Synthesis of 3-arylpiperazinylalkylpyrrolo[3,2-d]pyrimidine-2,4-dione derivatives as novel, potent, and selective α1-adrenoceptor ligands

Article abstract of DOI:10.1021/jm040870h

Novel compounds characterized by a pyrrolo[3,2-d]pyrimidine-2,4-dione (PPm) system connected through an alkyl chain to a phenylpiperazine (PPz) residue were designed as structural analogues of the α₁-adrenoceptor (α₁-AR) ligand RN5 (

Full text of DOI:10.1021/jm040870h

2420
J. Med. Chem. 2005, 48, 2420-2431
Synthesis of 3-Arylpiperazinylalkylpyrrolo[3,2-d]pyrimidine-2,4-dione
Derivatives as Novel, Potent, and Selective r1-Adrenoceptor Ligands‡
Emanuele Patane`,§ Valeria Pittala`,*,§ Francesco Guerrera,§ Loredana Salerno,§ Giuseppe Romeo,§
Maria Angela Siracusa,§ Filippo Russo,§ Fabrizio Manetti,# Maurizio Botta,# Ilario Mereghetti,†
Alfredo Cagnotto,† and Tiziana Mennini†
Dipartimento di Scienze Farmaceutiche, Universita` di Catania, viale A. Doria 6, 95125 Catania, Italy, Istituto di Ricerche
Farmacologiche “Mario Negri”, via Eritrea 62, 20157 Milano, Italy, and Dipartimento Farmaco Chimico Tecnologico,
Universita` degli Studi di Siena, via A. Moro, 53100 Siena, Italy
Received August 2, 2004
Novel compounds characterized by a pyrrolo[3,2-d]pyrimidine-2,4-dione (PPm) system connected
through an alkyl chain to a phenylpiperazine (PPz) residue were designed as structural
analogues of the R1-adrenoceptor (R1-AR) ligand RN5 (1). In this new series of derivatives an
arylpyrrolo moiety has replaced the indole nucleus of RN5. Several structural modifications
were performed on the PPm and PPz moieties and the connecting alkyl chain. These compounds
were synthesized and tested in radioligand binding experiments where many of them showed
interesting binding profiles. Some compounds, including 31, 34, and 36, displayed substantial
R1-AR selectivity with respect to serotoninergic 5-HT1A and dopaminergic D1 and D2 receptors.
Two different molecular modeling approaches (pharmacophoric mapping and quantitative
structure-affinity relationship analysis) have been applied to rationalize, at a quantitative
level, the relationships between affinity toward R1-ARs and the structure of the studied
compounds. Several QSAR models have been reported and described, accounting for the
influence of various molecular portions on such affinity data.
Chart 1
Introduction
The adrenoceptors (ARs) mediate the functional ef-
fects of catecholamines such as adrenaline and nor-
adrenaline, playing important roles in the control of
blood pressure, myocardial contractile rate and force,
airway reactivity, and a variety of metabolic functions.1,2
The ARs are divided into three principal families: R1,
R2, and â; all ARs are members of the superfamily of
G-protein-coupled receptors that transduce signal across
the cell membrane, thus initiating a variety of intra-
cellular biochemical events.1,2 The R1-ARs are presently
classified into three subtypes designated as R1A, R1B, R1D
and having high affinity for prazosin.3-5 In addition,
another subtype (R1L) has been proposed;6,7 this is
characterized by low affinity for prazosin, but efforts to
clone it are still unsuccessful. The R1-ARs play a central
role in the maintenance of smooth muscle tone in both
the cardiovascular system and the prostate; drugs
interacting with the R1-ARs are widely used in the
treatment of some diseases, including hypertension and
benign prostatic hyperplasia (BPH).8,9
Although in the past years pharmaceutical research
has focused the attention on the development of subtype-
selective ligands,10 comparable efforts have been made
to discover ligands selective for R1-ARs over other
related receptors, such as the serotoninergic 5-HT1A
(showing 45% of structural similarity) and the dopa-
minergic D1 and D2 receptors.11-13 Among the com-
pounds showing high affinity toward R1-ARs, relevant
attention has been devoted to molecules containing an
arylpiperazinyl moiety as the pharmacophoric por-
tion.12-16
In the course of our studies in the field of new and
selective R1-AR ligands, we have previously described
a class of pyrimido[5,4-b]indole-2,4-diones having an
arylpiperazinylalkyl chain at position 3 (Chart 1).17
Some of these compounds showed high affinity for R1-
ARs and good selectivity over the 5-HT1A receptors. In
particular RN5 (1), in which X is a methoxy group,
displayed the highest affinity, and compound 2, in which
X is a chlorine, showed a high level of selectivity. More
recently, some thienopyrimidine-2,4-dione derivatives
related to indole compounds have been described in a
patent and claimed as R1-AR ligands selective for R1D
with respect to R1A and R1B receptors.18
‡ This paper is dedicated to the memory of Dr. Emanuele Patane`.
* To whom correspondence should be addressed. Phone: +39 095
7384273. Fax: +39 095 222239. E-mail: vpittala@unict.it.
§ Universita` di Catania.
# Universita` degli Studi di Siena.
† Istituto di Ricerche Farmacologiche “Mario Negri”.
10.1021/jm040870h CCC: $30.25 © 2005 American Chemical Society
Published on Web 02/25/2005

Pyrimidine-2,4-dione Derivatives
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 7 2421
Scheme 1a
a Conditions: (a) ClCH2(CH2)nNCO, toluene, reflux; (b) 1-(substituted phenyl)piperazine, NaHCO3, NaI, THF, reflux; (c) KOH/MeOH,
reflux; (d) 1-(2-substituted phenyl)piperazine, 140 °C.
With the aim to obtain highly selective R1-AR ligands
over other related receptors, in this paper we describe
the synthesis and the binding data for R1-ARs and
5-HT1A, D1, D2 receptors of a series of new 3-arylpiper-
azinylalkylpyrrolo[3,2-d]pyrimidine-2,4-dione deriva-
tives (21-40) (Chart 1) in which an arylpyrrolo moiety
has replaced the indole nucleus in the structures of
compounds 1 and 2. In the new molecules, the aryl ring,
which is not condensed to form a planar tricyclic system,
is linked to the 6-position of the pyrrolo[3,2-d]pyrimi-
dine-2,4-dione (PPm) moiety and bears substituents of
different nature in the various positions. As in 1 and 2,
the PPm moiety is linked, through an alkyl chain, to a
phenylpiperazine (PPz) substructure bearing a methoxy
or a chloro substituent at the 2-position of the phenyl
ring.
New compounds were evaluated for their ability to
fit a pharmacophoric model for R1-AR antagonists
previously built by our research group and constituted
by five features representing three hydrophobic regions
(hereafter referred to as HY1-3 features), a hydrogen
bond acceptor (HBA), and a positively ionizable group
(PI).12 The model was able to predict affinity values of
6-unsubstituted (21-24), 6-phenyl (25-29), and 6-(2-
chlorophenyl) (30, 31) derivatives, in good agreement
with their corresponding experimental data. However,
the affinity of compounds 32-40, characterized by a
substituent at the meta or para position of the 6-phenyl
ring, was mispredicted. This was mainly due to the fact
that the pharmacophoric model was unable to account
for the influence of such a substituent on affinity. In
fact, substituents at the 3- or 4-position of the 6-phenyl
ring lie in an empty region of space, where any of the
pharmacophoric features was located. As a consequence
of this predictive limitation of the model, we have
pursued an alternative approach aimed at analyzing
how affinity values were correlated with structural
properties of the studied compounds. For this purpose,
we have applied the genetic function approximation
(GFA)19,20 to generate multiple QSAR equations able to
describe the affinity in terms of molecular descriptors.
As a result, affinity values were found to be dependent
on contributions from methylene groups, chlorine sub-
stituents, oxygen atoms as in the methoxy groups,
molecular refractivity, and positive and negative charge
surface areas.
Chemistry
The synthetic route to the pyrrolo[3,2-d]pyrimidine-
2,4-dione derivatives is described in Scheme 1. The
starting 3-amino-1H-pyrrole-2-carboxylic acid ethyl es-
ter 4 was prepared utilizing the procedure described by
Furneaux et al.21 Compounds 5-8 and 10 were synthe-
sized using substituted benzoyl acetonitriles as starting
material, as described by Chen et al.22-24 The same
synthetic method was used to obtain compound 9, which
is not reported in the literature.
Suitable pyrrolo amino esters (4-10) were reacted
with the appropriate chloroalkyl isocyanate in refluxing
toluene to give the (chloroalkyl)ureas (11-20). The
obtained ureas were then reacted with suitable 1-(2-
substituted phenyl)piperazine hydrochloride in THF in
the presence of NaI and NaHCO3 17 to afford intermedi-
ate ureas that, without isolation, were converted into
final products 21-40 in methanolic sodium hydroxide
(method A). Alternatively, some final derivatives were
prepared by reacting the appropriate (chloroalkyl)urea
derivative at 140 °C for 2 h with a large excess of the
suitable 1-(2-substituted phenyl)piperazine in the ab-
sence of solvent (method B).17
Results and Discussion
SAR Studies. The affinities and selectivities of
compounds 21-40 for R1-ARs and related receptors are
shown in Table 1. For the derivatives that showed the
highest affinity and selectivity toward R1-ARs over the
5-HT1A receptors, the affinity for dopaminergic D1 and
D2 receptors was also evaluated. The values are ex-
pressed as Ki (nM) values; RN5 (1) and 2 were selected
as reference substances.
Affinity data clearly showed that most of new syn-
thesized compounds were characterized by very inter-
esting binding profiles. In fact, with the exception of

