SAR studies of piperidine-based analogues of cocaine. Part 3: Oxadiazoles

Article abstract of DOI:10.1016/S0960-894X(01)00379-1

The synthesis of novel 4β-aryl-1-methyl-3α-(3-substituted-1,2,4-oxadiazol-5-yl) piperidines, bioisosteres of ester (+)-1, is described. The synthesized oxadiazoles were evaluated for their affinity to the DAT and their ability to inhibit monoamine reuptake at the DAT, NET, and 5HTT. The results show that affinity to the DAT and ability to inhibit the reuptake at the DAT, NET, and 5HTT is a function of the size of the substituent in the oxadiazole ring. (+)-(3R,4S)-4β-(4-Chlorophenyl)-1-methyl-3α-(3-methyl-1,2,4- oxadiazol-5-yl)piperidine [(+)-2a], which is structurally and pharmacologically most similar to the ester (+)-1 in this series, showed at least a 2-fold longer duration of action when compared to ester (+)-1.

Full text of DOI:10.1016/S0960-894X(01)00379-1

Bioorganic & Medicinal Chemistry Letters 11 (2001) 2079–2083
SAR Studies of Piperidine-Based Analogues of Cocaine.
Part 3: Oxadiazoles
Pavel A. Petukhov,a Mei Zhang,b Kenneth J. Johnson,b Srihari R. Tellaa
and Alan P. Kozikowskia,*
aDrug Discovery Program, Departments of Neurology and Pharmacology, Georgetown University Medical Center,
3900 Reservoir Road, N.W., Washington, DC 20007, USA
bDepartment of Pharmacology and Toxicology, University of Texas Medical Branch, Galveston, TX 77555-1031, USA
Received 30 March 2001; accepted 11 May 2001
Abstract—The synthesis of novel 4b-aryl-1-methyl-3a-(3-substituted-1,2,4-oxadiazol-5-yl)piperidines, bioisosteres of ester (+)-1, is
described. The synthesized oxadiazoles were evaluated for their affinity to the DAT and their ability to inhibit monoamine reuptake
at the DAT, NET, and 5HTT. The results show that affinity to the DAT and ability to inhibit the reuptake at the DAT, NET, and
5HTT is a function of the size of the substituent in the oxadiazole ring. (+)-(3R,4S)-4b-(4-Chlorophenyl)-1-methyl-3a-(3-methyl-
1,2,4-oxadiazol-5-yl)piperidine [(+)-2a], which is structurally and pharmacologically most similar to the ester (+)-1 in this series,
showed at least a 2-fold longer duration of action when compared to ester (+)-1. # 2001 Elsevier Science Ltd. All rights reserved.
Cocaine continues to be a widely abused drug. The
development of pharmacotherapies that could assist the
patient in initiating and maintaining abstinence and in
relapse prevention is important. Compounds that par-
tially mimic the effects of cocaine (substitute agonists)
and have a long duration of action could potentially be
the best candidates for relapse prevention treatment.1
since cocaine inhibits the transport of all three biogenic
amines with roughly equal potency, compounds that
selectively inhibit the uptake of DA and 5-HT8 or DA
and NE deserve to be evaluated in this context.
Recently, the synthesis, pharmacology, and locomotor
studies of some 3,4-substituted piperidine-based cocaine
analogues have been reported.8ꢀ11 Despite the lack of
the tropane nucleus in these compounds, they represent
a series of potent DA, NE, and 5-HT reuptake inhibi-
tors with Ki values in the nanomolar range. Moreover,
the results of different animal behavioral tests strongly
suggest that ester (+)-1, a modestly DAT/NET selective
compound, could be a possible medication for relapse
prevention during the treatment of cocaine addiction.12
In refining this approach, it is thought that the
development of compounds with a longer duration of
action remains desirable.
