Synthesis and structure-activity relationships of a new model of arylpiperazines. 2. Three-dimensional quantitative structure-activity relationships of hydantoin-phenylpiperazine derivatives with affinity for 5- HT(1A) and α1 receptors. A comparison of CoMFA models

Article abstract of DOI:10.1021/jm960744g

A series of 48 bicyclohydantoin-phenylpiperazines (1-4) with affinity for 5-HT(1A) and α1 receptors was subjected to three-dimensional quantitative structure-affinity relationship analysis using comparative molecular field analysis (CoMFA), in order to get insight into the structural requirements that are responsible for 5-HT(1A)/α1 selectivity. Good models (high cross-validation correlations and predictive power) were obtained for 5-HT(1A) and α1 receptors. The resulting 3D-QSAR models rationalize steric and electrostatic factors which modulate binding to 5-HT(1A) and α1 receptors. A comparison of these models gives an additional understanding for 5-HT(1A)/α1 selectivity: (a) Substitution at the ortho position by a group with negative potential is favorable to affinity for both receptors. (b) The meta position seems to be implicated in 5-HT(1A)/α1 selectivity. While the 5-HT(1A) receptor is able to accommodate bulky substituents in the region of its active site, the steric requirements of the α1 receptor are more restricted (optimum volume of substituent 11-25 A?³). (c) For both receptors the para position represents a region where the volume accessible by the ligands is limited. (d) The hydantoin moiety and the side chain length seem to modulate not only the affinity but also 5-HT(1A)/α1 selectivity. The 3D- QSAR models reveal an useful predictive information for the design of new selective ligands.

Full text of DOI:10.1021/jm960744g

1648
J . Med. Chem. 1997, 40, 1648-1656
Syn th esis a n d Str u ctu r e-Activity Rela tion sh ip s of a New Mod el of
Ar ylp ip er a zin es. 2.1 Th r ee-Dim en sion a l Qu a n tita tive Str u ctu r e-Activity
Rela tion sh ip s of Hyd a n toin -P h en ylp ip er a zin e Der iva tives w ith Affin ity for
5-HT1A a n d r1 Recep tor s. A Com p a r ison of CoMF A Mod els2
Mar´ıa L. Lo´pez-Rodr´ıguez,*,† Ma Luisa Rosado,† Bellinda Benhamu´,† Ma J ose´ Morcillo,‡ Esther Ferna´ndez,† and
Klaus-J u¨rgen Schaper§
Departamento de Qu´ımica Orga´nica I, Facultad de Ciencias Quı´micas, Universidad Complutense, Ciudad Universitaria s/ n,
28040 Madrid, Spain, Facultad de Ciencias, Universidad Nacional de Educacio´n a Distancia, 28040 Madrid, Spain, and
Borstel Research Center, D-23845 Borstel, Germany
Received October 25, 1996X
A series of 48 bicyclohydantoin-phenylpiperazines (1-4) with affinity for 5-HT1A and R1
receptors was subjected to three-dimensional quantitative structure-affinity relationship
analysis using comparative molecular field analysis (CoMFA), in order to get insight into the
structural requirements that are responsible for 5-HT1A/R1 selectivity. Good models (high cross-
validation correlations and predictive power) were obtained for 5-HT1A and R1 receptors. The
resulting 3D-QSAR models rationalize steric and electrostatic factors which modulate binding
to 5-HT1A and R1 receptors. A comparison of these models gives an additional understanding
for 5-HT1A/R1 selectivity: (a) Substitution at the ortho position by a group with negative potential
is favorable to affinity for both receptors. (b) The meta position seems to be implicated in
5-HT1A/R1 selectivity. While the 5-HT1A receptor is able to accommodate bulky substituents in
the region of its active site, the steric requirements of the R1 receptor are more restricted
(optimum volume of substituent 11-25 Å3). (c) For both receptors the para position represents
a region where the volume accessible by the ligands is limited. (d) The hydantoin moiety and
the side chain length seem to modulate not only the affinity but also 5-HT1A/R1 selectivity.
The 3D-QSAR models reveal an useful predictive information for the design of new selective
ligands.
In tr od u ction
Many endogenous substances such as some hormones
and neurotransmitters mediate their intracellular ef-
fects through signal transduction pathways that involve
G-protein-coupled receptor. The cloning of some of these
receptors has revealed their related structures, whose
most representative common feature is the presence of
seven transmembrane domains constituted by R-helices
recognition forces characterizing 5-HT1A and R1 receptor
of 20-25 hydrophobic amino acids.3-5 The R1 adreno-
sites. Our aim is to get insight into the structural
ceptor and the serotonin 5-HT1A receptor are represen-
tative members of this receptor superfamily. In spite
factors that are responsible for 5-HT1A/R1 selectivity, in
order to design new ligands with high selectivity for the
of their different pharmacology, they show some com-
5-HT1A receptor.
mon features in their binding sites. As a consequence
of such similarities some synthetic ligands (e.g. long-
chain arylpiperazines like NAN-190 and related com-
pounds6,7) bind at both receptors. This peculiarity was
also shown by some of the new hydantoin-phenylpi-
perazine derivatives (1-4) we have recently reported.1,8
The affinity of the phenylpiperazine derivatives for both
receptors has been attributed to interactions of their
aromatic moiety and the N4 of the piperazine ring with
the active sites.9-12
Here we report a study that applies CoMFA13,14
methodology to rationalize the relationships between
new hydantoin-phenylpiperazine structures (1-4) and
their binding-affinity data at 5-HT1A and R1 receptors.
This data set is used to derive CoMFA models that
describe the steric and electrostatic requirements for
Ma ter ia ls a n d Meth od s
Biologica l Activities. The compounds were evaluated for
in vitro 5-HT1A and R1 receptor affinity by radioligand binding
assays. All values have been obtained in rat cerebral cortex
membranes with [3H]-8-OH-DPAT and [3H]prazosin as the
specific radioligands. The compounds were first tested at the
fixed dose of 10-6 M, and for those that in this prescreening
process presented high activity (displacement of the radioli-
gand g55%), the dose-response curves were calculated.
However, for the members of the series showing low activity
(displacement <55%) the binding constants were not deter-
mined. Nevertheless, inactive compounds of a drug series
allow the activity scale to be expanded. Obviously a broad
range of activity data facilitates the recognition of QSAR
relationships. The evaluation of the IC50 values for the
compounds where only one point of the curve was available
was performed by the application of a method,15 which consists
in a simultaneous nonlinear regression analysis of all dose-
response curves (DRCs) of the drug series using eq 1.
† Departamento de Qu´ımica Orga´nica I, Universidad Complutense.
‡ Universidad Nacional de Educacio´n a Distancia.
§ Borstel Research Center.
%SU ) 100(1 - Cib/(IC50j + Cib))
(1)
b
X Abstract published in Advance ACS Abstracts, April 1, 1997.
S0022-2623(96)00744-3 CCC: $14.00 © 1997 American Chemical Society

Synthesis of Arylpiperazines
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 11 1649
where SU is the specific union of radioligand; b is the slope; i
) 1...n (measurements); j: 1...m (number of compounds).
