Synthesis, structural identification, and ligand binding of tropane ring analogs of paroxetine and an unexpected aza-bicyclo[3.2.2]nonane rearrangement product

Article abstract of DOI:10.1016/j.bmc.2005.01.046

The structural requirements for high affinity at the serotonin transporter (5-HTT) have been investigated through the preparation of rigid paroxetine analogs. Tropane-derived analogs (4a-i) of paroxetine (2) were designed and synthesized as potential inhibitors of serotonin reuptake based on the structural and biological similarity between the two compound classes. Overall, the affinity of tropane-derived analogs at the 5-HTT was found to be at least an order of magnitude lower than that of paroxetine and ranged from 2-400 nM. The reduced affinity at the 5-HTT may be attributed to the inability of the rigid tropane-derived analogs to adopt conformations favored by the 5-HTT. Within the series of tropane analogs, the 2β,3β- and 2β,3α-isomers, 4a and 4d, were the most potent at the DAT and NET and are also significantly more potent than paroxetine (2) suggesting that their reduced conformational flexibility maximizes residence time in conformations favored by these transporters. Examination of the previously published preparation and structural assignment of 4a by additional NMR and X-ray crystallographic data has established that nucleophilic addition to the intermediate 2β- methanesulfonyloxymethyl-3β-(4-fluorophenyl)tropane unexpectedly provided the aza-bicyclo[3.2.2]nonane derivative 10a.

Full text of DOI:10.1016/j.bmc.2005.01.046

Bioorganic & Medicinal Chemistry 13 (2005) 2439–2449
Synthesis, structural identification, and ligand binding of
tropane ring analogs of paroxetine and an unexpected
aza-bicyclo[3.2.2]nonane rearrangement product
Scott P. Runyon,a Jason P. Burgess,a Philip Abraham,a Kathryn I. Keverline-Frantz,a
b
Judy Flippen-Anderson, Jeffrey Deschamps, Anita H. Lewin, Hernan A. Navarro,
b
a
a
´
*
John W. Boja,c Michael J. Kuharc andF. Ivy Carroll
a,
aChemistry and Life Sciences, Research Triangle Institute, 3040 Cornwallis Road, Research Triangle Park, PO Box 12194,
North Carolina, NC 27709-2194, USA
bLaboratory for the Structure of Matter, Naval Research Laboratory, 4555 Overlook Avenue, Washington, DC 20375-5341, USA
cNeuroscience Branch, National Institute on Drug Abuse Addiction Research Center, 5500 Nathan Shock Drive, Baltimore,
Maryland 21224, USA
Received20 October 2004; accepted24 January 2005
Abstract—The structural requirements for high affinity at the serotonin transporter (5-HTT) have been investigatedthrough the
preparation of rigidparoxetine analogs. Tropane-derivedanalogs ( 4a–i) of paroxetine (2) were designed and synthesized as potential
inhibitors of serotonin reuptake basedon the structural andbiological similarity between the two compoundclasses. Overall, the
affinity of tropane-derived analogs at the 5-HTT was found to be at least an order of magnitude lower than that of paroxetine
and ranged from 2–400 nM. The reduced affinity at the 5-HTT may be attributed to the inability of the rigid tropane-derived ana-
logs to adopt conformations favored by the 5-HTT. Within the series of tropane analogs, the 2b,3b- and2 b,3a-isomers, 4a and 4d,
were the most potent at the DAT andNET andare also significantly more potent than paroxetine ( 2) suggesting that their reduced
conformational flexibility maximizes residence time in conformations favored by these transporters. Examination of the previously
publishedpreparation andstructural assignment of 4a by additional NMR and X-ray crystallographic data has established that
nucleophilic addition to the intermediate 2b-methanesulfonyloxymethyl-3b-(4-fluorophenyl)tropane unexpectedly provided the
aza-bicyclo[3.2.2]nonane derivative 10a.
ꢀ 2005 Elsevier Ltd. All rights reserved.
1. Introduction
serotonin (5-HT) neurotransmission by interacting at
the serotonin transporter (5-HTT) as well. Thus, the 5-
HTT may also be involvedin some way, perhaps via
modulation of cocaineÕs action.6,7
In order to understand the biochemical mechanism re-
lated to its addictive properties, the molecular site where
cocaine (1) interacts to produce its physiological effects
had to first be identified. In our 1992 perspective paper,1
we summarizedthe information that strongly suggested
the dopamine transporter (DAT) as the cocaine receptor
responsible for its abuse. Over the past several years
numerous studies have supported the hypothesis impli-
cating the DAT as a key target for cocaine abuse.2–5
However, it is well known that cocaine also influences
Paroxetine (2) is a selective 5-HTT reuptake inhibitor
with demonstrated antidepressant activity in humans.8
Since the 3b-phenylnortropane 3 also possesses high
affinity for the 5-HTT andshares some structural fea-
tures with paroxetine,9 we carriedout an investigation
of rigidtropane-derivedparoxetine analogs
4a–i10 (see
Table 2 for structures) to gain a better understanding
of the structural requirements necessary for high affinity
at the 5-HTT.
Keywords: Paroxetine; Monoamine transporters; Tropanes; Aza-
bicyclo[3.2.2]nonanes.
During a subsequent investigation of 3b-phenyltropanes
10
*
similar to those we hadreported,
Ogier et al. noted
Corresponding author. Tel.: +1 919 541 6679; fax: +1 919 541
8868; e-mail: fic@rti.org
that a key intermediate, 2b-methanesulfonyloxymethyl-
0968-0896/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2005.01.046

2440
S. P. Runyon et al. / Bioorg. Med. Chem. 13 (2005) 2439–2449
H
H3C
H
O
N
N
N
O
CO2CH3
CO2CH3
O
O
O
F
Cocaine
Paroxetine
1
2
3
F
O
O
O
O
O
CH3
H
N
H
S
O
N
O
O
O
H
X
N
F
CH3
4a
5
6a, X = CH3
b, X = F (1R)
c, X = F (1S)
O
O
R1
N
O
R2
F
N
10a
F
14,
R1 = CH3C6H4SO2, R2 = CO2CH3
15, R1 = CH3C6H4SO2, R2 = CH2OH
O
O
16,
R
1 = CH3C6H4SO2, R2 = CH2O
O
O
4d, R1 = H, R2 = CH2O
3b-(4-methylphenyl)tropane 5, couldnot be isolated
since it hadundergone an unanticipatedintermolecular
nucleophilic addition to provide the rearranged product
6a.11 This information promptedus to reinvestigate the
structures of the similar compounds that we had previ-
ously reported.10 Through X-ray crystallographic and
additional NMR analysis, we discovered that the com-
poundwe previously reportedas 4a has the structure
6b. In addition, we reinvestigated the previously re-
portedstructural assignments of the 2 b,3a tropane
derivatives and have determined that these compounds
did not undergo rearrangement.10
ethyl chloroformate in 1,2-dichloroethane at reflux pro-
vided the N-nor analog 6b in 62% yield. The structure
was establishedby single-crystal X-ray analysis ( Fig.