Cocaine possesses complex pharmacological properties,
which include the interaction with different neuro-
transmitter targets such as the dopamine (DA), ser-
otonin (5-HT), and norepinephrine (NE) transporters,
cholinergic and s receptors, and sodium and calcium
channels, and others.1ꢀ3 Despite the variety of these
pharmacological properties, the predominant model for
the origin of the reinforcing properties of cocaine is the
blockade of the dopamine transporter (DAT),4 while its
effect on serotonergic systems also has been implicated
in cocaine-seeking behavior in an animal model.5ꢀ7 The
complexity of the mechanisms mediating the addictive
properties of cocaine is such that it is difficult to define
appropriate targets for medication development. For
these purposes, compounds that only partially mimic
the actions of cocaine may be viable candidates. That is,
Oxadiazoles commonly serve as bioisosteric equivalents
of esters with increased stability to biochemical degra-
dation.13,14 Herein we describe the synthesis of new 4b-
(4-chlorophenyl)-3a-(3-substituted-1,2,4-oxadiazol-5-yl)-
piperidines, and we also report the results of pharma-
cological and preliminary behavioral studies.
The synthesis of the oxadiazoles is outlined in Scheme 1
a
previously reported procedure.15,16
*Corresponding author. Tel.: +1-202-687-0686; fax: +1-202-687-
5065; e-mail: kozikowa@georgetown.edu
and follows
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00379-1

2080
P. A. Petukhov et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2079–2083
Oxadiazoles (+)-2a, (+)-2b, (+)-2d–g, (ꢀ)-2b, and (ꢀ)-
2d–g were tested for their affinity as inhibitors of
[3H]WIN 35,248 binding to the DAT in studies using rat
striatal membranes. Compounds (+)-2a–g and (ꢀ)-2a–
g were also tested for their ability to inhibit high affinity
uptake of [3H]DA, [3H]5-HT, and [3H]NE using rat
nerve endings (synaptosomes) obtained from brain
regions enriched in DAT, 5HTT, and NET, respec-
tively, according to protocols described earlier.8,18 The
Ki data and selectivity profiles derived from the Ki
values are provided in Table 1.19
To explore the behavioral consequences of replacing the
ester with a bioisosteric oxadiazolyl group, we examined
compound (+)-2a in the locomotor test in mice and in a
drug-discrimination paradigm in rats according to pro-
tocols reported previously.12 We selected oxadiazole
(+)-2a for these studies because it was found to be the
most similar to the lead ester (+)-1 as regards its affinity
for the three biogenic amine transporters. The results
are shown in Figure 2.
Scheme 1. Reagents and conditions: (a) CH3C(¼NOH)NH2, NaH,
molecular sieves, THF, reflux, 74%; (b) 6 N HCl, reflux; (c) oxalyl
chloride, 1 mL of DMF, CH2Cl2; (d) RC(¼NOH)NH2, pyridine,
CHCl3, reflux; (e) AcOH, reflux.
Oxadiazoles (+)-2a and (ꢀ)-2a were synthesized in
good yield from esters (+)-1 and (ꢀ)-1 and acetamide
oxime.15 In all other cases, esters (+)-1 and (ꢀ)-1 were
converted to the corresponding acids (+)-3 and (ꢀ)-3,
and then to the acid chlorides (+)-4 and (ꢀ)-4.
In general, (ꢀ)-oxadiazoles (ꢀ)-2a–g with the 3S,4R-
configuration of substituents in their piperidine ring
show low potency at all transporters and do not bind to
the DAT, which is consistent with the lower potency of
(ꢀ)-ester 1 in comparison to (+)-ester 1 (see Table 1
and ref 9). Only oxadiazoles with relatively small sub-
stituents, such as methyl and i-propyl, possess residual
activity at micromolar levels at the DAT [for (ꢀ)-2a] or
at the DAT, NET, and 5HTT [for (ꢀ)-2b]. A depen-
dence of potency on the size of the 3-substituent of the
oxadiazole ring is observed for oxadiazoles (+)-2a–g in
3R,4S-series as well. Oxadiazoles (+)-2d–g with a
rather large 3-aryl substituent bearing an additional
substituent in its para-position have low potency at all
transporters (Ki>3 mM). It should be noted that our
lead compound, ester (+)-1, and its bioisostere, the
oxadiazole (+)-2a, exhibit similar pharmacological
profiles.