This analysis is performed under the assumption that all
derivatives present the same mechanism of action within the
given test model (i.e. parallel DRCs, identical Hill coeficients/
slopes). This approach requires that complete DRCs (g3 data
pairs) are available for some analogs. Missing IC50 values are
obtained from “fragmentary” DRCs by a computational parallel
shift of complete DRCs. IC50 values were converted to pKi
values (Tables 1 and 5) using the Cheng-Prusoff equation.16
Molecu la r Mod elin g a n d Align m en t Ru les. The entire
set of phenylpiperazine analogs was built de novo using the
SKETCH option in SYBYL 6.017 and fully geometry optimized
using the standard TRIPOS molecular mechanics force field,
with a 0.001 kcal/mol energy gradient convergence criterion
and a distance-dependent dielectric constant. The asymmetric
truncated at 15 and 10 kcal/mol, respectively. The grid used
in the CoMFA study had a resolution of 2.0 Å and extended
beyond the molecular dimensions by 4.0 Å in all directions.
The steric and electrostatic fields were subjected to scaling in
order to assign them the same weight (the command “scaling
CoMFA std” was used).
The QSAR table was built with the compounds as rows and
two types of column values: 5-HT1A and R1 pKi values
(dependent variables) and the steric and electrostatic field
potential values (independent variables). The introduction of
log P values did not improve the CoMFA models.
Partial least squares (PLS) methodology19,20 was used to
develop the relationship between the independent variables
and the pKi values. Five orthogonal latent variables were first
extracted by standard PLS algorithm and subsequently sub-
jected to a cross-validation in order to correlate them with the
dependent variable. If the analysis indicated that more latent
variables were required for an optimum description of the
variance in the data set, additional PLS runs were performed
considering a higher number of components. The “best” model
was the one that showed the sum of the squared differences
between predicted and observed values of the dependent
property to be a minimum from a leave-one-out cross-valida-
atom was arbitrarily oriented in the R configuration.
A
systematic conformational search was performed on the rotable
bonds using an increment at 10° in the SYBYL SEARCH
module; MAXIMIN2 steric energies were used to identify low-
energy conformations. While it is recognized that low-energy
conformations may not necessarily be adopted in the drug-
receptor complex, the use of a reasonable low-energy confor-
mation in the alignment is a useful starting point for statistical
comparisons of flexible structures within the SYBYL CoMFA
module.
The most crucial variable in CoMFA is the positioning of
the molecules within a fixed lattice. The best results were
obtained when the common pharmacophore portions of all the
molecules, comprising the phenylpiperazine moieties, were
aligned by a least-squares fit on the following common atoms:
(a) the centroid of the aromatic moiety; (b) the N1 and N4 of
the phenylpiperazine moiety; and (c) the chain carbon atom
adjacent to the N4. This choice of atoms produces a reasonable
overlap of the hydantoin substructures, and at the same time
allows the best superimposition of the phenylpiperazine
moieties.
In order to obtain the best superimposition of the hydantoin
moiety in each subfamily of compounds (n ) 1, 2, 3, and 4),
all members of the subfamily were superimposed using the
unsubstituted derivative (R ) H) with X ) -(CH2)3- as the
template molecule. They were aligned via root mean square
(RMS) fit of (a) the carbonyl groups of the hydantoin moiety,
(b) both hydantoinic nitrogens, (c) the centroid of the pyrrol-
idine and piperidine ring of the hydantoin substructure, and
(d) the common pharmacophore portions. The molecules were
aligned initially via RMS fit, followed by field fit optimization
to the template molecule. Field fitting sometimes forced the
molecules into high-energy conformations, in order to obtain
maximal similarity between the steric and electrostatic fields
of the template and the test molecules. The resulting struc-
tures were subsequently reoptimized without the field fit
option to allow relaxation of the molecule around the torsion
angle.
Low-energy conformers of the variously substituted mol-
ecules at the ortho, meta, and para positions of the phenyl ring
were chosen for optimal overlap with one another, under the
assumption that the substituents at the ortho and meta
positions always occupy the same cavity in the receptor. On
these basis all of the ortho and meta substituents of the phenyl
ring were oriented in the same direction. In our experience,
CoMFA fails to find a significant correlation if such an
approach is not used.
tion method. The r2
values listed for the different models
cross
are the maximal values, which were obtained considering the
number of components given in the tables. Following the
cross-validated analysis a non-cross-validated analysis was
performed using the optimum number of components previ-
ously identified. The non-cross-validated analyses were used
to make predictions of activities and to analyze the CoMFA
results. For both the cross-validated and non-cross-validated
analyses, the σ used was 2.0 as we found that σ ) 1 or 0.5 did
not significantly change the calculated r2 or SE.
The steric and electrostatic features of the final CoMFA
model were displayed as contour plots of the PLS regression
coefficients at each CoMFA region grid point. The steric
CoMFA contributions were contoured at the 75% and 25%
levels, with the “positive” steric contour (75%) colored green
and the “negative” steric contour (25%) colored yellow. The
electrostatic contribution contours were displayed in similar
fashion with red-colored positive contours (interaction of
ligands with the positive probe atom in these regions enhances
activity) at the 75% level, and blue-colored negative contours
(ligand interaction with the positive probe atom in these
regions lowers activity) at the 25% level.
Resu lts a n d Discu ssion
5-HT1A CoMF A Mod el. The CoMFA model derived
from the potencies of the whole set of analogs (1a -4l)
to inhibit [3H]-8-OH-DPAT binding at the 5-HT1A sites
was obtained using the alignment rule discussed in the
Methods section. The optimum number of components
was selected by identification of the point at which the
r2cross decreased and/or the SEcross increased significantly
with respect to the previous values. On the basis of this
criteria, eight components were selected (r2cross ) 0.840,
SEcross ) 0.365). Table 2 contains the statistics of the
CoMFA model derived for the 5-HT1A receptor. The
ratio of electrostatic and steric contributions to the final
model is 54.4:45.6, respectively. Observed and calcu-
lated pKi values are listed in Table 1 and plotted in
Figure 1.
In order to visualize the information content of the
derived 3D-QSAR model, CoMFA contour maps were
generated by interpolating the products between the 3D-
QSAR coefficients and their associated standard devia-
tions. It is worthwhile to remark that these types of
contours can only be found in the corresponding areas
of the lattice points characterized by variance of the
steric and electrostatic properties of the ligands. Thus,
their absence does not necessarily mean that a given
CoMF A Meth od . The comparative molecular field analysis
(CoMFA) was performed using the QSAR option of SYBYL
version 6.0 on a Silicon Graphics 4D/25 Personal Iris worksta-
tion. The partial atomic charges used in CoMFA were
computed using the AM118 semiempirical method available in
the MOPAC program. Single-point calculations were per-
formed on the geometries previously optimized with SYBYL/
MAXIMIN2.
The steric and electrostatic probe-ligand interaction ener-
gies (kcal/mol) were calculated with Lennard-J ones and
Coulomb potential functions of the Tripos force field using an
sp3 carbon probe carrying a charge of +1.0. The best results
were obtained when the steric and electrostatic energies were

1650 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 11
Lo´pez-Rodrı´guez et al.
