Molecular modeling, structure-activity relationships and functional antagonism studies of 4-hydroxy-1-methyl-4-(4-methylphenyl)-3-piperidyl 4-methylphenyl ketones as a novel class of dopamine transporter inhibitors

Article abstract of DOI:10.1016/S0968-0896(01)00090-6

We previously disclosed the discovery of 4-hydroxy-1-methyl-4-(4-methylphenyl)-3-piperidyl 4-methylphenyl ketone (3) as a novel class of dopamine transporter (DAT) inhibitors and showed that (±)-3 has a significant functional antagonism against cocaine in vitro. Our previous preliminary structure-activity relationship study led to identification of a more potent DAT inhibitor [(±)-4] but this compound failed to show any significant functional antagonism. To search for more potent analogues than 3 but still displaying significant functional antagonism, further SARs, molecular modeling studies and in vitro pharmacological evaluation of this novel class of DAT inhibitors were performed. Sixteen new analogues were synthesized in racemic form and evaluated as DAT inhibitors. It was found that seven new analogues are reasonably potent DAT inhibitors with K_i values of 0.041-0.30 and 0.052-0.16μM in [³H]mazindol binding and inhibition of DA reuptake. Chiral isomers of several potent DAT inhibitors were obtained through chiral HPLC separation and evaluated as inhibitors at all the three monoamine transporter sites. In general, the (-)- isomer is more active than the (+)-isomer in inhibition of DA reuptake and all the (-)-isomers are selective inhibitors at the DAT site. Evaluation of cocaine's effect on dopamine uptake in the presence and absence of (+)-3 and (-)-3 showed that (-)-3 is responsible for the functional antagonism obtained with the origin lead (±)-3 in binding and inhibition of DA reuptake, appeared to have improved functional antagonism as compared to (±)-3. Copyright

Full text of DOI:10.1016/S0968-0896(01)00090-6

Bioorganic & Medicinal Chemistry 9 (2001) 1753–1764
Molecular Modeling, Structure–Activity Relationships and
Functional Antagonism Studies of 4-Hydroxy-1-methyl-4-(4-
methylphenyl)-3-piperidyl 4-Methylphenyl Ketones as a Novel
Class of Dopamine Transporter Inhibitors
Shaomeng Wang,a,* Sukumar Sakamuri,b Istvan J. Enyedy,a Alan P. Kozikowski,b
Wahiduz A. Zamanc and Kenneth M. Johnsonc
aDepartments of Oncology and Neuroscience, Georgetown University Medical Center, Building D,
Room 235/237 4000 Reservoir Rd, Washington, DC 20007, USA
bDrug Discovery Program, Department of Neurology, Georgetown University Medical Center, 3900 Reservoir Rd,
Washington, DC 20007, USA
cDepartment of Pharmacology and Toxicology, University of Texas Medical Branch, Galveston, TX 77555-1031, USA
Received 6 November 2000; accepted 6 March 2001
Abstract—We previously disclosed the discovery of 4-hydroxy-1-methyl-4-(4-methylphenyl)-3-piperidyl 4-methylphenyl ketone (3)
as a novel class of dopamine transporter (DAT) inhibitors and showed that (Æ)-3 has a significant functional antagonism against
cocaine in vitro. Our previous preliminary structure–activity relationship study led to identification of a more potent DAT inhibitor
[(Æ)-4] but this compound failed to show any significant functional antagonism. To search for more potent analogues than 3 but
still displaying significant functional antagonism, further SARs, molecular modeling studies and in vitro pharmacological evalua-
tion of this novel class of DAT inhibitors were performed. Sixteen new analogues were synthesized in racemic form and evaluated
as DAT inhibitors. It was found that seven new analogues are reasonably potent DAT inhibitors with Ki values of 0.041–0.30 and
0.052–0.16 mM in [3H]mazindol binding and inhibition of DA reuptake. Chiral isomers of several potent DAT inhibitors were
obtained through chiral HPLC separation and evaluated as inhibitors at all the three monoamine transporter sites. In general, the
(À)-isomer is more active than the (+)-isomer in inhibition of DA reuptake and all the (À)-isomers are selective inhibitors at the
DAT site. Evaluation of cocaine’s effect on dopamine uptake in the presence and absence of (+)-3 and (À)-3 showed that (À)-3 is
responsible for the functional antagonism obtained with the original lead (Æ)-3. Out of the new compounds synthesized, analogue
(Æ)-20, which is 8- and 3-fold more potent than (Æ)-3 in binding and inhibition of DA reuptake, appeared to have improved
functional antagonism as compared to (Æ)-3. # 2001 Elsevier Science Ltd. All rights reserved.
Introduction
abuse that are currently being pursued have extensively
been reviewed.1
Cocaine (1) is one of the most addictive substances
known. Cocaine abuse remains one of the greatest con-
cerns of the American public today, and has therefore
become a concern of medical, social, and political lea-
ders. Despite intensive research efforts, development of
an effective cocaine therapy remains a difficult task and
no specific and effective therapy for the treatment of
cocaine abuse is currently available. Recently, the
potential pharmacotherapies for treatment of cocaine
Although cocaine (1) potently inhibits the reuptake of
both norepinephrine (NE) and serotonin (SER) (Table
3), many lines of evidence indicate that its ability to act
as a reinforcer stems from its ability to inhibit the
reuptake of dopamine (DA) into dopaminergic neu-
rons.2À9 Cocaine exerts this effect by binding to a spe-
cific site on the dopamine transporter (DAT) that can
be distinguished from the DA site by mutational analy-
sis.5 The consequent increase of dopaminergic trans-
mission in the reward mediating brain mesolimbic
system is the essence of the dopamine hypothesis of
reinforcement advanced by Wise10 and later elaborated
*Corresponding author. Tel.:+1-202-687-2028; fax: +1-202-687-
0617; e-mail: wangs@georgetown.edu
0968-0896/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(01)00090-6

1754
S. Wang et al. / Bioorg. Med. Chem. 9 (2001) 1753–1764
for cocaine by Kuhar.4 In pursuit of an effective ther-
apy for the treatment of cocaine abuse, significant
chemical, biochemical, pharmacological and behavioral
studies have been performed on several classes of DAT
inhibitors, including cocaine/tropane analogues,3,11À19
GBR compounds,20À22 methylphenidate analogues,23
mazindol analogues,24 and piperidine-based ana-
logues.25À29
uptake activities. The structures of these analogues (12–27)
are shown in Table 1 and their activities are summarized
in Table 2. For purpose of comparison, the structures
and activities of nine analogues (3–11) from our pre-
vious study,30 including the lead compound 3, are also
included in the tables.
Table 1. Chemical structures of piperidinols
We are interested in the discovery of novel DAT inhi-
bitors that can be used as either cocaine antagonists or
‘partial agonists’.30 For this purpose, we have used a
novel 3D-database pharmacophore searching approach
to discover novel DAT inhibitors.30 Novel DAT inhibi-
tors are then evaluated as potential cocaine antagonist
in an in vitro functional antagonism assay.30 A DAT
inhibitor with a significant functional antagonism sug-
gests that the inhibitor is capable of reducing the bind-
ing of cocaine, either by steric hindrance or by an
allosteric mechanism, while at the same time having a
relatively smaller effect on DA transport. Using this
approach, we have recently reported the discovery of
hydroxypiperdines as a novel class of DAT inhibitors
with significant functional antagonism.30 Functional
antagonism studies showed that the lead compound
[(Æ)-3] had the ability to significantly antagonize cocai-
ne’s inhibition of DA transport, representing a promis-
ing lead for further development. Initial SAR studies led
to the identification of a potent analogue [(Æ)-4], with
Ki values of 0.011 mM in [3H]mazindol binding and
0.051 mM in inhibition of DA reuptake but with no sig-
nificant functional antagonism against cocaine.30
Furthermore, since the antagonism of 3 was origin-
ally obtained with its racemic form, it was unclear
which stereoisomer is responsible for the observed
functional antagonism.
Compound
R1
R2
R3
R4
3
4
5
6
7
Me
Cl
H
F
Cl
H
Cl
H
H
H
H
H
H
H
Cl
Me
Me
Me
Me
Me
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Cl
Br
I
Ethyl
H
H
F
Me
Cl
Me
Cl
H
H
H
H
H
Cl
F
Me
Me
H
H
H
H
H
Cl
H
H
H
H
H
H
H
H
H
H
H
Me
Me
Me
Me
Me
Me
Me
Me
The major scopes of this investigation are to better
understand the SAR of this novel class of DAT inhi-
bitor and to identify and characterize new analogues
with significant functional antagonism.
Me
Ethyl
Ethyl
Ethyl
(CH2)2Ph
(CH2)2Ph
(CH2)3Ph
(CH2)3Ph
Results and Discussion
H
SARs of racemic compounds in inhibition of DA
reuptake
Cl
Cl
Br
Cl
Cl
Cl
H
Cl
H
A total of 16 new analogues were synthesized in racemic
form and tested as inhibitors at the DAT site for their
Br

