CODES, a novel procedure for ligand-based virtual screening: PDE7 inhibitors as an application example

Article abstract of DOI:10.1016/j.ejmech.2007.10.027

Phosphodiesterase (PDE) 7 is a high affinity cAMP-specific PDE whose functional role in T-cells has been the subject of some controversy. Recent findings on tissue distribution, however, support the hypothesis that PDE7 could be a good target for the treatment of airway diseases, T-cell related diseases or central nervous system (CNS) disorders. Therefore, the identification of selective inhibitors targeted against PDE7 enzyme has become an attractive area of research. We report here the first use of the descriptors generated by the CODES program for ligand-based virtual screening. This program codifies molecules from a topological point of view and the generated descriptors are related to the chemical nature of the atoms, the atomic bonds and the connectivity with the rest of the molecule. They are also able to distinguish among stereoisomers. By using this approach, 173 compounds were codified, and their similarity with the reference compound was analysed. Based on the analysis, new potential PDE7 inhibitors have been identified, synthesized and biologically evaluated confirming that CODES descriptors are valid for ligand-based virtual screening and provided new lead compounds for further optimization as potent and selective PDE7 inhibitors.

Full text of DOI:10.1016/j.ejmech.2007.10.027

Available online at www.sciencedirect.com
European Journal of Medicinal Chemistry 43 (2008) 1349e1359
http://www.elsevier.com/locate/ejmech
Original article
CODES, a novel procedure for ligand-based virtual screening:
PDE7 inhibitors as an application example
a
a
a,2
Ana Castro a, ,1, Marıa Jose Jerez , Carmen Gil , Felix Calderon ,
*
Teresa Domenech , Arsenio Nueda , Ana Martınez
´
´
´
´
b
b
a,1
´
´
a Instituto de Qu´ımica Me´dica (CSIC), Juan de la Cierva 3, 28006 Madrid, Spain
b Biology Department, Research Center, Almirall Prodesfarma, Laurea` Miro´ 408-410, 08980 Sant Feliu de Llobregat, Barcelona, Spain
Received 26 July 2007; received in revised form 18 October 2007; accepted 18 October 2007
Available online 21 December 2007
Abstract
Phosphodiesterase (PDE) 7 is a high affinity cAMP-specific PDE whose functional role in T-cells has been the subject of some controversy.
Recent findings on tissue distribution, however, support the hypothesis that PDE7 could be a good target for the treatment of airway diseases, T-
cell related diseases or central nervous system (CNS) disorders. Therefore, the identification of selective inhibitors targeted against PDE7 en-
zyme has become an attractive area of research. We report here the first use of the descriptors generated by the CODES program for ligand-based
virtual screening. This program codifies molecules from a topological point of view and the generated descriptors are related to the chemical
nature of the atoms, the atomic bonds and the connectivity with the rest of the molecule. They are also able to distinguish among stereoisomers.
By using this approach, 173 compounds were codified, and their similarity with the reference compound was analysed. Based on the analysis,
new potential PDE7 inhibitors have been identified, synthesized and biologically evaluated confirming that CODES descriptors are valid for
ligand-based virtual screening and provided new lead compounds for further optimization as potent and selective PDE7 inhibitors.
Ó 2007 Elsevier Masson SAS. All rights reserved.
Keywords: PDE7; Virtual screening; CODES
1. Introduction
One mechanism by which cAMP/cGMP may be elevated
within cells is by inhibition of cyclic nucleotide PDEs, which
are the only way to degrade them [3]. Based on the fact that
agents with the ability to elevate intracellular cAMP levels
have been demonstrated to possess immunosuppressive and
anti-inflammatory properties [4], the interest on the develop-
ment of specific PDEs inhibitors and their role on immunomo-
dulation processes have been renewed [5]. Thus, selective
inhibitors of cAMP-specific PDEs have been suggested as
therapies for the treatment of several human diseases [6], pre-
dominantly immunological disorders, such as multiple sclero-
sis [7], and inflammatory systems [8] and also disorders of the
CNS such as depression, ischemiaereperfusion injury, and
Alzheimer disease [9e12].
Phosphodiesterases (PDEs) comprise a large family of
metallophosphohydrolases enzymes that metabolize the ubiq-
uitous second messengers adenosine 30,50-cyclic monophos-
phate (cAMP) and guanosine 30,50-cyclic monophosphate
(cGMP) to their respective inactive 50-monophosphates [1].
cAMP and cGMP are generated through the action of adenylyl
cyclase and guanylyl cyclase, respectively, and serve to trans-
duce the action of many hormones, neurotransmitters, and
other cellular effectors [2].
* Corresponding author. Tel.: þ34 918061130; fax: þ34 918034660.
E-mail address: acastro@neuropharma.es (A. Castro).
From the large phosphodiesterases family, PDEs isoen-
zymes 3B, 4A, 4B, 4D and 7A1 are predominant in immune
cells [13e15]. To date, most of the research has been centered
on PDE4 inhibitors because PDE4 represents a major
1
Current address: NeuroPharma, S. A.; Avda. de la Industria 52, 28760 Tres
Cantos, Madrid, Spain.
2
´
Current address: Instituto de Quımica Organica General, Juan de la Cierva
´
3, 28006 Madrid, Spain.
0223-5234/$ - see front matter Ó 2007 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2007.10.027

1350
A. Castro et al. / European Journal of Medicinal Chemistry 43 (2008) 1349e1359
isoenzyme in most T-cell preparations and their selective in-
hibitors are able to reduce inflammatory cytokines production
[15,16]. However, a major drawback of these compounds has
been the significant side effect of emesis [17]. To avoid these
adverse effects several strategies have been considered to dis-
sociate the beneficial and detrimental effects of PDE4 inhibi-
tors with some degree of success and the second generation of
PDE4 inhibitors has shown better pharmacokinetics profiles
and less side effects [18e20].
An alternative approach is to target other cAMPePDE
families that are expressed in proinflammatory and immune
cells. Initial evidence indicated that PDE7 had an important
role in the activation of T-cells [21,22], however, results
based on the use of PDE7A (e/e) knock out mice failed
to confirm the role of PDE7A in T-cell proliferation and
suggested that this phosphodiesterase could have some other
role in the regulation of humoral immune responses [23].
Thus, the use of selective PDE7A inhibitors will be pivotal
to elucidate the true potential of PDE7A as a pharmacologi-
cal target in the context of the immune response [24].
Several years ago, our research group was the first one in
reported the first PDE7 selective inhibitors [25]. Since
then, a lot of efforts have been done to increase potency
and selectivity of this kind of compounds [26], conforming
a great variety of diverse chemical compounds with inter-
esting pharmacological profiles [27].
To evaluate the efficiency of such method, we conducted
a virtual screening focused on the finding of new phosphodies-
terase (PDE) 7 inhibitors. After the identification of lead com-
pounds, we synthesized new derivatives based on the results of
such screening and determined their inhibition on PDE7.
2. Materials and methods
Similarity searching begins with the identification of
a known bioactive molecule, the target structure. The structure
of this molecule is then compared with each of the structures
in a database using an objective measure of similarity. The tar-
get molecule and the database may comprise 2D and 3D struc-
tures, which are characterized by one or more descriptors that
describe some structural features of the molecules under scru-
tiny [28]. Later on a similarity coefficient is used to quantify
the degree of resemblance between the target structure and
each of the database structures. The database molecules are
then ranked in decreasing order of the calculated similarity
values, with the top-ranked molecules, being those that bear
the closest structural similarity to the target structure [34].
The effectiveness of the search depends on both the repre-
sentation and similarity coefficient used, which together com-
prise the overall similarity measure.
