Discovery of small-molecule nonpeptide antagonists of nociceptin/orphanin FQ receptor: The studies of design, synthesis, and structure-Activity relationships for (4-Arylpiperidine substituted-methyl)-[bicyclic (hetero)cycloalkanobenzene] derivatives

Article abstract of DOI:10.1016/j.ejmech.2016.02.014

Nociceptin/orphanin FQ (N/OFQ) and N/OFQ peptide (NOP) receptor are expressed and distributed in various regions such as central nervous system (CNS), peripheral nervous system, immune system, and peripheral tissues. N/OFQ and NOP receptor have important

Full text of DOI:10.1016/j.ejmech.2016.02.014

Accepted Manuscript
Discovery of small-molecule nonpeptide antagonists of nociceptin/orphanin FQ
receptor: The studies of design, synthesis, and structure–activity relationships for (4-
arylpiperidine substituted-methyl)-[bicyclic (hetero)cycloalkanobenzene] derivatives
Shigeo Hayashi, Katsuyo Ohashi, Sachiko Mihara, Eriko Nakata, Chie Emoto, Atsuko
Ohta
PII:
S0223-5234(16)30084-8
10.1016/j.ejmech.2016.02.014
EJMECH 8365
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 11 May 2015
Revised Date: 6 January 2016
Accepted Date: 4 February 2016
Please cite this article as: S. Hayashi, K. Ohashi, S. Mihara, E. Nakata, C. Emoto, A. Ohta, Discovery
of small-molecule nonpeptide antagonists of nociceptin/orphanin FQ receptor: The studies of design,
synthesis, and structure–activity relationships for (4-arylpiperidine substituted-methyl)-[bicyclic
(hetero)cycloalkanobenzene] derivatives, European Journal of Medicinal Chemistry (2016), doi: 10.1016/
j.ejmech.2016.02.014.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Running title: Design, synthesis, and SAR studies of nonpeptide NOP receptor antagonists
Discovery of small-molecule nonpeptide antagonists of
nociceptin/orphanin FQ receptor: The studies of design,
synthesis, and structure–activity relationships for (4-
arylpiperidine substituted-methyl)-[bicyclic
(hetero)cycloalkanobenzene] derivatives
Shigeo Hayashi*, Katsuyo Ohashi, Sachiko Mihara, Eriko Nakata, Chie Emoto, Atsuko Ohta
Pfizer Global Research & Development, Nagoya Laboratories, Pfizer Japan Inc., 5-2
Taketoyo, Aichi 470-2393, Japan
*
Corresponding author. E-mail address: Shigeo_Hayashi@nifty.com
Abstract
Nociceptin/orphanin FQ (N/OFQ) and N/OFQ peptide (NOP) receptor are expressed and distributed
in various regions such as central nervous system (CNS), peripheral nervous system, immune system,
and peripheral tissues. N/OFQ and NOP receptor have important roles on a variety of physiological,
pathophysiological, regulatory, and dysregulatory mechanisms in the living body. Both activation and
blockade of NOP receptor function have displayed clinical potential of NOP receptor agonists and
1

ACCEPTED MANUSCRIPT
antagonists for the treatment of various diseases or pathophysiological conditions, respectively. Potent
and selective NOP receptor agonists/antagonists are also useful tools to investigate the various
mechanisms mediated by NOP receptor–N/OFQ system. As the present study, a series of (4-
arylpiperidine substituted-methyl)-[bicyclic (hetero)cycloalkanobenzene] analogs was designed,
synthesized, and biologically evaluated in vitro to seek and identify potent and selective, small-
molecules of nonpeptide NOP receptor antagonists, which resulted in the discovery of novel potent
small-molecule 15 with high human NOP receptor selectivity over human µ receptor. The structure–
activity relationship (SAR) of the potency and selectivity, structure–metabolic stability relationship
(SMR), and SAR of hERG (human ether-a-go-go related gene) potassium ion channel binding affinity
for the analogs in the present studies in vitro provided or suggested significant and/or useful structural
determinants and insights for the respective purposes. The superior profiles of compound 15 are
discussed with a viewpoint of multisite interactions between ligand and NOP receptor, together with the
results of previous NOP receptor agonist/antagonist studies.
Keywords Nociceptin/Orphanin FQ (N/OFQ) · Human NOP (ORL1) receptor selective antagonist ·
Multisite interactions with NOP receptor · Metabolic stability · HERG potassium ion channel binding ·
Prolong QT interval issues.
Abbreviations: GPCR, G-protein-coupled receptor; N/OFQ, nociceptin/orphanin FQ; NOP receptor,
N/OFQ peptide receptor [or opioid-receptor-like-1 (ORL1) receptor]; hNOP receptor, human NOP
3
5
35
receptor; [ S]GTPγS, [ S]guanosine 5'-(γ-thiotriphosphate); SAR, structure–activity relationship; SMR,
structure–metabolic stability relationship; HPCOM, {1-[4-(2-{hexahydropyrrolo[3,4-c]pyrrol-2(1H)-yl}-
1
H-benzimidazol-1-yl)piperidin-1-yl]cyclooctyl}methanol; MCOPPB, 1-[1-(1-methylcyclooctyl)-4-
piperidinyl]-2-[(3R)-3-piperidinyl]-1H-benzimidazole; HEK, human embryonic kidney; CHO, Chinese
hamster ovary; HLM, human liver microsome; hERG, human ether-a-go-go related gene; WSCI, water
soluble carbodiimide, that is, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride; HOBT, 1-
hydroxybenzotriazole; LAH, lithium aluminum hydride; HBD, hydrogen bond donor.
1
. Introduction
2

ACCEPTED MANUSCRIPT
Nociceptin [1a]/orphanin FQ [1b] (N/OFQ), that is a metabolite of precursor polypeptide (prepro-
N/OFQ), has seventeen amino acid sequence. N/OFQ is an endogenous ligand of N/OFQ peptide (NOP)
receptor [or opioid-receptor-like-1 (ORL1) receptor]. NOP receptor has seven transmembrane domains
in its structures, and belongs to heterotrimeric GTP-binding protein (G-protein)-coupled receptors
(GPCRs) [1c–e]. N/OFQ and NOP receptor are widely expressed and distributed, especially, in the
central nervous system (CNS), peripheral nervous system (PNS), immune system, and peripheral tissues,
such as brain, spinal cord, peripheral sensory ganglia neurons, lymphocytes, monocytes, and smooth
muscles. NOP receptor activation by N/OFQ induces (i) inhibition of cyclic adenosine-3',5'-
2
+
+
monophosphate (cAMP) accumulation and of Ca channel currents, (ii) enhance of K channel currents,
(iii) regulation or attenuation of concentration levels of various signal messengers or neurotransmitters
such as tachykinins, noradrenaline, 5-hydroxytriptamine, dopamine, and proinflammatory cytokines, (iv)
suppression of excitatory synaptic transmissions of neurons, (v) activation of mitogen-activated protein
kinases (MAPKs) and of nuclear factor κB (NFκB), and (vi) modulation of immune functions and of
transcriptional activations. The biological responses are dependent on the regions, sites, tissues, systems,
and physiological conditions, respectively. Hence, NOP receptor–N/OFQ system has key roles in a
variety of physiological signal-transduction pathways and behavioral functions in the living body [1a–
e,2a–g] (for detailed roles of NOP receptor–N/OFQ system, see also refs. [3–7] and cited references
therein). Notably, the expression/distribution levels of (i) NOP receptor, its messenger RNA (mRNA),
and splice variants of the receptor, and of (ii) prepro-N/OFQ, N/OFQ, and other metabolites of prepro-
N/OFQ are widely varied dependent on (patho)physiological or disease conditions, and are dependent on
the regions, sites, tissues, and systems of the body in animal models or in humans [2c,8l–o].
Furthermore, N/OFQ is metabolically converted into inactive or variously-bioactive shorter-fragments
by (i) the corresponding processing-enzymes such as peptidases or (ii) further processing/truncation,
dependent on the regions/sites/tissues of the body [2a,3,5,9–14] (see also further discussion in refs.
[
3,5]).
3

