Design and synthesis of quinolinopropellane derivatives with selective δ opioid receptor agonism

Article abstract of DOI:10.1016/j.bmcl.2014.04.098

Indolopropellane 2 was reported to show almost no binding affinity to the δ opioid receptor (DOR) in spite of the fact that 2 has both the propellane fundamental skeleton (message part) with binding ability to the opioid receptors and a possible DOR address structure (indole moiety). We developed the working hypothesis that almost no binding affinity of 2 to the DOR would be derived from its possibly stable bent conformer. To enable the propellane skeleton to adopt an extended conformation which would reasonably interact with the DOR, quinolinopropellanes 3a-d were designed which had an additional pharmacophore, quinoline nitrogen. The calculated binding free energies of ligand-DOR complexes strongly supported our working hypothesis. The synthesized quinolinopropellane 3a was a selective DOR full agonist, confirming our working hypothesis and the results of in silico investigation.

Full text of DOI:10.1016/j.bmcl.2014.04.098

Bioorganic & Medicinal Chemistry Letters xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design and synthesis of quinolinopropellane derivatives
with selective d opioid receptor agonism
Hiroshi Nagase a,b, , Ryo Nakajima a,b,d, Naoshi Yamamoto a, Shigeto Hirayama b, Takashi Iwai b,
⇑
Toru Nemoto b, Hiroaki Gouda c,e, Shuichi Hirono c, Hideaki Fujii b
a International Institute for Integrative Sleep Medicine (WPI-IIIS), University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki 305-8577, Japan
b Laboratory of Medicinal Chemistry, School of Pharmacy, Kitasato University, 5-9-1 Shirokane, Minato-ku, Tokyo 108-8641, Japan
c Laboratory of Physical Chemistry for Drug Design, School of Pharmacy, Kitasato University, 5-9-1 Shirokane, Minato-ku, Tokyo 108-8641, Japan
d Graduate School of Pure and Applied Sciences, University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki 305-8571, Japan
e School of Pharmacy, Showa University, 1-5-8 Hatanodai, Shinagawa-ku, Tokyo 142-8555, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Indolopropellane 2 was reported to show almost no binding affinity to the d opioid receptor (DOR) in
spite of the fact that 2 has both the propellane fundamental skeleton (message part) with binding ability
to the opioid receptors and a possible DOR address structure (indole moiety). We developed the working
hypothesis that almost no binding affinity of 2 to the DOR would be derived from its possibly stable bent
conformer. To enable the propellane skeleton to adopt an extended conformation which would reason-
ably interact with the DOR, quinolinopropellanes 3a–d were designed which had an additional pharma-
cophore, quinoline nitrogen. The calculated binding free energies of ligand–DOR complexes strongly
supported our working hypothesis. The synthesized quinolinopropellane 3a was a selective DOR full ago-
nist, confirming our working hypothesis and the results of in silico investigation.
Received 29 March 2014
Revised 21 April 2014
Accepted 24 April 2014
Available online xxxx
Keywords:
Opioid
DOR
Propellane structure
Binding mode
Binding free energy
Ó 2014 Elsevier Ltd. All rights reserved.
The opioid receptor, with which potent prescription analgesic
morphine interacts, is one of the promising drug targets. Opioid
DOR, and KOR with binding affinities (Ki) of 58.2 nM, 448 nM,
and 17.4 nM, respectively. These results prompted us to develop
novel ligands with the propellane skeleton. However, indolopro-
pellane 2 (Fig. 3) was reported to show almost no affinity for opioid
receptors14 although 2 has not only a propellane skeleton as a mes-
sage structure but also an indole moiety as a possible DOR address
part like the selective DOR antagonist NTI.1,2 To explain these
observations, we developed the working hypothesis that 2 could
adopt two different conformations, bent and extended (Fig. 3).