2.1. Selection of the target molecule and the database
The search for new lead compounds in the pharmaceutical
industry increasingly makes use of virtual screening of data-
bases for drug discovery. Among the available methods, simi-
larity searching is a cheap and widely used method to distill
a pool of potentially interesting compounds from a large data-
base [28]. This method is based on the similarity principle, one
of the crucial principles of the rational drug design, that states
that structurally similar compounds are expected to have simi-
lar physicochemical and biological properties, and therefore,
they could have similar in vivo effects. Thus, once an active
molecule has been identified, the ligand virtual screening is
based in the search in codified databases of similar structures
regarding similarity or any other specific property. This has
two main advantages: a high speed in comparison to the direct
analysis of each molecule in a database, and the ability to find
molecules that at a first glance do look similar but, however,
they hide the same features that have defined the target mole-
cule [29].
In the search of new PDE7 inhibitors, we selected as target
molecule the previously published 1-[(3,4-dichlorophenyl)-
methyl]-2,1,3-benzothiadiazin-4(3H )-one 2,2-dioxide (Fig. 1),
because it was the first PDE7 inhibitor described with an
IC50 value of 8 mM [25].
The database was created with a set of 173 structurally di-
verse molecules (1e173). This collection of compounds in-
cluded 131 structurally diverse derivatives (1e131) selected
because they have a biologically privileged scaffold that
show different pharmacological and toxicological properties,
but it was also enriched with 31 PDE7 inhibitors (132e162),
including the target structure as 134, and 11 PDE4 inhibitors
(163e173) using in different biological assays. (All these
structures are included in Supplementary data.)
2.2. Codification of structures: CODES
In the present study, we introduce CODES as a novel pro-
cedure for ligand-based virtual screening. This program,
which generates topological descriptors based on calculations
with neural networks (NN), has never been used for virtual
screening purposes, although it has been widely used for
QSAR studies, as for example the prediction of the nemato-
cide action of pteridine derivatives [30], as well as new potas-
sium channels openers [31]. In this way, it has also been used
to carry out QSPR studies as the determination of pharmaco-
kinetic properties as for example the oral absorption and BBB
penetration of several drugs [32] and the mean life of antihis-
taminic drugs, classifying them in their corresponding thera-
peutic categories [33].
Databases in which the virtual screening is carried out have
their structures codified by descriptors, that is, indexes
O
NH
SO2
N
Cl
Cl
Fig. 1. Target structure for the virtual screening of new PDE7 inhibitors.

A. Castro et al. / European Journal of Medicinal Chemistry 43 (2008) 1349e1359
1351
calculated on the basis of one or several properties of those
3. Results and discussion
structures that possess essential information about the nature
and/or structure of those molecules. Descriptors can be derived
from 1-, 2- or 3-dimensional representations of the molecule
[29].
Results were ordered in a decreasing way depending on
their similarity with the target molecule (Table 1). Results
for those molecules whose coefficient was higher than the
threshold of 0.9 and of 0.95 are shown in Fig. 2. We only se-
lect those molecules because the capacity to find new active
compounds increases with the selected threshold [37]. As it
can be observed, the only molecule with Tanimoto coefficient
equal to 1 is 134, because it was the target molecule.
From a first analysis of these structures it is possible to ob-
serve that only 11 out of 173 molecules (6.3%) have a Tani-
moto coefficient higher than 0.95, and only 17 (9.8%) have
it higher than 0.90. Therefore this can be considered as a selec-
tive and efficient method for virtual screening by similarity.
Moreover molecules 136, 137, 135 and 132 belong to the ben-
zothienothiadiazine family, which is structurally similar to the
target molecule; 146 and 152 are PDE7 inhibitors, what indi-
cate that this process retrieves potentially active molecules
against a biological target; and finally the most interesting re-
sult falls on the structures 173, 164 and 163, because they are
PDE4 inhibitors, enzyme with a high structural homology with
PDE7.
On the basis of these results, we propose the synthesis not
only of new thiadiazine derivatives analogues to the target
molecule (174e180), but also quinazoline derivatives (181e
196), structurally analogues to nitraquazone (173) (Scheme 1).
For the preparation of the newly designed benzothiadiazine
derivatives, we alkylated the corresponding heterocycles ac-
cording to previously described procedures [25,38]. On the
other hand, the quinazoline derivatives proposed were pre-
pared following the procedures previously described [39e
46] by cyclocondesation of the corresponding functionalised
heterocycles with isocyanates or isothiocyanates.
These novel sets of compounds were tested for their inhib-
itory potencies against human recombinant PDE7 expressed in
Saccharomyces cerevisiae as described in Section 5. In this ex-
pression system the only cyclic nucleotide hydrolysing activity
present in cell extracts corresponded to human PDE7. Isoen-
zyme selectivity was obtained by comparing the IC50 values
or % of inhibition of the compounds against PDE7 with their
inhibitory activity against PDE4 and PDE3 (Table 2).
Some of the heterocyclic compounds evaluated exhibited
PDE7 inhibitory properties (IC50 at mM level) validating the
lead compounds identified by similarity virtual screening.
From the biological results, preliminary structure-activity rela-
tionships have emerged. Regarding the heterocyclic system,
the thioxoquinazoline ring appears more effective than the
other condensed heterocycles (185, 193 and 195 versus 175
and 179). Moreover, derivatives with a non-substituted phenyl
(183 and 192) or ortho-halogen are the most active ones of
each series (185 and 187).
In this work the structures have been codified using
CODES program. These descriptors are related to the chemi-
cal nature of the atoms, the atomic bonds and the connectivity
with the rest of the molecule. Moreover, apart of all the infor-
mation of the chemical structure, they show two additional ad-
vantages over other topological descriptors: they are able to
distinguish among stereoisomers and, at the same time, avoid
all the problems associated with the 3D molecular conforma-
tion. As it is described in section theoretical basis of CODES
of Supplementary data, the output of this program is a matrix
in which each of the columns represents one of the atoms of
the molecule.
2.3. Reduction of dimensions
In general, most similarity measures depend on dimension-
ality [35], mean that they can only compare two structures
with the same dimensions, that is, with the same number of
descriptors. In the case of CODES, the matrix depends on
the number of atoms of the molecule. Therefore, once all the
molecules have been codified, it is necessary to reduce the di-
mensions to homogenize them in all the data set, that is, to be
able to compare all the molecules independently of their na-
ture. A widely spread approach is to reduce the dimensionality
of the original descriptors by means of principal component
analysis (PCA), selecting only the few first most relevant prin-
cipal components. Components of this new variable have to be
a linear combination of the original descriptors and they must
explain a proportion of the variance of the data big enough
[36]. Therefore, it summarized all the available information
with the minimum loss of explaining capacity with respect
to the variance.
In the case of CODES’ matrixes, the reduction of dimen-
sions by PCA was done by Tsar (version 3.3. Oxford Molecu-
lar, Ltd.), finding that the two first components were able to
explain 95% of the variance and 99% of the times. That is
why we select these two variables to define each compound
of the database.
2.4. Similarity criteria
For the final comparison of all the structures of the data-
base a similarity criteria needs to be used. Among all the
similarity measures between continuous topological descrip-
tors, Tanimoto coefficient was selected in this work because
it has been shown to be the most effective one in the iden-
tification of active compounds from a target molecule [34].
Anyway, this measure is monotonic with the Dice coeffi-
cient, is highly related with the Cosine coefficient, and for
binary descriptors its complement is coincident with the
Soergel coefficient. So, any of these measures provides
a good indication of similarity [28].
4. Conclusions
We have shown the utility of CODES neural networks mo-
lecular codification as a useful tool for ligand-based virtual

1352
A. Castro et al. / European Journal of Medicinal Chemistry 43 (2008) 1349e1359
screening of compounds within a wide range of chemical di-
verse structures and have been efficiently applied to identify
new PDE7 inhibitors. A database of 173 no-congeneric com-
pounds has been codified by descriptors generated by CODES
program, the dimensionality of the original descriptors was
done by PCA and Tanimoto coefficient was selected as simi-
larity criteria. Final results of virtual screening provided new
leads based on thiadiazine and quinazoline scaffolds. Several
derivatives structurally related to these leads were synthesized
and evaluated enzymatically. The biological data revealed that
these novel compounds are equipotent to the target structure
but with a simple chemical structure. For this reason, these
new compounds may be considered as new prototypes for fur-
ther optimization. CODES program has been validated for li-
gand-based virtual screening, proving to be a selective and
efficient method for the selection of new PDE7 inhibitors.
