Synthesis and biological evaluation of 1-azabicyclo-[3.2.1]octanes: New dopamine transporter inhibitors

Article abstract of DOI:10.1016/S0960-894X(00)00308-5

The synthesis and biological activity of a series of 6-substituted 1-azabicyclo[3.2.1]octanes are described. 1-Azabicyclo[3.2.1]octanes represent a new class of compounds that exhibit monoamine transporter inhibitory activity highly dependent on the overall topology and the absolute stereochemistry of the molecule. (C) 2000 Elsevier Science Ltd. All rights reserved.

Full text of DOI:10.1016/S0960-894X(00)00308-5

Bioorganic & Medicinal Chemistry Letters 10 (2000) 1681±1686
Synthesis and Biological Evaluation of 1-Azabicyclo-
[3.2.1]octanes: New Dopamine Transporter Inhibitors
Amir P. Tamiz,a Miles P. Smith,b Istvan Enyedy,a Judith Flippen-Anderson,c
Mei Zhang,d Kenneth M. Johnsond and Alan P. Kozikowskia,*
aDrug Discovery Program Department of Neurology, Georgetown University Medical Center,
3970 Reservoir Road, NW, Washington, DC 20007-2197, USA
bZebra Pharmaceuticals, 160 Second Street, Cambridge, MA 02142, USA
cLaboratoryfor the Structure of Matter, Code 6030, Naval Research Laboratory, 4555 Overlook, SW, Washington, DC 20275±5000, USA
dDepartment of Pharmacology and Toxicology, University of Texas Medical Branch, Galveston, TX 77555-1031, USA
Received 12 April 2000; accepted 23 May 2000
AbstractÐThe synthesis and biological activity of a series of 6-substituted 1-azabicyclo[3.2.1]octanes are described. 1-Azabicy-
clo[3.2.1]octanes represent a new class of compounds that exhibit monoamine transporter inhibitory activity highly dependent on
the overall topology and the absolute stereochemistry of the molecule. # 2000 Elsevier Science Ltd. All rights reserved.
To date, a large number of compounds that bind at the
cocaine binding site of the dopamine (DA) transporter
(DAT) have been synthesized in order to better understand
the pharmacological properties of this drug.1 The ability of
cocaine to bind to the DAT and to inhibit the reuptake of
dopamine has been implicated in it's reinforcing proper-
ties.2,3 However, the precise binding interaction of these
analogues with speci®c monoamine transporters has been
a matter of much debate.4a,b
disclosed herein, such rigid compounds can be readily
synthesized in enantiomericaly pure form. The SAR
developed from this study reveals a better understanding
of the pharmacophore of the cocaine binding site at the
DAT and therefore can be used towards the design of
novel DAT inhibitors as drugs.
Chemistry
We have recently reported on the synthesis of a series of
rigid tricyclic tropane analogues in which the conformation
of the nitrogen lone pair is ®xed by means of a tether to
either the 3- or 2-carbon bridge of the tropane moiety.5a c
We demonstrated that the selectivity of these rigid tropane
analogues as monoamine transporter inhibitors was
in¯uenced by the orientation of the nitrogen lone pair.
Tropane based compounds (Fig. 1) with the nitrogen lone
pair localized over the 2-carbon bridge of the tropane ring
(front-bridged) exhibit improved potency and selectivity
for the serotonin (5-HT) transporter (SERT). We now
present additional structure±activity relationships (SAR)
in this series, which serve to further de®ne the binding site
topology for these molecules at the respective monoamine
transporters. Herein, we describe a facile synthesis of rigid
bicyclo[3.2.1]octan-6-one intermediates 4 and 5 (Scheme 1)
that are readily converted to O-alkyl oximes 6±11. As is
Bicyclooctanes 6±11 were synthesized as depicted in
Scheme 1. Brie¯y, N-demethylation of ( )-cis piperidine
16,7 was accomplished in two steps by treatment with 1-
chloroethyl chloroformate8 in dichloroethane in the
presence of proton sponge (1.5 equiv), followed by
methanolysis of the resulting carbamate intermediate to
a€ord ( )-2 in quantitative yield.