According to Carroll et al.,15 the reaction between 3b-
(substituted phenyl)tropane-2b-carboxylic acid chlor-
ides and different amidoximes in a 3:1 mixture of pyri-
dine and chloroform at reflux for 1.5 h led to the
corresponding 1,2,4-oxadiazoles. In this one-pot proto-
col, the formation of amidoxime esters16 was followed
by their spontaneous cyclization with formation of the
oxadiazoles. Surprisingly, however, these conditions
afforded in our case only the amidoxime esters (accord-
1
ing to H and 13C NMR). In order to effect cyclization,
the solvent was removed, and the crude amidoxime
esters were heated in glacial acetic acid16 to give the
desired 1,2,4-oxadiazoles (+)-2b–g and (ꢀ)-2b–g.17 The
3
chemical shifts and values of JHꢀH of the oxadiazoles
2a–g and of the starting ester 1 indicate that all of these
compounds exist, within the limits of error, in a single
chair conformation of the piperidine ring where the 3a-
oxadiazolyl and 4b-phenyl groups are equatorial.
Replacement of the methyl group in (+)-2a with an i-
propyl group in (+)-2b leads to a 2.5-fold reduction in
DAT binding and to a 2.1- and 3.0-fold reduction in
blocking the reuptake of DA and NE, respectively. The
Table 1. Ki’s and selectivities for compounds (+)-1, (ꢀ)-1, (+)-2a–g, (ꢀ)-2a–g, and cocaine
Ki (nM)bꢁSE (nM)
Selectivity
NE/DA
Compd
[3H]WINa Binding
[3H]DA Uptake
[3H]5-HT Uptake
[3H]NE Uptake
5-HT/DA
NE/5-HT
Cocaine
(+)-1
146ꢁ2.4
315ꢁ52.8
2005ꢁ610
201ꢁ2.00
N/Tc
91494ꢁ1033
2900ꢁ70
N/T
259ꢁ19.9
233.4ꢁ62.0
2930ꢁ580
187.2ꢁ3.0
954.9ꢁ7.6
ꢁ64.8
155ꢁ0.40
8490ꢁ1430
930ꢁ10
108ꢁ3.50
252ꢁ43.4
396ꢁ112
256ꢁ16.9
>1000d
778ꢁ40.37.49
2680ꢁ600
>10,000d
>3000d
0.37
36.4
0.32
29.5
—
1.99
0.34
—
0.20
1.08
0.14
1.27
—
0.26
0.75
—
0.70
0.03
0.43
0.04
—
(ꢀ)-1
(+)-2a
(ꢀ)-2a
5960ꢁ80
>3000d
(+)-2b
(ꢀ)-2b
2920ꢁ190
1220ꢁ230
>3000d
>3000d
3590ꢁ80
2.19
—
—
—
(+)-2c
(ꢀ)-2c
497.3ꢁ95.1
1713.2ꢁ141.5
>3000
N/T
>3000
—
—
—
—
(+)-2d-g, (ꢀ)-2d-g
>3000
>3000
aWIN 35,248.
bData are meanꢁstandard error of at least three experiments as described in ref 8.
cNot tested.
dIC50 is shown; data do not allow the determination of Ki.