5. Experimental
5.1. Chemistry
Melting points were determined with a ReicherteJung
Thermovar apparatus and are uncorrected. Flash column chro-
matography was carried out at medium pressure using silica
gel (E. Merck, Grade 60, particle size 0.040e0.063 mm,
230e240 mesh ASTM) with the indicated solvent as eluent.
1H NMR spectra were obtained on Varian XL-300 and Bruker
WP-300 spectrometers working at 300 MHz. Typical spectral
parameters were: spectral width 10 ppm, pulse width 9 ms
(57ꢀ), data size 32 K. NOE difference spectra were measured
under the same conditions, using a presaturation time of 3 s.
13C NMR experiments were carried out on Varian XL-300
and Bruker WP-300 spectrometers operating at 75 MHz. The
Table 1
Virtual screening results ordered by the Tanimoto coefficient
Entry
Compound
Tanimoto
coefficient
Entry
Compound
Tanimoto
coefficient
Entry
Compound
Tanimoto
coefficient
Entry
Compound
Tanimoto
coefficient
1
134
74
1
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
91
63
0.04707651
0.03985446
0.03355749
0.03342122
0.03325900
0.02953837
0.02821337
0.02419064
0.02066309
0.01809397
0.01642970
0.01586728
0.01488014
0.01329916
0.01130440
0.00802879
0.00758128
0.00667694
0.00588713
0.00539618
0.00470288
ꢁ0.00459139
ꢁ0.00871085
ꢁ0.01016302
ꢁ0.01708360
ꢁ0.02322586
ꢁ0.03710607
ꢁ0.04287797
ꢁ0.04553995
ꢁ0.05763491
ꢁ0.05789004
ꢁ0.06634058
ꢁ0.06888479
ꢁ0.07006056
ꢁ0.07416544
ꢁ0.07856270
ꢁ0.08798144
ꢁ0.08915842
ꢁ0.09353445
ꢁ0.19490619
ꢁ0.19490769
ꢁ0.19872345
ꢁ0.32664176
87
88
139
34
ꢁ0.33141429
ꢁ0.33188710
ꢁ0.33194858
ꢁ0.33196216
ꢁ0.33197592
ꢁ0.33208948
ꢁ0.33212217
ꢁ0.33216254
ꢁ0.33218093
ꢁ0.33227110
ꢁ0.33232989
ꢁ0.33246317
ꢁ0.33247253
ꢁ0.33248705
ꢁ0.33249802
ꢁ0.33252201
ꢁ0.33253986
ꢁ0.33258125
ꢁ0.33258309
ꢁ0.33259591
ꢁ0.33263054
ꢁ0.33263889
ꢁ0.33264401
ꢁ0.33265572
ꢁ0.33265593
ꢁ0.33267046
ꢁ0.33268213
ꢁ0.33269321
ꢁ0.33269378
ꢁ0.33269675
ꢁ0.33269874
ꢁ0.33269892
ꢁ0.33270931
ꢁ0.33271742
ꢁ0.33272102
ꢁ0.33272778
ꢁ0.33273716
ꢁ0.33274624
ꢁ0.33278687
ꢁ0.33279869
ꢁ0.33280073
ꢁ0.33280428
ꢁ0.33280748
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
82
93
ꢁ0.33281374
ꢁ0.33282288
ꢁ0.33283736
ꢁ0.33284900
ꢁ0.33285026
ꢁ0.33285890
ꢁ0.33286646
ꢁ0.33286934
ꢁ0.33288291
ꢁ0.33289077
ꢁ0.33289549
ꢁ0.33290870
ꢁ0.33292077
ꢁ0.33293108
ꢁ0.33293197
ꢁ0.33293548
ꢁ0.33293684
ꢁ0.33295283
ꢁ0.33296629
ꢁ0.33298047
ꢁ0.33298630
ꢁ0.33299286
ꢁ0.33299979
ꢁ0.33301542
ꢁ0.33302045
ꢁ0.33302122
ꢁ0.33302414
ꢁ0.33302877
ꢁ0.33305239
ꢁ0.33306977
ꢁ0.33308175
ꢁ0.33309358
ꢁ0.33310867
ꢁ0.33312856
ꢁ0.33313077
ꢁ0.33313179
ꢁ0.33316889
ꢁ0.33317586
ꢁ0.33318839
ꢁ0.33327559
ꢁ0.33328733
ꢁ0.33330564
ꢁ0.33332744
2
0.99760222
0.99686219
0.99518338
0.98935455
0.98190982
0.98030514
0.97672433
0.97370866
0.95967932
0.95421559
0.94632901
0.94273906
0.93849906
0.91761092
0.90380532
0.90319139
0.90238568
0.89445047
0.88085173
0.85530084
0.85049195
0.85010393
0.80705312
0.80670073
0.80465623
0.79510746
0.79455399
0.79168535
0.77674841
0.75853641
0.74918709
0.74918663
0.69955841
0.68273209
0.65609348
0.61968205
0.54941936
0.53978515
0.27538400
0.25228953
0.18791646
0.07573717
3
39
22
89
90
107
88
8
86
4
173
130
98
6
69
5
91
92
49
44
117
92
6
81
43
7
85
93
94
50
30
168
115
106
13
8
136
120
146
137
59
142
121
56
9
95
96
94
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
125
14
131
3
97
98
35
47
52
55
135
163
152
67
95
99
18
57
141
96
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
123
124
125
126
127
128
129
130
9
40
118
36
33
73
99
164
132
65
116
60
58
78
84
166
101
53
128
129
10
79
29
161
7
124
162
42
87
19
167
71
31
16
38
144
154
27
145
90
109
37
150
153
51
122
75
114
105
149
2
76
26
138
104
25
111
23
160
54
62
61
148
21
64
17
4
156
171
A28
113
158
155
157
153
100
103
45
65
97
110
170
48
147
119
80
140
127
143
89
162
126
159
41
15
11
12
83
102
151
70
169
72
108
124
46
32
68
20
165
24
5
1
77

A. Castro et al. / European Journal of Medicinal Chemistry 43 (2008) 1349e1359
1353
Cl
OH
H
N
O
O
OH
H
N
O
N
N
S
HO
O
O
O
N
SO2
NH
N
O
HO
OH
O
N3
S
OH
O
130 (0.989)
98 (0.981)
136 (0.977)
85 (0.980)
CN
NO2
N
H2NO2S
Cl
N
N
S
F3C
N
O
HN
N
NO2
SO2
NH
N
S
120 (0.974)
146 (0.960)
S
OH
O
137 (0.954)
59 (0.946)
O
CF3
O
H
N
CF3
N
O
O
N
N
SO2
NH
NH
Cl
H
N
HN
Cl
HO
N
H
O
S
O
Cl
N
135 (0.943)
163 (0.938)
152 (0.918)
67 (0.904)
CH3
O
O
F
O
F
O
O
Cl
N
HN
SO2
NH
S
Cl
O
Cl
Cl
164 (0.903)
132 (0.902)
O
N
Cl
Cl
N
NH2
Cl
H2N
Cl
Cl
O
N
Cl
SO2
NH
O
Cl
Cl
74 (0.998)
O
NO2
39 (0.997)
134 (1)
173 (0.995)
Fig. 2. Structures with Tanimoto coefficient higher than 0.90 (in brackets) obtained virtual screening by similarity with CODES.