2422 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 7
Patane` et al.
Table 1. Binding Properties of 3-Arylpiperazinylalkylpyrrolo[3,2-d]pyrimidine-2,4-dione Derivatives
Ki (nM)a
b
b
b
compd method
n
R
R1
2-CH3O 2.31 ( 0.22
2-Cl 0.94 ( 0.04
2-CH3O 28.01 ( 3.76 192.9 ( 30.9
2-Cl 59.78 ( 8.56 376.2 ( 68.8
2-CH3O 0.52 ( 0.05
2-Cl 1.08 ( 0.08
2-CH3O 1.23 ( 0.11
R1
5-HT1A
650.6 ( 109.1 ND
558.5 ( 124.1 >10000 322.1 ( 48.3
D1
D2
5-HT1A/R1
D1/R1
D2/R1
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
1c
A
A
A
A
B
B
A
A
B
B
A
B
B
B
B
B
A
A
A
A
1
1
2
2
1
1
2
2
1
1
1
1
1
1
1
1
2
2
1
1
H
ND
282
594
6.9
6.3
46
H
>10000 343
H
H
C6H5
C6H5
C6H5
C6H5
C6H5
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
24.01 ( 4.03
35.01 ( 7.05
22.36 ( 5.02
32
18
2-Cl
62.37 ( 3.99 129.8 ( 9.8
2.1
4-CH3O >10000
>10000
2-ClC6H4
2-ClC6H4
3-ClC6H4
4-ClC6H4
4-ClC6H4
4-CH3C6H4
4-CH3C6H4
4-CH3C6H4
4-CH3C6H4
2-CH3O 0.76 ( 0.09
16.14 ( 1.41
>10000
21
>20000
0.87
>1000
>800
>4500
>500
1.73
>90
682
>2500
238
>15000
2-Cl
0.46 ( 0.09
>10000 27.79 ( 3.89
ND ND
>20000 60
2-CH3O 19.32 ( 1.80 16.90 ( 3.02
2-CH3O 7.68 ( 1.63
2-Cl 11.70 ( 2.17 >10000
2-CH3O 2.05 ( 0.33 >10000
2-Cl 17.26 ( 4.14 >10000
>10000
>10000 3.64 ( 0.63
>10000 >10000
>1000
>800
>4500
>500
0.47
>800
9.7
>500
>10000 19.95 ( 1.51
>10000 >10000
2-CH3O 47.88 ( 2.80 82.99 ( 23.30 ND
2-Cl ND
ND
ND
110.1 ( 11.7 >10000
4-CH3OC6H4 2-CH3O 0.77 ( 0.10
4-CH3OC6H4 2-Cl
525.5 ( 94.1
>10000
50.00 ( 4.70
>10000
>10000 2.11 ( 0.35
>10000 2.23 ( 0.42
>10000 2.74
>2500 0.65
3.40 ( 0.54
0.21 ( 0.02
0.53 ( 0.18
2c
a Ki values were calculated as described in the Experimental Section and are the mean ( SD of three separate experiments. b Ratio
between Ki(5-HT1A)/Ki(R1), Ki(D1)/Ki(R1), and Ki(D2)/Ki(R1). c Data taken from ref 17.
compound 29, they exhibited good affinity toward R1-
ARs (some of them in the subnanomolar range) and
selectivity over the other tested receptors.
4-chloroderivative 34 and synthesized 3-chloro 32 and
2-chloro 30 and 31 substituted analogues. Compound
32 showed a slight decrease in affinity for R1-ARs with
a dramatic loss of selectivity, whereas compounds 30
and 31 showed an increase in affinity for R1-ARs (Ki )
0.76 and 0.46 nM, respectively). Moreover, 31 displays
an interesting selectivity profile because it showed no
affinity for the 5-HT1A and D1 receptors (5-HT1A/R1 and
D1/R1 greater than 20 000 nM) and only a moderate Ki
value at the D2 receptor (D2/R1 ) 60).
With the aim of investigating the role of the phenyl
ring linked to PPm, we synthesized compounds 21-24
in which this ring was removed. In binding assays these
compounds showed lower affinity for R1-ARs (compare
21 with 25, 23 with 27, and 24 with 28) with the
exception of 22 whose affinity and selectivity were
notable (Ki ) 0.94 nM, 5-HT1A/R1 ) 594, D1/R1 g 10 000,
D2/R1 ) 343).
Several R1-AR or 5-HT1A ligands reported in the
literature present a PPz moiety located at a different
distance from a second structural portion, generally
represented by a heterocyclic ring or system, containing
an amide group.13,25,26 The length of the alkyl chain
connecting these two substructures greatly influences
affinity and selectivity of such ligands. Compounds 23,
24, 27, 28, 37, and 38 were synthesized to evaluate the
influence of a propylene instead of an ethylene chain
on R1-AR affinity. Binding data clearly demonstrate that
the lengthening of this connecting chain determines a
general loss in both affinity and selectivity for R1-ARs.
The effects of the nature and the position of substit-
uents on the phenyl ring of the PPz moiety have been
extensively studied in many PPz-based R1-AR lig-
ands.12-16 As previously reported (and confirmed in this
Because our first aim was to verify whether the indole
moiety of pyrimido[5,4-b]indole-2,4-diones could be re-
placed by an arylpyrrole scaffold, first we synthesized
the 6-phenylpyrrolo[3,2-d]pyrimidine-2,4-dione deriva-
tives 25-28, more structurally related to 1 and 2.
Compounds 25 and 26 showed affinity toward R1-ARs
(Ki ) 0.52 and 1.08 nM, respectively) in the same range
as 1 and 2 showed (Ki ) 0.21 and 0.53 nM, respectively),
indicating that the pyrrole moiety is a good replacement
for the indole residue. These results encouraged us to
carry out additional modifications to improve both
affinity and selectivity. We then synthesized compounds
33-40 characterized by the presence of substituents,
such as chlorine, methyl, or methoxy, at the 4-position
of phenyl ring linked to PPm. As a general trend, these
substitutions afforded compounds showing a slight
decrease in affinity for R1-ARs with respect to com-
pounds 25 and 26 but, with exception of 37, with an
enhanced selectivity over the other receptors. In par-
ticular, compounds 35 and 39 showed the highest
affinity for R1-ARs (Ki ) 2.05 and 0.77 nM, respectively),
whereas compounds 34 and 36 were noteworthy for
their selectivity over all the other tested receptors (Ki
for 5-HT1A, D1, and D2 receptors greater than 10 000
nM). These results indicate that the presence of a
substituent at the 4-position of the phenyl ring linked
to PPm, although not essential for the affinity at R1-
ARs, can be critical to discriminate between R1-ARs and
all the other tested receptors. To investigate the effect
of the shift of the substituent on the phenyl ring from
the 4-position to the 3 and 2 positions, we selected

Pyrimidine-2,4-dione Derivatives
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 7 2423
series by the binding data of compound 29), a substitu-
ent in the 4-position of the aryl ring generally leads to
compounds with much lower affinity.17,27,28 On the other
hand, substituents in the 2-position usually give ligands
with higher affinity. Considering these premises and the
structures of compounds 1 and 2, we prepared and
evaluated 2-methoxy- or 2-chlorophenylpiperazinyl de-
rivatives. According to the trend previously encountered
in the pyrimido[5,4-b]indole-2,4-dione series,17 2-meth-
oxy substitution gave products with the highest affinity
on R1-ARs (25, 30, and 39) whereas 2-chloro substitution
gave products with lower affinity for R1-ARs but are
extremely selective because they are almost inactive on
the other tested receptors, particularly on 5-HT1A (22,
34, 36, and 40).
Molecular Modeling Calculations. 1. Superposi-
tion to a Pharmacophoric Model for R1-Adreno-
ceptor Antagonists. An additional effort to rationalize
the relationships between the structure of the new
compounds and their biological activity was performed
by analyzing the superposition pathways of such com-
pounds onto a pharmacophore model previously devel-
oped by us for R1-adrenoceptor antagonists.12 In prin-
ciple, with the exception of 29, all the compounds were
characterized by chemical features able to fill the five
features of the pharmacophore. Regarding their orienta-
tion with respect to the pharmacophore, the new
compounds showed a great similarity with RN5, which
constituted an element of the training set used to
generate the pharmacophore itself. In detail, the 2-meth-
oxy or 2-chlorophenylpiperazinyl moiety of the inhibitors
matched the HY1-HY2-PI system of the pharmaco-
phore, as depicted in Figure 1 for compounds 21, 25,
and 30, taken as representative examples of the whole
set of the new compounds. Moreover, one of the carbonyl
groups of the dione system corresponded to the hydrogen
bond acceptor group. In particular, the carbonyl group
at the 4-position of the PPm system was more frequently
found as the hydrogen bond acceptor. However, because
of a 180° rotation of the bicyclic moiety, the carbonyl
group at position 2 was sometimes allowed to interact
with HBA. Finally, different chemical substituents were
found to be able to fill the HY3 feature, allowing the
classification of the compounds into three subsets. (i)
The first subset was constituted by compounds lacking
any substituent at the pyrrole nucleus (i.e., 21-24), able
to fill HY3 with the CH-CH system at positions 6 and
7 of the pyrrole ring (Figure 1A). As a consequence of
the high fit between such compounds and the pharma-
cophore model, affinity values were predicted that were
in agreement with experimental data. As an example,
the pharmacophore model predicted an affinity of 2.0
nM for compound 21, comparable with the experimental
value of 2.3 nM. (ii) The second subset includes com-
pounds 25-28, bearing an unsubstituted phenyl group
at the 6-position of the PPm system, which was located
inside HY3 (Figure 1B). Compounds 25 and 26 showed
an enhanced fit with respect to the corresponding
unsubstituted counterparts 21 and 22, with a conse-
quent improvement of the predicted affinity. Compound
25 was predicted to have an affinity of 0.30 nM, in good
agreement with the experimental value of 0.52 nM. (iii)
Compounds with a 2-chlorophenyl group at the 6-posi-
tion of the PPm moiety (namely, 30 and 31) showed a
Figure 1. Superposition pattern of compounds 21, 25, and
30 into the pharmacophoric model for R1-AR antagonists.
Similar orientations of the o-methoxyphenylpiperazine systems
were found that were able to fill the HY1-2 features with the
substituted phenyl ring and the positive ionizable group PI
with the most basic nitrogen atom of the piperazine ring.
Moreover, one of the two carbonyl groups at the terminal
heterocyclic portion corresponded to the hydrogen bond ac-
ceptor group (HBA). In particular, HBA is represented by the
carbonyl at the 2-position of compound 21 (A), different from
compounds 25 (B) and 30 (C) where the same pharmacophoric
feature was the carbonyl at the 4-position. The major difference
was found in the region corresponding to HY3. In fact,
compound 21 partially matched the hydrophobic sphere with
the distal portion of the PPm fragment (that is, the C6-C7
sequence of the pirrole nucleus). Moreover, while the unsub-
stituted phenyl ring at the 6-position perfectly corresponded
to HY3 for compound 25, its 2-chloro derivative 30 fulfilled
the same feature with the halo substituent. Pharmacophoric
features are color coded as follows: blue for hydrophobics
(HY1-3), red for positive ionizable groups (PI), and green for
hydrogen bond acceptors (HBA).
conformational rearrangement with respect to the 6-
phenyl derivatives, allowing superposition of the chlo-
rine atom into the HY3 feature (Figure 1C). Also in this
case, the pharmacophore model predicted affinity values
well, for example, the predicted affinity of 30 being 0.75
nM, which is comparable with the experimental value
of 0.76 nM.
Binding data showed that elongation of the ethyl
spacer (linking the PPz and the PPm portions) to a

