Novel Aryloxyethyl Derivatives of 1-(1-Benzoylpiperidin-4-yl)methanamine as the Extracellular Regulated Kinases 1/2 (ERK1/2) Phosphorylation-Preferring Serotonin 5-HT1A Receptor-Biased Agonists with Robust Antidepressant-like Activity

Article abstract of DOI:10.1021/acs.jmedchem.9b00062

Novel 1-(1-benzoylpiperidin-4-yl)methanamine derivatives were designed as "biased agonists" of serotonin 5-HT_1A receptors. The compounds were tested in signal transduction assays (ERK1/2 phosphorylation, cAMP inhibition, Ca²⁺ mobiliz

Full text of DOI:10.1021/acs.jmedchem.9b00062

Subscriber access provided by TULANE UNIVERSITY
Article
Novel aryloxyethyl derivatives of 1-(1-benzoylpiperidin-4-
yl)methanamine as the Extracellular Regulated Kinases 1/2
ERK1/2) phosphorylation-preferring serotonin 5 HT1A receptor
biased agonists with robust antidepressant-like activity
(
Joanna Sniecikowska, Monika G#uch-Lutwin, Adam Bucki, Anna Wi#ckowska, Agata
Siwek, Magdalena Jastrz#bska-Wi#sek, Anna Partyka, Daria Wilczy#ska, Karolina
Pytka, Krzysztof Pociecha, Agnieszka Cios, El#bieta Wyska, Anna Weso#owska, Maciej
H. Paw#owski, Mark Varney, Adrian Newman-Tancredi, and Marcin Ko#aczkowski
J. Med. Chem., Just Accepted Manuscript • Publication Date (Web): 05 Feb 2019
Downloaded from http://pubs.acs.org on February 5, 2019
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 85
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 85
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
Novel Aryloxyethyl Derivatives of 1-(1-benzoylpiperidin-4-
yl)methanamine as the Extracellular Regulated Kinases 1/2 (ERK1/2)
Phosphorylation-preferring Serotonin 5-HT1A Receptor Biased
Agonists with Robust Antidepressant-like Activity
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
a
a
a
Joanna Sniecikowska , Monika Gluch-Lutwin , Adam Bucki , Anna Więckowska , Agata Siwek
a, Magdalena Jastrzebska-Wiesek , Anna Partyka , Daria Wilczyńska , Karolina Pytka ,
a
a
a
a
a
a
a
a
Krzysztof Pociecha , Agnieszka Cios , Elżbieta Wyska , Anna Wesołowska , Maciej Pawłowski
a
b
b
a*
, Mark Varney , Adrian Newman-Tancredi , Marcin Kolaczkowski
a
Faculty of Pharmacy, Jagiellonian University Medical College, 9 Medyczna St., 30-688 Kraków,
Poland;
b
Neurolixis Inc. 34145 Pacific Coast Highway #504, Dana Point, CA 92629 USA
ABSTRACT
Novel 1-(1-benzoylpiperidin-4-yl)methanamine derivatives were designed as ‘biased agonists’ of
serotonin 5-HT1A receptors. Compounds were tested in signal transduction assays (ERK1/2
2
+
phosphorylation, cAMP inhibition, Ca mobilization and β-arrestin recruitment) which identified
ERK1/2 phosphorylation-preferring aryloxyethyl derivatives. The novel series showed high
5-HT1A receptor affinity, >1000-fold selectivity versus noradrenergic α
1
, dopamine D
2
, 5-HT2A,
histamine H and muscarinic M receptors, and favorable drug-likeness properties (CNS-MPO,
1
1
Fsp3, LELP). The lead structure, (3-chloro-4-fluorophenyl)(4-fluoro-4-(((2-(pyridin-2-
yloxy)ethyl)amino)methyl)piperidin-1-yl)methanone (17, NLX-204), displayed high selectivity in
ACS Paragon Plus Environment

Page 3 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the SafetyScreen44™ panel (including hERG channel), high solubility, metabolic stability and
Caco-2 penetration, and did not block CYP3A4, CYP2D6 isoenzymes or P-glycoprotein.
Preliminary in vivo studies confirmed its promising pharmacokinetic profile. 17 also robustly
stimulated ERK1/2 phosphorylation in rat cortex and showed highly potent (MED=0.16 mg/kg)
and efficacious antidepressant-like activity, totally eliminating immobility in the rat Porsolt test.
These data suggest that the present 5-HT1A receptor biased agonists could constitute promising
antidepressant drug candidates.
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
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 85
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
INTRODUCTION
Despite the clinical availability of a large number of antidepressant drugs, the treatment of
depressive disorders still remains an area of unmet medical need. With few exceptions, current
1
antidepressants target inhibition of monoamine neurotransmitter reuptake into brain neurons. The
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
clinical efficacy of this strategy is less than optimal, as indicated by the substantial proportion of
patients that fail to respond to treatment or that show only partial treatment response.2,3
Furthermore, current antidepressants usually require several weeks of administration before
clinical efficacy is observed – this protracted delay in therapeutic onset is a source of distress to
families, of increased medical burden and of potential suicide risk for patients.4,5 Nevertheless,
recent findings suggest that the development of rapidly-acting and highly clinically-efficacious
novel antidepressants is a realistic prospect. Indeed, the discovery that ketamine, a general
anesthetic drug acting primarily as a non-competitive N-methyl-D-aspartate (NMDA) receptor
antagonist, can reverse depressive symptoms in profoundly-depressed and treatment-resistant
patients, has given new hope that depressive states can be treated efficaciously and rapidly.
Accordingly, the S-enantiomer of ketamine is currently being evaluated by the FDA as a rapid-
acting antidepressant for treatment-resistant depression.6–8 At a molecular level, ketamine was
found to mediate its antidepressant effects in rats via several mechanisms including AMPA and
mTOR activation and, of particular interest in the present context, Extracellular Regulated Kinase
(ERK) phosphorylation in prefrontal cortex. The latter mechanism has attracted considerable
attention as a substrate for ketamine’s antidepressant activity. Indeed, ketamine elicits robust
phosphorylation of ERK in this brain region 9–11 and, correspondingly, inhibition of ERK
phosphorylation using inhibitors of mitogen-activated proteins kinase (MEK) blocked the
antidepressant-like effects of ketamine in rat models, an observation that was also seen with
1
2
another NMDA receptor antagonist, Ro 25-6981. These results strongly suggest that mechanisms
3
ACS Paragon Plus Environment

