable bonds and lipophilicity. Compound 12 (CTW0415) was discovered as a 5-HT2CR PAM with improved pharmacokinetics and reduced off-target interactions relative to our previous series of molecules. The in vivo efficacy of compound 12 to potentiate the effects of a selective 5-HT2CR agonist was established in a drug discrimination assay. Thus, 12 is reported as a 5-HT2CR PAM with characteristics suitable for in vivo pharmacological studies to further probe the biological and behavioral mechanisms of allosteric modulation of a receptor important in several chronic diseases.

Full text of DOI:10.1021/acs.jmedchem.9b01953

Subscriber access provided by LUNDS UNIV
Article
Discovery of 4-Phenylpiperidine-2-Carboxamide
Analogues as Serotonin 5-HT2C Receptor Positive
Allosteric Modulators with Enhanced Drug-Like Properties
Eric A. Wold, Erik J. Garcia, Christopher Wild, Joanna M. Miszkiel, Claudia A Soto, Jianping Chen,
Konrad Pazdrak, Robert G Fox, Noelle C. Anastasio, Kathryn A. Cunningham, and Jia Zhou
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b01953 • Publication Date (Web): 22 Jun 2020
Downloaded from pubs.acs.org on June 22, 2020
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 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery of 4-Phenylpiperidine-2-Carboxamide Analogues as
Serotonin 5-HT Receptor Positive Allosteric Modulators with
2C
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
Enhanced Drug-Like Properties
Eric A. Wold,†,‡,ǁ Erik J. Garcia, Christopher T. Wild, Joanna M. Miszkiel, Claudia A.
†,ǁ
†,‡,ǁ
†
†
†,‡
†
†
†,‡,
Soto, Jianping Chen, Konrad Pazdrak, Robert G. Fox, Noelle C. Anastasio, * Kathryn A.
†
,‡,
†,‡,
Cunningham, * and Jia Zhou *
†
‡
Center for Addiction Research and Chemical Biology Program, and the Department of
Pharmacology and Toxicology, University of Texas Medical Branch, Galveston, TX 77555,
United States
ǁAuthors contributed equally to this work
*
Corresponding authors
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 56
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
Targeting the serotonin (5-HT) 5-HT2C receptor (5-HT2CR) allosteric site to potentiate endogenous
-HT tone may provide novel therapeutics to alleviate the impact of costly, chronic diseases such
5
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 obesity and substance use disorders. Expanding upon our recently described 5-HT2CR positive
allosteric modulators (PAMs) based on the 4-alkylpiperidine-2-carboxamide scaffold, we
optimized the undecyl moiety at the 4-position with variations of cyclohexyl- or phenyl-containing
fragments to reduce rotatable bonds and lipophilicity. Compound 12 (CTW0415) was discovered
as a 5-HT2CR PAM with improved pharmacokinetics and reduced off-target interactions relative
to our previous series of molecules. The in vivo efficacy of compound 12 to potentiate the effects
of a selective 5-HT2CR agonist was established in a drug discrimination assay. Thus, 12 is reported
as a 5-HT2CR PAM with characteristics suitable for in vivo pharmacological studies to further
probe the biological and behavioral mechanisms of allosteric modulation of a receptor important
in several chronic diseases.
ACS Paragon Plus Environment

Page 3 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Introduction
Neuropsychiatric and metabolic disorders currently encompass exceedingly costly and
intractable diseases both in the U.S. and worldwide. Indeed, obesity and substance use disorders
(SUDs) can manifest as chronic conditions that dramatically reduce life expectancy. The serotonin
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
(
5-HT) 5-HT2C receptor (5-HT2CR) is a predominately central nervous system (CNS) G protein-
coupled receptor (GPCR) that modulates key disease-related neurological pathways and
1
behaviors. For example, 5-HT2CR agonists suppress food intake and increase satiety in humans
and animals,2-6 and preclinical studies indicate a major role for the 5-HT2CR in modulating the
rewarding and incentive-salience value of psychostimulants (e.g., cocaine), opioids, and ethanol
as well as behavioral factors predictive of relapse during SUD recovery.7-11 In human studies, the
5-HT2CR agonist lorcaserin increased the rate of cessation from tobacco smoking, and decreased
corticolimbic activation evoked by presentation of palatable food cues, thus implicating the 5-
HT2CR as a target mechanism for disorders characterized by reward and cue-associated events.12,
1
3
Therefore, the 5-HT2CR is a favorable therapeutic target at the intersection of disinhibited
behaviors that promote obesity and SUDs, and a pharmacotherapy that could aid in the successful
treatment of SUDs would be a first-in-class medication.14, 15
Selective targeting of the 5-HT2CR remains challenging since each of the 13 5-HT GPCR
subtypes share a conserved orthosteric binding site for the endogenous agonist 5-HT. Among the
5-HT R subtypes (5-HT2AR, 5-HT2BR, and 5-HT2CR), 5-HT2CR selectivity is necessary to avoid
2
the potential for hallucinations or cardiac valvulopathy that may result as a consequence of 5-
HT2AR or 5-HT2BR stimulation, respectively.16, 17 One possible strategy to achieve selective 5-
1
8
HT2CR modulation is the rational design of positive allosteric modulators (PAMs). By targeting
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 56
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 binding site that is spatially distinct from the conserved 5-HT orthosteric site, divergent residues
or topological surfaces may be exploited to achieve 5-HT2CR selectivity while enhancing the
1
9
20
functional response to 5-HT. This strategy has been employed for numerous class A GPCRs
and specifically for the discovery of 5-HT2CR PAMs by our group and others (Figure 1).21-25
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. Highlighted compounds resulting from the search for synthetic small molecule 5-
HT2CR PAMs. Highlighted here are lead compounds from the published 5-HT2CR PAM drug
discovery campaigns. The ClogP, calculated in ChemDraw 18.0, for each compound displays a
trend in high lipophilicity.
Our previously described 5-HT2CR PAMs were obtained by an iterative exploration and
optimization process beginning with the initial 5-HT2CR PAM small molecule PNU-69176E (1).
The first priority in this process was the simplification of the α-D-galactopyranoside fragment,
termed the polar head (PH), to provide a more drug-like compound for in vitro and in vivo proof
of concept studies (Figure 2). From this effort, CYD-1-79 (3) was obtained and characterized as
2
3
a selective 5-HT2CR PAM (Figure 1). The PH moiety of 3 was modified to retain 5-HT2CR PAM
activity while engendering decreased molecular weight and complexity. Next, modifications to the
ACS Paragon Plus Environment