1354
A. Castro et al. / European Journal of Medicinal Chemistry 43 (2008) 1349e1359
F
F
F
H
N
F
N
Br
SO2
N
SO2
NH
NaHCO3/H2O
OH
O
174
175
H
N
N
Cl
SO2
N
SO2
NH
S
NaHCO3/H2O
S
OH
O
176
177
R
R
H
N
N
X
SO2
N
SO2
NH
NaHCO3/H2O
S
S
OH
O
178
R=2,5-diF 179
R=3,4-diCl 180
181
182
183
184
185
186
187
188
189
190
X = O R=H
X = O R=4-OMePh
X = S R=Ph
X = S R=4-OMePh
X = S R=2,6-diFPh
X = S R=2,3,4-triFPh
X = S R=2-BrPh
X = S R=Bn
H
N
X
R
NH2
OMe
Toluene, 130°C
+
+
R-N=C=X
R-N=C=S
N
N
O
O
X = S R=Naphthyl
X = S R=Me
H
N
S
R
NH2
Toluene, 130°C
OMe
S
S
O
O
191
R = Me
R = Ph
R = 2,6-diFPh
192
193
H
N
S
R
NH2
OMe
Toluene, 130°C
+
R-N=C=S
N
S
S
O
O
R = Ph
R = 2,6-diFPh
R = 2,3,4-triFPh
194
195
196
Scheme 1.

A. Castro et al. / European Journal of Medicinal Chemistry 43 (2008) 1349e1359
1355
Table 2
Biological activity (PDE7A, PDE3A, and PDE4B inhibitions) of thiadiazine
119.4 (C-4a), 116.1, 117.2, 118.9, 134.2, 155.6, 157.3 (AreC),
128.7 (C-5), 132.1 (C-7), 142.1 (C-8a), 165.6 (C-4). Anal.
C14H10F2N2O3S (C, H, N, S).
and quinazoline derivatives 174e196a
Compound
PDE3A (%)
PDE4B (%)
PDE7A (%)
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
a
0
0
12
5.1.2. 1-(1-Naphtylmethyl)-thieno[3,2-c]-1,2,6-thiadiazin-
4(3H )-one 2,2-dioxide (177)
14.1
0
12.2
1
4.1
0
To a solution of thienothiadiazine dioxide 176 [48]
(0.200 g, 1 mmol) in aqueous saturated solution of sodium bi-
carbonate (20 mL), 1-chloromethylnaphtyl chloride (0.176 g,
1 mmol) was added. The reaction mixture was refluxed for
24 h. After cooling to room temperature, the aqueous phase
was washed with CH2Cl2 (1 ꢃ 10 mL). The aqueous phase
was cooled at ꢁ4 ꢀC and the product was isolated by filtration
of aqueous phase. Purification: preparative thin layer chroma-
tography, using CH2Cl2:MeOH (10:1) as eluent; yield 0.070 g
82.5
0
7
26
16
3
30
12
13.2
26.2
1.7
4.4
21.7
6.3
37
3.6
25.1
8.4
0.2
5.5
2.4
5.5
48.5
1.9
5.4
2.1
7
59
4.4
3.3
1.7
6.5
2.5
2.7
8.8
1.9
20
13
56
1
(21%) as a solid; mp > 350 ꢀC. Purity 84% (by HPLC); H
16.4
5.9
12
NMR (CD3OD) d 4.98 (s, 2H, NeCH2), 6.70 (d, 1H,
JH6H7 ¼ 5.3 Hz, H-7), 6.97e7.46 (m, 7H, AreH), 7.57 (d,
1H, H-6); 13C NMR (CD3OD) d 54.5 (NeCH2), 119.7 (C-
7), 120.8 (C-4a), 124.5, 125.8, 126.9, 127.0, 127.4, 127.5,
127.9, 131.5, 133.5, 135.2 (AreC), 134.2 (C-6), 142.6 (C-
7a), 157.7 (C-4). Anal. C16H12N2O3S2 (C, H, N, S).
30.5
8
8.5
10
25
24
5
10
41.5
11
7
8.5
20.5
11.5
15.5
4.5
66.5
10
5.1.3. 1-[(2,5-Difluorophenyl)methyl]-benzo[4,5]thieno-
[3,2-c]-1,2,6-thiadiazin-4(3H )-one 2,2-dioxide (179)
The inhibitory potency of the synthesized compounds on the human
PDE3A, PDE4B and PDE7A activities was tested as described in Section 5.
Data are indicated as IC50 (mM) or percent inhibition at 10 mM (n ¼ 2e3).
To a solution of benzothienothiadiazine dioxide 178 [25]
(0.254 g, 1 mmol) in aqueous saturated solution of sodium
bicarbonate (20 mL), 2,5-difluorophenylmethyl bromide
(0.310 g, 1.5 mmol) was added. The reaction mixture was
refluxed for 3 h. After cooling to room temperature, the
aqueous phase was extracted with CH2Cl2 (4 ꢃ 10 mL).
The organic phase was dried over sodium sulphate and the
solvent evaporated under reduced pressure. The residue was
chromatographed by preparative thin layer chromatography,
using CH2Cl2:MeOH (8:1) as eluent; yield 0.040 g (10%)
as a yellow solid; mp 224e226 ꢀC. Purity 89% (by
acquisition parameters were: spectral width 16 kHz, acquisi-
tion time 0.99 s, pulse width 9 ms (57ꢀ), data size 32 K.
Chemical shifts are reported in d values (ppm) relative to in-
ternal Me4Si, and J values are reported in Hertz. Analytical
HPLC was carried out on a Waters 6000 system using a sym-
˚
metry C18 (5 mm, 100 A). Isocratic conditions were used: mo-
bile phase CH3CN/H2O (0.05% H3PO4 þ 0.04% Et3N); flow
rate 1 mL/min; detection, UV (254 nm). Elemental analyses
were performed by the analytical department at C.N.Q.O.
(CSIC) and the results obtained were within ꢂ0.4% of the the-
oretical values.
1
HPLC); H NMR (CD3OD) d 4.07 (s, 2H, NeCH2), 6.39e
6.52 (m, 3H, AreH), 7.49 (t, 1H, JH7H8 ¼ 6.9 Hz, H-7),
7.59 (t, 1H, JH8H9 ¼ 8.1 Hz, H-8), 7.79 (d, 1H, H-9), 7.89
(d, 1H, JH6H7 ¼ 7.3 Hz, H-6); 13C NMR (CD3OD) d 58.0
(NeCH2), 100.5, 109.7, 110.2, 145.7, 163.5, 163.6 (Are
C), 123.1 (C-9), 124.1 (C-6), 125.4 (C-4a), 125.9 (C-7),
128.7 (C-8), 131.6 (C-5a), 137.1 (C-9b), 140.3 (C-9a),
158.0 (C-4). Anal. C17H12F2N2S2O2 (C, H, N, S).