1). Compound 10b possessing the (1S) configuration
was preparedfrom (1 S)-2b-hydroxymethyl-3b-(4-fluoro-
phenyl)tropane following a route similar to that de-
scribedfor
10a. Demethylation of 10b with tri-
chloroethylformate and zinc dust provided the N-
nor analog 6c in 26% yield. For simplicity, only the
structures of the (1R) isomers are shown in Scheme 1.
The (1S) isomers (6c, 10b, and 17) are detailed in the
Experimental section.
1.1. Synthesis
The paroxetine analog 4a was preparedin a manner
analogous to that described by Ogier et al.11 as outlined
in Scheme 2. N-Demethylation of 7 with 1-chloroethyl
chloroformate was followedby formation of the tosyl-
amide 11 using p-toluenesulfonyl chloride in triethyl-
amine at 0 ꢁC. This protection step prevents
nucleophilic attack of the tropane nitrogen at C-9. Lith-
ium aluminum hydride reduction of resulting 2b-carbo-
methoxy-3b-(4-fluorophenyl)-8-(4-toluenesulfonyl)tro-
pane 11 provided the hydroxymethyl derivative 12 in
96% yield. Treatment of alcohol 12 with trifluorome-
thanesulfonic anhydride provided the intermediate mes-
ylate, which was treatedwith the sodium salt of sesamol
to affordthe desiredtropane analog 13 in 58% yield. The
p-toluenesulfonyl protecting group was removedwith
The previously publishedreaction sequence 10 describing
the synthesis of 4a has been modified to show the forma-
tion of the rearrangedproduct 6b (Scheme 1). Although
the addition of methanesulfonyl chloride to 8 in triethyl-
amine provided an intermediate mesylate, which could
be isolated, it rapidly underwent intramolecular nucleo-
philic addition to form 1-methyl-4-(4-fluorophenyl)-1-
azonium-tricyclo[4.3.0.03,9]nonane mesylate 9. Ring
opening of 9 occurredat the C-9 position with the so-
dium salt of sesamol at reflux and provided the rear-
rangedaza-bicyclo[3.2.2]nonane 10a in 44% yield. The
structure of 10a was confirmedby 1H NMR analysis
(see Table 1). N-Demethylation of 10a with 1-chloro-

S. P. Runyon et al. / Bioorg. Med. Chem. 13 (2005) 2439–2449
2441
H3C
H3C
N
N
H3C
O
S
OH
N
CO2CH3
b
a
O
CH3
F
O
F
F
7
8
O
O
c
O
O
-
O
CH3SO3
e
O
d
N+
F
H
F
N
F
N
9
6b, (1R)
c, (1S)
10a, (1R)
b, (1S)
22
Scheme 1. Reagents andconditions: (a) LiAlH 4, THF, 0 ꢁC; (b) (1) methanesulfonyl chloride, TEA, 0 ꢁC; (c) toluene, 1 h reflux; (d) Na, sesamol,
THF, reflux; (e) (1) 1-chloroethyl chloroformate, 1,2-dichloroethane, 24 h reflux; (2) MeOH, 24 h reflux.
Table 1. NMR assignments for 10a obtainedin d6-benzene
The corresponding intermediates, 14–16, were fully
1
5'
characterized. The H NMR spectrum of the product
4'
6'
1'
O
O
was consistent with the spectrum of 4d.10 The structure
of 4d was further confirmedby single-crystal X-ray anal-
ysis (Fig. 3).
7'
3'
O
2'
2
3
1
7
1.2. Biology
4
F
8
N
4"
1"
9
5
3"
2"
The binding affinities of the compounds at the DAT, 5-
HTT, andthe norepinephrine transporter (NET) were
determined via competition binding assays using the
Assignment
Carbon (d)
Proton (d)
HMBC correlations
4
22.29
27.14
32.65
1.14, 1.52
1.70, 1.91
1.82, 2.06
3.35
H3,H9,H5
H4,H1,H5
H4,H1,H8,H5
H7,H2,H1,H9
H7,H3,H8,H9
H7,H5
12
previously reportedprocedures.
The final concentra-
3
tion of radioligands in the assays were 0.5 nM
[3H]WIN35,428 for the DAT, 0.2 nM [3H]paroxetine
for the 5-HTT, and0.5 nM [ 3H]nisoxetine for the
NET. The binding data for (1R) 10a, (1S) 10b, (1R)
6b, and(1 S) 6c, along with data for previously reported
compounds for comparison, are given in Table 3.
9
8
34.44
42.89
1
2.09
2.14
NMe
7
43.08
50.24
1.94, 2.82
2.69
4.01
NMe,H2,H8,H5
NMe,H7,H4,H3,H8
H7,H4,H3,H1,H8
H50,H20
5
55.39
83.83
2
60
70
50
20
300
200
40
100
30
10
400
100.12
101.15
108.43
108.87
115.17
129.89
142.38
143.87
148.96
153.43
161.73
6.59
5.30
—
—
H60
2. Results and discussion
6.27
6.59
6.83
7.35
H300,H200
Tropane ring analogs of paroxetine (2) were designed as
potential inhibitors of serotonin reuptake basedon the
structural similarity andknown biological activity
among the two classes of compounds. Although 3-(4-flu-
orophenyl)tropanes (4) andparoxetine are structurally
similar, they possess distinct differences in conforma-
tional flexibility. Importantly, the difference between
paroxetine andthe tropane/paroxetine hybridstructures
(4a–i) is the presence of an ethylene bridge, which im-
parts increasedsteric bulk andreduces conformational
heterogeneity. Paroxetine is a flexible molecule capable
of existing in solution as an interconverting population
of the conformers, Caa, Bee, Cee, andB aa (Fig. 4).
The tropane/paroxetine hybridstructures can aodpt
analogous conformations, however, the chair form anal-
ogous to Caa (2b,3a) can only convert to the boat form
analogous to Bee andthe chair form analogous to C ee
(2a,3b) can only convert to the boat form analogous
toaa. Interconversion of Caa andC ee (andB ee/Baa) in
H300,H200
—
—
H70,H60
H300,H3,
—
—
H70,H50
H50,H20
—
H300,H200
5% sodium amalgam and disodium hydrogen phosphate
in methanol at reflux to provide the paroxetine analog
4a in 78% yield. The structure of 4a was confirmedby
1H NMR spectral analysis (Table 2) andsingle-crystal
X-ray analysis (Fig. 2).
In order to confirm the previous structural assignment10
of the 2b,3a paroxetine analogs 4b–e, a representative
compound 4d, was resynthesizedaccoridng to the
11
methodof Ogier et al.
as shown for 4a (Scheme 2).

2442
S. P. Runyon et al. / Bioorg. Med. Chem. 13 (2005) 2439–2449
Figure 1. Structure of compound 6b as determined by X-ray diffraction. Thermal ellipsoids are shown at the 30% probability level, hydrogen atoms
are shown as circles of an arbitrary radius.