Page 5 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
undecyl substituent at the 4-position of the piperidine, which is termed the lipophilic tail (LT),
were probed as an option for further improvements. In the current study, we chemically modified
the LT position with shorter, compact fragments to reduce lipophilicity while maintaining 5-
HT2CR PAM activity (Figure 2). Additionally, we rationalized that the compact LT substituents
would reduce the number of rotatable bonds and improve the drug-likeness of the 4-
alkylpiperidine-2-carboxamide scaffold, which was further demonstrated by the calculated
physicochemical properties for the synthesized test compounds herein (Table S1). Importantly,
modifying the undecyl LT position with compact, cyclic moieties – providing increased structural
rigidity – yielded small molecules that are advantageous for computational molecular modeling
and docking studies to ultimately illuminate features of the 5-HT2CR PAM binding site.
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. Medicinal chemistry strategy to achieve optimized 5-HT2CR PAMs from PNU-
69176E (1). A lincomycin derivative, PNU-69176E (1), was the first synthetic small molecule
reported to allosterically modulate the 5-HT2CR. Subsequent work has focused on chemical
modifications of the polar head (PH) and the lipophilic tail (LT) to engender drug-like properties.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 56
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
Chemistry. Allosteric modulation of the 5-HT2CR is a new endeavor with a small number of probe
compounds identified in recent years.23-25 Our chemical approach is focused on the optimization
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
of our previously reported compound 3. Elevation of intracellular calcium (Ca 2+
downstream impact of 5-HT2CR activation which is oft utilized as a readout in functional cellular
) release is a key
i
26
2+
assays. Compound 3 was shown to enhance Ca
i
release ~23% above the maximal effect of 5-
HT alone in Chinese hamster ovary (CHO) cells stably transfected with human (h)5-HT2CR. In the
absence of 5-HT, 3 had no effect on Ca 2+
i
release, indicating the PAM nature of this compound.
Compound 3 also exhibited in vivo efficacy to decrease locomotor activity and suppress cocaine-
seeking during abstinence from cocaine self-administration in rats; 3 potentiated the substitution
2
3
of low doses of the 5-HT2CR agonist WAY163909 in a drug discrimination assay. Therefore, the
,2-diol moiety in the PH position of 3 was considered a privileged fragment and was retained.
1
Additional PH fragments were employed in the following chemical series based upon their
previously observed activity or inactivity as a basis of comparison.
Figure 3. The volume, not length, of the undecyl moiety on PNU-69176E (1) and CYD-1-79
(3) may contribute to 5-HT2CR PAM activity. Compounds 1 and 3 favor similar minimized
conformations which leads to overlapping hydroxyl moieties in the PH and to a folded substructure
for the undecyl LT. The preferred, contracted substructure for the undecyl LT suggests that
hydrophobic volume is important.
ACS Paragon Plus Environment

Page 7 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Initial structure-activity relationship (SAR) studies of compound 1 showed that shorter
alkyl chains (methyl to n-methyl groups) as LTs failed to exert positive allosteric modulatory
21
actions at the 5-HT2CR; however we reasoned that the volume of the LT, not the length, may play
a role in compound activity. This hypothesis was based partially upon in silico ligand minimization
and electrostatic complementarity experiments comparing compound 1 with compound 3 (Figure
3). First, the modeling shows how the replacement of the α-D-galactopyranoside PH fragment with
the simplified 1,2-diol fragment found in 3 retains key hydroxyl group orientations and thus
potentially conserve H-bond partners and interaction distances. Second, the minimized alignment
of the undecyl LT fragments are shown in a conformation that suggests an exploitation of a
hydrophobic binding pocket due to lipophilic volume. Thus, we designed new small molecules to
incorporate shorter LTs with aromatic and aliphatic rings able to provide a comparable volume to
the longer alkyl chain counterparts. Crucially, these new analogues reduce the overall number of
rotatable bonds and engender enhanced drug-like properties and development potential.
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
Scheme 1. Synthesis of CYD-1-79 analogues with compact phenyl or cyclohexyl LTs
Reagents and conditions: a) phenylboronic acid, Pd(PPh
3
)
4
, Na
2
CO
3
, EtOH/H
2
O/PhMe = 2:1:1, reflux. b)
O (v/v = 2:1), rt,
H
2
, PtO , HCl/MeOH/H O, rt, 16 h. c) (Boc) O, Et N, MeOH, rt, 16 h. d) LiOH, THF/H
2
2
2
3
2
48 h. e) amino alcohols, HBTU, DIPEA, DMF, rt, 16 h. f) TFA, CH
2
Cl , rt.
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 56
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
Similar to our previously established synthetic route, a six-step protocol from 5 affords the
target compounds.22, 23 The iodinated ester of picolinic acid (5) underwent Suzuki (Scheme 1) or
Sonogashira coupling (Scheme 2) to provide incorporation of various LTs that included aromatic
rings (e.g., phenyl, phenethyl, 4’-methylphenethyl, and 4’-t-butylphenethyl fragments). Metal-
catalyzed hydrogenation, Boc-protection, and hydrolysis readily provided the (cis-2,4)-piperidyl
carboxylic acids 9-10 and 25-28. As a result of the metal-catalyzed reduction conditions, the
pyridine ring was consistently reduced, whereas the aromatic LTs either did not undergo
hydrogenation or were fully reduced into separable products. Therefore, isolable aliphatic versions
of the aromatic LTs were obtained, and upon HBTU-mediated coupling to suitable amines and N-
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
Boc deprotection, target compounds 11-17 and 29-39 were obtained. All isomers with the 2-amino-
23
1
-phenyl-1,3-propanediol PH were separable, consistent with our previous report. All other
compounds were an inseparable mixture of isomers and were tested as such (Schemes 1 & 2).
Scheme 2. Synthesis of CYD-1-79 analogues with phenethyl, ethylcyclohexyl, or substituted
phenethyl LTs
Reagents and conditions: a) Phenylacetylenes, CuI, Pd(PPh
HCl/MeOH/H O, rt, 16 h. c) (Boc) O, Et N, MeOH, rt, 16 h. d) LiOH, THF/H
amino alcohols, HBTU, DIPEA, DMF, rt, 16 h. f) TFA, CH Cl , rt.
3
)
2
Cl
2
, Et
3
N, rt, 12 h. b) H
2
, PtO ,
2
2
2
3
2
O (v/v = 2:1), rt, 48 h. e)
2
2
ACS Paragon Plus Environment

