Discovery of DSP-1053, a novel benzylpiperidine derivative with potent serotonin transporter inhibitory activity and partial 5-HT1A receptor agonistic activity

Article abstract of DOI:10.1016/j.bmc.2018.02.008

We have previously shown that SMP-304, a serotonin uptake inhibitor with weak 5-HT_1A partial agonistic activity, may act under high serotonin levels as a 5-HT_1A antagonist that improves the onset of paroxetine in the rat swimming test. However, SMP-304 is mostly metabolized by CYP2D6, indicating limited efficacy among individuals and increased side effects. To reduce CYP2D6 metabolic contribution and enhance SERT/5-HT_1A binding affinity, we carried out a series of substitutions at the bromine atom in the left part of the benzene ring of SMP-304 and replaced the right part of SMP-304 with a chroman-4-one. This optimization work led to the identification of the antidepressant candidate DSP-1053 as a potent SERT inhibitor with partial 5-HT_1A receptor agonistic activity. DSP-1053 showed low CYP2D6 metabolic contribution and a robust increase in serotonin levels in the rat frontal cortex.

Full text of DOI:10.1016/j.bmc.2018.02.008

Bioorganic & Medicinal Chemistry xxx (2018) xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Discovery of DSP-1053, a novel benzylpiperidine derivative with potent
serotonin transporter inhibitory activity and partial 5-HT1A receptor
agonistic activity
⇑
Hidefumi Yoshinaga , Tomoaki Nishida, Izumi Sasaki, Taro Kato, Hitomi Oki, Kazuki Yabuuchi,
Tomohiro Toyoda
Drug Research Division, Sumitomo Dainippon Pharma CO., Ltd., 3-1-98 Kasugade-naka, Konohana-ku, Osaka 554-0022, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
We have previously shown that SMP-304, a serotonin uptake inhibitor with weak 5-HT1A partial agonis-
tic activity, may act under high serotonin levels as a 5-HT1A antagonist that improves the onset of parox-
etine in the rat swimming test. However, SMP-304 is mostly metabolized by CYP2D6, indicating limited
efficacy among individuals and increased side effects. To reduce CYP2D6 metabolic contribution and
enhance SERT/5-HT1A binding affinity, we carried out a series of substitutions at the bromine atom in
the left part of the benzene ring of SMP-304 and replaced the right part of SMP-304 with a chroman-
4-one. This optimization work led to the identification of the antidepressant candidate DSP-1053 as a
potent SERT inhibitor with partial 5-HT1A receptor agonistic activity. DSP-1053 showed low CYP2D6
metabolic contribution and a robust increase in serotonin levels in the rat frontal cortex.
Ó 2018 Published by Elsevier Ltd.
Received 2 December 2017
Revised 6 February 2018
Accepted 7 February 2018
Available online xxxx
Keywords:
Serotonin transporter
5-HT1A receptor
Antidepressant
Fast onset
Benzylpiperidine
CYP2D6
1. Introduction
WAY-100635, a highly selective 5-HT1A receptor antagonist.11
Based on these findings, it is believed that 5-HT1A auto-receptor
Selective serotonin reuptake inhibitors (SSRIs) are prescribed as
first-line drugs for the treatment of major depressive disorder
(MDD), a worldwide growing heath issue. However, the use of
SSRIs has a number of drawbacks, including delayed onset and lim-
ited efficacy.1–4 In fact, it takes 2- to 3-week therapy with SSRIs to
produce beneficial antidepressant effect. The slow onset of SSRIs is
due to an initially low firing activity of serotonin-producing neu-
rons induced by a transient increase in serotonin via blockage of
SERT. As neuron low firing activity is due to activation of the soma-
todendritic 5-HT1A auto-receptor, serotonin levels in the synaptic
cleft are increased only when this receptor is desensitized.5,6 It
has been reported that pindolol, an adjuvant 5-HT1A and b adren-
ergic receptor antagonist, accelerates the onset of SSRIs.7,8 This
acceleration is believed to be triggered by blockage, via 5-HT1A
antagonism, of negative feedback inhibition in response to
increased serotonin level.9,10 Moreover, the role of 5-HT1A receptor
in this response is supported by non-clinical evidence showing a
rapid onset in the case of combination therapy with SSRIs and
antagonists can accelerate SSRIs onset. Along this line, we kicked
off an exploratory research program aimed at finding antidepres-
sant drug candidates with both serotonin transporter (SERT) inhi-
bitory activity and 5-HT1A receptor antagonism.
We have previously identified SMP-304 as a potent ligand with
dual binding affinity for SERT and 5-HT1A receptor.12 SMP-304
inhibits SERT and has weak partial agonistic activity for 5-HT1A
.
However, under high serotonin levels, SMP-304 may act as a 5-
HT1A antagonist that improves the onset of paroxetine, a represen-
tative SSRI, in the rat swimming test. Further biological studies
revealed that SMP-304 is mostly metabolized by CYP2D6, indicat-
ing possible fluctuation in blood concentration among individuals
and therefore limited efficacy with increased side effects.13 To
reduce CYP2D6 high contribution to SMP-304 metabolism, we ini-
tiated a structural optimization program aiming at first identifica-
tion of compounds with potent dual SERT and 5-HT1A binding
activity and then characterization of selected compounds as was
the case with SMP-304.12 This approach is considered both labor-
and cost-effective and can be supported by the following evidence.
1. It is reported that SERT binding activity correlates well with
serotonin uptake inhibitory activity14–16, indicating that a binding
⇑
Corresponding author.
E-mail address: hidefumi-yoshinaga@ds-pharma.co.jp (H. Yoshinaga).
https://doi.org/10.1016/j.bmc.2018.02.008
0968-0896/Ó 2018 Published by Elsevier Ltd.
Please cite this article in press as: Yoshinaga H., et al. Bioorg. Med. Chem. (2018), https://doi.org/10.1016/j.bmc.2018.02.008

2
H. Yoshinaga et al. / Bioorganic & Medicinal Chemistry xxx (2018) xxx–xxx
These modifications provided DSP-1053 (Fig. 1) as a promising
candidate. DSP-1053, having potent SERT inhibitory activity and
5-HT1A receptor partial agonistic activity, could act as 5-HT1A
auto-receptor antagonist with robust serotonin elevating effect.
Here, we describe our optimization of SMP-304 leading to the dis-
covery of DSP-1053 as a potent serotonin uptake inhibitor with 5-
HT1A partial agonistic activity and low CYP2D6 metabolic
contribution.
2. Result and discussion
2.1. Chemistry
Fig. 1. Discovery of DSP-1053 from the optimization of SMP-304.
The target compounds were synthesized by coupling various 4-
benzyl piperidine intermediates to 2-arylethyl tosylates. Modifica-
tion of the position of the bromine atom on the left phenyl group of
SMP-304 provided the benzyl piperidine intermediates 9, 16 and
24. Treatment of the 2-bromo-3-methoxybenzyl bromide with
triphenylphosphine provided the phosphonium salt 2. Wittig reac-
tion of the phosphonium salt 2 with tert-butyl 4-oxopiperidine-1-
carboxylate afforded the intermediate 3. Selective hydrogenation
of the resulting olefin bond in 3 over rhodium on carbon without
debromination followed by deprotection with hydrochloric acid
gave the benzyl piperidine 5. Demethylation of 5 with BBr3 fol-
lowed by N-protection with Boc2O gave the phenol 7 as precursor
for the next O-alkylation step. Alkylation with 2-methoxyethyl
assay for SERT could be used as a first assay to identify serotonin
uptake inhibitors. In fact, the throughput of our SERT radio binding
assay is higher than that of our functional assay system for sero-
tonin uptake inhibition.2 5-HT1A partial agonists can act as 5-
HT1A receptor antagonist under high serotonin levels in the synap-
tic cleft induced by serotonin uptake inhibition. It would therefore
be impractical to set intrinsic activity for 5-HT1A receptor as a cri-
terion in a functional assay.17
To reduce CYP2D6 contribution to the metabolism of SMP-304
and enhance SERT and 5-HT1A binding activity, we first moved, the
bromine atom at the 6-position of the benzene ring to the 4-posi-
tion and replaced the right part of SMP-304 with a chroman-4-one.
Scheme 1. Reagents and conditions: (i) PPh3, toluene, reflux, (ii) tert-butyl 4-oxopiperidine-1-carboxylate, K2CO3, 2-PrOH, reflux, (iii) H2, 5% Rh-C, EtOAc, EtOH, 45 °C, (iv) 4 N
HCl/1,4-dioxane, AcOH, 50 °C, (v) BBr3, CH2Cl2, 0 °C, (vi) (Boc)2O, 1,4-dioxane, H2O, r.t., (vii) MeO(CH2)2Br, K2CO3, KI, DMF, 80 °C.
Scheme 2. Reagents and conditions: (i) MeO(CH2)2OH, KOtBu, NMP, 90 °C, (ii) NaBH4, BF3-Et2O, THF, r.t., (iii) MsCl, Et3N, Me3NHCl, toluene, 0 °C, then LiBr, THF, reflux, (iv)
PPh3, toluene, reflux, (v) tert-butyl 4-oxopiperidine-1-carboxylate, K2CO3, 2-PrOH, reflux, (vi) H2, 5% Rh-C, EtOAc, EtOH, 45 °C, (vii) c HCl, AcOH, 50 °C.
Please cite this article in press as: Yoshinaga H., et al. Bioorg. Med. Chem. (2018), https://doi.org/10.1016/j.bmc.2018.02.008