Page 5 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
that elicit robust activation of cortical ERK phosphorylation could constitute an effective and novel
strategy to discover improved antidepressant drugs which, like ketamine, could (i) show efficacy
in profoundly-depressed and treatment-resistant patients, (ii) be effective following short durations
of treatment (or even acutely, like ketamine), and (iii) target new neurobiological mechanisms
distinct from classical monoamine reuptake inhibitors. Nevertheless, ketamine presents several
notable limitations that prevent its widespread use as an antidepressant, including the induction of
dissociative states which may be difficult to separate from the antidepressant effects,13,14
cardiovascular and urinary tract side-effects,15–17 as well as neurotoxicity and neurocognitive
dysfunction. These considerations highlight the ongoing medical need for novel and well-tolerated
antidepressant drugs. In the search for a suitable target to achieve such a profile, it is interesting
that ketamine’s mechanism of action involves serotonergic pathways and, notably, activation of
the 5-HT1A receptor subtype, at least in certain animal tests.18–20 Presynaptic 5-HT1A autoreceptors
have been implicated in the delay of therapeutic onset of antidepressants but post-synaptic 5-HT1A
receptors, notably in cortical brain regions, are known to mediate potent antidepressant-like effects
in animal models.21–25 In addition, a moderately-selective 5-HT1A receptor agonist, flesinoxan, was
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
6
shown to have promising antidepressant properties, whereas partial agonists at 5-HT1A receptors
e.g. buspirone) have failed to show robust antidepressant effects. It therefore appears that full-
(
agonist targeting of 5-HT1A receptors may be a promising strategy for identification of novel
antidepressants, potentially mimicking the remarkable antidepressant effects of ketamine.
However, targeting of 5-HT1A receptors may require preferential activation in specific brain
regions, notably frontal cortex. Although there is only a single gene sequence for 5-HT1A receptors,
which does not contain introns or splice variants, the effects of 5-HT1A receptor activation in
different brain regions can be quite diverse, and even divergent, due to coupling of the receptors
4
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 85
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 distinct signal transduction responses. Thus, although the amino acid sequence of the 5-HT1A
receptor itself is unchanged, the latter’s interaction with intracellular signaling proteins varies
depending on brain region. Thus, if agonists preferentially elicit coupling of the receptors to
particular G-proteins or second messengers, this would be anticipated to results in preferential
activation of distinct brain regions.
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 recent discovery and characterization of 5-HT1A receptor biased agonists offer a way forward
in this respect. F15599 (1), also known as NLX-101, is a highly-selective 5-HT1A receptor agonist
with a unique profile, preferentially activating ERK phosphorylation versus other signaling
27
28
pathways in vitro, as well as ex vivo in frontal cortex. Notably, 1 elicited strong antidepressant-
like activity in the rat forced swim test, completely abolishing immobility, even upon acute
administration – as is the case with ketamine, suggesting that the preferential activation of cortical
ERK phosphorylation via cortical 5-HT1A receptors translates to enhanced antidepressant
potential.29–31
However, 1 has so far been the only compound displaying such a promising ‘biased agonist’ profile
of activity, and some close chemical analogues (notably F13714 or befiradol) display quite
different signal transduction properties in vitro and in vivo.31–35 In addition, the latter compounds
do not exhibit preferential activation of cortical 5-HT1A receptors which, as discussed above, is
important for improved antidepressant activity.29,36,37 The structure-activity relationships of these
compounds with respect to preferential ERK phosphorylation have not been established and it
therefore remains to be determined whether increased ‘bias’ for ERK phosphorylation can elicit
even superior antidepressant profiles.
5
ACS Paragon Plus Environment

Page 7 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Based on these considerations, the present work used 1 as a model structure to explore the synthesis
and characterization of a series of novel 5-HT1A receptor-targeted antidepressants. In order to
establish their comprehensive functional selectivity profiles, the whole novel series was tested in
2
+
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
four signal transduction assays (ERK1/2 phosphorylation, cAMP inhibition, Ca mobilization and
β-arrestin recruitment), for the first time in the case of the 5-HT1A receptor agonists. Specifically,
we sought to identify the structural motifs that confer a ‘biased’ ERK-preferential activity of the
compounds as well as to determine whether they possess favorable drug-like properties that are
compatible with successful pharmaceutical development. Finally, to confirm the therapeutic
potential of the series, we tested a lead compound in a recognized in vivo model of antidepressant-
like activity.
6
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 85
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
Design, synthesis and biological evaluation of 1-(1-benzoylpiperidin-4-yl)methanamine
derivatives
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 a framework for the design of novel compounds we used docking-derived ligand-receptor
3
8
complex of 1 in the homology model of the 5-HT1A receptor (Figure 1). The analysis of the
binding mode revealed crucial interactions, commonly found for a vast majority of monoaminergic
ligands, namely (i) charge-reinforced hydrogen bond between the protonated central nitrogen atom
and the carboxyl group of the aspartic acid Asp3.32 as well as (ii) CH–π stacking between the
pyrimidine ring and phenylalanine Phe6.52. The additional interactions included (iii) a hydrogen
bond between the oxygen atom of the amide moiety and the NH group from the main chain of the
isoleucine Ile189 in the second extracellular loop, (iv) π–π stacking between the benzoyl moiety
and phenylalanine Phe3.28 as well as (v) cation–π interaction between the protonated central
nitrogen atom and phenylalanine Phe6.51.
Phe6.51
Tyr7.43
Trp6.48
Tyr2.64
ECL2
Phe6.52
Asp3.32
Phe3.28
Ile189
Ser5.42
Figure 1. The predicted binding mode of compound 1 in the site of serotonin 5-HT1A receptor. Amino acid residues
engaged in ligand binding (within 4Å from the ligand atoms) are displayed as sticks, whereas crucial residues, e.g.
forming H-bonds (dotted yellow lines), π-π/CH-π stacking (dotted cyan lines) and cation-π interactions (dotted green
7
ACS Paragon Plus Environment