ACCEPTED MANUSCRIPT
Significantly, both activation and inhibition of NOP receptor function have displayed clinical potential
of NOP receptor agonists and antagonists, respectively [2a,b,e]. Thus, utilization of NOP receptor
activation is an attractive target for the treatment of neuropathic pain, anxiety disorders, and other
diseases (see refs. [3–5] and cited references therein). Blockade against N/OFQ-induced biological
function of NOP receptor is an attractive target for the treatment of depressive disorders [15],
hyperphagia or hyperphagic obesity [16,17], and hypotension [18]. NOP receptor is distinct from the
receptors of classical opioid peptides, although the receptors share high structural and genetic-coding
3
5
similarities. NOP receptor–N/OFQ binding stimulated-[ S]guanosine 5'-(γ-thiotriphosphate)
3
5
(
[ S]GTPγS) binding response in the spinal cord of rodents was not affected by antagonists of classical
3
5
opioid receptors or of other various GPCRs; but the [ S]GTPγS binding response was inhibited by a
NOP receptor antagonist [19a]. The N/OFQ induced-inhibition of cAMP accumulation was not affected
by classical opioid receptor antagonists or agonists in vitro [1b]. Besides, classical opioid drugs have
been broadly prescribed with careful restriction or control in the medical practices, owing to classical
opioid receptor-derived serious adverse effects, such as respiratory restriction and sedation [19b–d].
Therefore, in drug-discovery studies of NOP receptor agonists/antagonists toward clinical drugs, the
assessments of their NOP receptor selectivities are very important, especially against µ receptor [or µ-
opioid peptide (MOP) receptor], to avoid the adverse effects.
Taken together, potent, selective, and metabolically stable NOP receptor agonists and antagonists are
important to investigate their (potential) clinical utilities. The compounds are also useful tools to
scrutinize a variety of (patho)physiological, regulatory, and disease mechanisms mediated by the
receptor and metabolically unstable N/OFQ, with confirming/verifying intrinsic activity of the
endogenous ligand in the mechanisms. In fact, design, synthesis, structure–activity relationship (SAR)
and structure―metabolic stability relationship (SMR) of distinctive nonpeptide NOP receptor agonists
and antagonists have been investigated, respectively, by Hayashi et al. [3–7]. The agonist studies
resulted in the discovery of NOP receptor full agonists, {1-[4-(2-{hexahydropyrrolo[3,4-c]pyrrol-2(1H)-
yl}-1H-benzimidazol-1-yl)piperidin-1-yl]cyclooctyl}methanol (HPCOM) as a peripheral and/or spinal
4

ACCEPTED MANUSCRIPT
cord-selective, systemically potent analgesic for the treatment of neuropathic pain [3] and 1-[1-(1-
methylcyclooctyl)-4-piperidinyl]-2-[(3R)-3-piperidinyl]-1H-benzimidazole (MCOPPB) as a brain
selective, orally potent anxiolytic [4,5], with respective unique and significant profiles in vitro and in
vivo. The antagonist studies resulted in the discovery of several significant and characteristic NOP
receptor antagonists in vitro [6,7]. Alternatively, various types of chemical entities of NOP receptor
antagonists are useful or helpful to explore further opportunity/potentiality. Moreover, different or new
types of ligands might be helpful to light on different or new aspects of interaction mechanisms between
NOP receptor and agonists/antagonists. Particularly, small-size ligands might be useful tools to search
for functionally/transductionally indispensable interaction point/region/residue of NOP receptor protein
[
7].
Hence, we explored small molecules of further chemotypes that were distinctive from the analogs of
the previous reports. Thus, design, synthesis, SAR, and SMR of a series of (4-arylpiperidine substituted-
methyl)-[bicyclic (hetero)cycloalkanobenzene] analogs were investigated in vitro in order to seek and
identify potent, selective, and metabolically stable, novel nonpeptide NOP receptor antagonists. Design
and SAR of the analogs for reducing hERG (human ether-a-go-go related gene) potassium ion channel
binding affinity were also investigated in vitro as a safe-drug study. Through the studies, key or
significant factors and molecular bases for getting or improving desired characteristics of the analogs
were investigated in fundamental and practical viewpoints for the present and further studies. Herein,
the unique and significant results and findings are described and discussed, together with the results of
the previous NOP receptor antagonist and agonist studies.
2
2
. Results and Discussion
.1. Chemistry
5

ACCEPTED MANUSCRIPT
Various types of (4-arylpiperidine substituted-methyl)-[bicyclic (hetero)cycloalkanobenzene] analogs
were prepared for the present design, SAR, and SMR studies.
First, as depicted in Scheme 1, tetrahydroisoquinoline 5 was synthesized. Thus, 4-(4-fluoro-2-
methylphenyl)piperidine 1 was prepared [6] then reacted with 2-(tert-butoxycarbonyl)-1,2,3,4-
tetrahydroisoquinoline-3-carboxylic acid 2 [4,20]. The N-Boc portion of the resulting amide 3 was
deprotected by acidic condition to afford compound 4. The carboxamide was subsequently reduced with
lithium aluminum hydride (LAH) [7] to afford the desired product 5, exclusively.
As depicted in Scheme 2, 5-chloro-tetrahydroisoquinolin-8-ol 15 was synthesized. Thus, diethyl
acetamidomalonate 6 and O-silylated 3-(bromomethyl)-4-chlorophenol 7 [21] were coupled using NaH
to afford compound 8 followed by treatment of concentrated HCl in acetic acid to give 2-chloro-5-
hydroxyphenylalanine 9. Pictet–Spengler tetrahydroisoquinoline synthesis was then performed with
formaldehyde to afford compound 10 [22]. The secondary-amine was N-Boc protected to afford
compound 11. The product was contained a by-product 12 that would plausibly be N,O-di-Boc
derivative of compound 11, that is, 2-(tert-butoxycarbonyl)-8-(tert-butoxycarbonyloxy)-5-chloro-
1
,2,3,4-tetrahydroisoquinoline-3-carboxylic acid. Finally, compounds 1 and 11 were exclusively
transformed into the requisite product by sequential reactions, that is, amidation to afford compound 13,
N-deprotection to afford compound 14, and borane reduction to afford compound 15.
Other tetrahydroisoquinolines 18 and 19 were prepared in a similar manner for the above synthesis of
compounds 4 and 5, respectively. Thus, after spiro[indane-1,4'-piperidine] 16 and compound 2 were
coupled, the resulting compound 17 was N-deprotected by acidic condition to afford compound 18. The
carboxamide was reduced with LAH to afford compound 19. Modifications around cycloamino moiety
within the tetrahydroisoquinoline skeleton of compound 19 were also carried out. Thus, N-Me derivative
2
0 was synthesized from compound 17 by simultaneous LAH reductions of the N-Boc portion [4] and of
the linker carbonyl portion. N-Me-1-oxo derivative 21 was synthesized by KMnO oxidation [23] of
4
compound 20. N-Ac derivative 22 was synthesized by acetylation of compound 19 (Scheme 3).
Furthermore, various modifications on the phenyl moiety within the tetrahydroisoquinoline skeleton
6

ACCEPTED MANUSCRIPT
were performed. 5-Chloro-8-hydroxy derivative 23 was prepared from compounds 16 and 11 by way of
amidation, N-deprotection, and borane reduction in the same way for the above synthesis of compound
1
5 (Scheme 4). Using the same manner, 5-bromo-8-hydroxy derivative 25 was obtained from the
corresponding 5-bromo-8-hydroxy-1,2,3,4-tetrahydroisoquinoline-3-carboxylic acid 24 [22] (Scheme 4).
-Bromo-8-hydroxy derivative 25 was N-Boc protected to prevent low conversion in the following
5
reaction; then hydrogenolysis of the 5-bromo group was performed similar to a synthesis of a N-Boc
des-bromo-tetrahydroisoquinoline derivative [22]. The resulting compound 26 was N-deprotected to
afford 8-hydroxy derivative 27. N-Boc compound 29, that was prepared from 6-hydroxy-1,2,3,4-
tetrahydroisoquinoline-3-carboxylic acid 28 [24], was converted into the corresponding 6-hydroxy
derivative 30 in the same way for the synthesis of compound 15 (Scheme 5).
On the other hand, chromane 33 was prepared from compound 16 and chromane-3-carboxylic acid 31
in a similar manner for the synthesis of compound 5 (Scheme 6). Tetrahydronaphthalene 38 was
prepared from compound 16 and 1,2,3,4-tetrahydronaphthalene-2-carboxylic acid 37 in high-speed
parallel synthesis of 1'-(substituted methyl)-spiro[indane-1,4'-piperidine] derivatives 36. As the
synthesis, compound 16 was coupled with carboxylic acids 34 by way of carboxylic acid activation
using polymer supported carbodiimide. The resulting amides 35 were reduced with LAH to afford
compounds 36 (Scheme 7, further data not shown).
2
.2. Design and SAR studies of NOP receptor antagonists in vitro
The design, synthesis, and SAR of the present (4-arylpiperidine substituted-methyl)-[bicyclic
(hetero)cycloalkanobenzene] analogs were investigated in vitro in order to seek and identify potent and
selective, novel small molecules of human NOP (hNOP) receptor antagonists, which was the primary
purpose of this study. Notably, the design study was also performed with pharmacokinetic study in vitro
and hERG binding study in vitro for the analogs in multiviewpoint strategies (see later sections). The in
vitro biological assays were performed with the described methods in the previous reports [6,7,25].
7