2424 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 7
Patane` et al.
propyl chain led to a marked decrease in affinity, also
accounted for by the pharmacophore model. In fact, the
ethyl chain seemed to be the optimal spacer to bring
the PPz system at the proper distance from the PPm
moiety to allow interactions with all the pharmacophoric
features. In contrast, lengthening of the spacer made
the superposition into the pharmacophore model more
difficult for the compounds. Consequently, only con-
formers at higher energy (i.e., with a stressed piperazine
ring or partially eclipsed Csp3-Csp3 systems) were
found to be able to partially fill the features of the
pharmacophore. Since the two variables (high energy
with respect to the minimum and a partial fit onto the
pharmacophore) were taken into account by the soft-
ware to predict affinity values, the last was calculated
to be 1-2 orders of magnitude lower with respect to the
values of the corresponding ethyl counterparts, in good
agreement with experimental data showing a drop in
affinity from an ethyl to a propyl spacer. As an example,
the affinity of compound 28 was predicted to be 20 nM
(versus an experimental value of 62 nM) with respect
to the affinity of 0.66 nM (versus an experimental value
of 1.1 nM) predicted for the corresponding ethyl deriva-
tive 26.
All the remaining compounds (32-40) bearing a
3-chlorophenyl or a 4-substituted phenyl ring at the
6-position of the PPm system could also be included in
the second subset, since their phenyl ring was usually
located inside the HY3 feature of the pharmacophore.
However, such compounds were characterized by an
additional group (at the 3- or 4-position of the phenyl
ring) that lies in an empty region of space where no
pharmacophoric feature was located. As a consequence,
the pharmacophore model was completely unable to
account for variations in affinity, if any, due to the
presence of such substituents. As an example, the
predicted activity for 33 was 0.45 nM, comparable to
the affinity predicted for compound 25 bearing an
unsubstituted phenyl ring.
2. QSAR Calculation by Means of the Genetic
Function Approximation (GFA). In an effort to
overcome the limitation of the pharmacophore model in
estimating the affinity of compounds such as 32-40 and
to produce improved theoretical models able to rational-
ize at the quantitative level the relationships between
affinity data and structural properties of the new
compounds, we have applied the genetic function ap-
proximation (GFA) algorithm19,20 as implemented in the
Cerius2 software.29 GFA has been largely applied in
recent years mainly because of its advantages over other
traditional multiple regression methods (i.e., it offers a
nonlinear approach to generate QSAR models using
different basis functions including spline terms and
produces multiple QSAR models rather than the usual
single model), and the robustness of this statistical
technique has been demonstrated by a large number of
literature reports also in the field of medicinal chemistry
(further details about GFA and its application can be
found on the Accelrys Web site, http://www.accelrys.com/
doc/life/cerius48/qsar/Output/stat_gfa_nf.html and in
recent literature).30-33
into a training set (14 compounds) and a test set (5
compounds), while their biological activity was used in
terms of -log Ki to have a (better) normal distribution
of biological data with respect to values expressed as
nanomolar concentrations. The pharmacophoric model
previously built for R1-AR antagonists was used as a
tool to perform molecular alignment, that is, to identify
(among all the conformers found) the best fitting
conformer for each molecule. In detail, each compound
was superposed to the pharmacophore by application
of the Best Fit option within the Compare/Fit menu of
Catalyst.29 As a result, the conformer chosen for each
compound was found to be oriented within the phar-
macophore in such a manner that the PPz moiety
matched HY1-HY2-PI and the oxygen atom of the
carbonyl group at the 4-position represented the HBA
feature. Finally, the PPm fragment and its phenyl
substituent at the 6-position were the molecular por-
tions filling HY3. Such an alignment was exported by
using the Export Workbench Alignments of Catalyst;
the partial atomic charges were calculated for each
compound by means of the Mopac software implemented
in the InsightII package,29 and charged compounds were
imported in Cerius2 for QSAR calculations. The original
set of descriptors (belonging to the classes of topological,
spatial, structural, thermodynamic, electronic, and elec-
trotopological state keys) calculated for each compound
was pruned in order to avoid the presence of collinear
regressors (descriptors) that could return inaccurate
results from a statistical point of view. For this purpose,
the correlation matrix was investigated to check the
linearity between variables (expressed as Pearson’s
correlation coefficient) and a cutoff of 0.7 was set to
discard molecular descriptors affected by collinearity.
As a consequence, the final set of descriptors used to
generate QSAR models was constituted by 21 variables
listed in Table 2.
The GFA algorithm is based on a genetic algorithm
to perform a search on the space of possible QSAR
equations and uses LOF (lack of fit) as the scoring
function to estimate the fitness of each model built. GFA
was applied to 21 descriptors, fixing to 100 the popula-
tion size (the number of QSAR models to be initially
generated by a random choice of descriptors). In the
models, only linear polynomial terms were allowed,
avoiding spline or quadratic terms, to better interpret
the final QSAR equations. A large number of crossover
operations (20 000) was chosen to generate progeny
models that were scored by the LOF parameter to
produce the 100 final QSAR models. As a result, all the
models constituting the final population possessed
roughly the same statistical parameters but each equa-
tion may be able to provide different insights to ratio-
nalize binding affinity data on the basis of each descrip-
tor participating in the equation itself. Table 3 reports
a summary of the eight best final GFA models scored
by increasing LOF. The best QSAR model generated
from 21 descriptors was chosen on the basis of its lowest
LOF value with respect to the other models character-
ized by comparable statistical parameters (in terms of
r2, internal CV r2,34 and F values). It was characterized
by an r2 of 0.920 and an LOF of 0.201. Analysis of the
best eight models allowed us to relate molecular de-
scriptors to the affinity values. In detail, affinity de-
From the original set of 20 molecules, compound 29,
with undefined affinity, was not included in GFA
calculations. The remaining 19 compounds were divided

Pyrimidine-2,4-dione Derivatives
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 7 2425
Table 2. Summary of the 21 Molecular Descriptors Used To Generate the QSAR Models
no.
descriptora
type
description
1
2
3
4
5
6
7
8
9
MR
Density
PMI-mag
Thermodynamic
Spatial
Molar refractivity
Molecular density
Principal moment of inertia
Dipole moment
Spatial
Dipole-mag Electronic
HOMO
LUMO
Sr
Electronic
Electronic
Electronic
Highest occupied molecular orbital energy
Lowest unoccupied molecular orbital energy
Superdelocalizability
Relative positive charge
Relative negative charge
Relative positive charge surface area
Relative negative charge surface area
Relative polar surface area
Jurs-RPCG Spatial
Jurs-RNCG Spatial
10 Jurs-RPCS
Spatial
11 Jurs-RNCS Spatial
12 Jurs-RPSA
13 S_sCH3
14 S_ssCH2
15 S_dssC
Spatial
Electrotopological Sum for a methyl group
Electrotopological Sum for a methylene group
Electrotopological Sum for a carbon atom with a double bond and two single bonds (i.e., a carbonyl carbon)
Electrotopological Sum for a carbon atom with a single bond and two aromatic bonds (i.e., a carbon of the
pyrrole ring)
16 S_aasC
17 S_ssNH
18 S_aaNH
Electrotopological Sum for a nitrogen atom linked to a hydrogen and involved in two additional single bonds
(i.e., the NH group of the pyrimidinedione system)
Electrotopological Sum for a nitrogen atom linked to a hydrogen and involved in two additional aromatic bonds
(i.e., the NH group of the pyrrole ring)
19 S_dO
20 S_ssO
21 S_sCl
Electrotopological Sum for an oxygen atom bound through a double bond (i.e., a carbonyl oxygen)
Electrotopological Sum for an oxygen atom bound through two single bonds (i.e., the oxygen of a methoxy group)
Electrotopological Sum for a chlorine atom
a Further details on molecular descriptors are available at the QSAR+ theory page of the Accelrys Web site http://www.accelrys.com/
doc/life/cerius481/qsar/Output/theory_d_nf.html.
Table 3. Overview of the Eight Best QSAR Models Obtained by Means of the GFA Algorithm on 21 Molecular Descriptors
Calculated for Compounds 21-40
no.
1
QSAR equation and statistical parametersa
Aff ) 21.0309 + 0.715137 HOMO - 1.53546 S_ssCH2 + 0.089448 MR - 0.00095 PMI-mag
r2 ) 0.920222, CV r2 ) 0.734, LOF ) 0.200777
Aff ) 30.9349 - 1.42254 S_ssCH2 - 0.00085 PMI-mag + 0.978227 HOMO
2
3
4
5
6
7
8
r2 ) 0.843812, LOF ) 0.221106
Aff ) 14.9317 - 1.28305 S_ssCH2 - 0.001135 PMI-mag + 0.590941 S_dO + 0.88154 HOMO
r2 ) 0.881520, LOF ) 0.298178
Aff ) -1.42385 + 0.218582 MR - 0.000713 PMI-mag - 1.41618 S_ssCH2 + 13.7639 Jurs-RPCG
r2 ) 0.881415, LOF ) 0.298444
Aff ) 35.7893 - 0.001058 PMI-mag + 1.20452 HOMO - 1.71175 S_ssCH2 + 0.430969 Jurs-RPCS
r2 ) 0.877877, LOF ) 0.307347
Aff ) 29.5356 - 0.033243 S_sCl - 1.4278 S_ssCH2 - 0.000758 PMI-mag + 0.880362 HOMO
r2 ) 0.870094, LOF ) 0.326934
Aff ) 18.8665 + 1.10796 Jurs-RNCS - 1.43661 S_ssCH2 + 0.093934 MR - 0.000938 PMI-mag + 0.664025 HOMO
r2 ) 0.941940, LOF ) 0.328768
Aff ) 30.219 - 0.000863 PMI-mag + 0.938424 HOMO - 1.40417 S_ssCH2 + 0.037602 S_ssO
r2 ) 0.869239, LOF ) 0.329087
a Aff represents log(1/Ki) where Ki is the affinity of compounds toward R1-ARs, expressed in nM. r2, CV r2, and LOF are the correlation
coefficient, the cross-validated correlation coefficient, and the lack-of-fit values, respectively. A brief description of molecular descriptors
is reported in Table 2.
pended on the electronic descriptor HOMO that should
roughly measure the nucleophilicity of a molecule. Thus,
molecules able to donate their electrons should be
characterized by enhanced affinity. In this context, it
is very important to note that compounds 22-40 are
all characterized by hydrogen bond acceptor systems
(one of them, the 4-carbonyl group, has been recognized
as a crucial element for interacting with the pharma-
cophore model) and a piperazine nitrogen atom able to
be a positive ionizable group. Moreover, eq 3 showed
that affinity was dependent on S_dO (dO represents an
oxygen atom bound through a double bond, that is, a
carbonyl oxygen), an electrotopological state key encod-
ing information about both the topological environment
of that atom and the electronic interactions due to all
other atoms in the molecule.35,36 The second term of the
best model was S_ssCH2 (describing a contribution from
a methylene group)35,36 that rationalized the presence
of an alkyl spacer. In particular, its negative coefficient
accounted for decreased affinity of compounds with a
propyl spacer with respect to the corresponding ethyl
derivatives (23 and 24 versus 21 and 22, 27 and 28
versus 25 and 26, 37 and 38 versus 35 and 36). In
addition, a molecular refractivity (MR) term, measuring
the volume and density of the compounds, was fre-
quently represented in QSAR equations found. Its
positive coefficient seemed to show that the binding site
for these compounds had a certain tolerance to the
bulkiness of the inhibitors, in agreement with the trend
of affinity showing that the o-methoxyphenylpiperazinyl
derivatives were more potent inhibitors with respect to
the corresponding o-chloro derivatives (with the excep-
tion of compound pairs 21 and 22 and compound pairs
30 and 31, where methoxy derivatives showed slightly
lower or comparable activity in comparison to chloro
analogues). This finding was also in agreement with the
variables S_sCl and S_ssO35,36 that appeared in eqs 6
and 8, respectively, indicating a negative contribution
to the activity from the chloro substituent but a positive
contribution from the methoxy group. From a quantita-