Page 9 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
line) are represented as thick sticks. ECL2 residues were hidden for clarity; ECL – extracellular loop. The homology
model of the 5-HT1A receptor prepared based on crystal structure of the 5-HT1B receptor (PDB ID: 4IAR).
Although the observed interactions seem to be optimal for binding to the 5-HT1A receptor, we have
noticed that the pyrimidine ring does not fully penetrate the hydrophobic binding pocket of the
receptor. At this point it is worth noting that this region of the binding site of GPCRs
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
39
accommodates moieties that are crucial for eliciting particular functional activity. Thus, we
hypothesized that diversification of the fragment of the molecule that binds within the pocket may
yield compounds with various functional profiles. In our attempt to develop novel ligands that
would more effectively fill this hydrophobic binding pocket, we initially focused on extension of
the one-unit methylene linker to three-unit linker between the central amine function and
pyrimidine ring of 1. We introduced a propylene linker and a number of analogues with one of the
methylene units replaced with heteroatoms such as nitrogen, sulfur and oxygen (Figure 2).
Figure 2. General structure of the new series of selective 5-HT1A receptor agonists.
We have validated our hypothesis in the docking studies where we compared the binding modes
of 1 and newly designed ligands, exemplified by compounds 16 and 17 (Figure 3). As expected,
the extension of the linker shifted the aromatic ring towards Phe6.52, markedly shortening the CH-
8
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 85
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
π bond (3.27 Å for 1 vs. 2.85 Å and 2.96 Å for compounds 16 and 17, respectively) and optimizing
the geometry of the interactions. The derivatives with three-unit linkers more effectively filled the
hydrophobic binding pocket near the transmembrane helix 6 (TMH6), which accounted for better
fitting to the receptor and promoted the reduction of binding energy. Encouraged by the results of
docking studies, we moved to the synthesis and biological evaluation of the compounds.
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
Tyr7.43
Trp6.48
Tyr2.64
Phe6.51
Phe6.52
Asp3.32
Phe3.28
ECL2
Ile189
Figure 3. The predicted binding mode of compound 1 (green) together with 16 (pink) and 17 (white) in the site of the
serotonin 5-HT1A receptor. Amino acid residues engaged in ligand binding (within 4Å from the ligand atoms) are
displayed as sticks, whereas crucial residues, e.g. forming H-bonds (dotted yellow lines), π-π/CH-π stacking (dotted
cyan lines) and cation-π interactions (dotted green line) are represented as thick sticks. ECL2 residues were hidden
for clarity; ECL – extracellular loop. The homology model of the 5-HT1A receptor prepared based on crystal structure
of the 5-HT1B receptor (PDB ID: 4IAR).
9
ACS Paragon Plus Environment

Page 11 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The synthetic pathway to the target compounds was based on the one previously developed and
reported, 40,41 as presented in Scheme 1.
Scheme 1. Synthesis of 1-(1-benzoyl-4-fluoropiperidin-4-yl)methanamine derivativesa
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Reagents and conditions: (a) Et
TBABr, DCM, r.t., 3 h, yield: 80–87%; (c) HF/pyridine, DCM, -10 °C, 30 min then r.t. 24–36 h, yield: 54–68%; (d)
DABCO, NaCNBH , FeSO × 7H O, molecular sieves, MeOH, r.t., 36–72 h, yield: 23–66%.
3
N, DCM, 0 °C then r.t., 12 h, yield: 47–60%; (b) chloroacetonitrile, 30% NaOH,
3
4
2
In the first step, the commercially available 4-piperidone was acylated by the appropriate benzoyl
chloride derivatives 2 or 3 in the presence of triethylamine. Then a Darzens reaction between the
obtained benzoylpiperidin-4-one derivatives
tetrabutylammonium bromide as a phase transfer catalyst, gave the corresponding cyanoepoxides
or 7. The epoxides 6 or 7 underwent a regioselective ring opening reaction with poly(hydrogen
4
or
5
and chloroacetonitrile, using
42
6
fluoride)pyridine43,44 to obtain the corresponding cyanohydrins 8 or 9. The final compounds 9–27
were prepared via reductive amination using cyanohydrins 8 or 9 and the appropriate amines in
the presence of a base and sodium cyanoborohydride as reducing agent and with addition of iron
sulfate heptahydrate.41
10
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 85
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
The affinity of the compounds for the 5-HT1A receptor was determined in radioligand binding
assay (Table 1). All but one (compound 11) displayed affinities higher than that of 1, with pK
values ranging from 8.73 to 10.83 vs 8.66 for 1. The affinity was most significantly increased for
the compounds with oxyethylene linker (16–19), which showed subnanomolar affinities (pK
0.19–10.83). These results unequivocally confirmed our hypothesis derived from molecular
i
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
i
=
1
modeling and allowed us to choose oxyethylene linker for further studies. The superiority of the
oxyethylene linker over other linkers tested may be due to the fact that oxygen atom is the best
bioisostere of the crucial nitrogen atom embedded in the aromatic moiety of 1 (electronegative, H-
bond accepting). At this stage we tested para position of the benzoyl fragment, by the replacement
of fluorine with chlorine atom (compounds 18 and 19). This modification increased binding
affinity (18 vs 16 and 19 vs 17), but comparison of LELP scores showed that it had resulted from
increase in lipophilicity and most probably non-specific interactions. Therefore, we used 3-chloro-
4-fluorobenzoil core in further studies.
11
ACS Paragon Plus Environment