ACCEPTED MANUSCRIPT
Briefly, the binding affinities (K
i
values) of the synthetic analogs to hNOP receptor were measured by
displacement of tritium-labeled N/OFQ for the recombinant hNOP receptor expressed in human
embryonic kidney (HEK)-293 cells. For the evaluation of the in vitro binding selectivities for NOP
receptor over µ receptor, the binding affinities of the analogs to human µ receptor expressed in Chinese
3
hamster ovary (CHO)-K1 cells were measured by displacement of radiolabeled ligand, [ H]DAMGO.
The hNOP receptor antagonist activities of the analogs were evaluated as inhibitory activities against
functional activity of hNOP receptor in response to N/OFQ binding. Thus, the activities (IC50 values)
3
5
were determined against [ S]GTPγS binding to α-unit of G-protein stimulated by hNOP receptor–
N/OFQ binding in HEK-293 cells. The in vitro SAR of the analogs is summarized in Table 1.
In the previous studies [6,7], effective structural units to get potent and selective hNOP receptor
antagonists were designed, and the respective significant features were illustrated in Scheme 8. For
example, 4-(4-fluoro-2-methylphenyl)piperidine unit was established as a preferable β-substituent
portion of 1-(β-amino substituted-β-alanyl)-N,N-dimethylindoline-2-carboxamide analogs for potency
and selectivity; especially attributed to the 4-fluoro and 2-methyl substituents of the unit for compound
3
9 (1-{3-[4-(4-fluoro-2-methylphenyl)piperidin-1-yl]propanoyl}-N,N-dimethylindoline-2-carboxamide)
[
6]. As another structural feature of the compound, two amido-portions and one bicyclo-portion of the
N,N-dimethylindoline-2-carboxamide unit restrict the changes of conformation/orientation of the
portions, whereas the whole structure of the molecule is some-extent large and intricate. By contrast,
compounds 40, thus, 1-(aza-alicyclic or carbo-alicyclic group substituted-propyl)-4-arylpiperidine
analogs have smaller and simpler structures than compound 39; and effectiveness of the monocyclic unit
at C-3 position of the three-methylene linker on the arylpiperidine unit for potency and selectivity were
displayed in comparison with various 1-(3'-substituted propyl)-4-arylpiperidine derivatives [7]. As an
imaginary ring-closure way that is illustrated for compound 39', one methylene can be inserted between
C-2 position of the linker portion on the piperidine ring and ortho-position of the phenyl ring. In this
case, it is interesting that the nitrogen atom of the piperidine ring and the ortho-carbon atom of the
8

ACCEPTED MANUSCRIPT
phenyl ring are connected with four single-bonds, similar to the four single-bonds of the linker portion
between the piperidine ring and the monocyclic unit of compounds 40. On the basis of the above results
and consideration, (i) combination of 4-(4-fluoro-2-methylphenyl)piperidine unit and cyclic unit by
connecting with four single-bonds and (ii) restriction of location/orientation of the cyclic unit by ring
closure between the unit and the linker portion, such as a tetrahydroisoquinoline ring, were expected to
be a promising or attractive design concept of potent and selective, small-molecules of NOP receptor
antagonists. In fact, newly designed compounds 41, that is, 3-{[4-(4-fluoro-2-methylphenyl)piperidin-1-
yl]methyl}-tetrahydroisoquinoline analogs exhibited superior characteristics as small molecules.
Tetrahydroisoquinoline analog 5 and 5-chloro-tetrahydroisoquinolin-8-ol analog 15 demonstrated potent
i
hNOP receptor binding affinities (K = 3.56 nM and 3.03 nM), high hNOP receptor binding selectivities
against human µ receptor (96.1-fold and greater than 143-fold), and potent hNOP receptor antagonist
activities (IC50 = 58.8 nM and 28.7 nM), respectively. Significantly, compared with 4-arylpiperidine
portion of tetrahydroisoquinoline analogs, the binding selectivities of 4-(4-fluoro-2-
methylphenyl)piperidine analogs 5 and 15 were much higher than that of the corresponding
spiro[indane-1,4'-piperidine] analogs 19 and 23, respectively. Notably, in comparison with unsubstituted
tetrahydroisoquinoline analogs, compound 19 exhibited 3.63-fold drop in binding selectivity and 4.22-
fold decrease in antagonist activity than that of compound 5, respectively.
With respect to the linkage part between 4-arylpiperidine portion and tetrahydroisoquinoline portion,
replacement of the methylene linker of compound 19 by carbonyl group, thus, amide analog 18, led to
loss of binding affinity (see later discussion).
SAR study of benzo-fused bicyclo portion using spiro[indane-1,4'-piperidine] template gave further
significant results.
Modifications on nitrogen atom of tetrahydroisoquinoline ring, even by small group, led to less potent
binding affinities and lower binding selectivities relative to compound 19. The rank of the orders of
binding affinities and binding selectivities was related to the size/volume of this portion, thus NH analog
1
9 > N-Me analog 20 > N-Ac analog 22, respectively. N-Me-1-oxo (lactam) analog 21 lost hNOP
9

ACCEPTED MANUSCRIPT
receptor binding affinity and retained weak human µ receptor binding affinity. N-Acetamide analog of
compound 19, that is, 2-[3-{(spiro[indane-1,4'-piperidin]-1'-yl)methyl}-3,4-dihydroisoquinolin-2(1H)-
yl]acetamide (not shown), led to inferior profiles as well, that is, K = 18.1 nM for hNOP receptor, K =
i
i
1
08 nM for human µ receptor, and 5.97-fold hNOP receptor binding selectivity against human µ
receptor. Substituent effects on the aryl moiety within the bicyclo portion, with the various modification
patterns (functional group, position, and/or number of substituent), highly influenced antagonist profiles.
For example, relative to compound 23, tetrahydroisoquinolin-8-ol analog 27 showed comparative
binding affinity and higher binding selectivity, but less potent antagonist activity. Exchanging 8-hydroxy
group (27) by 6-hydroxy group (30) gave much less potent binding affinity, higher binding selectivity,
and much less antagonist activity, which indicated regioselective effects of phenolic hydroxy group for
the profiles.
Furthermore, replacement of the azacyclic moiety within the tetrahydroisoquinoline ring by other
(hetero) alicyclic moiety was investigated to explore SAR. Exchanging by oxacyclic moiety, chromane
analog 33 exhibited potent binding affinity (K = 5.65 nM) with potent antagonist activity as an
i
unsubstituted bicyclic analog, although its binding selectivity was reduced relative to
tetrahydroisoquinoline analog 19. Relative to oxacyclic analog 33, carbocyclic analog 38 showed
comparative binding affinity and improved hNOP receptor binding selectivity (52.1-fold), but much
reduced antagonist activity (IC = 683 nM).
5
0
As presented in the SAR studies around (hetero) alicyclic moiety including the modifications around
azacyclic moiety of tetrahydroisoquinoline ring such as compounds 19–22, the characters of the moiety,
that is, atom species, functional group, basicity, steric hindrance, and modification pattern/mode, are
crucial for increasing or decreasing of binding affinities, binding selectivities, and antagonist activities.
In the previous study for comparison of substitutions on 1'-propyl portion of spiro[indane-1,4'-
piperidine] analogs [7] (see also compounds 40, Scheme 8), 1'-(3-phenylpropyl)-spiro[indane-1,4'-
piperidine] was a poor analog relative to 1'-[3-(unsubstituted aza-alicyclic or carbo-alicyclic group)
substituted-propyl]-spiro[indane-1,4'-piperidine] analogs. Significantly, the profiles of the 3-
1
0

