Novel isothiourea derivatives as potent neuroprotectors and cognition enhancers: synthesis, biological and physicochemical properties

Article abstract of DOI:10.1021/jm8012882

Various salts of 3-allyl-1,1-dibenzyl-2-ethyl-isothiourea, 1 (hydrochloride), 2 (hydrobromide), and 3 (hydroiodide), were synthesized. Ca-blocking properties of these salts were studied. Comparative analysis of the potentiating effects of 3 and cyclothiaz

Full text of DOI:10.1021/jm8012882

J. Med. Chem. 2009, 52, 1845–1852
1845
Novel Isothiourea Derivatives as Potent Neuroprotectors and Cognition Enhancers: Synthesis,
Biological and Physicochemical Properties
German L. Perlovich,*,†,‡ Alexey N. Proshin,‡ Tatyana V. Volkova,† Sergey V. Kurkov,† Vlaimir V. Grigoriev,‡
Ludmila N. Petrova,‡ and Sergey O. Bachurin‡
Institute of Solution Chemistry, Russian Academy of Sciences, 153045 IVanoVo, Russia, and Institute of Physiologically ActiVe Compounds,
Russian Academy of Sciences, 142432, ChernogoloVka, Russia
ReceiVed October 13, 2008
Various salts of 3-allyl-1,1-dibenzyl-2-ethyl-isothiourea, 1 (hydrochloride), 2 (hydrobromide), and 3
(hydroiodide), were synthesized. Ca-blocking properties of these salts were studied. Comparative analysis
of the potentiating effects of 3 and cyclothiazide (CT) on transmembrane currents caused by kainic acid
(KA) and glutamate in Purkinje neurons was performed. Analysis of the effects of 1 on N-methyl-D-aspartate
(NMDA) receptors was performed on primary culture of heterogeneous neurons of rat cerebral cortex. Single
crystals were grown and X-ray diffraction experiments solving the crystal structures of 1-3 were carried
out. Analysis of conformations of the molecules in the crystal lattices was performed. The temperature
dependencies of the solubility of 1-3 in water and n-octanol were obtained, and the thermodynamic
parameters of solubility process were calculated. The effect of halogen atoms on the solubility was analyzed.
The partitioning processes in the water-octanol system were studied at 25 °C. Chemical stability of tested
salts in pH 7.4 phosphate buffer was determined at 25 °C.
Introduction
The major goal of this study was focused on the properties
of various 3-allyl-1,1-dibenzyl-2-ethylisothiourea salts 1 (hy-
drochloride, IP5051_Cl), 2 (hydrobromide, IP5051_Br), and 3
(hydroiodide, IP5051_I) (Figure 1) that we synthesized and
developed.20
The N-methyl-D-aspartate (NMDAa) receptor belongs to a
family of ionotropic glutamate receptors and performs important
functions in the central nervous system. It is mainly involved
in neuronal signaling processes, memory consolidation, and
synaptic plasticity.1,2 The neurotoxicity that is induced by
NMDA hyperactivation leads to a number of pathological
conditions ranging from acute neurodegenerative disorders, such
as stroke and trauma, to chronic forms of neurodegenerative
diseases, i.e., Huntington’s disease, Parkinson’s disease, Alzhe-
imer’s disease (AD), and lateral amyotrophic sclerosis.3-6
It is generally recognized that specific inhibition of calcium
ion influx via hyperactivated glutamate(Glu)-receptor-channel
complex (GluR), particularly the NMDA-subtype receptor,
provides sufficient protection against a diverse group of
neurological stresses related to the Glu excitotoxicity, such as
brain ischemia, AD, etc.7 In the case of AD, it is considered
that specific inhibitors of Glu-stimulated calcium uptake in
nervous cells provides sufficient neuronal protection against the
neurotoxic effect associated with ꢀ-amyloid peptide (Aꢀ) which
is the major pathogen in AD, which is particularly realized by
potentiating of excitotoxic effects of endogenous glutamate.8
On the other hand, it is assumed that glutamate receptor agonists
(especially R-amino-3-hydroxy-5-methyl-4-isoxazolepropionic
acid (AMPA) subtype receptor agonists) exhibit strong cogni-
tion-enhancing effects due to the activation of glutamatergic
neurotransmission.9 Such a dualism in properties of glutamater-
gic compounds explains the strong interest in these compounds
as promising neuroprotectors and cognition enhancers.
In the present study, we used a variety of in vitro and in
vivo methods to perform the primary screening and selection
of potential neuroprotectors and cognition enhancers in wide
spectrum of GluR ligands. Particularly, we analyzed properties
of these salts to inhibit the total and specific glutamate-
stimulated calcium ions uptake (Glu-Ca-uptake) in rat brain
synaptosomes.
Interaction of the drug molecule with target receptors is a
key feature in the design drug molecules; however, drug delivery
plays an equally important role. Therefore, screening parameters
characterizing the solubility, partitioning, and passive transport
processes are an essential part of the drug design.
Hansch’s group10 suggested that the octanol-water partition-
ing (log Poct) is most commonly used method in QSAR studies.
The index around 2 represents the optimal lipophilic nature of
the tested molecule that is required for penetrating the CNS;
however, this rule is not based on the permeability rates or
equilibrium concentrations. Instead it only takes into account
tests of biological activity. Two decades ago Ganellin’s group11
failed to predict the brain penetration of H2 antagonists. The
log P data from the cyclohexane-water system (log Pchex) were
likewise inadequate. In contrast, Young et al.11 reported that
there is a significant correlation between the difference of log Poct
and log Pchex data (∆ log P) as well as the logarithm of the brain/
blood equilibrium concentration ratios of H2 antagonists. They
concluded that ∆ log P accounts for hydrogen bonding and
reflects two different events. The log Pchex value quantifies the
partitioning of the molecule into nonpolar regions of the brain,
whereas the log Poct value describes protein interactions in blood.
