Ligand conformation has a definitive effect on 5-HT1A and serotonin reuptake affinity

Article abstract of DOI:10.1016/j.bmcl.2004.06.050

Conformationally constrained cyclohexanes and cyclohexenes were prepared and tested for 5-HT_1A and SSRI activity. The conformation of the dioxyaryl ring with respect to the cyclohexane ring is a key factor for selecting the relative potency of the compounds at the two receptors studied. Conformationally constrained aryl cyclohexanes and cyclohexenes based on aryl cyclohexanols 1 were prepared. Locking the aryl ring in plane with the cyclohexane moiety provided potent SSRIs 3. Conversely, fixing the aryl ring perpendicular to the cyclohexane ring via a spiro lactone provided balanced 5-HT_1A antagonists with mid-nanomolar range SSRI potency (compounds 2).

Full text of DOI:10.1016/j.bmcl.2004.06.050

Bioorganic & Medicinal Chemistry Letters 14 (2004) 4467–4470
Ligand conformation has a definitive effect on 5-HT1A and serotonin
reuptake affinity
Kenneth M. Boy,* Michael Dee, Joseph Yevich, John Torrente, Qi Gao, Lawrence Iben,
Arlene Stark and Ronald J. Mattson
Bristol-Myers Squibb Pharmaceutical Research Institute, 5 Research Parkway, Wallingford, CT 06492, USA
Received 14 May 2004; revised 16 June 2004; accepted 16 June 2004
Available online 17 July 2004
Abstract—Conformationally constrained aryl cyclohexanes and cyclohexenes based on aryl cyclohexanols 1 were prepared. Locking
the aryl ring in plane with the cyclohexane moiety provided potent SSRIs 3. Conversely, fixing the aryl ring perpendicular to
the cyclohexane ring via a spiro lactone provided balanced 5-HT1A antagonists with mid-nanomolar range SSRI potency (comp-
ounds 2).
Ó 2004 Elsevier Ltd. All rights reserved.
1. Introduction
receptor prevents the decrease in neuronal firing nor-
mally observed upon autoreceptor occupancy by
serotonin. In keeping with the above discussion, clinical
co-administration of a SSRI and a 5-HT1A antagonist
reduces the time to onset of efficacy.7 The goal of our re-
search program is to incorporate both activities in a sin-
gle molecule.
Depression is a mood disorder, which affects an esti-
mated 121 million people worldwide.1 In the U.S. alone,
depression has a monetary impact in the tens of billions
of dollars per year in medical costs and lost productiv-
ity, and an immeasurable toll on quality of life.2 The
molecular basis for depression is not fully understood;
however, deficits in the activity of serotonin-mediated
neurons in the brain are clearly central to the disease.3
Other neurotransmitter systems, such as norepinephrine
and dopamine are also implicated.4 The introduction of
serotonin-selective reuptake inhibitors (SSRIÕs) to the
market has done much to treat the symptoms of depres-
sion; however, their efficacy is not yet optimal. One
major drawback of these compounds is the delayed on-
set of efficacy, which can be as long as two to four
weeks.5 During this time, patients may choose to discon-
tinue drug therapy, or try to harm themselves. It is
understood that the desired accumulation of serotonin
in the synapse as a result of serotonin reuptake blockade
also increases occupancy of the 5-HT1A autoreceptor on
the presynaptic neuron, causing a decrease in neuron fir-
ing. With continuous SSRI treatment, the autoreceptor
becomes desensitized, and neuronal firing is restored
contemporaneously with observed clinical benefit.6
Administration of an antagonist of the 5-HT1A auto-
2. Design and synthesis
In the course of previous investigations on serotonin
modulators, benzylpiperidine and benzylpiperazine moie-
ties attached to the 4-position of arylcyclohexanol (and
congeners thereof) (compounds 1) provided a basis for
O
A
A
n
O
O
O
n
O
N
NH
O
OH
O
Z
O
N
Y
N
Y
N
Y
N
Y
Z
Z
Z
2
3
4
1
Keywords: Serotonin; SSRI; 5-HT1A; Depression.
*
Figure 1. Prototype 1 and target molecules 2–4. ÔAÕ is dimethoxy or
(m)ethylenedioxy.
Corresponding author. Tel.: +1-203-677-6278; fax: +1-203-677-
7702; e-mail: boyk@bms.com
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.06.050