5.1.1. 1-[(2,5-Difluorophenyl)methyl]-
2,1,3-benzothiadiazin-4(3H )-one 2,2-dioxide (175)
To a solution of benzothiadiazine dioxide 174 [47] (0.210 g,
1 mmol) in aqueous saturated solution of sodium bicarbonate
(20 mL), 2,5-difluorophenylmethyl bromide (0.310 g, 1.5 mmol)
was added. The reaction mixture was refluxed for 3 h. After
cooling to room temperature, the aqueous phase was washed
with CH2Cl2 (1 ꢃ 10 mL). The aqueous phase was cooled at
ꢁ4 ꢀC and the product was isolated by filtration of aqueous
phase. Purification: preparative thin layer chromatography,
using CH2Cl2:MeOH (10:1) as eluent; yield 0.260 g (83%)
as a white solid; mp 131e132 ꢀC. Purity 86% (by HPLC);
1H NMR (CD3OD) d 5.16 (s, 2H, NeCH2), 6.71 (dd, 1H,
5.1.4. 1-[(3,4-Dichlorophenyl)methyl]-benzo[4,5]thieno-
[3,2-c]-1,2,6-thiadiazin-4(3H )-one 2,2-dioxide (180)
To a solution of benzothienothiadiazine dioxide 178 [25]
(0.254 g, 1 mmol) in aqueous saturated solution of sodium bi-
carbonate (20 mL), 3,4-dichlorophenylmethyl chloride (0.293 g,
1.5 mmol) was added. The reaction mixture was refluxed for
3 h. After cooling to room temperature, the aqueous phase
was washed with CH2Cl2 (4 ꢃ 10 mL). The organic phase
was dried over sodium sulphate and the solvent evaporated
under reduced pressure. The residue was chromatographed
by preparative thin layer chromatography, using CH2Cl2:
MeOH (8:1) as eluent; yield 0.060 g (15%) as a solid; mp
JH6H8 ¼ 1.2 Hz, JH7H8 ¼ 8.3 Hz, H-8), 6.89 (t, 1H, JH6H7
¼
7.3 Hz, H-6), 7.19 (t, 1H, JH5H7 ¼ 1.7 Hz, H-7), 7.06e7.99
(m, 3H, AreH), 7.90 (dd, 1H, JH5H6 ¼ 7.8 Hz, H-5); 13C
NMR (CD3OD) d 42.0 (CH2), 113.4 (C-8), 119.1 (C-6),

1356
A. Castro et al. / European Journal of Medicinal Chemistry 43 (2008) 1349e1359
227e228 ꢀC. Purity 91% (by HPLC); 1H NMR (CD3OD)
d 4.08 (s, 2H, NeCH2), 6.91e7.03 (m, 3H, AreH), 7.51 (t,
1H, JH7H8 ¼ 6.9 Hz, H-7), 7.59 (t, 1H, JH8H9 ¼ 8.1 Hz, H-8),
7.77 (d, 1H, H-9), 7.91 (d, 1H, JH6H7 ¼ 7.3 Hz, H-6); 13C
NMR (CD3OD) d 57.9 (NeCH2), 101.3, 128.9, 130.1,
133.02, 134.1, 141.9 (AreC), 122.9 (C-9), 124.3 (C-6),
125.6 (C-4a), 125.8 (C-7), 129.0 (C-8), 131.4 (C-5a), 137.1
(C-9b), 140.3 (C-9a), 158.5 (C-4). Anal. C16H10Cl2N2S2O3
(C, H, N, S).
indicated conditions in each case. After that, the product
was isolated by filtration.
5.2.1. 3-Phenyl-2-thioxo-(1H )-quinazolin-4-one (183)
Reagent: phenylisothiocyanate (0.135 g, 1 mmol). Condi-
tion: 24 h. Yield 0.090 g (35%) as a white solid; mp 300e
302 ꢀC (Ref. [49], 301 ꢀC). Purity 99% (by HPLC); 1H
NMR (CD3OD) d 7.20e7.60 (m, 7H, AreH, H-6, H-8), 7.80
(m, 1H, H-7), 8.06 (dd, 1H, JH5H6 ¼ 7.9 Hz, JH5H7 ¼ 1.1 Hz,
H-5); 13C NMR (CD3OD) d 115.8 (C-4a), 116.3 (C-8),
124.5 (C-6), 127.5 (C-5), 128.2, 129.0, 129.1, 139.4 (AreC),
135.7 (C-7), 139.9 (C-8a), 159.9 (C-4), 176.2 (C-2). Anal.
C14H10N2SO (C, H, N, S).
5.1.5. (1H )-Quinazolin-2,4-dione (181)
To a solution of potassium isocyanate (0.081 g, 1 mmol) in
1 mL of dry toluene, methyl anthranilate (0.151 g, 1 mmol)
was added. The reaction mixture was stirred at room tem-
perature and after 3 h extracted with CH2Cl2 (2 ꢃ 10 mL).
The organic phase was washed with an aqueous saturated so-
lution of sodium bicarbonate, dried over sodium sulphate, and
cooled at ꢁ4 ꢀC. After 24 h, the crude solid was filtered and
redissolved in 1 mL of ethanol and 1 mL of 10% NaOH.
The reaction was refluxed for 1 h. Acidification of the solution
with concentrate hydrochloric acid yielded 0.060 g (40%) of
181 as a white solid; mp 298e305 ꢀC (Ref. [39], 295e
5.2.2. 3-(4-Methoxyphenyl)-2-thioxo-(1H )-quinazolin-
4-one (184)
Reagent: 4-methoxyphenylisothiocyanate (0.165 g, 1 mmol).
Condition: 72 h. Yield 0.070 g (24%) as a white solid; mp
1
274e276 ꢀC (Ref. [43], 275 ꢀC). Purity 99% (by HPLC); H
NMR (CD3OD) d 3.76 (s, 3H, OCH3), 6.83e7.59 (m, 6H,
AreH, H-6, H-8), 7.77 (m, 1H, H-7), 7.93 (dd, 1H,
JH5H6 ¼ 7.9 Hz, JH5H7 ¼ 1.4 Hz, H-5); 13C NMR (CD3OD)
d 55.7 (OCH3), 115.7 (C-4a), 115.2 (C-8), 116.5, 116.7,
130.0, 130.8, 139.3, 159.3 (AreC), 125.5 (C-6), 127.6 (C-5),
135.8 (C-7), 140.2 (C-8a), 159.7 (C-4), 175.8 (C-2). Anal.
C15H12N2SO2 (C, H, N, S).
1
300 ꢀC). Purity 98% (by HPLC); H NMR (CD3OD) d 7.2e
7.6 (m, 2H, H-6, H-8), 7.59 (m, 1H, H-7), 7.83 (dd, 1H,
JH5H6 ¼ 7.9 Hz, JH5H7 ¼ 1.1 Hz, H-5); 13C NMR (CD3OD)
d 114.5 (C-4a), 115.5 (C-8), 122.5 (C-6), 127.1 (C-5), 135.1
(C-7), 141.0 (C-8a), 150.4 (C-4), 163.0 (C-2). Anal.
C8H6N2O2 (C, H, N).
5.2.3. 3-(2,6-Difluorophenyl)-2-thioxo-(1H )-quinazolin-
4-one (185)
5.1.6. 3-(4-Methoxyphenyl)-(1H )-quinazolin-
2,4-dione (182)
Reagent: 2,6-difluorophenylisothiocyanate (0.171 g, 1 mmol).
Condition: 4 days. Yield 0.160 g (55%) as a white solid; mp
262e263 ꢀC. Purity 99% (by HPLC); 1H NMR (CD3OD)
d 7.30e7.66 (m, 5H, AreH, H-6, H-8), 7.85 (m, 1H, H-7),
8.01 (dd, 1H, JH5H6 ¼ 7.7 Hz, JH5H7 ¼ 0.9 Hz, H-5); 13C
NMR (CD3OD) d 112.5, 112.7, 115.2, 131.8, 156.5, 159.0
(AreC), 115.5 (C-4a), 116.5 (C-8), 125.5 (C-6), 127.9
(C-5), 137.0 (C-7), 140.0 (C-8a), 159.8 (C-4), 175.0 (C-2).
Anal. C14H8F2N2SO (C, H, N, S).
To a solution of 4-methoxyphenylisocyanate (0.149 g,
1 mmol) in 1 mL of dry toluene, methyl anthranilate
(0.151 g, 1 mmol) was added. The reaction mixture was re-
fluxed and after 72 h extracted with CH2Cl2 (2 ꢃ 10 mL).
The organic phase was washed with an aqueous saturated so-
lution of sodium bicarbonate, dried over sodium sulphate, and
cooled at ꢁ4 ꢀC. After 24 h, the crude solid was filtered and
redissolved in 1 mL of ethanol and 1 mL of 10% NaOH.