Page 9 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Compound Screening In Vitro. A very well characterized intracellular signaling pathway of the
5
-HT2CR is the stimulation of phospholipase Cβ via Gαq/11 proteins leading to the production of
inositol-1,4,5-trisphosphate and diacylglycerol, resulting in Ca 2+
release from intracellular
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
27
2+
stores. Thus, a fluorescence-based, in vitro Ca
i
release assay was utilized to measure activation
of this 5-HT2CR signaling pathway and provide functional characterization of the newly
synthesized compounds. Chinese hamster ovary (CHO) cells stably transfected with the human 5-
HT2CR (unedited INI isoform; h5-HT2CR-CHO cells) were used for screening of compounds in
2+
2+
the Ca
i
release assay. The maximum 5-HT-induced Ca
i
release (Emax) was measured and
normalized to 100% response. All test compounds were assessed in 4-6 biological replicates, each
conducted in technical triplicates and results are shown relative to the Emax of 5-HT alone.
Compound 1 increased 5-HT-evoked release in titration studies and evoked a leftward shift of the
2
+
22
5
-HT-evoked Ca
i
response curve. Therefore, the initial compound screen was conducted at 1
nM to determine the extent to which compounds enhanced Ca 2+
release elicited by increasing
i
concentrations of 5-HT. Each test compound was added to the h5-HT2CR-CHO cells 15 min prior
to the addition of 5-HT to determine intrinsic agonist activity. None of the compounds tested
exhibited intrinsic activity to induce Ca 2+
Figure S1). The assessment of compounds tested in the Ca 2+
release assay in h5-HT2CR-CHO
release in h5-HT2CR-CHO prior to the addition of 5-HT
i
(
i
cells is shown in Table 1 and Table 2, and all concentration response curve shift plots are in
Figure S1.
Of the 22 compounds tested, several displayed 5-HT2CR PAM activity comparable to the
previously reported 3 which displayed an approximate 23% increase in the Emax for 5-HT-induced
2+
23
Ca
i
release in h5-HT2CR-CHO cells. Compounds (2S,4R)-11 and (2R,4S)-11 included a phenyl
ring as the LT and incorporated a (1S,2S)-2-amino-1-phenyl-1,3-propanediol PH (Table 1). In our
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 56
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
previous report, when combined with an undecyl LT, the (1S,2S)-2-amino-1-phenyl-1,3-
propanediol PH displayed negative allosteric modulatory (NAM) activity in (2S,4R)
2
3
conformation. However, (2S,4R)-11 showed no activity while (2R,4S)-11 displayed modest
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
2+
allosteric activity in the Ca
i
assay. Intriguingly, both 12 and 13, which incorporate an isomer
[
(S)-1,2-diol] of the same 1,2-diol moiety PH as compound 3, displayed 5-HT-evoked Emax values
of 127.4 ± 8.79% (p<0.05; Figure 4A) and 117.9 ± 4.71% (p<0.05), respectively, close to the
2
3
activity seen for 3 (123.2 ± 4.1%, p<0.05). This outcome was observed even though the new
analogues possess a phenyl ring and a cyclohexyl ring as the LT (12 and 13, respectively), which
is an extreme departure from the eleven-carbon tail in our previously reported compounds. These
findings provide additional confirmation that the 1,2-diol moiety is a privileged fragment. The 1,3-
propandiol as a PH did promote PAM activity when coupled with a phenyl LT (14, 118.8 ± 2.92%,
p<0.05; Figure 4B). The same PH with a cyclohexyl LT (15) and the retention of the phenyl or
cyclohexyl LT with incorporation of a morpholino with a two-carbon spacer (16 and 17,
respectively) did not result in positive allosteric effects.
Table 1. Effects of 4-alkylpiperidine-2-carboxamide derivatives (1 nM) on 5-HT-induced
2+
Ca
i
release in h5-HT2CR-CHO cells
Compound
LT
Emax RFU
(% 5-HT)
PH
a
Number
3
123.2 ± 4.1*
(2S,4R)-11
102.9 ± 7.67
(
2R,4S)-11
118.2 ± 2.39*
1
1
2
3
127.4 ± 8.79*
117.9 ± 4.71*
ACS Paragon Plus Environment

Page 11 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
4
5
118.8 ± 2.92*
119.1 ± 10.20
1
1
6
7
113.6 ± 7.72
121.4 ± 12.16
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. Effects of select compounds with shortened lipophilic tail (LT) moieties provide
evidence that hydrophobic volume contributes to target engagement. Data are plotted for
compounds 12 (A), 14 (B), 35 (C), and 37 (D). The effects of these compounds on 5-HT-induced
release in h5-HT2CR-CHO cells are shown in the absence (black) and presence of the test
compound (red) against the concentration-response curve for 5-HT. The assessment of vehicle
HBSS, blue circle), or vehicle with test compound (green triangle) is also illustrated.
2+
Ca
i
(
Utilization of the PHs while incorporating either phenethyl or ethylcyclohexyl LTs resulted
in a series of molecules with a range of activity (~94% - ~117%) dependent upon the type of PH
incorporated (Table 2). Positive allosteric effects were not seen with the 1,2-diol and 1,3-diol PHs
with either aforementioned LT (compounds 30 – 33). For both the ethylcyclohexyl and phenethyl
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 56
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
LT, the two-carbon spaced morpholino as the PH did not produce PAM effects (34, 36). However,
the addition of the morpholino with a three-carbon spacer did produce PAM effects, suggesting
there is a length requirement for morpholino fragment compounds (35, 118.2 ± 3.61%, p<0.05;
Figure 4C; 37, 117.9 ± 5.90%, p<0.05; Figure 4D). An increase in the length and volume of the
LT via the inclusion of a t-butyl or methyl in the 4-position of the phenethyl LT did not lead to
active compounds (38 – 39), an interesting finding given previous suggestions that LT length is a
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
21
critical component of PAM activity for this chemotype. These data reinforce the notion that a
diol moiety PH, specifically an (S)-1,2-diol fragment, is an important structural element in 5-
HT2CR PAMs (compounds 3 and 12) and that the 5-HT2CR PAM pharmacophore is accepting of
compact, cyclic LTs.
Table 2. Effects of 4-alkylpiperidine-2-carboxamide derivatives (1 nM) on 5-HT-induced
2
+
Ca
i
release in h5-HT2CR-CHO cells
Compound
Emax RFU
(%5-HT)
LT
PH
a
Number
3
123.2 ± 4.1*
(
(
2S,4R)-29
109.8 ± 2.29*
2R,4S)-29
108.7 ± 7.91
30
108.5 ± 7.41
109.7 ± 8.42
3
3
1
2
101.0 ± 5.26
106.4 ± 6.75
3
3
3
4
110.2 ± 6.68
ACS Paragon Plus Environment

