Bicyclo[2.2.1]heptanes as novel triple re-uptake inhibitors for the treatment of depression

Article abstract of DOI:10.1016/S0960-894X(03)00660-7

A series of substituted naphthyl containing chiral [2.2.1] bicycloheptanes were prepared utilizing asymmetric Diels-Alder chemistry. This paper describes structure-activity relationships in this series. The N-methyl 2-naphthyl analogue (16d) and its des-m

Full text of DOI:10.1016/S0960-894X(03)00660-7

Bioorganic & Medicinal Chemistry Letters 13 (2003) 3277–3280
Bicyclo[2.2.1]heptanes as Novel Triple Re-uptake Inhibitors for
the Treatment of Depression
Lorraine Axford,a,y John R. Boot,a Terrence M. Hotten,a Martine Keenan,a
Fionna M. Martin,a,* Sandra Milutinovic,a Nick A. Moore,a Michael F. O’Neill,a
Ian A. Pullar,a David E. Tupper,a Kristel R. Van Belleb and Vincent Viviena,{
aLilly Research Centre, Erl Wood Manor, Sunninghill Road, Windlesham, Surrey GU20 6PH, UK
bLilly Development Centre S.A, Parc Scientifique de Louvain-la-Neuve, Rue Granbonpre, 11-B-1348 Mont Saint-Guibert, Belgium
Received 21 May 2003; revised 20 June 2003; accepted 20 June 2003
Abstract—A series of substituted naphthyl containing chiral [2.2.1] bicycloheptanes were prepared utilizing asymmetric Diels–Alder
chemistry. This paper describes structure–activity relationships in this series. The N-methyl 2-naphthyl analogue (16d) and its des-
methyl analogue (17d) are active triple re-uptake inhibitors both in vivo and in vitro.
# 2003 Elsevier Ltd. All rights reserved.
Introduction
The use of selective serotonin re-uptake inhibitors
(SSRIs) in the treatment of depression is well known,
for example Prozac (fluoxetine 1). There are a number
of issues with the use of SSRIs: latency of onset (typi-
cally 2–3 weeks), lack of efficacy (ꢀ30% non-respon-
ders) and unwanted side effects such as sexual
dysfunction, sleep disturbances, nausea, anxiety and
reduced appetite. A newer strategy to address some of
these issues has been the development of dual re-uptake
inhibitors, a combination of 5HT and NE re-uptake
inhibition (SNRIs), for example Efexor (venlafaxine 2)
and Cymbalta (duloxetine 3).
Approaches to antidepressant therapy continue to be a
significant area of central nervous system (CNS)
research. Over the past 40 years monoamine re-uptake
inhibition has been an important neuropharmacological
strategy for the treatment of depression and in the
modulation of mood.1,2 The biogenic monoamines ser-
otonin (5HT), norephinephrine (NE) and dopamine
(DA) are amongst the most intensely studied of the
200+ chemical entities known to function as neuro-
transmitters in the mammalian CNS.
*Corresponding author: Tel.: +44-1276-483000; fax: +44-1276-
483305; e-mail: martin_fionna@lilly.com
yPresent address: School of Chemistry, University of Bristol, Can-
tock’s Close, Bristol BS8 1TS.
{Present address: Bioreason, Chateau de l’Ile, Strasbourg, France.
0960-894X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00660-7

3278
L. Axford et al. / Bioorg. Med. Chem. Lett. 13 (2003) 3277–3280
Although DA dysfunction has been implicated in core
depressive symptoms the potential of DA modulation in
antidepressant therapy is less certain. A potential con-
cern is the observation that drugs which act largely by
blocking uptake of DA, for example Ritalin (methyl-
phenidate 4) and cocaine 5 have a tendency to produce
euphoria or dysphoria and psychomimetic effects.3
Others have found, for example, that the use of Bro-
mocriptine (2-bromo-a-ergocryptine, an ergopeptine
derivative and a DA agonist) and Wellbutrin (bupro-
pion 6), which has DA re-uptake activity in vitro and in
vivo, show antidepressant efficacy. Other combination
studies4À8 have indicated that addition of some affinity at
the DA uptake site may have some clinical benefit.
synthesis and SAR of a series of naphthyl bicy-
clo[2.2.1]heptanes. These derivatives were prepared by
exploiting recent developments in asymmetric Diels–
Alder reactions.