Thus, in order to optimize the brain penetration of a drug
molecule, the overall H-bonding property of this molecule has
to be minimized. Abraham12-14 and colleagues interpret the
* To whom correspondence should be addressed. Phone: (+7) 4932
533784. Fax: (+7) 4932 336237. E-mail: glp@isc-ras.ru.
†
Institute of Solution Chemistry.
Institute of Physiologically Active Compounds.
‡
a Abbreviations: NMDA, N-methyl-D-aspartate; AD, Alzheimer’s disease;
GluR, glutamate(Glu)-receptor-channel complex; Glu, glutamate; Aꢀ,
ꢀ-amyloid peptide; AMPA, R-amino-3-hydroxy-5-methyl-4-isoxazolepro-
pionic acid; Glu-Ca-uptake, glutamate-stimulated calcium ions uptake; CNS,
central nervous system; CT, cyclothiazide; KA, kainic acid.
10.1021/jm8012882 CCC: $40.75
2009 American Chemical Society
Published on Web 03/10/2009

1846 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 7
PerloVich et al.
Figure 1. Structures of the compounds studied: (a) 1; (b) 2; (c) 3.
∆ log P value in a more complex way. They suggest not only
that the hydrogen bond acidity is important but also that the
solute hydrogen bond basicity and dipolarity/polarizability
contribute and lead to a more positive ∆ log P value, thus
favoring the penetration through the blood-brain barrier. In
other words, the larger is the solute, the more negative the
∆ log P value becomes, thereby favoring penetration into the
brain.
In the second part of this study we determined the solubility
of compounds that were selected in model solvents (water and
n-octanol) within a wide temperature range and analyzed the
impact of different types of anions (Cl-, Br-, and I-) on the
outlined processes. Moreover, the partitioning processes of
the compound in the model systems were described and the
relationship between the partitioning coefficients and biological
activity parameters was discussed.
Results and Discussion
Biological Experiments. Ca-Blocking Properties. The
results of the tested compounds for their ability to inhibit Glu-
Ca-uptake are the following: 3, KI ) 6.7 ( 1.8; 2, KI ) 3.0 (
1.2; 1, KI ) 4.2 ( 1.5.
Comparative Study of 1-3 and Cyclothiazide Action
on AMPA Receptors. 3 is the lead compound of the studied
salts in our analysis. Both 3 and 2 salts were investigated. The
goal of this stage was to determine the mechanism of the 1-3
effect. Our previous experiments revealed that the positive effect
of this compound on learning and memory in animal models is
probably linked to its ability to potentiate kainate-induced
currents in brain neurons.15
Nowadays, cyclothiazide (CT) is one of the most effective
positive modulators of AMPA receptors. The mechanism of its
effect is determined via the dose-dependent inhibition of
desensitization of AMPA receptors which is caused by their
agonists such as AMPA agonists themselves, glutamate, and to
a lesser extent kainate. Therefore, we performed a comparative
study of CT and 1-3 effects on AMPA receptors on the
transmembrane currents induced by the activation of these
receptors with their agonists. The comparison of CT and 1-3
was useful in revealing the diversity and/or similarity of their
interactions with AMPA receptors.
Experiments were performed on Purkinje neurons from rat
cerebellum that contained AMPA receptors. Since both glutamate
and AMPA caused fast and significant desensitization of the
currents, making it impossible to obtain the dose depended
response, kainate was used as agonist in our experiments (Figure
2).
The experiments demonstrated qualitatively similar effects
of CT and 3 on kainate and glutamate currents (Figure 3). To
find the quantitative difference between effects of these com-
pounds, the dose-dependent responses for both kainate and
kainate in the presence of CT and 3 were measured. As a result,
our analysis revealed an inverse dependence of the potentiating
effect of both of these compounds on the agonist dose; i.e., low
Figure 2. Transmembrane currents induced in Purkinje neurons by
different doses of glutamate and KA. (a) Currents induced by different
concentrations of glutamate: (1) 100 µΜ, (2) 500 µM, (3) 2000 µM.
(b) Currents induced by different concentrations of kainate: (1) 50 µΜ,
(2) 100 µΜ, (3) 500 µΜ, (4) 1000 µΜ, and (5) 2000 µΜ.
concentrations of the agonist led to more pronounced potentia-
tion of the receptors. Examples of the potentiating effect of 3
on the transmembrane currents of various amplitudes that were
caused by different concentrations of kainate are shown in Fig-
ure 4.
The kainate dose-dependent responses that were obtained in
the presence of 3 in its maximum soluble concentration (30 µM)
were compared with those measured in the presence of the same
concentration of CT (Figure 5). According to our results, both
3 and CT caused an unparallel shift of the kainate dose-
dependent response within wide range of concentration from
10 to 4000 µM. The response curve obtained in the presence
of 3 is shifted more to the left and it is higher than the response
obtained in the presence of CT, which demonstrates a higher
efficacy of 3 in comparison to CT.
In another series of experiments the concentrations of CT
and 1 were varied within the 0.5-50 µM range whereas kainate
dose remained constant (200 µM). The magnitude of this
potentiation significantly depended on kainate concentrations.
For low (20-50 µM) kainate doses the magnitude of potentia-

Isothiourea DeriVatiVes
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 7 1847
Figure 3. Comparative potentiation of KA and Glu responses in
Purkinje neurons by CT and 3. Transmembrane currents were induced
by 0.5 mM kainic acid and 0.5 mM glutamate. Both CT and 3 were
applied in 30 µM dose: (1) control, responses induced by 0.5 mM of
corresponding agonist; (2) 30 µM CT; (3) 30 µM 3 action on kainic
acid (a) and glutamate (b) responses.
