SAR and 3D-QSAR studies on thiadiazolidinone derivatives: Exploration of structural requirements for glycogen synthase kinase 3 inhibitors

Article abstract of DOI:10.1021/jm040895g

The 2,4-disubstituted thiadiazolidinones (TDZD) are described as the first ATP-noncompetitive GSK-3 inhibitors. Following an SAR study about TDZD, different structural modifications in the heterocyclic ring aimed to test the influence of each heteroatom o

Full text of DOI:10.1021/jm040895g

J. Med. Chem. 2005, 48, 7103-7112
7103
Articles
SAR and 3D-QSAR Studies on Thiadiazolidinone Derivatives: Exploration of
Structural Requirements for Glycogen Synthase Kinase 3 Inhibitors
Ana Martinez,*,†, Mercedes Alonso,,†, Ana Castro,,†, Isabel Dorronsoro,,†, J. Luis Gelp´ı,§ F. Javier Luque,|
Concepcio´n Pe´rez,† and Francisco J. Moreno‡
Instituto de Quı´mica Me´dica (CSIC), Juan de la Cierva 3, 28006 Madrid (Spain), Centro de Biolog´ıa Molecular “Severo
Ochoa” (CSIC-UAM), Universidad Auto´noma de Madrid, 28049 Madrid (Spain), Departamento de Bioqu´ımica y Biologı´a
Molecular Facultad de Quı´mica, Universidad de Barcelona, Martı´ i Franque´s 1, 08028 Barcelona (Spain), and Departamento
de Fisicoquı´mica, Facultad de Farmacia, Universidad de Barcelona, Av. Diagonal 643, 08028 Barcelona (Spain)
Received November 16, 2004
The 2,4-disubstituted thiadiazolidinones (TDZD) are described as the first ATP-noncompetitive
GSK-3 inhibitors. Following an SAR study about TDZD, different structural modifications in
the heterocyclic ring aimed to test the influence of each heteroatom on the biological study are
here reported here. Various compounds such as hydantoins, dithiazolidindiones, rhodanines,
maleimides, and triazoles were synthesized and screened as GSK-3 inhibitors. After an extensive
SAR study among these different heterocyclic families, TDZDs have been revealed as a
privileged scaffold for the selective inhibition of GSK-3. A CoMFA analysis was also performed
highlighting the molecular electrostatic field interaction in the interaction of TDZDs with GSK-
3. Moreover, first mapping studies indicate two binding modes which in turn might imply
relevant differences in the mechanism that underly the inhibitory activity of TDZDs.
Introduction
Glycogen synthase kinase-3 (GSK-3) is a ubiquitous
serine/threonine kinase which plays a key role in a large
number of cellular processes and diseases. Two func-
tional aspects make GSK-3 a peculiar kinase: its
activity is constitutive and down regulated after cell
activation by phosphorylation or interaction with inhibi-
tory proteins, and the enzyme prefers substrates that
are prephosphorylated by other kinases.1 Its pleiotropic
but unique activities have made GSK-3 a much sought-
after target for the treatment of prevalent human
diseases2 such as type 2 diabetes,3 Alzheimer’s disease,4
CNS disorders such as manic depressive disorder and
neurodegenerative diseases, and chronic inflammatory
Figure 1. Common GSK-3 inhibitors.
disorders.5
3â,13-17 rational drug optimization is now being applied
to discover new lead compounds.18
Nowadays, the search for GSK-3 inhibitors is a very
active research field for both academic centers and
pharmaceutical companies.6,7 The most frequent ap-
proach until now for discovering GSK-3 inhibitors
has exploited screening programs, specifically aimed
at finding new ‘hits’ which exhibit other pharmaco-
logical profiles. This is the case for kinase inhibitors
hymenialdisine,8 paullones,9 indirubines,10 maleim-
ides,11 and TDZDs12 (Figure 1). However, because of the
availability of X-ray crystallographic data of GSK-
Among the great diversity of chemical structures with
GSK-3 inhibition already found,19 the 2,4-disubstituted
thiadiazolidinones (TDZD) are presented as the first
ATP-noncompetitive GSK-3 inhibitors.12 These com-
pounds are of great interest since they did not show
inhibitory activity on other kinases such as PKA, PKC,
CK-2, and CDK1/cyclin B. Preliminary structure-
activity relationship (SAR) studies suggested that the
substituent attached to the N2 of the thiadiazolidinone
ring, as well as the nature of the substituent of N4,
seems to be important for GSK-3 inhibition. Moreover,
both C5 and C3 carbonyl groups could mediate favorable
interactions with the enzyme.
* Corresponding author. Tel 91-8061130 Fax: 91-8034660, e-mail:
amartinez@neuropharma.es.
† Instituto de Qu´ımica Me´dica (CSIC).
‡ Universidad Auto´noma de Madrid.
§ Departamento de Bioqu´ımica y Biolog´ıa Molecular Facultad de
Qu´ımica, Universidad de Barcelona.
In this study different structural modifications are
introduced in the heterocyclic ring with the aim to test
the influence of each heteroatom on the biological
| Departamento de Fisicoqu´ımica, Universidad de Barcelona.
Present address: NeuroPharma, S.A.; Avda. de la Industria 52,
28760 Tres Cantos, Madrid (Spain).
10.1021/jm040895g CCC: $30.25 © 2005 American Chemical Society
Published on Web 10/08/2005

7104 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23
Martinez et al.
Scheme 2a
a Reagents: (i) THF or toluene, r.t., 20 h.
Scheme 3a
Figure 2. Schematic representation of the structural differ-
ences in the heteroaromatic rings of the different classes of
compounds considered in the study.
Scheme 1a
a Reagents: (i) R′-X (X ) Cl, Br, I), alkaline medium.
using the commercial 4-phenyl-urazole (compound 31)
or hydantoin (compound 36), respectively, as starting
material under classical alkylation conditions (alkaline
medium and alkyl halide)23 (Scheme 3).
Biological Results and SAR Studies. Maleimides.
The maleimide derivatives are an important family of
compounds in chemical24 and biochemical areas.25,26 The
maleimide scaffold could be considered structurally
related to the TDZD, one in which two heteroatoms, the
sulfur atom, and the nitrogen directly attached to it are
eliminated. However, the planarity of the ring is pre-
served with the double bond between both carbon atoms.
Our previous study on TDZDs12 indicated that the
attachment of aromatic or hydrophobic substituents at
N4 leads to an increase of the activity. Therefore,
derivatives where the nitrogen atom is bonded through
different linkers to an aromatic ring containing func-
tional groups capable of establishing hydrogen-bond
interactions or an ester group were prepared. These new
derivatives (compounds 4-9), together with a wide set
of commercially available related compounds (com-
pounds 1-3 and 10-16), were evaluated as GSK-3
inhibitors following the method described in the Ex-
perimental Section.
In general, these compounds are GSK-3 inhibitors at
micromolar level (see Table 1). Inhibitory activity
increased when the nitrogen atom of the maleimide ring
was substituted by any aromatic or ester substituent
(compound 1 and 2 versus 3-5). Moreover, the length
of the linker between the aryl group and the heterocyclic
nitrogen does not influence the inhibitory activity
(compounds 6-9). Considering the nature of the atom
that bridges the two carbonyl groups, the inhibitory
activity decreases in the sequence N>C .. O (com-
pounds 1, 10, and 11). Finally, it is clear that the
presence of carbonyl groups and the double bond
between the two carbons attached to the carbonyl
groups are prerequisites for the inhibitory activity
(compounds 13-16).
a Reagents: (i) R-OH, Ph3P, DIAD or DEAD, THF, rt, 24 h; (ii)
R-X, alkaline medium, CH3COCH3.
activity. Thus, the biological activity of a large set of
different heterocyclic compounds structurally related to
TDZDs was evaluated. The common scaffold was a
pentagonal ring containing carbonyl and thiocarbonyl
groups in a 1,3 relative disubstitution. Among these
heterocycles we have screened as GSK-3 inhibitors
different compounds such as hydantoins, dithiazolidin-
diones, rhodanines, maleimides, and triazoles (Figure
2). The CoMFA methodology,20 in conjunction with
mapping studies, are used to gain insight into the
molecular determinants that mediate the GSK-3 inhibi-
tory activity of these compounds.
Chemistry. The chemical modifications on the dif-
ferent target heterocyclic systems, except for dithiazo-
lidinedione, were achieved using the more suitable
method of alkylation in each case. Thus, maleimides
4-9 and some rhodanine derivatives (compounds 22,
23, 26, and 27) were prepared using Mitsunobu reaction
modification,21 whereas N-alkylation of the rhodanine
ring to yield 24 and 25 derivatives occurred using a
weak base and the corresponding alkyl halide in acetone
(Scheme 1).
A previously described method22 using chlorocarbonyl-
sulfenyl chloride with different formamides was selected
for the preparation of dithiazolidinedione derivatives
28-30 (Scheme 2).