Page 13 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
3
3
5
6
118.2 ± 3.61*
94.2 ± 2.48
37
117.9 ± 5.90*
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
(2S,4R)-38
114.7 ± 10.10
106.7 ± 10.25
98.9 ± 4.15
(
2R,4S)-38
(2R,4S)-39
(
2S,4R)-39
99.3 ± 2.15
Compared to 3, in vitro characterization of 12 displayed a similar increase in the Emax of 5-
2
+
HT-induced Ca
i
release in the absence of a leftward shift in the EC50, indicating that 12 improved
the efficacy of 5-HT without a change in potency for 5-HT at the 5-HT2CR. Compound 12 was
counter-screened in h5-HT2AR-CHO cells and no alteration in the Emax or EC50 of 5-HT-induced
2
+
Ca
i
release was observed (Table 3). Likewise, the additional compounds [(2R,4S)-11, 13, 14,
29, 35, 37] that were validated in vitro as 5-HT2CR PAMs did not significantly alter the Emax or
2+
EC50 of 5-HT-induced Ca
i
release in h5-HT2AR-CHO cells. Thus, the herein identified 5-HT2C
R
PAMs exhibit selectivity against 5-HT2AR in vitro.
Table 3. Effects of active 5-HT2CR PAMs (1 nM) on 5-HT-induced Ca 2
CHO cells
+
release in h5-HT2AR-
i
Compound
Number
Emax RFU
LT
PH
a
(%5-HT)
(2R,4S)-11
102.6 ± 2.81
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 56
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
12
103.4 ± 5.79
97.8 ± 2.44
1
3
14
29
99.3 ± 1.58
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
103.9 ± 2.22
3
5
97.2 ± 4.64
97.9 ± 1.99
37
Molecular Modeling and Docking of 5-HT2CR PAMs. The narrow range of the elevated Emax
values resulting from our 5-HT2CR PAMs suggests a conserved positive allosteric modulatory
mechanism, likely originating from binding to a shared allosteric binding site. To probe possible
binding sites the Schrödinger Drug Discovery Suite was used for modeling the 5-HT2CR and for
docking the 5-HT2CR PAM molecules. This effort utilized the recently reported 5-HT2CR X-ray
28
crystal structure in complex with the orthosteric agonist ergotamine (PDB: 6BQG). Ergotamine
is larger than the 5-HT2CR endogenous agonist 5-HT and ergotamine interacts with numerous 5-
HT2CR residues throughout the transmembrane domains (TM) and extracellular loop 2 (ECL2). In
this context, PAM docking protocols are improved due to the stabilization and resultant enhanced
atomic resolution in the outer regions of the TM as well as the ECL2. However, to obtain a docking
model that is representative of the 5-HT2CR conformation encountered by 5-HT2CR PAMs in vitro
and in vivo, the Schrodinger Induced Fit Docking (IFD) utility was used to replace ergotamine
with 5-HT and allow ligand-induced conformational flexibility. Briefly, the X-ray crystal structure
of ergotamine-5-HT2CR (PDB: 6BQG) was preprocessed and optimized with the Schrödinger
Protein Preparation Wizard using default settings and each ligand (5-HT and 5-HT2CR PAMs) was
ACS Paragon Plus Environment

Page 15 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
prepared with the LigPrep tool to generate 3D conformations for docking. Serotonin was docked
to the 5-HT2CR orthosteric site via IFD with a Glide grid centered on the ergotamine indole N atom
and default values for sidechain and residue flexibility. The resultant 5-HT-5-HT2CR complex
model (Figure 5) was checked against published site-directed mutagenesis studies to assess 5-HT
orientation accuracy and was subsequently used for the remaining ligand docking.29
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. Molecular docking illustrates the considerable overlap in ligand poses of six 5-
HT2CR PAMs. Six small molecules with a 5-HT2CR PAM profile are represented in different
colors along with the endogenous agonist 5-HT. A representative pose for each 5-HT2CR PAM
was selected rationally from a top-scoring, enriched cluster docked to the 5-HT2CR X-ray crystal
structure (PDB: 6BQG). A measurement from the 5-HT (magenta) indole N atom to the piperidine
core N atom of compound 12 was found to be 13.5 Å (illustrated by the dashed line).
Selected compounds that displayed in vitro 5-HT2CR PAM functional activity were docked
to the 5-HT-5-HT2CR complex model using Glide XP precision and IFD without designated
exclusion parameters, as the model contained a 5-HT molecule to block the interaction of PAMs
with the orthosteric binding site. Scoring functions (GScore) and biological and chemical
rationales were used for ligand pose clustering and selection. The result upon docking six PAMs
[12, 13, 14, (2S,4R)-29, 35, 37] to the 5-HT-5-HT2CR model is a striking display of conformational
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 56
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
agreement at a spatially distinct allosteric site (Figure 5). The distance between the 5-HT indole
N-atom and the piperidine ring N-atom from the core of 12 is 13.5 Å. Surface mapping using
electrostatics and polarity was used for further docking site characterization with 12, which shows
that the allosteric site is positioned near extracellular-facing regions of the 5-HT2CR (Figure 6A).
Additionally, the PAM orientation reaches across the transmembrane bundle from a lipophilic
pocket between TM2 and TM3 towards polar residues located in TM6 and TM7 (Figure 6B).
Interestingly, the lipophilic and polar surfaces of the docking site correspond to the LT and PH
moieties that have been featured in all active 5-HT2CR PAMs throughout our discovery and
optimization efforts.
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. Docking site surface mapping illustrates the distinct location of the proposed
allosteric site in relation to 5-HT bound to the 5-HT2CR. (A) The side profile view of 5-HT2C
R
with the protein surface illustrated around the allosteric docking site in the transmembrane helical
bundle. 5-HT (magenta) can be seen below the surface illustration. (B) Docking pose of 12 (green)
is shown bridging the helical bundle at a site distinct from 5-HT. (5-HT2CR PDB: 6BQG)
The depth of the lipophilic pocket between TM2 and TM3 drew our attention and likely
provides a volume for anchoring the PAM LT. A comparison of docking poses of the 4-phenyl
substituted 12 and the 4-cyclohexyl substituted 13 (Figure 7A) reveals an aromatic, π-π stacking
ACS Paragon Plus Environment