H. Yoshinaga et al. / Bioorganic & Medicinal Chemistry xxx (2018) xxx–xxx
3
Scheme 3. Reagents and conditions: (i) MeO(CH2)2Br, K2CO3, DMF, 80 °C, (ii) n-BuLi, n-BuMgCl, toluene, 0 °C, then DMF, (iii) NaBH4, MeOH, r.t., (iv) MsCl, Et3N, Me3NHCl,
toluene, 0 °C, then LiBr, THF, reflux, (v) PPh3, toluene, reflux, (vi) tert-butyl 4-oxopiperidine-1-carboxylate, K2CO3, 2-PrOH, reflux, (vii) H2, 5% Rh-C, EtOAc, r.t., (viii) 10% HCl/
MeOH, MeOH, r.t.
bromide transformed 7 to the intermediate 8, which was depro-
tected with hydrochloric acid to furnish the benzyl piperidine 9
(Scheme 1). SNAr reaction of 4-bromo-3-fluorobenzoic acid with
2-methoxyethanol afforded the carboxylic acid 10. Reduction of
10 with NaBH4 and BF3-Et2O provided alcohol 11, which was first
mesylated and then brominated to give the benzyl bromide 12.
Treatment of 12 in a similar manner to that of 2-bromo-3-
methoxybenzyl bromide furnished the benzyl piperidine 16
(Scheme 2). Alkylation of 3,5-dibromophenol with 2-methoxyethyl
bromide followed by monoformylation afforded the aldehyde 18.
Reduction with sodium borohydride transformed 18 to the alcohol
19. Treatment of 19 in a similar manner to that of alcohol 11
yielded the benzyl piperidine 24 (Scheme 3). The target com-
pounds 27–31 were synthesized by coupling the benzyl piperidi-
nes 9, 16 and 24 to the tosylates 25 and 26, which were
synthesized as previously reported12 (Scheme 4).
To optimize the benzodioxane group on the right hand part of
SMP-304, we synthesized the tosylates 39, 42, 43, 51, 54, 60, 66
and 71. Alkylation of the methyl 4-hydroxyphenylacetate with 3-
bromo-1-propanol provided the alcohol 32. Oxidation with TEMPO
followed by cyclization with PPA transformed 32 to the ketone 34.
Reduction of 34 with NaBH4 afforded the alcohol 35, which was
dehydrated azeotropically under acidic condition and then hydro-
genated over palladium on carbon to yield the ester 37. Reduction
of 37 with LiAlH4 followed by tosylation furnished the tosylate 39
(Scheme 5). Acetalization of the ketone 34 produced the acetal 40,
which was reduced with LiAlH4 deprotected and tosylated to yield
the tosylate 42. Reduction of 42 with NaBH4 afforded the tosylate
43 (Scheme 6). The tosylates 51 and 54 were synthesized from
methyl 2-(3-hydroxyphenyl)acetate in a manner similar to that
for the production of the tosylates 39 and 42, respectively
(Schemes 7 and 8). Acetylation of ethyl 4-phenylbutyrate afforded
Scheme 4. Reagents and conditions: (i) K2CO3, MeCN, reflux.
Scheme 5. Reagents and conditions: (i) Br(CH2)3OH, K2CO3, MeCN, reflux, (ii) NaClO2, NaClO (cat.), TEMPO (cat.), MeCN, buffer, r.t., (iii) PPA, 80 °C, (iv) NaBH4, MeOH, r.t., (v)
p-TsOH (cat.), benzene, reflux, (vi) H2, 10% Pd-C, EtOAc, r.t., (vii) LiAlH4, THF, r.t., (viii) p-TsCl, Et3N, Me3NHCl, CH2Cl2, 0 °C.
Please cite this article in press as: Yoshinaga H., et al. Bioorg. Med. Chem. (2018), https://doi.org/10.1016/j.bmc.2018.02.008

4
H. Yoshinaga et al. / Bioorganic & Medicinal Chemistry xxx (2018) xxx–xxx
Scheme 6. Reagents and conditions: (i) HC(OMe)3, p-TsOH, MeOH, r.t., (ii) LiAlH4, THF, r.t., then 2 N aq. HCl, acetone, r.t., (iii) p-TsCl, Et3N, Me3NHCl, CH2Cl2, 0 °C, (v) NaBH4,
MeOH, THF, r.t.
Scheme 7. Reagents and conditions: (i) Br(CH2)3OH, K2CO3, MeCN, reflux, (ii) NaClO2, NaClO (cat.), TEMPO (cat.), MeCN, buffer, r.t., (iii) PPA, 80 °C, (iv) NaBH4, MeOH, r.t., (v)
p-TsOH (cat.), benzene, reflux, (vi) H2, 10% Pd-C, EtOAc, r.t., (vii) LiAlH4, THF, r.t., (viii) p-TsCl, Et3N, Me3NHCl, CH2Cl2, 0 °C.
Scheme 8. Reagents and conditions: (i) HC(OMe)3, p-TsOH, MeOH, r.t., (ii) LiAlH4, THF, r.t., then 2 N aq. HCl, acetone, r.t., (iii) p-TsCl, Et3N, Me3NHCl, CH2Cl2, 0 °C, THF, r.t.
Scheme 9. Reagents and conditions: (i) AcCl, AlCl3, CH2Cl2, r.t., (ii) morpholine, S8, reflux, then 15% aq. NaOH, 2-PrOH, reflux, (iii) PPA, 80 °C, (iv) LiAlH4, THF, reflux, (v) MnO2,
CH2Cl2, r.t., (vi) p-TsCl, Et3N, Me3NHCl, CH2Cl2, 0 °C.
compound 55, which was subsequently subjected to Willgerodt-
Kindler reaction to yield the dicarboxylic acid 56. Cyclization with
PPA, giving 57, followed by reduction with LiAlH4 transformed 57
to 58. Selective oxidation of 58 with MnO2 to promote tosylation
at the primary alcohol provided 59, which was then tosylated to
give compound 60 (Scheme 9). The dicarboxylic acid 62 was syn-
thesized by alkylation of methyl 4-hydroxyphenylacetate followed
by hydrolysis. Compound 62 was then transformed to the tosylate
66 as described for the synthesis of 60 (Scheme 10). Similarly, the
tosylate 71 was synthesized from methyl 4-hydroxyphenylacetate
as described for the synthesis of 42 with first cyclization of the sul-
fide 6818 to accelerate the formation of cyclic slufonium salt and
then dealkylation with triethlyamine and oxidation of the formed
tosylate 70 (Scheme 11). The target compounds 72–80 were syn-
thesized by coupling the benzyl piperidines 16 and 24 to the tosy-
late 39, 42, 43, 51, 54, 60, 66 and 71 (Scheme 12).
Please cite this article in press as: Yoshinaga H., et al. Bioorg. Med. Chem. (2018), https://doi.org/10.1016/j.bmc.2018.02.008