Figure 5. Effect of CT (cyclothiazide) and 3 on dose-dependent
response for the currents induced by kainic acid in Purkinje neurons.
All responses are normalized to those induced by maximum (4 mM)
concentration of kainate: (b) control; (O) in presence of 30 µM
cyclothiazide; (0) in presence of 30 µM 3.
Figure 4. Concentration dependence of 3 effect on the currents induced
by different kainate doses. Five doses of 3 were used: 1.0, 5.0, 10.0,
30.0, and 50 µM. Kainate was applied in 50 µM (b), 500 µM (O), and
1000 µM (0) doses. Magnitude of the kainate current in the absence
of 3 was taken as 1 for each of these three doses.
tion was increased to 19-fold, whereas at high kainate doses
(∼4000 µM) the increase was 1.5-fold relative to the maximum
value. Thus, 1 was found to be more effective than CT (Figure
6). For example, for the responses that were caused by 1000
µM kainate exposure, the potentiation was 150% in the presence
of 30 µM CT. In contrast, exposure at the same concentration
of 1 resulted in a 3.9-fold increase, which was equal to 290%.
In additional experiments we determined the potentiating effect
of the maximum doses of CT and 1 (30 µM) on the receptor
responses caused by the highest dose of kainate (∼4000 µM).
The magnitudes of potentiation caused by CT and 1 were equal
to 60% (n ) 4) and 132% (n ) 5), respectively (p < 0.05).
Because of the high solubility of 1, we measured the effects
of higher doses of this compound, compared these values with
CT, and examined the differences between 1 and 3. Further
experiments demonstrated that there was no difference between
1 and 3 at the 0.5-20 µM range. However, at 30 µM, 1 showed
a 5-10% increase of potentiation in response to kainate relative
to the potentiation of 3 at the same dose, which was probably
due to low solubility of 3 at this concentration. More significant
Figure 6. Comparative effect of different doses of CT and 1 on
currents, induced by the fixed concentration of KA. (a) Cyclothiazide
effect: (1) current induced by 200 µM kainate application (control);
(2-5) currents induced by 200 µM kainate application in presence of
1.0 µM (2), 10 µM (3), 30 µM (4), and 50 µM (5) of cyclothiazide. (b)
1 effect.
differences were detected between CT and 1 at 50 µM. For
example, CT led to 5.55-fold increase of 500 µM kainate-
induced response, whereas 1 resulted in a 12.2-fold increase at
the same concentration of kainate.
Both 1 and 3 potentiated the glutamate-induced currents in
the same way as CT; however, their effects were more
significant. Because of restrictions of compound applications
on neurons, it was hard to determine the peak response caused
by the glutamate and kainate. The glutamate application resulted

1848 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 7
PerloVich et al.
Table 1. Effect of 1 on NMDA-Induced Currents in Rat Cerebral
Cortex
currenta/responseb
50 µM
NMDA
100 µM
NMDA
5000 µM
NMDA
10000 µM
NMDA
control
325/100%
320/98%
220/68%
155/48%
110/34%
100/31%
35/11%
425/100%
420/99%
340/48%
235/55%
150/35%
100/23%
35/8%
710/100%
420/59%
250/35%
255/36%
145/20%
100/14%
30/4%
950/100%
425/45%
330/35%
245/26%
155/16%
100/10%
30/3%
0.5 µM 1
1 µM 1
2 µM 1
5 µM 1
10 µM 1
30 µM 1
a Current in pA. b % of the responses in the control.
in a flat curve and did not depend on the agonist dose. The
application of the studied compounds led to a significant
potentiation of the glutamate responses, whereas this potentiation
exceeded the kainate-induced responses. For example, 30 µM
CT and 3 caused 7.0- and 26.9-fold increases of receptor
responses after exposure to 500 µM glutamate, respectively.
Some differences were observed in the initiation of the
potentiating effect of CT and 3. The effect of CT evolves during
several seconds and remains unchanged, whereas the effect of
3 becomes constant only after 30-40 s. The time required for
recovery of the kainate-induced currents during washing of CT
and 3 also differs for these compounds. The amplitude of these
currents becomes normal after 1-3 min of washing of 30 µM
CT, whereas stabilization of currents at 30 µM 3 takes about
9-11 min after washing. All these experiments led to the
conclusion that CT and 3 showed similar effects on the
responses caused by the activation of AMPA receptors in
Purkinje neurons by different agonists. Therefore, 3 can potenti-
ate AMPA responses due to inhibition of desensitization caused
by agonists of these receptors.
Effects of 1 on Activity of NMDA Receptors. The effects
of 1 on NMDA receptors were analyzed on a heterogeneous
population of a primary culture of neurons isolated from rat
cerebral cortex. NMDA receptors (and more precisely receptors
in cortex neurons) were divided into two groups according to
their sensitivity to 1.
In the first group, NMDA receptors were effectively inhibited
by low concentrations of 1 (n ) 5). The efficacy of 1 inhibition
of NMDA-induced currents depends on agonist concentration
and amplitude of currents. Data are shown in Table 1. In general,
exposure to 1 µM 1 caused 32-65% inhibition of NMDA-
induced responses in neurons in the first group. The higher the
amplitude of the current, the more pronounced is the effect
detected. Exposure with 10 and 30 µM 1 resulted in 70-90%
and 89-97% inhibition of the receptor, respectively.