The resulting piperidine ( )-2 was then N-alkylated using
ethyl bromoacetate in EtOH to give the corresponding
diester ( )-3. Treatment of the diester ( )-3 with KOEt in
re¯uxing ethanol gave the b-ketoester intermediate
(structure not shown) which was reacted immediately
without further puri®cation with aqueous HCl (10%) to
give ketones ( )-4 and (+)-5, which were separated using
chromatography on silica gel. Similarly, ketones (+)-4
and ( )-5 were prepared from (+)-3. The absolute
stereochemistries of (+)-4 and (+)-5 as (1R)-( )-10-
camphorsulfonate salts were established by crystal-
lographic methods (Figs 2 and 3).9a,b
*Corresponding author. Tel.: +1-202-687-0686; fax: +1-202-687-
5065; e-mail: kozikowa@giccs.georgetown.edu
0960-894X/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00308-5

1682
A. P. Tamiz et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1681±1686
Figure 1.
Reaction of the ketones 4 and 5 with appropriate O-sub-
stituted hydroxylamine10 in re¯uxing toluene gave the E-
and Z-oximes, which are readily separated using chroma-
tography on silica gel.11 The Z-isomer was formed exclu-
sively upon condensation of 4 with hydroxylamines. The
stereochemistry of (+)-6 was established by crystal-
lographic methods (Fig. 4).9 Piperidines 13±15 were
prepared as shown in Scheme 2. Treatment of piperidine
( )-16,7 with NaOMe gave the trans piperidine (+)-12
quantitatively. Hydrolysis of piperidine (+)-12 using HCl
(10% in H2O) gave the acid intermediate (structure not
shown). Reaction of the acid intermediate with (COCl)2 in
CH2Cl2 gave the acid chloride which was then reacted
without puri®cation with the appropriate alcohol in the
presence of DMAP and Et3N in CH2Cl2 to give piperidines
13±15.11
Structure±Activity Relationships
All compounds were tested for their ability to inhibit
high anity uptake of DA, 5-HT and norepinephrine
(NE) into nerve endings (synaptosomes).12 The uptake
data based on the Ki values of these compounds are listed
in Table 1. All data are mean values Æ SEM of two to ®ve
separate experiments, each conducted with six drug con-
centrations in triplicate. Additionally, all ®nal compounds
reported in this series were tested for their ability to dis-
place [3H]mazindol binding. Mazindol has been shown
to label the cocaine binding sites on the dopamine trans-
porter of rat striatal membranes.13 This ligand binds with
high anity to a single, sodium-dependent site in striatal
membranes, representing the dopamine carrier.
Scheme 1. Reagents and conditions: (i) 1-Chloroethyl chloroformate,
1,2-dichloroethane, 1,8-bis-(dimethylamino)naphthalene; (ii) MeOH,
re¯ux; (iii) EtOCOCH2Br, K2CO3, EtOH; (iv) KH, ethanol, re¯ux;
(v) HCl (10%), re¯ux; (vi) O-Alkyl hydroxylamine, toluene, re¯ux.
In general, the mazindol binding anities parallel the
DAT uptake potencies. With exception of (+)-8 and
(+)-11, all O-alkyl oximes exhibit lower potency at the
DAT when compared to cocaine. In this series oximes
(+)-8 and ( )-8 exhibit dual potency at the DAT and the
norepinephrine transporter (NET) while oxime ( )-6
exhibits dual potency at the SERT and the NET. Inter-
estingly, oxime ( )-8 is 2-fold less potent at the SERT
than oxime (+)-8. Replacement of the methyl group in
(+)-6 with a t-butyl group gave (+)-9 and resulted in a
4- to 5-fold increase in potency at all three transporters.