H. Yoshinaga et al. / Bioorganic & Medicinal Chemistry xxx (2018) xxx–xxx
5
Scheme 13. Reagents and conditions: (i) H2, 10%Pd/C, MeOH, r.t. reflux.
the metabolism of SMP-304 is quite high reaching 86%. As CYP2D6
is a polymorphic drug-metabolizing enzyme with varying sub-
strate activity, drugs metabolized mainly by this enzyme have
exposure in poor metabolizers (PM) 3.5- to 53-fold higher than
that in extensive metabolizers (EM).13 Generally EMs eliminate
CYP2D6 substrates quickly while PMs do so more slowly.13 This
variation can affect the efficacy of SMP-304 due to fluctuation in
drug blood concentration and give rise to side effects.
Scheme 10. Reagents and conditions: (i) Br(CH2)3CO2Me, K2CO3, MeCN, reflux, (ii)
6 N aq. NaOH, MeOH, r.t., (iii) PPA, 80 °C, (iv) LiAlH4, THF, reflux, (v) MnO2, CH2Cl2, r.
t., (vi) p-TsCl, Et3N, Me3NHCl, CH2Cl2, 0 °C.
Most SSRIs show very potent binding affinity for SERT with Ki
values in the sub-nanomolar to single-degit nanomolar range.20
The affinity of SMP-304 for SERT (Ki = 32.7 nM) is more than ten-
fold weaker than the average value for known SSRIs, including
paroxetin, fluoxetine, sertraline and escitalopram. In a preliminary
study on SMP-304, a halogen atom was introduced into the 6-posi-
tion of the left phenyl group of compound A as it is reported that 4-
(3-methoxybenzyl)piperidine derivatives with a halogen atom at
6-position of the methoxy phenyl group possess potent SERT affin-
ity.12,21 We found that introduction of a bromine atom into the 6-
position of the left phenyl group of compound A enhances SERT
binding affinity, although optimization of this substitution remains
an issue (Fig. 2).
First, we changed the position of the bromine atom on the 6-
position of the benzene ring of SMP-304 to enhance SERT binding
affinity (Table 1). Although a bromine atom on the 4- or 5-position
of the benzene ring improved SERT binding affinity, it weakened
binding to the 5-HT1A receptor (SMP-304 vs. compound 28 and
29). Using the results of a previous study showing that compound
B (Y = H) has more potent binding affinity for 5-HT1A than com-
pound C (Y = Cl) (Fig. 3),12 the chlorine atom on the right hand part
of compound 28 and 29 was removed to afford compounds 30 and
31, respectively, with potent dual SERT and 5-HT1A binding affinity.
CYP2D6 contribution to the metabolism of these compounds was
slightly lower than that of SMP-304.
To better understand SAR at the 4-position of the benzene hav-
ing a methylene-piperidine, we synthesized the benzyl piperidines
81, 84a–c, 86 and 89a–b by direct transformation of the interme-
diate 15 or 16 with a bromine atom. Hydrogenolysis of 16 afforded
the benzyl piperidine 81 (Scheme 13). Direct halogenation of the
boronic acid 82 obtained from 1519 followed by deprotection with
hydrochloric acid afforded the benzyl piperidines 84a–c
(Scheme 14). The benzyl piperidine 86 was acquired by coupling
15 to methyl boronic acid and then treatment with hydrochloric
acid (Scheme 15). Hydrogenolysis of 87a–b obtained from 15
afforded 88a–b. Deprotection of 88a–b with hydrochloric acid fur-
nished the benzyl piperidines 89a–b (Scheme 16). Finally, the tar-
get compounds 90–96 were obtained by coupling of the benzyl
piperidines 81, 84a–c, 86 and 89a–b to the tosylate 42
(Scheme 17).
2.2. Evaluation of SERT and 5-HT1A receptor binding activity and
CYP2D6 metabolic contribution
In previous work, we found that SMP-304 possesses potent
SERT inhibitory activity with weak 5-HT1A partial agonistic activity
and induces faster onset of antidepressant-like effect than parox-
etine, a representative SSRI.12 However, CYP2D6 contribution to
Scheme 11. Reagents and conditions: (i) OH(CH2)2SMe, DEAD, PPh3, THF, r.t., (ii) Tf2O, CH2Cl2, ꢀ78 °C, then Et3N, MeCN, r.t., (iii) LiAlH4, THF, r.t., (iv) p-TsCl, Et3N, pyridine,
CH2Cl2, r.t., (v) mCPBA, CH2Cl2, r.t.
Scheme 12. Reagents and conditions: (i) K2CO3, MeCN, reflux.
Please cite this article in press as: Yoshinaga H., et al. Bioorg. Med. Chem. (2018), https://doi.org/10.1016/j.bmc.2018.02.008

6
H. Yoshinaga et al. / Bioorganic & Medicinal Chemistry xxx (2018) xxx–xxx
Scheme 14. Reagents and conditions: (i) n-BuLi, THF, ꢀ78 °C, then B(OMe)3, r.t., (ii) R = F: NaOH, AgOTf, MeOH, 0 °C, then F-TEDA-BF4, MS3A, acetone, 0 °C, R = Cl or I: NCS or
NIS, MeCN, r.t., (iii) 4 N HCl/EtOAc, CHCl3, r.t.
Scheme 15. Reagents and conditions: (i) MeB(OH)2, Pd(PPh3)4, 1.0 M aq. K2CO3, 1,4-dioxane, reflux (ii) 10% HCl/MeOH, r.t.
and 78 with a ketone bicyclic group showed much lower CYP2D6
metabolic contribution than compound 30. On the other hand,
introduction of a sulfone bicyclic group as hydrophilic substituent
into the right hand part of compound 30 had no effect on CYP2D6
metabolic contribution (compound 79). This finding indicates that
increasing hydrophilicity of the right hand part by introducing a
polar group does not necessarily lower CYP2D6 metabolic contri-
bution. Compounds 74 and 77 with a six-membered cyclic ketone
in the right hand part showed very potent SERT and 5-HT1A binding
activity. In particular, compound 74 exhibited a well-balanced
Scheme 16. Reagents and Conditions: (i) n-BuLi, THF, ꢀ78 °C, then DMA or acetone
(ii) H2 (0.4 MPa), 10% Pd-C, c. HCl, MeOH, r.t., (iii) 4 N HCl/EtOAc, CHCl3, r.t.
SERT and 5-HT1A binding affinity. Introduction of compound 74
right hand part into compound 31, giving compound 80, improved
CYP2D6 metabolic contribution to a level similar to that of com-
pound 74. However, compound 80 exhibited weaker SERT and 5-
HT1A binding affinity than compound 74.
Finally, we replaced the bromine atom of compound 74 with a
hydrogen atom, other halogen atoms, or lower alkyl groups
(Table 3). All substitutions, except for 2-propyl group had almost
no effect on 5-HT1A binding activity. However, binding to SERT
was significantly altered by displacement of the bromine atom.
For other halogen atoms, the order of SERT binding activity was
I > Br > Cl > F (compounds 91, 92, 74 and 93 in Table 3). This con-
forms to the expected activity ranking when a compound interacts
with a protein via halogen bonding defined as interaction of a halo-
It is known that CYP2D6 substrates have a basic ionized nitro-
gen at physiological pH and a hydrophobic region positioned 5–
7 Å from the nitrogen atom.22 In addition, metabolism analysis
revealed that metabolism of SMP-304 may affect the benzodioxane
ring (Fig. 4). Therefore, to reduce CYP2D6 contribution to SMP-304
metabolism, we focused our effort on optimizing the right hand
part of compound 30 (Table 2). First, the ether bond was changed
to a methylene bond, a hydroxy group or a ketone group (Fig. 4),
because this change promotes transformation of intermediates
34 and 46 to three tosylates 39, 42 and 43 and two tosylates 51
and 54, respectively (Scheme 5–8).
Changing the ether bond to a methylene bond improved
CYP2D6 metabolic contribution (compound 30 vs. compound 72
or 75). Introduction of a hydroxyl group or a ketone group at the
ether bond of compound 30 also ameliorated CYP2D6 metabolic
contribution (compound 30 vs. compound 73, 74 and 76). Espe-
cially, compound 74 shows stronger binding affinity for both SERT
and 5-HT1A than compound 30. Like compound 74, compound 77
gen atom with an oxygen or nitrogen via the
r-hole. The size of the
r
-hole rules the strength of the halogen bonding and increases
with increasing halogen size (I > Br > Cl > F).23–25 Actually, co-crys-
tal structures of leucine transporter LeuT, a bacterial SERT homo-
log, bound to SSRIs, such as sertraline and fluoxetine are
reported, suggesting that sertraline and fluoxetine interact with
Scheme 17. Reagents and conditions: (i) K2CO3, MeCN, reflux.
Please cite this article in press as: Yoshinaga H., et al. Bioorg. Med. Chem. (2018), https://doi.org/10.1016/j.bmc.2018.02.008

H. Yoshinaga et al. / Bioorganic & Medicinal Chemistry xxx (2018) xxx–xxx
7
Fig. 2. Strategy to enhance SERT binding activity.
Table 1
SERT and 5-HT1A receptor binding activity and CYP2D6 contribution in the optimization of Br and Y.
a
Compound
Br
Y
SERTa
5-HT1A
CYP2D6 contribution
86%
1
6-Br
Cl
32.7 3.4
9.4 0.20
(HCl salt: SMP-304)
27
28
29
30
31
2-Br
4-Br
5-Br
4-Br
5-Br
Cl
Cl
Cl
H
74.6 8.2
5.47 0.70
14.1 1.7
2.39 0.27
18.5 0.05
468 36
251 21
546 14
16.0 2.1
57.8 5.9
N.T.
75.6%
64.4%
55.6%
49.4%
H
N.T. = not tested.
a
Ki values (nM) are the means of two independent experiments.
Fig. 3. SAR information in the previous work.
Fig. 4. The strategy to reduce the contribution rate of CYP2D6 metabolism.
SERT via halogen bonding.26 Another possibility is that the halogen
atom affects the binding activity via hydrophobic or van der Waals
interaction. However, the strength of SERT binding affinity did not
increase with size or hydrophobicity in the case of lower alkyl
groups (90, 94, 95 and 96 in Table 3). Based on these findings, it
is believed that compound 74 and its analogues with different
Please cite this article in press as: Yoshinaga H., et al. Bioorg. Med. Chem. (2018), https://doi.org/10.1016/j.bmc.2018.02.008