The extended conformer, which resembles the stable conformation
of NTI, could bind to the DOR whereas the bent conformer could
not. Indeed the real binding conformation of NTI unveiled by the
X-ray crystallographic analysis of the NTI–DOR complex15 is an
extended form (Fig. 4). The lack of binding of 2 to the DOR may
be derived from the adoption of the bent conformer, which may
be the more stable form.
This working hypothesis suggests that the introduction of an
additional pharmacophore into the structure of 2, which can inter-
act with the DOR to stabilize the ligand–DOR complex, would
enhance the binding affinity to the DOR. In the course of develop-
ing the selective DOR agonist TAN-67,6 we assumed an interaction
between the quinoline nitrogen and the DOR and that the interac-
tion would trigger the precise conformational change of the DOR to
receptors are classified into three major types, MOR (
l opioid
receptor), DOR (d opioid receptor), and KOR ( opioid receptor).
j
For several decades, medicinal chemists have focused on develop-
ment of selective agonists and antagonists for each opioid receptor
type. The message-address concept is a useful guideline for design
of type selective opioid ligands.1 For example, DOR antagonists
such as NTI,2 NTB,3 BNTX,4 and SB-205588,5 DOR agonists such
as TAN-67,6 SB-219825,5 NS-28,7 and KNT-127,8 KOR antagonists
such as nor-BNI9 and 50-GNTI,10 and the KOR agonist nalfura-
fine11,12 were designed and synthesized according to this concept
(Fig. 1). The message part plays an important role in exerting opi-
oid functions. In contrast to the KOR ligands, which have the 4,5-
epoxymorphinan functionality as the common message structure,
the DOR ligands possess various message structures, including
4,5-epoxymorphinan, morphinan, and 4a-phenyldecahydroiso-
quinoline structures. Recently, we found propellane 1 (Fig. 2) as a
novel message skeleton.13 The propellane 1 bound to the MOR,
⇑
Corresponding author. Tel.: +81 29 853 6437.
E-mail addresses: nagase.hiroshi.gt@u.tsukuba.ac.jp, nagaseh@pharm.kitasato-u.
ac.jp (H. Nagase).
http://dx.doi.org/10.1016/j.bmcl.2014.04.098
0960-894X/Ó 2014 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Nagase, H.; et al. Bioorg. Med. Chem. Lett. (2014), http://dx.doi.org/10.1016/j.bmcl.2014.04.098

2
H. Nagase et al. / Bioorg. Med. Chem. Lett. xxx (2014) xxx–xxx
DOR antagonists
DOR agonists
OH
N
X
OH
H
H
H
Me
O
Et
Me
Et
Me
N
N
N
N
N
NEt2
X
N
O
N
H
N
N
H
O
O
OH
OH
NS28 (X=H)
KNT-127 (X=OH)
OH
OH
SB-205588
OH
OH
SB-219825
NTI (X=NH)
NTB (X=O)
BNTX
TAN-67
KOR agonists
KOR antagonists
OH
OH
H
N
OH
OH
NH2
NH
N
O
N
N
N
N
N
H
N
H
O
Me
O
O
O
O
OH
nalfurafine
OH
HO
OH
5'-GNTI
nor-BNI
Figure 1. Structures of DOR antagonists, DOR agonists, KOR antagonists, KOR agonist. The message structures of these ligands are indicated in red.
exert the DOR agonist activity. On the basis of the above consider-
ation, we designed quinolinopropellane 3a (Fig. 3) as a DOR agonist
which contains an additional possible pharmacophoric moiety like
TAN-67. Herein, we report the conformational analyses of indolo-
and quinolinopropellanes 2 and 3a and the evaluation of the bind-
ing free energies of the ligands to the DOR. We also describe the
synthesis of the designed quinolinopropellane derivatives 3a–d
and their in vitro profiles.
N
O
OH
propellane 1
Figure 2. Structure of propellane 1.