Chemistry
Compounds were prepared as described in Scheme 1
and associated catalyst details in Table 2. The requisite
dienophiles 12 were prepared by N-acylation of oxazo-
lidinones 11. The lithium salts of 11 were prepared by
reaction with nBuLi and were then reacted with the
mixed anhydride of acid 10, obtained by reaction with
pivaloyl chloride, to afford 12. Treatment of achiral 12
(R¼H, Ar=3-ClC6H4) with cyclopentadiene at low
This paper describes our efforts to obtain compounds
with equivalent functional activity at 5HT, NE and DA
uptake sites, a triple re-uptake inhibitor. Our starting
point was a historical series of Lilly compounds, trans-
(7), cis-(8) and planar-(9) bicyclo[2.2.2]octanes. These
compounds possess both NE and DA re-uptake inhi-
bitor activity (Table 1).
12
temperature using Yb(OTf)3 as Lewis acid gave race-
mic cycloadducts 13 (endo/exo ratio 4/1).13 endo/exo
isomers were separated by preparative HPLC or by
recrystallisation. Hydrogenation of the racemic endo
isomer, followed by basic hydrolysis gave the acid deri-
vative 15 which could be converted to either the dime-
thyl amino (16) or mono methyl amino (17) derivatives
in two further steps. The racemic endo isomer could also
be separated into its enantiomers by chiral HPLC14 and
We used pharmacophore models for each of the three
re-uptake binding sites to visualise the relative spatial
orientation of an aromatic moiety to a hydrophobe
(bicyclo-rings) and to the amine. These models were
built using the CatalystTM modeling package in its
common features approach.10 A training set from a
combination of in-house and literature molecules hav-
ing the basic structures described in Figure 1 and good
activity for the appropriate binding site was formed.2,10
The models predicted that the aryl group could be sub-
stituted aryl or bicycloaryl. Indeed, naphthyl had been
shown by other workers11aÀc to increase 5HT activity in
a tropane series. The models could not be used to pre-
dict the influence of amine substituents. Each bicyclo
system has cis, trans and planar isomers.
Both the bicyclo[2.2.1]heptanes and [3.2.1]octanes have
endo and exo isomers and two stereogenic centres. With
this in mind, we chose to focus our early efforts on the
bicyclo[2.2.1]heptane series. This paper will describe the
Table 1. In vitro receptor binding affinity and synaptosomal uptake9
Compd
Binding Ki (nM)a
Synap. Uptake (nM)
5-HT DA NE
5-HT
DA
NE
7
8
9
25
57
N/A
64
7.2
N/A
51
44
N/A
303 (590) 165 (340) 44 (140)
270 (3300)
(1100)
28 (69)
(40)
22 (110)
(48)
Scheme 1. Ar=1-naphthyl, 2-naphthyl, 2-naphthyl-6-OMe, 2-naph-
thyl-6-OH, 2-naphthyl-6-F. Reagents: (i) pivaloyl chloride, Et3N,
THF, À78 ꢁC (40–60%); (ii) *[cat], DCM, (80–90%); (iii) H2, Pd/C,
EtOAc (80–90%); (iv) LiOH, H2O2 (80–90%); (v) oxalyl chloride,
Me2NH or MeNH2 (60–70%); (vi) LiAlH4, Et2O(80–90%).
aValues are means of three experiments, mimium significant ratio
(MSR)<3. Numbers in brackets are historical values; N/A, no avail-
able compound for test.
Table 2. Catalyst details for Scheme 1
*[Cat] (up to 20 mol%)
Temp (ꢁC)
endo/exo ratio
Yb(OTf)3
0–20
rt
4/1
89/11
98/2
99/1
[Bis-(oxazoline)copper(II)]2SbF6
Et2AlCl and (S) aux
Et2AlCl and (R) aux
À20
À20
Figure 1. Bicyclo[2.2.2]octane, Bicyclo[2.2.1]heptane, Bicyclo[3.2.1]octane.

L. Axford et al. / Bioorg. Med. Chem. Lett. 13 (2003) 3277–3280
Table 3. In vitro receptor binding affinitya and synaptosomal uptakeb
3279
Ar
R
Stereo
Binding Ki (nM)a
Synaptosomal uptake IC50(nM)
DA
5-HTb
DAc
NEd
5-HT
NE
16a 1-Np
16b 2-Np
16c 2-Np
16d 2-Np
17b 2-Np
17d 2-Np
16e 2-Np-6-OMe
17e 2-Np-6-OMe
16f 2-Np-6-F
17f 2-Np-6-F
16g 2-Np-6-OH
Me
Me
Me
Me
H
H
Me
H
rac
rac
A
B
rac
B
B
B
B
B
52
430
18
38
25
25
41
64
60
50
72
6
183
111
149
152
266
99
93
10
20
9
50
9
nd
nd
nd
nd
nd
1514
95
93
76
87
86
nd
nd
nd
nd
nd
297
45
34
25
45
28
nd
nd
nd
nd
nd
2.8
5.8
2.3
4.9
3.6
25
68
12
120
385
316
50
Me
H
Me
2.6
4.4
47
B
nd, not determined; Np, naphthyl.
aValues are means of three experiments, MSR<3.
b3H-Citalopram.
c3H-Win35,428.
d3H-Nisoxetine.