Replacement of the methyl oxime in (+)-6 with phenyl-
alkyl oximes gave (+)-10 and (+)-11. Oxime 11 exhibits
excellent potency at the DAT compared to (+)-6. Simi-
lar to this e€ect, we had recently demonstrated that
certain boat tropanes bearing N-arylalkyl groups (mole-
cules that can be broadly viewed as hybrids of GBR1a
and cocaine), are able to retain high anity for the DAT,
while showing a good selectivity ratio for the DAT versus
the SERT.14

A. P. Tamiz et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1681±1686
1683
Figure 2. ORTEP drawing of bicyclooctane (+)-4. The ®gure, show-
ing the correct absolute con®guration, is drawn using the experimen-
tally determined coordinates with anisotropic thermal parameters at
the 20% probability level.9
Figure 4. ORTEP drawing of bicyclooctane (+)-6. The ®gure, show-
ing the correct absolute con®guration, is drawn using the experimen-
tally determined coordinates with anisotropic thermal parameters at
the 20% probability level.9
Figure 3. ORTEP drawing of bicyclooctane (+)-5. The ®gure, show-
ing the correct absolute con®guration, is drawn using the experimen-
tally determined coordinates with anisotropic thermal parameters at
the 20% probability level.9
Scheme 2. Reagents and conditions: (i) NaOMe, MeOH, re¯ux; (ii)
HCl (10%), re¯ux; (iii) (COCl)2, CH2Cl2; (iv) ROH, DMAP, Et3N,
CH2Cl2.
Compounds 13±15 were prepared in order to test the e€ect
of O-phenylalkyl substitution in the piperidine series.
Over the last few years, we have reported on the chemistry
and pharmacology of piperidine-based analogues of
cocaine as potent DAT inhibitors that lack the conven-
tional tropane skeleton. Piperidine based compounds such
as (+)-12 possess signi®cant DAT potency and selectivity,
while causing little locomotor stimulant activity in ani-
mals.6,7 In fact, one of these piperidine-based ligands (+)-
12 has recently been shown at the National Institute on
Drug Abuse (NIDA) to antagonize cocaine-induced
locomotor e€ects in rodents.
Molecular Modeling QSAR
Individual bicyclic oximes 6±11 were constructed using
the program QUANTA15a and the structures minimized
until convergence using the adopted-basis Newton
Raphson algorithm. Since compounds 10 and 11 have
¯exible O-phenylalkyl groups, the global minima were
found using a systematic conformational search with the
grid scan method from the conformational analysis
module of the QUANTA program.15a For each com-
pound two to four torsional angles were modi®ed during
the search, thus generating up to 2600 conformations.
Each conformation was minimized using the adopted-
basis Newton Raphson algorithm with a 2000 iteration
limit. A constant dielectric (equal to 1) was used in all
calculations. The lowest energy minimized conformation
was chosen as the most stable conformation for every
compound.
Interestingly, piperidines 13 and 14 are less active at the
DAT than the corresponding methyl ester 12. However,
piperidine 15 is more potent at the DAT than the methyl
ester 12. Piperidine 15 exhibits DA and NE reuptake
activities that are comparable to those of (+)-11, while it
is 6-fold less potent at the SERT. These similarities in
the biological pro®les of the bicyclic oximes in comparison
to the piperidines encouraged us to further examine the
quantitative SAR (QSAR) in this series of compounds.