Figure 7. Atom numeration of studied molecules and four conforma-
tionally independent fragments.
the shortest one corresponds to Cl (3.042 Å). This distance for
3 is 3.487 Å. For the mentioned compounds, the DsH· · ·A
angles have approximately the same value (within the experi-
mental error) of ∼153°, and they essentially differ from the
ideal geometric value of 180°. The following angles were chosen
for the quantitative comparison of the molecular conformations
(Table 2): S1-C4-N2-C7, N2-C7-C8-C9, N2-C14-
C15-C16, N2-C4-S1-C5, and C4-N1-C3-C2. The
projections of these molecules along the C4-N2 bond are shown
in Figure 8 as a pictorial rendering to highlight the geometric
differences. The bond between Ph1 and Ph2 fragments is
positioned in the background of the picture with regard to the
noted bond (dotted circles), whereas the R1 and R2 fragments
are located in the foreground (full lines).
It is obvious that the molecular conformations of the bromide
and iodide salts are similar; however, they essentially differ from
that of the chloride salt. As a result, these differences may
account for the physicochemical properties of these compounds
both in crystals and in solutions. For example, for 1 R1, R2 and
Ph2 fragments are positioned above the imaginary plane passing
through N1S1N2 atoms, whereas the Ph1 fragment is located
below the mentioned plane. For 2 and 3, the R1 and Ph2
fragments are positioned above the plane, but R2 and Ph1 are
below the plane. As a result of crystallization of these salts from
the solutions, the organic part of the molecules takes on special
conformation. Thus, we may assume that during solubilization
of crystals and dissociation of salts on cations and anions, the
organic part of the molecule can partly keep the conformational
state (inherited from the crystal) in the solution. Hence, such a
fact may explain the different molecular behavior during drug
delivery as well as interactions with the biological receptor.
The inhibition effect of 1 was significantly decreased in the
neurons in the second group (n ) 13). A decrease (about
10-30%) of NMDA-induced currents was observed for 10 µM
1 only. The magnitude of this effect was independent of the
amplitude of the responses. However, the exposure of 30 µM 1
led to 50-70% inhibition of the responses in these neurons.
Physicochemical Experiments. Crystal Structures of
1-3. In order to analyze the molecular conformations in crystals
of 1-3, single crystals were grown from the acetone-hexane
solutions and the complete crystal structures were resolved.
Atomic numbering of the studied molecules and four confor-
mationally independent fragments (Ph1, Ph2, R1, and R2) are
shown in Figure 7.
Solubility and Partitioning Properties. We carried out
solubility tests of salts in solvents modeling biological media.
Initially, we used phosphate buffer at pH 7.4. However, the
kinetic solubility experiments showed that the tested salts were
not chemically stable in this buffer; therefore, water was chosen
as the primary solvent (where the experimental values were
strictly reproducible) imitating delivery pathways. In order to
mimic the lipophilic medium (the model of transcellular delivery
pathways), we used n-octanol. The thermodynamic parameters
of solubility process were obtained on the basis of the temper-
ature dependencies of the solubility data within the wide
temperature range from 18 to 42 °C. Results of the solubility
experiments in water and n-octanol as well as the thermody-
These salts create a hydrogen bond between the halogen atom
and the N1-H1(N) fragment. The longest distance between N
and the halogen atom corresponds to Br (3.716 Å), whereas

Isothiourea DeriVatiVes
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 7 1849
Table 2. Angles (deg) Describing Molecular Conformations in the Crystal Lattices
compd
S1-C4-N2-C7
N2-C7-C8-C9
N2-C14-C15-C16
N2-C4-S1-C5
C4-N1-C3-C2
1
2
3
26.8
29.5
34.8
-69.6
-58.7
162.4
62.9
117.8
113.3
55.1
50.2
40.2
119.9
-94.8
-79.1
Table 3. Temperature Dependencies of Solubility (X2 Mole Fraction) of 1-3 in Water and n-Octanol
1
2
3
T (K)/t (°C)
water, X2 × 102
n-octanol, X2 × 102
water, X2 × 104
n-octanol, X2 × 103
water, X2 × 105
n-octanol, X2 × 104
291.15/18
293.15/20
295.15/22
298.15/25
300.15/27
303.15/30
310.15/37
315.15/42
8.20
8.60
9.10
10.2
10.7
11.3
4.23
4.89
3.44
4.96
2.82
3.46
7.28
9.85
5.66
7.23
8.17
6.74
10.4
13.2
4.27
5.81
6.88
12.6
18.3
24.8
Aa
Ba
Rb
σc
6.1 ( 0.5
2505 ( 153
0.9926
1.9 ( 0.3
2830 ( 86
0.9986
13.6 ( 0.5
5645 ( 139
0.9991
2.6 ( 0.2
3834 ( 62
0.9996
10.1 ( 0.4
5080 ( 121
0.9992
1.75 × 10-2
1.65 × 10-2
2.67 × 10-2
1.19 × 10-2
2.32 × 10-2
a Parameters of the correlation equation ln X2 ) A - B/T. b R: pair correlation coefficient. c σ: standard deviation.
Table 4. Thermodynamic Functions of Solubility Process of 1-3 in Studied Solvents
a
b
compd
X2 (mole fraction)
∆G0sol (kJ·mol-1
)
∆HXsol (kJ·mol-1
Water
)
∆H0sol (kJ·mol-1
)
T∆S0sol (kJ·mol-1
)
∆Ss0ol (J·mol-1 ·K-1
)
1
19.1 ( 0.3
25.2 ( 0.3
1·4H2O
2
3
5.05 × 10-4
3.46 × 10-5
18.8
25.5
23.5 ( 0.7
31.9 ( 0.5
4.7
6.4
16 ( 3
22 ( 2
n-Octanol
1
2
3
1.00 × 10-1
4.84 × 10-3
9.68 × 10-4
5.7
13.2
17.2
21 ( 1
47 ( 1
42 ( 1
15.3
33.8
24.8
51 ( 3
113 ( 4
83 ( 4
a Data obtained from solubility experiments. b Data obtained from calorimetric experiments. Enthalpy per mole of 1.
namic functions of the solubility processes are summarized in
Tables 3 and 4.