Different triazolidinedione (compounds 32-34) and
hydantoin (compound 37) derivatives were synthesized
Rhodanines and Thiazolidines. These compounds,
which have a broad pharmacological profile (antibacte-
rial,27 antiviral,28 anticonvulsive,29 and antihyper-
glicemic30 activity), differ from TDZDs by the presence
of only one nitrogen atom in the ring. Even though some

SAR and 3D-QSAR Studies on Thiadiazolidinones
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23 7105
Table 1. GSK-3 Inhibition of Maleimide-Related Compounds
Table 3. GSK-3 Inhibition of Dithiazolidindione,
Triazolidinedione, and Hydantoin Derivatives
IC50
(µM)
compd
X
Y
Z
A
B
R
compd
X
Y
R
R′
R′′
IC50 (µM)
1
CH CH
CH CH
CH CH
CH CH
CH CH
CH CH
CH CH
CH CH
CH CH
CH CH CH2 CdO CdO
CH CH CdO CdO
CH2 CH2 CH2 CdO CdO
CH2 CH2
CH2 CH2
CH2 CH2
CH CH
N
CdO CdO
H
6
5
1
3
2
N
N
N
N
N
N
N
N
CdO CdO CH3
28
29
30
31
32
33
34
35
36
37
S
S
S
S
CH3
Ph
-
-
-
H
-
-
-
H
100
12
12
100
100
100
100
100
100
100
3
CdO CdO CH2Ph
S
4
CdO CdO CH2CO2Et
CdO CdO CH2Ph-p-OCH3
CdO CdO (CH2)2Ph
S
CH2Ph
Ph
5
2.5
N
N
N
N
N
N
N
N
6
2
3
3
3
N
Ph
CH3
COPh
CH2COPh CH2COPh
CH3
COPh
7
CdO CdO (CH2)2Ph-p-OCH3
CdO CdO (CH2)3Ph
CdO CdO (CH2)5Ph
N
Ph
8
N
Ph
9
N
Ph
-
H
H
-
-
-
10
11
12
13
14
15
16
12
CH2
H
O
>100
>100
>100
>50
>100
>100
CH2 CH2Ph
N
N
N
N
CdO CdO H
aromatic substituent attached to the nitrogen atom,
flanked by the two carbonyl groups, seems to be
important for activity (compound 28 versus 29 and 30).
However, the activity, nevertheless, is completely lost
in triazolidindione and hydantoin compounds, which
suggests a crucial role for the sulfur atom of the TDZD
scaffold in modulating the inhibitory activity against
GSK-3.
CdO CdO CH2Ph
CdO CH2 CH2Ph
CH2 CH2 CH2Ph
Table 2. GSK-3 Inhibition of the Rhodanine-Related
Derivatives
Mode of Inhibition and Selectivity Studies. To
verify the ATP noncompetitive mechanism of action and
the selectivity of TDZD compounds and the rest of
derivatives examined in this study, several kinetic and
selectivity experiments were performed. Kinetic analysis
were carried out using combinations of six ATP concen-
trations (from 6.5 to 100 µM) and two inhibitor concen-
trations for two different GSK-3 inhibitors: a derivative
of dithiazolidinedione family (compound 30) and a
compound of the known TDZD family (compound 44).
Figure 3 shows that while the enzyme maximal rate
decreases in the presence of the inhibitor, the Michaelis
constant remains unaltered, thus indicating that com-
pounds 30 and 44 act as ATP-non competitive inhibitors.
Selectivity of kinase inhibition is critical for pathway
analysis in vivo and in cellular systems. In our previous
studies12 TDZD compounds did not show inhibitory
activity on PKA, PKC, CK-2, and CDK1/cyclin B. To
further examine the selectivity profile of the TDZD
family, some of the more representative TDZD com-
pounds (41, 43, 45, 50, 55, and 59) were evaluated
against a panel of seven related kinases: Abl-K, CAM
KII, EGF-TyrK, IRK MAPK, MEK1K, and PKp56. The
inhibitory assays were performed by using a 10 µM
concentration for the different inhibitors. The results
reported in Table 4 show that TDZD compounds do not
in general have a significant inhibitory effect on the
whole set of kinases, which in turn demonstrates the
remarkable specificity of TDZD with respect to GSK-3.
CoMFA and Mapping Studies. The preceding data
pointed out the important role played by the nature of
the heteroaromatic ring and by certain substituents in
mediating the GSK-3 inhibition. To further explore the
relationship between the chemical structure and the
biological activity of TDZDs, a 3D-QSAR study was
carried out by using CoMFA methodology. For this
purpose, we selected 48 compounds, which include a
variety of TDZDs (compounds 38-57; Table 5) as well
as representative members of the other heterocyclic
compd
X
Y
R
IC50 (µM)
17
18
19
20
21
22
23
24
25
26
27
S
S
S
S
O
S
S
S
O
S
S
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
CH2
H
100
100
100
100
100
25
35
50
100
65
100
CH3
CH2COOH
NH2
H
Bn
(CH2)2Ph
CH2COPh
Bn
Bn-p-OMe
(CH2)2Ph-p-OMe
of them are commercially available (compounds 17-21),
we introduced different aromatic or hydrogen-bonding
acceptor substituents at the nitrogen position in order
to increase the chemical diversity among these struc-
tures (compounds 22-27).
The GSK-3 inhibition data are given in Table 2. The
N-benzyl-rhodanine derivative (compound 22) was found
to have the best inhibitory activity, thus supporting the
important role played by the aromatic substituent at
this position. However, the activity diminishes when the
linker between the aromatic moiety and the nitrogen
atom is longer (compound 23). More remarkably, the
presence of a thiocarbonyl group is a prerequisite for
activity in these compounds, as its replacement by a
carbonyl group (compare compound 22-compound 25)
is detrimental.
Dithiazolidinediones, Triazolidinediones, and
Hydantoins. 1,2,4-Dithiazolidine-3,5-diones derivatives
bear a sulfur atom instead of the nitrogen atom at
position 2 in TDZDs. In constrast, in triazolidinedione
ring a nitrogen atom replaces the sulfur atom present
in TDZDs. Finally, some hydantoin compounds, where
the sulfur atom is replaced by carbon, were also included
in the study.
The results of the biological assays against GSK-3
(Table 3) corroborated our previous findings. Thus, an

7106 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23
Martinez et al.
Table 4. Inhibitory Activity (% inhibition)a Exhibited by Selected TDZD Compounds for Several Kinases
compd (10 µM)
Abl-K
CAM K II
EGF-TyrK
IRK
MAP K
MEK 1 K
PK p 56
41
43
45
50
55
59
-
-
-
-
-
-
23
16
35
-
-
25
-
17
40
19
-
20
19
36
46
50
-
-
-
-
-
35
27
-
-
-
-
10
-
-
-
-
15
19
-
a The symbol - indicates an inhibition of less than 10%.
Table 5. GSK-3 Inhibition of TDZD Compounds 38-59 Used
in the CoMFA Study
compd
R
R′
IC50 (µM)
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
Et
Et
Et
Et
Et
Bn
Bn
Bn
Ph
Ph-p-Br
Ph-m-Br
Ph-o-Br
Ph-p-CF3
Ph-p-Cl
Ph-p-F
Ph-p-OMe
Ph-p-Me
Ph-p-NO2
1-naphthyl
c-hex
Ph
CH2CO2Et
Me
Pr
5
10
35
10
25
2
iPr
c-hex
Et
Me
Et
7
Bn
Me
Me
Me
Me
Me
Me
Me
Me
Me
Et
10
2
1.1
4
6
6
4
4
2
5
8.5
2
Me
Me
Et
100
6
2
Me
Table 6. Statistical Resultsa of CoMFA Analysis
model
field
q2
n
r2
s
F
A
B
C
D
E
F
steric
0.403
0.495
0.194
0.654
0.274
5
3
2
5
2
4
5
0.862 0.251 49.85
0.775 0.313 48.22
0.393 0.507 13.92
0.922 0.189 94.32
0.498 0.462 21.30
0.780 0.313 36.33
0.859 0.253 48.84
electrostatic
h bond
steric+electrostatic
steric+lipophilic
electrostatic+ lipophilic 0.440
G
steric+electrostatic +
0.556
Figure 3. Double-reciprocal plot of kinetic data determined
for compounds 30 and 44. Each point is the mean of two
different experiments, each one analyzed in duplicate.
lipophilic
a q2, n, r2, s, and F are the squared cross-validated correlation
coefficient, the number of components, the squared correlation
coefficient, the standard deviation of regression equation, and the
F ratio, respectively.
compounds examined here (compounds 1-19, 22-25,
28, 31, 32, and 35).
Table 7. IC50 Data (µM) for GSK-3 Inhibitors Including in the
CoMFA Validation Set
CoMFA models were calculated for each molecular
field considered alone or in combination (see Table 6).