The reaction was refluxed for 1 h. Acidification of the solution
with concentrate hydrochloric acid yielded 0.110 g (43%) of
182 as a white solid; mp 297e299 ꢀC (Ref., 299 ꢀC). Purity
5.2.4. 3-(2,3,4-Trifluorophenyl)-2-thioxo-(1H )-quinazolin-
4-one (186)
Reagent: 2,3,4-trifluorophenylisothiocyanate (0.189 g, 1 mmol).
Condition: 14 days. Yield 0.200 g (67%) as a white solid; mp
260e261 ꢀC. Purity 99% (by HPLC); 1H NMR (CD3OD)
d 7.37e7.56 (m, 4H, AreH, H-6, H-8), 7.83 (m, 1H, H-7),
7.99 (dd, 1H, JH5H6 ¼ 7.2 Hz, JH5H7 ¼ 0.9 Hz, H-5); 13C
NMR (CD3OD) d 113.0, 124.4, 126.5, 145.3, 148.9, 152.2
(AreC), 115.8 (C-4a), 116.3 (C-8), 125.2 (C-6), 127.8
(C-5), 136.5 (C-7), 139.9 (C-8a), 159.6 (C-4), 175.5 (C-2).
Anal. C14H7F3N2SO (C, H, N, S).
1
99% (by HPLC); H NMR (CD3OD) d 3.43 (s, 3H, OCH3),
7.10e7.26 (m, 4H, AreH), 7.45 (m, 1H, H-6), 7.53 (dd, 1H,
JH8H7 ¼ 7.9 Hz, JH8H6 ¼ 1.1 Hz, H-8), 7.85 (m, 1H, H-7),
8.05 (dd, 1H, JH5H6 ¼ 7.9 Hz, JH5H7 ¼ 1.1 Hz, H-5); 13C
NMR (CD3OD) d 58.9 (OCH3), 116.3, 116.9, 123.8, 124.0,
125.2, 176.8 (AreC), 119.9 (C-4a), 120.7 (C-8), 128.0 (C-
6), 133.1 (C-5), 140.7 (C-7), 145.3 (C-8a), 164.4 (C-4),
167.9 (C-2). Anal. C15H12N2O3 (C, H, N).
5.2. General procedure for the synthesis of
thioxoquinazoline derivatives
5.2.5. 3-(2-Bromophenyl)-2-thioxo-(1H )-quinazolin-
4-one (187)
Reagents: 2-bromophenylisothiocyanate (0.214 g, 1 mmol).
Conditions: 36 h. Yield 0.170 g (52%) as a white solid; mp
252e254 ꢀC (Ref. [44], 256 ꢀC). Purity 98% (by HPLC);
1H NMR (CD3OD) d 7.20e7.80 (m, 7H, AreH, H-6, H-7, H-8),
To a solution of the corresponding isothiocyanate (1 mmol)
in 1 mL of dry toluene, methyl anthranilate (0.151 g, 1 mmol)
was added. The reaction mixture was refluxed under the

A. Castro et al. / European Journal of Medicinal Chemistry 43 (2008) 1349e1359
1357
7.79 (dd, 1H, JH5H6 ¼ 7.9 Hz, JH5H7 ¼ 1.1 Hz, H-5); 13C NMR
(CD3OD) d 115.8 (C-4a), 116.0 (C-8), 122.5, 128.8, 130.4,
131.4, 132.9, 138.3 (AreC), 124.7 (C-6), 127.6 (C-5), 136.9
(C-7), 139.8 (C-8a), 159.1 (C-4), 175.2 (C-2). Anal.
C14H9BrN2SO (C, H, N, S).
(C-4a), 115.0 (C-7), 117.5 (C-6), 137.5 (C-7a), 156.8 (C-4),
175.6 (C-2). Anal. C7H6N2SO2 (C, H, N, S).
5.3.2. 3-Phenyl-2-thioxo-(1H )-thieno[3,2-d]pyrimidin-
4-one (192)
Reagent: phenylisothiocyanate (0.135 g, 1 mmol). Condi-
tion: 5 days. Yield 0.030 g (13%) as a white solid; mp 265e
267 ꢀC. Purity 99% (by HPLC); 1H NMR (CD3OD) d 7.07 (d,
1H, JH6H7 ¼ 5.1 Hz, H-7), 7.23e7.78 (m, 5H, AreH), 7.84 (d,
1H, H-6); 13C NMR (CD3OD) d 115.7 (C-7), 117.7 (C-6),
126.0, 128.6, 129.3, 139.5 (AreC), 137.9 (C-7a), 145.6 (C-
4a), 157.0 (C-4), 176.7 (C-2). Anal. C12H8N2SO2 (C, H, N, S).
5.2.6. 3-Benzyl-2-thioxo-(1H )-quinazolin-4-one (188)
Reagent: benzylisothiocyanate (0.149 g, 1 mmol). Condi-
tion: 24 h. Yield 0.130 g (50%) as a yellow solid; mp 231e
1
233 ꢀC (Ref. [45], 231e233 ꢀC). Purity 99% (by HPLC); H
NMR (CD3OD) d 5.22 (s, 2H, CH2), 7.18e7.26 (m, 7H,
AreH, H-6, H-8), 7.74 (m, 1H, H-7), 7.81 (dd, 1H,
JH5H6 ¼ 7.9 Hz, JH5H7 ¼ 1.1 Hz, H-5); 13C NMR (CD3OD)
d 50.4 (CH2), 116.3 (C-8), 116.6 (C-4a), 124.5 (C-6), 126.7,
127.7, 129.3, 140.8 (AreC), 127.9 (C-5), 135.8 (C-7), 140.6
(C-8a), 159.7 (C-4), 175.2 (C-2). Anal. C15H12N2SO (C, H,
N, S).
5.3.3. 3-(2,6-Difluorophenyl)-2-thioxo-(1H )-
thieno[3,2-d]pyrimidin-4-one (193)
Reagent: 2,6-difluorophenylisothiocyanate (0.171 g, 1 mmol).
Condition: 17 days. Yield 0.005 g (2%) as a white solid; mp
270e272 ꢀC. Purity 98% (by HPLC); 1H NMR (CD3OD)
d 7.11 (d, 1H, JH6H7 ¼ 5.1 Hz, H-7), 7.21e8.30 (m, 3H,
AreH), 7.84 (d, 1H, H-6); 13C NMR (CD3OD) d 112.5,
112.7, 126.0, 159.9 (AreC), 114.4 (C-7), 117.9 (C-6), 139.5
(C-7a), 146.6 (C-4a), 156.6 (C-4), 175.5 (C-2). Anal.
C12H6F2N2SO2 (C, H, N, S).
5.2.7. 3-(1-Naphtyl)-2-thioxo-(1H )-quinazolin-4-one (189)
Reagent: 1-naphtylisothiocyanate (0.185 g, 1 mmol). Con-
dition: 7 days. Yield 0.050 g (18%) as a white solid; mp
1
226e228 ꢀC (Ref. [42], 228 ꢀC). Purity 99% (by HPLC); H
NMR (CD3OD) d 7.15e7.79 (m, 9H, AreH, H-6, H-8), 7.74
(m, 1H, H-7), 7.90 (dd, 1H, JH5H6 ¼ 7.9 Hz, JH5H7 ¼ 1.1 Hz,
H-5); 13C NMR (CD3OD) d 115.7 (C-4a), 115.9 (C-8),
124.1, 124.5, 124.9, 125.7, 125.8, 126.5, 126.6, 128.2,
134.6, 135.4, (AreC), 124.5 (C-6), 127.8 (C-5), 135.7 (C-7),
139.6 (C-8a), 159.7 (C-4), 175.7 (C-2). Anal. C18H12N2SO
(C, H, N, S).