S. Wang et al. / Bioorg. Med. Chem. 9 (2001) 1753–1764
1755
Table 2. The activity of racemic compounds in inhibition of DA
reuptake
Our preliminary SAR studies suggested that the sub-
stituents on each phenyl ring have considerable influ-
ence on their binding and uptake activities.30 To gain a
better understanding of the effects of phenyl substitu-
tions, analogues 12–20 were made and tested. Lead
compound 3 with a p-methyl substituent in both phenyl
rings has a Ki value of 0.36 mM in inhibition of DA
uptake. A p-fluoro substituent in both phenyl rings
resulted in an analogue (6) with a Ki value of 4.20 mM in
inhibition of DA reuptake, approximately 12-fold less
potent than 3, suggesting that a substituent at the para
position of both phenyl rings has a significant effect on
the uptake activity. To investigate further this effect,
analogues 12–15 were synthesized and evaluated. A p-
chloro substituent in both phenyl rings resulted in ana-
logue 12 with a Ki value of 0.16 mM in inhibition of DA
reuptake, approximately 2 times more potent than 3.
Compared to the unsubstituted analogue 5, the p-chloro
substituent in both phenyl rings improved the activity in
inhibition of DA reuptake by 16-fold. Analogue 13 with
a p-bromo substituent in both phenyl rings has a Ki
value of 0.085 mM in inhibition of DA uptake, repre-
senting a 30-fold improvement compared to the unsub-
stituted analogue 5. A p-iodo substituent in both phenyl
rings resulted in analogue 14, with a Ki value of
0.124 mM in inhibition of DA uptake, a 21-fold
improvement compared to the unsubstituted analogue 5
but slightly less potent than 13 with a p-bromo sub-
stituent in both phenyl rings. Analogue 15 with a p-ethyl
substituent in both phenyl rings has a Ki value of
8.09 mM in inhibition of DA uptake, a 22-fold loss in
activity as compared to the lead compound 3 with a
p-methyl substituent. Therefore, a bromo atom appears
to be optimal to achieving the highest potency in inhi-
bition of DA reuptake among analogues 3, 5, 6, 12, 13,
14, and 15 with a mono substituent at the para position
in both phenyl rings.
Compound
Ki (mM)
[3H]-DA uptake
1a
0.27Æ0.02b
0.36Æ0.03
0.051Æ0.008
2.59Æ0.23
4.20Æ0.07
5.85Æ1.08
15.80Æ0.65
4.22Æ0.02
7.52Æ0.09
9.48Æ0.74
0.16Æ0.02
0.085Æ0.019
0.12Æ0.00
8.09Æ0.53
23.34Æ1.06
0.16Æ0.01
0.89Æ0.10
0.10Æ0.01
0.11Æ0.02
1.14Æ0.03
0.57Æ0.03
0.052Æ0.004
0.39Æ0.01
1.49Æ0.01
0.50Æ0.02
1.40Æ0.15
3a
4a
5a
6a
7a
8a
9a
10a
11a
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
aData taken from ref 30.
bMean Æ standard error of 2–3 experiments, each conducted using six
concentrations of drug in triplicate.
The compounds were synthesized using known methods
from the literature. All substituted acetophenones were
purchased from commercial suppliers, except compound
29, which was prepared in 90% yield by reacting the
substituted phenacyl chloride 28 with SnCl2 and NaI in a
5:1 mixture of THF and H2O for 2 h (Scheme 1).35 The
general synthesis of the compounds 12–27 is illustrated in
Scheme 2. Briefly, reaction of an amine hydrochloride with
an excess of an aryl methyl ketone (29–44) and para-
formaldehyde in the presence of a catalytic amount of
hydrochloric acid led to compounds 12–27 (Scheme 2).36
To investigate the effect of the substitution positions
(ortho, meta, and para) in the phenyl rings, two ana-
logues, 16 and 17 with a mono chloro substituent at
either the ortho or meta position in both phenyl rings
were synthesized. They were tested as inhibitors of DA
reuptake and compared to analogue 12 with a chloro
substituent at the para position in both phenyl rings.
Analogue 16 with a chloro substituent at the ortho
position in both phenyl rings has a Ki value of 23.34 mM
in inhibition of DA uptake, a 65-fold less potent than
analogue 12. Analogue 17 with a chloro substituent at
the meta position in both phenyl rings has a Ki value of
0.16 mM in inhibition of DA uptake, comparable to the
Scheme 1. Synthesis of intermediate 29.
Scheme 2. Synthesis of analogues 12–27.

1756
S. Wang et al. / Bioorg. Med. Chem. 9 (2001) 1753–1764
activity of analogue 12. Since 12, 16, and 17 have very
similar hydrophobicity, the difference between 12, 16,
and 17 in their activities in inhibition of DA reuptake is
likely due to their difference in conformational profiles
and this has been further investigated through our
molecular modeling studies (see Molecular Modeling).
corresponding N-methyl analogue 12. The only excep-
tion is analogue 23 with a N-ethyl substituent and 3,4-
dichloro substituents in both phenyl rings, which has a
Ki value of 0.052 mM in inhibition of DA uptake, as
potent as its corresponding N-methyl analogue 4 (Ki
equal to 0.051 mM). With larger N-substituents, the
potency of the analogues decreases even more. Analo-
gue 24 with a N-ethylphenyl substituent and 3,4-
dichloro substituents in both phenyl rings has a Ki value
of 0.39 mM, 8-fold less potent than its corresponding N-
methyl analogue 4. Analogue 25 with a N-ethylphenyl
substituent and a 4-bromo substituent in both phenyl
rings has a Ki value of 1.49 mM, 18-fold less potent than
its corresponding N-methyl analogue 13. Analogue 26
with a N-propylphenyl substituent and 3,4-dichloro
substituents in both phenyl rings has a Ki value of
0.50 mM, 10-fold less potent than its corresponding N-
methyl analogue 4. Analogue 27 with N-propylphenyl
substituent and a 4-bromo substituent in both phenyl
rings has a Ki value of 1.40 mM, 17-fold less potent than
its corresponding N-methyl analogue 13. Therefore, it
appeared that a N-methyl substituent is optimal for the
activity in inhibition of DA reuptake with this class of
compounds.
Our previous SAR studies led to the identification of a
potent analogue (4) with 3,4-dichloro substituents in
both phenyl rings.30 To explore further the SARs of
compounds with two substituents in both phenyl rings,
analogues 18, 19, and 20 were made and tested. Analo-
gue 18 with 3,4-difluoro substituents has a Ki value of
0.89 mM in the inhibition of DA uptake, a 5-fold
improvement as compared to the 4-fluoro, mono-sub-
stituted analogue 6 but 17-times less potent than the 3,
4-dichloro substituted analogue 4. Analogue 19 with
3,4-dimethyl substituents has a Ki value of 0.10 mM in
inhibition of DA uptake, only 2-fold less potent than 4.
It is interesting to note that as compared to the 3,4-
dichloro substituted analogue 4, the activity of 19 in
inhibition of the reuptake of DA is only 2-fold less
potent, but the binding affinity for the [3H]mazindol
site is reduced by 28-fold (Table 3). Analogue 20 with
4-chloro-3-methyl-substituents has
a Ki value of
0.11 mM in inhibition of DA uptake, 2-fold less potent
than 4 but 3-fold more potent than the lead compound
3.
Selectivity of potent racemic compounds at the three
monoamine transporters (DAT, SERT, and NET)
Cocaine potently inhibits the reuptake of DA, but even
more potently inhibits the reuptake of serotonin (SER)
and norepinephrine (NE) (Table 3). To assess the
selectivity of several potent analogues (12, 13, 17,
19, 20, and 23), we evaluated them as reuptake
inhibitors at SERT and NET sites and the results
are summarized in Table 3. The selectivity between
DAT relative to SERT for analogues 4, 12, 13, 17,
19, 20, and 23 is 47-, 6-, 22-, 3-, 8-, 10-, and 43-
fold, respectively. Thus, analogues 4 and 23 have the
best selectivity between DAT and SERT, being 47- and
43-fold, respectively. The selectivity between DAT and
NET for analogues 4, 12, 13, 17, 19, 20, and 23 is 4-, 8-,
22-, 3-, 8-, 6-, and 11-fold, respectively. Therefore, in
contrast to cocaine, in general, these inhibitors are more
potent at the DAT site among these three transporter
sites.
Previous investigations of a tropane series of com-
pounds showed that fairly large N-substituents may be
tolerated for their activity in inhibition of DA reup-
take.31,32 To investigate the effect of N-substitutions in
this class of DAT inhibitors, a series of analogues were
synthesized and tested (21–27). In contrast to the SAR
of tropane analogues,31,32 replacement of the N-methyl
group with larger substituents in general decreases the
activity in inhibition of the reuptake of DA in this class
of compounds. Analogue 21 with a N-ethyl substituent
and a 4-methyl substituent at both phenyl rings has a Ki
value of 1.14 mM in inhibition of DA uptake, 3-fold less
potent than its corresponding N-methyl analogue 3.
Analogue 22 with a N-ethyl substituent and a 4-Cl sub-
stituent at both phenyl rings has a Ki value of 0.57 mM
in inhibition of DA uptake, 4-fold less potent than its
Table 3. The activities and selectivity of selected racemic compounds at three monoamine transporter sites
Compound
Ki (mM)
Selectivity
Binding
Uptake
SER
DA
NE
DA
([3H]-Mazindol)
([3H]-DA)
([3H]-SER)
([3H]-NE)
1a (cocaine)
0.23Æ0.02b
0.49Æ0.03
0.011Æ0.001
0.25Æ0.01
0.23Æ0.05
0.16Æ0.01
0.30Æ0.03
0.059Æ0.013
0.041Æ0.001
0.27Æ0.02
0.36Æ0.03
0.16Æ0.00
1.63Æ0.15
2.38Æ0.01
0.93Æ0.01
1.88Æ0.19
0.46Æ0.04
0.81Æ0.01
1.16Æ0.11
2.22Æ0.19
0.11Æ0.01
3.86Æ0.07
0.18Æ0.05
1.21Æ0.06
1.86Æ0.09
0.43Æ0.08
1.01Æ0.09
0.64Æ0.06
0.61Æ0.11
0.6
4.5
46.7
5.8
22.1
2.9
8.0
0.4
10.7
3.5
7.6
21.9
2.8
10.0
5.7
11.7
3a
4a
0.051Æ0.008
0.16Æ0.02
0.085Æ0.019
0.16Æ0.01
0.10Æ0.01
0.11Æ0.02
0.052Æ0.004
12
13
17
19
20
23
10.4
42.7
aData taken from ref 30.
bMean Æ standard error of 2–3 experiments, each conducted using six concentrations of drug in triplicate.