Page 13 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 1. 5-HT1A receptor affinity and developability parameters of compounds 10–19 and the references.
R
a
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
receptor affinity (pK
i
)
CNS
MPO
5-HT1Ab,d
c,e
b,f
Fsp3g
LELPh
i
Compd
10
X
R
α
1
D
2
F
9.32 ± 0.19
< 6.00
< 6.00
< 6.00
0.41
9.1
4.06
5.11
1
1
2
F
F
8.59 ± 0.01
8.73 ± 0.06
6.75 ± 0.02
0.43
0.38
6.8
7.0
1
< 6.00
< 6.00
< 6.00
4.93
5.08
13
F
8.90 ± 0.15
7.26 ± 0.07
0.40
5.5
14
15
16
17
F
F
F
F
9.25 ± 0.21
9.21 ± 0.04
10.21 ± 0.20
10.19 ± 0.19
NTj
< 6.00
< 6.00
< 6.00
< 6.00
0.38
0.40
0.38
0.40
8.6
7.2
6.7
5.5
3.86
4.86
4.89
5.42
7.63 ± 0.18
6.29 ± 0.13
6.78 ± 0.05
1
8
Cl
Cl
10.83 ± 0.15
10.47 ± 0.20
NT
< 6.00
< 6.00
0.38
0.40
7.2
6.2
4.31
4.93
1
1
9
6.89 ± 0.00
8.66 ± 0.07
9.09 ± 0.03
8.30 ± 0.05
< 6.00
NT
< 6.00
NT
0.42
0.62
0.71
5.2
5.9
4.3
5.44
3.41
5.82
8-OH-DPAT
Buspirone
NT
NT
a
i i
All binding affinity values are represented as pK i.e. –logK and expressed as means ± SEM from at least 3 experiments
b
c
performed in duplicates; Radioligand binding was performed using CHO-K1 cells, rat cortex; The affinity values were
d
3
e
3
f 3
i
determined using [ H]8-OH-DPAT, [ H]-prazosin, phentolamine pKi = 7.95, [ H]-methylspiperone, haloperidol pK = 8.85;
g
3
h
i
fraction of sp carbon atoms; ligand efficiency-dependent lipophilicity score; Central Nervous System Multiparameter
j
Optimization score; NT – not tested.
Next, we explored the optimal position of a nitrogen atom in the aryloxy moiety (Table 2). We
observed a gradual decrease in binding affinity to a pK
i
of 8.82 for pyridin-3-yloxy derivative (20)
and an unacceptably low binding value (pK 6.41) for pyridin-4-yloxy derivative (21).
i
12
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 85
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 2. 5-HT1A receptor affinity and developability parameters of compound’s 17 analogs 20–27.
a
receptor affinity (pK
i
)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
CNS
MPO
5-HT1Ab,d
α
1
c,e
D
2
b,f
Fsp3g
LELPh
i
Compd
Ar
1
2
2
2
2
2
2
2
7
0
1
2
3
4
5
6
7
10.19 ± 0.19
6.78 ± 0.05
< 6.00
< 6.00
< 6.00
< 6.00
< 6.00
< 6.00
< 6.00
< 6.00
< 6.00
< 6.00
< 6.00
< 6.00
< 6.00
< 6.00
< 6.00
< 6.00
< 6.00
0.40
5.5
5.42
5.65
5.41
5.46
5.47
5.44
5.44
4.90
5.37
8.82 ± 0.06
6.41 ± 0.13
9.28 ± 0.10
7.45 ± 0.10
6.68 ± 0.16
6.70 ± 0.12
8.42 ± 0.20
7.86 ± 0.13
0.40
0.40
0.42
0.42
0.42
0.42
0.43
0.45
5.0
6.9
3.4
5.9
4.5
3.7
8.2
7.1
2
a
i i
All the binding affinity values were represented as pK i.e. –logK and expressed as means ± SEM from at least 3 experiments
b
c
performed in duplicates; Radioligand binding was performed using CHO-K1 cells, rat cortex; The affinity values were
determined using [ H]8-OH-DPAT, [ H]-prazosin, phentolamine pK
fraction of sp carbon atoms; ligand efficiency-dependent lipophilicity score; Central Nervous System Multiparameter
Optimization score;
d
3
e
3
f
3
g
i
= 7.95, [ H]-methylspiperone , haloperidol pK
i
= 8.85;
3
h
i
Subsequently, we explored heterocycles with two nitrogen atoms (22–25). Only 2-pyrazine
derivative, compound 22 with pK
i
9.28 kept the desired level of affinity (pK
i
>8). Interestingly, 22
can be considered a hybrid of 2- and 3-pyridine and its pK
i
value is intermediate between that of
17 (10.19) and that of 20 (8.82). We have also prepared compound 26 that is a direct analog of
befiradol and compound 27 – an analogue of 1 to invesitgate the effect of methyl substituent at the
5
position of the aryl ring. Compound 26 displayed significantly decreased affinity compared to
compound 17 (pK
i
8.42 vs. 10.19, respectively) while compound 27 showed a modest increase in
13
ACS Paragon Plus Environment