4468
K. M. Boy et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4467–4470
SAR investigation (Fig. 1).8 Based on these early studies,
we developed a hypothesis that the angle formed between
the dialkoxyphenyl head group and the cyclohexane ring
was a determinant of potency at h-SERT and 5-HT1A
.
Specifically, we sought an optimized angle, which would
maximize both activities simultaneously.
Figure 2. ORTEP drawing of 2b with thermal ellipsoids at 40%
probability for non-H atoms and open circles for H-atoms.
We appreciated that compounds 1 sampled conforma-
tions about the cyclohexanol/aryl ring bond in solution.
We wished to add conformational constraint to this
bond to more closely match the (unknown) bound-state
conformation of the ligand. To study the issue in detail,
we envisioned two classes of molecules with fixed and
orthogonal relationships (compounds 2 and 3). In com-
pound 2, the aryl ring occupies a perpendicular (ꢀ90°)
relationship to the cyclohexanol unit, while in
compound 3, the aryl ring is fixed in a planar conforma-
tion with respect to the cyclohexene ring. Taken to-
gether, these molecules probe the binding requirements
of the h-SERT and 5-HT1A proteins.
no trans isomer detected. This was verified by single-
crystal X-ray analysis of compound 2b10 (Fig. 2).
Fused polycycles 3 were synthesized by effecting a reduc-
tive amination of 1,4-cyclohexanedione monoethylene
ketal with appropriately substituted benzylpiperazines
and benzylpiperidines. Deprotection of the ketal, fol-
lowed by enamine formation provided intermediates
14. Condensation with appropriate aryl isocyanates in
refluxing chloroform, followed by acid-mediated ring
closure provided the desired compounds 3 (Scheme 2).
Compounds 2 were synthesized starting from the
5-bromobenzo[d][1,3]dioxole-4-carbaldehyde9
Binding at the h-SERT and 5-HT1A receptors were
measured11 by published methodology12,13 (using [3H]-
citalopram and ( )-[3H]-8-hydroxydipropylaminotet-
ralin, respectively) and the data expressed as IC50 values
with the radioligand at the Kd concentration (Table 1).
Compounds 2 may be classified as dual 5-HT1A and h-
SERT modulators, the 5-HT1A activity being dominant
and in the lower-to mid-nanomolar range. Selectivities
range from ꢀ2.5-fold (2a) to ꢀ40-fold (2s, 2t). Consist-
ent with the SAR of the cyclohexanol analogs, 2-halo
substituents on the benzylpiperidine/piperazine moiety
known
(5) (Scheme 1). Silver nitrate oxidation to the corre-
sponding bromoacid and standard functional group
manipulation furnished the amide 7. Halogen–metal
exchange afforded a stable nucleophile, producing the
tertiary alcohol adduct 8 upon addition of 1,4-cyclohex-
anedione monoethylene ketal. Treatment with sodium
methoxide formed the lactone 9, which was then sub-
jected to 3M HCl and acetone to provide the ketone
10. Reductive amination with appropriately substituted
benzylpiperazines and benzylpiperidines utilizing a tita-
nium isopropoxide/sodium borohydride protocol pro-
vided the desired cis-cyclohexane spirolactones 2, with
O
O
O
O
O
O
O
O
O
n
n
n
H
N
O
O
O
a
b
c
OH
NEt2
a
b
c
Br
Br
N
Y
Br
O
N
Y
Y
5
6
7
Z
O
Z
O
O
Z
n
n
O
O
NEt2
d
e
O
11
12
13
O
HO
O
O
8
9
A
A
O
OH HN
NMe
NH
O
2
O
n
O
O
O
O
O
O
n
O
f
d
e
O
N
Y
N
Y
N
Y
NH
Y
N
10
2
Z
11
O
Z
Y
Z
Z
Z
15
14
3
Scheme 1. Reagents and conditions: (a) AgNO3, NaOH, rt, 1h; (b) i.
SOCl2, CH2Cl2, reflux, ii. HNEt2, CH2Cl2, reflux (88% from 5); (c) s-
BuLi, 1,4-cyclohexanedione monoethylene ketal, À78° C to rt; (d)
NaOMe, MeOH, (78%, two steps); (e) 3M HCl, acetone, reflux, 93%;
(f) i. Ti(Oi-Pr)4, amine, 120°C, ii. NaBH4, EtOH, (18%, two steps).
Scheme 2. Reagents and conditions: (a) 1,4-cyclohexanedione mono-
ethylene ketal, NaBH(OAc)3, THF, HOAc, 77%; (b) 3M HCl,
acetone, reflux, 100%; (c) Me2NSiMe3, p-TsOH, 100%; (d) aryl
isocyanate, CHCl3, reflux, 93%; (e) HCl, MeOH, reflux, 6%.