P. A. Petukhov et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2079–2083
2081
potency of (+)-2b in blocking 5-HT transport is 2.0-
fold higher than the potency of (+)-2a. The pharmaco-
logical differences between (+)-2c and (+)-2b are even
larger than those between (+)-2b and (+)-2a. With the
introduction of an aromatic substituent, oxadiazole
(+)-2c loses its activity at the 5-HT and NE transpor-
ters, while at the DAT, (+)-2c is only 2.6- and 1.3-fold
less potent than (+)-2a and (+)-2b, respectively.
differences in the assay conditions do not allow us to
compare our data directly with those obtained by Car-
roll’s group15,20 it is clear that the introduction of ali-
phatic or aromatic substituents into the oxadiazole ring
in the tropane series leads primarily to changes in selec-
tivity, making the compounds more NET or 5HTT
active with almost constant potency at the DAT. In the
piperidine series such changes are not observed. For
instance, the analogue of (+)-2a in the 3b-aryl-2b-oxa-
We observed large differences in the pharmacological
properties of our 4b-(4-chlorophenyl)-3a-oxadiazolylpi-
peridines relative to those reported earlier by Carroll
et al.15 for 3b-aryl-2b-oxadiazolyltropanes. Even if
diazolyltropane
series,
3b-(4-chlorophenyl)-2b-(3-
methyl-1,2,4-oxadiazol-5-yl)tropane, was DAT and NET
selective (binding to DAT/NET/5HTT, IC50=4.05, 363,
and 2580 nM,15 respectively), while the analogue of (+)-
2d in the tropane series, 3b-(4-chlorophenyl)-2b-(3-
methyl-1,2,4-oxadiazol-5-yl)tropane, was DAT and
5HTT selective (binding to DAT/NET/5HTT,
IC50=4.06, 4070, and 404 nM,15 respectively).21 It
should be noted, however, that the analogue of (+)-2a
in the 3b-aryl-2a-oxadiazolyltropane series was poorly
active at all transporters (binding to DAT/NET/5HTT,
IC50=1030, 71,000, and 33,100 nM,15 respectively).
Our results suggest that additional steric restrictions
should be included in the pharmacophore models for
the DAT,3,22 NET, and 5HTT binding sites in the case
of piperidine-based analogues of cocaine with certain
3a-substituents. We hypothesize that the binding pocket
for the 3a-substituent is only about 5.5 A long23 as
deduced from the distance between C3of the piperidine
ring and the carbon atom of the methyl substituent in
the oxadiazole ring in compound (+)-2a (Fig. 1A).
Figure 1. Pharmacophore model for 4b-aryl-3a-oxadiazolylpiperidines
(A); depiction of the space occupied by the oxadiazolyl group during
its rotation around the bond between C3of the piperidine ring and C5
of the oxadiazole ring (B).
Figure 2. Comparison of (+)-2a and cocaine in locomotor stimulation (A–D) and drug discrimination (E and F) tests.

2082
P. A. Petukhov et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2079–2083
Therefore compounds containing a 3a-substituent that
are not able to adjust the position in space of that sub-
stituent and thereby minimize its unfavorable interac-
tion with the binding pocket cannot effectively bind to
the transporter and will have low potency. In com-
pounds (+)-2a and (+)-2b, the 3a-substituent is small
enough to fit into the pocket (Fig. 1A) and thus to bind
to the transporter. In compounds (+)-2c–g, on the
other hand, the 3a-substituent can only occupy posi-
tions within a conical region of the space extending for a
distance of 10 A from the piperidine ring and reaching
into the prohibited region (Fig. 2B), thus making bind-
ing to the transporter difficult. On the other hand,
dimeric piperidine-based esters and amides bearing a
sufficiently long and flexible linker10 can easily adjust
the position in space of the 3a-substituent, thereby
avoiding unfavorable steric interactions. The exact pla-
cement of the 3a-substituent in space in these models is
unknown.
peridines. The greater duration of action of the oxadia-
zole (+)-2a in comparison to the ester (+)-1 combined
with its similar pharmacological and preliminary beha-
vioral profile make it an interesting candidate for addi-
tional study. The longer duration of action of the
oxadiazole (+)-2a may offer some therapeutic advantage.
Acknowledgements
The authors are indebted to NIH, National Institute of
Drug Abuse (DA11548 and 10458) for their support of
this work. We also thank Jamie Biswas (CTDP/NIDA/
NIH) and Aaron Janowsky (OHSU) for the human
transporter data for (+)-2a.