The best PLS analysis was obtained by combining steric
and electrostatic fields (see model D in Table 6), leading
to a good correlation between the IC50 values predicted
from the principal components extracted from these two
fields and the experimental data (r2 ) 0.922, q2 ) 0.654,
and n ) 5; see Figure 4). To further validate model D,
we have attempted to predict the activities of five
representative compounds not included in the training
set (compounds 4, 5, 7, 58, and 59). The predicted IC50
values are also in good agreement with experimental
values, which gives support to the predictive power of
the model.
compound
experimental
predicted
4
3
2.5
3
6
2
3.8
3.5
3
5
7
58
59
17
3.5
Since the relative contributions of steric and electro-
static molecular field were 42% and 58%, respectively,
it seems that the inhibitory activity in TDZD and
related compounds is modulated by a subtle balance of
the two fields. Figure 5 shows the graphical representa-
tion of CoMFA results. Sterically favorable regions

SAR and 3D-QSAR Studies on Thiadiazolidinones
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23 7107
Figure 6. Chemical structures of I-5, indirubin-3-monoxime,
and alsterpaullone
the X-ray structures determined for the enzyme bound
to those inhibitors. For the three inhibitors, we identi-
fied the best poses out of first 10 possible solutions by
comparing the positional root-mean square deviation
(rmsd) between predicted and experimentally deter-
mined solutions (a RMSD threshold of ∼2 Å is usually
considered to be the upper limit for drug discovery).33
For I-5, the best pose predicted correctly the experi-
mental binding mode of the inhibitor with an rmsd of
1.3 Å. Two other poses close in energy to the best pose
also reflected the alignment of the inhibitor with rmsds
of 1.7 and 2.2 Å. Similar results were found for indi-
rubin-3′-monoxime, where two poses reflected the ex-
perimentally determined solution with rmsds of 1.4 and
1.8 Å. Finally, six poses placed alsterpaullone at the
same binding site detected in the X-ray structure, but
with partial success, since the ligand was predicted to
be rotated by ∼180° along its main inertial axis, thus
inverting the position of the pyrrole NH and amide CdO
groups. This result can be mainly attributed to the
absence of explicit water molecules in the enzyme
structure used for docking, since binding of those groups
is mediated by bridging water molecules. On the basis
of these findings, though caution is required to deter-
mine accurately the binding mode of TDZDs, CMIP
calculations appear suitable to identify potential binding
sites in the enzyme.
The CMIP analysis was performed for TDZD com-
pound 45 and allowed us to identify two potential
binding sites. One of them is located in the vicinity of
the activation loop, near residues Arg96, Arg180, and
Lys205 (pocket on the right side of the enzyme as shown
in Figure 5). There seems to be a nice shape comple-
mentarity between the binding pocket and the ligand,
which adopts an eclipsed-like conformation with the two
phenyl rings pointing toward the same face of the TDZD
heteroaromatic ring. In this conformation cation-π
interactions can be formed between the phenyl rings and
the positively charged Arg92 and Arg96 residues. The
binding of TDZDs to this pocket, which would explain
the non-ATP competitive nature of these inhibitors (see
above) might hinder the proper alignment of the sub-
strate in the binding site, thus leading to inhibition of
the phosphorylation mechanism. On the other hand, an
alternative binding site is also found in the ATP-binding
cavity (pocket on the left side of the enzyme as shown
in Figure 7), where the ligand adopts an extended
conformation. This recognition mode might not a priori
justify the non-ATP competitive nature of TDZDs.
However, based on the sensitivity of the GSK3 inhibi-
tory data to the nature of the heteroaromatic ring and
particularly the requirement of the sulfur atom in the
Figure 4. Graphical representation of experimental and
predicted (model D) IC50 data for compounds included in the
training set.
Figure 5. Graphical representation of CoMFA analysis
(model D). (a) Steric contour plots: The yellow area represents
the region where an increase of steric bulk is predicted to
increase activity, whereas the green area represents the region
where an increase of steric bulk is predicted to decrease
activity. (b) Electrostatic contour plots: blue contour means
the region where a negative electrostatic potential enhances
the inhibitory activity.
(Figure 5a) are found around the aromatic substituent
attached to N4 as well as in the vicinity of N2. Moreover,
an increase of the steric field in the region between N2
and S is detrimental for the inhibitory activity. The
influence exerted by the nature of the heteroaromatic
ring is reflected in the electrostatic contour shown in
Figure 5b, which indicates that a negatively charged
region around the ring is favorable for activity.
Because of the availability of X-ray crystallographic
data for GSK-3, a mapping study was performed using
the docking module of CMIP program.31 To validate this
computational approach, we first explored the docking
of three GSK3 inhibitors (I-5, indirubin-3′-monoxime,
and alsterpaullone; Figure 6), whose binding mode has
been determined by solving the X-ray structure of the
complexes with GSK-3â (PDB entries 1Q4L, 1Q4I, and
1Q3W).32 To avoid any bias toward the X-ray solution,
CMIP calculations were performed by using the X-ray
structure of the unbound enzyme (PDB entry 1I09).
Present calculations, therefore, represent a stringent
test because the side chains of certain residues (for
instance, Arg141) exhibit notable differences between

7108 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23
Martinez et al.
rected. Flash column chromatography was carried out at
medium pressure using silica gel (E. Merck, Grade 60, particle
size 0.040-0.063 mm, 230-240 mesh ASTM) and preparative
centrifugal circular thin-layer chromatography (CCTLC) on a
circular plate coated with a 1 mm layer of Kieselgel 60 PF254,
Merk, by using a Chromatotron with the indicated solvent as
eluent. Compounds were detected with UV light (254 nm). 1H
NMR spectra were obtained on Varian XL-300 and Gemini-
200 spectrometers working at 300 and 200 MHz, respectively.
Typical spectral parameters: spectral width 10 ppm, pulse
width 9 µs (57°), data size 32 K. 13C NMR experiments were
carried out on the Varian Gemini-200 spectrometer operating
at 50 MHz. The acquisiton parameters: spectral width 16 kHz,
acquisition time 0.99 s, pulse width 9 µs (57°), data size 32 K.
Chemical shifts are reported in values (ppm) relative to
internal Me4Si and J values are reported in Hertz. Elemental
analyses were performed by the analytical department at C.
N. Q. O. (CSIC), and the results obtained were within (0.4%
of the theoretical values.
General Method for the Synthesis of N-Alkyl-male-
imides. A 50 mL round botton flask was charged with Ph3P
to which was added 25 mL of dry THF. The resulting clear
solution was cooled to -70 °C, under nitrogen atmosphere.
DIAD or DEAD depending of each case was added over 2-3
min. The yellow reaction mixture was stirred 5 min after which
the corresponding alkyl alcohol was added over 1 min and
stirred for 5 min. Then, maleimide was added to the reaction
mixture as a solid. The resulting suspension was allowed to
remain at -70 °C for 5 min during which time most of the
maleimide dissolved. The cooling bath was then removed, and
the reaction was stirred overnight at ambient temperature.
The solvent was evaporated to dryness in vacuo and the
residue purified by silica gel column chromatography using
as eluent mixtures of solvents in the proportions indicated for
each particulate case.
N-(Ethoxycarbonylmethyl)-maleimide (4). Reagents:
Ph3P (1.31 g, 5 mmol), DEAD (0.8 mL, 5 mmol), ethyl glycolate
(0.71 mL, 7.5 mmol), and maleimide (0.48 g, 5 mmol). Condi-
tions: room temperature, overnight. Purification: silica gel
column chromatography using AcOEt/hexane (1:3). Yield: 0.30
g (33%) as a colorless oil. 1H NMR (CDCl3): 1.25 (t, 3H,
CH2CO2CH2CH3, J ) 7.1 Hz); 4.20 (q, 2H, CH2CO2CH2CH3, J
) 7.1 Hz); 4.24 (s, 2H,. CH2CO2CH2CH3); 6.76 (d, 2H, CHdCH,
J ) 6.4 Hz). 13C NMR (CDCl3): 13.7 (CH2CO2CH2CH3); 38.4
(CH2CO2CH2CH3); 61.5 (CH2CO2-CH2CH3); 134.3 (CdC); 166.9
(CO2); 169.6 (CdO ). Anal. (C8H9NO4) C, H, N, S.
Figure 7. Preferred binding sites determined from CMIP
mapping studies for TDZD inhibitor 45 onto the X-ray crystal-
lographic structure of GSK3 (PDB entry 1I09).
TDZD scaffold for inhibitory activity, one might re-
interpret the electrostatic requirement found in the
CoMFA analysis in terms of the chemical susceptibility
of TDZDs to react with the thiol group of Cys199, which
lies close to the heteroaromatic ring of TDZD in some
of the structures docked into the ATP-binding site. This
finding would explain the apparent GSK3 specificity
exhibited by TDZDs. During the revision of this manu-
script, “selectivity filters” in the ATP-binding pocket of
protein kinases discovered by bioinformatics techniques
have been described.34 One of them is a reactive cysteine
residue within the active site. These data supports our
preliminary conclusion.
Conclusion
N-(p-Methoxy-benzyl)-maleimide (5). Reagents: Ph3P
(1.31 g, 5 mmol), DEAD (0.8 mL, 5 mmol), p-methoxybenzyl
alcohol (0.93 mL, 7.5 mmol), and maleimide (0.48 g, 5 mmol).