8
H. Yoshinaga et al. / Bioorganic & Medicinal Chemistry xxx (2018) xxx–xxx
Table 2
SERT and 5-HT1A receptor binding activity and CYP2D6 contribution in the optimization of Ar group.
a
Compound
Br
Ar
SERTa
5-HT1A
CYP2D6 contribution
55.6%
30
4-Br
2.39 0.27
16.0 2.1
43.1 4.0
32.8 8.4
72
73
4-Br
4-Br
1.74 0.07
8.44 0.28
0%
36.4%
74
4-Br
1.02 0.06
5.05 1.1
0%
75
76
4-Br
4-Br
5.40 0.19
1.41 0.11
19.0 6.2
42.8 3.9
6.8%
3.6%
77
78
4-Br
4-Br
0.67 0.07
1.17 0.13
12.4 2.0
11.8%
4.2%
71.9 10.1
79
4-Br
2.59 0.03
11.3 2.7
55.8%
31
80
5-Br
5-Br
18.5 0.05
5.32 0.68
57.8 5.9
44.3 1.9
49.4%
14%
a
Ki values (nM) are the means of two or three independent experiments.
halogen atoms bind to SERT via halogen bonding, although further
studies are needed to understand better the mechanism of this
interaction.
gle-digit nanomolar serotonin uptake inhibition, although all three
compounds exhibited equal 5-HT1A partial agonistic activity.
2.4. Evaluation of serotonin elevating effect in the rat frontal cortex
2.3. Evaluation of serotonin reuptake inhibition and 5-HT1A agonistic
activity
As compound 74, 92 and 94 showed serotonin uptake inhibition
and 5-HT1A partial agonistic activity, we proceeded to further selec-
tion using rat microdialysis experiments to evaluate their serotonin
elevating effect in the rat frontal cortex (Fig. 5). All three compounds
increased serotonin level following a single dosing at 10 mg/kg p.o.
compared to the vehicle or SMP-304. As for the order of potency
related to SERT binding affinity and serotonin uptake inhibition,
compound 74 had the most elevating effect on serotonin level. Based
on this finding and the results of multiple enzymes (such as CYP3A4
and CYP2J2) metabolism, we selected compound 74 (DSP-1053,
BsOH salt) as a potential candidate with low CYP2D6 metabolic con-
tribution for further evaluation in rat in vivo studies.
Based on our evaluation of compounds with good and balanced
(within 10 times) binding affinity for both SERT and 5-HT1A, we
chose compounds 74, 92 and 94 with single-digit nanomolar bind-
ing activity for both SERT and 5-HT1A and less than 60% CYP2D6
metabolic contribution for functional evaluation, including sero-
tonin reuptake inhibition and 5-HT1A agonistic activity (Table 4).
All three compounds showed serotonin reuptake inhibition in the
order of SERT binding activity, indicating that SERT binding assay
could be used as a quick method to find serotonin uptake inhibitors
(Ref. 14,15). Especially, compound 74 showed the most potent sin-
Please cite this article in press as: Yoshinaga H., et al. Bioorg. Med. Chem. (2018), https://doi.org/10.1016/j.bmc.2018.02.008

H. Yoshinaga et al. / Bioorganic & Medicinal Chemistry xxx (2018) xxx–xxx
9
Table 3
SERT and 5-HT1A receptor binding activity and CYP2D6 contribution in the transformation of X group.
SERTa
5-HT1A
CYP2D6 contribution
a
Compound
X
90
91
92
74
93
94
95
96
H
F
202 25
111 3.5
2.41 0.11
2.25 0.14
2.81 0.03
5.05 1.07
8.76 0.95
7.72 1.42
19.3 0.50
69.9 4.0
N.T.
N.T.
53.5%
0%
N.T.
48.5%
18.4%
N.T.
Cl
Br
I
Me
Et
2-Pr
3.37 0.11
1.02 0.06
0.41 0.06
5.25 1.25
9.85 1.36
246 27
N.T. = not tested.
a
Ki values (nM) are the means of two or three independent experiments.
Table 4
In vitro 5-HT uptake inhibition, 5-HT1A intrinsic activity and CYP2D6 contribution.
Compound
X
SERTa
(Ki, nM)
5-HT uptake inhibitionb
(IC50, nM)
5-HT1A
CYP2D6
contribution
(%)
bindinga
(Ki, nM)
I.A.c
(%)
74
92
94
Br
Cl
Me
1.02 0.06
3.37 0.11
5.25 1.25
2.71 0.41
16.4 4.96
45.3 12.0
5.05 1.1
2.81 0.03
7.72 1.42
70.0
60.8
49.5
0
53.5
48.5
a
b
c
Ki values (nM) are the means of two or three independent experiments.
IC50 values (nM) are the means of at least three independent experiments.
I.A.: intrinsic activity.
Fig. 5. Effects of compound 74, 92, 94 and SMP-304 on 5-HT levels in rat frontal cortex. (A) Time-course changes in 5-HT levels in the rat frontal cortex after compound 74,
92, 94 and SMP-304 p.o. administration. Each point with a vertical bar represents the mean + SEM of percentage baseline value. (B) Effects of compound 74, 92, 94 and SMP-
304 on extracellular 5-HT levels in the rat frontal cortex. Each column with vertical bar represents the mean + SEM of AUC of 5-HT percent over 3 h. *P < 0.05, compared to the
vehicle-treated group using parametric Tukey’s multiple comparison test. Vehicle group, n = 6; compound 74, 94 and SMP-304 10 mg kgꢀ1 groups, n = 3; and compound 92
10 mg kgꢀ1 group, n = 2.
Please cite this article in press as: Yoshinaga H., et al. Bioorg. Med. Chem. (2018), https://doi.org/10.1016/j.bmc.2018.02.008