S. Wang et al. / Bioorg. Med. Chem. 9 (2001) 1753–1764
1757
SARs of enantiomers and their selectivity at the three
transporter sites
two compounds is a favorable binding element for the
DAT and/or an unfavorable binding element relative to
the SERT and NET sites.
Each of these compounds shown in Table 1 was syn-
thesized in racemic form and the results are somewhat
complicated by the presence of two stereoisomers. To
solve this problem, we obtained the individual enantio-
mers of 4 potent analogues, 3 (the lead compound), 4
(the most potent analogue), 19 and 20 (potent analogues
possessing significant cocaine antagonism) through chiral
HPLC separation (see Experimental). These enantiomers
were then evaluated as inhibitors at all the three trans-
porter sites and the results are summarized in Table 4.
Molecular Modeling Studies
Our SAR studies showed that the nature and the posi-
tion of the substituent(s) on both phenyl rings play an
important role in the binding and reuptake activities for
this class of compounds. To achieve a better under-
standing of factors important to their activities and to
guide our further design of new analogues, we have
carried molecular modeling studies. Based upon the
data obtained for eight chiral compounds (Table 4), the
(À)-isomer is always the more active compound between
the two isomers and its activity is approximately 2 times
more potent than its racemic form. Because we have much
more extensive SAR with the racemic compounds and
have very limited data on chiral compounds, molecular
modeling study was based upon the activity in inhibi-
tion of DA reuptake of the racemic compounds. How-
ever, the molecular modeling results based upon the
SARs of the racemic compounds should be applicable
to both racemic compounds and their (À)-isomers.
Overall, the (À)-isomers are more potent inhibitors than
the (+)-isomers at the DAT site with each compound.
Thus, the (À)-isomer for each compound has an activity
in inhibition of DA reuptake approximately 2-fold
greater than its respective racemic form. The difference
in activities between the (+)- and (À)-isomers in inhi-
bition of DA reuptake is fairly large for 3 and 19, being
7.6- and 10.8-fold, respectively. Interestingly, the differ-
ence between the (+)- and (À)-isomers is relatively
small for 4 and 20, being only 1.4- and 2.5-fold, respec-
tively. Comparison between their structures showed that
3 and 19 have a para-methyl substituent in each phenyl
ring, while 4 and 20 have a para-Cl substituent in each
phenyl ring. Replacement of the para-methyl group in
19 with the para-Cl group resulted in 20. The improve-
ment in activities in inhibition of DA reuptake is 7.0-
fold from (+)-19 to (+)-20 but is only 1.6-fold from
(À)-19 to (À)-20. These results suggested that (+)- and
(À)-isomers may bind to the DAT in different manner
and a para-Cl group in each phenyl ring more sig-
nificantly improves the potency for the (+)-isomers
than for the (À)-isomers.
Analogues 12, 16, and 17 all have a mono chloro sub-
stituent on each phenyl ring. While 12 and 17 have
similar activities in inhibition of the reuptake of DA (Ki
values of 0.160 and 0.155 mM respectively), 16 (Ki value
of 23.34 mM) is approximately 150-fold less potent than
12 and 17. Since 12, 16 and 17 only differ in the position
of the chloro substituent in both phenyl rings and have
similar size and hydrophobicity, the substantial differ-
ence in their activities is thus likely due to their con-
formational preferences. Conformational analysis
showed that the lead compound 3, analogues 12 and 17
have virtually identical conformational profiles (i.e.,
they have essentially identical lowest energy conformations
and also have similar low minimum conformations).
The superposition of 3 and 12 in their lowest energy
conformations is shown in Figure 1A. However, analogue
16 has a significantly different conformational profile
from 3, 12, and 17. In its lowest energy conformation of
16, the phenyl ring attached to the carbonyl group
adopts a significantly different orientation from that in
3, 12, and 17, deviated by approximately 40ꢀ. The
Furthermore, the (+)- and (À)-isomers have different
selectivity profiles among the three transporters, espe-
cially for 3 and 19. While (+)-3 is only marginally
selective for DAT, (À)-3 is quite selective for DAT, with
a selectivity of 7.3-fold between DAT and SERT sites
and 33.7-fold between DAT and NET sites. Analogue
(+)-19 isomer is in fact more selective for SERT, but
(À)-19 is fairly selective for DAT. Both the (+)- and
(À)-isomers of 4 and 20 are selective for DAT, suggest-
ing the para-Cl substitutent in each phenyl ring in these
Table 4. Activities of chiral isomers of several potent inhibitors at the three transporter sites
Compound
Ki (nM)
Selectivity
Binding
Uptake
SER
DA
NE
DA
([3H]-Mazindol)
([3H]-DA)
([3H]-SER)
([3H]-NE)
(+)-3
(À)-3
1.93Æ0.20
0.27Æ0.01
1.64Æ0.22
0.22Æ0.02
0.62Æ0.03
1.58Æ0.02
1.12Æ0.10
0.90Æ0.09
0.37Æ0.01
0.497Æ0.05
0.97Æ0.09
0.93Æ0.05
5.17Æ0.44
7.27Æ1.32
0.38Æ0.04
0.18Æ0.01
3.55Æ0.07
0.79Æ0.08
0.55Æ0.04
0.52Æ0.07
2.5
7.3
21.9
24.2
0.5
6.9
8.8
21
2.9
33.7
7.4
4.9
4.6
10.8
4.9
11
(+)-4
(À)-4
0.020Æ0.002
0.013Æ0.002
1.57Æ2.00
0.051Æ0.007
0.037Æ0.002
0.78Æ0.13
(+)-19
(À)-19
(+)-20
(À)-20
0.060Æ0.003
0.082Æ0.014
0.022Æ0.005
0.072Æ0.005
0.11Æ0.01
0.045Æ0.008

1758
S. Wang et al. / Bioorg. Med. Chem. 9 (2001) 1753–1764
superposition of 3 and 16 in their lowest energy con-
formations is shown in Figure 1B. In order to adopt the
same orientation as that in 3, 12 and 17, analogue 16
would have to pay an energy penalty as much as 9 kcal/
mol. Therefore, although it is not absolutely certain that
the lowest energy conformations of 3, 12, and 17 must
be their active conformation in binding to the DAT, our
results suggest that conformational preference of these
compounds plays an important role for their activity in
inhibition of the reuptake of DA.
size of these seven compounds. The optimal value for
Vm is 322.4 A3, approximately equal to the volume of
compound 13 with a bromo substituent.
1
2
logð Þ ¼ 6:79 À 0:0013 Â ðVm À 322:4Þ
ð1Þ
Ki
N ¼ 7; r2 ¼ 0:76; F ¼ 15:5; LOF ¼ 0:42
Our SAR studies showed that among the analogues
with a mono substituent at the para position in both
phenyl rings, the nature of the substitutent has a sig-
nificant effect on the activity. For example, while 13 with
a p-bromo substituent has a Ki value of 0.085 mM in
inhibition of DA reuptake, 15 with a p-ethyl substituent
has a Ki value of 8.09 mM, 95-fold less potent than 13.
Conformational analysis of these compounds showed
that they have virtually identical conformational pro-
files, suggesting that their difference in activity is not
due to their conformational difference. Examination of
these substituents showed that they differ in size,
hydrophobicity and electronic properties. To investigate
if their size and hydrophobicity play a role in determin-
ing their activity in inhibition of the reuptake of DA, we
calculated the size (volume, Vm) and hydrophobicity
(partition coefficient between n-octanol and water, logP)
value for these seven mono-substituted analogues and
the data are provided in Table 5.
The correlation between AlogP (the calculated logP
value) and the Ki for reuptake of DA is, however, only
marginal [r2=0.25; eq (2)].
1
2
logð Þ ¼ 2:49 þ 1:66AlogP À 0:17ðAlogPÞ
ð2Þ
Ki
N ¼ 7; r2 ¼ 0:25; F ¼ 0:65; LOF ¼ 2:32
When using these two parameters, the best equation [eq
(3)] obtained has a marginal improvement as compared
to eq (1) using the volume alone.
1
2
logð Þ ¼ 4:67 À 0:0011 Â ðVm À 312:7Þ þ 0:57AlogP
Ki
Using the genetic function approximation algorithm42
implemented in the Cerius2 program,41 several quanti-
tative models were obtained for these seven analogues
with a mono substituent at the para position in both
phenyl rings. A strong parabolic correlation [eq (1)] was
found between the Ki value for reuptake of DA and the
N ¼ 7; r2 ¼ 0:89; F ¼ 15:6; LOF ¼ 0:79
ð3Þ
Furthermore, we have investigated if electronic terms
such as the sum of atomic polarizabilities and the dipole
moment may play a role. We found that inclusion of
these two terms in the QSAR analysis does not improve
the quality of the QSAR model (data not shown). Thus,
based upon our analyses, the size of the substituents
plays an important role in determining their potency in
inhibition of the reuptake of DA for analogues with a
mono substituent at the para position in both phenyl
rings, while the hydrophobicity (logP) play a less sig-
nificant role. Taken together, the results obtained from
current and previous studies30 indicated that for this
class of compounds, the pharmacophore, the con-
formational preference and the substituent size in the
phenyl rings play important roles in determining their
affinity for the DA reuptake site.
Figure 1. Superposition of the lowest energy conformations: (A) lead
compound 3 (yellow), on analogue 12 (brown); (B) lead compound 3
(yellow) on analogue 16 (brown).
Table 5. Calculated hydrophobicity (AlogP) and volume of para,
mono-substituted analogues used in the QSAR analysis
Functional Antagonism
Analogue
Substituent
–log Ki
Vm
ALogP
In our previous study, the lead compound 3 was shown
to display a significant functional antagonism but 4, a
more potent analogue, failed to show significant func-
tional antagonism. To search for more potent analogues
than 3 that possess significant functional antagonism, a
number of more potent new analogues synthesized in
the present study were evaluated for their cocaine
5
3
6
12
13
14
15
H
CH3
F
Cl
Br
I
CH3CH2
5.59
6.44
5.38
6.80
7.07
6.91
5.09
286.0
319.1
295.3
312.7
322.4
334.0
353.5
2.42
3.35
2.70
3.46
4.00
4.93
4.15