First, to investigate our proposal related to the bent and
extended conformers of indolo- and quinolinopropellanes 2 and
3a, we performed conformational analyses of NTI, 2, and 3a using
Conformational Analyzer with Molecular Dynamics And Sampling
(CAMDAS) 2.1 program.16 When the low-energy conformers of
NTI, 2, and 3a (those within 2.5 kcal/mol of the global minimum)
were superimposed (Fig. 5), we found that the low-energy con-
formers of both 2 and 3a adopted the bent form, while those of
NTI had the extended form, as expected. The extended forms of 2
and 3a roughly appeared at the energy difference of 3–5 kcal/mol
from the global minimum.
N
N
N
H
N
OH
OH
2
3a
NH
Next, the binding modes of 2 and 3a with the DOR and their
N
N
N
H
binding free energies (DGbind values) were examined by using a
combination method of the molecular-docking calculation17 and
the molecular mechanics Generalized-Born surface area (MM-
GBSA) free energy analysis.18,19 The resulting binding modes of 2
OH
bent form
OH
extended form
and 3a are displayed in Figure 6, and their calculated
DGbind values
Figure 3. Structures of indolopropellane 2, quinolinopropellane 3a, and the bent
and extended forms of 2.
are given in Table 1. Indolopropellane 2 was found to bind with the
DOR in its extended form (Fig. 6A). This result strongly supported
our working hypothesis that the extremely low affinity of 2 to
the DOR may be due to the fact that 2 could not bind to the DOR
when the ligand was in the low-energy bent form. In other words,
the binding of 2 to the DOR would require a considerable energy
penalty to adopt the high-energy extended form, which is suited
to bind to the DOR as shown in the crystal structure of the NTI–
DOR complex15 (Fig. 4). On the other hand, the binding mode of
quinolinopropellane 3a (Fig. 6B) suggested that the extended form
of 3a could also bind to the DOR.20 Interestingly, we found that the
lone electron pair on the nitrogen atom of the quinoline ring in 3a
could form a hydrogen bonding interaction with the NH+3 of the
Lys214 residue. A similar hydrogen bond was not observed in the
2–DOR complex, because 2 possessed the indole ring which lacks
a lone electron pair. Due to the additional hydrogen bonding inter-
L125
I304
L300
Y308
D128
NTI
W284
I277
F280
Y129
V281
K214
W274
M132
T285
H278
V217
action, the electrostatic interaction (
suggested to be much greater than that of 2 (Table 1). This situa-
tion inevitably led to a much better Gbind value for 3a. Taken
together, the above observations suggested that the additional
DEelec) of 3a with the DOR was
D
Figure 4. The binding mode of NTI observed in the X-ray structure of the NTI–DOR
complex.
Please cite this article in press as: Nagase, H.; et al. Bioorg. Med. Chem. Lett. (2014), http://dx.doi.org/10.1016/j.bmcl.2014.04.098

H. Nagase et al. / Bioorg. Med. Chem. Lett. xxx (2014) xxx–xxx
3
A
B
C
NTI
2
3a
Figure 5. The superimpositions of the low-energy conformers of NTI, 2, and 3a.
L300
I304
L125
A
O
N
N
N
Y308
D128
a)
W284
MeO
RO
I277
4a
5a (R=Me)
3a (R=H)
F280
b)
Y129
Scheme 1. Synthesis of quinolinopropellane 3a. Reagents and conditions: (a) 2-
aminobenzaldehyde, MeSO3H, EtOH, reflux; (b) HClꢁpyridine, 180 °C.
V281
K214
W274
V217
M132
T285
H278
I304
hydrogen bonding interaction in the 3a–DOR complex might com-
pensate for any energy penalty, allowing 3a to adopt the high-
energy extended form upon binding. The obtained binding mode
of quinolinopropellane 3a with the DOR included the hydrogen
bonding with the Lys214 residue, whereas a corresponding interac-
tion with the Lys214 residue was not observed in the crystal struc-
ture of the NTI (DOR antagonist)–DOR complex.15 In the course of
DOR agonist TAN-67 discovery, the hydrogen bonding with the
DOR was proposed to be important in producing the DOR agonist
activity.6 Therefore, quinolinopropellane 3a was expected to pro-
duce DOR agonism.