Results and Discussion
Table 4. In vivo test results
Stereoisomer
R
In Vivo
Two principal in vitro tests were used to measure re-
uptake activity of this series; binding affinity and
synaptosomal uptake. The binding affinities were mea-
sured as inhibitory binding constants of the test drug
against a known radiolabelled re-uptake inhibitor in a
similar manner as for a postsynaptic agonist or antago-
nist. Synaptosomes are ‘pinched-off’ nerve terminals
which contain much of the pre- and post-synaptic
machinery intact. The ability of the test drug to inhibit
the re-uptake of radiolabelled (3H-tritium) transmitter
into synaptosomes were measured. The in vitro poten-
cies of a series of naphthyl bicyclo[2.2.1]heptanes 16, 17
across the three uptake transporters are described in
Table 3. Activities are given for the racemate (rac) and
the two resolved isomers: 2-(R), 3-(R) (isomer A), 2-(S),
3-(S) (isomer B). Synaptosomal data were determined for
key compounds only. The inclusion of a bicyclic 2-naph-
thyl substituent increased activity at the 5-HT transporter.
The affinities at both DA and NE transporters were
maintained leading to an in vitro triple re-uptake inhi-
bitor. Interestingly, the corresponding 1-naphthyl deriva-
tive 16a was much less active across all three transporters.
The enantiomers of the 2-naphthyl analogues 16c,d
exhibited different activities across the three transporters.
5-HTP EDmin
(mg/Kg po)
MLA EDmin
(mg/Kg po)
APOED min
(mg/Kg po)
16c A
16d B
17d B
Me
Me
H
>20
5
20
nd
40
20
nd
25
25
nd, not determined.
then processed through to final materials for test. Typical
overall yields for this seven-step sequence were 10–20%.
An alternative approach to obtain chiral bicyclo[2.2.1]-
heptanes utilized a chiral Lewis acid catalyst following
the work of Evans.15 The best catalyst for this work was
the C2-symmetric [bis-(oxazoline)copper(II)]SbF6. This
gave the cycloadducts 13 (R¼H, Ar=naphthyl) with an
endo/exo ratio of 89/11 and an ee of 97% for the endo
adduct.16 Once again, chromatography or recrystallisa-
tion was used to separate the endo/exo Diels–Alder
adducts. Finally, we tried Evans chiral auxiliary Diels–
Alder methodology as a direct means to access chiral
bicyclo[2.2.1]heptanes. Replacement of achiral 11 with
chiral 2-oxazolidinone derivatives where R=(S)-benzyl
and R=(R)-benzyl (Ar=Naphthyl)17,18 gave us chiral
12. The most appropriate Lewis acid was diethylalumi-
num chloride and the reaction proceeded at À20 ꢁC in
80% yield to give directly 2-(S), 3-(S) and 2-(R), 3-(R)
Diels-Alder adducts 13. With (S)-4-benzyl-2-oxazolidi-
nones an endo/exo ratio of 98/2 was obtained with an ee
of 95% (endo). The corresponding (R)-4-benzyl-2-oxa-
zolidinones gave a ratio of endo/exo >99/1 with an ee
>99% (endo).19 All final test compounds were prepared
as described. The chiral auxiliary method was applicable
to scale-up with no loss of chiral integrity (0.149 mol
scale, 80% yield, >98% ee). Compounds described in
Table 3 were prepared using this methodology.
Microsomal metabolism studies20 on the tertiary amino
2-naphthyl derivatives 16c,d indicated N-demethylation
as one of the primary routes of metabolism. The des-
methyl analogue 17d was prepared and found to have in
vitro triple re-uptake inhibitor activity. Another route
of metabolism was found to be hydroxylation in the 6-
position of the naphthyl group. The metabolite (16g)
was subsequently prepared and shown to be inactive in
the 5-HT binding assay. In an attempt to block this
route of metabolism, substitution in the 6-position of
the 2-naphthyl moiety by F or OMe was attempted.