Page 15 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
affinity relative to its unsubstituted analog 23 (pK
i
7.86 vs. 7.45, respectively). These contrasting
results indicate that the substitution pattern at the aromatic ring will require further investigation.
Developability analysis – In silico-supported predictions
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
In order to provide an early assessment of the developability of the designed series, we analyzed
all the compounds using in silico-supported measures of drug-likeness. All the compounds fulfilled
Lipinski’s rule of five (Ro5) for good oral bioavailability and displayed exceptionally high Central
Nervous System Multiparameter Optimization (CNS MPO) values with median value of 5.10. 45,46
Another important parameter that is especially important in the context of compounds’ solubility
3
3
is the fraction of sp atoms (Fsp ). A majority of the compounds tested, mostly those with two
3
nitrogen atoms in the aryl ring, showed very high and favorable Fsp values of 0.42–0.45, while
all the others were characterized by acceptable values of 0.38–0.41. 47
Next, we used ligand efficiency dependent lipophilicity (LELP), a function that combines
physicochemical properties with compounds’ activity to predict druglikeness. The median LELP
value for all the compounds was 6.75, which is the desirable value for drug candidates.48,49 Six
compounds displayed excellent values ≤ 5.9, reaching an exceptionally favorable value for
compound 22 (LELP = 3.4) (Tables 1 and 2).
Finally, we have used SwissADME web tool 50–53 to identify Pan-Assay INterference compoundS
(PAINS) as well as toxicophores within the investigated group of compounds. The analysis
showed no alerts for any of the tested compounds.
14
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 85
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
Summing up, the vast majority of the designed compounds were characterized by favorable values
of physicochemical properties and multivariate scoring functions, as well as good results of safety
filters, which suggest a favorable developability potential of the designed series.
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
Developability analysis – In vitro assays
Next, to confirm the favorable developability profile predicted in silico, compound 17 was selected
as a lead structure, to be examined in a series of in vitro experiments. To this end, the
SafetyScreen44™ Panel (plus additional antitargets), the Bioavailability Panel, as well as
CYP3A4, CYP2D6 and P-glycoprotein (P-gp) inhibition assays, were performed at Eurofins
Pharma Discovery Services. SafetyScreen44™ Panel is one of the most commonly used set of
assays for in vitro pharmacological profiling of target selectivity. 54 Bioavailability Panel consists
of solubility, permeability, protein binding and metabolic stability assays, which are used for
prediction of ADME properties, a very important issue in the drug development process.
Data from the safety pharmacology profiling revealed excellent overall selectivity of the
investigated lead structure. Compound 17, tested at a concentration of 10 µM, showed substantial
binding inhibition (higher than 50%) for only 3 among 52 off-targets, namely: adrenergic α1A and
α2A and dopaminergic D2S receptors (91–99%) (Table 3). Compound 17 was subsequently tested
at a concentration of 1 µM for these antitargets and exhibited much lower binding inhibition at D
2
(9.5%) and α2A (29.9%) receptors but maintained a relatively high value for α1A receptor (87.6%).
The precise affinity for the latter was therefore determined and the pK was found to be 6.78, an
i
affinity which is more than 1000-fold lower than the affinity of compound 17 for the 5-HT1A
receptor (pK
i
= 10.19; see Table 1). In addition to testing the binding affinity of 17 for hERG
TM
channel in the SafetyScreen44 , we also tested it using an automated patch-clamp method. Patch
15
ACS Paragon Plus Environment

Page 17 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
clamp methods are considered to provide a more reliable assessment of hERG interaction, even at
an early stage of drug discovery process.55,56,57 This study was performed at 3 concentrations (10.0,
1.0 and 0.1 µM) showing only 0.1, 10.5 and 46.5% inhibition of tail current respectively. This
indicates that 17 has a low propensity for hERG channel blockade and suggests that the
investigated series has a low risk with respect to hERG-related cardiac safety.
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
In the Bioavailability Panel, compound 17 exhibited desirable druglike ADME attributes (Table
4
). It was characterized by high aqueous solubility in phosphate-buffered saline (PBS 7.4),
simulated intestinal fluid (SIF) and simulated gastric fluid (SGF) (192.2, 192.2 and 200 µM,
-6
respectively) as well as excellent permeability on Caco-2 cells (Papp > 10×10 cm/s). Moreover,
final efflux ratio value (ER = basolateral to apical [BA]/apical to basolateral [AB] permeability)
46
was lower than 2.5 (ER = 0.86), which means that compound 17 is not a P-gp substrate. Those
values are comparable with the reference drug propranolol, which belongs to the class 1 of
Biopharmaceutics Classification System (BCS) – high permeability/high solubility – suggesting
good oral absorption and reliable in vitro-in vivo correlation.58
16
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 85
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 3. SafetyScreen44™ Panel and additional antitargets results for compound 17.
% inhibition of Control
% inhibition of Control
Specific Binding
Specific Binding
a
a
Target
Assay
at 10 µM/1 µM
Target
Assay
at 10 µM/1 µM
GPCR
GPCR
ADENOSINE
ADRENERGIC
A
α
α
α
2A
11.8
99.7/87.6
NT/32.2
94.3/29.9
2.2
SEROTONIN
5-HT1B
-9.1
NT/25.5
9.0
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1A
5-HT1D
5-HT2A
5-HT2B
5-HT2C
1B
2A
39.7
NT/-26.2
β
1
2
β
3.8
0.4
5-HT
6
7
NT/12.1
NT/13.2
CANNABINOID
CB
1
2
5-HT
CB
-3.1*
TRANSPORTERS
DOPAMINE
CHOLECYSTOKININ
DOPAMINE
CCK
1
(CCK
A
)
-30.3
38.2
DAT
NET
20.4
-2.0
0.9
D
1
NOREPINEPHRINE
SEROTONIN
D
2S
91.3/9.5
SERT
D
3
NT/39.1
ION CHANNELS
ENDOTHELIN
HISTAMINE
ET
A
-13.8
-0.5
GABA CHANNEL
BZD (central)
-11.8
-1.8
H
H
H
1
GLUTAMATE CHANNEL
NICOTINIC CHANNEL
SEROTONIN CHANNEL
NMDA
2
-15.1
N neuronal α4β2
5-HT
Ca2+ channel (L)
-5.4
3
4
NT/-10.9
3
-12.7
H
NT/-10.9
12.9
Ca2+ CHANNEL
K+ CHANNELS
18.2
16.2
3.8
(dihydropyridine site)
3
MUSCARINIC
M
M
M
1
2
3
hERG ([ H]Dofetilide)
6.3
v
K channel
+
+
9.7
Na CHANNEL
Na channel (site 2)
48.8
M
4
NT/8.8
7.2
KINASES
OPIOID
& OPIOID LIKE
δ
2
(DOP)
Lck kinase
18.2
11.9
CTK
κ (KOP)
34.6
42.9
OTHER NON-KINASE ENZYMES
AA METABOLISM
µ (MOP)
COX
1
2
VASOPRESIN
V
1a
7.0
COX
-9.2
-2.9
-3.6
-3.9
-10.7
MONOAMINE &
NEUROTRANSMITTER
AChE
MAO-A
PDE3A
PDE4D2
NUCLEAR RECEPTORS
STEROID NUCLEAR
RECEPTORS
AR
GR
-20.5
-5.1
PHOSPHODIESTERASES
a
Assays, in which % inhibition of Control Specific Binding was over 50% in 10 µM, were repeated at 1 µM concentration. For
some targets only the test at 1 µM was performed – marked as NT/value. *Low to moderate negative values (< 50%) have
no real meaning and are attributable to variability of the signal around the control level;
17
ACS Paragon Plus Environment