L300
L125
B
Y308
W284
D128
I277
F280
Y129
To confirm the in silico outcomes, we synthesized quinolino-
propellanes 3a–d. A quinolone moiety was constructed by the
reaction of compound 4a with 2-aminobenzaldehyde in the pres-
ence of methanesulfonic acid (Friedländer quinoline synthesis21),
followed by demethylation of the O-Me group in 5a to provide
3a (Scheme 1). Via nor-compound 4b obtained from 4a by a reac-
tion with Troc-Cl and subsequent treatment with Zn/AcOH, the
V281
K214
W274
V217
M132
T285
H278
Figure 6. The binding modes of 2 (A) and 3a (B) with the DOR determined by our
docking procedure. Hydrogen-bonding interactions are indicated by red dashed
lines.
O
N
H
a)
b) or c)
or d)
4a
MeO
4b
Table 1
Energy contributions (kcal/mol) to the binding free energy of 2 and 3a to the DOR
Contribution
2
3a
Differencea
N
N
R
N
N
R
b
D
D
D
D
D
D
Eint
3.19
2.80
0.39
ꢀ1.44
13.54
ꢀ2.93
1.87
e) or f)
c
EVDW
ꢀ50.03
ꢀ11.93
11.06
ꢀ48.59
ꢀ25.47
13.99
d
Eelec
e
GGB
MeO
HO
f
GSA
ꢀ6.28
ꢀ8.15
5b (R=1-OH-CPM)
5c (R=Bn)
5d (R=Me)
3b (R=1-OH-CPM)
g
Gbind
ꢀ53.99
ꢀ65.42
11.43
3c (R=Bn)
3d (R=Me)
a
b
c
d
e
f
Differences of energy contributions of 2 and 3a.
Internal contributions from bond, angle, dihedral terms.
Nonbonded van der Waals.
Nonbonded electrostatics.
Electrostatic component to solvation.
Nonpolar component to solvation.
Scheme 2. Synthesis of quinolinopropellanes 3b–d. Reagents and conditions: (a) (i)
Troc-Cl, K2CO3, CH2Cl2, rt; (ii) Zn, AcOH, rt; (b) (i) 2-aminobenzaldehyde, MeSO3H,
EtOH, reflux; (ii) 1-acetoxy-1-cyclopropanecarboxylic acid, EDCIꢁHCl, DMAP, DMF,
rt; (iii) LiAlH4, H2SO4, THF, rt; (c) (i) 2-aminobenzaldehyde, MeSO3H, EtOH, reflux;
(ii) BnBr, K2CO3, DMF, rt; (d) (i) MeI, Et3N, CH2Cl2, rt; (ii) 2-aminobenzaldehyde,
MeSO3H, EtOH, reflux; (e) BBr3, CH2Cl2, 0 °C to rt; (f) HClꢁpyridine, 180 °C.
g
Total change of free energy in binding.
Please cite this article in press as: Nagase, H.; et al. Bioorg. Med. Chem. Lett. (2014), http://dx.doi.org/10.1016/j.bmcl.2014.04.098

4
H. Nagase et al. / Bioorg. Med. Chem. Lett. xxx (2014) xxx–xxx
Table 2
selective DOR full agonist, confirming our working hypothesis
and the results of in silico investigation.
Binding affinities of quinolinopropellanes 3a–d to the opioid receptorsa
Compound
R
Ki (nM)
Selectivity
Acknowledgments
MORb DORc
KORd MOR/DOR KOR/DOR
3a
3b
3c
3d
CPMe
112
0.941 84.6
119
378
2.42
1.63
89.9
801
18.8
104
We acknowledge the financial supports from Shorai Foundation
for Science, Uehara Memorial Foundation and Adaptable and
Seamless Technology Transfer Program through target-driven
R&D, JST (AS2315040G) (to H.N., S.H., T.I., T.N., and H.F.). We also
acknowledge the Institute of Instrumental Analysis of Kitasato Uni-
versity, School of Pharmacy for its facilities.