Page 17 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
interaction between 12 and TRP130 located in TM3 (Figure 7B), which is not present with the
non-aromatic 13 and may prove to be an important feature of the lipophilic pocket. Of note, when
viewing compounds 35 and 37 (Figure 5), which contain a less polar morpholine PH and differ
by an ethylcyclohexyl or phenethyl respectively, the presence of an aromatic interaction between
TRP130 and 37 may result in the phenethyl preferentially occupying the lipophilic pocket.
Contrarily, 35 is unable to form an aromatic interaction with TRP130 in the lipophilic pocket and
results in a flipped pose in which the morpholine occupies the lipophilic pocket. Consistent with
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
23
our previous docking results for 3, compound 12 forms contacts with SER334 in TM6 and a
residue in the ECL2 (Figure 7B). Reducing the length and rotatable bond number in the LT of 12
may have allowed for an optimized binding orientation with the number of molecular contacts
increasing accordingly compared to 3. Figure 7B shows the interaction diagram for 12 and
includes an aromatic interaction with TRP130 (TM3), an H-bond interaction between the ionizable
N atom of the piperidine ring and CYS207 (ECL2), H-bond interactions between the amide NH,
1
1
,2-diol hydroxy group and ASN351 (TM7), and an H-bond interaction between the remaining
,2-diol hydroxy group and SER334 (TM6). The 4-cyclohexyl substituted 13 (Figure 7C) lacks
the π-π stacking ability of 12 while retaining activity; thus, suggesting the aromatic interaction is
not a major contributor to functional activity. Additionally, the piperidine core configuration of 12
(
2R,4S) as docked is notably different than that of 13 (2S,4R), and, among 5-HT2CR PAMs that
returned rational poses, there was no consensus for the preferred conformation. Thus, these
findings suggest that the piperidine core may not be critical for binding or functional activity once
the PH and LT are in favorable positions.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 56
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 7. Docking demonstrates aromatic or aliphatic compact LTs orient towards the
hydrophobic pocket between TM2 and TM3 of the 5-HT2CR (5-HT2CR PDB: 6BQG). (A)
This view features a cutaway of the surface illustration to show how the phenyl and cyclohexyl
LTs of 12 (green) and 13 (blue) occupy a hydrophobic pocket within the transmembrane helical
bundle. (B) Predicted interactions formed by docking 12 including a possible aromatic interaction
with TRP130 in the hydrophobic pocket. (C) Predicted interactions formed by docking 13.
Following the prediction of a conserved allosteric docking site via the Schrodinger Drug
Discovery Suite, an additional in silico methodology was employed to predict possible allosteric
binding sites on the 5-HT2CR without a priori PAM structural information. AlloSitePro is a
computational tool featuring a support vector machine (SVM) decision engine based on
topological and physicochemical properties of the binding pocket.30 Thus, this methodology
provides a ligand-independent allosteric site prediction, whereas by definition, docking provides a
ligand-dependent allosteric site prediction. The web-based AlloSitePro was initiated using the
ergotamine-5-HT2CR complex X-ray crystal structure (PDB: 6BQG). The coordinates for
predicted allosteric binding site volumes were translated into magenta-colored sphere
representation and superimposed on the molecular docking model (Figure 8A-B). When viewed
simultaneously, the allosteric docking site surface map and the allosteric site prediction
coordinates are highly convergent, providing further support for the described, putative 5-HT2C
R
allosteric site (Figure 8A).
ACS Paragon Plus Environment

Page 19 of 56
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 8. Comparison of the allosteric binding site surface map and an independent allosteric
binding site prediction program for the 5-HT2CR model (5-HT2CR PDB: 6BQG). (A)
Composite model displaying docking site protein surface and predicted allosteric site volume
(
magenta balls) seen via cutaway to visualize the overlap between the models. (B) The volume
(magenta) determined to be a probable binding pocket capable of accommodating an allosteric
modulator. (C) From the same angle as panel B, the docking site protein surface is displayed.
In Vitro Assessment of Off-target Effects of Compound 12. To evaluate off-target interactions
for a panel of biologically relevant GPCRs and monoamine transporters, 12 was submitted to the
National Institute of Mental Health (NIMH) Psychoactive Drug Screening Program (PDSP). At
the assessed PAM concentration of 10 μM, none of the receptor targets returned inhibition greater
than 50%, a level considered above assay noise and indicative of test ligand interaction (Table 4).
This result improves upon our previously reported compound 3, (10 μM) which exhibited marked
inhibition of radioligand binding interactions at the dopamine transporter (DAT; 66.3%),
dopamine D3 receptor (71.7%), and α2A and α2B receptors (85.1% and 80.7%, respectively).23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 56
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
Interestingly, 12 maintains the same 1,2-propanediol moiety in the PH position as 3, differing only
the LT position by replacing the undecyl moiety with a phenyl substituent. The undecyl to phenyl
change results in a more compact, aromatic ligand, which could be expected to increase
promiscuity among monoamine GPCRs. However, it is likely that the removal of the undecyl
moiety reduces nonspecific lipophilic interactions, while the replacement with a phenyl group
allows further site-specific interactions to occur, as suggested by molecular modeling. Thus, the
selectivity profile of 12 against numerous GPCR orthosteric sites and the comparative analysis vs.
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
provides evidence that the functional allosteric site occupied by 12 is not an evolutionarily
residual binding pocket from another orthosteric site. From a development perspective the
selectivity profile strongly supports the advancement of 12 towards in vitro pharmacokinetic (PK)
evaluation and in vivo rodent behavioral characterization.
Table 4. Displacement of radioligand binding by compound 12 (10 μM) in a broad panel of
receptors and transporters
%
inhibition
Receptor / transporter
5-HT1A
Radioligand
(10 μM)a
28.4
12.3
15.3
6.3
K
i
(μM)b
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
3
[ H]-8-OH-DPAT
3
5
5
5
5
-HT1B
-HT1D
-HT1E
-HT2A
[ H]-GR125743
3
[ H]-GR125743
3
[ H]-5-HT
3
[ H]-Ketanserin
-4.3
-7.2
26.9
22.3
-4.3
2.5
2.2
9.7
7.5
3
5-HT2B
[ H]-LSD
3
5
5
5
-HT2C
[ H]-Mesulergine
3
-HT
3
[ H]-LY278584
3
-HT5A
[ H]-LSD
3
5-HT
-HT
6
[ H]-LSD
3
5
7
[ H]-LSD
[ H]-SCH23390
3
D
1
D
2
D
3
D
4
D
5
3
[ H]-N-Methylspiperone
[ H]-N-Methylspiperone
3
19.4
-9.3
9.4
3
[ H]-N-Methylspiperone
3
[ H]-SCH23390
3
DAT
[ H]-WIN35428
40.8
ACS Paragon Plus Environment