S. Wang et al. / Bioorg. Med. Chem. 9 (2001) 1753–1764
1759
antagonism using an in vitro functional assay.30 For the
purpose of comparison, the functional antagonism of
cocaine against cocaine was also determined using
unlabeled cocaine (cold cocaine). The results are sum-
marized in Table 6. Among the four new analogues
evaluated, analogues 12 and 20 showed significant
functional antagonism and 20 appeared to have an
improved functional antagonism as compared to 3. For
example, the experimental IC50 of cocaine in inhibition
of DA reuptake in the presence of 30 nM of 20 is
775 nM and this value is 213% of the calculated theo-
retical value of 363 nM using the same binding site
model. The experimental IC50 of cocaine in inhibition of
DA reuptake in the presence of 100 nM of 20 was
945 nM and this value is 176% of the calculated theo-
retical value of 538 nM using the same binding site
model. Very interestingly, cocaine itself has a very
moderate functional antagonism in this assay. Using
50 nM of cold cocaine, the experimental IC50 value of
the labeled cocaine is 415 nM, which is 122% of the
theoretical IC50 value calculated using the same site
binding model. Similarly, using 100 nM of cold cocaine,
the experimental IC50 value of the labeled cocaine is
511 nM, which is 131% of the calculated theoretical
IC50 value using the same site binding model. The
moderate antagonism of cold cocaine may be attributed
to the fact that cocaine binds to two different binding
sites at the DAT. Thus, as compared to 3 and cocaine
itself, analogues 20, a more potent DAT inhibitor than
3 and cocaine, displays a significant in vitro functional
antagonism.
ism. As can be seen from Table 7, both (À)-20 and (+)-
20 have similar functional antagonism. It should be
noted that no direct relationship between the potency in
inhibition of DA reuptake of a compound and its func-
tional antagonism was observed and the precise under-
lying molecular mechanism for the observed functional
antagonism is not entirely clear.
Summary
To further explore the SARs of 4-hydroxy-1-methyl-4-
(4-methylphenyl)-3-piperidyl 4-methylphenyl ketones as
a novel class of DAT inhibitors and to search for more
potent analogues than the lead compound that possess
significant functional antagonism, 16 new analogues
were synthesized in racemic form and evaluated as DAT
inhibitors. It was found that seven new analogues
are reasonably potent DAT inhibitors with Ki values of
0.052 to 0.16 mM in inhibition of DA reuptake. Mole-
cular modeling studies showed that the pharmacophore,
the conformational preference and the substituent size
in the phenyl rings play important roles in determining
their affinity for the DA reuptake site.
Chiral isomers of several potent DAT inhibitors were
obtained through chiral HPLC separation and eval-
uated as inhibitors at the three monoamine transporter
sites. In general, the (À)-isomer is more active than the
(+)-isomer in inhibition of DA reuptake and all the
(À)-isomers of potent analogues are selective inhibitors
at the DAT site. Evaluation of cocaine’s effect on
dopamine uptake in the presence and absence of (+)-3
and (À)-3 showed that (À)-3 is responsible for the
functional antagonism obtained with the original lead
(Æ)-3, although both isomers display significant func-
tional antagonism. Out of the new compounds synthe-
sized, analogue (Æ)-20, which is 8- and 3-fold more
potent than (Æ)-3 in binding and inhibition of DA
reuptake, appeared to have improved functional
antagonism as compared to (Æ)-3. Both stereoisomers
of (Æ)-20 display significant functional antagonism.
Therefore, analogue 20 represent a potential cocaine
antagonist for further evaluations. Further pharmaco-
logical and behavioral studies are currently under way
to investigate the mechanism of its functional antagon-
ism and its therapeutic potential for the treatment of
cocaine abuse.
The functional antagonism of 3 and 20 was obtained
with their racemates and it was not clear which chiral
isomer is responsible for the observed functional
antagonism. Thus, we have evaluated the functional
antagonism of the two stereoisomers of 3 and 20 in the
functional assay and the results are summarized in
Table 7. At concentrations of 50 and 200 nM of (Æ)-3,
(+)-3 and (À)-3 isomers presumably have 25 and
100 nM each in the mixture. Based upon their activity in
inhibition of DA reuptake (Table 4), at these con-
centrations, only (À)-3 has a significant activity in
binding to DAT, while (+)-3 has a minimal activity.
Hence, it is predicted that the functional antagonism of
(Æ)-3 may be primarily due to the (À)-3 isomer. Since
(À)-3 is approximately 2 times more potent than (Æ)-3
in inhibition of DA uptake, we tested the functional
antagonism of (À)-3 at 30 and 100 nM in order to com-
pare directly with the functional antagonism of (Æ)-3
obtained at 50 and 200 nM. As can be seen from Table
7, (À)-3 and (Æ)-3 have similar functional antagonism
at concentrations that produce similar inhibition of DA
reuptake, indicating that the functional antagonism
observed for (Æ)-3 at concentrations of 50 and 200 nM
is primarily due to the functional antagonism of the (À)-
3 isomer. However, (+)-3 also has significant functional
antagonism, although this isomer is a much weaker
DAT inhibitor. Although the chiral isomer (À)-20 is
twice as potent as (+)-20 in inhibition of DA reuptake,
both isomers are reasonably potent DAT inhibitors
with Ki values of 45 and 111 nM, respectively. These
two isomers were evaluated for the functional antagon-
Experimental
Molecular modeling
Conformational analysis was performed using the con-
formational analysis module in the QUANTA pro-
gram.38 Generally, if a compound has fewer than five
rotatable single bonds, the grid scan conformational
search protocol was employed. In this protocol, each
rotatable bond was systematically rotated to generate a
starting conformation, which was subsequently mini-
mized using the CHARMm program39 within
QUANTA. If a compound has more than five rotatable