ACCEPTED MANUSCRIPT
phenylpropyl analog were also inferior to that of the present tetrahydroisoquinoline analog 19, chromane
analog 33, and tetrahydronaphthalene analog 38 that have the same spiro[indane-1,4'-piperidine] unit.
Indeed, the 3-phenylpropyl analog showed low hNOP receptor binding affinity (K = 23.0 nM), low
i
hNOP receptor selectivity against human µ receptor (3.0-fold), and very weak hNOP receptor antagonist
activity (IC50 = 2,087 nM). The preference of the bicyclic analogs over the 3-phenylpropyl analog would
be due to or contributed by restriction/rigidity effect of the location and orientation for the aryl moiety as
a part of bicyclo portion, and/or size/volume effect of the (hetero) alicyclic moiety.
As an additional exploratory study of the bicyclo portion, 1'-[benzo-fused (hetero) aryl]methyl-
spiro[indane-1,4'-piperidine] derivatives were investigated. As a result, the inhibitory activities of the
analogs to hNOP receptor at 100 nM were 56% for 1'-(quinolin-3-ylmethyl) analog, 13% for 1'-
[
[
(naphthalen-2-yl)methyl] analog, 66% for 1'-[(1H-indol-3-yl)methyl] analog, and 9% for 1'-
(benzofuran-2-yl)methyl] analog, respectively (further data not shown). Their weak or very weak hNOP
receptor binding affinities would indicate disadvantage of high π-electron densities of their (hetero) aryl
moieties relative to the benzo-fused (hetero) alicyclic analogs.
It is significant to consider molecular bases for distinct pharmacological activities between agonists
and antagonists of NOP receptor.
In the previous in vitro SAR studies, (i) there were first-order correlations of the levels of functional
3
5
activities (EC50 values for [ S]GTPγS binding-induction) to that of respective binding affinities (K
i
values) for hNOP receptor full agonists, MCOPPB and its derivatives; whereas MCOPPB showed more
potent efficacy than N/OFQ, thus, Emax (efficacy) value of MCOPPB was 140% of maximal response of
N/OFQ [4]. (ii) There were first-order correlations of the levels of inhibitory activities (IC values)
5
0
3
5
against N/OFQ-stimulated [ S]GTPγS binding to that of respective binding affinities (K values) for
i
hNOP receptor antagonists, compound 39 and its derivatives, or representative derivatives of
compounds 40, respectively [6,7]. Consequently, it was indicated that the respective primary
mechanisms of (i) the G-protein activation responses by the agonists and of (ii) the inhibitions by the
1
1

ACCEPTED MANUSCRIPT
antagonists against the N/OFQ-induced biological response are due to the bindings of the respective
ligands to NOP receptor. The distinctive biological responses, that is, activation and inhibition via NOP
receptor binding of the respective ligands come from their differential interaction-patterns/-modes owing
to the respective whole and local structural features of the ligands. Indeed, through the present and
previous studies of potent and selective NOP receptor agonists/antagonists [3,4,6,7], their structural
features of size, shape, and conformational flexibility/rigidity, functional group, and electronic character
have been crucial for their binding affinities, selectivities, and biological activities, respectively.
As studies for the endogenous ligand, it has been reported that the binding regions of NOP receptor
protein for N/OFQ are in the second extracellular loop domain [26–29] and in the transmembrane
domain [26,30,31]. The binding site in the transmembrane domain works for NOP receptor activation by
N/OFQ, which are affected by mutation of amino acid residues of the protein. For example, Asp110,
Gln107, Asp130, Tyr131, Gln280, Asn133, and Gln286 were functionally indispensable to exhibit
N/OFQ binding induced-biological response of the receptor [30–32]. As well, replacement studies for
amino acid residues of N/OFQ have revealed various sites of the heptadecapeptide are required for both
binding and activation of NOP receptor or either of them. For example, the residues located at positions
1
–8, 12, 14, and 17 were important for the receptor activation [33,34]. Besides, it was strongly
suggested that binding mode of a peptide NOP receptor antagonist was different from that of N/OFQ;
and the binding site for the antagonist would be shared in part but not be shared in the other part of the
binding site for N/OFQ [29].
The requirements of the effective multisite-interactions between N/OFQ and NOP receptor in
orthosteric binding site for potent and selective binding and function are in line with the facts that potent
and selective, binding affinities and functional activities of synthetic NOP receptor agonists such as
HPCOM and MCOPPB were derived from their multisite characteristic portions as described in the
SAR studies [3,4]. On the contrary, to block N/OFQ-induced signal transduction of NOP receptor,
distinctive structural features of NOP receptor antagonists are significant in the orthosteric binding site
1
2

ACCEPTED MANUSCRIPT
in order to physically or functionally mask or interfere with the crucial points/regions involved in the
activation of the protein.
For elucidating or understanding detailed molecular mechanisms, the effective or crucial structural
determinants of potencies, selectivities, and activities for the present analogs would be important.
Particularly, the SAR of the present small molecules shows various triggers to set off the antagonism,
including some discrepancies for potency relationships between the binding affinities and the antagonist
activities dependent on their structures. Indeed, (i) relative to the potent antagonist activity of 4-(4-
fluoro-2-methylphenyl)piperidine analog 5, the activity of the corresponding spiro[indane-1,4'-
piperidine] analog 19 was much decreased. By conversion of its tetrahydroisoquinoline portion into
other benzo-fused (hetero) alicyclic portions, the less antagonist activity of compound 19 was some-
degree improved but not enough to reach the level of compound 5 in the case of compound 33, and
rather weakened in the case of compound 38, although the binding potencies were some-degree
improved in the both cases. These results indicated strong contribution of the character of 4-(4-fluoro-2-
methylphenyl)piperidine as 4-arylpiperidine portion for the potent antagonist activity of compound 5.
(ii) Compared with compound 5 (or compound 19), compound 15 (or compound 23) exhibited much
increased antagonist activity, relative to the change of binding affinities between them, indicating strong
contribution of chloro group and hydroxy group, that are appropriately located on the aryl moiety of
tetrahydroisoquinoline portion, for the activity. (iii) The presence/absence and the nature of (hetero)
alicyclic moiety within the bicyclo portion can be geared to evoke or lose antagonist activity, in addition
to the restriction/rigidity effect of location and orientation for the aryl ring moiety within bicyclo portion
as mentioned. As supporting SAR information of other synthetic NOP receptor antagonists in the
previous studies (see also Scheme 8), (i) the binding affinities and the antagonist activities of the
derivatives of N,N-dimethylindoline-2-carboxamide analog 39 were dependent on various structural
features of respective 4-phenylpiperidine portions [6]. (ii) The affinities and the activities of the 4-
arylpiperidine analogs 40 were dependent on various structural features of respective cyclic groups at
propyl portion [7]. (iii) Discrepancies between the potencies of binding affinities and antagonist
1
3

ACCEPTED MANUSCRIPT
activities have been found for some types of compounds. The discrepancies in the present and the
previous studies might be dependent on whether the ligand binding-induced physically or functionally
masking or interfering with functionally important points/regions of the receptor protein were effectual
or not [6,7].
Recently, an independent study with X-ray crystallography for complex of NOP receptor and an
antagonist C-24 (1-benzyl-N-{3-[spiroisobenzofuran-1(3H),4'-piperidin-1-yl]propyl}pyrrolidine-2-
carboxamide) was reported [32]. In the study, the binding site of the ligand was deep within seven-
transmembrane helical core of the receptor; the spiro[isobenzofuran-1,4'-piperidine] portion was buried
in a hydrophobic cavity; and the isobenzofuran group was interposed between Met134 and Tyr131.
Furthermore, the protonated nitrogen of piperidine ring of the ligand formed a salt bridge with Asp130
that is an important point required for both N/OFQ and NOP receptor antagonists [32]. The present
antagonists, compounds 5 and 15 (as well as compounds 19 and 23) might share similar orthosteric
binding sites in the transmembrane core owing to some structural similarities of the arylpiperidine
portions between the present compounds and the ligand, although the other portions are quite different.
Assuming that the arylpiperidine portions as respective scaffolds for the present analogs and the ligand
bind at almost the same site within the binding pocket and have similar interaction modes (location and
orientation), one of pharmacophore profiles of the present analogs might also be explicable. Thus, the
protonatable nitrogen of the piperidine ring of the respective present analogs might work for anchoring
salt bridge with Asp130 as a critical interaction, similar to the above manner. Notably, compound 18
that is amide analog of the pyridine ring did not show the binding affinity, which might explain
indispensability of the protonatable amino-moiety of the present analogs for their interactions with the
receptor protein, as well.
Taken together, as significant findings of the present studies, (i) the newly designed compound 15 was
discovered as a novel small-molecule nonpeptide hNOP receptor antagonist that demonstrated highly
potent hNOP receptor binding affinity, high hNOP receptor binding selectivity against human µ receptor,
and highly potent hNOP receptor antagonist activity in vitro. (ii) The whole and multisite structural or
1
4