K. M. Boy et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4467–4470
Table 1. Inhibition assay results for compounds 2–4
4469
Compds
n
Y
Z
h-SERT, IC50, nMa
5-HT1A, IC50, nMa
18
9.6
2a
2b
2c
2d
2e
2f
1
N
2-I
2-I
48
54
2
N
1
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
N
2-Br
2-Br
45
73
9.6
9.8
2
1
2-Cl
2-Cl
140
100
410
480
180
510
150
59
10
2
19
14
12
67
43
2g
2h
2i
1
2-F
2-F
2
1
2-CF3
2-CF3
3-OMe
3-OMe
2-Br-5-F
2-Br-5-F
2,5-Di-F
2,5-Di-F
2-F-5-OMe
2-F-5-OMe
2-Br-5-OMe
2,5-Di-Cl
2-I
2j
2
2k
2l
1
6.7
4.3
14
2
2m
2n
2o
2p
2q
2r
2s
2t
1
82
77
2
11
22
14
1
250
110
400
430
91
2
1
NT
2
11
1
2.4
2.3
1
92
23
3a
3o
3u
4o
1
>1000
440
1
CH
N
2,5-Di-F
2-I
2,5-Di-F
5.0
5.5
14
(OMe)2
1
>1000
450
CH
a See Ref. 11; (NT=not tested).
provide potently active molecules at 5-HT1A, with the 3-
methoxy substituent also providing potent agents (2k,
2l). Superposition of these substituents (2s) provided a
very potent 5-HT1A modulator, however the
h-SERT activity remained at a modest level. Ring size
at the alkylated catechol (n in Scheme 1) effected binding
in an unpredictable manner, although the effects at 5-
HT1A were less than at h-SERT (2c vs 2d and 2g vs
2h). Compounds 3 showed a reversal in selectivity, being
devoid of activity at 5-HT1A, while enjoying potent
activities at h-SERT. O-Methylation of the lactam gave
the corresponding cyclic imidate 4o, which retained po-
tency at h-SERT. The retention of activity in 4 versus 3
demonstrates that the hydrogen-bonding network by the
lactam functionality itself is not responsible for the high
levels of h-SERT activity observed. Taken together, this
data provides evidence that the conformation of the aryl
ring relative to the cyclohexyl ring is an important factor
in determining relative potencies at h-SERT and
By manipulating the topology of the ligand via judicious
conformational constraint, potent, dual antagonists at
the 5-HT1A and h-SERT receptors were obtained.
References and notes
1. World Health Organization WHO Fact Sheet Number
265; The World Health Organization: Geneva, Switzer-
land, 2001.
2. Thase, M. E. J. Clin. Psychiatry 2003, 64(Suppl. 13),
18–25.
3. Fava, M. J. Clin. Psychiatry 2003, 64(Suppl. 13), 26–
29.
4. Stahl, S. M.; Grady, M. M. J. Clin. Psychiatry 2003,
64(Suppl. 13), 13–17.
5. Schatzberg, A. F. J. Clin. Psychiatry 2003, 64(Suppl. 13),
30–37.
6. Blier, P.; Bergeron, R.; de Montigny, C. Neuropsycho-
pharmacology 1997, 16, 333–338.
7. Blier, P.; Bergeron, R. J. Clin. Psychopharmacol. 1995, 15,
217–222.
5-HT1A
.
Compounds 2a and 2l were shown to be partial anta-
gonists at the 5-HT1A autoreceptor by published
methods,14 having intrinsic activities of 0.1 and 0.37,
respectively.
8. Mattson, R. J.; Catt, J. D.; Sloan, C. P.; Gao, Q.; Carter,
R. B.; Gentile, A.; Mahle, C. D.; Matos, F. F.; McGovern,
R.; VanderMaelen, C. P.; Yocca, F. D. Bioorg. Med.
Chem. Lett. 2003, 13, 285–288.
9. Mattson, R. J.; Sloan, C. P.; Lockhart, C. C.; Catt, J. D.;
Gao, Q.; Huang, S. J. Org. Chem. 1999, 64, 8004–
8007.
10. Crystallographic data has been deposited with the Cam-
bridge Crystallographic Data Centre as supplementary
publication number CCDC 241821. Copies of the data can
be obtained, free of charge, via the Internet at http://
www.ccdc.cam.ac.uk.
Manipulation of the global conformation of serotonin
modulators provided two unique series of compounds.
In the first, a balance of nanomolar 5-HT1A (antagonist)
and h-SERT (antagonist) activities was achieved by fix-
ing the aryl ring in a perpendicular orientation via a spiro
lactone. The SSRI activity was shown to improve dra-
matically at the expense of 5-HT1A activity upon rotation
of the aryl ring to an orthogonal, planar, conformation.
11. 5-HT1A and SERT IC50 values are the mean of two
determinations run at five different concentrations with the

4470
K. M. Boy et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4467–4470
radioligand at the Kd concentration. Each experiment was
carried out from 1 to 6 times. Standard errors were
typically within 20% of the mean value.
H. Synapse 2003, 50, 233–239; (b) Poss, M. A.; Tortolani,
D. R.; Mattson, R. J.; Yevich, J. P. U.S. Patent 6,225,324,
example 11.
12. Mahle, C. D.; Wiener, H. L.; Yocca, F. D.; Maayani, S.
J. Pharmacol. Exp. Ther. 1992, 263, 1275.
13. (a) Rothman, R. B.; Jayanthi, S.; Wang, X.; Dersch, C.
M.; Cadet, J. L.; Prisinzano, T.; Rice, K. C.; Baumann, M.
14. (a) Yocca, F. D.; Iben, L.; Meller, E. Mol. Pharmacol.
1992, 41, 1066–1072; (b) Zgombick, J. M.; Beck, S. G.;
Mahle, C. D.; Craddock-Royal, B.; Maayani, S. Mol.
Pharmacol. 1989, 35, 484–494.