1760
S. Wang et al. / Bioorg. Med. Chem. 9 (2001) 1753–1764
Table 6. Functional antagonism of racemate compounds against cocaine
Drugs
IC50 (nM) [3H]-Dopamine uptake
(Experimental)
IC50 (nM) [3H]-Dopamine uptake
(Theoretical)
Cocaine alone
297Æ22a
415Æ30
511Æ28
470Æ25
717Æ49
382Æ13
647Æ75
506Æ8
Cocaine+cold cocaine (50 nM)
Cocaine+cold cocaine (100 nM)
Cocaine+(Æ)-3 (50 nM)
Cocaine+(Æ)-3 (200 nM)
Cocaine+(Æ)-4 (10 nM)
Cocaine+(Æ)-4 (50 nM)
Cocaine+(Æ)-12 (50 nM)
Cocaine+(Æ)-12 (100 nM)
Cocaine+(Æ)-19 (30 nM)
Cocaine+(Æ)-19 (75 nM)
Cocaine+(Æ)-20 (30 nM)
Cocaine+(Æ)-20 (100 nM)
Cocaine+(Æ)-23 (10 nM)
Cocaine+(Æ)-23 (30 nM)
339Æ11
390Æ14
331Æ11
438Æ22
346Æ17
567Æ32
379Æ19
466Æ24
373Æ20
483Æ23
363Æ18
538Æ29
344Æ18
451Æ23
616Æ38
463Æ11
595Æ45
775Æ17
945Æ97
453Æ20
555Æ38
aSE was based upon 2–3 experiments.
Table 7. Functional antagonism of chiral isomers against cocaine
Drugs
IC50 (nM) [3H]-Dopamine uptake
(Experimental)
IC50 (nM) [3H]-Dopamine uptake
(Theoretical)
Cocaine alone
297Æ22a
415Æ30
511Æ28
482Æ8
Cocaine+cold cocaine (50 nM)
Cocaine+cold cocaine (100 nM)
Cocaine+(À)-3 (30 nM)
Cocaine+(À)-3 (100 nM)
Cocaine+(+)-3 (300 nM)
Cocaine+(+)-3 (1000 nM)
Cocaine+(À)-20 (10 nM)
Cocaine+(À)-20 (30 nM)
Cocaine+(+)-20 (10 nM)
Cocaine+(+)-20 (30 nM)
339Æ11
390Æ14
329Æ16
430Æ22
343Æ17
461Æ24
345Æ18
479Æ27
328Æ13
365Æ18
663Æ20
464Æ2
701Æ58
483Æ12
647Æ11
451Æ20
564Æ32
aSE was based upon 2–3 experiments.
bonds, a random sampling protocol was used to gen-
erate conformations. Up to 5000 conformations were
generated and minimized. Energy minimization of each
conformation was computed with 5000 iterations or
until convergence, defined as an energy gradient of
0.001 kcal molÀ1 AÀ1 or less. An adopted basis Newton–
Raphson algorithm, implemented in the CHARMm
program, was used for energy minimization. A constant
dielectric constant (equal to 1) was used throughout all
the calculations. Upon the completion of conformation
generation and energy minimization, the most stable
conformation was identified (the global minimum in
vacuum). It is noted, however, that the lowest energy
conformation may not be the bioactive conformation,
as was shown previously.40 For this reason, other low
energy conformations, typically within 5 kcal/mol of the
global minimum were identified. Cluster analysis was
performed to determine the number of truly unique
conformations (clusters), using the cluster analysis
module available in the QUANTA program. These low
energy conformational clusters together are likely to
include the bioactive conformations for a compound.
mented in the Cerius2 molecular modeling package. The
molecular volume was calculated using the QSAR
module as implemented in the Cerius2 molecular mod-
eling package.
Chemistry
General methods. THF was freshly distilled under
nitrogen from sodium benzophenone. 1H and 13C NMR
spectra were obtained with a Varian Unity Inova
1
instrument at 300 and 75.46 MHz, respectively. H che-
mical shifts (d) are reported in ppm downfield from
internal TMS. 13C chemical shifts are referenced to
CDCl3 (central peak, d=77.0 ppm). Melting points were
determined in Pyrex capillaries with a Thomas–Hoover
Unimelt apparatus and are uncorrected. Mass spectra
were measured in the EI mode at an ionization potential
of 70 eV. TLC was performed on Merck silica gel
60F254 glass plates; column chromatography was per-
formed using Merck silica gel (60–200 mesh). The fol-
lowing abbreviations are used: THF=tetrahydrofuran;
DCM=dichloromethane; ether=diethyl ether.
QSARs were derived using the genetic approximation
algorithm42 as implemented in the QSAR module of the
Cerius2 molecular modeling package.41 The partition
coefficient between n-octanol and water (logP) values
were calculated using the ALOGP method,43 as imple-
4-Chloro-3-methylacetophenone (29). A solution of 4-
(28)
23.6 mmol), sodium iodide (14.2 g, 94.6 mmol) and
SnCl2 (14.4 g, 75.7 mmol) in a 5:1 mixture of THF and
H2O was refluxed for 2 h and cooled to room tempera-
chloro-3-methyl
phenacylchloride
(4.8 g,

S. Wang et al. / Bioorg. Med. Chem. 9 (2001) 1753–1764
1761
ture. The organic layer was separated, and the aqueous
layer was extracted with ether. The combined organic
layers were dried (Na2SO4), concentrated, and purified
by passing through a small bed of silica gel using ether
as eluent to afford the title compound as a colorless
J=7.5 Hz), 1.83 (1H, br d, J=13.9 Hz), 2.02–2.11 (1H,
m), 2.41 (3H, s), 2.56 (2H, q, J=15.1, 7.6 Hz), 2.63–2.81
(5H, m), 2.94 (1H, dd, J=10.8, 2.5 Hz), 4.40 (1H, dd,
J=11.5, 3.7 Hz), 5.23 (1H, d, J=2.4 Hz), 7.08 (2H, d
J=8.0 Hz), 7.26 (2H, d, J=8.3 Hz), 7.41 (2H, d,
J=8.3 Hz), 7.85 ( 2H, d, J=8.1 Hz); 13C NMR (CDCl3)
d 14.9, 15.1, 28.5, 40.0, 45.8, 50.2, 51.4, 54.6, 72.3, 124.4,
127.6, 128.2, 128.5, 133.5, 142.3, 144.4, 150.9, 203.7.
Anal. (C23H29NO2) C, H, N.
1
liquid (3.55 g, 90%). H NMR (CDCl3) d 2.42 (3H, s),
2.60 (3H, s), 7.42 (1H, d, J=8.3 Hz), 7.72 (1H, dd,
J=8.3, 1.7 Hz), 7.83 (1H, s); 13C NMR (CDCl3) d 20.0,
26.5, 127.0, 129.1, 130.6, 135.4, 136.4, 139.6, 197.1.
Anal. (C9H9ClO) C, H, N.
2-Chlorophenyl 4-(2-chlorophenyl)-4-hydroxy-1-methyl-
3-piperidyl ketone (16). White solid; yield 71%; mp 83–
General procedure for the synthesis of compounds 12–27.
To an equimolar mixture of aryl methyl ketone and
paraformaldehyde in acetonitrile (15 mL/g of ketone),
was added amine hydrochloride (0.25 equiv), and the
mixture was refluxed for 20 h in the presence of a cata-
lytic amount of hydrochloric acid (37% w/v, 0.02 mL/
1 g of ketone). The reaction mixture was cooled to room
temperature, and volatiles were removed under reduced
pressure. The resulting mass was dissolved in DCM,
washed with aq NaHCO3 solution, water, and brine,
dried (Na2SO4), and concentrated under reduced pres-
sure. The crude compounds were purified by column
chromatography using triethylamine/diethyl ether as
eluent to afford the following compounds:
ꢀ
1
85 C. H NMR (CDCl3) d 1.58 (1H, br d, J=13.7 Hz),
2.42 (3H, s), 2.66–2.82 (3H, m), 2.87–2.98 (1H, m), 3.04
(1H, dd, J=11.0, 3.4 Hz), 4.60 (1H, d, J=2.7 Hz), 5.22
(1H, dd, J=11.5, 3.9 Hz), 6.94 (1H, d, J=7.1 Hz), 7.07–
7.12 (3H, m), 7.25–7.36 (3H, m), 7.92 (1H, d,
J=8.1 Hz); 13C NMR (CDCl3) 33.9, 46.1, 50.9, 51.4,
53.2, 73.1, 126.5, 127.0, 128.6, 128.8, 128.9, 130.3, 131.3,
131.5, 132.1, 138.2, 142.1, 207.2. Anal. (C19H19Cl2NO2)
C, H, N.
3-Chlorophenyl 4-(3-chlorophenyl)-4-hydroxy-1-methyl-
3-piperidyl ketone (17). Light-yellow solid; yield 68%;
mp 115–117 ꢀC. H NMR (CDCl3) d 1.80 (1H, br d,
1
J=11.5 Hz), 1.98–2.09 (1H, m), 2.41 (3H, s), 2.63–2.72
(2H, m), 2.79 (1H, broad d, J=7.0 Hz), 2.92 (1H, dd,
J=11.2, 2.9 Hz), 4.30 (1H, dd, J=11.5, 3.7 Hz), 5.01
(1H, d, J=2.4 Hz), 7.11–7.20 (2H, m), 7.30 (1H, d,
J=7.0 Hz), 7.39 (1H, t, J=8.1 Hz), 7.53 (2H, d,
J=4.8 Hz), 7.70 (1H, d, J=7.8 Hz), 7.84 (1H, s); 13C
NMR (CDCl3) d 39.7, 45.8, 50.8, 51.1, 54.3, 72.4, 122.4,
125.2, 126.3, 127.0, 128.3, 129.5, 130.1, 133.9, 134.4,
135.2, 137.1, 149.1, 202.6. Anal. (C19H19Cl2NO2) C, H,
N.
4-Chlorophenyl 4-(4-chlorophenyl)-4-hydroxy-1-methyl-
3-piperidyl ketone (12). White solid; yield 88%; mp
168–170 ꢀC. 1H NMR (CDCl3) d 1.82 (1H, br t,
J=13.9 Hz), 1.94–2.07 (1H, m), 2.41 (3H, s), 2.64–2.73
(2H, m), 2.80 (1H, br d, J=7.1 Hz), 2.90 (1H, dd,
J=11.2, 2.7 Hz), 4.30 (1H, dd, J=11.5, 3.7 Hz), 5.10
(1H, d, J=2.7 Hz), 7.22 (2H, d, J=8.6 Hz), 7.42 (4H,
dd, J=8.8, 2.9 Hz), 7.82 (2H, d, J=8.5 Hz); 13C NMR
(CDCl3) d 39.8, 45.2, 50.5, 51.2, 54.5, 72.3, 125.9, 128.4,
129.2, 129.6, 132.6, 133.9, 140.8, 145.5, 202.7. Anal.
(C19H19Cl2NO2) C, H, N.
3,4-Difluorophenyl 4-(3,4-difluorophenyl)-4-hydroxy-1-
methyl-3-piperidyl ketone (18). White solid; yield 59%;
ꢀ
1
4-Bromophenyl 4-(4-bromophenyl)-4-hydroxy-1-methyl-
3-piperidyl hetone (13). Yellow solid; yield 74%; mp
181 ꢀC. 1H NMR (CDCl3) d 1.80 (1H, br d, J=13.9 Hz),
1.96–2.08 (1H, m), 2.40 (3H, s), 2.66 (2H, t, J=11.2 Hz),
2.78 (1H, br d, J=7.5 Hz), 2.90 (1H, dd, J=11.3,
3.2 Hz), 4.28 (1H, dd, J=11.3, 3.4 Hz), 5.08 (1H, d,
J=2.2 Hz), 7.29–7.40 (4H, m), 7.58 (2H, d, J=8.3 Hz),
7.75 (2H, d, J=8.3 Hz); 13C NMR (CDCl3) d 39.8, 45.9,
50.4, 51.1, 54.4, 72.3, 120.8, 126.3, 129.6, 129.7, 131.4,
132.2, 134.3, 146.0, 202.9. Anal. (C19H19Br2NO2) C, H, N.
mp 110–112 C. H NMR (CDCl3) d 1.80 (1H, br d,
J=13.9 Hz), 1.95–2.06 (1H, m), 2.41 (3H, s), 2.66 (2H,
t, J=11.2 Hz), 2.80 (1H, br dd, J=8.8, 1.8 Hz), 2.92
(1H, dd, J=11.2, 3.1 Hz), 4.23 (1H, dd, J=11.2,
3.4 Hz), 5.03 (1H, d, J=2.5 Hz), 6.99–7.15 (2H, m),
7.21–7.37 (2H, m), 7.68–7.74 (2H, m); 13C NMR
(CDCl3) d 39.7, 45.8, 50.5, 51.1, 54.4, 72.1, 114.1, 114.3,
116.9, 117.2, 117.5, 117.7, 118.0, 120.1, 120.2, 120.3,
125.3, 125.4, 125.5, 125.6, 132.6, 144.2, 147.5, 148.5,
150.6, 152.1, 152.3, 155.9, 201.2. Anal. (C19H17F4NO2)
C, H, N.
4-Hydroxy-4-(4-iodophenyl)-1-methyl-3-piperidyl 4-iodo-
phenyl ketone (14). Light-yellow solid; yield 66%; mp
176 ꢀC. 1H NMR (CDCl3) d 1.80 (1H, br d, J=14.0 Hz),
1.95–2.03 (1H, m), 2.40 (3H, s), 2.62–2.71 (2H, m), 2.78
(1H, br d, J=7.5 Hz), 2.90 (1H, dd, J=11.0, 3.2 Hz),
4.26 (1H, dd, J=11.5, 3.4 Hz), 5.06 (1H, d, J=2.4 Hz),
7.22 (2H, d, J=8.3 Hz), 7.58 (4H, d, J=8.3 Hz), 7.82
(2H, d, J=8.3 Hz); 13C NMR (CDCl3) d 39.8, 45.9,
50.3, 51.1, 54.4, 72.4, 92.5, 102.6, 126.6, 129.5, 134.8,
137.3, 138.2, 146.8, 203.2. Anal. (C19H19I2NO2) C, H, N.
3,4-Dimethylphenyl 4-(3,4-dimethylphenyl)-4-hydroxy-1-
methyl-3-piperidyl ketone (19). Light-yellow solid; yield
76%; mp 90–92 ꢀC. 1H NMR (CDCl3) d 2.18 (1H, br d,
J=13.9 Hz), 1.99–2.10 (1H, m), 2.16 (3H, s), 2.20 (3H,
s), 2.30 (3H, s), 2.31 (3H, s), 2.41 (3H, s), 2.63–2.80 (3H,
m), 2.92 (1H, dd, J=11.0, 2.8 Hz), 4.38 (1H, dd,
J=11.4, 3.6 Hz), 5.25 (1H, d, J=2.7 Hz), 6.99 (1H, d,
J=7.8 Hz), 7.19 (2H, t, J=7.1 Hz), 7.33 (1H, br s),
7.68 (2H, d, J=7.4 Hz); 13C NMR (CDCl3) d 19.2,
19.7, 19.9, 20.0, 40.3, 45.9, 50.1, 51.5, 54.7, 72.3, 121.6,
126.1, 126.2, 129.4, 129.5, 129.9, 133.7, 134.8, 136.2,
137.1, 143.7, 144.7, 204.0. Anal. (C23H29NOꢁ HCl) C, H,
N.
4-Ethylphenyl 4-(4-ethylphenyl)-4-hydroxy-1-methyl-3-
piperidyl ketone (15). Colorless thick syrup; yield 78%.
1H NMR (CDCl3) d 1.16 (3H, t, J=7.6 Hz), 1.25 (3H, t,