Ta ble 1. 5-HT1A Receptor Binding Dataa with CoMFA Predictions for Derivatives 1-4
compd
X
n
R
Ki (nM) ( SEMb
pKi obsd (nM)
pKi calcd (nM)
resid
1a
1b
1c
1d
1e
1f
1g
1h
1i
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
1
1
1
1
1
1
1
1
1
1
1
1
H
85.3 ( 3.1
34.9 ( 0.7
58.4 ( 1.1
120 ( 10
500 ( 60
6095
-1.93
-1.54
-1.77
-2.08
-2.70
-3.79
-2.01
-1.49
-1.76
-1.90
-2.65
-3.78
-1.92
-1.59
-1.86
-2.04
-2.91
-3.72
-2.13
-1.36
-1.61
-2.02
-2.71
-3.87
-0.01
0.05
0.09
-0.04
0.21
-0.07
0.12
-0.13
-0.15
0.12
o-OCH3
m-Cl
m-CF3
p-F
p-NO2
H
o-OCH3
m-Cl
m-CF3
p-F
101 ( 8
31.1 ( 1.7
57.7 ( 5.7
78.6 ( 7.5
444 ( 52
5970
1j
1k
1l
0.06
0.09
p-NO2
2a
2b
2c
2d
2e
2f
2g
2h
2i
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
2
2
2
2
2
2
2
2
2
2
2
2
H
7550
234 ( 20
418 ( 60
123 ( 11
23932
501187
1349
45.5 ( 4.6
128 ( 10
65.8 ( 3.1
19274
-3.88
-2.37
-2.62
-2.09
-4.38
-5.70
-3.13
-1.66
-2.11
-1.82
-4.28
-4.96
-3.94
-2.33
-2.87
-2.54
-4.16
-5.36
-3.17
-1.73
-2.05
-1.70
-4.14
-4.89
0.06
-0.04
0.25
o-OCH3
m-Cl
m-CF3
p-F
p-NO2
H
o-OCH3
m-Cl
m-CF3
p-F
0.45
-0.22
-0.34
0.04
0.07
-0.06
-0.12
-0.14
-0.07
2j
2k
2l
p-NO2
93590
3a
3b
3c
3d
3e
3f
3g
3h
3i
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
3
3
3
3
3
3
3
3
3
3
3
3
H
19.2 (1.5
4.4 ( 0.6
55.9 (9.1
3.8 ( 0.5
1183
-1.28
-0.64
-1.75
-0.58
-3.07
-5.23
-2.19
-0.60
-1.73
-0.76
-2.78
-4.49
-1.31
-0.68
-1.72
-0.67
-2.99
-5.29
-2.35
-0.51
-1.37
-0.75
-2.94
-4.47
0.03
0.04
-0.03
0.09
-0.08
0.06
0.16
-0.09
-0.36
-0.01
0.16
o-OCH3
m-Cl
m-CF3
p-F
p-NO2
H
o-OCH3
m-Cl
m-CF3
p-F
168260
154 ( 10
4.1 ( 0.6
53.6 ( 1.5
5.7 ( 0.7
598 ( 70
30618
3j
3k
3l
p-NO2
-0.02
4a
4b
4c
4d
4e
4f
4g
4h
4i
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
4
4
4
4
4
4
4
4
4
4
4
4
H
24.8 ( 1.4
5.5 ( 0.7
11.3 ( 1.0
2.4 ( 0.6
89.9 ( 5.2
23932
78.5 ( 6.8
8.8 ( 0.9
7.2 ( 0.6
9.9 ( 0.9
57.9 ( 3.2
2582
-1.39
-0.74
-1.05
-0.38
-1.95
-4.38
-1.89
-0.95
-0.85
-0.99
-1.76
-3.41
-1.38
-0.82
-0.95
-0.28
-1.85
-4.04
-1.44
-0.96
-0.95
-0.69
-2.17
-3.81
-0.01
0.08
o-OCH3
m-Cl
m-CF3
p-F
p-NO2
H
o-OCH3
m-Cl
m-CF3
p-F
-0.10
-0.10
-0.10
-0.34
-0.45
0.01
0.10
-0.30
0.41
4j
4k
4l
p-NO2
0.40
a
b
Ki ( SEM values are derived from two to four experiments performed in triplicate. SEM is indicated when Ki values are obtained
from complete DRCs.
Ta ble 2. CoMFA-PLS Analysis Statistics for pKi Data at
5-HT1A Sites
SEcross
cross
0.365
0.840
0.206
0.981
245.613
no. of compounds
no. of components
steric fraction
48
8
0.456
0.544
r2
SE
r2
F
electrostatic fraction
pharmacophoric element is actually unimportant, but
only that all the examined compounds exert in that area
more or less the same steric or electronic influence.
Figures 2-4 show the CoMFA steric and electrostatic
contour map using compounds 4b (X ) -(CH2)3-, n )
4, R ) o-OCH3), 4j (X ) -(CH2)4-, n ) 4, R ) m-CF3),
and 2f (X ) -(CH2)3-, n ) 2, R ) p-NO2) as reference
structures. The green and yellow polyhedra describe
regions of space whose occupancy by the ligands respec-
tively increases or decreases affinity for the 5-HT1A
receptor. The green contours around the ortho and meta
positions of the phenyl ring indicate that bulky substit-
uents are tolerated in these positions. However the
F igu r e 1. Calculated vs observed pKi values at 5-HT1A
receptor sites (n ) 48, r ) 0.990, s ) 0.190, p < 0.001).
para position is surrounded by a yellow region. This
“unfavorable” contour is occupied by the p-NO2 group

Synthesis of Arylpiperazines
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 11 1651
F igu r e 2. Substitution at the ortho position by bulky substituents with negative potential is favorable for 5-HT1A affinity. Molecule
displayed is 4b (X ) -(CH2)3-, n ) 4, R ) o-OCH3).
F igu r e 3. Substitution at the meta position by bulky substituents with negative potential is favorable for 5-HT1A affinity. Molecule
displayed is 4j (X ) -(CH2)4-, n ) 4, R ) m-CF3).
F igu r e 4. The yellow and blue zones close to the para position indicate that bulky and electron-withdrawing substituents can
make a negative contribution. Yellow contours close to the hydantoin moiety represent an unfavorable bulk interaction for
compounds with n ) 1 and especially n ) 2. Molecule displayed is 2f (X ) -(CH2)3-, n ) 2, R ) p-NO2).
and is related to the presence of a steric restriction in
the receptor cavity delimiting the volume accessible to
the ligands. If this hypothesis is correct, bulky substit-
uents at this position, no matter their electrostatic
properties, would cause a strong decrease of potency.
In order to prove this assumption, a new ligand bearing
an electron-donating amino group at this position, 4m
(X ) -(CH2)3-, n ) 4, R ) p-NH2), was synthesized,
leading, as expected, to an inactive compound (Ki >
10 000 nM). Finally, another important yellow region
is located around the hydantoin moiety of the short-
chain molecules. This appears to suggest that an
unfavorable bulky interaction would be involved in the
decrease of affinity observed for derivatives with n ) 1
and especially with n ) 2.