5.4. General procedure for the synthesis of
benzothienopyrimidinone derivatives
To a solution of the corresponding isothiocyanate (1 mmol)
in 1 mL of dry toluene, 3-amino-benzo[b]thiophene-2-carboxy-
lic acid methyl ester [51] (0.207 g, 1 mmol) was added. The re-
action mixture was refluxed under the indicated conditions in
each case. After that, the product was isolated by filtration.
5.2.8. 3-Methyl-2-thioxo-(1H )-quinazolin-4-one (190)
Reagent: methylisothiocyanate (0.073 g, 1 mmol). Condi-
tion: 72 h. Yield 0.090 g (46%) as a yellow solid; mp 262e
1
264 ꢀC (Ref. [50], 259e260 ꢀC). Purity 90% (by HPLC); H
5.4.1. 3-Phenyl-2-thioxo-(1H )-benzo[4,5]thieno[3,2-d]-
pyrimidin-4-one (194)
NMR (CD3OD) d 3.45 (s, 3H, CH3), 7.30e7.39 (m, 2H, H-
6, H-8), 7.73 (m, 1H, H-7), 7.95 (dd, 1H, JH5H6 ¼ 8.0 Hz,
JH5H7 ¼ 0.7 Hz, H-5); 13C NMR (CD3OD) d 33.6 (CH3),
115.6 (C-4a), 115.9 (C-8), 124.7 (C-6), 127.6 (C-5), 135.6
(C-7), 139.6 (C-8a), 159.9 (C-4), 175.7 (C-2). Anal.
C9H8N2SO (C, H, N, S).
Reagent: phenylisothiocyanate (0.135 g, 1 mmol). Condi-
tion: 5 days. Yield 0.040 g (14%) as a yellow solid; mp
255e256 ꢀC. Purity 95% (by HPLC); 1H NMR (CD3OD)
d 7.24e7.47 (m, 5H, AreH), 7.56 (m, 1H, H-8), 7.64 (t, 1H,
JH7H8 ¼ 6.0 Hz, H-7), 8.10 (d, 1H, JH6H7 ¼ 6.3 Hz, H-6), 8.60
(d, 1H, JH8H9 ¼ 5.4 Hz, H-9); 13C NMR (CD3OD) d 114.5 (C-
4a), 124.0 (C-9), 124.0 (C-6), 125.6 (C-8), 128.0, 129.0, 139.7
(AreC), 129.3 (C-7), 129.9 (C-9a), 140.5 (C-5a), 142.6 (C-
9b), 157.7 (C-4), 177.0 (C-2). Anal. C16H10N2S2O (C, H, N, S).
5.3. General procedure for the synthesis of
thienopyrimidinone derivatives
To a solution of the corresponding isothiocyanate (1 mmol)
in 1 mL of dry toluene, 3-amino-thiophene-2-carboxylic acid
methyl ester [48] (0.157 g, 1 mmol) was added. The reaction
mixture was refluxed under the indicated conditions in each
case. After that, the product was isolated by filtration.
5.4.2. 3-(2,6-Difluorophenyl-2-thioxo-(1H )-benzo[4,5]-
thieno[3,2-d]-pyrimidin-4-one (195)
Reagent: 2,6-difluorophenylisothiocyanate (0.171 g, 1 mmol).
Condition: 24 h. Yield 0.040 g (12%) as a yellow solid; mp
280e282 ꢀC. Purity 99% (by HPLC); 1H NMR (CD3OD)
d 7.22e7.33 (m, 3H, AreH), 7.52e7.62 (m, 2H, H-7, H-8),
8.14 (d, 1H, JH6H7 ¼ 8.06 Hz, H-6), 8.60 (d, 1H, H-9); 13C
NMR (CD3OD) d 117.6, 130.8, 142.6, 160.3 (AreC), 123.1
(C-4a), 124.0 (C-6), 124.2 (C-9), 125.7 (C-8), 128.6 (C-9a),
129.7 (C-7), 140.9 (C-5a), 143.7 (C-9b), 156.4 (C-4), 175.9
(C-2). Anal. C16H8F2N2S2O (C, H, N, S).
5.3.1. 3-Methyl-2-thioxo-(1H )-thieno[3,2-d]pyrimidin-
4-one (191)
Reagent: methylisothiocyanate (0.073 g, 1 mmol). Condi-
tion: 5 days. Yield 0.040 g (21%) as a brown solid; mp
234e236 ꢀC. Purity 99% (by HPLC); 1H NMR (CD3OD)
d 2.97 (s, 3H, CH3), 7.01 (d, 1H, JH6H7 ¼ 5.1 Hz, H-7), 8.15
(d, 1H, H-6); 13C NMR (CD3OD) d 33.6 (CH3), 145.3

1358
A. Castro et al. / European Journal of Medicinal Chemistry 43 (2008) 1349e1359
[3] W. Ritcher, Proteins 46 (2002) 278e286.
5.4.3. 3-(2,3,4-Trifluorophenyl-2-thioxo-(1H )-
[4] A.C. Allison, Immunopharmacology 47 (2000) 63e83.
[5] D.M. Essayan, Biochem. Pharmacol. 57 (1999) 965e973.
[6] C. Lugnier, Pharmacol. Ther. 109 (2006) 366e398.
[7] H.J. Dyke, J.G. Montana, Exp. Opin. Invest. Drugs 11 (2002) 1e13.
[8] D. Spina, Drugs 63 (2003) 2575e2594.
benzo[4,5]thieno[3,2-d]-pyrimidin-4-one (196)
Reagent: 2,3,4-trifluorophenylisothiocyanate (0.189 g,
1 mmol). Condition: 17 days. Yield 0.030 g (9%) as a yellow
solid; mp 268e269 ꢀC. Purity 99% (by HPLC); 1H NMR
(CD3OD) d 7.15e7.38 (m, 2H, AreH), 7.49e7.65 (m, 2H,
H-7, H-8), 8.14 (d, 1H, JH6H7 ¼ 8.02 Hz, H-6), 8.60 (d, 1H,
H-9); 13C NMR (CD3OD) d 115.3, 118.8, 130.1, 130.2,
131.4, 167.0 (AreC), 123.1 (C-4a), 124.2 (C-9), 124.3 (C-6),
125.5 (C-8), 128.7 (C-9a), 129.6 (C-7), 140.8 (C-5a), 143.8
(C-9b), 156.4 (C-4), 175.8 (C-2). Anal. C16H7F3N2S2O
(C, H, N, S).
[9] F.S. Menniti, W.S. Faraci, C.J. Schmidt, Nat. Rev. Drug Discov. 5 (2006)
660e670.
[10] J.M. O’Donnell, H.T. Zhang, Trends Pharmacol. Sci. 25 (2004) 158e
163.
[11] S. Birk, L. Edvinsson, J. Olesen, C. Kruuse, Eur. J. Pharmacol. 489
(2004) 93e100.
´
[12] S. Perez-Torres, R. Cortes, M. Tolnay, A. Probst, J.M. Palacios,
G. Mengod, Exp. Neurol. 182 (2003) 322e334.
[13] D. Ekholm, B. Hemmer, G. Gao, M. Vergelli, R. Martin, V. Manganiello,
J. Immunol. 159 (1997) 1520e1529.
5.5. Measurement of PDE activities
[14] S. Erdogan, M.D. Houslay, Biochem. J. 321 (1997) 165e175.
[15] M.A. Giembycz, C.J. Corrigan, J. Seybold, R. Newton, P.J. Barnes, Br. J.
Pharmacol. 118 (1996) 1945e1958.