10
H. Yoshinaga et al. / Bioorganic & Medicinal Chemistry xxx (2018) xxx–xxx
The results of further evaluation of DSP-1053 in rat in vivo
4.2 Hz), 7.44 (1H, brd, J = 8.1 Hz), 7.54 (1H, brs), 7.70 (1H, d, J =
8.1 Hz), 13.2 (1H, brs).
studies have already been reported in our previous work.27
Although in vitro experiments indicated that DSP-1053 acts as a
5-HT1A partial agonist, more elaborate rat microdialysis experi-
ments revealed that this compound could work as a 5-HT1A auto-
receptor antagonist.27–29 In fact, in the rat forced swimming test
and olfactory bulbectomy model, DSP-1053 showed a faster onset
antidepressant-like effect than paroxetine, a representative SSRI,
and had minimal undesirable effects, such as nausea and emesis
in animal models. Based on these findings, we proceeded with its
development as a clinical candidate.
4.1.2. [4-Bromo-3-(2-methoxyethoxy)phenyl]methanol (11)
To a suspension of sodium borohydride (8.08 g, 214 mmol) in
THF (100 mL) was added boron trifluoride diethyl etherate (35.0
mL, 285 mmol). After stirring at room temperature for 1 h, a solu-
tion of intermediate 10 (19.5 g, 71.2 mmol) in THF (300 mL) was
added dropwise to the reaction mixture for 30 min, and the whole
was stirred at room temperature for 3 h. H2O (200 mL) was then
added dropwise to the reaction mixture for 20 min, and the mix-
ture was extracted with toluene (200 mL ꢁ 2). The combined
organic layer was first washed with 3% aqueous NaHCO3 (200
mL) and H2O (200 mL), and then concentrated to give 18.2 g
(98%) of the title compound as a colorless oil. 1H NMR (300 MHz,
CDCl3) d: 1.70 (1H, t, J = 5.9 Hz), 3.49 (3H, s), 3.82 (2H, t, J = 4.8
Hz), 4.20 (2H, t, J = 4.8 Hz), 4.66 (2H, d, J = 5.7 Hz), 6.83 (1H, brd,
J = 8.1 Hz), 6.97 (1H, brs), 7.50 (1H, d, J = 8.1 Hz).
3. Conclusion
In this study, we optimized SMP-304 to ameliorate its high
CYP2D6 metabolism. Changing the position of the bromine atom
on the 6-position of the benzene ring of SMP-304 enhanced SERT
binding affinity. In addition, replacement of the benzodioxane ring
in the right hand part with a chromon-4-one ring lowered CYP2D6
metabolic contribution. These efforts led to the identification of
compounds 74 (DSP-1053, BsOH salt), 92 and 94 as potent SERT
inhibitors with 5-HT1A partial agonistic activity and low CYP2D6
metabolic contribution. In rat microdialysis experiments, DSP-
1053 exhibited the most potent serotonin elevating effect in the
rat frontal cortex. Based on these findings, we chose DSP-1053 as
a potential clinical candidate for the treatment of depression.
4.1.3. 1-Bromo-4-(bromomethyl)-2-(2-methoxyethoxy)benzene (12)
To a mixture of intermediate 11 (18.0 g, 68.9 mmol), trimethy-
lamine hydrochloride (0.467 g, 7.12 mmol) and triethylamine
(19.8 mL, 142 mmol) in toluene (90 mL) was added dropwise
methanesulfonyl chloride (8.56 g, 74.7 mmol) at 0 °C for 30 min.
After stirring at 0 °C for 2 h, the reaction mixture was added to
5% aqueous KHSO4 (180 mL) at 0 °C. The organic layer was sepa-
rated, and the water layer was extracted with toluene (90 mL).
The combined organic layer was then washed with H2O (180 mL)
and concentrated. To a solution of the residue in THF (100 mL)
was added lithium bromide (18.5 g, 214 mmol). After reflux for 1
h, H2O (100 mL) was added to the reaction mixture, and the whole
was extracted with toluene (100 mL ꢁ 2). The combined organic
layer was washed with 5% aqueous NaHCO3 (100 mL) and H2O
(100 mL) and then concentrated to give 19.5 g (88%) of the title
compound as a white solid. 1H NMR (300 MHz, CDCl3) d: 3.49
(3H, s), 3.82 (2H, t, J = 4.6 Hz), 4.20 (2H, t, J = 4.6 Hz), 4.43 (2H, s),
6.87 (1H, dd, J = 8.1, 1.7 Hz), 6.97 (1H, d, J = 1.8 Hz), 7.49 (1H, d, J
= 8.3 Hz).
4. Experimental section
4.1. Chemistry
Melting points were determined on a Stanford Research Sys-
tem; OptiMelt MPA 100 without correction. NMR spectra were
recorded at ambient temperature on a JEOL JNM-AL400 FT NMR
spectrometer. Chemical shifts are expressed in d values (ppm) rel-
ative to the internal standard tetramethylsilane, and signals are
expressed as: s (singlet), d (doublet), t (triplet), m (multiplet) or
br (broad). IR spectra were recorded on a JEOL JIR-SPX60 spectrom-
eter as attenuated total reflection (ATR). High-resolution mass
spectra (HRMS) were recorded on a Thermo Fisher Scientific LTQ
orbitrap Discovery MS equipment. Elemental analysis was per-
formed on a CE Instrument EA1110 and a Yokokawa analytical sys-
tem IC7000. In general, reagents and solvents were used as
obtained from commercial suppliers without further purification.
Reaction progress was determined by thin layer chromatography
(TLC) analysis on a Merck silica gel 60 F254 precoated glass plate.
Visualization was done with UV light (254 nm) or iodine. Flash col-
umn chromatography was conducted using Merck silica gel 60
(70–230 mesh). All reactions were carried out under a nitrogen
atmosphere unless otherwise mentioned.
4.1.4. [4-Bromo-3-(2-methoxyethoxy)benzyl](triphenyl)phosphonium
bromide (13)
The title compound was prepared in a manner similar to that
for the preparation of 2 using intermediate 12. Compound 13
was obtained in 78% yield as a white solid. 1H NMR (300 MHz,
DMSO-D6) d: 3.27 (3H, s), 3.52 (2H, t, J = 5.0 Hz), 3.66 (2H, t, J =
4.6 Hz), 5.12 (2H, d, J = 15.8 Hz), 6.55 (1H, brd, J = 7.9 Hz), 6.59
(1H, brs), 7.47 (1H, d, J = 7.9 Hz), 7.60–7.81 (12H, m), 7.90 (3H,
brdd, J = 7.7, 7.7 Hz).
4.1.5. tert-Butyl 4-[4-bromo-3-(2-methoxyethoxy)benzylidene]
piperidine-1-carboxylate (14)
The title compound was prepared in a manner similar to that
for the preparation of 3 using intermediate 13. Compound 14
was obtained quantitatively as a pale yellow oil. 1H NMR (300
MHz, CDCl3) d: 1.48 (9H, s), 2.32 (2H, t, J = 5.6 Hz), 2.43 (2H, t, J
= 6.7 Hz), 3.40 (2H, t, J = 5.9 Hz), 3.46–3.55 (5H, m), 3.81 (2H, t, J
= 5.0 Hz), 4.16 (2H, t, J = 4.4 Hz), 6.28 (1H, brs), 6.68 (1H, dd, J =
8.1, 1.7 Hz), 6.74 (1H, d, J = 2.1 Hz), 7.46 (1H, d, J = 8.1 Hz).
4.1.1. 4-Bromo-3-(2-methoxyethoxy)benzoic acid (10)
To a mixture of 2-methoxyethanol (16.5 g, 217 mmol) and
potassium tert-butoxide (24.3 g, 217 mmol) in N-methyl-2-pyrro-
lidinone (175 mL) was added 4-bromo-3-fluorobenzoic acid
(19.0 g, 86.8 mmol). After stirring at 90 °C for 6 h, the reaction mix-
ture was cooled to room temperature and added dropwise to a
solution of conc. HCl (25 mL) in H2O (500 mL) for 40 min. After stir-
ring at room temperature for 1 h, the resulting solid was filtered,
rinsed with H2O (20 mL ꢁ 2) and MeCN (20 mL ꢁ 2), and then col-
lected. The white solid was crystallized using MeCN (380 mL) and
stirred at room temperature for 1 h. The resulting solid was fil-
tered, rinsed with MeCN (20 mL ꢁ 2) and collected to give 20.1 g
(85%) of the title compound as a white solid. 1H NMR (300 MHz,
DMSO-D6) d: 3.33 (3H, s), 3.70 (2H, t, J = 4.2 Hz), 4.23 (2H, t, J =
4.1.6. tert-Butyl 4-[4-bromo-3-(2-methoxyethoxy)benzyl]piperidine-
1-carboxylate (15)
The title compound was prepared in a manner similar to that
for the preparation of 4 using intermediate 14. Compound 15
was obtained in 64% yield as a colorless oil. 1H NMR (300 MHz,
CDCl3) d: 1.04–1.20 (2H, m), 1.45 (9H, s), 1.51–1.70 (3H, m), 2.48
(2H, d, J = 6.6 Hz), 2.62 (2H, brt, J = 13.2 Hz), 3.49 (3H, s), 3.81
Please cite this article in press as: Yoshinaga H., et al. Bioorg. Med. Chem. (2018), https://doi.org/10.1016/j.bmc.2018.02.008