The low energy structures obtained after conformational
analysis were used to generate SAR models using the
QSAR+ module from the Cerius2 program.15b In all

1684
A. P. Tamiz et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1681±1686
Table 1. Ki values for the 1-azabicyclo[3.2.1]octane analogues in mazindol binding and dopamine, norepinephrine and serotonin uptake experi-
ments (nM)
5-HT
Compound
Structure
Ð
[3H]Mazindol
bindinga
[3H]DA
uptakea
[3H]NE
uptakea
[3H]5-HT
uptakea
DA
cocaine
230 Æ 16
301 Æ 50
186 Æ 23
305 Æ 1.0
413 Æ 81
503 Æ 21
1.4
(
)-6
>5000
3640 Æ 250
0.15
0.076
1.0
(+)-6
(+)-7
>36,000
1430 Æ 10
2790 Æ 120
796 Æ 121
434 Æ 34
641 Æ 44
>45,000
2520 Æ 470
2660 Æ 330
292 Æ 18
>4000
1790 Æ 30
2640 Æ 450
2530 Æ 10
856 Æ 109
>2000
871 Æ 97
652 Æ 180
240 Æ 18
232 Æ 19
287 Æ 25
(
)-7
0.95
2.9
(+)-8
(
)-8
304 Æ 12
6.6
(+)-9
725 Æ 103
137 Æ 25
0.19
(+)-10
(+)-11
1000 Æ 40
53 Æ 6.0
1860 Æ 270
41 Æ 4.0
1550 Æ 270
483 Æ 11
117 Æ 9.5
292 Æ 33
0.06
7.1
(+)-13
(+)-14
(+)-15
(+)-12
Ðb
Ðb
Ðb
Ðb
1490 Æ 190
420 Æ 31
52 Æ 6
1420 Æ 90
1160 Æ 40
53 Æ 2.6
72 Æ 29
3140 Æ 510
382 Æ 15
90 Æ 5.2
2160 Æ 10
2590 Æ 40
1790 Æ 10
5880 Æ 440
1.5
2.2
34
26
248 Æ 18
228 Æ 30
aNumbers represent the mean Æ SEM from two to ®ve independent experiments, each conducted with at least three determinations.
bStructure shown in Scheme 2.
models log Ki (M) of DA, NE, or 5-HT uptake inhibition
was correlated with the following parameters: (1) dis-
tance between O6B and C4A (D, Fig. 4); (2) the distance
between Cl1 and C6C for compounds 6±8 (D2), (for
compounds 9±11 D2 was measured using the distance
between the Cl1 and the average distance to the hydro-
phobic t-butyl group or phenyl groups); (3) the angle
between O6B-C4-C4A (c); (4) the dihedral angle between
O6B-N6A-C4-C4A (j); and ®nally (5) the molecular
volume (Vm) of each compound. The ®rst three para-
meters characterize the relative position of the sub-
stituent at O6B compared to the phenyl group at C4.
The di€erences in volume (ÁVm) between compounds
9±11 and 6 was introduced to take in account the size
of the substituent at O6B. In the QSAR equations
ÁVm was used instead of Vm because compounds 6±8
have the same volumes, and ÁVm is directly related
to the size of the substituent at the oxygen atom.
The values used for generating the models are listed in
Table 2.
The genetic function approximation method15c was used
to generate several QSAR models. Herein we list ®ve
such equations with r2 values greater than 0.8.

A. P. Tamiz et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1681±1686
1685
Inhibition of DA Uptake
the DAT while showing moderate potency at the NET
and the SERT. The present series of compounds that
embody some elements of structural rigidity provide
further insight into the design of ligands showing certain
levels of transporter selectivity that may be useful in
the development of medications for the treatment of a
variety of neurological disorders.
logKi ˆ 2:83651 ‡ 0:020935DVm ‡ 6:4  10 5DV m2
0:006636' ‡ 0:026145 ; r2 ˆ 0:78
ꢀ1†
logKi ˆ 3:3879 ‡ 0:02499DVm 0:006672'
ꢀ1a†
‡ 0:035676' 0:1857 D2;r2 ˆ 0:81
Acknowledgements
We are indebted to the National Institutes of Health,
National Institute on Drug Abuse (DA10458, and
11548), and to the Oce of Naval Research for their
support of these studies.