The electrostatic contour maps are represented by red
and blue polyhedra describing regions where a high
electron density within the ligand structure increases

1652 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 11
Lo´pez-Rodrı´guez et al.
Ta ble 3. Statistical Parameters Derived for the 5-HT1A
Receptor CoMFA-PLS Analysis Model Used for Predictions
enhance the potency. However the strong steric hin-
drance at this position restricts completely the substitu-
tion.
SEcross
cross
0.360
0.825
0.202
0.978
229.726
no. of compounds
no. of components
steric fraction
42
7
r2
In order to evaluate the predictive capacity of the
CoMFA model, six molecules (1d , 2f, 2k , 3c, 4b, and
4g) were removed from the original data set (n ) 48)
used in the previous CoMFA. The molecules were
chosen on the basis of their affinity (one with high, two
with low, and three with moderate affinity). CoMFA
was redone for the remaining 42 compounds. The
CoMFA results for the 42 compounds are presented in
Table 3. This CoMFA was applied to the six omitted
compounds. A comparison of the observed and predicted
values for the six compounds tested shows an acceptable
agreement overall (Table 4).
SE
r2
F
0.399
0.601
electrostatic fraction
Ta ble 4. Observed vs Predicted 5-HT1A Receptor Binding
Values (pKi, nM)
compd
X
n
R
pKi obsd pKi pred
resid
1d
2f
2k
3c
4b
4g
-(CH2)3-
-(CH2)3-
-(CH2)4-
-(CH2)3-
-(CH2)3-
-(CH2)4-
1
2
2
3
4
4
m-CF3
p-NO2
p-F
m-Cl
o-OCH3
H
-2.08
-5.70
-4.28
-1.75
-0.74
-1.89
-1.89
-5.00
-3.88
-1.40
-0.89
-1.34
-0.19
-0.70
-0.40
-0.35
0.15
-0.55
r1 CoMF A Mod el. A set of 42 molecules was used
in the CoMFA study (Table 5). The molecules 1h , 2a ,
2e, 2g, 2k , and 4l were not included in the analysis in
order to keep a balance between active and inactive
molecules. These compounds were used later as a test
set to investigate the predictive power of the derived
model. The stepwise F-test performed on the PRESS
values justified seven components with SEcross ) 0.287
and r2cross ) 0.809. As expected, the corresponding non-
or decreases, respectively, the affinity. The red contours
surrounding the ortho and meta positions show that
negative potentials in such positions increase the 5-HT1A
affinity. The blue polyhedron in the para position
indicates that electron-withdrawing substituents exert
a negative effect on the affinity. This seems to suggest
that positive potentials in that region of the phenyl ring
Ta ble 5. R1 Receptor Binding Dataa with CoMFA Predictions for Derivatives 1-4
compd
X
n
R
Ki (nM) ( SEMb
pKi obsd (nM)
pKi calcd (nM)
resid
1a
1b
1c
1d
1e
1f
1g
1i
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
1
1
1
1
1
1
1
1
1
1
1
H
2588
500 ( 65
292 ( 15
3296
1193
260000
2075
135 ( 5
3296
869 ( 75
260000
-3.41
-2.70
-2.47
-3.52
-3.08
-5.41
-3.32
-2.13
-3.52
-2.94
-5.41
-3.19
-2.76
-2.57
-3.15
-3.38
-5.44
-3.46
-2.22
-3.47
-2.82
-5.30
-0.22
0.06
0.10
-0.37
0.30
0.03
0.14
0.09
o-OCH3
m-Cl
m-CF3
p-F
p-NO2
H
m-Cl
m-CF3
p-F
1j
1k
1l
-0.05
-0.12
-0.11
p-NO2
2b
2c
2d
2f
2h
2i
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
2
2
2
2
2
2
2
2
o-OCH3
m-Cl
m-CF3
p-NO2
o-OCH3
m-Cl
190 ( 38
500 ( 65
3935
260000
131 ( 28
375 ( 16
4325
-2.28
-2.70
-3.60
-5.40
-2.12
-2.57
-3.64
-5.30
-2.40
-2.81
-3.65
-5.24
-2.31
-2.67
-3.52
-5.21
0.12
0.11
0.05
-0.16
0.19
0.10
-0.12
-0.09
2j
2l
m-CF3
p-NO2
199000
3a
3b
3c
3d
3e
3f
3g
3h
3i
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
3
3
3
3
3
3
3
3
3
3
3
3
H
15.4 ( 3.9
3.1 ( 0.5
21.6 ( 1.1
109 ( 9
-1.19
-0.49
-1.33
-2.04
-1.40
-3.88
-1.06
-1.00
-1.24
-1.96
-1.41
-3.82
-1.15
-0.52
-1.44
-1.95
-1.31
-3.70
-0.81
-0.90
-1.32
-2.07
-1.54
-4.04
-0.04
0.03
0.11
-0.09
-0.09
-0.18
-0.25
-0.10
0.08
o-OCH3
m-Cl
m-CF3
p-F
p-NO2
H
o-OCH3
m-Cl
m-CF3
p-F
25.2 ( 1.1
7533
11.4 ( 0.9
9.9 ( 1.0
17.9 ( 1.1
90.4 ( 5.1
25.8 ( 1.1
6652
3j
3k
3l
0.11
0.13
0.22
p-NO2
4a
4b
4c
4d
4e
4f
4g
4h
4i
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
4
4
4
4
4
4
4
4
4
4
4
H
26.4 ( 1.9
8.3 ( 0.3
9.6 ( 0.9
64.9 ( 2.6
47.2 ( 1.8
247 ( 63
18.6 ( 3.1
8.6 ( 1.0
12.1 ( 1.2
72.4 ( 8.0
10.4 ( 0.6
-1.42
-0.92
-0.98
-1.81
-1.67
-2.39
-1.27
-0.94
-1.08
-1.86
-1.02
-1.29
-0.90
-0.97
-1.72
-1.53
-2.63
-1.10
-0.98
-0.81
-2.05
-1.41
-0.13
-0.02
-0.01
-0.09
-0.14
0.24
-0.17
0.04
-0.27
0.19
0.39
o-OCH3
m-Cl
m-CF3
p-F
p-NO2
H
o-OCH3
m-Cl
m-CF3
p-F
4j
4k
a
b
Ki ( SEM values are derived from two to four experiments performed in triplicate. SEM is indicated when Ki values are obtained
from complete DRCs.

Synthesis of Arylpiperazines
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 11 1653
Ta ble 6. CoMFA-PLS Analysis Statistics for pKi Data at R1
Sites
is 61:39, respectively. Observed and calculated pKi
values are reported in Table 5 and plotted in Figure 5.
SEcross
cross
0.287
0.809
0.179
0.985
329.391
no. of compounds
no. of components
steric fraction
42
7
0.390
0.610
Figures 6-8 show the contributions of the steric and
electrostatic fields to the CoMFA model. The green
polyhedra around the ortho and meta positions suggest
that the occupation of those regions increases affinity.