H3C
Ts
Ts
a,b
c
N
N
N
CO CH
2
CO CH
2
OH
3
3
F
F
F
7
11
12
d,e
O
O
O
O
H
N
Ts
N
O
O
f
F
F
4a
13
Ts = CH C H SO
2
3
6
4
Scheme 2. Reagents and conditions: (a) (1) 1-chloroethyl chloroformate, 1,2-dichloroethane, reflux; (2) MeOH reflux; (b) p-toluenesulfonyl chloride,
triethylamine; (c) LiAlH4, THF, rt; (d) trifluoromethanesulfonic anhydride, pyridine, CH2Cl2; (e) Na, sesamol, THF; (f) 5% Na/Hg amalgam,
Na2HPO4, MeOH.
the tropane analogs is impossible due to the ethylene
bridge. NMR analysis indicates that in solution a flat-
by either 4d or 4i. Unfortunately, the remaining two
conformers of paroxetine are not mimickedin tro-
pane-paroxetine hybridstructures.
tenedboat conformation is preferredfor
4d (2b,3a)
andthat 4i (2a,3b) exists predominately as a chair.10
Confirmation of the NMR data was obtained through
single crystal X-ray analysis, which demonstrated that
4d exists in the boat conformation (Fig. 3). Therefore,
compound 4d represents a rigidanalog of the B ee con-
formation of paroxetine andcompound 4i represents
the Cee paroxetine conformation. Radioligand binding
studies showed that when compared to paroxetine, 4d,
has a 20-foldreducedaffinity for the 5-HTT and156-
and38-foldenhancement of affinity for the DAT and
the NET, respectively. The affinities of 4i at the 5-
HTT andNET are reducedby 170- and3-fold, respec-
tively, comparedto paroxetine; however, affinity for
the DAT is enhanced 22-fold. The striking difference
in the affinity profiles between 4d,i, andparoxetine
may be attributable to increasedsteric bulk introduced
by the ethylene bridge, or may imply that the active con-
formation of paroxetine at the 5-HTT is not mimicked
In an attempt to further elucidate the effects of stereo-
chemistry on the binding affinity of rigid paroxetine-like
structures, tropane analogs 4e and 4h were preparedto
represent the mirror images of paroxetine conforma-
tions Bee and Cee, respectively. This change in stereo-
chemistry, hadonly a moedst effect at 5-HTT, but
significantly reduced affinity at DAT and NET (Table
3). Radioligand binding studies of the 2b,3b-isomer 4a,
revealedrelatively high affinity at the 5-HTT
(Ki = 0.2 nM), DAT (3 nM), andNET ( Ki = 4.0 nM).
Since several (1R)-3b-(para-substituted-phenyl)tropane-
2b-carboxylic acidesters possess high affinity at the 5-
HTT;13 this observation is not surprising. However,
the relatively small differences between the affinity of
4a (2b,3b) and 4d (2b,3a) at all three transporters is
striking; it suggests that stereochemistry at C-3 is irrele-
vant in this case. Indeed, an overlay of the X-ray struc-

S. P. Runyon et al. / Bioorg. Med. Chem. 13 (2005) 2439–2449
Table 2. NMR assignments for 4a obtainedin CDCl
2443
3
7'
O
3'
O
2'
1'
8
5
4'
H
5'
N
O
9
6'
7
1
6
2
4
F
3
4"
1"
2"
3"
Assignment
Carbon (d)
Proton (d)
HMBC correlations
6
28.93
30.16
33.98
1.79, 1.86
1.86, 2.00
1.59, 2.12
3.24
H4,H7,
H6,H4,H5
H3,H6
H4,H5,H9,H200
H9,H9,H3,H4
H4,H6,
H9,H9
H3,H1
H60
7
4
3
35.10
44.91
2
2.06
3.68
5
54.70
56.68
1
3.81
9
67.67
97.93
3.41, 3.97
6.31
5.85
20
70
50
60
300
200
100
30
400
1
Figure 3. Structure of compound 4d as determined by X-ray diffrac-
tion. Thermal ellipsoids are shown at the 30% probability level,
hydrogen atoms are shown as circles of an arbitrary radius.
100.96
105.70
107.76
115.11
128.66
138.30
141.40
148.03
154.35
161.34
—
—
6.11
6.60
—
H600
H300,H200
H300,H200
H50
6.99
7.13
Taken together these results suggest that tropane/par-
oxetine hybridstructures 4a–i interact differently than
paroxetine at monoamine transporters. The reduced
conformational flexibility of 4a and 4d may contribute
to the high affinity of these analogs by maximizing their
residence in conformations favored for these transport-
ers. The greater flexibility of paroxetine permits confor-
mational heterogeneity therefore allowing paroxetine to
adopt a conformation favored by the 5-HTT, which can-
not be achievedby the structurally-rigidtropane ring
analogs 4a and 4d.
—
—
—
—
H200
H5
400
—
H200
Compounds 10a, 10b, 6a, and 6c were preparedas a re-
sult of an unexpected intermolecular nucleophilic addi-
tion andsubsequent ring opening with sesamol. The
affinity andselectivity of these compounds is not parti-
cularly noteworthy with the exception of (1S) 10b. Com-
pound 10b possesses moderate affinity for the 5-HTT
anddisplays selectivity for the 5-HTT (11- and119-fold
relative to the DAT andNET). Although these com-
pounds were prepared inadvertently, some interesting
biological data may be gleaned from further study of
this novel ring system.
The previously reportedattempt to synthesize 4a re-
sultedin the isolation of the rearrangedproduct 10a in-
steadof the expectedintermediate
N-methyl derivative
of 4a. Furthermore, we incorrectly identified 10a as
the expectedtropane. Our incorrect identification was
made based on the similarity of the proton and carbon
Figure 2. Structure of compound 4a as determined by X-ray diffrac-
tion. Thermal ellipsoids are shown at the 30% probability level,
hydrogen atoms are shown as circles of an arbitrary radius.
NMR spectra between the isolatedproduct
10a and
the anticipatedstructure N-methyl 4a. Specifically, the
carbon resonance at 50.24 ppm hadbeen attributedto
the alkoxy methylene carbon C-9 in structure 4a (Table
2) insteadof the methylene C-7 of 10a (Table 1). On this
basis the chemical shifts andproton coupling patterns
determined via a COSY spectrum were self-consistent
andtherefore further supportedthe incorrect assign-
ment. Subsequently, we have employeda more detailed
tures of 4a and 4d (Fig. 5) demonstrates that the orien-
tation of the aromatic rings is remarkably similar despite
the difference in stereochemistry at C-3. This is of partic-
ular interest since the side-chains of 4a and 4d are flexi-
ble andcouldpotentially interact with the transporters
in a similar manner.