Conditions: room temperature, overnight. Purification: silica
gel column chromatography using AcOEt/hexane (1:3). Yield:
0.50 g (46%) as white solid; mp 99-102 °C. 1H NMR (CDCl3):
3.74 (s, 3H, OCH3); 4.58 (s, 2H, CH2Ph); 6.65 (d, 2H, CHdCH,
J ) 6.4 Hz); 6.8-7.2 (m, 4H, arom). 13C NMR (CDCl3): 40.4
(CH2Ph-OCH3); 54.8 (OCH3); 113.6; 128.2; 129.5; 158.8 (C
arom); 133.8 (CdC); 170.1 (CdO) Anal. (C12H11NO3) C, H, N,
S.
N-Phenethyl-maleimide (6). Reagents: Ph3P (0.65 g, 2.5
mmol), DEAD (0.4 mL, 2.5 mmol), phenethyl alcohol (0.5 mL,
3.75 mmol), and maleimide (0.24 g, 2.5 mmol). Conditions:
room temperature, overnight. Purification: silica gel column
chromatography using AcOEt/hexane (1:4). Yield: 0.15 g (30%)
as a white-yellow solid; mp 109-110 °C (lit.: 112 °C).21 1H
NMR (CDCl3): 2.90 (m, 2H, CH2CH2Ph); 3.77 (m, 2H,
CH2CH2Ph); 6.66 (d, 2H, CHdCH, J ) 6.4 Hz); 7.19-7.32
(m, 5H, arom). 13C NMR (CDCl3): 34.5 (CH2CH2Ph); 38.9
(CH2CH2Ph); 126.6; 128.5; 128.7; 137.8 (C arom); 133.9 (CdC);
170.5 (CdO). Anal. (C12H11NO2) C, H, N, S.
Inspection of the results obtained for the different
series of TDZD-related compounds examined here points
out the important role played by the nature of the
heteroaromatic ring in modulating the GSK3 inhibitory
activity of these compounds. After an extensive SAR
study among different heterocyclic families, TDZDs have
been revealed as a privileged scaffold for the selective
inhibition of GSK-3. Moreover, the SAR study also
proves the influence played by aromatic substituents
attached to the two nitrogen atoms of the TDZD ring.
Mapping studies permit us to hypothesize two binding
modes, which in turn might imply relevant differences
in the mechanism that underly the inhibitory activity
of TDZDs. Further studies, including directed mutagen-
esis and proteomic analysis, are in progress to clarify
the mechanistic implications provided by these results.
In the meantime, TDZDs have proved their in vivo
activity, reducing the hyperphosphorylation of tau
protein in a tet/GSK-3 double transgenic model,35 being
valuables candidates for drug development in the next
future Alzheimer’s therapy.
N-(p-Methoxy-phenethyl)-maleimide (7). Reagents: Ph3P
(1.31 g, 5 mmol), DEAD (0.8 mL, 5 mmol), p-methoxyphenethyl
alcohol (1.2 g, 7.5 mmol), and maleimide (0.48 g, 5 mmol).
Conditions: room temperature, overnight. Purification: silica
gel column chromatography using AcOEt/hexane (1:5). Yield:
0.71 g (60%) as a yellow solid; mp 79-81°C. 1H NMR (CDCl3):
Experimental Section
Chemical Procedures. Melting points were determined
with a Reichert-Jung Thermovar apparatus and are uncor-

SAR and 3D-QSAR Studies on Thiadiazolidinones
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23 7109
2.80 (m, 2H, CH2CH2Ph); 3.70 (m, 2H, CH2CH2Ph); 3.75 (s,
3H, OCH3); 6.63 (d, 2H, CHdCH, J ) 6.4 Hz); 6.8-7.1 (m,
4H, arom). 13C NMR (CDCl3): 33.5 (CH2CH2Ph-OCH3); 39.2
(CH2CH2Ph-OCH3); 55.5 (OCH3); 114.0; 129.7; 129.9; 158.4 (C
arom); 133.9 (CdC); 170.5 (CdO). Anal. (C13H13NO3) C, H, N,
S.
N-(3-Phenyl-propyl)-maleimide (8). Reagents: Ph3P (0.65
g, 2.5 mmol), DIAD (0.5 mL, 2.5 mmol), 3-phenyl-1-propanol
(0.48 mL, 3.75 mmol), and maleimide (0.24 g, 2.5 mmol).
Conditions: room temperature, overnight. Purification: silica
gel column chromatography using AcOEt/hexane (1:4). Yield:
0.20 g (37%) as a white solid; mp 79-80 °C. 1H NMR (CDCl3):
1.92 (q, 2H, CH2CH2CH2Ph, J ) 7.1 Hz); 2.60 (t, 2H,
CH2CH2CH2Ph, J ) 7.1 Hz); 3.55 (t, 2H, CH2CH2CH2Ph, J )
7.1 Hz); 6.27 (d, 2H, CHdCH, J ) 6.4 Hz); 7.12-7.28 (m, 5H,
arom). 13C NMR (CDCl3): 29.6 (CH2CH2CH2Ph); 32.8
(CH2CH2CH2Ph); 37.4 (CH2CH2CH2Ph); 125.8; 128.1; 128.2;
140.7 (C arom); 133.7 (CdC); 170.6 (CdO) Anal. (C13H13NO2)
C, H, N, S.
N-(5-Phenyl-pentanyl)-maleimide (9). Reagents: Ph3P
(0.65 g, 2.5 mmol), DIAD (0.5 mL, 2.5 mmol), 5-phenyl-1-
pentanol (0.63 mL, 3.75 mmol), and maleimide (0.24 g, 2.5
mmol). Conditions: room temperature, overnight. Purifica-
tion: silica gel column chromatography using AcOEt/hexane
(1:4). Yield: 0.32 g (52%) as white-yellow solid; mp 49-51 °C.
1H NMR (CDCl3): 1.20-138 (m, 2H, CH2CH2CH2CH2-CH2Ph);
1.52-2.02 (m, 4H, CH2CH2CH2CH2CH2Ph); 2.57 (t, 2H,
CH2CH2CH2CH2CH2Ph, J ) 7.3 Hz); 3.5 (t, 2H, CH2CH2CH2-
CH2CH2Ph, J ) 7.3 Hz); 6.65 (d, 2H, CHdCH, J ) 6.4 Hz);
7.11-7.28 (m, 5H, arom). 13C NMR (CDCl3): 25.9 (CH2-
CH2CH2CH2CH2Ph); 28.0 (CH2CH2CH2CH2CH2Ph); 30.6
(CH2CH2CH2CH2CH2Ph); 35.4 (CH2CH2CH2CH2CH2Ph); 37.3
(CH2CH2CH2CH2CH2Ph); 125.4; 127.9; 128.0; 142.0 (C arom);
133.6 (CdC); 170.5 (CdO). Anal. (C15H17NO2) C, H, N, S.
General Method for the Synthesis of Alkyl-rhoda-
nines. Method A: To an equimolecular solution of rhodanine
with a base, 1 equiv of alkyl agent (alkyl halide) was added,
with the conditions indicated for each particular case. The
solvent was evaporated to dryness in vacuo and the residue
purified by silica gel column chromatography using the ap-
propriated eluent.
Method B: A 50 mL round-bottom flask was charged with
Ph3P to which was added 25 mL of dry THF. The resulting
clear solution was cooled to -70 °C, under nitrogen atmo-
sphere. DIAD or DEAD depending of each case was added over
2-3 min. The yellow reaction mixture was stirred 5 min after
which the corresponding alkyl alcohol was added over 1 min
and stirred for 5 min. Then, rhodanine was added to the
reaction mixture as solid. The resulting suspension was
allowed to remain at -70 °C for 5 min during, which time most
of the rhodanine dissolved. The cooling bath was then removed,
and the reaction was stirred overnight at ambient tempera-
ture. The solvent was evaporated to dryness in vacuo and the
residue purified by silica gel column chromatography using
as eluent mixtures of solvents in the proportions indicated for
each particulate case.
N-Benzyl-rhodanine (22). Method B: Reagents: Ph3P
(1.31 g, 5 mmol), DIAD (1 mL, 5 mmol), benzyl alcohol (0.78
mL, 7.5 mmol), and rhodanine (0.67 g, 5 mmol). Conditions:
12 h at room temperature. Purification: silica gel column
chromatography using AcOEt/hexane (1:4). Yield: 0.25 g (22%)
as yellow solid, mp 84-85 °C (lit.: 85 °C).36 1H NMR (CDCl3):
3.9 (s, 2H); 5.2 (s, 2H, CH2Ph); 7.3-7.4 (m, 5H, arom). 13C NMR
(CDCl3): 35.4 (CH2); 47.6 (CH2Ph); 128.2; 128,6; 129.1; 134.7
(C arom); 153.8 (CdO); 173.8 (CdS). Anal. (C10H9NS2O) C, H,
N, S.
(CH2CH2Ph); 35.3 (CH2); 45.7 (CH2CH2Ph); 126.8; 128,6; 128.6;
137.4 (C arom); 173.5 (CdO); 200.9 (CdS). Anal. (C11H11NS2O)
C, H, N, S.