H. Yoshinaga et al. / Bioorganic & Medicinal Chemistry xxx (2018) xxx–xxx
11
(2H, t, J = 5.3 Hz), 4.17 (2H, t, J = 4.6 Hz), 3.98–4.23 (2H, m), 6.63
3.56 (2H, s), 3.65 (3H, s), 4.25–4.32 (2H, m), 6.73 (1H, d, J = 8.4
(1H, dd, J = 7.9, 1.7 Hz), 6.70 (1H, d, J = 1.8 Hz), 7.42 (1H, d, J =
Hz), 7.10 (1H, dd, J = 8.4, 2.1 Hz), 7.41 (1H, d, J = 2.1 Hz).
8.1 Hz).
4.1.12. 6-(2-Hydroxyethyl)-2,3-dihydro-4H-chromen-4-one (41)
To a suspension of lithium aluminum hydrate (2.58 g, 68.1
mmol) in THF (190 mL) was added dropwise a solution of interme-
diate 40 (12.1 g, 45.4 mmol) for 20 min. After stirring at room tem-
perature for 1 h, H2O (2.54 mL), 15% aqueous NaOH (2.54 mL) and
H2O (7.63 mL) was added to the reaction mixture at 0 °C. After stir-
ring at 0 °C for 1 h, the reaction mixture was filtered, and the fil-
trate was concentrated. To a solution of the residue (10.7 g) in
acetone (100 mL) was added 2 N aqueous HCl (100 mL). After stir-
ring at room temperature for 1 h, acetone was evaporated from the
reaction mixture, and the remaining mixture was extracted with
EtOAc (500 mL). The organic layer was washed with H2O (400
mL), and the aqueous layers were combined and extracted with
EtOAc (300 mL). The combined organic layer was dried over
MgSO4, filtered and concentrated to give 8.55 g (98%) of the title
compound as a pale yellow oil. 1H NMR (400 MHz, CDCl3) d:
2.78–2.86 (4H, m), 3.86 (2H, t, J = 6.4 Hz), 4.52 (2H, t, J = 6.4 Hz),
6.94 (1H, d, J = 8.4 Hz), 7.36 (1H, dd, J = 8.4, 2.2 Hz), 7.76 (1H, d, J
= 2.2 Hz).
4.1.7. 4-[4-Bromo-3-(2-methoxyethoxy)benzyl]piperidine
hydrochloride (16)
The title compound was prepared in a manner similar to that
for the preparation of 5 using intermediate 15. Compound 16
was obtained in 88% yield as a white solid. 1H NMR (300 MHz,
CDCl3) d: 1.47–1.94 (5H, m), 2.55 (2H, d, J = 5.5 Hz), 2.79 (2H, t, J
= 12.0 Hz), 3.47 (2H, t, J = 13.0 Hz), 3.51 (3H, s), 3.81 (2H, t, J =
4.8 Hz), 4.15 (2H, t, J = 4.8 Hz), 6.62 (1H, dd, J = 7.9, 1.8 Hz), 6.68
(1H, d, J = 1.7 Hz), 7.43 (1H, d, J = 8.1 Hz), 9.50 (2H, brs); Mp:
171–172 °C.
4.1.8. Methyl [4-(3-hydroxypropoxy)phenyl]acetate (32)
To a mixture of methyl 4-hydroxyphenylacetate (26.6 g, 160
mmol) and potassium carbonate (26.5 g, 192 mmol) in MeCN
(500 mL) was added 3-bromo-1-propanol (17.4 mL, 192 mmol).
After reflux for 14 h, the reaction mixture was filtered, and the fil-
trate was concentrated. The residue was purified by silica gel chro-
matography using 34% EtOAc/hexane as eluent to give 23.7 g (66%)
of the title compound as a colorless oil. 1H NMR (300 MHz, CDCl3)
d: 1.75 (1H, t, J = 5.5 Hz), 2.00–2.09 (2H, m), 3.57 (2H, s), 3.68 (3H,
s), 3.87 (2H, td, J = 5.5, 5.5 Hz), 4.12 (2H, t, J = 5.9 Hz), 6.84–6.90
(2H, m), 7.16–7.22 (2H, m).
4.1.13. 2-(4-oxo-3,4-dihydro-2H-chromen-6-yl)ethyl 4-
methylbenzenesulfonate (42)
The title compound was prepared in a manner similar to that
for the preparation of 39 using intermediate 41. Compound 42
was obtained in 89% yield as a white solid. 1H NMR (300 MHz,
CDCl3) d: 2.44 (3H, s), 2.79 (2H, t, J = 6.5 Hz), 2.91 (2H, t, J = 6.9
Hz), 4.18 (2H, t, J = 6.9 Hz), 4.51 (2H, t, J = 6.4 Hz), 6.88 (1H, d, J =
8.4 Hz), 7.26 (1H, dd, J = 8.5, 2.5 Hz), 7.30 (2H, d, J = 8.5 Hz), 7.60
(1H, d, J = 2.2 Hz), 7.71 (2H, d, J = 8.3 Hz).
4.1.9. 3-[4-(2-Methoxy-2-oxoethyl)phenoxy]propanoic acid (33)
To a mixture of intermediate 32 (23.7 g, 105 mmol), 2,2,6,6-
tetramethyl-1-piperidinyloxy, radical (2.06 g, 13.2 mmol), sodium
chlorite (19.0 g, 210 mmol), 0.25 M aqueous KH2PO4 (130 mL),
0.25 M aqueous Na2HPO4 (130 mL) and MeCN (260 mL) was added
5% aqueous NaClO (2.85 mL, 2.10 mmol). After stirring at room
temperature for 3.5 h, 10% aqueous NaHSO3 (200 mL) was added
to the reaction mixture, and MeCN was evaporated. The remaining
mixture was then extracted with EtOAc (500 mL ꢁ 2), and the
combined organic layer was dried over MgSO4, filtered and con-
centrated. The residue was purified by silica gel chromatography
using 16–100% EtOAc/hexane as eluent to give 24.7 g (99%) of
the title compound as a white solid. 1H NMR (300 MHz, CDCl3) d:
2.85 (2H, t, J = 6.3 Hz), 3.57 (2H, s), 3.69 (3H, s), 4.24 (2H, t, J =
6.3 Hz), 6.87 (2H, d, J = 8.5 Hz), 7.20 (2H, d, J = 8.5 Hz).
4.1.14. 6-(2-{4-[4-Bromo-3-(2-methoxyethoxy)benzyl]piperidin-1-yl}
ethyl)-2,3-dihydro-4H-chromen-4-one (74)
The title compound was prepared in a manner similar to that
for the preparation of 27 using intermediates 16 and 42. Com-
pound 74 was obtained in 99% yield as a colorless oil. 1H NMR
(300 MHz, CDCl3) d: 1.24–1.39 (2H, m), 1.40–1.73 (3H, m), 1.93
(2H, t, J = 10.6 Hz), 2.40–2.61 (2H, m), 2.48 (2H, d, J = 7.2 Hz),
2.66–2.87 (2H, m), 2.79 (2H, t, J = 6.4 Hz), 2.95 (2H, d, J = 11.7
Hz), 3.49 (3H, s), 3.81 (2H, t, J = 4.9 Hz), 4.17 (2H, t, J = 4.9 Hz),
4.51 (2H, t, J = 6.4 Hz), 6.64 (1H, dd, J = 8.1, 1.8 Hz), 6.71 (1H, d, J
= 1.8 Hz), 6.89 (1H, d, J = 8.4 Hz), 7.32 (1H, dd, J = 8.4, 2.2 Hz),
7.41 (1H, d, J = 8.1 Hz), 7.70 (1H, d, J = 2.2 Hz); HRMS (ESI) m/z
calcd for C26H33BrNO4 [M + H]+ 502.1587; found 502.1591. The
product was converted to the corresponding benzenesulfonate salt.
Mp: 142–143 °C. Anal. Calcd for C32H38BrNO7S: C, 58.18; H, 5.80; N,
2.12; S, 4.85; Br, 12.10. Found: C, 58.09; H, 5.80; N, 2.25; S, 4.85; Br,
12.10.
4.1.10. Methyl (4-oxo-3,4-dihydro-2H-chromen-6-yl)acetate (34)
A mixture of intermediate 33 (22.4 g, 94.0 mmol) and polyphos-
phoric acid (112 g) was stirred at 80 °C for 2 h. Toluene (300 mL)
and H2O (1000 mL) were then added to the reaction mixture, and
the whole was extracted with toluene (500 mL) and Et2O (500
mL). The combined organic layer was dried over MgSO4, filtered
and concentrated, and the residue was purified by silica gel chro-
matography using 25% EtOAc/hexane as eluent to give 19.0 g
(92%) of the title compound as a pale yellow oil. 1H NMR (300
MHz, CDCl3) d: 2.81 (2H, t, J = 6.5 Hz), 3.60 (2H, s), 3.70 (3H, s),
4.53 (2H, t, J = 6.5 Hz), 6.96 (1H, d, J = 8.5 Hz), 7.41 (1H, dd, J =
8.5, 2.2 Hz), 7.78 (1H, d, J = 2.2 Hz).
4.1.15. [4-{[1-(tert-Butoxycarbonyl)piperidin-4-yl]methyl}-2-(2-
methoxyethoxy)phenyl]boronic acid (82)
To a solution of intermediate 15 (1.00 g, 2.33 mmol) in THF (10
mL) was added dropwise 2.6 M n-BuLi in n-hexane (0.99 mL, 2.57
mmol) at ꢀ78 °C. After stirring at ꢀ78 °C for 30 min, trimethyl
borate (0.779 mL, 6.99 mmol) was added to the reaction mixture,
and the whole was stirred at room temperature for 5 h. 10% aque-
ous KHSO4 (20 mL) was added to the reaction mixture, and the new
mixture was extracted with EtOAc (20 mL ꢁ 2). The combined
organic layer was dried over MgSO4, filtered and concentrated,
and the residue was purified by silica gel chromatography using
32–53% EtOAc/hexane as eluent to give 0.360 g (39%) of the title
compound as a white solid. 1H NMR (300 MHz, CDCl3) d: 1.06–
1.22 (2H, m), 1.45 (9H, s), 1.60–1.71 (3H, m), 2.53 (2H, d, J = 7.0
Hz), 2.56–2.70 (2H, m), 3.46 (3H, s), 3.75–3.79 (2H, m), 3.99–4.14
4.1.11. Methyl (4,4-dimethoxy-3,4-dihydro-2H-chromen-6-yl)acetate
(40)
To a mixture of intermediate 34 (10.0 g, 45.4 mmol), trimethyl
orthoformate (199 mL) and MeOH (106 mL) was added p-toluene-
sulfonic acid monohydrate (0.864 g, 4.54 mmol). After stirring at
room temperature for 22 h, the reaction mixture was added to
sat. aqueous NaHCO3 (500 mL) and extracted with EtOAc (500
mL). The organic layer was dried over Na2SO4, filtered and concen-
trated to give 12.1 g (quant.) of the title compound as a pale yellow
oil. 1H NMR (300 MHz, CD3OD) d: 2.08–2.14 (2H, m), 3.24 (6H, s),
Please cite this article in press as: Yoshinaga H., et al. Bioorg. Med. Chem. (2018), https://doi.org/10.1016/j.bmc.2018.02.008