3280
L. Axford et al. / Bioorg. Med. Chem. Lett. 13 (2003) 3277–3280
Whilst 5-HT affinity was maintained in the 2-naphthyl-
6-fluoro analogue (16f) NE affinity is reduced. The 2-
naphthyl-6-methoxy derivative (16e) was shown to have
reduced activities across all three transporters. Com-
pounds 16c,d and 17d were then selected for in vivo
evaluation (Table 4). Three in vivo models were used:
potentiation of 5-hydroxytryptophan-(5-HTP, the pre-
cursor of serotonin) induced serotonin syndrome-like
behaviour in mouse for 5-HT activity,21 mouse loco-
motor activity test (MLA) for DA activity22 and the
mouse apomorphine hypothermia (APO) for NE activ-
ity.23 In vivo triple re-uptake activity was confirmed for
16d and 17d. Whilst the in vitro binding profiles for the
enantiomeric pair (16c,d) had only modest differences, in
vivo the differences were more pronounced with isomer
16c weakly active (Table 4).
11. (a) Davies, H. M. L.; Kuhn, L. A.; Thornley, C.; Matasi,
J. J.; Sexton, T.; Childers, S. R. J. Med. Chem. 2001, 44, 1509.
(b) Davies, H. M. L.; Gilliat, V.; Kuhn, L. A.; Saikali, E.; Ren,
P.; Hammand, P. S.; Sexton, T.; Childers, S. R. J. Med. Chem.
1996, 39, 2554. (c) Javanmard, S.; Deutsch, H. M.; Collard,
M. M.; Kuhar, M. J.; Schweri, M. M. J. Med. Chem. 1999, 42,
4836.
12. Kobagashi, S.; Hachiya, I.; Araki, M.; Isuitani, H. Tetra-
hedron Lett. 1993, 34, 4535.
13. The endo/exo ratio of model compound 3-{[3-(3-chloro-
phenyl)bicyclo[2.2.1]hept-5-en-2-yl]carbonyl}-1,3-oxazolidin-2-
one was determined by NMR. The endo isomer refers to the
compound with the oxazolidinone on the opposite side to the
one-carbon bridge. 1H, 13C, gHSQC, COSY and gHMBC
experiments were used to confirm the proposed structure.
endo/exo isomers can be distinguished using the alkene pro-
tons: endo 1H (ppm) d 5.95 (d, 2H), 6.55 (d, 2H); exo 1H (ppm)
d 6.05 (d, 2H), 6.50 (d, 2H), ratio 79:21. The major isomer was
confirmed as endo by nOe: interactions are observed between
the bridgehead proton and the aromatic ring; such an inter-
action is not observed for the exo isomer.
In conclusion, we have demonstrated potent in vitro
triple re-uptake inhibition in a series of bicyclo[2.2.1]-
heptanes. Further SAR studies on the related planar
[2.2.1]; [2.2.2] and [3.2.1] bicycloalkanes, and their rela-
tionship to the trans endo series will be the subject of a
separate communication.
14. HPLC separation conditions for endo/exo isomers of 3-
carbonyl}-
{[3-(3-chlorophenyl)bicyclo[2.2.1]hept-5-en-2-yl]
1,3-oxazolidin-2-one: (KR60-5SIL; 90:10 hexane/DCM (ea.
2% EtOH), 40 ꢁC, 1 mL/min) exo tr 10.3 min, endo tr 11.28
min. Separation of endo/exo enantiomers by chiral HPLC:
(CHIRAPAK-AD; 95:5 hexane/IPA, (ea. 0.2 TEA), 40 ꢁC, 0.5
mL/min) exo tr (E1) 14.85 min, tr (E2) 19.32 min; endo tr (E1)
15.82 min, tr (E2) 22.2 min. Analytical column size: 250ꢂ4.6
mm 1D and for preparative work: 250ꢂ20 mm 1D.
15. (a) Evans, D. A.; Murry, J. A.; von Matt, P.; Norcross,
R. D.; Miller, S. J. Angew. Chem., Int. Ed. Engl. 1995, 34, 798.
(b) Evans, D. A.; Miller, S. J.; Lectka, T. J. Am. Chem. Soc.
1993, 115, 6460.
16. Separation of endo/exo enantiomers by chiral HPLC:
(CHIRACEL-OD; 80:20 hexane/IPA, 40 ꢁC, 0.5 mL/min) exo
tr (E1) 10.85 min, tr (E2) 14.99 min; endo tr (E1) 11.65 min, tr
(E2) 40.89 min. Absolute stereochemistry not assigned. Abso-
lute stereochemistry was assigned later by X-ray crystal-
lographic analysis of Ar=naphthyl, 16 d.