ACCEPTED MANUSCRIPT
chemical characters such as size, shape, volume, flexibility/rigidity, atom species, functional group, and
electronic state of the present small-molecule analogs were significant or crucial to determine their
profiles in vitro, which would be associated with the multisite interactions of the ligands to the receptor.
Especially, the characteristic portions were indispensable for compound 15 to be a superior small-
molecule hNOP receptor antagonist, respectively. The significant results, information, and the present
analogs per se would be useful to conduct further NOP receptor antagonist studies and might be helpful
to elucidate dynamic mechanisms of a variety of biological effects mediated by NOP receptor–ligand
system, together with the results and the significant analogs per se of the previous studies for the
distinctive series of hNOP receptor agonists [3–5] and antagonists [6,7].
2
.3. Design and SMR studies of metabolic stability in vitro
Alternatively, the design and SMR of the present analogs were investigated as a pharmacokinetic
study in vitro. The half-lives of the compounds were evaluated in human liver microsome (HLM). The
in vitro SMR of the analogs is shown in Table 2.
Significantly, the above 4-(4-fluoro-2-methylphenyl)piperidine unit was designed in the previous
detailed SAR and SMR investigations with multiviewpoint and integrated drug-design strategy for N,N-
dimethylindoline-2-carboxamide analogs [6], expecting to get compatibility of potent and selective
hNOP receptor antagonist activity and metabolic stability. Owing to a successful result in the study, the
unit was also thought to be effectual for the present study. Indeed, 4-(4-fluoro-2-
methylphenyl)piperidine analogs of the tetrahydroisoquinoline derivatives were investigated. As a
significant result, 5-chloro-tetrahydroisoquinolin-8-ol analog 15 was more metabolically stable (21.6
min) than tetrahydroisoquinoline analog 5. When the arylpiperidine portion was substituted by
spiro[indane-1,4'-piperidine], 5-chloro-tetrahydroisoquinolin-8-ol 23 was also more stable than the
corresponding tetrahydroisoquinoline analog 19. As well, SMR study around tetrahydroisoquinoline
portion of 1'-[(tetrahydroisoquinolin-3-yl)methyl]-spiro[indane-1,4'-piperidine] analogs was investigated.
1
5

ACCEPTED MANUSCRIPT
Relative to 5-chloro-tetrahydroisoquinolin-8-ol analog 23, tetrahydroisoquinolin-8-ol analog 27 and
tetrahydroisoquinolin-6-ol analog 30 exhibited comparative stabilities. 5-Bromo-tetrahydroisoquinolin-
8
-ol analog 25 showed shorter half-life than that of compound 23. Generally, phenolic analogs 15, 23,
5, 27, and 30 were more metabolically stable compared to the corresponding unsubstituted analogs 5
2
and 19. Besides, chromane analog 33 was less stable than tetrahydroisoquinoline analog 19. As
supporting data for the above SMR of 1'-(substituted alkyl) portion of spiro[indane-1,4'-piperidine]
analogs, compound 19 was more stable than 1'-(3-phenylpropyl)-spiro[indane-1,4'-piperidine] (3.0 min)
of which the monocyclic phenyl portion was very weak against metabolic reaction in HLM as described
in the previous study [7].
It is significant to consider contributing-factor or -effect for enhancing/improving metabolic stability in
the viewpoint of structural features or chemical properties, with the results of the previous SMR studies
for a variety of compounds in HLM [3,5–7]. First, in the previous studies, the structural differences of
the 4-arylpiperidine portions influenced respective stabilities of compounds. As reported, 4-(4-fluoro-2-
methylphenyl)piperidine was much preferable to 4-(2-methylphenyl)piperidine as 4-arylpiperidine
portion of N,N-dimethylindoline-2-carboxamide analogs; and the 4-fluoro group effect on the aryl
moiety for enhancing stability in the comparison was in accord with SMR of other analogs in the study
as well as a result of major-metabolite identification study for a related compound, thus 4-position of its
aryl moiety was hydroxylated in liver microsome [6]. In another study, 4-(2-methylphenyl)piperidine
was preferable to spiro[indane-1,4'-piperidine] as 4-arylpiperidine portion of (piperidin-1-yl)propyl
analogs [7]. The preference of 4-(4-fluoro-2-methylphenyl)piperidine for compound 15 over
spiro[indane-1,4'-piperidine] for compound 23 in the present study was in line with the results of those
studies, thus the differences of metabolic stabilities between compound 15 and compound 23 were
attributable to their 4-arylpiperidine portions as well. Second, as the SMR for the opposite part of 4-
arylpiperidine portion of the present compounds, the structural differences of the benzo-fused bicyclic
portion variously affected their respective stabilities as described (for example, compound 15 >
compound 5). As substituents on propyl group for 1'-propyl-spiro[indene-1,4'-piperidine] analogs, γ-
1
6

ACCEPTED MANUSCRIPT
hydroxypiperidine and 2-piperidone were much preferable to piperidine, while cyclohexane and benzene
were much less favorable over piperidine [7]. These results have indicated that the structural features
and/or chemical properties of both 4-arylpiperidine portion and the opposite part would contribute to the
overall metabolic stability or lability of the respective compounds. Noteworthy, the respective structural
features and/or chemical properties of specific moiety or portion have been more apparently responsible
for the differences of metabolic stabilities between analogs in the comparisons on condition that the
features and/or the properties of the other moiety or portion of the analogs were similar. Assuming that
the unsubstituted tetrahydroisoquinoline portions of compounds 5 and 19 are important sites involved in
the metabolically labilities of the compounds, the various results of the SMR study around bicyclo
portion of the present analogs might show further potential contributing-factors/-effects in terms of
metabolic stabilities of the analogs; although identification study of possible metabolites of compound 5
or its analogs would also be important or helpful to understand them. Thus, more favorable characters of
secondary-amino alicyclic group and phenolic group for metabolic stability in comparison of the analogs
might be due to or contributed by hydrogen bond donor (HBD) functionality (or reduced local
lipophilicity) of them, compared to the corresponding oxacyclic group and des-hydroxy-phenyl group
analogs that have no HBD functionality, respectively. Supportingly, HBD functionality effects [or
reduced (local) lipophilicity effects attributable to (local) structural differences] for getting/enhancing
metabolic stability in HLM can be seen for (i) plasma-to-brain selective HPCOM, that is NH-
hexahydropyrrolo[3,4-c]pyrrole analog of (hydroxymethyl)cyclooctyl-type benzimidazole derivative,
against the corresponding N-Me-piperazine analog [3], for (ii) brain selective MCOPPB, that is NH-
piperidine analog of methylcyclooctyl-type benzimidazole derivative, against the corresponding N-Me-
piperidine analog [5], and for (iii) (β- or γ-)hydroxypiperidine analogs of 1-(3'-substituted propyl)-4-
arylpiperidine derivatives against the corresponding unsubstituted piperidine analog [7], respectively.
Furthermore, introduction of hydroxy group onto phenyl moiety within tetrahydroisoquinoline portion
for the present analogs heightens oxidation-state and acidity of the moiety, which might potentially
increase resistivity of the moiety against oxidation by major metabolic enzymes, cytochrome P450s
1
7

ACCEPTED MANUSCRIPT
(CYPs) in HLM. Electron withdrawing characters of both 4-fluoro group of arylpiperidine portion and
5
-chloro group of tetrahydroisoquinoline portion of compound 15 would affect π-electron density
(distribution) and polarizability of the respective aryl cores, and might affect resistivity against
electrophilic oxidative-reaction by CYPs. Besides, appropriately located substituent-effects for
metabolic stability in this study might also be associated with their blocking effects against potential
metabolic reactions on the respective sites/portions, similar to the potential blocking effect produced by
additional substituent for the arylpiperidine analogs in the previous study [6].
In conclusion, (i) compound 15 was most preferable among the present analogs in the viewpoint of the
compatibility of metabolic stability in HLM and profiles as highly potent and highly selective hNOP
receptor antagonist in vitro. (ii) The results of this study suggested significant or potential contributing-
factors/-effects for enhancing/improving metabolic stability in comparison of the present analogs, which
might be beneficial or helpful to conduct further or other drug-discovery studies of metabolically stable
drugs.
2
.4. Design and SAR studies of hERG ion channel binding in vitro for drug safety
The bindings of (candidate) clinical drugs to hERG potassium ion channel are serious concerns in the
pharmacotherapies. The concerns relate to prolongation of QT interval in the electrocardiogram in vivo,
which may lead to Torsades de Pointes in the human heart or sudden death at worst. Therefore, the
assessments of the binding affinities to the channel for drugs are very important in drug-discovery
strategies toward clinical medications. Indeed, various types of compounds such as HPCOM [3],
MCOPPB [5], and other compounds [7], that have little potential risks of the hERG potassium ion
channel binding issues, have been discovered in the respective safe-drug studies.
As well, the design and SAR of the present analogs for reducing hERG potassium ion channel binding
affinity were investigated in vitro. The inhibitory activities of the analogs (IC values) were determined
5
0
1
8