2426 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 7
Patane` et al.
Table 4. Calculated (Estimated or Predicted) and Actual
Affinity Data for Compounds 21-40
cant if its statistical parameters are better than those
derived from the scrambled sets of data. A summary of
the results obtained through the randomization test
from 19 trials follows; r from the nonrandom model was
0.959, markedly better than the values obtained from
randomized data (mean value of r from random trials
was 0.625, with a standard deviation of 0.202). All the
scrambled sets produced models with r lower than 0.959
at the 95% confidence level. (iii) A leave-one-out cross-
validation encompassing both the optimization of the
choice of descriptors and the optimization of regression
coefficients was also performed on the best QSAR model.
Statistical parameters for the cross-validated model
were CV r2 of 0.835, PRESS of 1.066, and a sum of
squares deviation of 6.471.
affinity toward R1-adrenoreceptora
GFA residuals
compd
actual
calculatedb
activityc
21
22
23
24
25
26
27
28d
29e
30
31d
32d
33
34
35
36d
37
38d
39
40
1d
8.64
9.03
7.55
7.22
9.28
8.97
8.91
7.21
<4
9.12
9.34
7.71
8.11
7.93
8.69
7.76
7.32
6.96
9.11
8.47
9.68
9.28
7.89
9.06
8.79
7.41
7.19
9.11
8.96
8.82
7.22
-0.42
0.24
0.14
0.03
0.17
0.01
0.09
-0.01
8.91
9.35
8.40
8.08
7.91
8.67
8.76
7.58
6.87
9.07
8.80
9.26
8.97
7.50
0.21
-0.01
-0.69
0.03
0.02
0.02
-1.00
-0.26
0.09
0.04
-0.33
0.42
0.31
0.39
Conclusions
A new series of 3-arylpiperazinylalkylpyrrolo[3,2-d]-
pyrimidine-2,4-diones, structurally related to RN5 and
characterized by remarkable R1-AR affinity coupled to
high selectivity over 5-HT1A, D1, and D2 receptors, has
been discovered.
2d
A structure-affinity relationships analysis revealed
some structural features that, in this class of com-
pounds, are important to ensure high affinity and
selectivity. Replacement of a methoxy group with a
chlorine atom in the 2-position of the phenyl ring in the
PPz moiety gave compounds (22, 31, 34, 36, 40) with
the best 5-HT1A/R1 selectivity ratio. Among them, com-
pounds 34 and 36 are noteworthy for their R1-AR
selectivity over all the other tested receptors. An eth-
ylene alkyl chain connecting the PPz to the PPm moiety
seems to be optimum for R1-AR affinity. The length of
this spacer is critical because its elongation by addition
of a methylene unit invariably led to a remarkable loss
in both R1-AR affinity and selectivity. With reference
to structural modifications on the PPm moiety, most
compounds bearing a substituent at the 4- or 2-position
of the phenyl ring (31, 33-36, 39, 40) are endowed with
higher selectivity for R1-ARs with respect to the corre-
sponding unsubstituted analogues 25 and 26. However,
a notable R1-AR selectivity is also displayed by deriva-
tive 22, which lacks the phenyl ring at the 6-position of
the PPm portion. Thus, the pattern of substitutions of
the PPm moiety seems to play a significant role in
determining R1-AR selectivity over the other tested
receptors.
Moreover, in an effort to correlate R1-AR affinity of
the new compounds with their three-dimensional struc-
ture, two different molecular modeling approaches were
applied. In the first attempt, analysis of how well such
compounds fitted a pharmacophoric model for R1-AR
antagonists allowed us to rationalize affinity data for
compounds bearing a hydrogen atom or a phenyl ring
or a 2-chlorophenyl moiety at the 6-position of the PPm
moiety. In fact, predicted affinities were in good agree-
ment with experimental data. However, the pharma-
cophoric model was unable to make satisfactory predic-
tions for compounds bearing a 3- or 4-substituted phenyl
ring. To get over this limitation, a GFA-based QSAR
analysis was performed and several equations were
generated. They described affinity as dependent on the
length of the alkyl spacer and on the presence of oxygen
atoms able to be hydrogen bond acceptors, in agreement
3d
a Expressed as log(1/Ki) where Ki is the affinity of compounds
toward R1-ARs in nM. b Estimated (training set) or predicted (test
set) affinity values using eq 1 in Table 3. c Difference between
actual and calculated affinity values. d Compounds belonging to
the test set. With the exception of 29 (see below), the remaining
compounds were included in the training set. e Compound not
considered in the GFA QSAR analysis.
tive point of view, the positive effect of MR on affinity
was counterbalanced by the opposite effect due to the
S_ssCH2 variable. Next, the principal moment of inertia
(PMI, a spatial descriptor) accounted for the fact that
orientation of the molecules and distribution of their
chemical groups were important for affinity. Finally, two
additional spatial descriptors (relative positive charge
surface, RPCS, and relative negative charge surface,
RNCS, belonging to the Jurs family of descriptors,
calculated by mapping atomic partial charges on solvent-
accessible surface areas of each atom) participated in
eqs 5 and 7, respectively, to describe affinity on the basis
of both shape and electronic information characterizing
the molecules.
3. Validation of the QSAR Models. With the aim
of validating the QSAR analysis, three different ap-
proaches have been chosen. (i) Eight compounds (namely,
1-3,17 28, 31, 32, 36, and 38) were used to generate a
test set to be submitted to affinity prediction through
the best QSAR model (model no.1 in Table 3). Results
of affinity prediction was summarized in Table 4. In
detail, the QSAR equation produces predicted affinity
values for compounds 28, 31, and 38 (7.22, 9.35, and
6.87, respectively) in very good agreement with experi-
mental data (7.21, 9.34, and 6.96, respectively). On the
other hand, poorer predictions were found for 1-3, 32,
and 36 (9.26 vs 9.68, 8.97 vs 9.28, 7.50 vs 7.89, 8.40 vs
7.71, and 8.76 vs 7.76, respectively). (ii) To perform a
randomization test, affinity values were repeatedly
scrambled to generate 19 new spreadsheets containing
both the dependent variable (scrambled affinity) and
independent variables (molecular descriptors). The origi-
nal QSAR model can be considered statistically signifi-

Pyrimidine-2,4-dione Derivatives
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 7 2427
General Procedure for the Preparation of N-Chloro-
alkyl-N′-[3-(2-ethoxycarbonyl-5-substituted)-1H-pyrrolyl]-
ureas (11-20). A suspension of the appropriate 3-amino-5-
substituted-1H-pyrrole-2-carboxylic acid ethyl ester (20 mmol)
and the appropriate chloroalkyl isocyanate (30 mmol) in 40
mL of toluene was stirred for 16 h at 22 °C. The reaction
mixture was then filtered, and the obtained solid was washed
several times with EtOH to give the desired product as a pure
solid.
with what was found by means of the pharmacophore
model. In addition, QSAR equations correlated affinity
with the principal moment of inertia and molar refrac-
tivity, two molecular properties that a pharmacophoric-
based study is unable to evaluate. Finally, results from
the QSAR validation steps suggested the GFA model
as a predictive tool to evaluate the R1-AR affinity of new
compounds.
In conclusion, within this new class of R1-AR ligands,
compound 31 emerges over the other described deriva-
tives because of its subnanomolar R1-AR affinity and
high selectivity over the serotoninergic 5-HT1A and
dopaminergic D1 and D2 receptors. An outstanding R1-
AR selectivity also characterizes the binding profiles of
compounds 34 and 36. These two derivatives, along with
31, have been selected and an additional study is in
progress in order to investigate their activities in
functional assays and their affinities for the R1-AR
subtypes.
By use of this procedure, the subsequent compounds were
obtained.
N-(2-Chloroethyl)-N′-[3-(2-ethoxycarbonyl)-1H-pyrrol-
yl]urea (11). The title compound was obtained as a white
powder (90%): mp 177-179 °C; 1H NMR (DMSO-d6) δ 1.30
(t, J ) 7.0 Hz, 3 H, OCH2CH3), 3.40-3.43 (m, 2 H, NHCH2),
3.63 (t, J ) 6.0 Hz, 2 H, CH2Cl), 4.27 (q, J ) 7.0 Hz, 2 H,
OCH2CH3), 6.65-6.68 (m, 1 H, pyrrole), 6.80-6.83 (m, 1 H,
pyrrole), 7.61 (br t, J ) 5.4 Hz, 1 H, NHCONHCH2 which
exchanges with D2O), 8.43 (br s, 1 H, NHCONHCH2 which
exchanges with D2O), 11.24 (br s, 1 H, NH which exchanges
with D2O). Anal. (C10H14ClN3O3) C, H, N.
N-(3-Chloropropyl)-N′-[3-(2-ethoxycarbonyl)-1H-pyr-
rolyl]urea (12). The title compound was obtained as a white
Experimental Section
1
powder (80%): mp 78-80 °C; H NMR (DMSO-d6) δ 1.29 (t,
Chemistry. Melting points were determined in a Gallen-
kamp apparatus with an MFB-59 digital thermometer in glass
capillary tubes and are uncorrected. Elemental analyses for
C, H, N, and S were within (0.4% of theoretical values and
were performed on a Carlo Erba elemental analyzer model
J ) 7.0 Hz, 3 H, OCH2CH3), 1.83-1.89 (m, 2 H, CH2CH2CH2),
3.13-3.25 (m, 2 H, NHCH2), 3.67 (t, J ) 6.4 Hz, 2 H, CH2Cl),
4.26 (q, J ) 7.0 Hz, 2 H, OCH2CH3), 6.65-6.68 (m, 1 H,
pyrrole), 6.79-6.82 (m, 1 H, pyrrole), 7.33 (br t, J ) 5.4 Hz,
1 H, NHCONHCH2 which exchanges with D2O), 8.34 (br s,
1 H, NHCONHCH2 which exchanges with D2O), 11.20 (br s,
1 H, NH which exchanges with D2O). Anal. (C11H16ClN3O3)
C, H, N.
1
1108 apparatus. H NMR spectra were recorded at 200 MHz
on a Varian Inova Unity 200 spectrometer in DMSO-d6
solution. Chemical shifts are given in δ values (ppm), using
tetramethylsilane as the internal standard; coupling constants
(J) are given in hertz (Hz). Signal multiplicities are character-
ized as s (singlet), d (doublet), t (triplet), q (quartet), m
(multiplet), br (broad signal). All the synthesized compounds
were tested for purity on TLC (aluminum sheet coated with
silica gel 60 F254, Merck) and visualized by UV (λ ) 254 and
366 nm). All chemicals and solvents were reagent grade and
were purchased from commercial vendors. Compounds 4-8
and 10 were synthesized as previously reported.21-24
N-(2-Chloroethyl)-N′-[3-(2-ethoxycarbonyl-5-phenyl)-
1H-pyrrolyl]urea (13). The title compound was obtained as
1
a white powder (79%): mp 227-228 °C; H NMR (DMSO-d6)
δ 1.34 (t, J ) 7.0 Hz, 3 H, OCH2CH3), 3.35-3.43 (m, 2 H,
NHCH2), 3.66 (t, J ) 6.0 Hz, 2 H, CH2Cl), 4.32 (q, J ) 7.0 Hz,
2 H, OCH2CH3), 7.12 (d, 4J ) 2.8 Hz, 1 H, pyrrole), 7.29-7.44
(m, 3 H, aromatic), 7.69 (br t, J ) 5.6 Hz, 1 H, NHCONHCH2
which exchanges with D2O), 7.77-7.81 (m, 2 H, aromatic), 8.52
(br s, 1 H, NHCONHCH2 which exchanges with D2O), 11.42
(br s, 1 H, NH which exchanges with D2O). Anal. (C16H18-
ClN3O3) C, H, N.
3-Amino-5-(4-methylphenyl)-1H-pyrrole-2-carboxylic
Acid Ethyl Ester (9). To a solution of 4-methylbenzoyl
acetonitrile (3.90 g, 24.54 mmol) and p-toluenesulfonic anhy-
dride (9.60 g, 29.45 mmol) in 50 mL of dry CH2Cl2 was added,
dropwise, 5.10 mL of Et3N (36.81 mmol). After 16 h of stirring
at 22 °C, the reaction mixture was diluted with H2O and
extracted with CH2Cl2 (4 × 50 mL). The combined organic
layers were dried over Na2SO4, filtered, and concentrated in
vacuo to give an orange solid. The obtained crude material
was recrystallized from EtOH to give 4.80 g of 2-cyano-1-(4-
methylphenyl)vinyl 4-methylbenzenesulfonate as a white solid
N-(3-Chloropropyl)-N′-[3-(2-ethoxycarbonyl-5-phenyl)-
1H-pyrrolyl]urea (14). The title compound was obtained as
1
a white powder (76%): mp 190-191 °C; H NMR (DMSO-d6)
δ 1.34 (t, J ) 7.0 Hz, 3 H, OCH2CH3), 1.86-1.93 (m, 2 H,
CH2CH2CH2), 3.17-3.26 (m, 2 H, NHCH2), 3.70 (t, J ) 6.4
Hz, 2 H, CH2Cl), 4.32 (q, J ) 7.0 Hz, 2 H, OCH2CH3), 7.13 (d,
4J ) 2.8 Hz, 1 H, pyrrole), 7.26-7.44 (m, 3 H + 1 H, aromatic
+ NHCONHCH2 which exchanges with D2O), 7.77-7.81 (m,
2 H, aromatic), 8.44 (br s, 1 H, NHCONHCH2 which exchanges
with D2O), 11.39 (br s, 1 H, NH which exchanges with D2O).
Anal. (C17H20ClN3O3) C, H, N.
1
(62%): mp 125-127 °C; H NMR (DMSO-d6) δ 2.33 (s, 3 H,
CH3), 2.42 (s, 3 H, CH3), 6.55 (s, 1 H, CH), 7.19-7.28 (m, 2 H,
aromatic), 7.44-7.51 (m, 4 H, aromatic), 7.78-7.89 (m, 2 H,
aromatic). Anal. (C17H15NO3S) C, H, N, S.
N-(2-Chloroethyl)-N′-[3-[2-ethoxycarbonyl-5-(2-chlo-
rophenyl)-1H-pyrrolyl]]urea (15). The title compound was
obtained as a white powder (85%): mp 156-158 °C; 1H NMR
(DMSO-d6) δ 1.32 (t, J ) 7.0 Hz, 3 H, OCH2CH3), 3.38-3.44
(m, 2 H, NHCH2), 3.65 (t, J ) 6.0 Hz, 2 H, CH2Cl), 4.30 (q,
J ) 7.0 Hz, 2 H, OCH2CH3), 7.08 (d, 4J ) 2.8 Hz, 1 H, pyrrole),
7.33-7.44 (m, 2 H, aromatic), 7.52-7.62 (m, 2 H, aromatic),
7.69 (br t, J ) 5.6 Hz, 1 H, NHCONHCH2 which exchanges
with D2O), 8.51 (br s, 1 H, NHCONHCH2 which exchanges
with D2O), 11.48 (br s, 1 H, NH which exchanges with D2O).
Anal. (C16H17Cl2N3O3) C, H, N.
N-(2-Chloroethyl)-N′-[3-[2-ethoxycarbonyl-5-(3-chlo-
rophenyl)-1H-pyrrolyl]]urea (16). The title compound was
obtained as a white powder (82%): mp 170-172 °C; 1H NMR
(DMSO-d6) δ 1.35 (t, J ) 7.0 Hz, 3 H, OCH2CH3), 3.40-3.46
(m, 2 H, NHCH2), 3.66 (t, J ) 6.0 Hz, 2 H, CH2Cl), 4.33 (q,
J ) 7.0 Hz, 2 H, OCH2CH3), 7.18 (d, 4J ) 2.8 Hz, 1 H, pyrrole),
7.31-7.47 (m, 2 H, aromatic), 7.67-7.75 (m, 1 H + 1 H,
aromatic + NHCONHCH2 which exchanges with D2O), 7.97
(m, 1 H, aromatic), 8.53 (br s, 1 H, NHCONHCH2 which
To a solution of sodium ethoxide freshly prepared from Na
(1.03 g, 44.64 mmol) and absolute EtOH (28 mL) under
nitrogen atmosphere was slowly added a suspension of 2-cy-
ano-1-(4-methylphenyl)vinyl 4-methylbenzenesulfonate (4.66
g, 14.88 mmol) and diethyl aminomalonate hydrochloride (3.15
g, 14.88 mmol) in 60 mL of EtOH and 6 mL of THF. After the
addition was completed, the reaction mixture was stirred at
22 °C for 2 h and concentrated under reduced pressure to give
an orange solid. The obtained crude material was diluted with
50 mL of water and extracted with EtOAc (3 × 50 mL). The
combined organic layers were dried over Na2SO4 and concen-
trated under reduced pressure. The obtained orange solid was
recrystallized from EtOH to give 1.27 g of a white powder
(35%): mp 167-169 °C; 1H NMR (DMSO-d6) δ 1.30 (t, J ) 7.0
Hz, 3 H, OCH2CH3), 2.30 (s, 3 H, CH3), 4.23 (q, J ) 7.0 Hz, 2
H, OCH2CH3), 5.11 (br s, 2 H, NH2 which exchanges with D2O),
6.01 (d, 4J ) 2.8 Hz, 1 H, pyrrole), 7.21-7.28 (m, 2 H,
aromatic), 7.74-7.78 (m, 2 H, aromatic), 10.78 (br s, 1 H, NH
which exchanges with D2O). Anal. (C14H16N2O2) C, H, N.