Acknowledgements
The authors would like to thank both the Spectroscopy
and Separation Sciences groups at Erl Wood for their
invaluable contributions to this work. Thanks also to
Magnus Walter for helping with the preparation of this
manuscript.
References and Notes
17. Evans, D. A.; Chapman, K. T.; Bisaha, J. J. Am. Chem.
Soc. 1988, 110, 1238.
1. Hindmarch, I. Hum. Psychopharmacol. Clin. Exp. 2001, 16,
203.
2. Owens, M. J.; Morgan, W. N.; Plott, S. J.; Nemeroff, C. B.
J. Pharmacol. Exp. Ther. 1997, 283, 1305.
3. Baldessarini, R. J. Drugs and the Treatment of Psychiatic
Disorders. In The Pharmacological Basis of Therapeutics, 7th
ed.; Goodman, A. G., Gilman L. S., Rall, T. W., Murad, F.,
Eds.; Macmillan: New York, 1985, p 416.
4. Nelson, J. C. J. Clin. Psychiatry 1998, 59, 65.
5. Masand, P. S.; Anand, V. S.; Tanquary, J. F. Depression
Anxiety 1998, 7, 89.
6. Bodkin, J. A.; Kasser, R. A.; Wines, J. D.; Gardner, D. M.;
Baldessarini, R. J. J. Clin. Psychiatry 1997, 58, 137.
7. Dewan, M. J.; Anand, V. S. J. Nervous Mental Dis. 1999,
187, 96.
8. Ashton, A. K.; Rosen, R. C. J. Clin. Psychiatry 1998, 59,
112.
18. Evans, D. A.; Chapman, K. T.; Bisaha, J. J. Am. Chem.
Soc. 1984, 106, 4261.
19. (S)-Auxilary: exo isomer not evident in NMR spectrum;
1
endo isomer alkene protons: endo H (ppm) d 6.0 (d, 2H), 6.6
(d, 2H). Separation of endo/exo enantiomers by chiral HPLC:
(CHIRACEL-OD; 90:10 hexane:IPA, 40 ꢁC, 0.5 mL/min) exo
tr(E1)À, tr(E2) 11.62 min; endo tr (E1) 14.69 min, tr (E2) 19.06
min. (R)-Auxiliary: exo isomer not evident in NMR spectrum;
1
endo isomer alkene protons: endo H (ppm) d 6.0 (d, 2H), 6.6
(d, 2H). Separation of endo/exo enantiomers by chiral HPLC:
(CHIRACEL-OD; 90:10 hexane/IPA, 40 ꢁC, 0.5 mL/min) exo
tr (E1)À, tr (E2) 10.57 min; endo tr (E1) 12.23 min, tr (E2) 15.21
min.
20. Microsomes incubated for 0.5 h (10 mM, Et2NH) HPLC/
MS profiling.
21. (a) Grahame-Smith, D. G. J. Neurochem. 1971, 18, 856.
(b) Grahame-Smith, D. G. J. Neurochem. 1971, 18, 1053. (c)
Bogeso, K. P.; Christensen, A. V.; Hyttel, J.; Liljeforss, T. J.
Med. Chem. 1985, 28, 1817.
22. Agoston, G. E.; Wu, J. H.; Izenwasser, S.; George, C.;
Kline, R. H.; Hauck Newman, A. J. Med. Chem. 1997, 40,
4329.
9. Wedney, S.; Howard, J. L.; Large, B. T.; Pullar, I. A. Bio-
chem. Pharm. 1978, 27, 2907.
10. The training set for each pharmacophore model com-
prised only active molecules. Six molecules per set with the
following in vitro cut offs; Ki (nM): 5HT <5, DA<20 and
NE<50. CatalystTM, version 4.5 (Accelerys, San Diego, CA,
USA). Common feature approach generated within Catalyst,
using BEST with an energy limit at 10 Kcal/mol, default dic-
tionary definition, feature spacing set to 150 picometer with
variable weight and tolerance.
23. (a) Peuch, A. J.; Chermant, R.; Poncelet, M.; Doare, L.;
Simon, P. Psychopharmacology 1981, 75, 84. (b) Redrobe, J. P.;
Bourin, M.; Colombel, M. C.; Baker, G. B. Psychopharmaco-
logia 1998, 138, 1.