PDE3A (purified from human platelets), PDE4B (human
recombinant) and PDE7A (human recombinant) activities
were monitored by measuring the hydrolysis of [3H]-cAMP
to [3H]-AMP using a PDEeSPA kit (Amersham). Extracts
containing the corresponding human phosphodiesterases
were incubated in ‘‘low binding’’ plates (Costar 3604) for
60 min at room temperature. The assay mixture (80 mL) con-
tains 15 nM [3H]-cAMP (1 mCi/mL) in the assay buffer
(50 mM tris, pH 7.5, 8.3 mM MgCl2, 1.7 mM EGTA) and
10 mL of test compound. These compounds were resuspended
in DMSO (final DMSO concentration 5% (v/v)) and tested at
different concentrations varying from 1 mM to 1 nM to calcu-
late an IC50 and/or % activity inhibition. These dilutions were
done in 96-well plates.
[16] C.D. Manning, M. Burman, S.B. Christensen, L.B. Cieslinski,
D.M. Essayan, M. Grous, T.J. Torphy, M.S. Barnette, Br. J. Pharmacol.
128 (1999) 1393e1398.
[17] D.M. Essayan, A. Kagey-Sobotka, L.M. Lichtenstein, S.K. Huang, J.
Pharmacol. Exp. Ther. 282 (1997) 505e512.
[18] M.D. Houslay, P. Schafer, K.Y.J. Zhang, Drug Discov. Today 10 (2005)
1503e1519.
[19] K.F. Chung, Eur. J. Pharmacol. 533 (2006) 110e117.
[20] M. Giembycz, Curr. Opin. Pharmacol. 5 (2005) 238e244.
[21] L. Li, C. Yee, J.A. Beavo, Science 283 (1999) 848e851.
[22] A. Nakata, K. Ogawa, T. Sasaki, N. Koyama, K. Wada, J. Kotera,
H. Kikkawa, K. Omori, O. Kaminuma, Clin. Exp. Immunol. 128
(2002) 460e466.
[23] G. Yang, K.W. McIntyre, R.M. Townsend, H.H. Shen, W.J. Pitts,
J.H. Dodd, S.G. Nadler, M. McKinnon, A.J. Watson, J. Immunol. 171
(2003) 6414e6420.
Hydrolysis of [3H]-cAMP was initiated by adding 10 mL of
a solution containing the corresponding phosphodiesterase ex-
tract, and the plate was then incubated under agitation at room
temperature. The reaction was stopped after 60 min (with 10e
20% substrate conversion) by addition of 50 mL phosphodi-
esterase scintillation proximity assay (SPA) beads. All reactions
were carried out in duplicate. After incubation the reaction was
stopped with 50 mL (0.89 mg) of PDE SPA beads (Amersham
Pharmacia Biotech), and the resulting mixture was allowed to
settle for 20 min before counting in a microtitre plate counter.
[24] S.J. Smith, L.B. Cieslinski, R. Newton, L.E. Donnelly, P.S. Fenwick,
A.G. Nicholson, P.J. Barnes, M.S. Barnette, M.A. Giembycz, Mol. Phar-
macol. 66 (2004) 1679e1689.
´
[25] A. Martınez, A. Castro, C. Gil, M. Miralpeix, V. Segarra, T. Domenech,
´
´
J. Beleta, J.M. Palacios, H. Ryder, X. Miro, C. Bonet, J.M. Casacuberta,
´
~
´
F. Azorın, B. Pina, P. Puigdomenech, J. Med. Chem. 43 (2000) 683e689.
´
[26] A. Castro, M.I. Abasolo, C. Gil, V. Segarra, A. Martınez, Eur. J. Med.
Chem. 36 (2001) 333e338.
[27] A. Castro, M.J. Jerez, C. Gil, A. Martınez, Med. Res. Rev. 25 (2005)
229e244.
´
[28] M. Whittle, V.J. Gillet, P. Willett, J. Chem. Inf. Comput. Sci. 44 (2004)
1840e1848.
[29] F. Barbosa, H. Dragos, Curr. Top. Med. Chem. 4 (2004) 589e600.
´
´
[30] C. Ochoa, J. Rodrıguez, M. Rodrıguez, A. Chana, M. Stud, P. Alonso-
Villalobos, M.M. Martınez-Grueiro, Med. Chem. Res. 7 (1998) 530e
Acknowledgment
´
545.
[31] A. Martınez, A. Castro, M. Stud, J. Rodrıguez, I. Cardelus, J. Llenas,
´
[32] I. Dorronsoro, A. Chana, M.I. Abasolo, A. Castro, C. Gil, M. Stud,
The authors gratefully acknowledge the financial support of
´
NeuroPharma S.A., CICYT (SAF2003/02262) and Fundacion
´
´
´
A. Fernandez, J.M. Palacios, Med. Chem. Res. 8 (1999) 171e180.
´
Medica Mutua Madrilena (MPY 1387/04) for this project.
~
´
A. Martınez, Quant. Struct. -Act. Relat. 23 (2004) 89e98.
´
lat. 19 (2000) 448e454.
~
´
[33] C. Quinones, J. Caceres, M. Stud, A. Martınez, Quant. Struct. -Act. Re-
Appendix. Supplementary data
[34] P. Willett, J. Chem. Inf. Comput. Sci. 38 (1998) 983e996.
[35] A. Bender, H.Y. Mussa, G.S. Gill, R.C. Glen, J. Med. Chem. 47 (2004)
6569e6583.
Theoretical basis of CODES and structures of the database.
Supplementary data associated with this article can be found
in the online version, at doi:10.1016/j.ejmech.2007.10.027.
[36] H. Xu, D.K. Agafiotis, Curr. Top. Med. Chem. 2 (2002) 1305e1320.
[37] T. Poter, H. Matter, J. Med. Chem. 41 (1998) 478e488.
¨
´
´
[38] A. Martınez, C. Gil, A. Castro, A.M. Bruno, C. Perez, C. Prieto, J. Otero,
Antiviral Chem. Chemother. 14 (2003) 107e114.
References
[39] Y.-Q. Wu, D.C. Limburg, D.E. Wilkinson, G.S. Hamilton, J. Heterocycl.
Chem. 40 (2003) 191e193.
[1] M. Conti, S.L. Jin, Prog. Nucleic Acid Res. Mol. Biol. 63 (1999) 1e38.
[2] M. Vig, A. George, R. Sen, J. Durdik, S. Rath, V. Bal, Mol. Pharmacol.
62 (2002) 1471e1481.
´
[40] J. Garin, E. Melendez, F.L. Merchan, T. Tejero, E. Villarroya, Synthesis 5
´
(1983) 406e408.

A. Castro et al. / European Journal of Medicinal Chemistry 43 (2008) 1349e1359
[41] H.Y.P. Choo, M. Kim, S.K. Lee, Bioorg. Med. Chem. 10 (2002) 517e523. [47] E. Cohen, B. Klarberg, J. Am. Chem. Soc. 84 (1962) 1994e2002.
1359
´
[42] N.B. Das, A.S. Mittra, J. Indian Chem. Soc. 56 (1979) 398e400.
[43] P.N. Bhargava, P. Ram, Bull. Chem. Soc. Jpn. 38 (1965) 342e344.
[44] S. Muthusamy, V.T. Ramakrishnan, Synth. Commun. 22 (1992) 519e
533.
[48] A. Martınez, A.I. Esteban, A. Castro, C. Gil, G. Andrei, R. Snoeck,
J. Balzarini, E. De Clerq, Antiviral Chem. Chemother. 11 (2000) 221e
230.
[49] D.H. Raheja, U.P. Fulwadhva, N.B. Laxmeshwar, J. Indian Chem. Soc.
61 (1984) 584e585.
[45] R.M. Shafik, A.A.B. Hazzaa, N.S. Habib, Pharmazie 34 (1979) 148e150.
[46] D.H.R. Barton, D. Crich, G. Kretzschmar, J. Chem. Soc., Perkin Trans. 1
(1986) 39e53.
[50] J.P. Ferris, S. Singh, T.A. Newton, J. Org. Chem. 44 (1979) 173e178.
[51] R.B. Beck, J. Org. Chem. 37 (1972) 3224e3226.