N-Phenacyl-rhodanine (24). Method A: Reagents: Rhoda-
nine (0.13 g, 1 mmol), K2CO3 (exceed), and acetophenone
bromide (0.2 g, 1 mmol) in 25 mL of acetone. Conditions:
Stirred at room temperature for 3 h. Isolation: The carbonate
was filtered and the solvent evaporated to dryness in vacuo.
Purification: preparative centrifugal circular thin-layer chro-
matography (CCTLC) using CH2Cl2. Yield: 0.04 g (15%) as
brown oil. 1H NMR (CDCl3): 3.9 (s, 2H); 4.2 (s, 2H, CH2COPh);
7.4-7.9 (m, 5H, arom). 13C NMR (CDCl3): 37.6 (CH2); 45.3
(CH2COPh); 128.6; 128,7; 133.5; 135.3 (C arom); 170.5 (Cd
O); 194.1 (CH2COPh); 197.6 (CdS). Anal. (C11H9NS2O2) C, H,
N, S.
4-Benzyl-2,4-thiazolidine-3,5-dione (25). Method A: Re-
agents: 2,4-Thiazolidinedione (0.12 g, 1 mmol), sodium hy-
dride (1 mmol, 0.04 g), and benzyl bromide (0.12 mL, 1 mmol)
in 15 mL of DMF. Conditions: Stirred at room temper-
ature for 8 h. Isolation: the solvent evaporated to dryness in
vacuo. Purification: silica gel column chromatography using
1
CH2Cl2.Yield: 0.14 g (66%) as white solid. mp 39-41 °C. H
NMR (CDCl3): 3.8 (s, 2H, CH2); 4.7 (s, 2H, CH2Ph); 7.2-7.3
(m, 5H, arom); 13C NMR (CDCl3): 33.6 (CH2); 45.1 (CH2Ph);
128.1; 128.5; 128.7; 134.9 (C arom); 171.0 (4-CdO); 171.5 (2-
CdO). Anal. (C10H9NO2S) C, H, N, S.
N-(p-Methoxy-benzyl)-rhodanine (26). Method B: Re-
agents: Ph3P (0.65 g, 2.5 mmol), DEAD (0.4 mL, 2.5 mmol),
p-methoxybenzyl alcohol (0.5 mL, 3.75 mmol), and rhodanine
(0.33 g, 2.5 mmol). Conditions: room temperature, overnight.
Purification: silica gel column chromatography using AcOEt/
hexane (1:3). Yield: 0.15 g (32%) as a yellow oil; 1H NMR
(CDCl3): 3.5 (s, 3H, CH2Ph-p-OCH3); 3.8 (s, 2H); 4.9 (s, 2H,
CH2Ph-p-OCH3); 7.5-7.8 (m, 5H, arom). 13C NMR (CDCl3):
37.5 (CH2); 46.5 (CH2Ph-p-OCH3); 56.3 (CH2Ph-p-OCH3); 113.9;
126.5; 128.7; 159.8 (C arom); 171.2 (CdO); 197.6 (CdS). Anal.
(C10H9NO2S) C, H, N, S.
N-(p-Methoxy-phenethyl)-rhodanine (27). Method B:
Reagents: Ph3P (0.65 g, 2.5 mmol), DIAD (0.4 mL, 2.5 mmol),
p-methoxyphenethyl alcohol (0.3 mL, 5 mmol), and rhodanine
(0.33 g, 2.5 mmol). Conditions: room temperature, overnight.
Purification: silica gel column chromatography using AcOEt/
hexane (1:4). Yield: 0.22 g (33%), as a brown solid. mp 104-
105 °C. 1H NMR (CDCl3): 2.8 (m, 2H, CH2CH2Ph); 3.7 (s, 3H,
OCH3); 3.9 (s, 2H, CH2); 4.1 (s, 2H, CH2CH2Ph); 6.8-7.2
(m, 5H, arom, J1 ) 2.8, J2 ) 8.6). 13C NMR (CDCl3): 31.5
(CH2CH2Ph); 35.1 (CH2CH2Ph); 45.6 (CH2); 55.1 (OCH3); 113.9;
129.2; 129.7; 158.9 (C arom); 173.4 (CdO); 200.8 (CdS). Anal.
(C12H13NO2S2) C, H, N, S.
General Method for the Synthesis of Alkyl-dithiazoli-
dinediones. To a solution of N-alkyl-formamide in dry toluene
under nitrogen atmosphere, 2 equiv of chlorocarbonylsulfenyl
chloride was slowly added. The resulting mixture was stirred
at room temperature for 12 h. Then, the solvent was evapo-
rated to dryness in vacuo and the residue purified by silica
gel column chromatography using the appropriate eluent.
4-Methyl-1,2,4-dithiazolidine-3,5-dione (28). Reagents:
N-Methyl-formamide (0.23 mL, 4 mmol) and chlorocarbonyl-
sulfenyl chloride (0.676 mL, 8 mmol) in 20 mL of dry toluene.
Conditions: Stirred at room temperature for 12 h. Purifica-
tion: silica gel column chromatography using AcOEt/hexane
(1:4) as eluent. Yield: 0.15 g (32%) as yellow solid. mp 37-38
°C. (lit.: 38-39 °C).37 1H NMR (CDCl3): 3.2 (s, 3H, CH3);13C
NMR (CDCl3): 32.0 (CH3); 167.1 (CdO). Anal. (C3H3NO2S2)
C, H, N, S.
4-Phenyl-1,2,4-dithiazolidine-3,5-dione (29). Reagents:
formanilide (0.49 g, 4 mmol) and chlorocarbonylsulfenyl
chloride (0.676 mL, 8 mmol) in 20 mL of dry toluene. Condi-
tions: Stirred at room temperature for 12 h. Purification: silica
gel column chromatography using AcOEt/hexane (1:15) as
eluent. Yield: 0.15 g (17%) as a white-yellow solid; mp 165-
167 °C (lit.: 168-169 °C).371H NMR (CDCl3): 7.14-7.5 (m,
5H, arom). 13C NMR (CDCl3): 28.1 (2 CH2); 42.3 (CH2Ph);
N-Phenethyl-rhodanine (23). Method B: Reagents: Ph3P
(1.31 g, 5 mmol), DIAD (1 mL, 5 mmol), phenethyl alcohol (0.78
mL, 7.5 mmol), and rhodanine (0.67 g, 5 mmol). Conditions:
Stirred at room temperature for 12 h. Purification: silica gel
column chromatography using AcOEt/hexane (1:4). Yield: 0.55
g (42%) as yellow solid. mp 108-110 °C. 1H NMR (CDCl3): 2.9
(t, 2H, CH2CH2Ph, J ) 8.1), 3.9 (s, 2H); 4.2 (s, 2H, CH2CH2Ph,
J ) 8.1); 7.4-7.9 (m, 5H, arom). 13C NMR (CDCl3): 32.6

7110 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23
Martinez et al.
127.5; 129,7; 129.9; 134.2 (C arom); 166.9 (2 CdO). Anal. (C8H5-
NO2S2) C, H, N, S.
10 h. Purification: silica gel column chromatography using
AcOEt/hexane (1:4). Yield: 0.26 g (11%) as a yellow solid; mp
1
117-118 °C. H NMR (CDCl3): 1.34 (t, 3H, CH2CH3, J ) 7.1
4-Benzyl-1,2,4-dithiazolidine-3,5-dione (30). Reagents:
N-Benzylformamide (540 mg, 4 mmol) and chlorocarbonyl-
sulfenyl chloride (0.676 mL, 8 mmol) in 20 mL of dry toluene.
Conditions: Stirred at room temperature for 12 h. Purifica-
tion: silica gel column chromatography using AcOEt/hexane
(1:15) as eluent. Yield: 146 mg (17%), as a white solid. mp
90-92 °C (Lit.: 93-94 °C).37 1H NMR (CDCl3): 7.4-7.3 (m,
5H, arom); 4.9 (s, 2H, CH2Ph). 13C NMR (CDCl3): 49.3
(CH2Ph); 133.9; 128.8; 128.6; 128.5 (C arom); 167.1 (CdO).
1,2-Dimethyl-4-Phenyl-1,2,4-triazolidine-3,5-dione (32).
Reagents: 4-Phenyl-urazole (31) (0.09 g, 1 mmol), K2CO3
(exceed), and iodomethane (0.14 g, 1 mmol) in 25 mL of
acetone. Conditions: Refluxed for 6 h. Isolation: The carbonate
was filtered and the solvent evaporated to dryness in vacuo.
Purification: preparative centrifugal circular thin-layer chro-
matography (CCTLC) using CH2Cl2. Yield: 0.06 g (56%) as a
white solid; mp 121-123 °C (lit.: 123-124 °C).38 1H NMR
(CDCl3): 3.2 (s, 6H, 2 CH3); 7.3-7.5 (m, 5H, arom). 13C NMR
(CDCl3): 32.3 (2 CH3); 125.4; 128.0; 129.0; 131.5 (C arom);
153.5 (CdO). Anal. (C10H11N3O2) C, H, N, S.