1-OH-CPMe 415
1.10
31.6
1.88
879
594
195
Bn
Me
76.3
3.06
a
b
c
Binding assays were carried out in duplicate.
[3H] DAMGO was used.
[3H] DPDPE was used.
d
e
[3H] U-69593 was used.
CPM: cyclopropylmethyl.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2014.04.
098. These data include MOL files and InChiKeys of the most
important compounds described in this article.
conversion of N-substituents was carried out by two methods: (1)
a Friedländer quinoline synthesis and followed by amidation with
carboxylic acid and reduction of the obtained amide by alane,22
and (2) a sequential Friedländer quinoline synthesis and alkylation
(Scheme 2).
References and notes
The binding affinities of the prepared quinolinopropellanes 3a–
d to the opioid receptors were evaluated by competitive binding
assays (Table 2). As we expected, all the tested quinolinopropell-
anes 3a–d exhibited high binding affinities and selectivities for
the DOR. Quinolinopropellane 3a with the N-cyclopropylmethyl
group had the highest binding affinity for the DOR, while N-(1-
hydroxycyclopropylmethyl) derivative 3b showed the highest
selectivity for the DOR, although its binding affinity for the DOR
was slightly decreased compared with that of 3a. Although a pro-
pellane 1 derivative with the N-methyl substituent was reported
to be a strong binder to the MOR (Ki = 3.6 nM) with 122- and 71-
fold greater selectivities over the DOR and KOR,13 respectively, N-
methylquinolinopropellane 3d exerted low but evident selectivity
for the DOR.
1. Portoghese, P. S. Trends Pharmacol. Sci. 1989, 10, 230.
2. Portoghese, P. S.; Sultana, M.; Takemori, A. E. J. Med. Chem. 1990, 33, 1714.
3. Sofuoglu, M.; Portoghese, P. S.; Takemori, A. E. J. Pharmacol. Exp. Ther. 1991, 257,
676.
4. Portoghese, P. S.; Sultana, M.; Nagase, H.; Takemori, A. E. Eur. J. Pharmacol.
1992, 218, 195.
5. Dondio, G.; Ronzoni, S.; Eggleston, D. S.; Artico, M.; Petrillo, P.; Petrone, G.;
Visentin, L.; Farina, C.; Vecchietti, V.; Clarke, G. D. J. Med. Chem. 1997, 40, 3192.
6. Nagase, H.; Kawai, K.; Hayakawa, J.; Wakita, H.; Mizusuna, A.; Matsuura, H.;
Tajima, C.; Takezawa, Y.; Endo, T. Chem. Pharm. Bull. 1998, 46, 1695.
7. Nagase, H.; Osa, Y.; Nemoto, T.; Fujii, H.; Imai, M.; Nakamura, T.; Kanemasa, T.;
Kato, A.; Gouda, H.; Hirono, S. Bioorg. Med. Chem. Lett. 2009, 19, 2792.
8. Nagase, H.; Nemoto, T.; Matsubara, A.; Saito, M.; Yamamoto, N.; Osa, Y.;
Hirayama, S.; Nakajima, M.; Nakao, K.; Mochizuki, H.; Fujii, H. Bioorg. Med.
Chem. Lett. 2010, 20, 6302.
9. Portoghese, P. S.; Nagase, H.; Lipkowski, A. W.; Larson, D. L.; Takemori, A. E. J.
Med. Chem. 1988, 31, 836.
10. Stevene, W. C.; Jones, R. M., Jr.; Subramanian, G.; Metzger, T. G.; Ferguson, D.
M.; Portoghese, P. S. J. Med. Chem. 2000, 43, 2759.
11. Nagase, H.; Hayakawa, J.; Kawamura, K.; Kawai, K.; Takezawa, Y.; Matsuura, H.;
Tajima, C.; Endo, T. Chem. Pharm. Bull. 1998, 46, 366.