2444
S. P. Runyon et al. / Bioorg. Med. Chem. 13 (2005) 2439–2449
Table 3. Comparison of transporter binding affinity for isomers of 10, 6, and 4
O
O
O
O
R
N
O
O
F
R
N
F
6, 10
4
CompdR
Stereo-
IC50, nM (Ki, nM)a
chemistry
2
3
5-HTT
DAT
NET
[3H]Paroxetine
[3H]WIN 35,428
[3H]Nisoxetine
Paroxetineb
(1R)-10a
(1S)-10b
(1R)-6b
0.3 0.02 (0.03 0.001)
294 18 (27 1.6)
88 3 (8 0.3)
623 25
308 20
1050 45
835 90
1210 33
535 15 (322 9.0)
CH3
5300 450 (3200 270)
27,600 1100 (16,600 660)
37,400 1400 (22,500 840)
17,300 1800 (10,400 1080)
CH3
H
480 21 (44 1.9)
424 15 (39 1.4)
(1S)-6c
H
(1R)-4a
H
b
b
b
b
b
a
a
a
a
b
a
a
a
a
b
b
b
b
2
0.7 (0.2 0.06)
3
3
0.6
0.2
6
123 9.5 (74 5.7)
0.3 (4 0.2)
(1R)-4bb
(1S)-4cb
(1R)-4db
(1S)-4eb
(1R)-4fb
(1S)-4gb
(1R)-4hb
(1S)-4ib
CH3
CH3
H
422 16 (38 1.5)
178 13 (16 1.2)
298 17
0.2
12,400 720 (7500 430)
14 1.3 (8.4 0.78)
6
0.2 (0.5 0.02)
4
H
19 1.8 (1.7 0.2)
53 3.6 (4.8 0.3)
447 47 (41 4.3)
90 3.4 (8.2 0.3)
56 5.6 (5.1 0.5)
407 33
172 8.8
1500 74
142 13
28 2.4
1990 176 (1200 106)
26,600 1200 (16,000 720)
2916 1950 (1750 1170)
2500 250 (1500 150)
1690 150 (1020 90)
CH3
CH3
H
H
a Data are mean standard error of three or four experiments with triplicate values at each concentration.
10
b IC50 values were obtainedfrom previous studies.
analysis using gradient enhanced COSY, HSQC, and
HMBC NMR spectra for structural assignment. The
complete proton andcarbon assignments of 10a are pre-
sentedin Table 1. Stereochemical assignments were
made using a ROESY NMR spectrum. The relative ste-
reochemistry is identified by observation of a through
space correlation between H-2 andone of the H-7 pro-
tons andbetween the aromatic proton H-2 00andthe
other H-7 proton. This sets the stereochemistry of both
H-2 andthe 4-fluorophenyl group as syn to the azameth-
ylene bridge.
ene carbon identified C-4. The protons on C-4 show a
vicinal coupling to a methine proton identified as H-5.
Their mutual proton couplings andtheir proximity to
either H-5 or H-1 assignedthe remaining low fieldpro-
tons as H-6 andH-7, respectively. The methylene pro-
tons at 5.85 ppm were assignedto the methylenedioxy
moiety C-70 basedon the characteristic chemical shift.
The complete assignment of the proton andcarbon
spectra of 4a is presentedin Table 2. The stereochemis-
try was identified by the observation of a correlation in
the ROESY spectrum between H-9 andH-4 b. This
interaction is possible only if C-9 andH4 b are diaxial.
Furthermore, an interaction between H-200 andboth
H-9 andH-4 b indicates that the fluorophenyl group is
equatorial. An axial C-9 andequatorial 4-fluorophenyl
group is consistent with the expected2 b,3b tropane.
The structure of 4a was also determined by gradient en-
hancedCOSY, HSQC, HMBC, andROESY NMR
spectra (Table 2). The protons of the sesamoyl and4-flu-
orophenyl group were identified by the characteristic
1,3,5-trisubstitiution and1,4-idsubstitution coupling
patterns, respectively, with the latter being modified by
fluorine–proton coupling interactions. H-3 was identi-
fiedby the long-range proton–carbon correlations be-
tween H-3 andthe carbons of the 4-fluorophenyl
group observedin a gradient HMBC experiment. Car-
bon C-9 was identified as the methylene carbon with a
chemical shift of 67.67 ppm, which is reasonable for an
oxymethylene carbon. A gradient COSY spectrum
shows a vicinal coupling of both H-9 protons to a
methine proton, which in turn is coupled, to H-3 thus
identifying the methine proton as H-2. Likewise a cou-
pling between H-2 anda methine proton at 3.81 ppm
identifies the methine resonance as H-1. A long-range
proton–carbon correlation between H-3 anda methyl-
3. Conclusion
Tropane ring analogs 4a–i of paroxetine (2) were pre-
pared in order to gain a better understanding of the con-
formational requirements for high affinity andselectivity
at the 5-HTT. Ligand binding inhibition data demon-
strate that tropane-derived analogs of paroxetine (4a–
i) interact with the 5-HTT in a manner distinct from par-
oxetine. Paroxetine displays higher affinity and selectiv-
ity than the two most potent tropane analogs 4a (2b,3b)
and 4d (2b,3a). Both 4a and 4d have a nonselective pro-
file with respect to the 5-HTT, NET, andDAT andpos-
sess moderate affinity at all three transporters. These

S. P. Runyon et al. / Bioorg. Med. Chem. 13 (2005) 2439–2449
2445
H
O
H
N
F
N
O
O
O
O
O
Caa
(1R)4d
Chair
F
O
O
O
O
H
N
O
O
F
H
N
(1R)4d
Boat
F
Bee
F
F
H
H
H
N
N
O
Baa
H
O
O
O
O
O
(1S)4i
Boat
N
F
N
O
F
O
Cee
(1S)4i
Chair
O
O
O
O
Figure 4. Potential conformations of paroxetine 2 andthe analogous tropanes 4d and 4i. B andC indicate boat andchair conformations while aa and
ee are axial–axial andequatorial–equatorial.
4. Experimental
Melting points were determined using a Thomas Hoover
melting point apparatus andwere uncorrected. Mag-
netic resonances (1H NMR and 13C NMR) were ob-
tainedwith a Bruker AMX 500 MHz anda Bruker
Avance DPX 300 MHz spectrometer using tetrameth-
ylsilane as an internal standard. All optical rotations
were determined at the sodium D line using a Rudolph
Research Autopol lII polarimeter (1 dm cell). Thin layer
chromatography (TLC) was carriedout using Whatman
andMerck silica gel 60 TLC plates. Medium pressure
column chromatography was performedusing silica
gel 60, 0.040–0.063 mm (230–400 mesh), EM Science.
All chemicals were purchasedfrom Aldrich Chemical
Co., Inc. andwere usedwithout further purification.
Elemental analysis was conducted by Atlantic Microlab,
Norcross, GA. Results were within 0.4% of the critical
Figure 5. A superimposition of the X-ray crystal structures of 4a and
4d.
values
4.1. (1R)-2b-[(3,4-Methylenedioxy)phenoxy]methyl-3b-
(4-fluorophenyl)nortropane tartrate (4a)
results suggest that greater flexibility of paroxetine per-
mits conformational heterogeneity therefore allowing
paroxetine to obtain a conformation favoredby the 5-
HTT, which cannot be achievedby the structurally-rigid
tropane ring analogs 4a and 4d.