Inhibition of NE Uptake
logKi ˆ 3:71107 ‡ 0:02131DVm 0:004941'
0:02369 ‡ 0:59815 D2; r2 ˆ 0:95
ꢀ2†
ꢀ3†
References and Notes
Inhibition of 5-HT Uptake
1. (a) Carroll, F. I.; Howell, L. L.; Kuhar, M. J. J. Med.
Chem. 1999, 42, 2721. (b) Smith, M. P.; Hoepping, A.; John-
son, K. M.; Trzcinska, M.; Kozikowski, A. P. Drug Des.
Today 1999, 4, 322. (c) Singh, S. Chem. Rev. 2000, 100, 925.
2. Volkow, N. D.; Wang, G. L.; Fowler, J. S.; Logan, J.;
Gatley, S. J.; Hitzemann, R.; Chen, A. D.; Dewey, S. L.;
Pappas, N. Nature 1997, 386, 830.
3. Rothman, R. B.; Elmer, G. I.; Shippenberg, T. S.; Rea, W.;
Baumann, M. H. Ann. N. Y. Acad. Sci. 1998, 844, 59.
4. (a) Carroll, F. I.; Gao, Y. G.; Rahman, M. A.; Abraham,
P.; Parham, K.; Lewin, A. H.; Boja, J. W.; Kuhar, M. J.
J. Med. Chem. 1991, 34, 2719. (b) Carroll, F. I.; Lewin,
A. H.; Boja, J. W.; Kuhar, M. J. In Drug Des. Neuroscience;
Kozikowski, A. P. Ed.; Raven; New York; 1993 pp 149.
5. (a) Smith, M. P.; Johnson, K. M.; Zhang, M.; Flippen-
Anderson, J. L.; Kozikowski, A. P. J. Am. Chem. Soc. 1998,
120, 9072. (b) Smith, M. P.; George, C.; Kozikowski, A. P.
Tetrahedron Lett. 1998, 39, 197. (c) Tamiz, A. P.; Smith, M.
P.; Kozikowski, A. P. Bioorg. Med. Chem. Lett. 2000, 10, 297.
6. Kozikowski, A. P.; Araldi, G. L.; Boja, J.; Meil, W. M.;
Johnson, K. M.; Flippen-Anderson, J. L.; George, C.; Saiah,
E. J. Med. Chem. 1998, 41, 1962.
logKi ˆ 5:6438 ‡ 0:03201DVm 0:000218
DV 2m 0:00364' ‡ 0:0661 D2; r2 ˆ 0:87
As predicted, all equations derived in this study show that
the relative position and the size of the substitutions are
very important in determining activity at the three
monoamine transporters studied in this series. The QSAR
equation developed for the NET shows a linear depen-
dence of log Ki on the size of the O-alkyl substitution,
while the DAT and the SERT exhibit a parabolic depen-
dence. In case of the DAT, the distance between O6B
and C4B does not improve the quality of the QSAR
model suggesting that this parameter is not important
for inhibition of the DAT. However, the distance
between Cl1 and the substitution at O6B signi®cantly
improves the quality of the QSAR equation. Hence, the
relative spatial orientation of the two hydrophobic sub-
stituents is important for DAT inhibition.
7. Tamiz, A. P.; Zhang, J.; Flippen-Anderson, J. L.; Zhang,
M.; Johnson, K. M.; Tella, S.; Kozikowski, A. P. J. Med.
Chem. 2000, 43, 1215.