References and Notes
1. Carroll, F. I.; Howell, L. L.; Kuhar, M. J. J. Med. Chem.
1999, 42, 2721.
2. Smith, M. P.; Hoepping, A.; Johnson, K. M.; Trzcinska,
M.; Kozikowski, A. P. Drug Discov. Today 1999, 4, 322.
3. Carroll, F. I.; Lewin, A. H.; Boja, J. W.; Kuhar, M. J. J.
Med. Chem. 1992, 35, 969.
4. Boja, J. W.; Vaughan, R.; Patel, A.; Shaya, E. K.; Kuhar,
M. J. In Dopamine Receptors and Transporters; Niznik, H. B.,
Ed.; Marcel Dekker: New York, 1994; p 611.
5. Schenk, S. Pharmacol. Biochem. Behav. 2000, 67, 363 and
reference therein.
6. Castanon, N.; Scearce-Levie, K.; Lucas, J. J.; Rocha, B.;
Hen, R. Pharmacol. Biochem. Behav. 2000, 67, 559.
7. Singh, S. Chem. Rev. 2000, 100, 925.
8. Tamiz, A. P.; Zhang, J.; Flippen-Anderson, J. L.; Zhang,
M.; Johnson, K. M.; Deschaux, O.; Tella, S.; Kozikowski,
A. P. J. Med. Chem. 2000, 43, 1215.
9. Kozikowski, A. P.; Araldi, G. L.; Boja, J.; Meil, W. M.;
Johnson, K. M.; Flippen-Andersen, J. L.; George, C.; Saiah,
E. J. Med. Chem. 1998, 41, 1962.
10. Tamiz, A. P.; Zhang, J.; Zhang, M.; Wang, C.; Johnson,
K. M.; Kozikowski, A. P. J. Am. Chem. Soc. 2000, 122, 5393.
11. Tamiz, A. P.; Conti, P.; Zhang, M.; Johnson, K. M.;
Kozikowski, A. P. Bioorg. Med. Chem. Lett. 2000, 10, 2741.
12. Kozikowski, A. P.; Deschaux, O.; Bandyopadhyay, B. C.;
Araldi, G. L.; Munzar, P.; Fishman, A. J.; Smith, M. P.;
Babich, J. W.; Johnson, K. M.; Balster, R. L.; Beardsley, P.
M.; Tella, S. R. J. Pharmacol. Exp. Ther. submitted.
13. Patani, G. A.; LaVoie, E. J. Chem. Rev. 1996, 96, 3147.
14. Cramer, R. D.; Clark, R. D.; Patterson, D. E.; Ferguson,
A. M. J. Med. Chem. 1996, 39, 3060.
The oxadiazole (+)-2a is the most potent of the present
compounds, and it is completely cocaine-like in both
locomotor and drug-discrimination tests. Both cocaine
(5.6–30 mg/kg, Fig. 2A and B) and oxadiazole (+)-2a
(5.6–100 mg/kg, Fig. 2C and D) produced dose-depen-
dent locomotor stimulation in mice. Both cocaine (56
mg/kg) and (+)-2a (156 mg/kg) produced convulsions
within 5 min following drug injection. This suggests that
(+)-2a readily enters the brain similar to cocaine.
Within the range of nonconvulsant doses, there were
significant differences in the duration of locomotor
effects of cocaine versus (+)-2a. The locomotor stimu-
lation by (+)-2a at moderate to high doses (15.6–100
mg/kg) lasted at least 4 h, while cocaine produced sti-
mulation that lasted only 2 h at the maximal non-
convulsant dose (30 mg/kg). However, cocaine and
compound (+)-2a had similar efficacies in increasing
the distance traveled as indicated by their similar max-
imal effects. Oxadiazole (+)-2a appeared to have
greater efficacy in increasing stereotypic behavior than
cocaine.
In rats trained to discriminate cocaine from saline, both
(+)-2a (1–10 mg/kg), and cocaine (1–10 mg/kg, Fig.