(1R)-2b-[(3,4-Methylenedioxy)phenoxy]methyl-3b-(4-flu-
orophenyl)-8-(4-methyl benzenesulfonyl)tropane 13 (0.5 g,
0.98 mmol) was added to a suspension of Na2HPO4

2446
S. P. Runyon et al. / Bioorg. Med. Chem. 13 (2005) 2439–2449
(0.69 g, 4.9 mmol) and5% Na/Hg amalgam (10 g,
50 mmol) in anhydrous MeOH (35 mL). The reaction
was heatedat reflux for 3 h andallowedto cool to room
temperature. Water (100 mL) andNH 4OH (100 mL)
were added, and the mixture was extracted with CH2Cl2
(3 · 100 mL). The extracts were combined, dried
(MgSO4), and concentrated under reduced pressure to
provide a tan solid. The solid was purified using medium
pressure column chromatography (CHCl3/MeOH/
NH4OH, 9:0.9:0.1) to provide a tan solid (0.27 g,
78%). Mp 122–123 ꢁC. The tartrate salt was prepared
by adding D(À)-tartaric acidto the amine in 2-propanol.
The suspension was heated until the salt dissolved,
trate salt using D(À)-tartaric acidin EtOAc. The solid
was collectedandrecrystallizedfrom EtOH/Et 2O to
22
provide 6b as a white powder. Mp 183–184 ꢁC. ½aꢀ
D
+10.6 (c 0.36, MeOH). 1H NMR (CD3OD): d 1.98–
2.37 (m, 7H), 3.24–3.42 (m, 2H), 3.72 (m, 1H), 3.87 (s,
1H), 4.42 (s, 2H), 4.55 (s, 1H), 5.7 (s, 2H), 6.39 (dd,
1H), 6.58 (d, 1H), 6.68 (d, 1H), 7.05 (m, 2H), 7.33 (m,
2H). Anal. Calcdfor C 29H28FNO9: C, 59.40; H, 5.58;
N, 2.77. Found: C, 59.27; H, 5.63; N, 2.70.
4.4. (1S,2R,5R,8R)-2-[(3,4-Methylenedioxy)phenoxy]-8-
(4-fluorophenyl)-6-aza-bicyclo[3.2.2]nonane tartrate (6c)
cooled, and the solid precipitate 4a was collected. Mp
Trichloroethyl chloroformate (0.08 g, 3.77 mmol) was
added under N2 to a suspension of potassium carbonate
(0.05 g, 0.36 mmol) and 10b (0.39 g, 1.07 mmol) in tolu-
ene (10 mL). The suspension was heatedat reflux for
22
D
183–185 ꢁC; ½aꢀ À10.3 (c 0.46, MeOH). 1H NMR
(CD3OD): d 1.88 (m, 1H), 2.11–2.33 (m, 5H) 2.63 (dd,
J = 13.8, 13.8 Hz, 1H), 3.53–3.67 (m, 2H), 3.80 (dd,
J = 3.3, 10 Hz, 1H), 4.25 (m, 1H), 4.3 (m, 3H), 5.8 (s,
2H, O–CH2–O), 6.29 (dd, J = 2.4, 8.4 Hz, 1H), 6.56 (d,
J = 2.4 Hz, 1H), 6.65 (d, J = 8.4 Hz, 1H), 7.03 (m, 2H,
Ar–H), 7.31 (m, 2H, Ar–H). Anal. Calcdfor
C29H28FNO9: C, 59.40; H, 5.58; N, 2.77. Found: C,
59.03; H, 5.58; N. 2.70.
24 h, andH 2O (50 mL) andCHCl
(100 mL) were
3
added. The biphasic mixture was extracted with CHCl3
(3 · 100 mL), dried (Na2SO4), andconcentratedunder
reduced pressure. The oil was then dissolved in glacial
acetic acid (6 mL), and Zn dust (1.0 g) was added in
small portions. The slurry was then allowedto stir at
room temperature for 12 h. Water (50 mL) andCHCl
(100 mL) were added, and the reaction was filtered
through Celite. The biphasic mixture was made basic
3
4.2. (1R)-2b-[(3,4-Methylenedioxy)phenoxy]methyl-3a-
(4-fluorophenyl)nortropane tartrate (4d).
with NH4OH andextractedwith CHCl
(3 · 100 mL).
3
The title compoundwas preparedin a manner analo-
gous to 4a. The resulting oil was purifiedusing medium
pressure column chromatography (CHCl3/MeOH/
NH4OH, 90:9:1) to provide a colorless oil (0.22 g,
63%). The tartrate salt was prepared by the addition
of 1.1 equiv of D(À)-tartaric acidto the amine in 2-pro-
panol. The suspension was heateduntil the soliddis-
solved, cooled, and the salt precipitated with
anhydrous Et2O. The solidwas collectedandrecrystal-
The organic extracts were combined, dried (K2CO3),
and concentrated under reduced pressure to provide
an oil. The oil was purifiedusing medium pressure col-
umn chromatography (50% 9:1, Et2O/Et3N; 50%
8:1.8:0.2, CHCl3/MeOH/NH4OH) to provide an oil
(0.12 g, 26%). The oil was convertedto a tartrate salt
using L(+)-tartaric acidin EtOH. The solidwas collected
andrecrystallizedfrom EtOH/Et 2O to provide 6c as a
22
tan powder. Mp 180–184 ꢁC. ½aꢀ À10.3 (c 0.30,
D
lizedfrom EtOH/Et O to provide 4d as a white solid.
2
MeOH). 1H NMR (CD3OD): d 1.98–2.37 (m, 7H),
3.24–3.42 (m, 2H), 3.72 (m, 1H), 3.87 (s, 1H), 4.42 (s,
2H), 4.55 (s, 1H), 5.7 (s, 2H), 6.39 (dd, 1H), 6.58 (d,
1H), 6.68 (d, 1H), 7.05 (m, 2H), 7.33 (m, 2H). Anal.
Calcdfor C 29H28FNO9: C, 59.40; H, 5.58; N, 2.77.
Found: C, 59.22; H, 5.61; N. 2.74.
22
D
Mp 177–179 ꢁC. ½aꢀ À82.12 (c 0.47, MeOH). 1H
NMR (CD3OD): d 1.58–1.67 (m, 1H), 1.89–2.24 (m,
5H), 2.45–2.50 (m, 1H), 2.83–2.89 (m, 1H), 3.54–3.69
(m, 2H), 5.77 (s, 2H), 6.18–6.21 (dd, J = 2.4, 8.4 Hz,
1H), 6.40–6.41 (d, J = 2.4 Hz, 1H), 6.94–7.0 (m, 2H),
7.21–7.26 (m, 2H).
4.5. (1R)-2b-Hydroxymethyl-3b-(4-fluorophenyl)tropane
(8)
4.3. (1R,2S,5S,8S)-2-[(3,4-Methylenedioxy)phenoxy]-8-
(4-fluorophenyl)-6-aza-bicyclo[3.2.2]nonane tartrate (6b)
(1R)-2b-Carbomethoxy-3b-(4-fluorophenyl)tropane 710
(1.01 g, 3.61 mmol) in Et2O (10 mL) was added under
N2 in a dropwise manner to an ice-cold solution of
LiAlH4 (0.19 g, 5.01 mmol) in Et2O (20 mL). The sus-
pension was allowedto warm to room temperature while
stirring was continued(3 h). The suspension was cooled
to 0 ꢁC, andsaturatedNH 4Cl (ꢁ3 mL) was added. The
suspension was allowedto warm to room temperature,
water was added, and the suspension was extracted with
Et2O (3 · 100 mL). The organic extracts were combined,
dried (Na2SO4), and concentrated under reduced pres-
sure to provide a white solid. The solid was recrystallized
from hexane to provide 8 (0.56 g, 62%) as a white solid.