Hz); 3.77 (c, 2H, CH2CH3, J ) 7.1 Hz); 7.6-8.4 (m, 5H, arom).
13C NMR (CDCl3): 13.6 (CH2CH3); 40.3 (CH2CH3); 124.3; 127.6;
137.9; 147.1 (C arom.); 150.9 (3-CdO); 164.8 (5-CdO) Anal.
(C10H9N3SO4) C, H, N, S.
2-Ethyl-4-phenyl-1,2,4-thiadiazolidine-3,5-dione (58).
Reagents: Phenyl isothiocyanate (0.78 mL, 6.5 mmol), 35%
HCl (3.1 mL), KMnO4 (0.5 g), and ethyl isocyanate (0.51 mL,
6.5 mmol). Conditions: room temperature, 8 h. Isolation:
filtration of reaction mixture. Purification: recrystallization
from methanol. Yield: 0.43 g (30%) as a white solid; mp 174-
1
179 °C. H NMR (CDCl3): 1.34 (t, 3H, CH2CH3 , J ) 7.2 Hz);
3.76 (c, 2H, CH2CH3, J ) 7.2 Hz); 7.34-7.50 (m, 5H, arom.).
13C NMR (CDCl3): 13.9 (CH2CH3 ); 40.3 (CH2CH3);127.1; 129.0;
129.2; 132.6 (C arom.); 151.9 (3-CdO); 165.3 (5-CdO). Anal.
(C10H10N2SO2) C, H, N, S.
4-(Ethoxycarbonylmethyl)-2-methyl-1,2,4-thiadiazoli-
dine-3,5-dione (59). Reagents: Ethyl isothiocyanatoacetate
(0.8 mL, 6.5 mmol), 35% HCl (3.1 mL), KMnO4 (0.5 g), and
methyl isocyanate (0.38 mL, 6.5 mmol). Conditions: room
temperature, 8 h. Isolation: filtration of reaction mixture.
Purification: recrystallization from hexane. Yield 0.28 g (20%)
as white solid; mp 67-69 °C; 1H NMR (CDCl3): 1.3 (t, 3H,
CH2CO2CH2CH3, J ) 7.1 Hz); 3.2 (s, 3H, CH3); 4.2 (c, 2H,
CH2CO2CH2CH3, J ) 7.1 Hz); 4.4 (s, 2H,. CH2CO2CH2CH3).
13C NMR (CDCl3): 14.0 (CH2CO2CH2CH3); 31.5 (CH3); 42.7
(CH2CO2CH2CH3); 62.1 (CH2CO2CH2CH3); 152.6 (3-CdO);
166.4 (5-CdO); 166.4 (CO2 ). Anal. (C7H10N2SO3) C, H, N, S.
Biological Procedures. GSK-3 Inhibition Assay. GSK-3
(rabbit GSK-3, Sigma) was incubated at 37 °C for a 20 min
period in buffer [Tris pH ) 7.5 (50 mM), EDTA (1 mM), EGTA
(1 mM), and MgCl2 (10 mM)] supplemented by the synthetic
peptide GS-1 (15 µM final concentration) as substrate, ATP
(15 µM), [γ-32P]ATP (0.2 µCi), and different concentrations of
the product to be tested.39 After that, aliquots of the reaction
mixtures are added on phosphocellulose p81 papers. These
papers are washed three times with phosphoric acid 1%, and
the radioactivity incorporated in the GS-1 peptide is measured
in a liquid scintillation counter. Dose-response profiles were
generated, and the IC50 value (concentration at which a 50%
of enzyme inhibition is shown) for inhibition of GSK-3 by the
test compound was calculated using a four-parameter logistic
function.
Selectivity Panel Protein Kinase Assays. The pan-
kinase screen was performed at CEREP (France). The mem-
bers of the kinase selectivity panel were assessed as was
described previously: Abl kinase assay,40 CAM kinase assay,41
EGF-tyrosine kinase assay,42 IRK assay,43 MAP kinase (ERK
42) assay,44 MEK 1 kinase assay,45 and protein kinase p56
assay.46
The activated protein kinases were assayed for their ability
to phosphorylate the appropriate peptide/protein substrate in
the presence of 10 µM of test substance, respectively. Results
are given as the mean of duplicate determinations for %
inhibition compared to control incubations, in which the
inhibitors were substituted with DMSO and reference com-
pound (staurosporine).
1,2-Dibenzoyl-4-phenyl-1,2,4-triazolidine-3,5-dione (33).
Reagents: 4-phenyl-urazole (31) (0.18 g, 1 mmol), sodium
hydride (0.08 g, 2 mmol), and benzoyl chloride (0.18 mL, 1.5
mmol) in 15 mL of dimethylformamide (DMF). Conditions:
Stirred at room temperature for 6 h. Purification: silica gel
column chromatography using AcOEt/hexane (1:2) as eluent.
Yield: 0.09 g (24%). mp 276-280 °C (Lit.: 272-273 °C),38 1
H
NMR (CDCl3): 7.4-7.5 (m, 5H, Ph); 7.5-8.1 (m, 10H, 2 × Bz).
13C NMR (CDCl3): 126.0; 128.2; 130.9; 134.4 (C arom Ph);
129.5; 128.4; 130.2; 133.7 (C arom Bz); 164.5 (CdO); 172.3
(COPh). Anal. (C22H15N3O4) C, H, N, S.
1,2-Bis-(phenylcarbonylmethyl)-4-phenyl-1,2,4-triazo-
lidine-3,5-dione (34). Reagents: 4-phenylurazole (31) (0.35
g, 2 mmol), potassium hydroxide (0.11 g, 2 mmol), and
acetophenone bromide (0.4 g, 2 mmol) in 15 mL of ethanol.
Conditions: Refluxed for 10 h. Isolation: filtration of reaction
mixture. Purification: recrystallization from EtOH/ CH2Cl2.
Yield: 0.24 g (28%), as a white solid. mp 208-210 °C (lit.:
205-206 °C).38 1H NMR (CDCl3): 5.0 (s, 4H, CH2COPh); 7.4-
7.9 (m, 15 H, arom). 13C NMR (CDCl3): 53.0 (CH2COPh); 126.0;
128.7; 128.4; 128.9; 129.2; 131.6; 134.1; 134.2 (C arom); 155.6
(CdO); 192.0 (COPh). Anal. (C24H20N3O4) C, H, N, S.
4-Benzyl-hydantoin (37). Reagents: Hydantoin (0.1 g, 1
mmol), sodium hydride (0.04 g, 1 mmol) and benzyl bromide
(0.12 mL, 1 mmol) in 15 mL of DMF. Conditions: Stirred at
room temperature for 12 h. Isolation: Evaporated to dryness
in vacuo the solvent. Purification: silica gel column chroma-
tography using AcOEt/hexane (1:1). Yield: 0.1 g (50%) as white
solid; mp 138-139 °C. 1H NMR (CDCl3): 3.9 (s, 2H, CH2); 4.6
(s, 2H, CH2Ph); 5.4 (br, 1H, NH); 7.3-7.4 (m, 5H, arom). 13C
NMR (CDCl3): 42.0 (CH2); 46.5 (CH2Ph); 127.9; 128.3; 128.6;
135.8 (C arom.); 158.5 (2-CdO); 171.2 (4-CdO). Anal. (C10H10-
N2O2) C, H, N
General Procedure for the Synthesis of 1,2,4-Thiadia-
zolidine-3,5-dione. Chlorine was bubbled slowly through a
solution of aryl or alkyl isothiocyanate in dry hexane (25 mL),
under nitrogen atmosphere, at -15 °C to -10 °C. Chlorine
was generated by the addition of 35% HCl to KMnO4. The
temperature of the reaction mixture was carefully controlled
during the addition step. At this point, the N-aryl- or N-alkyl-
S-chloroisothiocarbamoyl chloride was formed. Afterward,
alkyl isocyanate was added, yielding, immediately, sparingly
soluble 3-oxathiadiazolium salts. The mixture was stirred at
room temperature between 8 and 10 h, producing the hydroly-
sis of these heterocyclic salts. After this time, the resulting
product was purified by suction filtration and recrystallization
or silica gel column chromatography using the appropriate
eluent.
CoMFA Analysis. The compounds were built from frag-
ments in the Sybyl database. Each structure was initially fully
optimized using the standard Tripos molecular mechanics force
field using the default bond distances and angles as well as
Gasteiger-Marsili charges.47 Conformational analysis was
carried out to define the low-energy conformations. The
structures were fully geometrically optimized with the semi-
empirical AM1 Hamiltonian48 using MOPAC-6.0.49 TDZD 43
was used as template molecule on which to align the other
compounds using the minimum-energy conformation obtained
as described above.