1 is very soluble in water, and it creates crystallohydrate(s)
in the bottom phase. Therefore, it is not possible to obtain
reproducible solubility measurements. In contrast, the results
for the remaining salts were reproducible in water and n-octanol
solutions, since there was no formation of any crystallohydrates/
crystallosolvates in the bottom phases. On the basis of the
thermogravimetric experiments, the stoichiometry of 1 crystal-
lohydrate was derived: [1·4H2O].
The chloride salt is the most soluble both in water and in
n-octanol, whereas the iodide one is less soluble in these solvents
(Table 3). The impact of the halogen atom on the solubility
data at 25 °C can be roughly estimated. The bromide salt is 14
times more soluble in water in comparison with the iodide one.
In turn, the bromide salt is 5 times more soluble in n-octanol
compared to iodide salt, whereas the chloride salt is 21 times
more soluble in contrast with the bromide salt. The entropies
of the solubility processes both in water and in n-octanol have
positive signs, and this fact testifies that the studied molecules
are distributed in a more disordered state in the solutions than
in the crystals. It is worth noting that the octanolic solutions
are more disordered than the aqueous ones.
In order to estimate the difference of crystal lattice energies
of the crystallohydrate [1·4H2O] and the unhydrated phase (in
other words, the stabilization of crystal lattice energy by solvent
(water) molecules), we used our previously described method.16
The standard values of solution enthalpies (per mole of 1) for
the hydrated and for the unhydrated phases were determined
in the same solvent (water) by the calorimetric method. The
results are presented in Table 4. As a result, the crystal lattice
energy of the crystallohydrate is 6.1 ( 0.6 kJ ·mol-1 higher than
energy of the unhydrated phase.
Figure 8. Projection of the studied molecules along the C4-N2 bond:
(a) 1; (b) 2; (c) 3. Dotted line corresponds to molecular fragment located
behind the C4-N2 bond (second plane), whereas solid line corresponds
to molecular fragment located in front of the bond.
The partition coefficients were studied within the wide
concentration range of n-octanol phase solutions at 25 °C. It is

1850 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 7
PerloVich et al.
conformations in this buffer, which they “inherit” from the
crystal lattices during the solubility process.
Conclusions
As a result of our studies, we determined Ca-blocking
properties of 1-3 and arranged them in the following manner:
(KI(3) ) 6.7 ( 1.8) > (KI(1) ) 4.2 ( 1.5) > (KI(2) ) 3.0 (
1.2).
A comparative study of the potentiating effect of 3 and
cyclothiazide (CT) as a well-known positive modulator of
AMPA receptors transmembrane currents induced by kainic acid
(KA) and glutamate in Purkinje neurons was performed. In order
to determine the quantitative differences between effects of these
compounds, we measured the dose-dependent responses for
kainate alone and for kainate in the presence of CT and 3.
Analysis of these results revealed an inverse correlation of the
potentiating effect for both of these compounds on the agonist,
i.e., low concentration of the agonist led to more pronounced
potentiation. As a result of these experiments, we concluded
that effects of CT and 3 on responses induced by activation of
AMPA receptors in Purkinje neurons by different agonists are
similar. Therefore, 3 potentiates AMPA responses because of
inhibition of desensitization caused by agonists of these
receptors.
Figure 9. Kinetic dependencies of the decomposition processes of 1-3
in phosphate buffer with pH 7.4 at 25 °C.
worth noting that we did not detect any concentration depend-
encies of partition coefficients, which indicates that the associa-
tive processes are negligible and activity coefficients are equal
to 1.
The partitioning coefficients have the following values: P(1)
) 2.1, P(2) ) 4.4, and P(3) ) 0.42. Thus, the iodide salt is
mainly distributed in the aqueous phase, in contrast to the
chloride and bromide salts. In other words, the iodide salt
“prefers” water enriched pathways (paracellular pathways) for
the delivery process. Probably the chloride and bromide salts
undertake the lipophilic (transcellular) delivery pathways.
The effects of 1 on NMDA receptors were analyzed on
heterogeneous population of primary culture of neurons isolated
from rat cerebral cortex. NMDA receptors were divided into
two groups according to their sensitivity to 1.
Single crystals were grown and X-ray diffraction experiments
were performed to solve the crystal structures of 1-3. The
molecular conformations of the bromide and iodide salts are
similar, whereas they essentially differ from that of the chloride
salt.
Stability of 1-3 in Phosphate Buffer. Chemical stability
of the tested salts was studied in phosphate buffer (pH 7.4) in
the following way. Three test tubes were filled with the same
mass of each salt and diluted with the same volume of buffer
followed by intensive stirring by mechanical stirrer at 25 °C.
The analyzed probes were collected (2 mL) from the initial test
tube mixtures at a definite time interval and mixed with 1 mL
of deuterated chloroform (CDCl3). The obtained mixture was
stirred in a thermostat for 30 min, and finally the chloroform
phase was separated for NMR analysis. It is worth noting that
30 min after the beginning of stirring, the tested salts dissociated
to the organic phase and the halogen anion (no peak was
detected corresponding to the quaternary proton on nitrogen at
∼10.2 ppm), whereupon the organic part of the molecule started
to decay and resulted in decomposition products. One of these
products was a dibenzylamine fragment (peaks, ∼4.6-4.9 ppm).