ACCEPTED MANUSCRIPT
3
against [ H]dofetilide binding to potassium ion channel from hERG expressed in HEK-293 cells
[
3,5,7,25]. The in vitro SAR of the analogs is shown in Table 2.
In comparison of 4-(4-fluoro-2-methylphenyl)piperidine analogs, 5-chloro-tetrahydroisoquinolin-8-ol
analog 15 exhibited less potent binding affinity than tetrahydroisoquinoline analog 5. Besides, in
comparison of spiro[indane-1,4'-piperidine] analogs, tetrahydroisoquinoline analog 19 and 5-
halogenated tetrahydroisoquinolin-8-ol analogs 23 and 25 exhibited less potent binding affinities than
tetrahydroisoquinolin-6-ol analog 27 and tetrahydroisoquinolin-8-ol analog 30. Compared with
substituent effects on the aryl moiety within tetrahydroisoquinoline portion for spiro[indane-1,4'-
piperidine] analogs, the rank of the orders of the binding affinities was 5-chloro-tetrahydroisoquinolin-8-
ol 23 < 5-bromo-tetrahydroisoquinolin-8-ol 25 < tetrahydroisoquinolin-8-ol 27 < tetrahydroisoquinolin-
6
-ol 30. These results indicated that (i) the orders of the reducing affinity effects of 5-substituents for the
respective tetrahydroisoquinolin-8-ol analogs was in accord with the order of the electron negativities of
the respective 5-substituents, that is, chloro group (23) > bromo group (25) > hydrogen group (27). By
contrast, (ii) phenolic hydroxy groups at 8- or 6-position (27 or 30), that have π-electron donating
character for the respective aromatic-ring moieties, showed increasing affinity effects compared with
hydrogen group (19). In addition, the phenolic hydroxy group effects of compounds 27 and 30 displayed
regioselectivities for their affinities, that is, 6-OH group (30) > 8-OH group (27). As supporting data in
the previous study in vitro in terms of the electron effects, MCOPPB that has NH-piperidine (electron-
poor ring) as 2-substituent on its benzimidazole moiety showed little hERG ion channel binding affinity,
contrary to the corresponding 2-aryl group (electron-rich ring) analogs that showed higher binding
affinities [5]. In comparison of (hetero) alicyclic moiety within the bicyclo portion of the present analogs
in Table 2, N-Me-1-oxo (lactam) analog 21 was very successful to achieve much lower binding affinity
(IC50 = 6,960 nM); however N-Ac analog 22 showed higher affinity. Compared to the other groups for
the moiety, the preference of the lactam group for reducing affinity might be associated with its
character, that is, electron withdrawing effect of the group for reducing electron density of the adjacent
1
9

ACCEPTED MANUSCRIPT
aryl moiety, electron distribution effect or polarity effect for the hetero alicyclic moiety (that is, lactam
group itself) or for the aryl moiety affected by the group, and/or position-dependent steric effect of the
amido group, against the ligand binding to the channel. Fuerthermore, local lipophilicity around the
(hetero) alicyclic moiety might contribute to the binding affinity. Thus, tetrahydronaphthalene analog 38
that has higher lipophilic carbocyclic moiety showed higher affinity relative to tetrahydroisoquinoline
analog 19 that has lower lipophilic secondary-amino alicyclic moiety. The possible lower lipophilicity
effect for reducing affinity was in line with previously described lower lipophilicity effects for reducing
in vitro hERG ion channel binding affinity of (i) 1,2-substituted benzimidazole analogs (for example,
MCOPPB that is NH-piperidine analog against the corresponding aryl analogs, see also the above
discussion) [5] and of (ii) 1-(azacyclic or carbocyclic group substituted-propyl)-4-arylpiperidine analogs
(
for example, piperidine analog had much less affinity than cyclohexane analog) [7] (for further results
of the detailed SAR studies, see the original reports [5,7], respectively). Interestingly, oxacyclic analog
3 was preferable to carbocyclic analog 38 for reducing affinity. The oxacyclic group effect of
compound 33 against hERG ion channel binding might be associated with -electron withdrawing effect
3
σ
of the ether oxygen atom for reducing electron density of the adjacent aryl moiety, lower lipophilic
effect that is similar to the above examples, and/or electron lone-pair effect of the oxygen atom,
compared to the carbocyclic moiety of compound 38.
Overall, i) several structures were found to be effectual for reducing hERG potassium ion channel
binding affinity in this study. ii) As significant or potential contributing-factors for reducing affinity,
various whole/local structural features and/or chemical properties, such as shape, electron state, polarity,
atom species, functional group, and position of substituent/group, were suggested in comparison of the
present analogs, which might be associated with interaction levels between the compounds and hERG
ion channel. The results and findings might provide useful or helpful information to conduct further or
other drug-discovery studies for safe drugs, together with the results of the previous studies for diverse
structural types of compounds [3,5,7].
2
0

ACCEPTED MANUSCRIPT
3
. Conclusions
In the present study, the design, synthesis, and SAR of a series of (4-arylpiperidine substituted-
methyl)-[bicyclic (hetero)cycloalkanobenzene] analogs were investigated in vitro in order to seek and
identify potent and selective, novel small-molecules of hNOP receptor antagonists. Among them,
compounds 5 and 15 were discovered as novel nonpeptide hNOP receptor antagonists that demonstrated
highly potent hNOP receptor binding affinities, high hNOP receptor binding selectivities against human
µ receptor, and highly potent hNOP receptor antagonist activities in vitro. In the design and SMR
studies in HLM, compound 15 exhibited more metabolic stability than compound 5. The SAR and SMR
studies indicated or suggested various whole and local features as key and/or useful contributing-
factors/-effects to achieve or improve the profiles. In the design and SAR studies of the analogs against
hERG ion channel binding affinity in vitro, several effective structures and significant and/or useful
contributing-factors for reducing affinities were found or suggested. The above results and findings
including the significant small-molecule analogs per se might provide useful or helpful
information/tools to conduct further studies of hNOP receptor antagonists, elucidation studies of NOP
receptor–N/OFQ system-mediated pharmacological/functional/transductional mechanisms, and other
various studies for metabolically stable and safe drugs.
4
4
4
. Experimental section
.1. Chemistry
.1.1. General procedures
2
1

ACCEPTED MANUSCRIPT
In general, reagents, solvents, and other chemicals were used as purchased without further
purification. All reactions with air- or moisture-sensitive reactants and solvents were carried out under
nitrogen atmosphere. Flash column chromatography (medium pressure liquid chromatography)
purifications were carried out using Merck silica gel 60 (230–400 mesh ASTM). Preparative thin-layer
chromatography (PTLC) purifications were carried out on Merck silica gel 60 F254 precoated glass plates
at a thickness of 0.5 or 1.0 mm. The structures of all isolated compounds were assured by the following
1
1
techniques, such as NMR, IR, MS or elementary analysis. H nuclear magnetic resonance ( H NMR)
data were determined at 270 MHz on a JNM-LA 270 (JEOL) spectrometer and at 300 MHz on a JNM-
1
LA300 (JEOL) spectrometer. Chemical shifts are expressed in
δ
(ppm). H NMR chemical shifts were
determined relative to tetramethylsilane (TMS) as internal standard. NMR data are reported as follows:
chemical shift, number of atoms, multiplicities (s, singlet; d, doublet; t, triplet; dd, double doublet; m,
multiplet; and br, broadened), and coupling constants. Infrared spectra were measured by an IR-470
(Shimadzu) infrared spectrometer. Low-resolution mass spectral data (EI) were obtained on an
Automass 120 (JEOL) mass spectrometer. Low-resolution mass spectral data (ESI) were obtained on a
Quattro II (Micromass) mass spectrometer–Agilent 1100 HPLC system. Melting point was obtained
using Exstar 6000 (Seiko Instruments Inc.) and was uncorrected.
After preparation by manual synthesis, compounds 5, 15, 18–23, 25, 27, 30, and 33 were converted
into salts of citrate, hydrochloride or 4-tosylate, respectively, as shown in the subsections for the
compounds, respectively. The purities of citrates of the compounds 5, 15, 19, 23, 25, 27, and 30, a
hydrochloride of compound 22, and a 4-toluenesulfonate of compound 33 were confirmed by elementary
analysis to be within ±0.4% of calculated values, respectively. Elementary analysis of hydrochloride
salts of compounds 18, 20, and 21 was not performed owing to the requisite amounts for the analysis of
1
the respective salts; however the purity of salt-free compounds 18, 20, and 21 were assured by H NMR
analysis that had no significant impurity signal except small amounts of solvent signals. Compound 38
prepared by high-speed parallel synthesis was obtained as a formate salt. These salts of the compounds
were used for pharmacological and/or pharmacokinetic evaluations.
2
2