2E,F) produced dose-dependent and full substitution
and had similar potencies. The doses of cocaine and (+)-
2a necessary to produce 50 percent cocaine-appropriate
responding were 3.98 (95% confidence limits: 3.16–5.13)
mg/kg and 3.55 (95% confidence limits: 2.63–4.57) mg/kg,
respectively.
15. Carroll, F. I.; Gray, J. L.; Abraham, P.; Kuzmenko,
M. A.; Lewin, A. H.; Boja, J. W.; Kuhar, M. J. J. Med. Chem.
1993, 36, 2886.
16. Eloy, F.; Lenaers, R. Chem. Rev. 1962, 62, 155.
17. All oxadiazoles are characterized by 1H and 13C NMR, IR
and MS of the free bases and by elemental analysis of their
hydrochloride salts. The results are in agreement with the
Being DAT-NET selective, oxadiazole (+)-2a produces
behavioral effects that are similar to those of the lead
compound, ester (+)-1.12 The results of the present
behavioral tests are consistent with the compound’s DA
reuptake inhibitory activity, however, further studies
will be required to evaluate the effects if any of the
compound’s lack of 5HTT activity.
3
assigned structures. For (+)-2a all JHꢀH coupling constants
were determined using proton–proton homonuclear decoupling.
18. Wang, S.; Sakamuri, S.; Enyedy, I. J.; Kozikowski, A. P.;
Deschaux, O.; Bandyopadhyay, C.; Tella, S. R.; Zaman,
W. A.; Johnson, K. M. J. Med. Chem. 2000, 43, 351.
19. Oxadiazole (+)-2a was independently tested by NIDA
within the framework of the CTDP screening program for its
effects on radioligand ([125I]RTI-55) binding to and [3H]dopa-
mine uptake by HEK cells expressing cDNA for the human
dopamine transporter (HEK-hDAT cells), its effect on
In conclusion, the results of the SAR studies suggest
that additional steric restrictions, such as the prohibited
conical region for the 3a-substituent, should be included
in the pharmacophore model for these oxadiazolylpi-

P. A. Petukhov et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2079–2083
2083
radioligand ([125I]RTI-55) binding and [3H]serotonin uptake
by HEK cells expressing cDNA for the human serotonin
transporter (HEK-hSERT cells), and its effect on radioligand
([125I]RTI-55) binding and [3H]norepinephrine uptake by
HEK cells expressing cDNA for the human norepinephrine
transporter (HEK-hNET cells). In these assays (+)-2a was
highly selective for the dopamine transporter (IC50=183nM;
cocaine, IC50=263nM) with no significant activity at the ser-
otonin (Ki>10 mM; cocaine IC50=471 nM) and nor-
epinephrine transporters (IC50=1370 nM; cocaine, IC50=281
nM). The experiments were performed as described in Eshle-
man, A. J.; Carmolli, M.; Cumbay, M.; Martens, C. R.; Neve,
K. A.; Janowsky, A. J. Pharmacol. Exp. Ther. 1999, 289, 877.
20. See discussion in Kozikowski, A. P.; Araldi, G. L.; Pra-
kash, K. R. C.; Zhang, M.; Johnson, K. M. J. Med. Chem.
1998, 41, 4973.
21. Changing the relative orientation of the nitrogen lone pair
provides another way for tuning the selectivity of monoamine
transporter inhibitors. See discussion in Smith, M. P.; John-
son, K. M.; Zhang, M.; Flippen-Anderson, J. L.; Kozikowski,
A. P. J. Am. Chem. Soc. 1998, 120, 9072.
22. Lieske, S. F.; Yang, B.; Eldefrawi, M. E.; MacKerell,
A. D., Jr.; Wright, J. J. Med. Chem. 1998, 41, 864.
23. Molecular modeling calculations were performed using
MM2 force field in Chem3D 5.0 Pro software, Cambridge Soft
Corp.