However the meta position presents an optimum volume
of substituent between 11 and 25 Å3 (the VW of chloro
and CF3 group, respectively). Thus, while a chloro
substituent exerts a modest positive effect, the introduc-
tion of a CF3 group decreases slightly the affinity.
Figure 7 shows that the m-CF3 group reaches an
adjacent yellow zone, which extends around the whole
para position. The CoMFA model describes the loss of
potency associated with the p-nitro substitution as a
strong unfavorable bulk interaction.
r2
SE
r2
F
electrostatic fraction
The red contours surrounding the ortho and meta
positions show that the presence of negative potentials
in those regions leads to an increase of affinity. The
interpretation of the inner red and the outer blue
polyhedra located near the para position is more com-
plex. The presence of a fluoro substituent at the para
position of the phenylpiperazine system leads to a
moderate increase of affinity. This effect is described
by the red contours located in the corresponding areas
of the p-fluoro derivatives. The blue polyhedra take into
account the about 100-fold drop of affinity observed
when a nitro group is introduced at this position.
However, such a contour should not be overemphasized
since the loss of activity of the p-NO2 derivatives can
F igu r e 5. Calculated vs observed pKi at the R1 receptor site
(n ) 42, r ) 0.992, s ) 0.164, p < 0.001).
cross-validated analysis yields a better data fitting (SE
) 0.179 and r2 ) 0.985) (Table 6). The ratio of
electrostatic and steric contributions to the final model
F igu r e 6. Substitution at the ortho position by bulky substituents with negative potential is favorable for affinity. Molecule
displayed is 3b (X ) -(CH2)3-, n ) 3, R ) o-OCH3).
F igu r e 7. The yellow region close to the meta position indicates that too bulky groups decrease affinity. Molecule displayed is
4d (X ) -(CH2)3-, n ) 4, R ) m-CF3).

1654 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 11
Lo´pez-Rodrı´guez et al.
F igu r e 8. A bulky substituent with negative potential at the para position can make a negative contribution. Molecule displayed
is 2l (X ) -(CH2)4-, n ) 2, R ) p-NO2).
Ta ble 7. Observed vs Predicted R1 Receptor Binding Values
(pKi, nM)
chain derivatives bearing bulky substituents with nega-
tive potential at the meta position. Thus, the new ligand
4n (X ) -(CH2)3-, n ) 4, R ) m-NHCOPri) was
designed and synthesized. This analog bound at 5-HT1A
sites (Ki ) 102 ( 8 nM) and exhibited high selectivity
over the R1 receptor (Ki > 10000 nM). Although the
potency of 4n at the 5-HT1A receptor is moderate, the
high 5-HT1A/R1 selectivity supports our CoMFA models
and affords insights for the design of new selective
compounds. In order to increase the affinity for the
5-HT1A receptor, further synthesis and biological evalu-
ation of new derivatives structurally related to 4n are
in progress, and the results will be reported in due
course.
In conclusion, the CoMFA method has been success-
fully applied to a set of recently described hydantoin-
phenylpiperazines with affinity for 5-HT1A and R1
receptors. The resulting 3D-QSAR models provide
significant correlation of steric and electrostatic fields
with the biological affinities. The CoMFA coefficient
contour plots provide a self-consistent picture of the
main chemical features responsible for the pKi varia-
tions and also result in predictions which agree with
experimental values. By comparison of the CoMFA
models, suggestions can be made about the improve-
ment of 5-HT1A/R1 selectivity. This information seems
to be very useful for the design of new agents possessing
high selectivity for 5-HT1A vs R1 receptors.
compd
X
n
R
pKi obsd
pKi pred
1h
2a
2e
2g
2k
4l
-(CH2 )4-
-(CH2 )3-
-(CH2 )3-
-(CH2 )4-
-(CH2 )4-
-(CH2 )4-
1
2
2
2
2
4
o-OCH3
H
p-F
H
p-F
<-3
<-3
<-3
<-3
<-3
<-3
-2.94
-3.11
-3.35
-2.90
-3.46
-2.92
p-NO2
be due to an undesired bulk interaction. This is
supported by the loss of affinity exhibited by the ligand
4m (R ) p-NH2) (Ki > 1000 nM). This proves that the
potency of the para-substituted derivatives is modulated
mainly by steric factors and not by electrostatic ones.
Finally, the blue contour in the proximity of the hydan-
toin moiety describes the detrimental effect on affinity
caused by the shortening of the chain.
As already mentioned, compounds 1h , 2a , 2e, 2g, 2k ,
and 4l allow us to evaluate the efficiency of the derived
CoMFA model in estimating binding affinity values for
structures outside the training set. Table 7 lists the
observed pKi values of the compounds belonging to the
test set together with the corresponding pKi values
predicted by the CoMFA model.
Com p a r ison of CoMF A Mod els. Comparison of the
CoMFA contour maps generated for both analyses gives
an additional understanding for 5-HT1A/R1 selectivity,
leading to four important conclusions. (a) Substitution
at the ortho position by a group with negative potential
is favorable to affinity for both receptors.
Exp er im en ta l Section
Ch em istr y. Melting points (uncorrected) were determined
on a Gallenkamp electrothermal apparatus. Infrared (IR)
spectra were obtained on a Perkin-Elmer 781 infrared spec-
trophotometer. 1H- and 13C-NMR spectra were recorded on a
Varian VXR-300S or Bruker 250-AM instrument. Chemical
shifts (δ) are expressed in parts per million relative to internal
tetramethylsilane; coupling constants (J ) are in hertz. El-
emental analyses (C, H, N) were determined within 0.4% of
the theoretical values. Thin-layer chromatography (TLC) was
run on Merck silica gel 60 F-254 plates. For flash chroma-
tography, Merck silica gel type 60 (size 230-400 mesh) was
used. Unless stated otherwise, starting materials were used
as high-grade commercial products.
(b) The meta position seems to be implicated in
5-HT1A/R1 selectivity. While the 5-HT1A receptor is able
to accommodate bulky substituents in the region of its
active site, the steric requirements of the R1 receptor
at this position are more restricted. In addition the R1
receptor exhibits an optimum volume of substituent
(11-25 Å3).
(c) For both receptors the para position represents a
region where the volume accessible by the ligands is
limited. Only very small substituents, like fluoro, can
be accommodated in the receptor pocket.
(d) Finally, both structural features, the hydantoin
moiety and the side chain length, seem to modulate not
only the affinity, but also the 5-HT1A/R1 selectivity.
Thus the compounds with n ) 1 present a moderate
potency at the 5-HT1A receptor and are much less active
at the R1 receptor.