1762
S. Wang et al. / Bioorg. Med. Chem. 9 (2001) 1753–1764
4-Chloro-3-methylphenyl 4-(4-chloro-3-methylphenyl)-4-
hydroxy-1-methyl-3-piperidyl ketone (20). White solid;
4-Bromophenyl 4-(4-bromophenyl)-4-hydroxy-1-(2-phe-
nylethyl)-3-piperidyl ketone (25). Yellow thick syrup;
yield 74%. 1H NMR (CDCl3) d 1.82 (1H, br d,
J=13.9 Hz), 1.97–2.08 (1H, m), 2.74–3.04 (8H, m), 4.28
(1H, dd, J=11.2, 3.4 Hz), 5.08 (1H, d, J=2.4 Hz), 7.21–
7.38 (9H, m), 7.60 (2H, d, J=8.6 Hz), 7.74 (2H, d,
J=8.5 Hz); 13C NMR (CDCl3) d 33.6, 39.7, 48.9, 50.4,
52.5, 60.0, 72.9, 120.8, 126.1, 126.4, 128.4, 128.6, 129.6,
129.7, 131.4, 132.2, 134.3, 139.9, 146.1, 203.0. Anal.
(C26H25Br2NO2) C, H, N.
ꢀ
1
yield 82%; mp 114–116 C. H NMR (CDCl3) d 1.80
(1H, d, J=13.9 Hz), 1.99–2.08 (1H, m), 2.30 (3H, s),
2.40 (3H, s), 2.42 (3H, s), 2.64–2.81 (3H, m), 2.92 (1H,
dd, J=11.9 Hz, 3.4 Hz), 4.32 (1H, dd, J=11.5, 3.7 Hz),
5.10 (1H, d, J=2.4 Hz), 7.18 (2H, s), 7.40 (2H, t,
J=5.1 Hz), 7.66 (1H, dd, J=8.3 Hz, 2.0 Hz), 7.75 (1H,
s); 13C NMR (CDCl3) d 20.1, 20.2, 39.8, 45.8, 50.3, 51.2,
54.4, 72.2, 123.1, 126.9, 127.4, 128.8, 129.5, 130.6, 132.7,
134.0, 135.7, 136.9, 140.9, 145.5, 203.0. Anal.
ꢁ
(C21H23Cl2NO2 HCl) C, H, N.
3,4-Dichlorophenyl 4-(3,4-dichlorophenyl)-4-hydroxy-1-
(3-phenylpropyl)-3-piperidyl ketone (26). Colorless thick
1
1-Ethyl-4-hydroxy-4-(4-methylphenyl)-3-piperidyl 4-me-
thylphenyl ketone (21). White solid; yield 60%; mp
syrup; yield 62%. H NMR (CDCl3) d 1.77–2.02 (4H,
m), 2.52 (2H, t, J=7.8 Hz), 2.61–2.71 (4H, m), 2.86–
2.96 (2H, m), 4.18 (1H, dd, J=11.5, 3.7 Hz), 4.96 (1H,
d, J=2.4 Hz), 7.20–7.35 (7H, m), 7.56 (1H, d,
J=8.3 Hz), 7.63 (1H, d, J=1.9 Hz), 7.71 (1H, dd,
J=8.6, 2.0 Hz), 7.95 (1H, d, J=1.7 Hz); 13C NMR
(CDCl3) d 28.6, 33.7, 39.6, 48.7, 50.6, 52.5, 57.6, 72.7,
123.7, 125.8, 127.1, 128.3, 130.2, 130.3, 131.0, 131.1,
132.6, 133.8, 135.0, 139.1, 141.8, 147.4, 201.6. Anal.
(C27H25Cl4NO2) C, H, N.
115–117 ꢀC. 1H NMR (CDCl3)
d 1.15 (3H, t,
J=7.3 Hz), 1.84 (1H, br d, J=13.9 Hz), 1.99–2.10 (1H,
m), 2.57 (3H, s), 2.41 (3H, s), 2.53–2.71 (4H, m), 2.90
(1H, broad d, J=11.3 Hz), 3.00 (1H, dd, J=11.2,
2.6 Hz), 4.36 (1H, dd, J=11.5, 3.5 Hz), 5.24 (1H, d,
J=2.5 Hz), 7.06 (2H, d, J=8.1 Hz), 7.24 (2H, d,
J=8.1 Hz), 7.40 (2H, d, J=8.1 Hz), 7.82 (2H, d,
J=8.0 Hz); 13C NMR (CDCl3) d 12.0, 20.7, 21.5, 40.0,
48.6, 50.1, 52.0, 52.5, 72.9, 124.3, 128.4, 128.8, 129.4,
133.3, 136.0, 144.3, 144.8, 203.9. Anal. (C22H27NO2) C,
H, N.
4-Bromophenyl 4-(4-bromophenyl)-4-hydroxy-1-(3-phe-
nylpropyl)-3-piperidyl ketone (27). Yellow thick syrup;
yield 69%. 1H NMR (CDCl3) d 1.77–2.03 (4H, m), 2.52
(2H, t, J=8.3 Hz), 2.63–2.71 (4H, m), 2.87 (1H, br d,
J=11.0 Hz), 2.95 (1H, dd, J=11.0, 2.9 Hz), 4.26 (1H,
dd, J=11.2, 3.2 Hz), 5.08 (1H, d, J=2.4 Hz), 7.20–7.41
(9H, m), 7.60 (2H, d, J=8.3 Hz), 7.74 (2H, d,
J=8.6 Hz); 13C NMR (CDCl3) d 28.7, 33.7, 39.7, 48.9,
50.4, 52.6, 57.6, 72.9, 120.8, 125.8, 126.3, 128.3, 128.4,
129.6, 129.7, 131.3, 132.2, 134.3, 141.8, 146.2, 203.0.
Anal. (C27H27Br2NO2) C, H, N.
4-Chlorophenyl 4-(4-chlorophenyl)-1-ethyl-4-hydroxy-3-
piperidyl ketone (22). White solid; yield 91%; mp
151 ꢀC. 1H NMR (CDCl3) d 1.14 (3H, t, J=7.0 Hz),
1.82 (1H, br d, J=13.9 Hz), 1.95–2.06 (1H, m), 2.53–
2.71 (4H, m), 2.90 (1H, dd, J=9.5, 1.7 Hz), 2.98 (1H,
dd, J=11.0, 2.5 Hz), 4.28 (1H, dd, J=11.2, 3.4 Hz), 5.10
(1H, d, J=2.4 Hz), 7.22 (2H, d, J=8.5 Hz), 7.42 (4H,
dd, J=8.5, 1.7 Hz), 7.82 (2H, d, J=8.8 Hz); 13C NMR
(CDCl3) d 12.0, 39.7, 48.4, 50.5, 52.0, 52.3, 72.9, 125.9,
128.4, 129.1, 129.6, 132.6, 133.9, 140.7. 145.6, 202.9.
Anal. (C20H21Cl2NO2) C, H, N.
HPLC separation of enantiomers
Racemic piperidinols were separated into their indivi-
dual enantiomers by using a chirex brand HPLC col-
umn (Chirex 3018, purchased from Phenomenex, Inc.,
CA) which contains (S)-proline covalently bound to g-
aminopropyl silinized silica gel (5 mm particle size) and
derivatized via an urea linkage with (R)-1-(a-naphthyl)-
ethylamine as a chiral stationary phase.44 The chiral
HPLC was performed on a Shimadzu SCL-10A-VP
system at a flow rate of 5.0 mL/min at room tempera-
ture using hexane/DCM/ethanol-TFA (20-1) as a
mobile phase in 87.5:10:2.5 ratio and UV detection at
254 and 280 nm. The sample for injection was prepared
by dissolving racemic piperidinol (5 mg/1 mL) in mobile
phase and the separation was carried out by injecting
250 mL on a 250Â10 mm chiral column. Retention times
and rotations of each isomer were given in Table 8.
3,4-Dichlorophenyl
4-(3,4-dichlorophenyl)-1-ethyl-4-
hydroxyꢀ-3-piperidyl ketone (23). White solid; yield 71%;
mp 140 C. 1H NMR (CDCl3) d 1.11 (3H, t, J=7.3 Hz),
1.78 (1H, br d, J=14.0 Hz), 1.93–2.01 (1H, m), 2.50–
2.67 (4H, m), 2.87 (1H, br d, J=11.2 Hz), 2.95 (1H, dd,
J=11.0, 2.7 Hz), 4.19 (1H, dd, J=11.2, 3.4 Hz), 4.97
(1H, d, J=2.2 Hz), 7.21–7.30 (2H, m), 7.50 (1H, d,
J=8.3 Hz), 7.61 (1H, d, J=1.9 Hz), 7.69 (1H, dd,
J=8.3, 1.9 Hz), 7.92 (1H, d, J=2.0 Hz); 13C NMR
(CDCl3) d 11.9, 39.5, 48.1, 50.5, 51.9, 52.1, 72.6, 123.6,
127.1, 130.1, 130.2, 130.8, 130.9, 132.5, 133.7, 134.9,
138.9, 147.3, 201.4. Anal. (C20H19Cl4NO2) C, H, N.
3,4-Dichlorophenyl 4-(3,4-dichlorophenyl)-4-hydroxy-1-
(2-phenylethyl)-3-piperidyl ketone (24). Light-yellow
thick syrup; yield 67%. H NMR (CDCl3) d 1.83 (1H,
1
br d, J=13.9 Hz), 1.96–2.07 (1H, m), 2.73–3.04 (8H, m),
4.22 (1H, dd, J=11.2, 3.7 Hz), 4.98 (1H, d, J=2.5 Hz),
7.22–7.35 (7H, m), 7.56 (1H, d, J=8.6 Hz), 7.65 (1H, d,
J=2.2 Hz), 7.72 (1H, dd, J=8.3, 2.0 Hz), 7.96 (1H, d,
J=2.2 Hz); 13C NMR (CDCl3) d 33.6, 39.6, 48.7, 50.6,
52.4, 59.9, 72.7, 123.7, 126.1, 127.1, 127.2, 128.4, 128.6,
130.2, 130.3, 131.0, 132.7, 133.8, 135.0, 139.1, 139.8,
Pharmacology
[3H]Mazindol binding. Binding assays were conducted as
previously described.30 Briefly conventional P2 mem-
brane pellets were prepared by differential centrifuga-
tion from rat striatum. The P2 pellet was resuspended in
Krebs–Ringer-HEPES (KRH) buffer consisting of (in
mM): NaCl (125), KCl (4.8), MgSO4 (1.2), CaCl2 (1.3),
ꢁ
147.3, 201.5. Anal. (C26H23Cl4NO2 HCl) C, H, N.