ACCEPTED MANUSCRIPT
4
.1.2.
tetrahydroisoquinoline (5).
.1.2.1. tert-Butyl
Synthesis
of
3-{[4-(4-fluoro-2-methylphenyl)piperidin-1-yl]methyl}-1,2,3,4-
4
3-{[4-(4-fluoro-2-methylphenyl)piperidin-1-yl]carbonyl}-3,4-
dihydroisoquinoline-2(1H)-carboxylate (3). To a suspension of 1,2,3,4-tetrahydroisoquinoline-3-
carboxylic acid hydrochloride (2.137 g, 10.0 mmol) in H O (54.0 mL)–tert-butyl alcohol (45.0 mL) was
2
added 2 N NaOH (11.0 mL, 22.0 mmol) at room temperature under N . Di-tert-butyl dicarbonate
2
(
Boc O) (2.20 g, 10.1 mmol) was added to the resulting solution at a time, stirred at room temperature
2
under N for 16 h [4,20]. The reaction solution was poured into 10% aqueous citric acid at room
2
temperature, then the resulting acidic mixture was extracted with chloroform three times. The combined
extracts were washed with H
.5196 g of 2-(tert-butoxycarbonyl)-1,2,3,4-tetrahydroisoquinoline-3-carboxylic acid 2 [20] as a slight
yellowish-white solid (crude).
2 4
O, dried over anhydrous MgSO , and concentrated in vacuo to afford
3
A mixture of 4-(4-fluoro-2-methylphenyl)piperidine 1 (386.5 mg, 2.00 mmol) [6], 2-(tert-
butoxycarbonyl)-1,2,3,4-tetrahydroisoquinoline-3-carboxylic acid 2 (554.6 mg, 2.00 mmol), water
soluble carbodiimide, that is, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (WSCI)
(
766.8 mg, 4.00 mmol), hydroxybenzotriazole (HOBT) (540.5 mg, 4.00 mmol), and dry Et N (836 µL,
3
6
.00 mmol) in dry CH
reaction solution was poured into H O (30 mL), and the resulting mixture was extracted with CH Cl
2
2 2 2
Cl (16.0 mL) was stirred at room temperature under N for 24 h [7]. The
2
2
(
30 mL × 2). The combined extracts were dried over anhydrous MgSO , and concentrated in vacuo. The
4
residue was purified by column chromatography (silica gel, hexane/AcOEt = 3:1) to afford 833.5 mg of
1
the title product 3 in 92% yield as a colorless oil. H NMR (300 MHz, CDCl
3
)
δ
7.27–7.00 (5H, m),
6
9
.90–6.85 (2H, m), 5.40–4.16 (5H, m), 3.25–2.60 (5H, m), 2.35 (3H, s), 2.00–1.37 (13H, m, including
+
H, s, at 1.49 ppm). MS (ESI positive) m/z: [M+H] 453.
2
3

ACCEPTED MANUSCRIPT
4.1.2.2. 3-{[4-(4-Fluoro-2-methylphenyl)piperidin-1-yl]carbonyl}-1,2,3,4-tetrahydroisoquinoline
(4). A mixture of compound 3 (833.5 mg, 1.84 mmol) in 10% HCl in MeOH (80.0 mL) was stirred at
room temperature under N for 40 h, then concentrated in vacuo. The residue was partitioned between
2
CH Cl (60.0 mL) and saturated aqueous NaHCO (100 mL). The organic layer was separted and the
2
2
3
aqueous layer was extracted with CH Cl (60 mL × 2). The organic layers were combined, dried over
2
2
1
anhydrous MgSO
4
, and concentrated in vacuo to afford 603.3 mg of the title product 4 in 93% yield. H
NMR (300 MHz, CDCl3)
δ
7.20–7.02 (5H, m), 6.92–6.83 (2H, m), 4.86 (1H, d, J = 12.5 Hz), 4.13 (3H,
m), 3.99 (1H, dd, J = 11.0 Hz, J = 4.23 Hz), 3.24–2.66 (5H, m), 2.35 (3H, s), 1.94–1.50 (5H, m). MS
+
(
ESI positive) m/z: [M+H] 353.
4.1.2.3.
3-{[4-(4-Fluoro-2-methylphenyl)piperidin-1-yl]methyl}-1,2,3,4-tetrahydroisoquinoline
(5). To a solution of compound 4 (603.3 mg, 1.71 mmol) in anhydrous THF (15.0 mL) was added
lithium aluminum hydride (LAH) (194.9 mg, 5.14 mmol) at 0 °C under N [7]. The reaction mixture
2
was stirred at the temperature under N for 1 h, allowed to room temperature, and stirred under N for 24
2
2
h. The reaction mixture was cooled to 0 °C, AcOEt (80 mL) was added dropwise, stirred at the same
temperature under N for 20 min, allowed to warm to room temperature, and stirred for 40 min. The
2
mixture was cooled to 0 °C, ice-cooled H O (80 mL) was added dropwise, then the resulting mixture
2
was stirred at the temperature for 1.5 h. The organic layer was separated, and the aqueous layer was
extracted with AcOEt (80 mL × 2). The organic layers were combined, dried over anhydrous MgSO4,
and concentrated in vacuo. The residue was purified by column chromatography (silica gel,
CH Cl /MeOH = 14:1), followed by PTLC twice (silica gel, CH Cl /MeOH = 14:11 for the first,
2
2
2
2
CH Cl /AcOEt/2-propanol/25% aqueous ammonia = 16:40:4:0.4 for the second) was performed to
2
2
+
1
afford 130.2 mg of the title product 5 in 22% yield. MS (ESI positive) m/z: [M+H] 339.24. H NMR
(300 MHz, CDCl3)
δ 7.22–7.03 (5H, m), 6.90–6.83 (2H, m), 4.10 (2H, s), 3.15–2.96 (3H, m), 2.76–1.70
(15H, m, including 3H, s, at 2.32 ppm). Citric acid salt formation, general procedure [6,7]:
2
4

ACCEPTED MANUSCRIPT
Compound 5 (122.5 mg, 0.362 mmol) and one equivalent of citric acid (69.5 mg, 0.362 mmol) was
dissolved in dry CH
2 2
Cl (8.0 mL) and dry MeOH (12.0 mL), and the resulting solution was stirred at
room temperature for 2 h, then concentrated in vacuo. The residue was resuspended in CH Cl , collected
2
2
by filtration, and dried under vacuum at 50 °C for 5 h to afford citrate salt as a white solid. Monocitrate
of compound 5: Mp: 185.7 °C. IR (KBr): 3430, 2932, 1717, 1589, 1499, 1387, 1246, 1188, 972, 864,
–
1
8
20, 752, 613 cm . Anal. (C H FN ·C H O ) C, H, N.
22 27 2 6 8 7
4
.1.3. Synthesis of 5-chloro-3-{[4-(4-fluoro-2-methylphenyl)piperidin-1-yl]methyl}-1,2,3,4-
tetrahydroisoquinolin-8-ol (15).
.1.3.1. Diethyl (acetylamino)(5-{[tert-butyl(dimethyl)silyl]oxy}-2-chlorobenzyl)malonate (8). To
stirred a solution of diethyl acetamidomalonate (15.9 g, 73.2 mmol) in dry DMF (80.0 mL) was added
NaH (60% in oil, 2.92 g, 73.0 mmol) at 0 °C under N . The reaction mixture was stirred at room
temperature under N for 30 min, then the resulting yellow solution was cooled to 0 °C. A solution of [3-
bromomethyl)-4-chlorophenoxy](tert-butyl)dimethylsilane 7, that was prepared according to the
reported method [21] (24.5 g, 73.0 mmol), in dry DMF (20.0 mL) was added to the above solution at 0
C under N . The reaction solution was allowed to warm to room temperature, stirred under N for 2 h,
4
2
2
(
°
2
2
then H O (200 mL) was added. The resulting mixture was extracted with AcOEt twice. The combined
2
extracts were washed with H O twice and then brine, dried over anhydrous MgSO , and concentrated in
2
4
vacuo. The residue was purified by column chromatography (silica gel, hexane/AcOEt = 2:1) to afford
1
1
7
4
8.45 g of the title product 8 as an oil in 54% yield from compound 7. H NMR (300 MHz, CDCl
3
)
δ
.16 (1H, d, J = 8.6 Hz), 6.66 (1H, dd, J = 8.6 Hz, J = 2.8 Hz), 6.57 (1H, d, J = 2.8 Hz), 6.54 (1H, br s),
.37–4.07 (4H, m), 3.78 and 3.74 (total 2H, each s), 2.05 and 2.01 (total 3H, each s), 1.29 (6H, t, J = 8.8
Hz), 0.96 (9H, s), 0.17 (6H, s).
4.1.3.2. 2-(tert-Butoxycarbonyl)-5-chloro-8-hydroxy-1,2,3,4-tetrahydroisequinoline-3-carboxylic
acid (11). A mixture of compound 8 (18.45 g, 39.1 mmol), concentrated HCl (70.0 mL), and acetic acid
2
5