The following compounds were synthesized by published
procedures: 1,3-dioxoperhydropyrrolo[1,2-c]imidazole,21 1-(m-
nitrophenyl)piperazine,22 2-(4-bromobutyl)-1,3-dioxoperhydro-
pyrrolo[1,2-c]imidazole,1 and 2-[4-[4-(p-nitrophenyl)piperazin-
1-yl]butyl]-1,3-dioxoperhydropyrrolo[1,2-c]imidazole (4f).1
2-[4-[4-(m -Nitr op h en yl)p ip er a zin -1-yl]bu tyl]-1,3-d iox-
op er h yd r op yr r olo[1,2-c]im id a zole (4o). To a suspension
of 2-(4-bromobutyl)-1,3-dioxoperhydropyrrolo[1,2-c]imidazole
(2.6 g, 9.5 mmol) and 1-(m-nitrophenyl)piperazine (3.0 g, 14.6
These results suggest that a good way to improve
5-HT1A/R1 selectivity would be the synthesis of long-

Synthesis of Arylpiperazines
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 11 1655
mmol) in acetonitrile (19 mL) was added 2.0 mL of triethyl-
amine (1.5 g, 14.6 mmol). The mixture was refluxed for 20-
24 h. Then, the solvent was evaporated under reduced
pressure, and the residue was resuspended in water and
extracted with dichloromethane (3 × 100 mL). The combined
organic layers were washed with water and dried over MgSO4.
After evaporation of the solvent the crude oil was purified by
column chromatography (ethyl acetate/ethanol, 9:1) to give
2.21 g (58%) of 4o, which was converted to its hydrochloride
4n , which was converted to its hydrochloride salt: mp 189-
192 °C (methanol/ethyl ether); IR (CHCl3, cm-1) 3060-3300
(NH), 1770 (CON), 1700 (NCON), 1650 (CONH), 1610, 1500,
1450 (Ar); 1H-NMR (CDCl3, as free base) δ 1.16 (d, J ) 6.6
Hz, 6H, 2CH3), 1.42-1.67 (m, 5H, -(CH2)2-, H7), 1.97-2.04
(m, 2H, 2H6), 2.18-2.22 (m, 1H, H7), 2.32 (t, J ) 7.5 Hz, 2H,
CH2-Npip), 2.41-2.49 (m, 5H, 2CH2-pip, CH), 3.10-3.21 (m,
5H, 2CH2-pip, H5), 3.42 (t, J ) 7.5 Hz, 2H, NCH2), 3.60 (dt, J
) 11.1, 7.5 Hz, 1H, H5), 4.00 (dd, J ) 7.5, 7.2 Hz, 1H, H7a),
6.57 (dd, J ) 8.4, 2.1 Hz, 1H, H6-phenyl), 6.76 (d, J ) 7.8 Hz,
1H, H4-phenyl), 7.08 (t, J ) 7.8 Hz, 1H, H5-phenyl), 7.34 (s,
1H, H2-phenyl), 7.41 (s, 1H, CONH); 13C-NMR (CDCl3, as free
base) δ 19.8 (2CH3), 24.0, 26.1 (-(CH2)2-), 27.1 (C6), 27.6 (C7),
36.8 (CH), 38.9 (NCH2), 45.6 (C5), 48.9 (2CH2-pip), 53.2 (2CH2-
pip), 58.0 (CH2-Npip), 63.4 (C7a), 107.4 (C2-phenyl), 110.6, 111.5
(C4- and C6-phenyl), 129.4 (C5-phenyl), 139.2 (C3-phenyl), 152.0
(C1-phenyl), 160.9 (C3), 174.1 (C1), 175.5 (CONH). Anal.
(C24H35N5O3‚2HCl‚H2O) C, H, N.
salt: mp 197-198 °C (methanol/ethyl ether); IR (CHCl3, cm-1
)
1770 (CON), 1710 (NCON), 1620, 1580, 1500, 1450 (Ar), 1530
(NO2); 1H-NMR (CDCl3, as free base) δ 1.43-1.66 (m, 5H,
-(CH2)2-, H7), 1.94-2.06 (m, 2H, 2H6), 2.13-2.22 (m, 1H, H7),
2.35 (t, J ) 7.4 Hz, 2H, CH2-Npip), 2.52 (t, J ) 5.1 Hz, 4H,
2CH2-pip), 3.15-3.23 (m, 5H, 2CH2-pip, H5), 3.43 (t, J ) 7.0
Hz, 2H, NCH2), 3.61 (dt, J ) 11.7, 7.7 Hz, 1H, H5), 4.02 (dd, J
) 9.0, 7.7 Hz, 1H, H7a), 7.10 (dd, J ) 8.3, 2.2 Hz, 1H,
H6-phenyl), 7.29 (t, J ) 8.2 Hz, 1H, H5-phenyl), 7.56 (dd, J )
8.0, 1.2 Hz, 1H, H4-phenyl), 7.62 (t, J ) 2.1 Hz, 1H, H2-phenyl);
13C-NMR (CDCl3, as free base) δ 24.0, 26.0 (-(CH2)2-), 27.1
(C6), 27.6 (C7), 38.8 (NCH2), 45.6 (C5), 48.3 (2CH2-pip), 52.9
(2CH2-pip), 57.9 (CH2-Npip), 63.4 (C7a), 109.6 (C2-phenyl),
113.6 (C4-phenyl), 121.1 (C6-phenyl), 129.7 (C5-phenyl), 149.3
(C1-phenyl), 151.9 (C3-phenyl), 160.9 (C3), 174.1 (C1). Anal.
(C20H27N5O4‚HCl) C, H, N.
Gen er a l P r oced u r e. P r ep a r a tion of Com p ou n d s 4m
a n d 4p . To a solution of 4f,o (5 mmol) in methanol (18 mL)
was added 0.1 g of 10% Pd(C). The mixture was hydrogenated
(35 psi) at room temperature for 2 h. The reaction mixture
was filtered over Celite and evaporated to dryness to afford
the pure amines. The free base was converted to its hydro-
chloride salt.
2-[4-[4-(p-Am in op h en yl)p ip er a zin -1-yl]bu tyl]-1,3-d iox-
op er h yd r op yr r olo[1,2-c]im id a zole (4m ): yield 1.62 g (87%);
mp 281-284 °C (methanol/ethyl ether); IR (KBr, cm-1) 3420,
3350, 3220 (NH2), 1770 (CON), 1700 (NCON), 1520, 1450 (Ar);
1H-NMR (CDCl3, as free base) δ 1.40-1.66 (m, 5H, -(CH2)2-,
H7), 1.91-2.04 (m, 2H, 2H6), 2.12-2.22 (m, 1H, H7), 2.33 (t, J
) 7.2 Hz, 2H, CH2-Npip), 2.51 (t, J ) 4.8 Hz, 4H, 2CH2-pip),
2.97 (t, J ) 4.8 Hz, 4H, 2CH2-pip), 3.12-3.20 (m, 1H, H5), 3.32
(sa, 2H, NH2), 3.42 (t, J ) 7.2 Hz, 2H, NCH2), 3.60 (dt, J )
11.4, 7.5 Hz, 1H, H5), 3.99 (dd, J ) 8.1, 7.5 Hz, 1H, H7a), 6.57
(d, J ) 8.7 Hz, 2H, H3- and H5-phenyl), 6.73 (d, J ) 8.7 Hz,
2H, H2- and H6-phenyl); 13C-NMR (CDCl3, as free base) δ 23.9,
25.9 (-(CH2)2-), 26.9 (C6), 27.4 (C7), 38.7 (NCH2), 45.4 (C5),
50.7 (2CH2-pip), 53.2 (2CH2-pip), 57.9 (CH2-Npip), 63.2 (C7a),
116.0 (C2- and C6-phenyl), 118.4 (C3- and C5-phenyl), 140.0 (C4-
phenyl), 144.4 (C1-phenyl), 160.7 (C3), 173.8 (C1). Anal.