S. Wang et al. / Bioorg. Med. Chem. 9 (2001) 1753–1764
Table 8. Retention times and rotations of each isomer
1763
glass pestle in ice-cold 0.32 M sucrose and centrifuged
for 10 min at 1000g. The supernatant was centrifuged at
17,500g for 20 min. This P2 synaptosomal pellet was
resuspended in 30 volumes of ice-cold modified KRH
buffer. An aliquot of the synaptosomal suspension was
preincubated with the buffer and drug for 30 min at
37 ꢀC, and uptake initiated by the addition of
[3H]dopamine (5 nM, final concn.). After 5 min, uptake
was terminated by adding 5 mL of cold buffer contain-
ing glucosamine as a substitute for NaCl and then
finally by rapid vacuum filtration over GF-C glass fiber
filters, followed by washing with two 5 mL volumes of
ice-cold, sodium-free buffer. Radioactivity retained on
the filters was determined by liquid scintillation spec-
trometry. Specific uptake is defined as that which is
sensitive to inhibition by 30 mM cocaine. It is identical
to that calculated by subtracting the mean of identical
tubes incubated at 0 ꢀC. [3H]SER and [3H]NE uptake
were measured in an entirely analogous fashion using
synaptosomes prepared from rat midbrain and parietal/
occipital cortex, respectively. Also, specific uptake of
[3H]SER and [3H]NE were defined in the presence of
10 uM fluoxetine and 1 uM desipramine, respectively.
Compound
Isomer
Retention
time
(tR) min
Optical
rotation
(ꢀ)
Concentration
(%) and
solvent
3
(+)-isomer
(À)-isomer
(+)-isomer
(À)-isomer
(+)-isomer
(À)-isomer
(+)-isomer
(À)-isomer
13.5
17.2
14.9
18.2
12.1
14.8
18.2
22.4
+54.0
À54.0
+93.3
À93.3
+44.2
À44.2
+62.3
À62.3
0.5, DCM
0.5, DCM
0.4, DCM
0.4, DCM
1.0, DCM
1.0, DCM
0.7, DCM
0.7, DCM
4
19
20
KH2PO4 (1.2), glucose (5.6), nialamide (0.01), and
HEPES (25) (pH 7.4)37 and centrifuged again. Finally,
the pellet was resuspended in 30 volumes of buffer, pel-
leted at 15,000g and frozen at À80 ꢀC until used. The
striatal homogenates were thawed by resuspension in
the buffer described above at 75–125 mg protein/mL and
incubated with [3H]mazindol, with or without compet-
ing drugs, for 60 min in a 4 ꢀC cold room. Nonspecific
binding was determined with 30 mM cocaine. The bound
and free [3H]mazindol were separated by rapid vacuum
filtration over Whatman GF/C filters, using a Brandel
M24R cell harvester, followed by two washes with 5 mL
of cold buffer. Radioactivity on the filters was then
extracted by allowing to sit over night with 5 mL of
scintillant. The vials were vortexed and counted. IC50
values were determined using the computer program
LIGAND.
IC50 values were determined by a computer assisted,
iterative fit to a four-parameter sigmoidal equation
(ALLFIT). These values were then converted to Ki
values according to the Cheng–Prusoff equation assum-
ing classical competitive inhibition. Pre-incubation of
the drug and synaptosomes at 37 ꢀC for 30 min was used
to approximate equilibrium conditions as necessary to
satisfy the requirements of the Cheng–Prusoff equation.
Synaptosomal uptake of [3H]DA, [3H]SER, and [3H]NE.
The effect of candidate compounds in antagonizing
dopamine high-affinity uptake was determined using a
method previously employed.30 For [3H]DA uptake,
dissected rat striata were homogenized with a Teflon-
Functional antagonism. The IC50 value of cocaine in the
presence of approximate IC10–C50 concentrations of
candidate antagonist compounds was then compared to
the IC50 value of cocaine alone. Significant differences in
Elemental analysis
Compound
Molecular formula
Calculated
C
Found
H
N
C
H
N
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
29
C19H19Cl2NO2
C19H19Br2NO2
C19H19I2NO2
C23H29NO2
62.65
50.36
41.71
78.59
62.65
62.65
62.12
71.21
58.82
78.30
63.50
53.75
55.79
57.48
60.35
58.19
64.11
5.26
4.23
3.50
8.32
5.26
5.26
4.66
7.79
5.64
8.06
5.60
4.28
4.32
4.64
4.69
4.88
5.38
3.85
3.09
2.56
3.99
3.85
3.85
3.81
3.61
3.27
4.15
3.70
3.13
2.50
2.58
2.61
2.51
62.51
50.57
41.80
78.44
62.63
62.77
61.99
71.11
58.69
78.21
63.35
53.66
55.61
57.53
60.47
58.07
64.02
5.16
4.35
3.59
8.29
5.34
5.28
4.72
7.71
5.60
8.00
5.44
4.31
4.31
4.71
4.81
4.74
5.30
3.63
3.12
2.66
3.88
3.72
3.88
3.79
3.72
3.15
4.14
3.60
3.07
2.39
2.64
2.69
2.42
C19H19Cl2NO2
C19H19Cl2NO2
C19H17F4NO2
ꢁ
C23H29NO2 HCl
ꢁ
C21H23Cl2NO2 HCl
C22H27NO2
C20H21Cl2NO2
C20H19Cl4NO2
C26H23Cl4NO2 HCl
ꢁ
C26H25Br2NO2
C27H25Cl4NO2
C27H27Br2NO2
C9H9ClO