Page 19 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Compound 17 also exhibited high microsomal stability, with intrinsic clearance (CLint) lower than
1
15 [μL/min/mg] and a half-life higher than 60 minutes. Noteworthy, it is the best possible result
in the given experimental setting, suggesting high metabolic stability and therefore a low risk of
significant first pass effect. Overall, the above-mentioned results suggest favorable oral
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
bioavailability of compound 17 and are in line with the favorable CNS MPO (5.42), Fsp (0.40)
and LELP (5.5) values predicted in silico for compound 17. Furthermore, compound 17 displayed
no P-gp inhibition (6.5% of inhibition at 10 µM) as well as negligible inhibition of CYP3A4 and
CYP2D6 (23.2% and 15.9% at 10 µM, respectively). Together with the moderate plasma protein
binding (93.2%), this suggests a relatively low drug–drug interaction potential. Taken together,
these data provide compelling evidence supporting desirable druglike properties of compound 17,
suggesting its good developability potential and promising similarly favorable properties for the
whole series.
18
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 85
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 4. In vitro Bioavailability Panel and CYP3A4, CYP2D6 and P-gp inhibition assays results for compound 17.
Assay
Compound 17 (10 μM*)
192.2 μM
Reference
PBSa
SGFb
180.8 μM (metoprolol)
Solubility 45
200 μM
–
–
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
SIFc
192.2 μM
A–Bd
-6
-6
25.90 [10 cm/s]
23.70 [10 cm/s] (propranolol)
Permeability 59
Metabolic
B–Ad
22.45 [10 cm/s]
-6
27.40 [10 cm/s] (propranolol)
-6
Intrinsic
Clearance (CLint
<
115 [μL/min/mg]
< 115 [μL/min/mg] (propranolol)
)
e 60
stability
Half-Life (T1/2
)
> 60 [min]
93.2%
6.5%
> 60 [min] (propranolol)
99% (warfarin)
Protein bindingf 61
P-gp inhibitiong 62
CYP3A4 inhibitionh 63
CYP2D6 inhibitionh 64
15 μM [IC50] (verapamil)
0.24 μM [IC50] (ketoconazole)
0.0081 μM [IC50] (quinidine)
23.2%
15.9%
*
Metabolic stability study was performed at the concentration of 0.1 μM; Solubility study was performed in a 200 μM
a
b
c
concentration; aqueous solubility at buffer pH7.4; aqueous solubility in simulated gastric fluid; aqueous solubility at
d
e
f
g
simulated intestinal fluid; Caco-2, pH 6.5/7.4; liver microsomes–human; plasma human; MDR1-MDCKII cell line;
h
Human recombinant
Selectivity vs. key antitargets
As the goal of the present study was to obtain selective 5-HT1A receptor agonists, it was essential
to verify the selectivity of the novel ligands. To this end, the affinity for the most important
antitargets was determined for all of the obtained compounds. The determination of the affinities
for α
1
and D receptors was prioritized because activities at such receptors may interfere in vivo
2
with the activity elicited by 5-HT1A receptor activation.65–67 Moreover, the pertinence of testing
those targets was also highlighted by the results of SafetyScreen44 for compound 17, which
showed relatively higher affinity for α
1
and D
2
receptors than for other off-targets. Concerning α
1
receptor affinity, the majority of the compounds showed substantial selectivity for 5-HT1A receptor
(over 1000–10000-fold). Selectivity dropped lower than 100-fold only for compounds 11, 13 and
1
5, which are pyridine analogs with non-oxygen linker. In the case of the D
2
receptor, none of the
19
ACS Paragon Plus Environment

Page 21 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
compounds had pK
i
over 6 (Tables 1 and 2). Additionally, we examined the binding affinity of
each compound for serotonin 5-HT2A , muscarinic M
1
and histamine H receptors (see Supporting
1
Information Table S3). None of the tested compounds exceed 50% of control specific binding at a
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
concentration of 1 µM in these assays.
Considering the above and bearing in mind that the selected lead structure 17 showed substantial
selectivity also vs. the other 49 off-targets (Table 3), it may be concluded that the designed series
is characterized by a high selectivity for 5-HT1A receptors.
Structure-Functional Activity Relationships
To investigate the functional profiles of the presented series we evaluated functional activity of
twelve selected compounds (10–17, 19, 20, 22 and 27) in the following pathways engaged in
5
-HT1A receptor signal transduction: ERK1/2 phosphorylation (pERK1/2), adenylyl cyclase
2
+
inhibition (cAMP), β-arrestin recruitment (β-arrestin) and calcium mobilization (Ca ) (Figure 4).
20
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 85
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. Signaling pathways measured after activation of 5-HT1A receptor; adopted from Chilmonczyk et. al. 68 and
modified based on Rojas & Fiedler.
69
To our knowledge, the present study is the first in which such a large number of 5-HT1A receptor
ligands was tested in such a number of intracellular signaling tests. This is also only the second
7
0
study in which 5-HT1A receptor ligands were examined for β-arrestin recruitment. Here, we
assumed that Emax values higher than 80% relative to the maximal effect of serotonin are
characteristic of a full agonist, Emax values between 79% and 21% indicate a partial agonist and
Emax values 20% or less are indicative of an antagonist. The experiments were carried out using
cell lines expressing the recombinant human 5-HT1A receptor. Although such cell lines may differ
from neuronal systems e.g. as concerns their signal transduction machinery or receptor expression
21
ACS Paragon Plus Environment