2-Ethyl-4-(4-nitrophenyl)-1,2,4-thiadiazolidine-3,5-di-
one (55). Reagents: 4-Nitrophenyl isothiocyanate (1.17 g, 6.5
mmol), 35% HCl (3.1 mL), KMnO4 (0.5 g), and ethyl isocyanate
(0.51 mL, 6.5 mmol) in THF. Conditions: room temperature,
The CoMFA analysis is employs carried out using a grid
spacing of 2.0 Å in the x, y, and z directions, and the grid region
was automatically generated to encompass all molecules with
an extension of 4.0 Å in each direction. A sp3 carbon and a

SAR and 3D-QSAR Studies on Thiadiazolidinones
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23 7111
(9) Leost, M.; Schultz, C.; Link, A.; Wu, Y. Z.; Biernat, J.; Man-
delkow, E. M.; Bibb, J. A.; Snyder, G. L.; Greengard, P.;
Zaharevitz, D. W.; Gussio, R.; Senderowicz, A. M.; Sausville, E.
A.; Kunick, C.; Meijer, L. Paullones are potent inhibitors of
glycogen synthase kinase-3beta and cyclin-dependent kinase
5/p25. Eur. J. Biochem. 2000, 267, 5983-5994.
(10) Leclerc, S.; Garnier, M.; Hoessel, R.; Marko, D.; Bibb, J. A.;
Snyder, G. L.; Greengard, P.; Biernat, J.; Wu, Y. Z.; Mandelkow,
E. M.; Eisenbrand, G.; Meijer, L. Indirubins inhibit glycogen
synthase kinase-3 beta and CDK5/p25, two protein kinases
involved in abnormal tau phosphorylation in Alzheimer’s dis-
ease. A property common to most cyclin-dependent kinase
inhibitors. J. Biol. Chem. 2001, 276, 251-260.
(11) Smith, D. G.; Buffet, M.; Fenwick, A. E.; Haigh, D.; Ife, R. J.;
Saunders: M.; Slingsby, B. P.; Stacey, R.; Ward, R. W. 3-Anilino-
4-arylmaleimides: Potent and selective inhibitors of Glycogen
synthase kinase-3 (GSK-3). Bioorg. Med. Chem. Lett. 2001, 11,
635-639
(12) Martinez, A.; Alonso, M.; Castro, A.; Perez, C.; Moreno, F. J.
First non-ATP competitive glycogen synthase kinase 3 beta
(GSK-3beta) inhibitors: thiadiazolidinones (TDZD) as potential
drugs for the treatment of Alzheimer’s disease. J. Med. Chem.
2002, 45, 1292-1299.
(13) Ter Haar, E.; Coll, J. T.; Austen, D. A.; Hsiao, H. M.; Swenson,
L.; Jain, J. Structure of GSK-3beta reveals a primed phospho-
rylation mechanism. Nat. Struct. Biol. 2001, 8, 593-596.
(14) Dajani, R.; Fraser, E.; Roe, S. M.; Young, N.; Good, V.; Dale, T.
C.; Pearl, L. H. Crystal Structure of Glycogen Synthase Kinase
3â: Structural Basis for Phosphate-Primed Substrate Specificity
and Autoinhibition. Cell 2001, 105, 721-732.
(15) Bertrand, J. A.; Thieffine, S.; Vulpetti, A.; Cristiani, C.; Valsa-
sina, B.; Knapp, S.; Kalisz, H. M.; Flocco, M. Structural
characterization of the GSK-3 active site using selective and
nonselective ATP-mimetic inhibitors. J. Mol. Biol. 2003, 33,
393-407.
(16) Bhat, R.; Xue, Y.; Berg, S.; Hellberg, S.; Ormo, M.; Nilsson, Y.;
Radesater, A. C.; Jerning, E.; Markgren, P. O.; Borgegard, T.;
Nylof, M.; Gimenez-Cassina, A.; Hernandez, F.; Lucas, J. J.;
Diaz-Nido, J.; Avila, J. Structural insights and biological effects
of glycogen synthase kinase 3-specific inhibitor AR-A014418. J.
Biol. Chem. 2003, 278, 45937-45945.
(17) Meijer, L.; Skaltsounis, A. L.; Magiatis, P.; Polychronopoulos,
P.; Knockaert, M.; Leost, M.; Ryan, X. P.; Vonica, C. A.;
Brivanlou, A.; Dajani, R.; Crovace, C.; Tarricone, C.; Musacchio,
A.; Roe, S. M.; Pearl, L.; Greengard, P. GSK-3 selective inhibitors
derived from Tyrian purple indirubins. Chem. Biol. 2003, 10,
1255-1266.
(18) Polychronopoulos, P.; Magiatis, P.; Skaltsounis A.; Myriantho-
poulos, V.; Mikros, E.; Tarricone, A.; Musacchio, A.; Roe, S. M.;
Pearl, L.; Leost, M., Greengard, P.; Meijer, L. Structural basis
for the synthesis of indirubins as potent and selective inhibitors
of glycogen synthase kinase-3 and cyclin-dependent kinases. J.
Med. Chem. 2004, 47, 935-946.
(19) Alonso, M.; Mart´ınez, A. GSK-3 inhibitors: discoveries and
developments. Curr. Med. Chem. 2004, 11, 753-761.
(20) Cramer, R. D., III; Patterson, D. E.; Bunce, J. D. Comparative
molecular field analysis (ComFA). 1. Effect of shape on binding
of steroids to carrier proteins. J. Am. Chem. Soc. 1988, 110,
5959-5967.
(21) Walker, A. M. A high yielding synthesis of N-alkyl maleimides
using a novel modification of the Mitsunobu reaction. J. Org.
Chem. 1995, 60, 5352-5355.
(22) Zumach, G.; Weiss, W.; Ku¨hle, E. Belgian Pat. 682820. Farben-
fabriken Bayer A. G., June 1966.
(23) MacLauchlin, C.; May, I. H.; Izydore, R. A. Synthesis and
Cytotoxic action of 1-Oxoalkyl and 1,2-Dioxoalkyl-1,2,4-triazo-
lidine-3,5-diones in Murine and Human tissue cultured Cells.
Arch. Pharm. Pharm. Med. Chem. 1999, 332, 225-232.
(24) Hoshitsugu, A.; Matsui, M.; Fujii, A.; Kontani, T.; Koshiyuki,
O.; Koizumi, T.; Shiro. M. Asymmetric Diels-Alder Reaction of
optically active R-(2-exo-Hydroxy-O-bornyl)-sulfinylmaleimides
and its applicattion to optically active 5-functionalized Pyrrolines
via Retro-Diels-Alder Reaction. J. Chem. Soc., Perkin. Trans.
1 1994, 25-39.
(25) Corrie, J. E. T. Thiol-reactive Fluorescent Probes for Protein
Labelling. J. Chem. Soc., Perkin. Trans. 1 1994, 2975-2982.
(26) Janda, K. D.; Ashley, J. A.; Jones, T. M.; McLeod, D. A.;
Schloeder, D. M.; Weinhous. M. I. Inmobilized Catalytic Anti-
bodies in Aqueous and Organic Solvents. J. Am. Chem. Soc.
1990, 112, 8886-8888.
(27) Foye, W. O.; Tovivich, P. N-Glucopyranosyl-5-Aralkylidene-
rhodanines: Synthesis and Antibacterial and Antiviral Activi-
ties. J. Pharm. Sci. 1977, 66, 1607-1611.
(28) Sudo, K.; Matsumoto, Y.; Matsushima, M.; Fujiwara, M.; Konno,
K.; Shimotohno, K.; Shigeta, S.; Yokota, T. Novel Hepatitis C
Virus Protease Inhibitors: Thiazolidine Derivatives. Biochem.
Biophys. Res. Commun. 1997, 238, 643-647.
charge +1 were used as probes to generate the interaction
energies at each lattice point. The default value of 30 kcal.mol-1
was used as the maximum electrostatic and steric energy
cutoff. Additionally, the molecular lipophilicity potential (MLP)
was determined by using the method of Gaillard50 to include
hydrophobic effects into the CoMFA study. Partial least square
(PLS) method was used to identify the relationship between
the inhibitory activity data and the field potential values. Final
non-cross-validated models were chosen on the basis of the best
combinations of mentioned three fields.
All molecular modeling techniques and CoMFA studies
described here were performed on Silicon Graphics workstation
using the Sybyl 6.6 molecular modeling software from Tripos,
Inc. St. Louis, MO.51
CMIP Calculations. Preferential binding sites of TDZDs
on GSK3 protein surface were obtained using the docking
module of the program CMIP.31 The protein (PDB entry 1I09)
was mapped onto a 3D grid of the appropriate size with a 0.5
Å regular spacing (ca. 4000000 grid positions). Binding was
assayed by exhaustive search of 8000 orientations (using
nonredundant Euler angles) for every grid position. When
necessary, flexible molecules were assayed as a family of
standard rotamers. Docked positions were scored using inter-
action energies between the ligand and the protein, which were
evaluated by adding electrostatic and van der Waals contribu-
tions (eq 1).
∆Gint ) ∆Gelect + ∆Gvdw
(1)
The solvent-screened electrostatic interaction was determined
by using the Mehler-Solmajer sigmoidal distance-dependent
dielectric model,52 and the van der Waals component was
evaluated by using a Lennard-Jones expression. Parameters
for protein atoms were obtained from the amber98 force field.53
For the ligand molecules, RESP atomic charges54 determined
at the HF/6-31G(d) level were used in conjunction with van
der Waals parameters taken for related atoms in the amber98
force field.