12
H. Yoshinaga et al. / Bioorganic & Medicinal Chemistry xxx (2018) xxx–xxx
(2H, m), 4.18–4.23 (2H, m), 5.84 (2H, brs), 6.67 (1H, d, J = 1.3 Hz),
6.83 (1H, dd, J = 7.5, 1.3 Hz), 7.73 (1H, d, J = 7.5 Hz).
1.94 (2H, t, J = 12.0 Hz), 2.49 (2H, d, J = 7.2 Hz), 2.50–2.55 (2H, m),
2.72–2.80 (2H, m), 2.79 (2H, t, J = 6.4 Hz), 2.96 (2H, d, J = 11.5 Hz),
3.49 (3H, s), 3.81 (2H, t, J = 4.8 Hz), 4.18 (2H, t, J = 4.8 Hz), 4.51 (2H,
t, J = 6.4 Hz), 6.69 (1H, dd, J = 8.0, 1.8 Hz), 6.74 (1H, d, J = 1.8 Hz),
6.89 (1H, d, J = 8.5 Hz), 7.24 (1H, d, J = 8.0 Hz), 7.31 (1H, dd, J = 8.5,
4.1.16. tert-Butyl 4-[4-chloro-3-(2-methoxyethoxy)benzyl]piperidine-
1-carboxylate (83b)
To a solution of intermediate 82 (0.100 g, 0.254 mmol) in MeCN
(1.0 mL) was added N-chlorosuccinimide (0.0340 g, 0.254 mmol)
and cuprous chloride (0.0252 g, 0.254 mmol). After reflux for 3 h,
H2O (20 mL) was added to the reaction mixture, and the whole
was extracted with EtOAc (30 mL ꢁ 2). The combined organic layer
was dried over MgSO4, filtered and concentrated, and the residue
was purified by silica gel chromatography using 20–41% EtOAc/
hexane as eluent to give 0.0775 g (79%) of the title compound as
a colorless oil. 1H NMR (400 MHz, CDCl3) d: 1.05–1.19 (2H, m),
1.45 (9H, s), 1.52–1.69 (3H, m), 2.49 (2H, d, J = 6.8 Hz), 2.56–2.69
(2H, m), 3.48 (3H, s), 3.80 (2H, t, J = 4.9 Hz), 3.97–4.14 (2H, m),
4.17 (2H, t, J = 4.9 Hz), 6.68 (1H, dd, J = 8.0, 2.0 Hz), 6.71 (1H, d, J
= 7.8 Hz), 6.73 (1H, d, J = 2.0 Hz).
2.2 Hz), 7.70 (1H, d, J = 2.2 Hz); HRMS (ESI) m/z calcd for C26H33-
ClNO4 [M + H]+ 458.2093; found 458.2100.
4.1.21. 6-(2-{4-[3-(2-Methoxyethoxy)-4-methylbenzyl]piperidin-1-yl}
ethyl)-2,3-dihydro-4H-chromen-4-one (94)
The title compound was prepared in a manner similar to that
for the preparation of 27 using intermediates 42 and 86. Com-
pound 94 was obtained in 93% yield as a colorless oil. 1H NMR
(400 MHz, CDCl3) d: 1.24–1.37 (2H, m), 1.44–1.54 (1H, m), 1.61–
1.69 (2H, m), 1.89–1.97 (2H, m), 2.20 (3H, s), 2.46–2.55 (4H, m),
2.72–2.82 (4H, m), 2.91–2.98 (2H, m), 3.47 (3H, s), 3.77 (2H, t, J
= 4.9 Hz), 4.12 (2H, t, J = 4.9 Hz), 4.51 (2H, t, J = 6.5 Hz), 6.61 (1H,
d, J = 1.2 Hz), 6.65 (1H, dd, J = 7.3, 1.5 Hz), 6.89 (1H, d, J = 8.5 Hz),
7.03 (1H, d, J = 7.6 Hz), 7.32 (1H, dd, J = 8.5, 2.4 Hz), 7.70 (1H, d, J
4.1.17. 4-[4-Chloro-3-(2-methoxyethoxy)benzyl]piperidine
hydrochloride (84b)
= 2.2 Hz); HRMS (ESI) m/z calcd for
C
27H36NO4 [M + H]+
438.2639; found 438.2643.
The title compound was prepared in a manner similar to that
for the preparation of 84a using intermediate 83b. Compound
84b was obtained quantitatively as a white solid. 1H NMR (400
MHz, DMSO-D6) d: 1.26–1.39 (2H, m), 1.65–1.72 (2H, m), 1.74–
1.84 (1H, m), 2.52 (2H, d, J = 4.9 Hz), 2.73–2.82 (2H, m), 3.18–
3.25 (2H, m), 3.34 (3H, s), 3.67–3.70 (2H, m), 4.15–4.18 (2H, m),
6.77 (1H, dd, J = 8.0, 1.7 Hz), 6.99 (1H, d, J = 1.7 Hz), 7.32 (1H, d, J
= 8.0 Hz), 8.64 (1H, br s).
4.2. Biological experiments
4.2.1. Radioreceptor assay
Radioreceptor assays were conducted according to the methods
of Owens et al. for SERT30 and Yabuuchi et al. for 5-HT1A receptor.31
Briefly, the followings were used. For serotonin transporter assays,
cell membrane expressing human serotonin transporter and [3H]ci-
talopram. For 5-HT1A receptor assays, cell membrane expressing
human 5-HT1A receptor and [3H]8-OH-DPAT. The inhibition con-
stant (Ki) was calculated in MicrosoftOffice Excel 2003 (Microsoft
Corporation) using the Cheng–Prusoff equation [Ki = IC50/(1 + ([L]/
Kd)], where L is the concentration of radioligand in the assay and
Kd is the dissociation constant of the radioligand for the receptor.
4.1.18. tert-Butyl 4-[3-(2-methoxyethoxy)-4-methylbenzyl]
piperidine-1-carboxylate (85)
A mixture of intermediate 15 (5.00 g, 11.7 mmol), methyl-
boronic acid (0.978 g, 16.0 mmol), tetrakis(triphenylphosphine)-
palladium (0.674 g, 0.583 mmol), 1 M aqueous K2CO3 (35 mL)
and 1,4-dioxane (80 mL) was refluxed for 4 h. After evaporation
of 1,4-dioxane from the reaction mixture, H2O was added, and
the whole was extracted with EtOAc. The combined organic layer
was washed with brine, dried over MgSO4 and concentrated, and
the residue was purified by silica gel chromatography using 20%
EtOAc/hexane as eluent to give 3.46 g (82%) of the title compound
as a colorless oil. 1H NMR (400 MHz, CDCl3) d: 1.05–1.19 (2H, m),
1.45 (9H, s), 1.58–1.68 (3H, m), 2.20 (3H, s), 2.48 (2H, d, J = 6.6
Hz), 2.57–2.69 (2H, m), 3.47 (3H, s), 3.77 (2H, t, J = 4.9 Hz), 4.00–
4.13 (4H, m), 6.60 (1H, d, J = 1.2 Hz), 6.64 (1H, dd, J = 7.4, 1.6 Hz),
7.03 (1H, d, J = 7.8 Hz).
4.2.2. [3H]5-HT uptake assay
Phosphate-buffered saline containing 0.1 mmol/L CaCl2 and 1
mmol/L MgCl2 was used as reaction buffer. Dimethyl sulfoxide
(1
lL) or test substance was mixed with reaction buffer (199
lL),
and 50
lL of the mixture was added to human serotonin trans-
porter-expressing CHO cells cultured in 96-well assay plates. The
plates were pre-incubated at 37 °C for 10 min. During that time,
dimethyl sulfoxide or test substance was diluted with [3H]5-HT
solution in another 96-well plate. After cells pre-incubation, 50
l
L of the prepared [3H]5-HT solution containing dimethyl sulfox-
ide or test substance was added to the wells, and the mixture
4.1.19. 4-[3-(2-Methoxyethoxy)-4-methylbenzyl]piperidine (86)
A mixture of intermediate 85 (3.46 g, 9.50 mmol) and 10% HCl/
MeOH (40 mL) was stirred at room temperature for 12 h and then
concentrated. Sat. aqueous NaHCO3 was added to the residue, and
the whole was extracted with CHCl3. The organic layer was dried
over Na2SO4, filtered and concentrated to give 2.56 g (quant.) of
the title compound as a white solid. 1H NMR (300 MHz, CDCl3) d:
1.39–1.58 (2H, m), 1.58–1.74 (1H, m), 1.77 (2H, d, J = 13.9 Hz),
2.20 (3H, s), 2.52 (2H, d, J = 7.1 Hz), 2.70 (2H, t, J = 12.6 Hz), 3.33
(2H, d, J = 12.4 Hz), 3.47 (3H, s), 3.77 (2H, t, J = 4.8 Hz), 4.10 (2H, t,
J = 4.8 Hz), 6.59 (1H, s), 6.63 (1H, d, J = 7.6 Hz), 7.03 (1H, d, J = 7.6
Hz).
was incubated at 37 °C for 20 min. After incubation, the liquid layer
was discarded, and the cells were rinsed twice with 200
lL reac-
tion buffer before being lysed with 100 L of the solvable solution.
l
Radioactivity in each lysate sample was measured as described in
the previous section.
4.2.3. Guanosine 50-(
HT1A receptor
c
-thio) triphosphate, [35S]-(GTP
c
S) assay for 5-
To make up a total volume of 500
GTP S (2 mM, to measure nonspecific binding), DMSO (to measure
basal [35S]GTP
S binding) or serotonin (20 mM, to measure maxi-
mal [35S]GTP
S binding); 50 L of reaction buffer [HEPES-NaOH
buffer (20 mM, pH 7.4) containing 100 mM NaCl, 10 mM MgCl2,
0.1 mM DTT, and 1 S; and
M GDP] containing 0.5 nM [35S]GTP
447.5 L of cell membranes expressing human 5-HT1A receptors
lL, 2.5 lL of test compound,
c
c
c
l
4.1.20. 6-(2-{4-[4-Chloro-3-(2-methoxyethoxy)benzyl]piperidin-1-yl}
ethyl)-2,3-dihydro-4H-chromen-4-one (92)
l
c
l
The title compound was prepared in a manner similar to that for
the preparation of 27 using intermediates 42 and 84b. Compound 92
was obtained in 98% yield as a colorless oil. 1H NMR (400 MHz,
CDCl3) d: 1.23–1.37 (2H, m), 1.43–1.56 (1H, m), 1.60–1.78 (2H, m),
were mixed. The following manipulation was carried out as
described in the above 5-HT transporter binding assay. Intrinsic
activity was expressed as relative value of the activity of 100
serotonin, which was considered to be 100%.
lM
Please cite this article in press as: Yoshinaga H., et al. Bioorg. Med. Chem. (2018), https://doi.org/10.1016/j.bmc.2018.02.008