2428 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 7
Patane` et al.
exchanges with D2O), 11.56 (br s, 1 H, NH which exchanges
with D2O). Anal. (C16H17Cl2N3O3) C, H, N.
exchanges with D2O), 11.91 (br s, 1 H, NH which exchanges
with D2O). Anal. (C19H23N5O3) C, H, N.
N-(2-Chloroethyl)-N′-[3-[2-ethoxycarbonyl-5-(4-chlo-
rophenyl)-1H-pyrrolyl]]urea (17). The title compound was
obtained as a white powder (70%): mp 248-249 °C; 1H NMR
(DMSO-d6) δ 1.34 (t, J ) 7.0 Hz, 3 H, OCH2CH3), 3.36-3.45
(m, 2 H, NHCH2), 3.66 (t, J ) 5.8 Hz, 2 H, CH2Cl), 4.32 (q,
J ) 7.0 Hz, 2 H, OCH2CH3), 7.14 (d, 4J ) 2.8 Hz, 1 H, pyrrole),
7.43-7.48 (m, 2 H, aromatic), 7.69 (br t, J ) 5.4 Hz, 1 H,
NHCONHCH2 which exchanges with D2O), 7.80-7.84 (m,
2 H, aromatic), 8.52 (br s, 1 H, NHCONHCH2 which exchanges
with D2O), 11.50 (br s, 1 H, NH which exchanges with D2O).
Anal. (C16H17Cl2N3O3) C, H, N.
N-(2-Chloroethyl)-N′-[3-[2-ethoxycarbonyl-5-(4-meth-
ylphenyl)-1H-pyrrolyl]]-urea (18). The title compound was
obtained as a white powder (89%): mp 180-182 °C; 1H NMR
(DMSO-d6) δ 1.34 (t, J ) 7.0 Hz, 3 H, OCH2CH3), 2.31 (s, 3 H,
CH3), 3.40-3.45 (m, 2 H, NHCH2), 3.66 (t, J ) 6.0 Hz, 2 H,
CH2Cl), 4.31 (q, J ) 7.0 Hz, 2 H, OCH2CH3), 7.08 (d, 4J ) 2.8
Hz, 1 H, pyrrole), 7.19-7.23 (m, 2 H, aromatic), 7.67-7.71 (m,
2 H + 1 H, aromatic + NHCONHCH2 which exchanges with
D2O), 8.52 (br s, 1 H, NHCONHCH2 which exchanges with
D2O), 11.37 (br s, 1 H, NH which exchanges with D2O). Anal.
(C17H20ClN3O3) C, H, N.
N-(3-Chloropropyl)-N′-[3-[2-ethoxycarbonyl-5-(4-meth-
ylphenyl)-1H-pyrrolyl]]urea (19). The title compound was
obtained as a white powder (61%): mp 212-214 °C; 1H NMR
(DMSO-d6) δ 1.33 (t, J ) 7.0 Hz, 3 H, OCH2CH3), 1.85-1.92
(m, 2 H, CH2CH2CH2), 2.31 (s, 3 H, CH3), 3.16-3.25 (m, 2 H,
NHCH2), 3.69 (t, J ) 6.4 Hz, 2 H, CH2Cl), 4.30 (q, J ) 7.0 Hz,
2 H, OCH2CH3), 7.07 (d, 4J ) 2.6 Hz, 1 H, pyrrole), 7.18-7.22
(m, 2 H, aromatic), 7.40 (br t, J ) 5.6 Hz, 1 H, NHCONHCH2
which exchanges with D2O), 7.66-7.70 (m, 2 H, aromatic), 8.44
(br s, 1 H, NHCONHCH2 which exchanges with D2O), 11.31
(br s, 1 H, NH which exchanges with D2O). Anal. (C18H22-
ClN3O3) C, H, N.
N-(2-Chloroethyl)-N′-{3-[2-ethoxycarbonyl-5-(4-meth-
oxyphenyl)-1H-pyrrolyl]}urea (20). The title compound was
obtained as a white powder (80%): mp 178-180 °C; 1H NMR
(DMSO-d6) δ 1.33 (t, J ) 7.0 Hz, 3 H, OCH2CH3), 3.40-3.45
(m, 2 H, NHCH2), 3.66 (t, J ) 6.0 Hz, 2 H, CH2Cl), 3.78 (s, 3
H, OCH3), 4.31 (q, J ) 7.0 Hz, 2 H, OCH2CH3), 6.94-7.03 (m,
1 H + 2 H, pyrrole + aromatic), 7.65-7.75 (m, 2 H + 1 H,
aromatic + NHCONHCH2 which exchanges with D2O), 8.52
(br s, 1 H, NHCONHCH2 which exchanges with D2O), 11.27
(br s, 1 H, NH which exchanges with D2O). Anal. (C17H20-
ClN3O4) C, H, N.
Method A: General Procedure for the Preparation of
3-[[4-(Substituted phenyl)piperazin-1-yl]alkyl]-6-substi-
tuted-1,5-dihydropyrrolo[3,2-d]pyrimidine-2,4-diones (21-
24, 27, 28, 31, 37-40). A suspension of the appropriate (ω-
chloroalkyl)urea (0.50 mmol), the appropriate 1-(substituted
phenyl)piperazine hydrochloride (0.75 mmol), NaHCO3 (2.0
mmol), and a catalytic amount of NaI in 2 mL of THF was
refluxed for 4 days. After this period the obtained mixture was
concentrated and the residue was diluted with water (50 mL)
and extracted with CH2Cl2 (3 × 50 mL). The combined organic
layers were dried over Na2SO4, filtered, and evaporated. The
obtained crude material was refluxed in methanolic KOH (1
N, 1.5 mL) for 2 h. The reaction mixture was then allowed to
cool at 22 °C, acidified (pH 4) with glacial AcOH, and
neutralized with saturated NaHCO3. A brown precipitate was
isolated, washed with water and EtOH, and finally recrystal-
lized from DMF/water (2:1) to afford the desired products as
solids.
3-[2-[4-(2-Methoxyphenyl)piperazin-1-yl]ethyl]-1,5-di-
hydropyrrolo[3,2-d]pyrimidine-2,4-dione (21). The title
compound was isolated as a white powder (79%): mp 308-
309 °C; 1H NMR (DMSO-d6) δ 2.35-2.79 (m, 4 H + 2 H,
NCH2 + CONCH2CH2N), 2.85-3.09 (m, 4 H, NCH2), 3.74 (s,
3 H, CH3), 4.02 (t, J ) 6.6 Hz, 2 H, CONCH2CH2N), 5.85 (d,
J ) 2.6 Hz, 1 H, pyrrole), 6.79-6.97 (m, 4 H, aromatic), 7.28
(d, 1 H, J ) 2.6 Hz, pyrrole), 11.09 (br s, 1 H, NH which
3-[2-[4-(2-Chlorophenyl)piperazin-1-yl]ethyl]-1,5-dihy-
dropyrrolo[3,2-d]pyrimidine-2,4-dione (22). The title com-
pound was isolated as a white powder (85%): mp 285-287
1
°C; H NMR (DMSO-d6) δ 2.49-2.77 (m, 4 H + 2 H, NCH2 +
CONCH2CH2N), 2.85-3.09 (m, 4 H, NCH2), 4.01 (t, J ) 6.6
Hz, 2 H, CONCH2CH2N), 5.85 (d, J ) 2.6 Hz, 1 H, pyrrole),
6.98-7.42 (m, 4 H + 1 H, aromatic + pyrrole), 11.11 (br s, 1
H, NH which exchanges with D2O), 11.92 (br s, 1 H, NH which
exchanges with D2O). Anal. (C18H20ClN5O2) C, H, N.
3-[3-[4-(2-Methoxyphenyl)piperazin-1-yl]propyl]-1,5-
dihydropyrrolo[3,2-d]pyrimidine-2,4-dione (23). The title
compound was isolated as a white powder (70%): mp 225-
227 °C; 1H NMR (DMSO-d6) δ 1.63-1.77 (m, 2 H, CONCH2CH2-
CH2N), 2.34-2.59 (m, 4 H + 2 H, NCH2 + CONCH2CH2CH2N),
2.86-2.99 (m, 4 H, NCH2), 3.76 (s, 3 H, CH3), 3.91 (t, J ) 6.6
Hz, 2 H, CONCH2CH2CH2N), 5.84 (d, J ) 2.6 Hz, 1 H, pyrrole),
6.83-6.94 (m, 4 H, aromatic), 7.15 (d, 1 H, J ) 2.6 Hz, pyrrole),
11.12 (br s, 1 H, NH which exchanges with D2O), 11.95 (br s,
1 H, NH which exchanges with D2O). Anal. (C20H25N5O3) C,
H, N.
3-[3-[4-(2-Chlorophenyl)piperazin-1-yl]propyl]-1,5-di-
hydropyrrolo[3,2-d]pyrimidine-2,4-dione (24). The title
compound was isolated as a white powder (89%): mp 223-
225 °C; 1H NMR (DMSO-d6) δ 1.71-1.77 (m, 2 H, CH2CH2-
CH2), 2.34-2.60 (m, 4 H + 2 H, NCH2 + CONCH2CH2CH2N),
2.82-2.97 (m, 4 H, NCH2), 3.92 (t, J ) 6.6 Hz, 2 H, CONCH2-
CH2CH2N), 5.86 (d, J ) 2.6 Hz, 1 H, pyrrole), 6.80-7.41 (m, 4
H + 1 H, aromatic + pyrrole), 11.09 (br s, 1 H, NH which
exchanges with D2O), 11.88 (br s, 1 H, NH which exchanges
with D2O). Anal. (C19H22ClN5O2) C, H, N.
3-[3-[4-(2-Methoxyphenyl)piperazin-1-yl]propyl]-6-phen-
yl-1,5-dihydropyrrolo[3,2-d]pyrimidine-2,4-dione (27). The
title compound was isolated as a white powder (69%): mp
324-325 °C; 1H NMR (DMSO-d6) δ 1.65-1.79 (m, 2 H,
CONCH2CH2CH2N), 2.32-2.58 (m, 4 H + 2 H, NCH2
+
CONCH2CH2CH2N), 2.82-2.94 (m, 4 H, NCH2), 3.76 (s, 3 H,
CH3), 3.95 (t, J ) 6.6 Hz, 2 H, CONCH2CH2CH2N), 6.31 (s, 1
H, pyrrole), 6.80-6.93 (m, 4 H, aromatic), 7.29-7.46 (m, 3 H,
aromatic), 7.86-7.90 (m, 2 H, aromatic), 11.15 (br s, 1 H, NH
which exchanges with D2O), 12.24 (br s, 1 H, NH which
exchanges with D2O). Anal. (C26H29N5O3) C, H, N.
3-[3-[4-(2-Chlorophenyl)piperazin-1-yl]propyl]-6-phen-
yl-1,5-dihydropyrrolo[3,2-d]pyrimidine-2,4-dione (28). The
title compound was isolated as a white powder (54%): mp
278-279 °C; 1H NMR (DMSO-d6) δ 1.66-1.80 (m, 2 H,
CONCH2CH2CH2N), 2.34-2.61 (m, 4 H + 2 H, NCH2
+
CONCH2CH2CH2N), 2.81-2.95 (m, 4 H, NCH2), 3.96 (t, J )
6.6 Hz, 2 H, CONCH2CH2CH2N), 6.29 (s, 1 H, pyrrole), 6.31-
7.46 (m, 7 H, aromatic), 7.86-7.89 (m, 2 H, aromatic), 11.16
(br s, 1 H, NH which exchanges with D2O), 12.20 (br s, 1 H,
NH which exchanges with D2O). Anal. (C25H26ClN5O2) C, H,
N.
3-[2-[4-(2-Chlorophenyl)piperazin-1-yl]ethyl]-6-(2-chlo-
rophenyl)-1,5-dihydropyrrolo[3,2-d]pyrimidine-2,4-di-
one (31). The title compound was isolated as a light-pink
powder (51%): mp 280-281 °C; 1H NMR (DMSO-d6) δ 2.44-
2.82 (m, 4 H + 2 H, NCH2 + CONCH2CH2N), 2.87-3.16 (m,
4 H, NCH2), 4.07 (t, J ) 6.6 Hz, 2 H, CONCH2CH2N), 6.25 (s,
1 H, pyrrole), 6.99-7.69 (m, 8 H, aromatic), 11.17 (br s, 1 H,
NH which exchanges with D2O), 12.25 (br s, 1 H, NH which
exchanges with D2O). Anal. (C24H23Cl2N5O2) C, H, N.
3-[3-[4-(2-Methoxyphenyl)piperazin-1-yl]propyl]-6-(4-
methylphenyl)-1,5-dihydropyrrolo[3,2-d]pyrimidine-2,4-
dione (37). The title compound was isolated as a white powder
(50%): mp 272-273 °C; 1H NMR (DMSO-d6) δ 1.65-1.91 (m,
2 H, CONCH2CH2CH2N), 2.28-2.59 (m, 4 H + 2 H + 3 H,
NCH2 + CONCH2CH2CH2N, CH3), 2.83-2.92 (m, 4 H, NCH2),
3.75 (s, 3 H, OCH3), 3.94 (t, J ) 6.6 Hz, 2 H, CONCH2CH2-
CH2N), 6.24 (s, 1 H, pyrrole), 6.77-6.90 (m, 4 H, aromatic),
7.20-7.24 (m, 2 H, aromatic), 7.73-7.77 (m, 2 H, aromatic),