The amount of the decay product was estimated by integral
values of the proton peaks. The kinetic dynamics of the
decomposition process of the studied compounds are shown in
Figure 9 where the organic parts of these molecules (obtained
from different salts) have various chemical stabilities in the
phosphate buffer. In other words, the organic phase that is
derived from the iodide salt is the most stable to the decay
process, whereas the organic phase from the chloride salt
demonstrates less stability to the decomposition. Interestingly,
the Ca-blocking properties (a parameter describing biological
activity) of tested salts correlate with their chemical stability
(i.e., more stable molecules have higher KI): (KI(3) ) 6.7 (
1.8) > (KI(1) ) 4.2 ( 1.5) > (KI(2) ) 3.0 ( 1.2). Since the
organic phases of these molecules in phosphate buffer have
similar chemical composition, the above-mentioned differences
in the chemical stability can be linked to various molecular
The temperature dependencies of 1-3 solubility in water and
n-octanol were obtained, and thermodynamic parameters of
solubility process were calculated. The chloride salt is mostly
soluble in both water and n-octanol, whereas the iodide salt is
less soluble in these solvents. The bromide salt is 14 times more
soluble in water compared to the iodide salt. In turn, the bromide
salt is 5 times more soluble in n-octanol compared to iodide
salt, whereas the chloride salt is 21 times more soluble in
contrast with the bromide salt. The partition coefficients were
determined within a wide concentration range of the n-octanol
phase solutions at 25 °C. The partition coefficients have the
following values: P(1) ) 2.1; P(2) ) 4.4; P(3) ) 0.42. Thus,
the iodide salt is mainly distributed in the aqueous phase in
contrast to the chloride and bromide salts.
1 is very soluble in water, and it creates crystallohydrate in
the bottom phase of the solubility experiments. On the basis of
the thermogravimetric experiments, we determined the stoichi-
ometry of 1 crystallohydrate: [1·4H2O]. The crystal lattice
energy of the crystallohydrate is 6.1 ( 0.6 kJ ·mol-1 higher than
the unhydrated phase.
Chemical stabilities of tested salts in phosphate buffer (pH
7.4) were studied at 25 °C. The salts dissociated into pure
organic phase and the halogen anion 30 min after the beginning
of stirring. The organic phase, which is obtained from the iodide
salt, is mostly stable to the decay process, whereas the organic
phase obtained from the chloride salt demonstrates less stability
to the decomposition. Interestingly, the Ca-blocking properties

Isothiourea DeriVatiVes
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 7 1851
Scheme 1
Co. Phosphate buffer solution (pH 7.4) was prepared by mixing of
solutions of appropriate sodium and potassium salts of phosphoric
acid. All chemicals were AR grade. The pH values were measured
with a Toledo MP 220 pH meter (Mettler) standardized with pH
1.68 and 9.22 solutions.
of the discussed salts correlate with the chemical stability: more
stable molecules have higher KI.
Experimental Section
Chemical Procedures. The standard and accessible methods of
the synthesis of tetra-substituted isothioureas were developed by
the Institute of Physiologically Active Compounds of Russian
Academy of Science.20 Synthetic approaches to novel isothioureas
were carried out according to Scheme 1.
Solubility Determination. Solubilities of these compounds were
determined within a wide temperature range (18-42) ( 0.1 °C by
the isothermal saturation technique. The solid phase precipitated
upon centrifugation, the surfactant was filtered (Acrodisc CR syringe
filter, PTFE, 0.2 µm pore size), and the bulk solution was measured
spectrophotometrically by UV-2401 PC spectrophotometer (Shi-
madzu, Japan) according to a previously described protocol.17 The
solubility measurements are reported as averages of at least three
replicated experiments with statistical errors within 2.5%.
Determination of Partition Coefficients. The method of
isothermal saturation was used for the determination of partition
coefficients P (D7.4). We prepared n-octanol solutions with fixed
concentrations, varying the concentrations between one-tenth and
the maximum solubility value. The respective solution was left in
a thermostated ampule, with the same volume of added water. The
equilibration was carried out for 24 h under continuous stirring.
The initial concentration and final concentration of the tested
compound in the octanol solutions were determined by spectro-
photometry before and after the equilibration. All experiments were
carried out at 25 °C.
S-Ethyl-N-allyl-N′,N′-dibenzylisothiourea Hydrochloride (1).
S-Ethyl-N-allyl-N′,N′-dibenzylisothiourea hydroiodide (4.52 g, 0.01
mol) was added to a stirred mixture of 10% aqueous solution NaOH
(50 mL) and benzene (50 mL). The mixture was stirred at 0-5 °C
until the salt was completely dissolved. After that, the benzene layer
was separated and washed out with water (2 × 50 mL). The solution
was dried (Na2SO4), and the solvent was evaporated. S-Ethyl-N-
allyl-N′,N′-dibenzylisothiourea acted as base which was quickly
1
used in the next step because of its lability. H NMR [200 MHz,
DMSO-d6] δ: 1.20 (t, 3H, CH3), 2.30 (q, 2H, SCH2), 4.15 (m, 2H,
NCH2), 4.60 (s, 4H, CH2Ph), 5.25 (m, 2H, dCH2), 5.80 (m, 1H,
-CHd), 7.20-7.40 (m, 10H, ArH).
S-Ethyl-N-allyl-N′,N′-dibenzylisothiourea (1.6 g, 0.005 mol) was
dissolved in isopropanol (30 mL) followed by addition (by drops)
of isopropanolic HCl solution (0.006 M, 20 mL) to the stirred
solution. The reaction mixture was kept under stirring at 20 °C for
8 h. The grown crystals were filtered, and the yield was 1.2 g
(66.6%). Mp 143-145 °C. Anal. (C20H25ClN2S) C, H, N, S.
S-Ethyl-N-allyl-N′,N′-dibenzylisothiourea Hydrobromide (2).
Bromoethane (1.31 g, 0.012 mol) was added to a solution of N-allyl-
N′,N′-dibenzylthiourea (2.96 g, 0.01 mol) in acetone (50 mL). The
reaction mixture was kept for 72 h at 40 °C. The grown crystals
were filtered and recrystallized from isopropanol (30 mL) and
yielded 3.1 g (76.5%). Mp 156-158 °C. Anal. (C20H25BrN2S) C,
H, N, S.