1-Chloroethyl chloroformate (0.62 g, 4.63 mmol) was
added under N2 to a suspension of 1,8-bis(dimethyl-
amino)naphthalene (0.02 g, 0.09 mmol) and 10a (0.14 g,
0.68 mmol) in CH2Cl2 (10 mL). The suspension was
heatedat reflux for 24 h andwas concentratedunder re-
duced pressure to provide a red oil. The oil was dis-
solvedin MeOH (10 mL) andheatedat reflux for
24 h. The solvent was removed under reduced pressure,
andthe reaction was made basic with NH 4OH. The bi-
phasic mixture was then extractedwith CH 2Cl2
(3 · 75 mL). The organic extracts were combined, dried
(MgSO4), and concentrated under reduced pressure to
provide an orange oil. The oil was purified using med-
ium pressure column chromatography (50% 9:1, Et2O/
Et3N; 50% 8:1.8:0.2, CHCl3/MeOH/NH4OH) to provide
a pink oil (0.08 g, 61.9%). The oil was convertedto a tar-
1
Mp 75–78 ꢁC. H NMR (CDCl3): d 1.46 (m, 1H), 1.58–
1.67 (m, 1H), 1.72 (s, 1H), 1.75 (s, 1H), 2.16 (m, 2H),
2.28 (s, 3H), 2.50 (m, 1H), 3.07 (m, 1H), 3.35 (m, 2H),
3.46 (m, 1H), 3.75 (m, 1H), 7.01 (m, 2H), 7.32 (m, 2H).

S. P. Runyon et al. / Bioorg. Med. Chem. 13 (2005) 2439–2449
2447
4.6. (3R,4S,6S,9R)-1-Methyl-4-(4-fluorophenyl)-1-azo-
nium-tricyclo[4.3.0.03,9]nonane mesylate (9)
tricyclo[4.3.0.03,9]nonane mesylate 17 (1.66 g, 7.6 mmol)
in THF (20 mL) was added at 0 ꢁC in a dropwise man-
ner over 10 min. The reaction was allowedto warm to
room temperature andwas heatedat reflux for 3 h.
Water (100 mL) andconcentratedNH 4OH were added
until the aqueous layer was basic (pH ꢁ 10), andthe bi-
phasic mixture was extractedwith CH 2Cl2 (3 · 150 mL).
The organic extracts were combined, dried (MgSO4),
and concentrated under reduced pressure to provide a
yellow oil. The oil was purifiedusing medium pressure
column chromatography (30% 9:1 Et2O/EtN, 70% hex-
ane) to provide an oil (1.02 g, 69%). The oil was con-
vertedto a tartrate salt with D(À)-tartaric acidin
EtOAc. The resulting solidwas collectedandrecrystal-
Methanesulfonyl chloride (0.30 g, 2.6 mmol) was added
dropwise at 0 ꢁC under N2 to a solution of triethylamine
(0.22 g, 2.16 mmol) and 8 (0.46 g, 1.86 mmol) in CH2Cl2
(15 mL). The reaction was allowedto stir at 0 ꢁC for
0.5 h andthen at room temperature for 3 h. Water
andCH Cl2 were added, and the suspension was made
2
basic (pH 10) with NH4OH. The CH2Cl2 was removed,
andthe mixture was extractedwith CH 2Cl2 (3 · 75 mL).
The organic extracts were combined, dried (Na2SO4),
and concentrated under reduced pressure to provide a
colorless oil. Toluene (100 mL) was added, and the sus-
pension was heatedat reflux for 1 h. The toluene was re-
moved under reduced pressure to provide 9 (0.58 g,
95%) as a white semisolid, which was used without fur-
lizedfrom 2-propanol/Et O to provide 10b as a white
2
25
D
powder. Mp 90 ꢁC (fusion). ½aꢀ
À26.8 (c 0.63,
1
MeOH). H NMR (CD3OD): d 2.0–2.43 (m, 7H), 2.95
(s, 3H), 3.48 (m, 2H), 3.68 (m, 2H), 4.61 (m, 1H), 5.86
(s, 2H), 6.39 (dd, 1H), 6.49 (d, 1H), 6.68 (dd, 1H),
7.03 (m, 2H), 7.36 (m, 2H).
1
ther purification. H NMR (CDCl3): d 2.09 (m, 1H),
2.30 (m, 3H), 2.57 (m, 1H), 2.78 (s, 3H), 2.93 (m, 2H),
3.28 (s, 3H), 3.37 (m, 1H), 4.45 (m, 1H), 4.56 (m, 1H),
4.69 (m, 1H), 5.35 (m, 1H), 7.02 (m, 2H), 7.22 (m, 2H).
4.9. (1R)-2b-Carbomethoxy-3b-(4-fluorophenyl)-8-(4-
methyl benzenesulfonyl)tropane (11)
4.7. (1R,2S,5S,8S)-2-[(3,4-Methylenedioxy)phenoxy]-8-
(4-fluorophenyl)-6-methyl-6-aza-bicyclo[3.2.2]nonane tar-
trate (10a)
Tosylchloride (4 g, 20.9 mmol) was added under N2 to a
solution of (1R)-2b-carbomethoxy-3b-(4-fluorophen-
yl)nortropane9 (5 g, 19 mmol) in triethylamine
(150 mL) at À10 ꢁC. The reaction was allowedto stir
Sesamol (0.56 g, 4.08 mmol) in anhydrous THF (10 mL)
was added under N2 at 0 ꢁC to a suspension of NaH
(60% dispersion, 0.17 g, 4.30 mmol, washed 2 · 10 mL
with hexanes) in anhydrous THF (10 mL). The reaction
was allowedto warm to room temperature over 1 h.
(3R,4S, 6S,9R) 1-Methyl-4-(4-fluorophenyl)-1-azonium-
tricyclo[4.3.0.03,9]nonane mesylate 9 (0.58 g, 2.6 mmol)
in THF (10 mL) was added at 0 ꢁC in a dropwise
manner over 10 min. The reaction was allowedto warm
to room temperature andwas heatedat reflux for
3 h. Water (100 mL) andconcentratedNH 4OH were
added until the aqueous layer was basic (pH ꢁ 10),
andthe biphasic mixture was extractedwith CH 2Cl2
(3 · 100 mL). The organic extracts were combined, dried
(MgSO4), and concentrated under reduced pressure to
provide a yellow oil. The oil was purified using medium
pressure column chromatography (30% 9:1 Et2O/EtN,
70% hexane) to provide an oil (0.33 g, 44%). The oil
was convertedto a tartrate salt with D(À)-tartaric acid
in EtOAc. The resulting solidwas collectedandrecrys-
for 10 h before water (200 mL) andCH Cl2 (150 mL)
2
were added. The mixture was then extracted with
CH2Cl2 (3 · 150 mL). The organic extracts were com-
bined, dried (MgSO4), and concentrated under reduced
pressure to provide a white solid. The solid was then
recrystallizedfrom MeOH/toluene to proveid
11
(6.27 g, 79%) as a white solid. Mp 195–197 ꢁC. 1H
NMR (CDCl3): d 1.6–2.1(m, 5H) 2.42 (s, 3H, Ar–
CH3), 2.67 (ddd, J = 3, 9.6, 9.6 Hz, 1H), 2.92 (m, 1H),
3.1 (m, 1H), 3.35 (s, 3H, OCH3), 4.45 (m, 1H), 4.59
(m, 1H), 6.93 (m, 2H, Ar–H), 7.17 (m, 2H, Ar–H),
7.28 (d, J = 9 Hz, 2H, Ar–H), 7.77 (d, J = 9 Hz, 2H,
Ar–H).