ACCEPTED MANUSCRIPT
2
(30.0 mL) was warmed to reflux conditions, stirred under N for 14 h, then more concentrated HCl (40
mL) was added. The reaction mixture was stirred under reflux conditions for 10 h, and then more
concentrated HCl (10 mL) was added. The reaction mixture was stirred under reflux conditions for 16 h,
then cooled to room temperature. The reaction mixture was diluted with H O (50 mL), washed with
2
CH Cl twice, and then concentrated in vacuo to afford 10.2 g of 2-chloro-5-hydroxyphenylalanine
2
2
hydrochloride 9 as a white solid (crude).
A mixture of compound 9 (10.2 g, crude), 37% formalin (40.0 mL), and H O (40.0 mL) was warmed
2
to 90 °C, and stirred under N for 1.5 h, then cooled to 0 °C [22]. The resulting white precipitate was
2
collected by filtration, and dried under vacuum to afford 7.88 g of 5-chloro-8-hydroxy-1,2,3,4-
tetrahydroisoquinoline-3-carboxylic acid hydrochloride 10 in 76% yield as a solid. MS (ESI negative)
–
m/z: [M–H] 226.02.
To a stirred mixture of compound 10 (7.88 g, 29.8 mmol) and 2 N NaOH (75.0 mL, 150 mmol) in 1,4-
dioxane (80.0 mL) was added Boc O (7.81 g, 35.8 mmol) at room temperature under N . The reaction
2
2
mixture was stirred at the temperature for 18 h, then concentrated in vacuo. The residue was dissolved in
H O (60.0 mL), then MeOH (30.0 mL) was added followed by 2 N NaOH (20.0 mL). The mixture was
2
stirred at room temperature for 6 h, then concentrated in vacuo. The residue was acidified by adding 2 N
HCl and 10% aqueous citric acid, and then extracted with CH Cl –MeOH (10:1) three times. The
2
2
combined extracts were concentrated in vacuo to afford 8.0 g of the title product 11 as a brown solid
crude), containing a by-product 12 that would plausibly be N,O-di-Boc derivative of compound 11, that
(
is,
2-(tert-butoxycarbonyl)-8-(tert-butoxycarbonyloxy)-5-chloro-1,2,3,4-tetrahydroisoquinoline-3-
–
carboxylic acid, as detected by mass spectroscopy: MS (ESI negative) m/z: [M–H] 426 (not isolated).
1
Compound 11: H NMR (300 MHz, DMSO-d
6
) 10.00 (1H, s), 7.44 and 7.13 (total 1H, each d, J = 8.7
Hz and J = 8.6 Hz), 7.18 and 6.72 (total 1H, each d, J = 8.8 Hz), 5.08–5.00 (0.5H, m), 4.90–4.80 (0.5H,
m), 4.50–4.35 (1H, m), 4.30–4.15 (1H, m), 3.50–2.80 (2H, m), 1.50, 1.46, and 1.41 (total 9H, each s).
–
MS (ESI negative) m/z: [M–H] 326.
2
6

ACCEPTED MANUSCRIPT
4.1.3.3. tert-Butyl 5-chloro-3-{[4-(4-fluoro-2-methylphenyl)piperidin-1-yl]carbonyl}-8-hydroxy-
3
,4-dihydroisoquinoline-2(1H)-carboxylate (13). A mixture of compound 1 (162.3 mg, 0.840 mmol),
compound 11 (250.0 mg, crude), WSCI (322.1 mg, 1.68 mmol), HOBT (227.0 mg, 1.68 mmol), and dry
Et N (351 µL, 2.52 mmol) in dry CH Cl (7.0 mL) was stirred at room temperature under N for 36 h,
3
2
2
2
then poured into H O (50 mL). The resulting mixture was extracted with CH Cl (50 mL × 2). The
2
2
2
combined extracts were dried over anhydrous MgSO , and concentrated in vacuo. The residue was
4
purified by column chromatography (silica gel, hexane/AcOEt = 2:1) to afford 218.9 mg of the title
product 13 in 52% yield from compound 1. TLC: R =0.35, AcOEt/hexane = 1:1. MS (ESI positive) m/z:
f
+
[
M+H] 501.
4.1.3.4.
5-Chloro-3-{[4-(4-fluoro-2-methylphenyl)piperidin-1-yl]carbonyl}-1,2,3,4-
tetrahydroisoquinolin-8-ol (14). To a stirred solution of compound 13 (218.9 mg, 0.435 mmol) in
anhydrous THF (2.0 mL) was added trifluoroacetic acid (4.0 mL) at 0 °C under N . The reaction
2
solution was stirred at the temperature for 1 h, allowed to room temperature, stirred overnight, then
concentrated in vacuo. The residue was partitioned between CH Cl (20 mL) and saturated aqueous
2
2
3 2 2
NaHCO (50 mL). The organic layer was separated, and the aqueous layer was extracted with CH Cl
(
50 mL × 4). The organic layers were combined, dried over anhydrous MgSO , and concentrated in
4
vacuo. The residue was purified by PTLC twice (silica gel, CH Cl /MeOH = 15:1 for each) to afford
2
2
1
7
3
6.3 mg of the title product 14 in 44% yield. H NMR (270 MHz, CDCl ) δ 7.20–6.76 (4H, m), 6.65
(1H, d, J = 8.24 Hz), 4.84 (1H, d, J = 13.0 Hz), 4.22–3.78 (4H, m), 3.30–3.15 (1H, m), 2.97–2.90 (2H,
+
m), 2.78–2.58 (2H, m), 2.35–2.33 (3H, m), 1.90–1.50 (4H, m). MS (ESI positive) m/z: [M+H] 403.
4.1.3.5.
5-Chloro-3-{[4-(4-fluoro-2-methylphenyl)piperidin-1-yl]methyl}-1,2,3,4-
tetrahydroisoquinolin-8-ol (15), general procedure of borane reduction of amide for amine
synsthesis. To a solution of compound 14 (76.3 mg, 0.189 mmol) in anhydrous THF (3.7 mL) was
2
7

ACCEPTED MANUSCRIPT
added a solution of BH3
·
SMe (1.0 M in THF, 90 µL, 90 µmol) at room temperature under N . The
2
2
reaction solution was warmed up to reflux conditions, stirred under N for 5 h, cooled to room
2
temperature, then more BH
solution was warmed up to reflux conditions, stirred for 24 h, then cooled to room temperature, then
more BH3 SMe (1.0 M in THF, 90 µL, 90 µmol) was added. The reaction solution warmed up to reflux
3 2
·SMe (1.0 M in THF, 90 µL, 90 µmol) was added. Again, the reaction
·
2
conditions, stirred for 16 h, cooled to 0 °C, then 6 N HCl (1.26 mL), 2 N HCl (1.0 mL), and MeOH (3.7
mL) were added dropwise. The resulting mixture was warmed up to reflux conditions, stirred under N2
for 24 h, then allowed to room temperature. The resulting solution was diluted with MeOH (10 mL),
then concentrated in vacuo. For the residue, a procedure, that consists of the addition of MeOH (5 mL)
followed by the concentration of the resulting mixture in vacuo, was repeated six times. The residue was
dissolved in CH Cl (25 mL)–MeOH (3 mL), then poured into saturated aqueous NaHCO (15 mL)–
2
2
3
H
2
O (10 mL). The organic layer was separated, and the aqueous layer was extracted with CH
2 2
Cl –
MeOH (10:1, 25 mL × 2), and with CH Cl (25 mL × 2). The organic layers were combined, dried over
2
2
anhydrous MgSO , and concentrated in vacuo. The residue was purified by column chromatography
4
(
silica gel, CH Cl /MeOH = 15:1, then CH Cl /MeOH/25% aqueous ammonia = 15:1:0.15) followed by
2 2 2 2
2 2
PTLC (silica gel, CH Cl /MeOH/25% aqueous ammonia = 18:1:0.1 × 2, that is, twice developments on
the same plate) was performed to afford 26.3 mg of the title product 15 in 36% yield as a slight
1
yellowish-white solid. H NMR (300 MHz, CDCl )
δ 7.19–7.14 (1H, m), 6.99 (1H, d, J = 8.40 Hz),
3
6
1
.90–6.84 (2H, m), 6.47 (1H, d, J = 8.43 Hz), 4.30 (2H, br s), 4.14 (1H, d, J = 15.9 Hz), 3.76 (1H, d, J =
+
5.9 Hz), 3.13–1.73 (17H, m, including 3H, s, at 2.32 ppm). MS (ESI positive) m/z: [M+H] 389 for
O·C ) C, H, N.
salt-free form. Monocitrate of compound 15: Anal. (C22
H
26ClFN
2
6 8 7
H O
4
.1.4. Synthesis of 1'-(1,2,3,4-tetrahydroisoquinolin-3-ylcarbonyl)-2,3-dihydro-1'H-spiro[indene-
,4'-piperidine] (18) and 1'-[(1,2,3,4-tetrahydroisoquinolin-3-yl)methyl]-2,3-dihydro-1'H-
spiro[indene-1,4'-piperidine] (19).
1
2
8