In conclusion, a series of rigid 6-substituted azabicyclo-
octanes were prepared in their enantiomerically pure
form using a Dieckmann cyclization method. In this
series, oxime (+)-11 exhibits potent activity at the DAT
and moderate activity at the SERT and the NET. The
present work demonstrates that ``front bridged'' bicyclic
piperidines can be tailored to exhibit good potency at
8. Koreeda, M.; Luengo, J. I. J. Org. Chem. 1984, 49, 2081.
9. For (+)-4: monoclinic crystal (0.08Â0.48Â0.58 mm) in
space group P21, a=6.946(2), b=10.766(2), and c=15.631(1)
A, b=96.42(2)ꢀ, V=1162(1) A3, Z=2. Dx=1.34 g cm
,
3
m=2.58 mm 1, F(000)=496, T=293 K. Final agreement fac-
tors were R(F)=0.046 and wR(F2)=0.125 for 1762 observed
data. For (+)-5: monoclinic crystal (0.06Â0.14Â0.20 mm) in
space group P21, a=11.809(1), b=11.942(1), and c=16.926(1)
A, b=104.22(2)ꢀ, V=2314(1) A3, Z=4. Dx=1.34 g cm
,
3
Table 2. Values used for building QSAR equations. Angles and
dihedrals are in degrees, distances are in A, and Vm is in A3
m=2.59 mm 1, F(000)=992, T=293 K. Final agreement fac-
tors were R(F)=0.056 and wR(F2)=0.120 for 2384 observed
data. For (+)-6: monoclinic crystal (0.04Â0.16Â0.84 mm) in
Compd
j
c
D
D2
ÁVm
space group P21, a=12.999(6), b=7.150(3), and c=19.413(6)
(
)-6
(+)-6
)-7
(+)-7
)-8
114.6
114.6
176.3
177.5
35.45
35.37
109.2
109.8
108.8
87.6
87.6
4.74
4.74
5.80
4.36
5.84
4.36
4.77
4.49
4.82
7.26
7.26
10.88
10.9
9.57
9.57
7461
8.93
4.79
0.00
0.00
0.00
0.00
0.00
0.00
49.9
71.5
105.7
A, b=108.93(2)ꢀ, V=1712(1) A3, Z=4. Dx=1.38 g cm
,
3
m=2.23 mm 1, F(000)=744, T=293 K. Final agreement fac-
tors were R(F)=0.057 and wR(F2)=0.140 for 2058 observed
data. All atomic coordinates have been deposited with the
Cambridge Crystallographic Data Centre, and can be
obtained on request from the Director, Cambridge Crystal-
lographic Data Centre, 12 Union Road, Cambridge CB2
1EW, UK or by e-mail from software@ccdc.cam.ac.uk.
(
156.3
160.0
125.6
125.6
89.6
(
(+)-8
(+)-9
(+)-10
(+)-11
90.0
84.1

1686
A. P. Tamiz et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1681±1686
10. Heinisch, G.; Holzer, W.; Kunz, F.; Langer, T.; Luka-
versusky, P.; Pechlaner, C.; Weissenberger, H. J. Med. Chem.
1996, 39, 4058.
11. All compounds described were puri®ed by chromato-
graphy on silica gel, and were characterized by NMR and GC-
MS (EI mode, ionization potential 70 eV). All ®nal com-
pounds gave satisfactory C, H, N analyses.
12. Wang, S.; Sakamuri, S.; Enyedy, I.; Kozikowski, A. P.;
Deschaux, O.; Bandyopadhyay, B. C.; Tella, S. R.; Zaman, W.
A.; Johnson, K. M. J. Med. Chem. 2000, 43, 351.
Pharmacol. 1984, 26, 35. (b) McElvain, J. S.; Schenk, J. O.
Biochem. Pharmacol. 1992, 43, 2189.
14. Prakash, K. R. C.; Tamiz, A. P.; Araldi, G. L.; Zhang,
M.; Johnson, K. M.; Kozikowski, A. P. Bioorg. Med. Chem.
Lett. 1999, 9, 3325.
15. (a) QUANTA, a molecular modeling system, is supplied
by Molecular Simulations Inc., 9685 Scranton Road, San
Diego, CA 92121. (b) Cerius2, a molecular modeling system, is
supplied by Molecular Simulations Inc., 9685 Scranton Road,
San Diego, CA 92121. (c) Rogers, D.; Hop®nger, A. J. J.
Chem. Inf. Comput. Sci. 1994, 34, 854.
13. (a) Javitch, J. A.; Blaustein, R. O.; Snyder, S. H. Mol.