4.10. (1R)-2b-Hydroxymethyl-3b-(4-fluorophenyl)-8-(4-
methyl benzenesulfonyl)tropane (12)
(1R)-2b-Carbomethoxy-3b-(4-fluorophenyl)-8-(4-methyl
benzenesulfonyl)tropane 11 (6.0 g, 14.37 mmol) in
anhydrous THF (50 mL) was added under N2 to a sus-
pension of LiAlH4 (1.64 g, 43.11 mmol) in anhydrous
THF (300 mL) at 0 ꢁC. The reaction was allowedto stir
at room temperature for 1 h. Water (1.65 mL) and10%
NaOH (1.65 mL) were cautiously added followed by
Celite (2 g). The slurry was then filteredthrough a sin-
teredglass filter, andthe filtrate was concentratedun-
der reduced pressure to provide a white solid. The
solidwas recrystallizedfrom toluene/ethanol to provide
tallizedfrom 2-propanol/Et 2O to provide 10a as a white
25
powder. Mp 97 ꢁC (fusion). ½aꢀ +11.8 (c 0.27, MeOH).
D
1H NMR (CD3OD): d 2.0–2.43 (m, 7H), 2.95 (s, 3H),
3.48 (m, 2H), 3.68 (m, 2H), 4.61 (m, 1H), 5.86 (s, 2H),
6.39 (dd, 1H), 6.49 (d, 1H), 6.68 (dd, 1H), 7.03 (m,
2H), 7.36 (m, 2H).
4.8. (1S,2R,5R,8R)-1-[(3,4-Methylenedioxy)phenoxy]-8-
(4-fluorophenyl)-6-methyl-6-aza-bicyclo[3.2.2]nonane tar-
trate (10b)
1
12 (5.4 g, 96%) as a white solid. Mp 154–156 ꢁC. H
Sesamol (1.4 g, 10.2 mmol) in anhydrous THF (10 mL)
was added under N2 at 0 ꢁC to a suspension of NaH
(60% dispersion, 0.41 g, 10.2 mmol, washed 2 · 10 mL
with hexanes) in anhydrous THF (10 mL). The reaction
was allowedto warm to room temperature over 1 h.
(3S,4R,6R,9S)-1-Methyl-4-(4-fluorophenyl)-1-azonium-
NMR (CDCl3): d 1.4–1.7 (m, 5H) 2.01 (m, 1H), 2.34
(ddd, J = 3, 10.2, 10.2 Hz, 1H), 2.44 (s, 3H, Ar-CH3),
3.03 (m, 1H), 3.16 (m, 1H), 3.81 (m, 1H) 4.45 (m,
2H), 6.98 (m, 2H, Ar–H), 7.09 (m, 2H, Ar–H), 7.32
(d, J = 8.5 Hz, 2H, Ar–H), 7.77 (d, J = 8.5 Hz, 2H,
Ar–H).

2448
S. P. Runyon et al. / Bioorg. Med. Chem. 13 (2005) 2439–2449
4.11. (1R)-2b-[(3,4-Methylenedioxy)phenoxy]methyl-3b-
(4-fluorophenyl)-8-(4-methyl benzenesulfonyl)tropane (13)
3H), 2.46–2.66 (m, 2H), 3.51–3.53 (d, J = 6 Hz, 2H),
4.28–4.33 (m, 2H), 6.89–7.02 (m, 2H), 7.11–7.18 (m,
2H), 7.28–7.31 (d, J = 8.1 Hz, 2H), 7.74–7.77 (d,
J = 8.1 Hz).
Trifluoromethanesulfonic anhydride (2.17 g, 7.70 mmol)
was added dropwise at À78 ꢁC under N2 to a solution of
12 (2.0 g, 5.13 mmol) in pyridine (0.49 g, 6.16 mmol)
andCH Cl2 (50 mL). The reaction was allowedto stir
at À78 ꢁ2C for 1 h andthen warmedto room tempera-
ture. Water (100 mL) was added, and the mixture was
4.14. (1R)-2b-[(3,4-Methylenedioxy)phenoxy]methyl-3a-
(4-fluorophenyl)-8-(4-methyl benzenesulfonyl)tropane (16)
The title compoundwas preparedin a manner analo-
gous to that of 13. The resulting solidwas purifiedusing
medium pressure column chromatography (petroleum
ether/acetone, 8:2). The solidwas recrystallizedfrom tolu-
ene/petroleum ether to provide 16 as a white solid
(0.60 g, 77%). Mp 170–172 ꢁC. 1H NMR (CDCl3): d
1.27–1.37 (m, 1H), 1.45–1.62 (m, 3H), 1.83–1.99 (m,
2H), 2.39 (s, 3H), 2.46–2.63 (m, 2H), 3.49–3.53 (m,
1H), 3.56–3.66 (m, 1H), 4.29–4.31 (d, J = 7.5 Hz, 1H),
4.36–4.40 (dd, J = 5.7, 13 Hz, 1H), 5.89 (s, 2H), 6.12–
6.15 (dd, J = 3, 8.4 Hz, 1H), 6.32 (d, J = 3 Hz, 1H),
6.64–6.70 (d, J = 18 Hz, 1H), 6.88–6.99 (m, 2H), 7.11–
7.16 (m, 2H), 7.22–7.25 (d, J = 8.4 Hz, 2H), 7.73–7.76
(d, J = 8.1 Hz, 2H).
extractedwith CH Cl2 (3 · 75 mL). The extracts were
2
combined, dried (MgSO4), andconcentratedunedr
reduced pressure to provide crude (1R)-2b-trifluoro-
methanesulfonyloxymethyl-3b-(4-fluorophenyl)-8-(4-meth-
yl benzenesulfonyl)tropane as a white solid. The desired
compoundwas usedin the next step without further
purification.