Page 21 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
3
SERT
NET
[ H]-Citalopram
47.5
74.4
-16.2
-11.0
-4.8
NT
2.5
NT
NT
NT
NT
NT
3
[ H]-Nisoxetine
3
α
α
α
α
α
1A
1B
1D
2A
2B
[ H]-Prazosin
[ H]-Prazosin
3
3
[ H]-Prazosin
[ H]-Rauwolscine
[ H]-Rauwolscine
3
-7.6
14.5
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
3
3
α
2C
[ H]-Rauwolscine
-1.8
NT
1
25
β
β
β
1
2
3
[ I]-Pindolol
-7.7
-15.9
-11.2
-
30.6
-6.1
13.4
23.3
7.0
NT
NT
NT
NT
NT
NT
NT
NT
NT
3
[ H]-CGP12177
3
[ H]-CGP12177
CB
CB
1
2
-
3
[ H]-CP55940
[ H]-DADLE
[ H]-U69593
3
δOR
κOR
μOR
GABA
3
3
[ H]-DAMGO
3
A
[ H]-Muscimol
a
The percent (%) inhibition at multiple targets was tested at 10 μM of compound 12. Inhibition
results > 50% are considered to be a reliable indication that the test compound displaced the
radioligand at the target binding site. For inhibition results > 50%, a K
non-linear regression analysis of radioligand competition isotherms. K
from best fit IC50 values using the Cheng-Prusoff equation. NT = not tested, The average K
repeated experiments was determined for NET.
b
i
value was obtained via a
values were calculated
i
i
from
In Vitro and In Vivo Pharmacokinetic (PK) Profiling and In Silico Toxicological Profiling.
The PK profile of 12 was assessed to determine whether this compound maintained favorable drug-
like properties for utility as an in vivo probe. Additionally, in silico toxicological prediction scoring
was used as a cautionary guide before in vivo rat behavioral assessments were made. Initially, 12
was subject to a battery of in vitro PK assessments, which returned an excellent profile (Table 5).
In rat liver microsomal fractions spiked with NADPH, clearance (CLint) was < 9.6 µL/min mg and
the half-life (t1/2) was > 240 min. Rat plasma protein binding studies displayed a free (unbound)
fraction of 66% for 12. Compound 12 was highly soluble in physiologically relevant conditions
and did not display meaningful inhibition of any cytochrome P450 enzyme tested. In an efflux
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 56
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
substrate assay utilizing a semi-permeable membrane of Madin-Darby Canine Kidney (MDCK)
cells expressing the P-glycoprotein (P-gp) efflux protein and the multidrug resistance mutation 1
(MDR1) gene (MDCK-MDR1), compound 12 exhibited a tolerable efflux ratio of 30.9. In silico
toxicity profiling was performed using the ProTox-II web-based prediction tool and 12 was
predicted to belong to a low toxicity class with a high degree of safety predicted for the doses
utilized for in vivo assessments (below).31 We also calculated the CNS multiparameter
optimization (MPO) value for compound 12 to be 4.7. CNS MPO values range from 0 to 6 with
the higher MPO scores predicted to be more desirable for CNS-penetrant medication candidates
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
32
(
Table S2). Lastly, the in vivo PK profile of 12 was obtained in male Sprague-Dawley rats (Table
6) and was observed as overall favorable and in line with our in vitro observations (Table 5).
Specifically, 10 mg/kg of 12 administered iv (t1/2 = 1.25 ± 0.02 h) or 20 mg/kg administered po
(t1/2=4.4 ± 0.5 h) resulted in appropriate blood plasma concentrations (Table 6). The brain
concentrations of 12 administered at 10 mg/kg iv were 96.4 ± 7.2 ng/g at 15 min and 94.0 ± 1.3
ng/g) at 1 hr. The brain-to-plasma ratios at 10 mg/kg iv were 0.07 ± 0.0012 (15 min) and 0.11 ±
0
.0056 (1 hr). Brain-to-plasma concentration ratios >0.04 are consistent with CNS penetration.33
These cumulative findings suggested that compound 12 would have adequate target exposure for
in vivo rat behavioral assessments.
Table 5. In Vitro Pharmacokinetic Profile and In Silico Toxicity Prediction for Compound
a
1
2
Compound 12
In Vitro Pharmacokinetics
In Silico Toxicity (ProTox-II)
Predicted LD50 3990 mg/kg
Rat Liver Microsomal
Clearance (NADPH)
CLint = <9.6 µL/min mg
1/2 = >240 min
t
Rat Plasma Protein Binding
CYP3A4 Inhibition
Fuplasma = 66%
Toxicity Class 1-6 (1 = high, 6 = low)
Prediction Accuracy
5
0.0% (10 µM)
69.26%
ACS Paragon Plus Environment