Pyrimidine-2,4-dione Derivatives
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 7 2429
11.13 (br s, 1 H, NH which exchanges with D2O), 12.14 (br s,
1 H, NH which exchanges with D2O). Anal. (C27H31N5O3) C,
H, N.
which exchanges with D2O), 12.23 (br s, 1 H, NH which
exchanges with D2O). Anal. (C25H27N5O3) C, H, N.
6-(2-Chlorophenyl)-3-[2-[4-(2-methoxyphenyl)piperazin-
1-yl]ethyl]-1,5-dihydropyrrolo[3,2-d]pyrimidine-2,4-di-
one (30). The title compound was isolated as a white powder
(43%): mp 291-292 °C; 1H NMR (DMSO-d6) δ 2.45-2.76 (m,
4 H + 2 H, NCH2 + CONCH2CH2N), 2.85-3.10 (m, 4 H,
NCH2), 3.77 (s, 3 H, CH3), 4.05 (t, J ) 6.6 Hz, 2 H, CONCH2-
CH2N), 6.25 (s, 1 H, pyrrole), 6.82-7.00 (m, 4 H, aromatic),
7.39-7.44 (m, 2 H, aromatic), 7.56-7.69 (m, 2 H, aromatic),
11.17 (br s, 1 H, NH which exchanges with D2O), 12.23 (br s,
1 H, NH which exchanges with D2O). Anal. (C25H26ClN5O3) C,
H, N.
6-(3-Chlorophenyl)-3-[2-[4-(2-methoxyphenyl)piperazin-
1-yl]ethyl]-1,5-dihydropyrrolo[3,2-d]pyrimidine-2,4-di-
one (32). The title compound was isolated as a white powder
(39%): mp 297-298 °C; 1H NMR (DMSO-d6) δ 2.45-2.76 (m,
4 H + 2 H, NCH2 + CONCH2CH2N), 2.85-3.00 (m, 4 H,
NCH2), 3.77 (s, 3 H, CH3), 4.04 (t, J ) 6.6 Hz, 2 H, CONCH2-
CH2N), 6.43 (s, 1 H, pyrrole), 6.86-6.91 (m, 4 H, aromatic),
7.36-7.49 (m, 2 H, aromatic), 7.83-7.87 (m, 1 H, aromatic),
8.30-8.39 (m, 1 H, aromatic), 11.26 (br s, 1 H, NH which
exchanges with D2O), 12.35 (br s, 1 H, NH which exchanges
with D2O). Anal. (C25H26ClN5O3) C, H, N.
6-(4-Chlorophenyl)-3-[2-[4-(2-methoxyphenyl)piperazin-
1-yl]ethyl]-1,5-dihydropyrrolo[3,2-d]pyrimidine-2,4-di-
one (33). The title compound was isolated as a white powder
(55%): mp 334-335 °C; 1H NMR (DMSO-d6) δ 2.40-2.65 (m,
4 H + 2 H, NCH2 + CONCH2CH2N), 2.80-3.00 (m, 4 H,
NCH2), 3.76 (s, 3 H, CH3), 4.03 (t, J ) 6.6 Hz, 2 H, CONCH2-
CH2N), 6.35 (s, 1 H, pyrrole), 6.85-6.91 (m, 4 H, aromatic),
7.45-7.50 (m, 2 H, aromatic), 7.88-7.93 (m, 2 H, aromatic),
11.21 (br s, 1 H, NH which exchanges with D2O), 12.30 (br s,
1 H, NH which exchanges with D2O). Anal. (C25H26ClN5O3) C,
H, N.
3-[3-[4-(2-Chlorophenyl)piperazin-1-yl]propyl]-6-(4-
methylphenyl)-1,5-dihydropyrrolo[3,2-d]pyrimidine-2,4-di-
one (38). The title compound was isolated as a white powder
(49%): mp 294-296 °C; 1H NMR (DMSO-d6) δ 1.68-1.84 (m,
2 H, CONCH2CH2CH2N), 2.30-2.56 (m, 4 H + 2 H + 3 H,
NCH2 + CONCH2CH2CH2N, CH3), 2.84-2.95 (m, 4 H, NCH2),
3.94 (t, J ) 6.6 Hz, 2 H, CONCH2CH2CH2N), 6.25 (s, 1 H,
pyrrole), 6.97-7.08 (m, 2 H, aromatic), 7.17-7.25 (m, 3 H,
aromatic), 7.36-7.40 (m, 1 H, aromatic), 7.74-7.78 (m, 2 H,
aromatic), 11.14 (br s, 1 H, NH which exchanges with D2O),
12.15 (br s, 1 H, NH which exchanges with D2O). Anal. (C26H28-
ClN5O2) C, H, N.
3-[2-[4-(2-Methoxyphenyl)piperazin-1-yl]ethyl]-6-(4-
methoxyphenyl)-1,5-dihydropyrrolo[3,2-d]pyrimidine-
2,4-dione (39). The title compound was isolated as a white
powder (59%): mp 310-312 °C; 1H NMR (DMSO-d6) δ 2.43-
2.75 (m, 4 H + 2 H, NCH2 + CONCH2CH2N), 2.85-3.04 (m, 4
H, NCH2), 3.76 (s, 3 H, CH3), 3.79 (s, 3 H, CH3), 4.03 (t, J )
6.6 Hz, 2 H, CONCH2CH2N), 6.20 (s, 1 H, pyrrole), 6.86-7.00
(m, 6 H, aromatic), 7.80-7.84 (m, 2 H, aromatic), 11.15 (br s,
1 H, NH which exchanges with D2O), 12.07 (br s, 1 H, NH
which exchanges with D2O). Anal. (C26H29N5O4) C, H, N.
3-[2-[4-(2-Chlorophenyl)piperazin-1-yl]ethyl]-6-(4-meth-
oxyphenyl)-1,5-dihydropyrrolo[3,2-d]pyrimidine-2,4-di-
one (40). The title compound was isolated as a white powder
(65%): mp 315-317 °C; 1H NMR (DMSO-d6) δ 2.44-2.78 (m,
4 H + 2 H, NCH2 + CONCH2CH2N), 2.85-3.04 (m, 4 H,
NCH2), 3.79 (s, 3 H, CH3), 4.04 (t, J ) 6.6 Hz, 2 H, CONCH2-
CH2N), 6.21 (s, 1 H, pyrrole), 6.86-7.00 (m, 6 H, aromatic),
7.80-7.84 (m, 2 H, aromatic), 11.15 (br s, 1 H, NH which
exchanges with D2O), 12.07 (br s, 1 H, NH which exchanges
with D2O). Anal. (C25H26ClN5O3) C, H, N.
3-[2-[4-(2-Chlorophenyl)piperazin-1-yl]ethyl]-6-(4-chlo-
rophenyl)-1,5-dihydropyrrolo[3,2-d]pyrimidine-2,4-di-
one (34). The title compound was isolated as a white powder
(51%): mp 337-338 °C; 1H NMR (DMSO-d6) δ 2.42-2.75 (m,
4 H + 2 H, NCH2 + CONCH2CH2N), 2.90-3.11 (m, 4 H,
NCH2), 4.04 (t, J ) 6.6 Hz, 2 H, CONCH2CH2N), 6.36 (s, 1 H,
pyrrole), 6.99-7.15 (m, 4 H, aromatic), 7.25-7.50 (m, 2 H,
aromatic), 7.89-7.94 (m, 2 H, aromatic), 11.23 (br s, 1 H, NH
which exchanges with D2O), 12.32 (br s, 1 H, NH which
exchanges with D2O). Anal. (C24H23Cl2N5O2) C, H, N.
3-[2-[4-(2-Methoxyphenyl)piperazin-1-yl]ethyl]-6-(4-
methylphenyl)-1,5-dihydropyrrolo[3,2-d]pyrimidine-2,4-
dione (35). The title compound was isolated as a white powder
Method B: General Procedure for the Preparation of
3-[2-[4-(Substituted phenyl)piperazin-1-yl]ethyl]-6-aryl-
1,5-dihydropyrrolo[3,2-d]pyrimidine-2,4-diones (25, 26,
29, 30, 32-36). A mixture of the appropriate (2-chloroethyl)-
urea (2.0 mmol) and the appropriate 1-(substituted phenyl)-
piperazine (10 mmol) was heated in an oil bath for 2 h at 140
°C. After cooling, the reaction mixture was treated with 15
mL of warm EtOH and the precipitate was filtered off, washed
with EtOH and successively with water, and dried. Recrys-
tallization from DMF/water (2:1) gave the desired products as
solids.
3-[2-[4-(2-Methoxyphenyl)piperazin-1-yl]ethyl]-6-phen-
yl-1,5-dihydropyrrolo[3,2-d]pyrimidine-2,4-dione (25). The
title compound was isolated as a white powder (50%): mp
328-329 °C; 1H NMR (DMSO-d6) δ 2.40-2.79 (m, 4 H + 2 H,
NCH2 + CONCH2CH2N), 2.82-3.10 (m, 4 H, NCH2), 3.77 (s,
3 H, CH3), 4.04 (t, J ) 6.6 Hz, 2 H, CONCH2CH2N), 6.32 (s, 1
H, pyrrole), 6.81-6.95 (m, 4 H, aromatic), 7.35-7.48 (m, 3 H,
aromatic), 7.80-7.98 (m, 2 H, aromatic), 11.19 (br s, 1 H, NH
which exchanges with D2O), 12.24 (br s, 1 H, NH which
exchanges with D2O). Anal. (C25H27N5O3) C, H, N.
1
(45%): mp 349-350 °C; H NMR (DMSO-d6) δ 2.32 (s, 3 H,
CH3), 2.45-2.66 (m, 4 H + 2 H, NCH2 + CONCH2CH2N),
2.85-3.04 (m, 4 H, NCH2), 3.77 (s, 3 H, OCH3), 4.03 (t, J )
6.6 Hz, 2 H, CONCH2CH2N), 6.26 (s, 1 H, pyrrole), 6.81-6.95
(m, 4 H, aromatic), 7.18-7.25 (m, 2 H, aromatic), 7.71-7.88
(m, 2 H, aromatic), 11.17 (br s, 1 H, NH which exchanges with
D2O), 12.16 (br s, 1 H, NH which exchanges with D2O). Anal.
(C26H29N5O3) C, H, N.
3-[2-[4-(2-Chlorophenyl)piperazin-1-yl]ethyl]-6-(4-meth-
ylphenyl)-1,5-dihydropyrrolo[3,2-d]pyrimidine-2,4-di-
one (36). The title compound was isolated as a white powder
3-[2-[4-(2-Chlorophenyl)piperazin-1-yl]ethyl]-6-phen-
yl-1,5-dihydropyrrolo[3,2-d]pyrimidine-2,4-dione (26). The
title compound was isolated as a white powder (59%): mp
331-332 °C; 1H NMR (DMSO-d6) δ 2.56-2.65 (m, 4 H + 2 H,
NCH2 + CONCH2CH2N), 2.89-3.01 (m, 4 H, NCH2), 4.05 (t,
J ) 6.6 Hz, 2 H, CONCH2CH2N), 6.32 (s, 1 H, pyrrole), 7.03-
7.15 (m, 2 H, aromatic), 7.24-7.46 (m, 5 H, aromatic), 7.86-
7.90 (m, 2 H, aromatic), 11.17 (br s, 1 H, NH which exchanges
with D2O), 12.23 (br s, 1 H, NH which exchanges with D2O).
Anal. (C24H24ClN5O2) C, H, N.
1
(48%): mp 337-338 °C; H NMR (DMSO-d6) δ 2.32 (s, 3 H,
CH3), 2.41-2.75 (m, 4 H + 2 H, NCH2 + CONCH2CH2N),
2.86-3.09 (m, 4 H, NCH2), 4.04 (t, J ) 6.6 Hz, 2 H, CONCH2-
CH2N), 6.26 (s, 1 H, pyrrole), 6.94-7.49 (m, 6 H, aromatic),
7.76-7.91 (m, 2 H, aromatic), 11.18 (br s, 1 H, NH which
exchanges with D2O), 12.15 (br s, 1 H, NH which exchanges
with D2O). Anal. (C25H26ClN5O2) C, H, N.
Pharmacology. Binding Assays. Male CRL:CD(SD)BR-
COBS rats weighing about 150 g were killed by decapitation,37
and their brains were rapidly dissected (hippocampus for
5-HT1A; striatum for D1 and D2; cortex for R1-ARs), frozen, and
stored at -80 °C until the day of assay.
Tissue was homogenized in about 50 volumes of ice-cold 50
mM Tris-HCl buffer (pH 7.4) using an Ultra Turrax TP-1810
(2 × 20 s) and centrifuged at 50000g for 10 min (Beckman
3-[2-[4-(4-Methoxyphenyl)piperazin-1-yl]ethyl]-6-phen-
yl-1,5-dihydropyrrolo[3,2-d]pyrimidine-2,4-dione (29). The
title compound was isolated as a white powder (50%): mp
371-372 °C; 1H NMR (DMSO-d6) δ 2.45-2.74 (m, 4 H + 2 H,
NCH2 + CONCH2CH2N), 2.82-3.10 (m, 4 H, NCH2), 3.67 (s,
3 H, CH3), 4.04 (t, J ) 6.6 Hz, 2 H, CONCH2CH2N), 6.31 (s, 1
H, pyrrole), 6.82-6.90 (m, 4 H, aromatic), 7.32-7.42 (m, 3 H,
aromatic), 7.85-7.89 (m, 2 H, aromatic), 11.19 (br s, 1 H, NH