(C20H29N5O2‚2HCl‚1/2H2O) C, H, N.
2-[4-[4-(m -Am in op h en yl)p ip er a zin -1-yl]bu tyl]-1,3-d iox-
op er h yd r op yr r olo[1,2-c]im id a zole (4p ): yield 1.73 g (93%);
mp 92-94 °C (methanol/ethyl ether); IR (CHCl3, cm-1) 3120-
3600 (NH2), 1770 (CON), 1700 (NCON), 1600, 1500, 1450 (Ar);
1H-NMR (CDCl3, as free base) δ 1.45-1.65 (m, 5H, -(CH2)2-,
H7), 1.96-2.04 (m, 2H, 2H6), 2.13-2.23 (m, 1H, H7), 2.35 (t, J
) 7.5 Hz, 2H, CH2-Npip), 2.52 (t, J ) 5.1 Hz, 4H, 2CH2-pip),
3.10 (t, J ) 4.8 Hz, 4H, 2CH2-pip), 3.15-3.21 (m, 1H, H5), 3.42
(t, J ) 7.2 Hz, 2H, NCH2), 3.59 (dt, J ) 11.1, 7.5 Hz, 1H, H5),
4.00 (dd, J ) 9.0, 7.5 Hz, 1H, H7a), 6.14 (dd, J ) 8.1, 1.5 Hz,
1H, H4-phenyl), 6.17 (t, J ) 1.8 Hz, 1H, H2-phenyl), 6.27 (dd,
J ) 7.8, 1.8 Hz, 1H, H6-phenyl), 6.96 (t, J ) 7.8 Hz, 1H, H5-
phenyl); 13C-NMR (CDCl3, as free base) δ 23.7, 25.9 (-(CH2)2-),
26.9 (C6), 27.4 (C7), 38.6 (NCH2), 45.4 (C5), 48.8 (2CH2-pip),
53.0 (2CH2-pip), 57.8 (CH2-Npip), 63.2 (C7a), 102.7 (C2-phenyl),
106.7, 106.9 (C4- and C6-phenyl), 129.7 (C5-phenyl), 147.2 (C3-
phenyl), 152.3 (C1-phenyl), 160.7 (C3), 173.9 (C1). Anal.
(C20H29N5O2‚3HCl) C, H, N.
Ra d ioliga n d Bin d in g Assa ys. For all receptor binding
assays, male Sprague-Dawley rats (Rattus norvegicus albi-
nus), weighing 180-200 g, were killed by decapitation and the
brains rapidly removed and dissected.
5-HT1A Recep tor . The receptor binding studies were
performed by a modification of a previously described proce-
dure.23 The cerebral cortex was homogenized in 10 volumes
of ice-cold Tris-buffer (50 mM Tris-HCl, pH 7.7, at 25 °C) and
centrifuged at 28000g for 15 min. The membrane pellet was
washed twice by resuspension and centrifugation. After the
second wash the resuspended pellet was incubated at 37 °C
for 10 min. Membranes were then collected by centrifugation,
and the final pellet was resuspended in 50 mM Tris-HCl, 5
mM MgSO4, and 0.5 mM EDTA buffer (pH 7.4 at 37 °C).
Fractions of the final membrane suspension (about 1 mg of
protein) were incubated at 37 °C for 15 min with 0.6 nM [3H]-
8-OH-DPAT [8-hydroxy-2-(di-n-propylamino)tetralin] (133 Ci/
mmol) in the presence or absence of several concentrations of
the competing drug, in a final volume of 1.1 mL of assay buffer
(50 mM Tris-HCl, 10 nM clonidine, 30 nM prazosin, pH 7.4 at
25 °C). Nonspecific binding was determined with 10 µM 5-HT.
r1 Ad r en ocep tor . The radioligand receptor binding stud-
ies were performed according to a previously described pro-
cedure.24 The cerebral cortex was homogenized in 20 volumes
of ice-cold buffer (50 mM Tris-HCl, 10 mM MgCl2, pH 7.7 at
25 °C) and centrifuged at 30000g for 15 min. Pellets were
washed twice by resuspension and centrifugation. Final
pellets were resuspended in the same buffer. Fractions of the
final membrane suspension (about 250 µg of protein) were
incubated at 25 °C for 30 min with 0.2 nM [3H]prazosin (23
Ci/mmol) in the presence or absence of several concentrations
of the competing drug, in a final volume of 2 mL of buffer.
Nonspecific binding was determined with 10 µM phentolamine.
Ack n ow led gm en t. This work was supported by
DGICYT (PB940289) and the Universidad Complutense
(PR218/94-5657). The authors are grateful to MEC for
an FPI grant to M. L. Rosado, and to UNED for
predoctoral grants to B. Benhamu´ and E. Ferna´ndez.
Refer en ces
(1) Lo´pez-Rodr´ıguez, M. L.; Rosado, M. L.; Benhamu´, B.; Morcillo,
M. J .; Sanz, A. M.; Orensanz, L.; Beneytez, M. E.; Fuentes, J .
A.; Manzanares, J . Synthesis and Structure-Activity Relation-
ships of a New Model of Arylpiperazines. 1. 2-[[4-(o-Methoxy-
phenyl)piperazin-1-yl]methyl]-1,3-dioxoperhidroimidazo[1,5-a]-
pyridine: A Selective 5-HT1A Receptor Agonist. J . Med. Chem.
1996, 39, 4439-4450.
(2) Portions of this research were presented at the 11th European
Symposium on Quantitative Structure-Activity Relationships:
Computer-Assisted Lead Finding and Optimization, Lausanne,
Switzerland, September 1-6, 1996; abstract p P-29.D.
(3) Fargin, A.; Raymond, J . R.; Lohse, M. J .; Kobilka, B. K.; Caron,
M. G.; Lefkowitz, R. J . The Genomic Clone G-21 which Re-
sembles a â-Adrenergic Receptor Sequence Encodes the 5-HT1A
Receptor. Nature 1988, 335, 358-360.
2-[4-[4-[m -(2-Meth ylp r op a n a m id o)p h en yl]p ip er a zin -1-
yl]bu tyl]-1,3-d ioxop er h yd r op yr r olo[1,2-c]im id a zole (4n ).
To a solution of 4p (1.7 g, 4.6 mmol) in pyridine (52 mL) at 0
°C was added dropwise 0.5 mL of isobutyryl chloride (0.49 g,
4.6 mmol). After being stirred at room temperature for 1.5 h,
the mixture was washed with a saturated aqueous solution of
CuSO4, water, and a saturated aqueous solution of NaCl
(brine). The organic layer was dried (NaSO4) and the solvent
evaporated under reduced pressure to afford 1.25 g (62%) of

1656 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 11
Lo´pez-Rodrı´guez et al.
(4) Libert, F.; Parmentier, M.; Lefort, A.; Dinsart, C.; Van Sande,
J .; Maenhaut, C.; Simons, M.-J .; Dumont, J . E.; Vassart, G.