Sodium (0.14 g, 6.20 mmol) was added under N2 to a
solution of sesamol (0.79 g, 5.64 mmol) in anhydrous
THF (35 mL). The reaction was allowedto stir at room
temperature until all of the sodium had dissolved (about
2 h). The sodium sesamol solution was then added under
N2 to a solution of (1R)-2b-trifluoromethanesulfonyl-
oxymethyl-3b-(4-fluorophenyl)-8-(4-methyl
benzene-
sulfonyl)tropane (2.68 g, 5.13 mmol) in CH2Cl2
(15 mL). The solution was allowedto stir at room tem-
perature for 12 h. Water (100 mL) and10% NaOH
(20 mL) were added, and the mixture was extracted with
CH2Cl2 (3 · 75 mL). The extracts were combined, dried
(MgSO4), and concentrated under reduced pressure to
provide a white solid. The solid was recrystallized from
MeOH/toluene to provide 13 (1.56 g, 58%) as a white so-
lid. mp 173–175 ꢁC. 1H NMR (CDCl3): d 1.7–2.1(m,
5H) 2.20 (m, 2H), 2.30 (s, 3H, Ar–CH3), 3.2 (dd,
J = 4, 9 Hz, 1H), 3.27 (m, 1H), 3.69 (dd, J = 12.6,
12.6 Hz, 1H), 4.47 (m, 1H), 4.57 (m, 1H), 5.87 (s, 2H,
O–CH2–O), 5.96 (dd, J = 2.7, 8.7 Hz, 1H), 6.12 (d,
J = 2.7 Hz, 1H), 6.27 (d, J = 12 Hz, 1H), 7.01 (m, 2H,
Ar–H), 7.09 (m, 4H, Ar–H), 7.68 (d, J = 8.4 Hz, 2H,
Ar–H).
4.15. (3S,4R,6R,9S)-1-Methyl-4-(4-fluorophenyl)-1-azo-
nium-tricyclo[4.3.0.03,9]nonane mesylate (17)
Methanesulfonyl chloride (0.74 g, 6.46 mmol) was
added dropwise at 0 ꢁC under N2 to a solution of tri-
ethylamine (0.73 g, 7.2 mmol) and(1 S)-2b-hydroxy-
methyl-3b-(4-fluorophenyl)tropane10
(1.0 g,
4.02
mmol) in CH2Cl2 (15 mL). The reaction was allowed
to stir at 0 ꢁC for 0.5 h andthen at room tem-
perature for 3 h. Water andCH Cl2 were added, and
2
the suspension was made basic (pH 10) with NH4OH.
The CH2Cl2 was removed, and the mixture was ex-
tractedwith CH 2Cl2 (3 · 100 mL). The organic ex-
tracts were combined, dried (Na2SO4), and
concentrated under reduced pressure to provide a col-
orless oil. Toluene (100 mL) was added, and the sus-
pension was heatedat reflux for 1 h. The toluene
was removed under reduced pressure to provide crude
17 (1.67 g) as a yellow semisolidthat was usedwithout
further purification. 1H NMR (CDCl3): d 2.09 (m,
1H), 2.30 (m, 3H), 2.57 (m, 1H), 2.78 (s, 3H), 2.93
(m, 2H), 3.28 (s, 3H), 3.37 (m, 1H), 4.45 (m, 1H),
4.56 (m, 1H), 4.69 (m, 1H), 5.35 (m, 1H), 7.02 (m,
2H), 7.22 (m, 2H).
4.12. (1R)-2b-Carbomethoxy-3a-(4-fluorophenyl)-8-(4-
methyl benzenesulfonyl)tropane (14)
The title compoundwas preparedin a manner analo-
gous to that of 11. The resulting solidwas recrystallized
from toluene to provide 14 as a white solid(0.70 g,
1
68%). Mp 187–189 ꢁC. H NMR (CDCl3): d 1.04–1.70
(m, 4H), 1.95–1.98 (m, 1H), 2.43 (s, 3H), 2.47–2.57 (m,
2H), 3.19–3.28 (dd, J = 9.9, 17 Hz, 1H), 4.33–4.38 (dd,
J = 7.2, 15 Hz, 1H), 4.43–4.46 (d, J = 7.2 Hz, 1H),
6.90–6.98 (m, 2H), 7.10–7.15 (m, 2H), 7.29–7.32 (d,
J = 8.4 Hz, 2H), 7.78–7.80 (d, J = 8.2 Hz).
Supporting information available: Crystallographic data
for the structures 4a,d, and 6b have been deposited with
the Cambridge Crystallographic Data Centre as supple-
mentary publication numbers CCDC 252218, 252219,
and252220, respectively. Copies can be obtainedfree
of charge by application to CCDC, 12 Union Road,
Cambridge, CB2 1EZ, UK (fax: + 44-(0) 1223-336033
or e-mail: deposit@ccdc.cam.ac.uk).
4.13. (1R)-2b-Hydroxymethyl-3a-(4-fluorophenyl)-8-(4-
methyl benzenesulfonyl)tropane (15)
The title compoundwas preparedin a manner analo-
gous to that of 12. The resulting solidwas recrystallized
from toluene to provide 15 as a white solid(0.56 g,
90%). Mp 165–167 ꢁC. H NMR (CDCl3): d 1.33–1.57
(m, 4H), 1.72–1.79 (m, 2H), 2.03 (br s, 1H), 2.43 (s,
Acknowledgements
1
This research was supportedby the National Institute
on Drug Abuse Grant DA05477. We thank Dr. Gilles

S. P. Runyon et al. / Bioorg. Med. Chem. 13 (2005) 2439–2449
2449
8. Fuller, R. W. Prog. Drug Res. 1995, 45, 167–
204.
Tamagnan for supplying us with a copy of his manu-
script11 before publication.
9. Boja, J. W.; Kuhar, M. J.; Kopajtic, T.; Yang, E.;
Abraham, P.; Lewin, A. H.; Carroll, F. I. J. Med. Chem.
1994, 37, 1220–1223.
References and notes
10. Keverline-Frantz, K. I.; Boja, J. W.; Kuhar, M. J.;
Abraham, P.; Burgess, J. P.; Lewin, A. H.; Carroll, F. I.
J. Med. Chem. 1998, 41, 247–257.
1. Carroll, F. I.; Lewin, A. H.; Boja, J. W.; Kuhar, M. J. J.
Med. Chem. 1992, 35, 969–981.
2. Giros, B.; Jaber, M.; Jones, S. R.; Wightman, R. M.;
Caron, M. G. Nature 1996, 379, 606–612.
3. Hyman, S. E. Neuron 1996, 16, 901–904.
4. Nestler, E. J. Nat. Rev. Neurosci. 2001, 2, 119–128.
5. White, F. J.; Kalivas, P. W. Drug Alcohol Depen. 1998, 51,
141–153.
11. Ogier, L.; Turpin, F.; Baldwin, R. M.; Riche, F.; Law, H.;
Innis, R. B.; Tamagnan, G. J. Org. Chem. 2002, 67, 3637–
3642.
12. Carroll, F. I.; Gray, J. L.; Abraham, P.; Kuzemko, M. A.;
Lewin, A. H.; Boja, J. W.; Kuhar, M. J. J. Med. Chem.
1993, 36, 2886–2890.
6. McMahon, L. R.; Filip, M.; Cunningham, K. A. J.
Neurosci. 2001, 21, 7781–7787.
7. Czoty, P. W.; Ginsburg, B. C.; Howell, L. L. J. Pharmacol.
Exp. Ther. 2002, 300, 831–837.
13. Carroll, F. I.; Lewin, A. H.; Mascarella, S. W. In
Neurotransmitter Transporters: Structure, Function, and
Regulation, 2nd ed.; Reith, M. E. A., Ed.; Humana:
Totowa, NJ, 2001; pp 381–432.