2430 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 7
Patane` et al.
model J-21B refrigerated centrifuge). The pellet was resus-
pended in the same volume of fresh buffer, incubated at 37 °C
for 10 min, and centrifuged again at 50000g for 10 min. The
pellet was then washed once by resuspension in fresh buffer
and centrifuged as before. The pellet was then resuspended
in the appropriate incubation buffer (50 mM Tris-HCl, pH 7.7,
containing 10 µM pargyline and 4 mM CaCl2 for 5-HT1A
receptors; 50 mM Tris-HCl, pH 7.7, containing 10 µM pargyline
and 0.1% ascorbic acid for R1-ARs; 50 mM Tris-HCl, pH 7.1,
containing 10 µM pargyline, 120 mM NaCl, 5 mM KCl, 2 mM
CaCl2, 1 mM MgCl2, and 0.1% ascorbic acid for D1 and D2
receptors) just before the binding assay.
convergence, a plot of the number of times a variable has been
used versus the number of crossovers (variable usage graphic)
was analyzed. In particular, when the frequency of descriptor
usage remained unaltered, it suggested the optimal value of
crossover operations (∼20 000 in this study). Moreover, the
fact that descriptors found in the best final QSAR equations
(reported in Table 3) were the most frequently represented in
this graphic indicated that the GFA algorithm reached con-
vergence.
During GFA analysis, only linear polynomial terms were
allowed to be included in the final models. Default values of
mutation and smoothing parameters were kept, and the least-
squares method was used for regression.
Binding assays were done as previously described.38 Briefly,
the following incubation conditions were used. 5-HT1A: [3H]-
8-OH-DPAT (sp act. 157 Ci/mmol, NEN) final concentration 1
nM, 30 min, 25 °C (nonspecific binding, 5-HT 10 µM). D1: [3H]-
SCH23390 (sp act. 71.1 Ci/mmol, NEN) final concentration
0.4 nM, 15 min, 25 °C (nonspecific binding, (-)-cis-flupentixol
10 µM). D2: [3H]spiperone (sp act. 19.0 Ci/mmol, NEN) final
concentration 0.2 nM, 15 min, 37 °C (nonspecific binding,
(-)-sulpiride 100 µM). R1-ARs: [3H]prazosin (sp act. 71.8 Ci/
mmol, NEN) final concentration 0.2 nM, 30 min, 25 °C
(nonspecific binding, phentolamine 3 µM).
Incubations were stopped by rapid filtration under vacuum
through GF/B filters which were then washed with 12 mL
(4 × 3 times) of ice-cold 50 mM Tris-HCl (pH 7.4) using a
Brandel M-48R apparatus and counted in 4 mL of Filter Count
(Packard) in a LKB 1214 RACKBETA liquid scintillation
spectrometer with counting efficiency of about 60%. Dose-
inhibition curves were analyzed by the “Allfit” program to
obtain the concentration of unlabeled drugs that inhibited
ligand binding by 50%.39 The Ki values were derived from the
IC50 values.40
Supporting Information Available: Elemental analysis
data. This material is available free of charge via the Internet
at http://pubs.acs.org.
References
(1) Hieble, J. P.; Bondinell, W. E.; Ruffolo, R. R. R- and â-Adreno-
ceptors: From the Gene to the Clinic. 1. Molecular Biology and
Adrenoceptor Subclassification. J. Med. Chem. 1995, 38, 3415-
3444.
(2) Ruffolo, R. R., Jr.; Bondinell, W.; Hieble, J. P. R- and â-Adreno-
ceptors: From the Gene to the Clinic. 2. Structure-Activity
Relationships and Therapeutic Applications. J. Med. Chem.
1995, 38, 3681-3716.
(3) Ford, A. P. D. W.; Williams, T. J.; Blue, D. R.; Clarke, D. E. R1-
Adrenoceptor Classification: Sharpening Occam’s Razor. Trends
Pharmacol. Sci. 1994, 15, 167-170.
(4) Hieble, J. P.; Bylund, D. B.; Clarke, D. E.; Eikemburg, D. C.;
Langer, S. Z.; Lefkowiz, R. J.; Minneman, K. P.; Ruffolo, R. R.
International Union of Pharmacology. X. Recomandation for
Nomenclature of R1-Adrenoceptors: Consensus Update. Phar-
macol. Rev. 1995, 47, 267-270.
(5) Zhong, H.; Minneman, K. P. R1-Adrenoceptor Subtypes. Eur. J.
Pharmacol. 1999, 375, 261-276.
Molecular Modeling. All calculations and graphic ma-
nipulations were performed on an SGI Origin300 server and
an Octane 12K workstation by means of the Catalyst (version
4.8) and Cerius2 (version 4.8.1) software packages.
(6) Flavahan, N. A.; Vanhoutte, P. M. R-Adrenoceptor Subclassifica-
tion in Vascular Smooth Muscle. Trends Pharmacol. Sci. 1986,
7, 347-349.
All the compounds used in this study were built using the
2D-3D sketcher of Catalyst. A representative family of
conformations was generated for each molecule using the
poling algorithm and the “best quality conformational analysis”
method. Conformational diversity was emphasized by selection
of the conformers that fell within 20 kcal/mol above the lowest
energy conformation found. The Compare/Fit command has
been used to predict activity values of the studied compounds.
Particularly, the Best Fit option has been selected which
manipulates the conformers of each compound to find, when
possible, different mapping modes of the ligand within the
model. Affinity values for test set compounds were calculated
by the program on the basis of the following equation:
(7) Muramatsu, I.; Ohmura, T.; Kigoshi, S.; Hashimoto, S.; Ospita,
M. Pharmacological Subclassification of R1-Adrenoceptors in
Vascular Smooth Muscle. Br. J. Pharmacol. 1990, 99, 197-201.
(8) Grassi, G. Sympathetic Deactivation as a Goal of Nonpharma-
cologic and Pharmacologic Antihypertensive Treatment: Ratio-
nale and Options. Curr. Hypertens. Rep. 2003, 5, 277-280.
(9) Vaughan, E. D. Jr. Medical Management of Benign Prostatic
Hyperplasia. Are Two Drugs Better Than One? N. Engl. J. Med.
2003, 349, 2449-2451.
(10) Hancock, A. H. R1 Adrenoceptor Subtypes: A Synopsis of Their
Pharmacology and Molecular Biology. Drug Dev. Res. 1996, 39,
54-107.
(11) Trumpp-Kallmayer, S.; Hoflack, J.; Bruinvels, A.; Hibert, M.
Modeling of G-Protein-Coupled Receptors: Application to Dopa-
mine, Adrenaline, Serotonin, Acetylcholine, and Mammalian
Opsin Receptors. J. Med. Chem. 1992, 35, 3448-3462.
(12) Barbaro, R.; Betti, L.; Botta, M.; Corelli, F.; Giannaccini, G.;
Maccari, L.; Manetti, F.; Strappaghetti, G.; Corsano, S. Synthe-
sis, Biological Evaluation, and Pharmacophore Generation of
New Pyridazinone Derivatives with Affinity towards R1- and R2-
Αdrenoceptors. J. Med. Chem. 2001, 44, 2118-2132.
(13) (a) Betti, L.; Floridi, M.; Giannaccini, G.; Manetti, F.; Strappa-
ghetti, G.; Tafi, A.; Botta, M. R1-Adrenoceptor Antagonists. 5.
Pyridazinone-arylpiperazines. Probing the Influence on Affinity
and Selectivity of Both ortho-Alkoxy Groups at the Arylpipera-
zine Moiety and Cyclic Substituents at the Pyridazinone Nucleus.
Bioorg. Med. Chem. Lett. 2003, 13, 171-173. (b) Betti, L.; Corelli,
F.; Floridi, M.; Giannaccini, G.; Maccari, L.; Manetti, F.; Strap-
paghetti, G.; Botta, M. R1-Adrenoceptor Antagonists. 6. Struc-
tural Optimation of Pyridazinone-Arylpiperazines. Study of the
Influence on Affinity and Selectivity of Cyclic Substituents at
the Pyridazinone Ring and Alkoxy Groups at the Arylpiperazine
Moiety. J. Med. Chem. 2003, 46, 3555-3558.
(14) Carroll, W. A.; Sippy, K. B.; Esbenshade, T. A.; Buckner, S. A.;
Hancock, A. A.; Meyer, M. D. Two Novel and Potent 3-[(o-
Methoxyphenyl)piperazinylethyl]-5-phenylthieno[2,3-d]pyrimi-
dine-2,4-diones Selective for the R1D Receptor. Bioorg. Med.
Chem. Lett. 2001, 11, 1119-1121.
(15) Barlocco, D.; Cignarella, G.; DalPiaz, V.; Giovannoni, M. P.; De
Benedetti, P. G.; Fanelli, F.; Montesano, F.; Poggesi, E.; Leo-
nardi, A. Phenylpiperazinylalkylamino Substituted Pyridazin-
ones as Potent R1 Adrenoceptor Antagonists. J. Med. Chem.
2001, 44, 2403-2410.
affinity ) 10-(Fit+Intercept)
where the term Fit quantifies how well each compound
was able to fit the pharmacophoric features, while the
term Intercept represents the intercept of the regression
equation on the y axis estimating the affinity of training set
compounds. Further details can be found on the Accelrys Web
site: http://www.accelrys.com/doc/life/catalyst49/help/catalyst/
VHToolsmenu.doc.html#306504.
The Compare/Fit command has also been used to perform
molecular alignment of compounds to be submitted to QSAR
analysis. In detail, RN5 was fitted onto the pharmacophoric
model and its best fitted conformation was used as a template
to align compounds 21-40 through the Best Fit option. Next,
for each molecule in such an alignment, the atomic partial
charges were calculated by means of the Mopac software and
AM1 approximation, as implemented in the InsightII (version
2000.1) software package. Molecules with their charges were
imported in Cerius2 to perform a QSAR analysis based on the
genetic function approximation (GFA) algorithm. The GFA
parameter set used in this study was based on an initial
population of 100 equations that were evolved through 20 000
crossover operations. To evaluate if the GFA algorithm reached