S-Ethyl-N-allyl-N′,N′-dibenzylisothiourea Hydroiodide (3).
Iodoethane (1.87 g, 0.012 mol) was added to a solution of N-allyl-
N′,N′-dibenzylthiourea (2.96 g, 0.01 mol) in acetone (50 mL). The
reaction mixture was kept for 48 h at 40 °C. The grown crystals
were filtered and recrystallized from isopropanol (30 mL) and
yielded 3.5 g (77.3%). The melting point was 150-152 °C. Anal.
(C20H25IN2S) C, H, N, S.
The partition coefficient was calculated using the following
formula:
P ) Cow/Cwo
(1)
where Cow and Cwo are the molar concentrations of the solute in
the mutually saturated phases of octanol and water. The accuracy
of the P-value was verified by checking the mass balance of the
starting amount of the compound i compared to the total amount
of the compound partitioned between the two phases,
mi ) mow + mwo
(2)
where mi ) CiVi is the starting mass (in moles) of the compound,
mow ) CowVow is the mass of the substance dissolved in the water-
saturated octanol phase, and mwo ) CwoVwo is the mass of the
substance dissolved in the octanol-saturated water phase. Each
experiment was performed at least three times.
Materials and Solvents. 1-Octanol (n-octanol, CH3(CH2)7OH,
MW ) 130.2, lot no. 11K3688), ARG, was from Sigma Chemical

1852 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 7
PerloVich et al.
TG Experiments. Thermogravimetric studies were carried out
within the temperature range 25-200 °C with a heating rate of 10
°C/min and a flow rate of dry argon gas at 5.4 L/h using NETZSCH
TG 209 F1 (Germany). The calibration of the TGA cell was
performed using a sample of calcium oxalate monohydrate in an
atmosphere of flowing dry argon. The measurements were repeated
three times. The determinations were off by 3%.
X-ray Diffraction Experiments. Single-crystal X-ray measure-
ments were carried out using a Nonius CAD-4 diffractometer with
graphite-monochromated Mo KR radiation (λ ) 0.710 69 Å). The
intensity data were collected at 25 °C by means of a ω-2θ scanning
procedure. The crystal structures were solved using direct methods
and refined by means of a full-matrix least-squares procedure.
CAD-4 Software (1989)18 was applied for data collection, data
reduction, and cell refinement. Programs SHELXS-97 and SHELXL-
9719 were used to solve and to refine structures, respectively.
NMR Experiments. NMR data were obtained by standard
methods using Bruker AMX-200.
Biological Methods. Calcium-Blocking Property Experi-
ments. Interaction between the compounds and glutamate-dependent
calcium uptake system was studied on newborn (8-11 days old) rat
brain synaptosomal P2-fraction isolated according to the following
method: synaptosomes were suspended in the incubation buffer A (132
mM NaCl, 5 mM KCl, 5 mM HEPES), pH 7.4, and stored at 0 °C
during all experiments. Aliquots of synaptosomes (50 µL) were
transposed to buffer A containing the testing compound and 45Ca
samples. Calcium ion uptake was stimulated by introducing 200 µM
glutamate solution. After 5 min of incubation at 37 °C, the process
was terminated by filtration on GF/B filters. The sample was washed
three times with cold buffer B (145 mΜ KCl, 10 mΜ Tris, 5 mΜ
Trilon B) followed by measurements of radioactivity with scintillator
counter SL-4000 Intertechnic.
Effects of 1-3 on NMDA receptors were analyzed on a
heterogeneous population of a primary culture of neurons isolated
from rat cerebral cortex.
Acknowledgment. The present research was funded as a part
of the basic research program established by the Presidium of
Russian Academy of Sciences “Fundamental Sciences for Medi-
cine”.
Supporting Information Available: Elemental analysis data,
X-ray data, and NMR spectra of compounds 1-3. This material is
available free of charge via the Internet at http://pubs.acs.org.
References
(1) Parsons, C. G.; Danysz, W.; Quack, G. Glutamate in CNS disorders
as a target for drug development: an update. Drug News Perspect.
1998, 11, 523–579.
(2) Nakanishi, N. S. Molecular diversity of glutamate receptors and
implications for brain function. Science 1992, 258, 31–37.
(3) Dingledine, R.; Borges, K.; Bowie, D.; Traynelis, S. F. The glutamate
receptor ion channels. Pharmacol. ReV. 1999, 51, 7–61.
(4) Bra¨uner-Osborne, H.; Egebjerg, J.; Nielsen, E. Ø.; Madsen, U.;
Krogsgaard-Larsen, P. Ligands for glutamate receptors: design and
therapeutic prospects. J. Med. Chem. 2000, 43, 2609–2645.
(5) Petrov, V. I.; Piotrovsky, L. B.; Grigoriev, I. A. Exitatory Amino Acids;
Volgograd Medical Academy: Volgograd, Russia, 1997; p 168.
(6) Zefirova, O. N.; Zefirov, N. S. Physiologically active compounds acting
on glutamate receptors. Zh. Org. Khim. 2000, 36, 1273–1300.
(7) Khachaturian, Z. S. Calcium hypothesis of Alzheimer’s disease and
brain aging. Ann. N.Y. Acad. Sci. 1994, 747, 1–11.
(8) Mattson, M. P.; Cheng, B.; Davis, D.; Bryant, K.; Lieberburg, I.; Rydel,
R. E. ꢀ-Amyloid peptides destabilize calcium homeostasis and render
human cortical neurons vulnerable to excitotoxicity. J. Neurosci. 1992,
12, 376–389.