12. Kawai, K.; Hayakawa, J.; Miyamoto, T.; Imamura, Y.; Yamane, S.; Wakita, H.;
Fujii, H.; Kawamura, K.; Matsuura, H.; Izumimoto, N.; Kobayashi, R.; Endo, T.;
Nagase, H. Bioorg. Med. Chem. 2008, 16, 9188.
13. Yamamoto, N.; Fujii, H.; Nemoto, T.; Nakajima, R.; Momen, S.; Izumimoto, N.;
Hasebe, K.; Mochizuki, H.; Nagase, H. Bioorg. Med. Chem. Lett. 2011, 21, 4104.
14. Li, F.; Gaob, L.; Yin, C.; Chen, J.; Liu, J.; Xie, X.; Zhang, A. Bioorg. Med. Chem. Lett.
2009, 19, 4603. We also obtained the same experimental results at the same
time as those.
15. Granier, S.; Manglik, A.; Kruse, A. C.; Kobilka, T. S.; Thian, F. S.; Weis, W. I.;
Kobilka, B. K. Nature 2012, 485, 400.
16. Tsujishita, H.; Hirono, S. J. Comput. Aided Mol. Des. 1997, 11, 305.
17. Docking was done with the induced fit docking protocol of Schrödinger Suite
2010.
18. Massova, I.; Kollman, P. A. Perspect. Drug Discovery Des. 2000, 18, 113.
19. Kollman, P. A.; Massova, I.; Reyes, C.; Kuhn, B.; Huo, S.; Chong, L.; Lee, M.; Lee,
T.; Duan, Y.; Wang, W.; Donini, O.; Cieplak, P.; Srinivasan, J.; Case, D. A.;
Cheatham, T. E., 3rd. Acc. Chem. Res. 2000, 33, 889.
20. The stable conformers of trans-isomers of morphinans are expected to be
extended conformations and could fit to delta receptor. On the other hand, the
stable ones of cis-compounds like propellanes may be bent form and could not
fit to.
21. Cheng, C.-C.; Yan, S.-J. In Org. React.; Dauben, W. G., Ed.; John Wiley & Sons Inc.:
Canada, 1982; Vol. 28, pp 37–201.
22. Greiner, E.; Folk, J. E.; Jacobson, A. E.; Rice, K. C. Bioorg. Med. Chem. 2004, 12,
233.
We next assessed the functional activities of a selected com-
pound 3a, which exhibited the highest binding affinity for the
DOR, by [35S]GTP
cS binding assays. As we expected, 3a exhibited
DOR full agonist activity (EC50 (DOR) = 2.50 nM, Emax
(DOR) = 88%, EC50 (MOR) = 197 nM, Emax (MOR) = 56%, EC50
(KOR) = 836 nM, Emax (KOR) = 57%). The functional DOR selectivi-
ties of 3a were comparable to or higher than its binding DOR selec-
tivities (EC50 ratio: MOR/DOR = 78.8, KOR/DOR = 334). The
outcomes of in vitro evaluations strongly supported our working
hypothesis and the in silico experimental results. Moreover, these
observations suggest that the hydrogen bonding interaction
between a ligand and the Lys214 residue in the DOR plays a crucial
role in not only obtaining strong binding ability but also exerting
DOR agonist activity.
In conclusion, we have developed the working hypothesis that
almost no binding affinity of indolopropellane 2 to the DOR would
be derived from its possibly more stable bent conformer. To enable
the propellane skeleton to adopt an extended conformation, which
could reasonably be expected to interact with the DOR, quinolino-
propellanes 3a–d were designed which had an additional pharma-
cophore, the quinoline nitrogen. The calculated binding free
energies of ligand–DOR complexes strongly supported our working
hypothesis. The synthesized quinolinopropellane 3a was
a
Please cite this article in press as: Nagase, H.; et al. Bioorg. Med. Chem. Lett. (2014), http://dx.doi.org/10.1016/j.bmcl.2014.04.098