Page 23 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
levels (336 fmol/mg for CHO-K1 cells and 78 fmol/mg for U2OS cells), they provide an
experimentally robust and widely accepted means of comparing drug activities and conducting
SAR studies. In the case of the investigated series, the aryloxyethanamine derivatives (16, 17 and
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
9) were the ones that showed the highest potency in functional activity assays (Table 5), as
discussed in greater detail below. Dose-response curves for functional activity of all the
compounds are presented in Supporting Information.
ERK1/2 phosphorylation
Derivatives containing S, NH or CH in the linker were over one order of magnitude weaker than
2
their aryloxyethanamine derivatives, with the compound 14 (phenylsulfanylethanamine
derivative), exhibiting the lowest intrinsic activity in this pathway (pEC50 = 6.77). Except for
compounds 14 and 15, no potency differentiation between phenyl derivatives and their pyridine
counterparts, as well as between 3-chloro-4-fluorobenzoyl derivative (17) and its 3,4-
dichlorobenzoyl analog (19) were observed. The potencies of aryloxyethanamines were higher
than those of all references, including 1.
In contrast to buspirone, agonist efficacy (Emax) of all the tested compounds was high, i.e. between
7
7% and 95%. The highest agonist efficacy, comparable with 8-OH-DPAT, was observed for
compounds 11 and 16, however none of the compounds reached the 100% efficacy level of 1.
22
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 85
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 5. Functional activity of compounds 10–17, 19 and references at 5-HT1A receptors.
5
-HT1A receptor functional activitya
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
ERK1/2b
cAMPb
β-arrestinc
Ca2+ b
Compd
X
R
E
MAX
pEC50
E
MAX
pEC50
E
MAX
pEC50
E
MAX
pEC50
1
0
1
F
77%
95%
7.88
91%
74%
7.10
99%
6.35
6.94
74%
87%
7.07
1
F
F
7.90
7.59
7.33
6.73
101%
7.84
6.98
1
2
3
81%
83%
79%
86%
104%
99%
5.70
6.58
65%
52%
1
F
7.76
7.26
7.13
1
4
5
F
F
F
F
Cl
F
F
F
85%
84%
92%
81%
85%
99%
100%
72%
6.77
7.89
9.10
9.37
9.06
7.62
8.31
7.28
71%
76%
90%
89%
71%
100%
90%
41%
5.61
7.18
8.09
8.05
8.04
7.32
7.68
6.03
70%
102%
96%
98%
99%
101%
97%
75%
5.44
6.54
7.98
7.68
6.76
6.87
7.60
5.56
NT
NT
1
73%
52%
57%
58%
98%
89%
NTd
6.95
7.39
7.76
7.41
7.34
7.54
NT
16
17
19
20
22
27
1
8
100%
93%
8.33
8.09
7.82
7.48
92%
63%
7.22
7.50
7.14
7.51
98%
101%
100%
100%
6.71
7.84
6.73
60%
NT
6.61
NT
-OH-DPAT
Buspirone
44%
49%
NT
NT
Serotonin
100%
100%
6.89 100%
7.30
a
All the functional activity values were expressed as means from at least 3 experiments performed in duplicates.
For the sake of clarity, the SEM values were omitted in this table and presented in the supporting information –
b
c
d
Table 3; The functional assay was performed using CHO-K1 cells, U2OS cells (Tango LiveBLAzer assay kit) ; NT –
not tested
23
ACS Paragon Plus Environment

Page 25 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Inhibition of cAMP production (cAMP)
Among the most potent derivatives with oxyethylene linker, no differences between the phenyl
derivative and the pyridine analog were observed. By contrast, in the case of aminoethylene and
sulfanylethylene linker, the pyridine counterparts showed higher cAMP potency than their phenyl
derivatives (15 and 13 vs. 14 and 12). Potencies of the most active compounds (16, 17 and 19)
were higher than those of reference compounds.
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
In the case of agonist efficacy, most of the compounds behaved as partial agonists, however their
efficacy was higher (71–79%) than for the reference partial agonists, buspirone and 8-OH-DPAT
(49% and 63% respectively). Compounds 10, 13, 16, and 17 were full agonists with efficacy (86–
9
1%) comparable to 1. There were no marked differences between the pyridine and phenyl
derivatives, except for the analogs with propylene linker, where phenyl derivative 10 was a full
agonist (91%) while its pyridine counterpart 11 behaved as a partial agonist (74%).
β-arrestin recruitment (β-arrestin)
In the case of the analogs with oxyethylene linker, the pyridine derivatives were weaker than their
phenyl counterparts, whereas in the case of other linkers they had higher potency. Similarly to
previously discussed pathways, 3,4-dichlorobenzoyl analog showed lower potency than its 3-
chloro-4-fluoro- counterpart. In this pathway, the most active compounds showed potency
comparable to that of 8-OH-DPAT, and higher than 1 and buspirone.
Only a phenylsulphanyl derivative 14 behaved as a partial agonist (70%). The other compounds
exhibited Emax values close to 100%, similar to that seen for the reference compounds.
Calcium mobilization (Ca2+).
24
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 85
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
In this assay, the pyridine derivatives showed higher potency than their phenyl analogues. All the
tested compounds had pEC50 values higher than that of 1, with the most potent being 11 and 17.
In comparison to other pathways, efficacies in the calcium mobilization assay were markedly
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
lower, with all compounds except 11 being partial agonists (52–74%). The propylene derivatives
1
0 and 11 displayed the highest efficacy (74% and 87% respectively).
Functional activity of the compound 17 analogs: 20, 22 and 27
Changing the position of nitrogen atom in the aryloxyl moiety from 2 to 3, as well as replacement
of 2-pyridine by 2-pyrimidine and 2-pyrazine resulted in decreased potency in each of the
pathways. Pyrazin-2-yloxy derivative 22 was the most potent compound in this group and
exhibited potency in a range from 7.54 in calcium mobilisation assay to 8.31 in the ERK1/2
phosphorylation assay. Similarly to affinity results, the pEC50 values of compound 22 are
intermediate between pEC50 values for pyridin-2-yloxy (17) and pyridin-3-yloxy (20) derivatives.
The 5-methylpyrimidin-2-yloxy derivative 27 showed the lowest potency.Compounds 20 and 22
displayed high agonist efficacy (89–101%) in all the performed assays, showing higher efficacy
levels than 17, especially in the case of ERK1/2 phosphorylation and calcium mobilisation. By
contrast, compound 27 behaved as a partial agonist in all the pathways (41–75%, with exception
of calcium mobilisation, where it was not tested).
Bias factors
In order to determine functional selectivity of the tested ligands, bias factors between particular
pathways were calculated, according to the following equation: 71–74
25
ACS Paragon Plus Environment