In preliminary experiments, all compounds were tested at 100
µM for their ability to inhibit the glutamate stimulated Ca uptake.
If inhibition of the Glu-Ca-uptake was 50% or more, then further
studies were carried out to determine the concentration dependence
of its inhibition and the corresponding value of KI was measured
according to the following equation:
(9) Bleakman, D.; Lodge, D. Neuropharmacology of AMPA and kainate
receptors. Neuropharmacology 1998, 37, 1187–1204.
(10) Hansch, C.; Bjorkroth, J. P.; Leo, A. Hydrophobicity and central
nervous system agents: on the principle of minimal hydrophobicity
in drug design. J. Pharm. Sci. 1987, 76, 663–686.
(11) Young, R. C.; Mitchell, R. C.; Brown, T. H.; Ganellin, C. R.; Griffiths,
R.; Jones, M.; Rana, K. K.; Saunders, D.; Smith, I. R.; Sore, N. E.;
Wilks, T. J. Development of a new physicochemical model for brain
penetration and its application to the design of centrally acting H2
receptor histamine antagonists. J. Med. Chem. 1988, 31, 656–671.
(12) Abraham, M. H.; Chadha, H. S.; Mitchell, R. C. Hydrogen bonding.
33. Factors that influence the distribution of solutes between blood
and brain. J. Pharm. Sci. 1994, 83, 1257–1268.
(13) Abraham, M. H.; Chadha, H. S.; Mitchell, R. C. Hydrogen bonding.
Part 36. Determination of blood brain distribution using octanol-water
partition coefficients. Drug Des. DiscoVery 1995, 13, 123–131.
(14) Abraham, M. H.; Chadha, H. S. Applications of a Solvation Equation
to Drug Transport Properties. In Lipophilicity in Drug Action and
Toxicology; Pliska, V., Testa, B., van de Waterbeemd, H., Eds.;
Methods and Principles in Medicinal Chemistry, Vol. 4; VCH
Publishers: Weinheim, Germany, 1995.
(15) Bachurin, S. O. Medicinal chemistry approaches for the treatment and
prevention of Alzheimer’s disease. Med. Res. ReV. 2003, 13 (1), 48–88.
(16) Perlovich, G. L.; Hansen, L. K.; Volkova, T. V.; Mirza, S.; Manin,
A. N.; Bauer-Brandl, A. Thermodynamic and structural aspects of
hydrated and unhydrated phases of 4-hydroxy-benzamide. Cryst.
Growth Des. 2007, 7 (12), 2643–2648.
(17) Perlovich, G. L.; Volkova, T. V.; Bauer-Brandl, A. Towards an
understanding of the molecular mechanism of solvation of drug
molecules: a thermodynamic approach by crystal lattice energy,
sublimation and solubility exemplified by hydroxybenzoic acids.
J. Pharm. Sci. 2006, 95 (7), 1448–1458.
(18) CAD-4 Software, version 5.0; Enraf-Nonius: Delft, The Netherlands,
1989.
(19) Sheldrick, G. M. SHELXL97 and SHELXS97; University of Go¨ttingen:
Go¨ttingen, Germany, 1997.
(20) Tkachenko, S. E.; Proshin, A. N.; Bachurin, S. O.; Petrova, L. N.;
Zefirov, N. S. Derivatives of N,S-Substituted N′-1-[(Hetero)arylalkyl]-
N′-[(hetero)aryl]-methylisothioureas, Method of Synthesis of Their
Bases and Salts, Physiologically Active N′-1-[(hetero)arylalkyl]-N′-
[(hetero)aryl]-methylisothioureas, Pharmaceutical Formulation and
Method of Treatment. Russian Patent 2223952, February 20, 2004
(according to the Patent Application 2001114922 of June 4, 2001).
KI ) K(43/21) ) [(Ca4 - Ca3)/(Ca2 - Ca1)] × 100
(3)
where Ca1 is the Ca2+ influx in blank experiment (without glutamate
and tested compounds). Ca2 is the Ca2+ influx in the presence of
glutamate only (Glu-Ca-uptake), Ca3 is the Ca2+ influx in the
presence of tested compound (without glutamate), and Ca4 is the
Ca2+ influx in the presence of both glutamate and tested compound.
Electrophysiological Experiments. The AMPA-ligands activity
studies were carried out on isolated Purkinje neurons that contain
a broad population of AMPA-receptors. The main series of
experiments were performed with kainic acid (KA) as receptor
agonist because glutamate and especially AMPA induce fast and
strong desensitization of receptors, which makes the study of
dose-response relationships very difficult. Electrophysiological
experiments were carried out on freshly isolated neurons from
different areas of the brain of 14-18 days old rats using the patch-
clamp technique. The preparation of isolated cells was performed
as follows: a selected region of the brain was cut into slices of
0.4-0.6 mm width followed by the incubation in saline (150 mM
NaCl, 5 mM KCl, 2 mM CaCl2, 2 mM MgCl2, 10 mM HEPES, 10
mM glucose, pH 7.4) for 1 h. The slices were transferred to fresh
saline solution with 2 mg/mL Pronase (Serva) and 1 mg/mL
collagenase (Sigma). Slices were incubated at 34 °C and pregassed
with 100% O2. Finally, the slices were mechanically dissociated
into individual cells by means of Pasteur pipettes. The composition
of extracellular saline was 150 mM NaCl, 5 mM KCl, 2.6 mM
CaCl2, 10 mM HEPES, 10 mM glucose, pH 7.32. The composition
of the intracellular saline was 140 mM KCl, 10 mM HEPES, 5
mM EGTA, 1 mM MgCl2, 1 mM ATP. The transmembrane currents
were registered in the configuration of “whole cell”. Compounds
were exposed to neurons by the method of fast perfusion.
JM8012882