Page 23 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
CYP2C9 Inhibition
CYP2D6 Inhibition
CYP2C19 Inhibition
0.0% (10 µM)
Organ Toxicity (probability)
Carcinogenicity (probability)
Immunotoxicity (probability)
Mutagenicity (probability)
Cytotoxicity (probability)
Inactive (0.83)
Inactive (0.71)
Inactive (0.99)
Inactive (0.66)
Inactive (0.76)
1.9% (10 µM)
0.0% (10 µM)
11.8% (10 µM)
CYP1A2 Inhibition
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
Kinetic Solubility (PBS; pH
>
100 µM (2 hr)
7
.4)
A→B
app
= 0.04 • 10-6
P
Aryl hydrocarbon Receptor
(probability)
Inactive (0.96)
Inactive (0.90)
Inactive (0.95)
cm/s
PappB→A = 1.3 • 10-6
MDCK-MDR1
Permeability
Estrogen Receptor Alpha (probability)
cm/s
Efflux ratio = 30.9
Androgen Receptor (probability)
aIn vitro PK included assessment of 10 µM of compound 12 in liver microsomal clearance performed under NADPH-
dependent conditions and cytochrome P450 enzymatic inhibition assays performed and represented as percent
inhibition. The P-gp efflux substrate experiment employed Madin-Darby Canine Kidney (MDCK) cells expressing
multidrug resistance mutation 1 (MDR1) gene. In silico toxicity prediction profile shown along with statistical
probabilities (more information at http://tox.charite.de/protox_II/).31
Table 6. In Vivo Pharmacokinetic Profile and Brain Penetrability of Compound 12a
Pharmacokinetic profile of 12
Dose
(mg/kg)
CL
AUC0−inf
T1/2 (h)
T
max (h)
V
ss (L/kg)
C
max (ng/mL)
F (%)
(L h−1 kg )
−1
(ng·h·mL )
2408.6 ± 216.8 3065.6 ± 184.7
−1
1
0, iv
1.25 ± 0.02
4.4 ± 0.5
b
3.28 ± 0.2
5.4 ± 0.3
b
2
0, po
1.3 ± 0.34
10.89 ± 3.5
70.6 ± 23.7
411.3± 105.7
2244.7 ± 667.8 36.3 ± 11
Brain penetration analysis of 12
Plasma conc.
ng/mL)
Dose
mg/kg)
Brain conc.
Brain:Plasma
Time (h)
(
(
(ng/g)
Ratio
10, iv
0.25
1.0
96.4 ± 7.2
1486.5 ± 89.6
892.1 ± 39.1
0.07 ± 0.001
1
0, iv
94.0 ± 1.3
0.11 ± 0.005
aValues are the average of results (± SEM) from male Sprague-Dawley rats (n = 3/treatment group). Vehicle, 10%
DMSO:90% 2-hydroxypropyl-β-cyclodextrin (HP-β-CD). T1/2, half-life; Tmax, time of maximum concentration; CL,
plasma clearance; Vss, volume of distribution; Cmax, maximum concentration; AUC0−inf, area under the plasma
concentration−time curve; F, oral bioavailability; Time, hours after dose for brain collection; Brain conc., averaged
b
concentration of 12 in tissue sample. Not determined.
Analysis of Compound 12 in a Drug Discrimination Assay. The drug discrimination assay is a
powerful in vivo tool for establishing the interoceptive (subjective) effects of novel compounds as
well as for establishing their mechanisms of action in humans and animals.23, 34-43 This assay has
been employed to reliably characterize the neurobiological profiles of 5-HT2CR agonists (i.e., Ro
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 56
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
6
0-0175, WAY163909) as well as GPCR allosteric modulators.23, 35, 38-42 In the present series of
studies, we trained rats (n=13) to discriminate the selective 5-HT2CR agonist WAY163909 [0.75
mg/kg, intraperitoneal (ip); 15 min pretreatment] from an equivalent volume (1 ml/kg) of saline
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
(0.9% NaCl) in a two-lever drug discrimination protocol under a fixed ratio 20 (FR 20) schedule
2
3
of water reinforcement. In substitution tests, animals were tested for lever selection after the
administration of various doses of the training drug WAY163909 or compound 12. In combination
tests, animals were administered 12 (0.5-2 mg/kg) with a dose of WAY163909 (0.5 mg/kg) that
produced ~50% WAY163909•appropriate responding when given alone. We also assessed the
ability of the selective 5-HT2CR antagonist SB242084 to block the substitution of compound 12
plus a low dose of WAY163909.44
All rats acquired the WAY163909 vs. saline discrimination to criterion. The mean number
of sessions required to meet the acquisition criterion (defined as 10 consecutive sessions with ≥
8
0% of stimulus-appropriate responding) for the WAY163909 vs. saline discrimination was 57
training sessions (range: 39-68). Substitution of a test drug for the training stimulus is typically a
quantal rather than continuous function of stimulus similarity in the two-choice drug
discrimination assay.23, 34-42 Thus, intermediate doses of WAY163909 evoke responding confined
to one lever. Full substitution was defined as ≥ 80% of rats selecting the WAY163909-appropriate
lever. Partial substitution was defined as ≥ 40% and < 80% drug-appropriate responding. As can
be seen in Figure 9A, 0% of rats (0/13 rats) chose the WAY163909-associated lever when
administered saline or 0.125 mg/kg, while 92% (12/13 rats) and 100% (13/13) of rats chose the
WAY163909-associated lever at the training dose (0.75 mg/kg) and 1.0 mg/kg of WAY163909,
respectively (vs. saline, p < 0.05). Intermediate doses of WAY163909 (0.25-0.625 mg/kg) resulted
in a graded, quantal increase in the number of rats selecting the WAY163909-associated lever.
ACS Paragon Plus Environment

Page 25 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The response rate (responses/min) observed following saline administration was significantly
higher than following injection of 0.75 mg/kg (t12 = 2.22, p < 0.05) and 1.0 mg/kg of WAY163909
(t12 = 3.74, p <0.05). The response rates analyzed for remaining doses of WAY163909 tested
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
(0.125, 0.25, 0.50, 0.625, 1.0 mg/kg, ip) did not differ relative to 0.75 mg/kg of WAY163909 (n.s.).
Here, 0.5 and 0.625 mg/kg doses of WAY163909 resulted in 46% and 54% of rats selecting the
WAY163909-appropriate lever. Log-probit analyses indicate that the dose of WAY163909
predicted to result in 50% drug-associated responses (ED50) is 0.51 mg/kg, in agreement with our
previous study (0.53 mg/kg).23
Compound 12 did not exhibit intrinsic activity as a 5-HT2CR agonist in vitro (Figure 4A).
This result shaped the hypothesis that 12 would fail to substitute in the WAY163909 vs. saline
discrimination. As expected from in vitro analyses, 12 (0.5 - 5 mg/kg), administered ip 30 min
prior to testing, evoked saline-lever responding with no change in response rates (n.s.; Figure S2).
The observed lack of substitution supports the premise that 12 and the full 5-HT2CR agonist
WAY163909 have dissociable discriminative stimulus effects.
We next tested the hypothesis that 12 would enhance the discriminative stimulus effects of
a subthreshold dose of WAY163909 (0.5 mg/kg). Thus, rats were pretreated with 12 (0, 0.5, 1 or
2
mg/kg) 15 min before WAY163909 (0.5 mg/kg), followed 15 min later by placement in the
chambers. Figure 9B demonstrates that ~53% of rats (7/13 rats) selected the drug-associated lever
following WAY163909 (0.5 mg/kg). Pretreatment with 1 or 2 mg/kg of 12 plus WAY163909 (0.5
mg/kg) evoked ~77% (10/13 rats) and ~92% (12/13 rats) of rats selecting the WAY163909 lever,
respectively, suggesting a synergistic substitution for the training drug (Figure 9B). The selective
5-HT2CR antagonist SB242084 completely reversed the substitution of 12 (2 mg/kg) plus
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 56
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
WAY163909 (0.5 mg/kg) in 13/13 rats. In summary, 12 did not evoke intrinsic 5-HT2CR agonist
actions, but potentiated a behavioral marker associated with 5-HT2CR agonist-mediated signaling
in vivo.
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 9. Compound 12 dose-dependently augments the stimulus effects of the selective 5-
HT2CR agonist WAY163909. (A) The dose-response relationship for WAY163909 is shown (n =
1
3 rats). Filled magenta circles denote the rats that chose either the saline- or WAY163909-
associated lever for each condition/dose. Closed black circles denote the mean percentage of rats
selecting the WAY163909-associated lever (Y-axis) [*p<0.05 vs. saline (SAL)]. (B) Compound
(Cmpd) 12 (0-2 mg/kg) or WAY163909 (0.5 mg/kg) was administered alone or combination as
illustrated, with or without pretreatment with SB242084 (0.5 mg/kg). Open bars illustrate the mean
percentage of rats selecting the WAY163909-associated lever during combination tests (Y-axis)
[
*p < 0.05 vs. WAY163909 (0.5 mg/kg) alone; ^p < 0.05 vs. combination of compound 12 plus
WAY163909 following pretreatment with SB242084 (0.5 mg/kg). The details of the statistical
analyses are found in the Methods section.
Conclusions
A series of structurally optimized analogues was rationally designed and synthesized based
on the scaffold of our previously reported 5-HT2CR PAM CYD-1-79 (3), bearing privileged diol
PH moieties. Specifically, the reported investigation focused on the replacement of the undecyl
LT moiety with hydrocarbon fragments (cyclic in nature) with a decreased length while
approximately maintaining a similar molecular volume to maintain activity and improve drug-
ACS Paragon Plus Environment