Page 27 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
relative activityퟏ2,ꢀꢁꢂ
Bias factor = log (
) =
relative activityퟏ2,ꢃꢄꢅ
퐸mꢇxpꢈꢉh1 × EC50−ꢊꢇꢋꢌ2
퐸mꢇxpꢈꢉh1 × EC50−ꢊꢇꢋꢌ2
= log ꢆ(
)
÷ (
)
ꢏ
EC50−ꢊꢇꢋꢌꢍ × 퐸mꢇxpꢈꢉhꢎ
EC50−ꢊꢇꢋꢌꢍ × 퐸mꢇxpꢈꢉhꢎ
ꢀꢁꢂ
ꢃꢄꢅ
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
Bias factor ensures equal comparison between all the pathways, considering the differences in
EC50 and Emax values of both a test ligand and a reference agonist, which was serotonin in this case.
Results for all the compounds were presented in Table 6. Compounds which displayed bias factor
higher than 1 (equaling to >10× preference) were highlighted in green (for positive values) and in
blue (for negative values). Those compounds which showed significant bias, but with pEC50 values
which were relatively low (<8 in all assays), were considered less interesting and marked in gray.
26
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 85
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 6. Bias factors of compounds 10–17, 19, 20, 22, 27 and references at 5-HT1A receptors.
5-HT1A receptor bias factor (logarithmic value)
ERK1/2
vs. cAMP
ERK1/2 vs. ERK1/2
cAMP vs.
β-arrestin
cAMP
vs. Ca
β-arrestin
vs. Ca
2+
2+
2+
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
Compd
X
R
β-arrestin
vs. Ca
1
1
1
0
1
2
F
0.74
0.71
0.90
0.51
1.27
0.78
1.05
1.30
1.13
0.33
0.70
1.52
1.17
0.79
0.66
0.00
0.82
0.65
0.09
-0.37
0.28
-0.09
-0.17
-0.42
-0.66
0.15
–
F
F
F
F
F
F
F
Cl
F
F
F
0.34
1.19
0.50
0.82
0.67
0.51
1.01
1.63
0.15
0.12
1.10
1.02
-0.39
0.14
0.00
-0.09
0.52
0.65
–a
-0.80
-0.38
0.15
–
13
-0.01
-0.45
-0.12
-0.54
-0.30
0.51
1
4
5
1
0.81
1.78
1.57
1.63
0.10
0.63
–
0.03
0.73
0.27
0.50
-0.23
-0.07
–
0.15
1.27
0.57
0.00
-0.05
0.51
–
16
1
7
9
1
20
-0.18
-0.58
-0.42
-0.15
-1.18
-0.52
0.00
2
2
1
8
2
7
1.76
–
0.59
–
0.74
–
-OH-DPAT
Buspirone
Serotonin
–
–
–
0.00
0.00
0.00
a- no data for Ca2+ assay
In general, the novel series showed highest bias for ERK1/2 phosphorylation, especially vs. cAMP
2
+
and Ca (mean bias factors 0.91 and 0.83, respectively). The bias was most pronounced in the
case of the aryloxyethanamine derivatives, where high preference (bias factors >1) was observed
27
ACS Paragon Plus Environment

Page 29 of 85
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
in majority of cases. In this respect, the novel aryloxyethanamines were the most similar to the
reference 5-HT1A receptor biased agonist 1, which previously showed preference for ERK1/2
phosphorylation, particularly over cAMP inhibiton.27
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
Noteworthy, as mentioned in the Introduction, it has been suggested that intensification of signal
transduction through phosphorylation of ERK1/2 kinases may be important for antidepressant
effects.75–77 In fact, in post-mortem examinations of suicide victims with previously diagnosed
depression, reduction of ERK1/2 expression and phosphorylation in brain structures was
observed.77,78 Furthermore, depression caused by chronic stress was found to be associated with
ERK phosphorylation deficits,79–82 and the chronic administration of ERK inhibitors caused
8
1
anhedonia and anxiety behavior. On the other hand, the inhibition of cAMP synthesis in the
hippocampus resulting from the activation of the 5-HT1A receptor may lead to reduced
2
+
phosphorylation of the Ca /calmodulin-dependent protein kinase II (CaMKII) and thus result in
impairment of cognitive functions.83 The above considerations further support the need for
selective stimulation of a desired intracellular signaling in order to achieve optimal therapeutic
benefit. In this context, the newly discovered aryloxyethyl derivatives of 1-(1-benzoylpiperidin-4-
yl)methanamine represent a favorable extension of the prototype ERK1/2-preferring 5-HT1A
receptor biased agonist 1, which is, so far, the sole representative of such a functional profile. The
present study also evidenced clear difference in functional activity between these compounds and
the widely used reference agonist 8-OH-DPAT and the partial agonist buspirone.
28
ACS Paragon Plus Environment