H. Yoshinaga et al. / Bioorganic & Medicinal Chemistry xxx (2018) xxx–xxx
13
4.2.3.1. Data analyses. The following formulae were used:
1) Basal [35S]GTP
a rat under pentobarbital anesthesia [80 mg kgꢀ1, intraperitoneal
(i.p.)]. Microdialysis was conducted on the day after surgery. A
dialysis probe (A-I-4-03; EICOM) was inserted into the guide can-
nula under light anesthesia with isoflurane and continuously per-
fused by Ringer solution (147 mmol/L NaCl, 4 mmol/L KCl, 2.3
cS binding
Basal [35S]GTP
cS binding (dpm) = Binding activity in the
DMSO group (dpm) ꢀ Binding activity in the GTP
cS group
(dpm)
mmol/L CaCl2) at 2
l
L minꢀ1 using a microsyringe pump. Micro-
L) were continuously collected for 5 min
2) Basal [35S]GTP
c
S binding
dialysate samples (10 l
Maximal [35S]GTP
c
S binding (dpm) = Binding activity in the
at 20-min interval and automatically injected into the HPLC sys-
tem. Test-compounds or vehicle were orally administered to the
rats at least 3 h after the start of perfusion, that is, when stable
HPLC baseline values for 5-HT were obtained in the dialysate sam-
ples. Measurement continued for 3 h after drug or vehicle
administration.
serotonin group (dpm) ꢀ Binding activity in the GTP
cS
group (dpm)
3) Specific binding of the test substance
Specific binding of the test substance (dpm) = Binding activ-
ity in test substance group (dpm) ꢀ Binding activity in
GTPcS group (dpm)
4) Maximal specific binding
4.2.5.2. Chromatography. The collected microdialysate samples (10
Maximal specific binding of the test substance was deter-
mined using Dx calculation (logistic curve fitting) with the
‘‘measurement value input” function in Stat Preclinica Client
Version 1.0. The direct estimation method was used. Maximal
specific binding was calculated using the logistic curve of test
substance concentrations and the specific binding values.
5) Intrinsic activity of the test substance
lL) were separated by HPLC using a PP-ODS column (EICOM), a
mobile phase containing 0.1 mol/L phosphate buffer (pH 6.0), 1%
methanol, 50 mg Lꢀ1 ethylenediamine tetraacetic acid disodium,
and 500 mg Lꢀ1 sodium 1-decanesulfonateat, and a flow rate of
0.5 mL minꢀ1. The peaks corresponding to 5-HT and dopamine
were amperometrically detected using a graphite electrode set at
400 mV with an Ag/AgCl reference electrode (RE-100; EICOM).
Online data acquisition was performed using PowerChrom soft-
ware (Version 2.2; AD Instruments Pty Ltd., Nagoya, Aichi, Japan).
Before performing the microdialysis, the retention time of the
HPLC peak for 5-HT was determined using a standard solution.
The peak height (mV) of 5-HT at each measurement was converted
into a percentage of the average of the last 4 pre-drug baseline val-
ues (percentage of baseline).
When increment in maximal [35S]GTP
c
S binding (Maximal
[
35S]GTP
c
S binding ꢀ Basal [35S]GTP
cS binding) was consid-
ered as 100%, intrinsic activity of the test substance, which is
equivalent to percentage increment in maximal specific
binding of the test substance (Maximal specific binding of
test substance ꢀ Basal [35S]GTP
cS binding), was calculated
using the following formula.
Intrinsic activity of the test substance (%) = 100 ꢁ {[Maximal
Acknowledgments
specific binding of test substance (dpm) ꢀ Basal [35S]GTP
cS bind-
ing (dpm)]/[Maximal
GTPcS binding (dpm)]}.
[
35S]GTP
c
S
binding (dpm) ꢀ Basal
[
35S]
We thank Daisuke Tanaka for his valuable suggestions through-
out the preparation of this manuscript; Hisayo Emoto for conduct-
ing the receptor binding assays; and Yoshie Nakatsuji for providing
the high resolution MS spectra.
4.2.4. Assay for CYP2D6 metabolic contribution in human liver
microsomes
A. Supplementary data
Potassium phosphate buffer solution (0.2 mL, 50 mM, pH 7.4)
containing NADPH (final concentration: 3 mM, Oriental Yeast Co.,
Ltd.), 1 mg/mL human liver microsomes (XENOTECH, LLC), and a
Supplementary data associated with this article can be found, in
the online version, at https://doi.org/10.1016/j.bmc.2018.02.008.
These data include MOL files and InChiKeys of the most important
compounds described in this article.
1 lM test substance was heated in a water bath set at 37 °C. After
15–30 min, methanol was added in a volume 3 times that of the
reaction solution, and the mixture was stirred to terminate the
reaction. The reaction solution was next centrifuged for protein
precipitation, and the supernatant was collected and subjected to
LC–MS/MS analysis. The results were as follows:
References
1. Kessler RC, Chiu WT, Demler O, Walters EE. Arch Gen Psychiatry. 2005;62:617.
2. Rush AJ, Trivedi MH, Wisniewski SR, et al. N Engl J Med. 2006;354:1231.
3. Warden D, Rush AJ, Trivedi MH, Fava M, Wisniewski SR. Curr Psychiatry Rep.
2007;9:449.
4. Reeves H, Batra S, May RS, Zhang R, Dahl DC, Li X.
2008;69:1228.
5. Artigas F, Romero L, de Montigny C, Blier P. Trends Neurosci. 1996;19:378.
6. Arborelius L, Nomikos GG, Hertel P, et al. Naunyn Schmiedebergs Arch Pharmacol.
1996;353:630.
7. Artigas F, Perez V, Alvarez E. Arch Gen Psychiatry. 1994;51:248.
8. Perez V, Gilaberte I, Faries D, Alvarez E, Artigas F. Lancet. 1997;349:1594.
9. Rutter JJ, Gundlah C, Auerbach SB. Neurosci Lett. 1994;171:183.
10. Arborelius L, Nomikos GG, Grillner P, et al. Naunyn Schmiedebergs Arch
Pharmacol. 1995;352:157.
11. Gartside SE, Umbers V, Hajos M, Sharp T. Br J Pharmacol. 1995;115:1064.
12. Yoshinaga H, Masumoto S, Koyama K, et al. Bioorg Med Chem. 2017;25:293.
13. Gibbs JP, Hyland R, Youdim K. Drug Metab Dispos. 2006;34:1516.
14. Marcusson JO, Norinder U, Hogberg T, Ross SB. Eur J Pharmacol. 1992;215:191.
15. Owens MJ, Knight DL, Nemeroff CB. Biol Psychiatry. 2001;50:345.
16. Marcusson JO, Ross SB. Psychopharmacology. 1990;102:145.
17. Bolonna AA, Kerwin RW. Br J Psychiatry. 2005;186:7.
– The test substance was quantified, and time-dependent change
in the amount of the substance remaining was logarithmically
plotted. Metabolic rate was calculated from the slope.
– The ratio of metabolic rate obtained by addition of quinidine
(final concentration: 4 lM) to the reaction solution to metabolic
rate obtained without quinidine addition was used as the rate of
metabolic contribution of enzymes other than CYP2D6. The rate
of metabolic contribution obtained by subtracting the rate of
contribution of other enzymes from the total was used as the
rate of CYP2D6 metabolic contribution. Specifically, it was cal-
culated according to the equation [rate of contribution (%) =
{1 ꢀ (metabolic rate [with quinidine]/metabolic rate [without
quinidine])} ꢁ 100].
J Clin Psychiatry.
4.2.5. Rat microdialysis
18. Shevchenko NE, Nenajdenko VG, Balenkova ES. Synthesis. 2003;8:1191.
19. Wu H, Hynes Jr J. Org Lett. 2010;12:1192.
20. Owens MJ, Knight DL, Nemeroff CB. Biol Psychiatry. 2001;50:345.
21. Poss MA, Tortolani D R, Mattson RJ, Yevich JP, PCT Int Appl WO2000/44376,
2000.
4.2.5.1. Surgery. This experiment was performed using 5–6-week
old male rats (Crlj:WI). A vertical guide cannula (AG-04; EICOM)
was implanted into the right side of the frontal cortex (3.7 mm
anterior, 3.0 mm lateral, and 1.5 mm ventral from the bregma) of
Please cite this article in press as: Yoshinaga H., et al. Bioorg. Med. Chem. (2018), https://doi.org/10.1016/j.bmc.2018.02.008

14
H. Yoshinaga et al. / Bioorganic & Medicinal Chemistry xxx (2018) xxx–xxx
22. Smith DA, Jones BC. Biochem Pharmacol. 1992;44:2089.
23. Lu Y, Shi T, Wang Y, et al. J Med Chem. 2009;52:2854.
24. Ford MC, Ho PS. J Med Chem. 2016;59:1655.
25. Wilcken R, Zimmermann MO, Lange A, Joerger AC, Boeckler FM. J Med Chem.
2013;56:1363.
28. Huges ZA, Starr KR, Scott CM, et al. Psychopharmacology. 2007;192:121.
29. Millan MJ, Newman-Tancredi A, Rivet J-M, et al.
1997;282:132.
J
Pharmacol Exp Ther.
30. Owens MJ, Morgan WN, Plott SJ, Nemeroff CB.
1997;283:1305.
J Pharmacol Exp Ther.
26. Zhou Z, Zhen J, Karpowich NK, Law CJ, Reith MEA, Wang D-N. Nat Struct Mol
Biol. 2009;16:652.
31. Yabuuchi K, Tagashira R, Ohno Y. Biog Amines. 2004;18:319.
27. Kato T, Matsumoto Y, Yamamoto M, et al. Pharmacol Res Perspect. 2015;3:
e00142.
Please cite this article in press as: Yoshinaga H., et al. Bioorg. Med. Chem. (2018), https://doi.org/10.1016/j.bmc.2018.02.008