1764
S. Wang et al. / Bioorg. Med. Chem. 9 (2001) 1753–1764
IC50 values were compared to theoretical IC50 values
expected from models of ‘same site’ antagonism.33,34
IC50 values greater than those expected for ‘same site’
antagonism was taken as evidence of functional
antagonism. Compounds demonstrating antagonism
were tested at lower concentrations to determine their
minimum effective concentration. This test was per-
formed under the preincubation conditions described
above to allow slowly equilibrating compounds to reach
equilibrium. Further, any artifactual differences in Ki
due to differences in temperature, tissue preparation,
and so on, were negated in this assay as binding of
cocaine and the putative antagonists to both the cocaine
binding site and the transporter occurred under iden-
tical conditions. This insures that a right shift in the
cocaine inhibition curve beyond what is expected for
two drugs acting at the same site is a true measure of
functional antagonism.
17. Smith, M. P.; Johnson, K. M.; Zheng, M.; Flippen-Ander-
son, J. L.; Kozikowski, A. P. J. Am. Chem. Soc. 1998, 120, 9072.
18. Lieske, S. F.; Yang, B.; Eldefrawi, M. E.; MacKerell,
A. D., Jr.; Wright, J. J. Med. Chem. 1998, 41, 864.
19. Chen, Z.; Izenwasser, S.; Katz, J. L.; Klein, C. L.; Zhu,
N.; Trudell, M. L. J. Med. Chem. 1996, 39, 4744.
20. Glowa, J. R.; Fantegrossi, W. E.; Lewis, D. B.; Matecka,
D.; Rice, K. C.; Rothman, R. B. J. Med. Chem. 1996, 39, 4689.
21. Matecka, D.; Rothman, R. B.; Radesca, L.; de Costa, B.;
Dersch, C. M.; Partilla, J. S.; Pert, A.; Glowa, J. R.; Wojnicki,
F. H. M.; Rice, K. C. J. Med. Chem. 1996, 39, 4704.
22. Matecka, D.; Lewis, D.; Rothman, R. B.; Dersch, C. M.;
Wojnicki, F. H. E.; Glowa, J. R.; DeVries, A. C.; Pert, A.;
Rice, K. C. J. Med. Chem. 1997, 40, 705.
23. Deutsch, H. M.; Shi, Q.; Gruszecka-Kowalik, E.; Schweri,
M. M. J. Med. Chem. 1996, 39, 1201.
24. Houlihan, W. J.; Boja, J. W.; Parrino, V. A.; Kopajtic,
T. A.; Kuhar, M. J. J. Med. Chem. 1996, 39, 4935.
25. Singh, S.; Basmadjian, G. P.; Avor, K. S.; Pouw, B.; Seale,
T. W. J. Med. Chem. 1997, 40, 2474.
26. Avor, K. S.; Singh, S.; Seale, T. W.; Pouw, B.; Basmad-
jian, G. P. J. Med. Chem. 1998, 41, 2380.
Acknowledgements
27. (a) 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. (b) 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.
28. Sakamuri, S.; Enyedy, I. J.; Kozikowski, A. P.; Zaman,
W. A.; Johnson, K. M.; Wang, S. Bioorg. Med. Chem. Lett.
2001, 11, 495.
This work is supported by a grant (DA-11545) from the
National Institute on Drug Abuse, National Institutes
of Health. We wish to thank Dr. Tivadar Farkas of
Phenomenex, Inc. 2320 W. 205th Street, Torrance, CA
90501, USA for his help on the design of the chiral
HPLC column used in this work.
29. Smith, M. P.; Hoepping, A.; Johnson, K. M.; Trzcinska,
M.; Kozikowski, A. P. Drug Discovery Today 1999, 7, 322.
30. Wang, S.; Sakamuri, S.; Enyedy, I. J.; Kozikowski, A. P.;
Deschaux, O.; Bandyopadhyay, B. C.; Tella, S. R.; Zaman,
W. A.; Johnson, K. M. J. Med. Chem. 2000, 43, 351.
31. 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.
References and Notes
1. Carroll, F. I.; Howell, L. L.; Kuhar, M. J. J. Med. Chem.
1999, 42, 2721.
2. Ritz, M. C.; Lamb, R. J.; Goldberg, S. R.; Kuhar, M. J.
Science 1987, 237, 1219.
3. Carroll, F. I.; Lewin, A. H.; Boja, J. W.; Kuhar, M. J. J.
Med. Chem. 1992, 35, 969.
32. Meltzer, P. C.; Blundell, P.; Madras, B. K. Med. Chem.
Res. 1998, 8, 12.
4. Kuhar, M. J.; Ritz, M. C.; Boja, J. W. Trends Neurosci.
1991, 14, 299.
33. Huang, L.-Y. M.; Ehrenstein, G. J. Gen. Physiol. 1981, 77,
37.
5. Lin, Z.; Wang, W.; Kopajtic, T.; Revay, R.; Uhl, G. Mol.
Pharmacol. 1999, 56, 434.
34. Reith, M. E. A.; Selmeci, F. Naunyn-Schmiedeberg’s Arch.
Pharmacol. 1992, 345, 309.
6. Woolverton, W. L.; Johnson, K. M. TIPS 1992, 13, 193.
7. Woolverton, W. L.; Kleven, M. S. NIDA Research Mono-
graph 1988, 88, 160.
8. Spealman, R. D.; Madras, B. K.; Bergman, J. Pharmacol.
Exp. Ther. 1989, 251, 142.
35. Ono, A. Fujimoto, E.; Ueno, M. Synthesis 1986, 570.
36. Dimmock, J. R.; Vashishtha, S. C.; Wilson Quail, J.;
Pugazhenthi, U.; Zimpel, Z.; Sudom, A. M.; Allen, T. M.;
Kao, G. Y.; Balzarini, J.; De Clercq, E. J. Med. Chem. 1998,
41, 4012.
9. Carroll, F. I.; Lewin, A. H.; Biswas, J. Pharm. News 1994,
1, 11.
37. Slusher, B. S.; Tiffany, C. W.; Olkowski, J. L.; Jackson,
P. F. Drug Alcohol Depend. 1997, 48, 43.
10. Wise, R. A. Brain Res. 1978, 152, 215.
11. Carroll, F. I.; Kotian, P.; Dehghani, A.; Gray, J. L.;
Kuzemko, M. A.; Parham, K. A.; Abraham, P.; Lewin, A. H.;
Boja, J. W.; Kuhar, M. J. J. Med. Chem. 1995, 38, 379.
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.
38. QUANTA, a molecular modeling system, is supplied by
Molecular Simulations Inc., 9685 Scranton Road, San Diego,
CA 92121-3752, USA.
39. Brooks, B. R.; Bruccoleri, R. E.; Olafson, B. D.; States,
D. J.; Swaminathan, S.; Jaun, B.; Karplus, M. J. Comput.
Chem. 1983, 4, 187.
40. Nicklaus, M. C.; Wang, S.; Driscoll, J.; Milne, G. W. A.
Bioorg. Med. Chem. 1995, 3, 411.
41. Cerius2, a molecular modeling system, is supplied by
Molecular Simulations Inc., 9685 Scranton Road, San Diego,
CA 92121-3752, USA.
42. Rogers, D.; Hopfinger, A. J. J. Chem. Inf. Comput. Sci.
1994, 34, 854.
43. Ghose, A. K.; Pritchett, A.; Crippen, G. M. J. Comput.
Chem. 1988, 9, 180.
44. Cleveland, T. J. Liq. Chromatogr. 1995, 18, 649.
13. Meltzer, P. C.; Liang, A. Y.; Madras, B. K. J. Med. Chem.
1994, 37, 2001.
14. Newman, A. H.; Kline, R. H.; Allen, A. C.; Izenwasser,
S.; George, C.; Katz, J. L. J. Med. Chem. 1995, 38, 3933.
15. Kozikowski, A. P.; Roberti, M.; Xiang, L.; Bergmann,
J. S.; Callahan, P. M.; Cunningham, K. A.; Johnson, K. M. J.
Med. Chem. 1992, 35, 4764.
16. Kozikowski, A. P.; Saiah, M. K. E.; Bergmann, J. S.;
Johnson, K. M. J. Med. Chem. 1994, 37, 3440.