Pyrimidine-2,4-dione Derivatives
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 7 2431
(16) Chern, J.; Tao, P.; Wang, K.; Gutcait, A.; Liu, S.; Yen, M.; Chien,
S.; Rong, J. Studies on Quinazolines and 1,2,4-Benzothiadiazine
1,1-Dioxides. 8.1,2 Synthesis and Pharmacological Evaluation of
Tricyclic Fused Quinazolines and 1,2,4-Benzothiadiazine 1,1-
Dioxides as Potential R1-Adrenoceptor Antagonists. J. Med.
Chem. 1998, 41, 3128-3141.
(17) Russo, F.; Romeo, G.; Guccione, S.; De Blasi, A. Pyrimido[5,4-
b]indole Derivatives. 1. A New class of Potent and Selective R1
Adrenoceptors Ligands. J. Med. Chem. 1991, 34, 1850-1854.
(18) Meyer, M. D.; Carrol, W. A. Piperazinyl Pyrimidine Dione
Compounds Selective for Adrenoceptors. U.S. 6,153,614, No-
vember 28, 2000.
(19) Rogers, D.; Hopfinger, A. J. Application of Genetic Function
Approximation to Quantitative Structure-Property Relation-
ships. J. Chem. Inf. Comput. Sci. 1994, 34, 854-866.
(20) Rogers, D. Some Theory and Examples of Genetic Function
Approximation with Comparison to Evolutionary Techniques.
In Genetic Algorithms in Molecular Modeling; Devillers, J., Ed.;
Academic Press: London, 1996; pp 87-107.
(30) Fan, Y.; Shi, L. M.; Kohn, K. W.; Pommier, Y.; Weinstein, J. N.
Quantitative Structure-Antitumor Activity Relationships of
Camptothecin Analogues: Cluster Analysis and Genetic Algo-
rithm-Based Studies. J. Med. Chem. 2001, 44, 3254-3263.
(31) Gokhale, V. M.; Kulkarni, V. M. Understanding the Antifungal
Activity of Terbinafine Analogues Using Quantitative Structure-
Activity Relationship (QSAR) Models. Bioorg. Med. Chem. 2000,
8, 2487-2499.
(32) Lee, K. W.; Briggs, J. M. Comparative Molecular Field Analysis
(CoMFA) Study of Epothilones-Tubulin Depolymerization In-
hibitors: Pharmacophore Development Using 3D QSAR Meth-
ods. J. Comput.-Aided Mol. Des. 2001, 15, 41-55.
(33) Wang, S.; Sakamuri, S.; Enyedy, I. J.; Kozikowski, A. P.; Zaman,
W. A.; Johnson, K. M. Molecular Modeling, Structure-Activity
Relationships and Functional Antagonism Studies of 4-Hydroxy-
1-methyl-4-(4-methylphenyl)-3-piperidyl 4-Methylphenyl Ke-
tones as a Novel Class of Dopamine Transporter Inhibitors.
Bioorg. Med. Chem. 2001, 9, 1753-1764.
(21) Furneaux, R. H.; Tyler, P. C. Improved Synthesis of 3H,5HPyrrolo-
[3,2-d]pyrimidines. J. Org. Chem. 1999, 64, 8411-8412.
(22) Chen, N.; Lu, Y.; Gadamasetti, K.; Hurt, C. R.; Norman, M. H.;
Fotsch, C. A Short, Facile Synthesis of 5-Substituted 3-Amino-
1H-pyrrole-2-carboxylates. J. Org. Chem. 2000, 65, 2603-2605.
(23) Norman, M. H.; Chen, N.; Han, N.; Liu, L.; Hurt, C. R.; Fotsch,
C. H.; Jenkins, T. J.; Moreno, O. A. Bicyclic Pyridine and
Pyrimidine Derivatives as Neuropeptide Y Receptor Antagonists.
WO 9940,091, August 12, 1999.
(24) Norman, M. H.; Chen, N.; Chen, Z.; Fotsch, C.; Hale, C.; Han,
N.; Hurt, R.; Jenkins, T.; Kincaid, J.; Liu, L.; Lu, Y.; Moreno,
O.; Santora, V. J.; Sonnenberg, J.; Karbon, W. Structure-
Activity Relationships of a Series of Pyrrolo[3,2-d]pyrimidine
Derivatives and Related Compounds as Neuropeptide Y5 Recep-
tor Antagonists. J. Med. Chem. 2000, 43, 4288-4312.
(25) Goetz, A. S.; King, H. K.; Ward, S. D. C.; True, T. A.; Rimele, T.
J.; Saussy, D. L., Jr. BMY 7378 Is a Selective Antagonist of the
D Subtype of R1-Adrenoceptor. Eur. J. Pharmacol. 1995, 272,
R5-R6.
(34) Internal or standard cross-validation in GFA encompasses only
the optimization of regression coefficients but does not encom-
pass the optimization of the choice of molecular descriptors.
(35) Kier, L. B.; Hall, L. H. An Electrotopological-State Index for
Atoms in Molecules. Pharm. Res. 1990, 7, 801-807.
(36) Hall, L. H.; Mohney, B.; Kier, L. B. The Electrotopological
State: An Atom Index for QSAR. Quant. Struct.-Act. Relat.
1991, 10, 43-51.
(37) Procedures involving animals and their care were conducted in
conformity with the institutional guidelines that are in compli-
ance with national (D.L. n. 116, G.U., Suppl. 40, February 18,
1992) and international laws and policies (EEC Council Directive
86/609, OJ L 358, 1, December 12, 1987; Guide for the Care and
Use of Laboratory Animals, U.S. National Research Council,
1996).
(38) Caccia, S.; Confalonieri, S.; Guiso, G.; Bernasconi, P.; Cagnotto,
A.; Skorupscka, M.; Mennini, T. Brain Uptake and Distribution
of the Potential Memory Enhancer CL 275,838 and Its Main
Metabolites in Rats: Relationship between Brain Concentrations
and in Vitro Potencies on Neurotransmitter Mechanisms. Psy-
chopharmacology 1994, 115, 502-508.
(26) Yocca, F. D.; Hyslop, D. K.; Smith, D. W.; Maayani, S. BMY 7378,
a Buspirone Analogue with High Affinity, Selectivity and Low
Intrinsic Activity at the 5-HT1A Receptor in Rat and Guinea Pig
Hippocampal Membranes. Eur. J. Pharmacol. 1987, 137, 293-
294.
(39) De Lean, K. W.; Munson, P. J.; Rodbard, D. Simultaneous
analysis of families of sigmoidal curves: application to bioassay,
radioligand assay, and physiological dose-response curves. Am.
J. Physiol. 1978, 235, E97-E102.
(40) Cheng, Y.; 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.
(27) Russell, R. K.; Press: J. B.; Rampulla, R. A.; McNally, J. J.;
Falotico, R.; Keiser, J. A.; Bright, D. A.; Tobia, A. Thiopene
Systems. 9. Thienopyrimidinedione Derivatives as Potential
Anthypertensive Agents. J. Med. Chem. 1988, 31, 1786-1793.
(28) Romeo, G.; Russo, F.; Guccione, S. Heterocyclic Systems Con-
taining the Pyrimido-2,4-dione Ring as Selective Ligands for the
R1-Adrenoceptors. Farmaco 1995, 6, 471-477.
(29) Cerius2 (version 4.8.1), Catalyst (version 4.8), and InsightII
(version 2000.1) softwares are distributed by Accelrys, Scranton
Road, San Diego, CA.
JM040870H