Page 27 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
likeness. Compound 12 (CTW0415) displayed significant PAM activity at the 5-HT2CR using a
2
+
fluorescence-based, 5-HT-evoked Ca
i
release assay and was shown to be inactive as an agonist
or allosteric modulator at the 5-HT2AR. An additional six compounds were characterized as 5-
HT2CR PAMs, enriching our computational molecular modelling studies by providing an
interesting array of chemical diversity. The molecular modelling and docking studies provide
compelling support for a unique, spatially distinct 5-HT2CR allosteric binding site that bridges a
hydrophobic pocket between the TM2 and TM3 helices to polar contacts on the TM6 and TM7
helices. One aspect to consider regarding the location of the suggested allosteric site is the evidence
of metastable binding sites within transmembrane helical bundles that can briefly accommodate
orthosteric ligands.45-47 The transient nature of these binding pockets may provide a site or a
portion of a binding site for specific allosteric modulators to exert negative or positive effects.
Future computational and pharmacological experiments are needed to address these ideas.
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
Compound 12 was ultimately chosen for progression towards in vivo behavioral studies
due to superior physicochemical properties, an appropriate MPO value, highly selective off-target
profile, and promising pharmacokinetics. Moreover, the structure of 12 lends itself to additional
medicinal chemistry efforts as the compound is well within the boundaries of Lipinski’s Rule of
Five. Compound 12 dose-dependently potentiated the stimulus effects of the 5-HT2CR agonist
WAY163909. Importantly, this effect was completely blocked by the 5-HT2CR antagonist
SB242084, supporting the critical involvement in the synergism between 12 and a selective 5-
HT2CR agonist in vivo. Taken together, 12 is an ideal lead compound for further optimization as a
5-HT2CR PAM and the continued development of this class of molecules may be aided by the
structural presentation of a putative 5-HT2CR allosteric binding site.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 56
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
EXPERIMENTAL SECTION
Chemistry
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
General. All commercially available starting materials and solvents were reagent grade and used
without further purification. Reactions were performed under a nitrogen atmosphere in dry
glassware with magnetic stirring. Preparative column chromatography was performed using silica
gel 60, particle size 0.063-0.200 mm (70-230 mesh, flash). Analytical TLC was carried out
employing silica gel 60 F254 plates (Merck, Darmstadt). Visualization of the developed
chromatograms was performed with detection by UV (254 nm). NMR spectra were recorded on a
1
13
1
13
Bruker-600 ( H, 600 MHz; C, 150 MHz) spectrometer or Bruker-300 ( H, 300 MHz; C, 75
1
13
MHz). H and C NMR spectra were recorded with TMS as an internal reference. Chemical shifts
were expressed in ppm, and J values were given in Hz. High-resolution mass spectra (HRMS)
were obtained from Thermo Fisher LTQ Orbitrap Elite mass spectrometer. Parameters include the
following: Nano ESI spray voltage was 1.8 kV; Capillary temperature was 275 ˚C and the
resolution was 60,000; Ionization was achieved by positive mode. Melting points were measured
on a Thermo Scientific Electrothermal Digital Melting Point Apparatus and uncorrected. The
purity of final compounds was determined by analytical HPLC using a Shimadzu HPLC system
(
model: CBM-20A LC-20AD SPD-20A UV/VIS). HPLC analysis conditions: Waters Bondapak
C18 (300 × 3.9 mm); flow rate 0.5 mL/min; UV detection at 270 and 254 nm; linear gradient from
0% acetonitrile in water to 100% acetonitrile in water in 20 min followed by 30 min of the last-
1
named solvent (0.1% TFA was added into both acetonitrile and water). All biologically evaluated
ACS Paragon Plus Environment

Page 29 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
13
compounds have been characterized with H NMR, C NMR, HRMS, and HPLC analyses for a
quality control to ensure a purity of > 95%.
General procedure for the synthesis of 11-17
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
To a solution of 9 or 10 (0.18 mmol) and amino alcohols (0.18 mmol) in 4 mL of DMF was added
HBTU (0.23 mmol) and DIPEA (0.45 mmol). The resulting mixture was stirred at room
temperature for 16 h. The DMF was removed under vacuum to give a brown oily residue, which
was partitioned between CH
layer was separated and washed with saturated aqueous NaHCO
anhydrous Na SO , the solvent was removed under vacuum to give an oily residue. This residue
was purified with a silica gel column (5% MeOH in CH Cl ) and afforded the corresponding Boc-
protected amide. The amide (0.13 mmol) was dissolved in CH Cl (1 mL), followed by the addition
2
Cl
2
(50 mL) and 10% citric aqueous solution (10 mL). The organic
3
(10 mL). After drying over
2
4
2
2
2
2
of TFA (250 μL). The resulting mixture was stirred at room temperature. After 2 h, TLC showed
that the starting material had disappeared. The solvent was removed under vacuum to give an oily
residue. The residue was partitioned between CH
solution (10 mL). The organic layer was dried over anhydrous Na
to give an oily residue. This residue was purified with a silica gel column (eluting with 10% MeOH
in CH Cl ), affording compound 11-17.
2S,4S)-N-((1S,2S)-1,3-Dihydroxy-1-phenylpropan-2-yl)-4-phenylpiperidine-2-carboxamide
(2S,4S)-11) and (2S,4R)-N-((1S,2S)-1,3-dihydroxy-1-phenylpropan-2-yl)-4-
2
Cl
2
(30 mL) and saturated NaHCO
3
aqueous
2
SO , filtered, and concentrated
4
2
2
(
(
phenylpiperidine-2-carboxamide ((2S,4R)-11).48 Compounds (2R,4S)-11 (36 mg, 40%) and
(2S,4R)-11 (38 mg, 43%) were prepared from 9 (2 steps), respectively. These two isomers could
1
be separated by preparative TLC as colorless amorphous gel. Compound (2R,4S)-11: H NMR
(600 MHz, CDCl
3
) δ 7.42 (m, 2H), 7.30 (m, 4H), 7.22 (m, 2H), 7.12 (d, 2H, J = 7.2 Hz), 5.06 (d,
ACS Paragon Plus Environment