Acknowledgment. The financial support from
NeuroPharma is kindly acknowledged. Financial sup-
port from the Direccio´n General de Investigacio´n Cien-
t´ıfica y Te´cnica (SAF2002-04282) is acknowledged.
Supporting Information Available: Plots of the best
poses obtained from CMIP calculations for GSK-3â inhibitors.
This material is available free of charge via the Internet at
http://pubs.acs.org.
References
(1) Wauwe, J. V.; Haefner, B. Glycogen synthase kinase-3 as drug
target: From wallflower to center of attention. Drug New
Perspect. 2003, 16, 557-565.
(2) Cohen, P.; Goedert, M. GSK-3 inhibitors: Development and
therapeutic potential. Nat. Rev. Drug Discovery 2004, 3, 479-
487.
(3) Kaidanovich, O.; Eldar-Finkelman, H. The role of glycogen
synthase kinase-3 in insulin resistance and Type 2 diabetes.
Expert Opin. Ther. Targets 2002, 6, 555-561.
(4) Grimes, C. A.; Jope, R. S. The multifaceted roles of glycogen
synthase kinase 3beta in cellular signalling. Prog. Neurobiol.
2001, 65, 391-426.
(5) Hoeflich, K. P.; Luo, J.; Rubie, E. A.; Tsao, M. S.; Jin, O.;
Woodgett, J. Role of glycogen synthase kinase-3 in TNF-alpha-
induced NF-kappaB activation and apoptosis in hepatocytes.
Nature 2000, 406, 86-90.
(6) Martinez, A.; Castro, A.; Dorronsoro, I.; Alonso, M. Glycogen
Synthase Kinase 3 Inhibitors as new promising drugs for
diabetes, neurodegeneration, cancer, and inflamation. Med. Res.
Chem. 2002, 22, 373-384.
(7) Dorronsoro, I.; Castro, A.; Martinez, A. Inhibitors of glycogen
synthase kinase-3: Future therapy for unmet medical needs?.
Expert Opin. Ther. Pat. 2002, 12, 1527-1536
(8) Meijer, L.; Thunnissen, A. M.; White, A. W.; Garnier, M.; Nikolic,
M.; Tsai, L. H.; Walter, J.; Cleverley, K. E.; Salinas, P. C.; Wu,
Y. Z.; Biernat, J.; Mandelkow, E. M.; Kim, S. H.; Pettit, G. R.
Inhibition of cyclin-dependent kinases, GSK-3â and CK1 by
hymenaldisine, a marine sponge constituent. Chem. Biol. 2000,
7, 51-63.20

7112 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23
Martinez et al.
(29) Mishra, S.; Srivastava, S. K.; Srivastava, S. D. Synthesis of
5-arylidene-2-aryl-3-(phenothiazino/ benzo-triazoloacetamidyl)-
1,3-thiazolidine-4-ones as antiinflamatory, anticonvulsant, an-
algesic and antimicrobial agents. Indian. J. Chem. Sect. B. 1997,
36, 826-830.
(30) Peet, N. P. Genomics Research: Where are the drugs? Idrugs
2000, 3, 131-132.
(31) Gelpi, J. L., Kalko, S. G., Barril, X., Cirera, J., de la Cruz, X.,
Luque, F. J., Orozco, M. Classical molecular interaction poten-
tials: improved setup procedure in molecular dynamics simula-
tions of proteins. Proteins 2001, 45, 428-437.
protein kinase II by its ATP-binding domain. J. Neurochem.
2001, 76, 1066-1072.
(42) Carpenter, G.; King, L.; Cohen, S. Rapid enhancement of protein
phosphorylation in A-431 cell membrane preparation by epider-
mal growth factor. J. Biol. Chem. 1979, 254, 4884-4891.
(43) Al-Hasani, H.; Pablack, W.; Klein, H. W. Phosphoryl exchange
is involved in the mechanism of the insulin receptor kinase.
FEBS Lett. 1994, 349, 17-22.
(44) Robbins, D. J.; Zhen, E.; Owaki, H. Regulation and properties
of extracellular signal-regulated protein kinases 1 and 2 in vitro.
J. Biol. Chem. 1993, 26, 5097-5106.
(45) Williams, D. H.; Wilkinson, S. E.; Purton, T.; Lamont, A.; Flotow,
A.; Murray, E. J. Ro 09-2210 exhibits potent anti-proliferative
effects on activated T cells by selectively blocking MKK activity.
Biochemistry 1998, 37, 9579-9585.
(46) Cheng, H. C.; Nishio, H.; Hatase, O.; Ralph, S.; Wang, J. H. A
synthetic peptide derived from p34cdc2 is a specific and efficient
substrate of Src-family tyrosine kinases. J. Biol. Chem. 1992,
13, 9248-9256.
(47) Gastaiger, J.; Marsili, M. Iterative partial equalization of orbital
electronegativity-rapid access to atomic charges. Tetrahedron
1980, 36, 3219-3228.
(48) Dewar, M. J. S., Zoebisch, E. G.; Healy, E. F.; Stewart, J. J.
AM1: a new general purpose quantum mechanical molecular
model. J. Am. Chem. Soc. 1985, 107, 3902-3909.
(49) Stewart, J. J. MOPAC: a semiempirical molecular orbital
program. J. Comput.-Aided Mol. Des. 1990, 4, 1-105.
(50) Gaillard, P., Carrupt, P. A.; Testa, B.; Boudon, A. Molecular
lipophilicity potential, a tool in 3D QSAR: method and applica-
tions. J. Comput.-Aided Mol. Des. 1994, 8, 83-96
(51) SYBYL Molecular Modelling Package, version 6.6; Tripos St.
Louis, MO.
(52) Mehler, E. L.; Solmajer, T. Electrostatic effects in proteins:
comparison of dielectric and charge models. Protein Eng. 1991,
4, 903-910.
(53) Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K.
M.; Fergurson, D. M.; Spellmeyer, D. C.; Fox, T.; Caldwell, J.
W.; Kollman, P. A. A second generation force field for the
simulation of proteins nucleic acids and organic molecules. J.
Am. Chem. Soc. 1995, 117, 5179-5197.
(54) Bayly, C. I.; Cieplak, P.; Cornell, W. D.; Kollman, P. A. A well-
behaved electrostatic potential based method using charge
restraints from deriving atomic charges: the RESP model. J.
Phys. Chem. 1993, 97, 10268-10280.
(32) Bertrand, J. A.; Thieffine, S.; Vulpetti, A.; Cristiani, C.; Valsa-
sina, B.; Knapp, S.; Kalisz, H. M.; Flocco, M. Structural
characterization of the GSK-3â active site using selective and
nonselectve ATP-mimetic inhibitors. J. Mol. Biol. 2003, 333,
393-407.
(33) Paul, N.; Rognan, D. ConsDock:
A new program for the
consensus analysis of protein-ligand interactions. Proteins 2002,
47, 521-533.
(34) Cohen, M. S.; Zhang, C.; Shokat, K. M.; Taunton, J. Structural
bioinformatics-based design of selective, irreversible kinase
inhibitors. Science 2005, 308, 1318-1321.
(35) Mart´ın-Aparicio, E., Fuertes, A., Perez-Puerto, M. J., Perez, D.,
Alonso, M., Martinez, A., Medina, M. TDZDs: GSK-3 inhibitors
as therapeutic agents for Alzheimer’s disease and others
tauopathies. Neurobiol. Aging 2004, 25, S596.
(36) Girard, M. L.; Dreux, C. Possibilite´s Re´actionnelles et Struc-
turales de de´rive´s de la Thiazolidine. I.-Addition, Substitution,
Hydrolyse. Bull. Soc. Chim. Fr. 1968, 3, 3461-3468.
(37) Zumach, G.; Weiss, W.; Ku¨hle, E., British Patent 1,136,737
(June, 1966); Chem. Abstr. 1969, 70, 7795.
(38) MacLauchlin, C.; May, I. H.; Izydore, R. A. Synthesis and
cytotoxic action of 1-oxoalkyl and 1,2-dioxoalkyl-1,2,4-triazoli-
dine-3,5-diones in murine and human tissue cultured cells. Arch.
Pharm. Pharm. Med. Chem. 1999, 332, 225-232.
(39) Woodget, J. R. Use of peptide substrates for affinity purification
of protein-serine kinases. Anal. Biochem. 1989, 180, 237-241.
(40) Parker, G. J.; Law, T. L.; Lenoch, F. J.; Bolger, R. E. Develop-
ment of High Through-put screening assays using fluorescence
polarization: nuclear receptor-ligand-binding and kinase*/
phosphatase assays. J. Biomol. Screen. 2000, 5, 77-88.
(41) Lengyel I.; Nairn, A. C.; MC Cluskey, A.; Toth, B.; Penke, B.;
Rostas, J. A. P. Auto-inhibition of Ca2+/calmodulin-dependent
JM040895G