Selective Amplification and Cloning of Four New Members of
the G Protein-Coupled Receptor Family. Science 1989, 244, 569-
572.
(14) Cramer, R. D., III; DePriest, S. A.; Patterson, D. E.; Hecht, P.
The Developing Practice of Comparative Molecular Field Analy-
sis. In 3D QSAR in Drug Design. Theory Methods and Applica-
tions; Kubinyi, H., Ed.; ESCOM Science Publishers: Leiden,
1993; pp 443-485.
(15) Schaper, K.-J .; Emig, P.; Engel, J .; Fleischhauer, I.; Kutscher,
B.; Rosado, M. L.; Lo´pez-Rodr´ıguez, M. L. Dose Response
Relationships, Biotest Intercorrelations, QSAR and the Saving
of Animals Experiments. 10th European Symposium on Structure-
Activity Relationships: QSAR and Molecular Modelling. Bar-
celona, 1994.
(16) Cheng, Y. C.; Prusoff, W. H. Relationship between the Inhibition
Constant (Ki) and the Concentration of Inhibitor which Causes
50 Per Cent Inhibition (IC50) of an Enzymatic Reaction. Biochem.
Pharmacol. 1973, 22, 3099-3108.
(17) SYBYL 6.0, Tripos Associates, Saint Louis, MO 63144.
(18) Dewar, M. J . S.; Zoebisch, E. G.; Healy, E. F.; Stewart, J . J . P.
AM1: A New General Purpose Quantum Mechanical Molecular
Model. J . Am. Chem. Soc. 1985, 107, 3902-3909.
(19) Wold, S.; Albano, C.; Dunn, W. J ., III; Edlund, U.; Esbenson,
K.; Geladi, P.; Hellberg, S.; J ohansson, E.; Lindberg, W.; Sjo¨s-
tro¨m, M. In Chemometrics: Mathematics and Statistics in
Chemistry; Kowalsky, B. R., Ed.; Reidel: Dordrecht, 1984; pp
17-95.
(20) Dunn, W. J ., III; Wold, S.; Edlund, U.; Hellberg, S.; Gasteiger,
J . Multivariate Structure-Activity Relationship Between Data
from a Battery of Biological Tests and an Ensemble of Structure
Descriptors: The PLS Method. Quant. Struct.-Act. Relat. 1984,
3, 131-137.
(21) Dakin, H. D. Amino Acids of Gelatin. J . Biol. Chem. 1920, 44,
499-529.
(22) Martin, G. E.; Elgin, R. J ., J r.; Mathiasen, J . R.; Davis, C. B.;
Kesslick, J . M.; Baldy, W. J .; Shank, R. P.; DiStefano, D. L.;
Fedde, C. L.; Scott, M. K. Activity of Aromatic Substituted
Phenylpiperazines Lacking Affinity for Dopamine Binding Sites
in a Preclinical Test of Antipsychotic Efficacy. J . Med. Chem.
1989, 32, 1052-1056.
(23) Clark, R. D.; Weinhardt, K. K.; Berger, J .; Fisher, L. E.; Brown,
C. M.; MacKinnon, A. C.; Kilpatrick, A. T.; Spedding, M. 1,9-
Alkano-bridged 2,3,4,5-Tetrahydro-1H-3-benzazepines with Af-
finity for the R2-Adrenoceptor and the 5-HT1A Receptor. J . Med.
Chem. 1990, 33, 633-641.
(5) Hoflack, J .; Trumpp-Kallmeyer, S.; Hibert, M. F. Molecular
Modelling of G Protein-Coupled Receptors. In 3D QSAR in Drug
Design. Theory Methods and Applications; Kubinyi, H., Ed.;
ESCOM Science Publishers: Leiden, 1993; pp 355-372.
(6) El-Bermawy, M.; Raghupathi, R.; Ingher, S. P.; Teitler, M.;
Maayani, S.; Glennon, R. A. 4-[4-(1-Noradamantane-
carboxamido)butyl]-1-(2-methoxyphenyl)piperazine: a High-Af-
finity 5-HT1A-Selective Agent. Med. Chem. Res. 1992, 2, 88-95.
(7) Mokrosz, M. J .; Chojnacka-Wo´jcik, E.; Tatarczyn´ska, E.; Klodz-
in´ska, A.; Filip, M.; Boksa, J .; Charakchieva-Minol, S.; Mokrosz,
J . L. 1-(2-Methoxyphenyl)-4-[(4-succinimido)butyl]piperazine
(MM-77): A New, Potent, Postsynaptic Antagonist of 5-HT1A
Receptors. Med. Chem. Res. 1994, 4, 161-169.
(8) Lo´pez-Rodr´ıguez, M. L.; Morcillo, M. J .; Rosado, M. L.; Benhamu´,
B.; Sanz, A. M. 2-[[4-(o-Methoxyphenyl)piperazin-1-yl]methyl]-
1,3-dioxoperhidroimidazo[1,5-a]pyridine as
a New Selective
5-HT1A Receptor Ligand. Bioorg. Med. Chem. Lett. 1996, 6, 689-
694.
(9) Hibert, M. F.; Gittos, M. W.; Middlemiss, D. N.; Mir, A. K.;
Fozard, J . R. Graphics Computer-Aided Receptor Mapping as a
Predictive Tool for Drug Design: Development of Potent, Selec-
tive and Stereospecific Ligands for the 5-HT1A Receptor. J . Med.
Chem. 1988, 31, 1087-1093.
(10) El-Bermawy, M. A.; Lotter, H.; Glennon, R. A. Comparative
Molecular Field Analysis of the Binding of Arylpiperazines at
5-HT1A Serotonin Receptors. Med. Chem. Res. 1992, 2, 290-297.
(11) Mokrosz, J . L.; Duszynska, B. Structure-Activity Relationships
Studies of CNS Agents. Part VII. The Effect of the Imidazole
Fragment in 2-Substituted 1-[3-(4-Aryl-1-piperazinyl)propyl]-
imidazoles on their Interaction Modes with 5-HT1A and 5-HT2
Receptors. Pol. J . Pharmacol. Pharm. 1992, 44, 527-538.
(12) Campbell, S. F.; Davey, M. J .; Hardstone, J . D.; Lewis, B. N.;
Palmer, M. J . 2,4-Diamino-6,7-dimethoxyquinazolines. 1. 2-[4-
(1,4-Benzodioxan-2-ylcarbonyl)piperazin-1-yl] Derivatives as R1-
Adrenoceptor Antagonists and Antihypertensive Agents. J . Med.
Chem. 1987, 30, 49-57.
(24) Ambrosio, E.; Montero, M. T.; Ferna´ndez, I.; Azuara, M. C.;
Orensanz, L. M. [3H]Prazosin Binding to Central Nervous
System Regions of Male and Female Rats. Neurosci. Lett. 1984,
49, 193-197.
(13) Cramer, R. D., III; Patterson, D. E.; Bunce, J . D. Comparative
Molecular Field Analysis (CoMFA). 1. Effect of Shape on Binding
of Steroids to Carrier Proteins. J . Am. Chem. Soc. 1988, 110,
5959-5967.
J M960744G